Nuclear Physics (Irving Kaplan)

NUCLEAR
PHYSICS
To Ruth
NUCLEAR
PHYSICS
by
IRVING KAPLAN
Department of Nuclear Engineering
Massachusetts Znstitute of Technology
SECOND EDITION
ADDISON-WESLEY PUBLISHING COMPANY

Reading. Massachusetts . Menlo Park, California . London Sydney . Manila
THIS BOOK IS AN ADDISON-WESLEY
WORLD STUDENT SERIES EDITION
EIGHTH PRINTING 1977
A complete and unabridged reprint of the original American textbook,

this World Student Series edition may be sold only in those countries to
which it is consigned by Addison-Wesley or its authorized trade distribu-
tors. It may not be re-exported from the country to which it has been
consigned, and it may not be sold in the United States of America or
its possessions.
Copyright© 1962 by Addison-Wesley Publishing Company, lnc.

Ali rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any forro or by any
means, electronic, mechanical, photocopying, recording, or other-
wise, without the prior written permission of the publisher. Original
edition printed in the United States. Published simultaneously in
Canada. Library of Congress Catalog Card Number 62-9402.
PREFACE TO THE FIRST EDITION
I have tried to write an elementary, yet coherent, account of nuclear

physics suitable as an introduction to this field. The book is elementary
in the sense that the reader is assumed to have had no previous exposure
to atomic and nuclear physics. It is supposed that he has had a two-year
college coÚrse in physics covering approximately the material contained
in Sears' books: Mechanics, Heat, and Sound; Electricity and Magnetism;
and Oplics, in the Addison-Wesley Physics Series. It is also assumed that
the reader is familiar with the differential and integral calculus; a one-
year course in general chemistry would be helpful, if not essential. This
book should therefore be useful in an advanced undergraduate course in
nuclear physics, to engineers interested in the large-scale applications of
nuclear physics grouped together under the name "nuclear engineering,"
or to anyone else with the indicated preparation who might be interested
in nuclear physics.
The concepts and techniques of nuclear physics are not elementary in
the sense that they are easily mastered. They have evolved through a
great deal of experimental and theoretical research and c1,,nnot be expressed
or explained in simple nontechnícal terms. I have tried, therefore, to show
how our present knowledge of atoros and nuclei has been developed, and
how sorne of this knowledge is applied. In treating the experiments that
yield information about atoros and nuclei and the ways in which these ex-
periments are interpreted in terms of theory, I have tried to avoid both the
devil of experimental complication and the deep blue sea of mathematical
detail. The emphasis is on the principies underlying the experiments and
on the experimental results rather than on the apparatus, on the physical
ideas rather than on the details of the calculations. Thus, derivations have
been included when they involve straightforward applications of physical
principies in terms of mathematics not beyond the integral calculus.
When it is necessary to use the results of more advanced methods, such
as quantum mechanics, the details of the derivations are omitted; but the
physical ideas contained in the results are discussed in sorne detail and
related to the experimental data.
The plan of the book is based on the advice of the King of Hearts:
"Begin at the beginning," the King said gravely, "and go on till you come
to the end: then stop." The need for a reasonable lirnit on the size of the
book makes it necessary, however, to define the beginning and end arbi-
trarily and to leave sorne things out on the way. The first part of the
book, consisting of the first seven chapters, is devoted to the background
V
PREFACE
of nuclear physics and begins with the chemical foundations of atomic

theory. The second part, Chapters 8 through 17, treats the physics of
the nucleus in a way that seems to me to be logical, yet consistent with
the elementary nature of the book. The third part, Chapters 18 through
22, deals with special topics and applications and includes neutron physics
and nuclear fission, which do not fit conveniently into the scheme of the
second part and at the same time lead into the most spectacular application
of nuclear physics-"nuclear energy." The subjects of charged-particle
accelerators and isotope separation, although not really nuclear physics,
are closely related to important branches of this field; they are included
for this reason and because they form, along with nuclear energy, the
main part of nuclear engineering. I have not discussed such subjects as
cosmic rays, mesons, and nuclear moments because I think that they
need a more advanced treatment than the material covered.
I have included, for several reasons, a large number of literature refer-
ences. For the skeptical or curious reader, the references will supply the
experimental and theoretical details omitted from the text, as well as
proofs of results that are stated without proof. A more important reason
for including these references stems from the fact that atomic and nuclear
physics are recently developed fields, and the original literature is more
readily available than that of classical physics. The student can read for
himself the original papers and books of the pioneers in this field-Ruther-
ford, Bohr,. Millikan, Moseley, Aston, Chadwick, Fermi, and others. The
articles in which the basic discoveries were first published give, more than
any textbook can, a feeling for the imagination and beauty inherent in
modern physics. The current literature adds a feeling for the gradual, and
sometimes painful, accumulation of experimental and theoretical infor-
mation, and for the problems which have yet to be solved.
This book has been developed from the notes for a course of lectures
given at the Brookhaven National Laboratory. I am indebted to col-
leagues who attended the lectures or used the notes for many valuable
comments and suggestions, and to many workers in the field of nuclear
physics for permission to cite and use their published results. Above all,
I find it hard to express adequately my indebtedness and gratitude to
Jean Harless for her unfailing patience, good humor, and skill in preparing
the manuscript.
IRVING KAPLAN
East Patchogue, N ew York
June, 1954
PREFACE TO THE SECOND EDITION
I have tried to bring the book up-to-date and to improve the presenta-
tion of the material. The rapid accumulation of information during the
past seven years has made it necessary to change many of the numbers
and sorne of the ideas, while the use of the book as a text, by friends and
colleagues as well as by myself, has resulted in many suggestions for im-
provement and clarification.
For the convenience of those who used the first edition, the major
changes are listed. Section 2-8, on the detection and measurement of
radiation, has been expanded. In the treatment of special relativity, in
Chapter 6, derivations of the Lorentz transformation equations and of
the formula for the variation of mass with velocity have been included.
Section 7-9 has been added to Chapter 7, in which the Schroedinger
equation is applied to the problem of a particle in a one-dimensional box;
it is hoped that this addition will help the student develop a little more
insight into sorne of the elementary ideas of wave mechanics. Two new
sections have been added to Chapter 14: Section 14-8 deals with the
neutrino and its detection; Section 14-9 with symmetry laws and the
nonconservation of parity in beta-decay. The last two sections of Chap-
ter 15, dealing with gamma-decay, nuclear energy levels and isomerism,
have been completely rewritten. A short section, 16--7, on the limitations
of the compound nucleus theory has been added to Chapter 16. In Chap-
ter 17, the treatment of nuclear models has been extended; in particular,
the shell and collective models are discussed in somewhat more életail
than in the first edition, and the optical model for nuclear reactions is
introduced. A short section, 17-4, on the nuclear radius has been added.
Chapter 19, on nuclear fission, has been modernized to include material
which was classified at the time the first edition was written but is now
part of the open literature. Corresponding changes have been made in
Chapter 20, with the result that sorne of the elementary properties of
nuclear reactors can be introduced in a more satisfactory way; a short
section on controlled thermonuclear reactions has also been added. To
compensate in part for the new material, Chapter 22 on isotope separation
has been dropped; this subject seems to attract little interest in courses
in nuclear physics, and is now treated thoroughly in other books. Finally,
I have tried to improve the problems by adding to them and by replacing
sorne of the "substitution" problems by "thought" problems.
I am indebted to many friends, colleagues, and students who have
pointed out errors and suggested changes, and I regret that I cannot list
vii
viii PREFACE
them individually because of the lack of space. I have, however, exer-

cised the author's privilege of disregarding sorne advice and criticism, and
have only myself to blame for the shortcomings of the revised book.
I am indebted to Rachel Sprinsky for her invaluable assistance in the
preparation of the manuscript.
IRVING KA.PLAN
Cambridge, Mas8.
January, 196f4
CONTENTS
PART l. THE BACKGROUND OF NUCLEAR PHYSICS

Page
CHAPTER l. THE CHEMICAL FOUNDATIONS OF ATOMIC THEORY 3

1-1 The laws of chemical combination 3
1-2 Dalton's atomic hypothesis 5
1-3 Avogadro's hypothesis and the molecule 7
1-4 Molecular and atomic weights of gaseous elements 8
1-5 The standard atomic weight . 10
1-6 Atomic weights of nongaseous elements 11
1-7 Weights and sizes of atoms and molecules 12
1-8 The periodic system of the elements 14
CHAPTER 2. ATOMS, ELECTRONS, AND RADIATIONS 19

2-1 Faraday's laws of electrolysis and the electron 19
2-2 The conduction of electricity in gases 20
2-3 Cathode rays 20
2-4 The electron: the determination of its charge 24
2-5 Positive rays 31
2-6 X-:ays 31
2-7 Radioactivity: alpha-, beta-, and gamma-rays 32
2-8 The detection and measurement of radiation 35
CHAPTER 3. THE NUCLEAR ATOM 51

3-1 The Thomson atom 51
3-2 Rutherford's theory of the scattering of alpha-particles 53
3-3 The experimental test of the Rutherford scattering theory 58
3-4 Sorne characteristics of the atomic nucleus 62
CHAPTER 4. X-RAYS AND ATOMIC STRUCTURE. 67

4-1 Sorne properties of x-rays . 67
4-2 The scattering of x-rays by atoms and the number
of electrons per atom . 70
4-3 The diffraction of x-rays and Bragg's law . 72
4-4 Characteristic x-ray spectra. Moseley's law 76
CHAPTER 5. THE QUANTUM THEORY OF RADIATION 84

5-1 The failure of classical physics to describe atomic phenomena 84
ix
X COXTEXTS
Page
5-2 The emission and absorption of thermal radiation 86
5-3 The classical theory of thermal radiation 89
5-4 The failure of the classical theory of thermal radiation 93
5-5 Planck's quantum theory of thermal radiation 96
5-6 Quantum theory and the photoelectric effect 99
CHAPTER 6. THE SPECIAL THEORY OF RELATIVITY 104

6--1 The role of the special theory of relativity in atornic
and nuclear physics 104
6--2 The ether and the problem of absolute velocity 105
6--3 The Michelson-Morley experiment . 107
6--4 The problem of the invariant form of physical theories 110
6--5 The special theory of relativity; the variation of masa
with velocity 113
6--6 The relationship between mass and energy . 124
6--7 The Compton effect 127
CHAPTER 7. ATOMIC SPECTRA AND ATOMIC STRUCTURE 135

7-1 Atomic spectra . 135
7-2 The Bohr theory of atornic spectra and atomic structure 139
7-3 The stationary states of an atom 145
7-4 Extension of the Bohr theory: elliptic orbits 149
7-5 Failure of the Bohr theory: wave mechanics and the correct
quantum numbers . 154
7-{i Atomic theory and the periodic table 156
7-7 Atomic theory and characteristic x-ray spectra 159
7-8 The ha.sic ideas of wave mechanics 164
7-9 The solution of the Schroedinger equation: sorne uaeful
examples and results 171
PART 11. THE NUCLEUS
CHAPTER 8. THE CoNSTITUTION OF THE NucLEUS 183

8-1 The proton-electron hypothesis of the constitution of the nucle118 183
8-2 The angular momentum of the nucleus; failure of the
proton-electron hypothesis 185
8-3 Nuclear transmutation and the discovery of the neutron . 189
8-4 The proton-neutron hypothesis 191
8-5 Magnetic and electric properties of the nucleus 193
8--{) Additional properties of atornic nuclei 195
CHAPTER 9. ISOTOPES 197

9-1 Natural radioactivity and isotopes 197
CONTENTS xi
Page
9-2 Positive-ray analysis and the existence of isotopes 198

9-3 Isotopic masses and abundances: the mass spectrograph
and mass spectrometer 203
9-4. The stable isotopes of the elements and their
percentage abundances 209
9-5 Atomic masses: packing fractions and binding energies 218
CHAPTER 10. NATURAL RADIOACTIVITY AND THE LAWS

0F RADIOACTIVE TRANSFORMATION . 229
10-1 The basis of the theory of radioactive disintegration 229

10-2 The disintegration constant, the half-life, and the mean life 234
10-3 Successive radioactive transformations 239
lo-4 Radioactive equilibrium 244
10-5 The natural radioactive series 248
lo-6 Units of radioactivity . 253
CHAPTER 11. ARTIFICIAL NUCLEAR DISINTEGRATION . 258
11-1 Transmutation by alpha-particles: alpha-proton reactions 258

11-2 The balance of mass and energy in nuclear reactions 261
11-3 The neutron: alpha-neutron reactions 265
11-4 The acceleration of charged particles 268
11-5 Transmutation by protons 269
11-6 Transmutation by deuterons 272
11-7 Transmutation by neutrons 273
11-8 Transmutation by photons 275
11-9 Nuclear chemistry: nuclear masses 276
CHAPTER 12. ARTIFICIAL RADIOACTIVITY 283

12-1 The discovery of artificial radioactivity . 283
12-2 The artificial radionuclides. Electron and positron
emission. Orbital electron capture . 285
12-3 The transuranium elements 290
12-4 The artificial radionuclides: alpha-emitters . 295
12-5 Isotope tables and nuclide charts 297
CHAPTER 13. ALPHA-DECAY 304

13-1 The velocity and energy of alpha-particles 304
13-2 The absorption of alpha-particles: range, ionization,
and stopping power 307
13-3 Range-energy curves . 317
13-4 Alpha-part1cle spectra. Long-range particles and fine structure . 319
xii CONTENTS
PQJJe
13-5 Nuclear energy levels . 322
13-6 The theory of alpha-decay 326
CHAPTER 14. BETA-DECAY 340

14-1 The velocity and energy of beta-particles 340
14-2 The absorption of beta-particles. Range, ionization,
and energy loss . 346
14-3 Range-energy relations for beta-particles 350
14-4 Beta-particle spectra. The continuous spectrum 352
14-5 The theory of beta-decay. Basis of the theory 358
14-6 The theory of beta-decay. Resulte and comparison
with experiment 363
14-7 Energy levels and decay schemes 371
14-8 The neutrino 374
14-9 Symmetry laws and the nonconservation of parity in ¡S-decay 381
CHAPTER 15. GAMMA-RAYS AND GAMMA-DECAY 395

15-1 The absorption of gamma-rays by matter: experimental data 39~
15-2 The interaction of gamma-rays with matter 400
15-3 Photoelectric absorption . 402
15-4 Compton scattering 405
15-5 Electron-positron pair formation 410
15-6 The absorption of gamma~rays by matter. Comparison
of experimental and theoretical resulte . 414
15-7 The measurement of gamma-ray energies 418
15-8 Gamma-decay: interna) conversion . 425
15-9 Gamma-decay and nuclear energy Ievels: theory 428
15-10 Gamma-decay and nuclear energy levels: experimental resulte
and nuclear isomerism 433
CHAPTER 16. NUCLEAR REACTIONS • 447

16-1 Nuclear reactions.and excited states of nuclei 447
16-2 The compound nucleus 452
16-3 Cross sections for nuclear reactions . 457
16-4 Experimental resulte. General considerations 463
16-5 Neutron-induced reactions 465
16-6 Reactions induced by protons and alpha-particles . 477
16-7 Limitations of the compound nucleus theory 484
16-8 Deuteron-induced reactions: intermediate nuclei, high energy 487
16-9 Reactions induced by gamma-rays . 491
16-10 Reactions at ultra-high energies . 496
16-11 Reactions with light nuclei: energy levels of light nuclei 496
CONTENTS xiii
Page
CHAPTER 17. NUCLEAR FORCES AND NUCLEAR 8TRUCTURE . 508
17-1 Nuclear binding energies and the saturation of nuclear forces 508
17-2 Nuclear stability and the forces between nucleons 510
17-3 Energy levels of light nuclei and the hypothesis of the charge
independence of nuclear forces 512
17-4 The nuclear radius . 517
17-5 The interaction between two nucleons 518
17-6 The status of the problem of nuclear forces 523
17-7 Nuclear models. The shell, or independent particle, model 526
17-8 The liquid drop model and the semiempirical binding
energy formula . 537
17-9 The collective nuclear model . 542
17-10 The optical model for nuclear reactions . 544
PART III. SPECIAL TOPICS AND APPLICATIONS
ÜHAPTER 18. NEUTRON PHYSICS . 559
18-1 The production of neutrons 559

18-2 The detection of neutrons 563
18-3 The interaction of neutrons with matter in bulk: slowing down 565
18-4 Thermal neutrons . 574
18-5 The diffusion of thermal neutrons 580
18-6 Cross sections for neutron-induced reactions: measurement
of the total cross section . 584
18-7 Scattering, absorption, and activation cross sections 594
CHAPTER 19. NUCLEAR F1sSION . 606

19-1 The discovery of nuclear fission 606
19-2 Fission cross sections and thresholds 608
19-3 The fission products 614
19-4 The mass and energy distributions of the fission products 616
19-5 Neutron emission in fission 619
19-6 The energy distribution of the neutrons emitted in fission 623
19-7 The energy release in fission . 625
19-8 The theory of the fission process . 626
CHAPTER 20. NuctEAR ENERGY SouRcEs 637

20-1 Nuclear fission as a source of energy 637
20-2 The chain-reacting system or nuclear reactor 639
20-3 Thermal nuclear reactors. The neutron cycle 644
20-4 The calculation of the multiplication factor for a
homogeneous thermal reactor 648
xiv CONTENTS
Page
20-5 The heterogeneous thermal reactor . 656
20-6 The critica! size oí a thermal reactor 660
20-7 Power and breeding 664
20-8 Energy production in stars. Thermonuclear reactions 666
20-9 Controlled thermonuclear reactions . 671
CHA.PTER 21. THE AccELERA.TION OF CHARGED PARTICLES 679

21-1 The Cockcroft-Walton machine . 679
21-2 The electrostatie generator or Van de Graaff machine 681
21-3 The cyclotron . 683
21-4 The frequency-modulated cyclotron or synchrocyclotron . 689
21-5 The acceleration of electrons. The betatron and
the electron synchrotron 692
21-6 The proton synehrotron ·. 697
21-7 Linear accelerators 700
21-8 The alternating-gradient synchrotron 701
APPENDIX l. ALPHABETICAL LIST OF THE ELEMENTS 713
APPENDIX II. VALUES OF PHYSICAL CONSTANTS . 715
APPENDIX III. JouRNAL ABBREVIATIONS FOR REFERENCES 717
APPENDIX IV. THE RUTHERFORD SCATI'ERING FQRMULA. 718
APPENDJX V. NucLIDE CHART lnside back cover
ANSWERS TO PROBLEMS • 727
lNDEX • 759
Part I
The Background of Nuclear Physics
CHAPTER 1
THE CHEMICAL FOUNDATIONS OF ATOMIC THEORY
The foundations of modern atomic theory were laid in the late 18th
and 19th centuries in the attempt to understand the chemical properties
of matter. Two of the fundamental quantities of atomic and nuclear
physics-atomic weight and atomic number-had their origins in the
correlation of the results of chemical experiments and in the systemization
of the properties of the chemical elements. The laws of chemical combina-
tion were unified at the beginning of the 19th century by Dalton's atomic
theory, which introduced the concept of atomic weight. The develop-
ment of methods for determining the atomic weights of the elements and
the investigation of chemical reactions were major contributions of 19th
century chemistry. The systemization of the atomic weights and prop-
erties of the elements then led to the formulation of the periodic system
and to the concepts of atomic number and atomic structure.
1-1 The laws of chemical combination. The experimental information

which gave rise to the atomic theory of matter can be summarized in a
few basic laws. The first of these was deduced from investigations such
as those of Lavoisier, who showed that when tin is made to react with
áir in a closed vessel, the weight of the vessel and its contents before and
after heating is the same. This constancy of the weight, which has been
found to be true for all chemical reactions, is expressed in the law of con-
servation of mass, which state$ that the mass of a system is not affected by
any chemical change within the system.
It was found that when various metals are oxidized in excess air, one
part by weight of oxygen always combines with 1.52 parts by weight of
magnesium, 2.50 parts of calcium, 1.12 parts of aluminum, 3.71 parts of
tin, 3.97 parts of copper, and so on. If, as turns out to be more convenient,
the weight of oxygen is taken to be 8, the following combining weights
are obtained.
Oxygen Magnesium Calcium Aluminum Copper Tin
8 12.16 20.04 8.96 31.76 29.68
The same combining weights are obtained when the oxides are prepared
by methods other than simple oxidation. These experimental results,
and many others like them, are expressed by the law of definite propor-
tions, which states that a particular chemical compound always contains
the same elements united together in the same proportions by weight.
3
4 CHEMICAL FOUNDATIONS OF ATOMIC THEORY [cHAP. 1
It was also found that two elements can combine to forro more than one
compound. For example, at least five distinct oxides of nitrogen are
known in which the relative proportions by weight of nitrogen and oxygen
are listed in the table.
Nitrogen Oxygen
14 8 = 1X 8
14 16 = 2 X 8
14 24=3X8
14 32=4X8
14 40=5X8
The different weights of oxygen which can combine with the same weight
of nitrogen are integral multiples of 8. Resulta of this kind are sum-
marized by the law of multipl.e proportiom, which states that if two ele-
ments combine to fonn more than one compound, the different weights of
one which combine with the same weight of the other are in the ratio of
small whole numbers.
The further study of the quantitative relations between the elements
led to a fourth law of chemical combination. The law of reciprocal pro-
portions states that the weights of two (or more) substances which react
separately with identical weights of a third are also the weights which
react with each other, or simple multiples of them. In other words, if
each of two substances A and B combines with a substance C, then A and
B can combine with each other only in those proportions in which they
combine with C, or in sorne simple multiple of those proportions. This
law is illustrated by the reactions which take place between oxygen and
sulfur, oxygen and zinc, and sulfur and zinc. Thus, 8 parts by weight
of oxygen combine with 8.015 parts of sulfur to fonn an oxide of sulfur,
and with 32.69 parts of zinc to fonn an oxide oí zinc; in the fonnation
of zinc sulfide, 65.38( = 2 X 32.69) parts oí zinc combine with
32.06( = 4 X 8.015) of sulfur.
Another generalization deduced from the analysis of chemical compounds
is that of the combining weight or chemical equivalent of an element.
This concept was touched upon in the discussion oí the oxides oí various
meta.Is in connection with the law of definite proportions. Consider a
number oí chemical compounds such as those tabulated on the next page.
If oxygen = 8 is taken as the standard, the amount of ea.ch of the other
elements which combines with this standard amount of oxygen can be
calculated. Thus, 28.53 parts by weight oí oxygen combine with 71.47
parts of calcium, and
28.53:8 = 71.47:x, or x = 20.04 for calcium,
1-2] DALTON'.S ATOMIC HYPOTHESIS 5
Compound Percent Percent
1. Calcium oxide Calcium 71.47 Oxygen 28.53

2. Water Hydrogen 11.19 Oxygen 88.81
3. Hydrogen chloride Hydrogen 2.76 Chlorine 97 .23
4. Magnesium chloride Magnesíum 25.53 Chlorine 74.47
5. Silver chloride Silver 75.26 Chlorine 24.71
6. Silver iodide Silver 45.94 Iodine 54.06
where x is the amount of the element, in this case calcium, which combines
with 8 parts of oxygen. From the experimental data for water,
88.81:8 = 11.19:x, or x = 1.008 for hydrogen.
In hydrogen chloride, if the weight of hydrogen is taken as 1.008, that of
chlorine is 35.45; in magnesium chloride, if the weight of chlorine is 35.45,
that of magnesium is 12.16. When this procedure is continued, it is found
that a number can be assigned to each element which represents the number
of parts by weight of the given element that can combine with 8 parts by
weight of oxygen or 1.008 parts by weight of hydrogen.
Oxygen Calcium Hydrogen Chlorine Magnesium Silver Iodine
8 20.04 1.008 35.45 12.16 107.88 126.9
The value obtained for magnesium by the above indirect process is the
same as that obtained earlier from data on the direct oxidation of mag-
nesium. The numbers obtained for the different elements are called the
comllining weights or equivalent weights of the elements.
The four laws of chemical corobination, conservation of mass, definite
proportions, multiple proportions, and reciproca! proportions, together
with the concept of combining weights, summarized the basic experi-
mental facts of chemical combination as known about 1800, and led to
the atomic hypothesis proposed early in the 19th century by Dalton.
1-2 Dalton's atomic hypothesis. Dalton's atomic hypothesis was

proposed (1803) to account for the facts expressed by the laws of cheroical
combination and was based on the following postulates:
l. The chemical eleroents consist of discrete particles of matter, atoros,
which cannot be subdivided by any known chemical process and which
preserve their individuality in chemical changes.
2. Ali atoms· of the saroe eleroent are identical in all respects, particularly
in weight or roass; different elements have atoms differing in weight.
Each element is characterized by the weight of its atoro, and the corobining
weights of the elements represent the combining weights of their respective
atoms.
6 CHEMICAL FOUNDATIONS OF ATOMIC THEORY [CHAP. 1
3. Chemical compounds are formed by the union of atoros of different

elements in simple numerical proportions, e.g., 1:1, 1:2, 2:1, 2:3.
It is easy to show that the laws of chemical combination can be deduced
from these postulates. Since atoros undergo no change during a chemical
process, they preserve their masses, and the mass of a compound is the
sum of the masses of its elements; the result is the law of conservation
of mass. Since ali atoms of the same element are identical in weight,
and a compound is formed by the union of atoms of different elements in
a simple numerical proportion, the proportions by weight in which two
elements are combined in a given compound are always the same, giving
the law of definite proportions.
Consider next the case in which two elements A and B can forro two
different compounds. Suppose that the first compound contains m atoms
of A and n atoms of B, and that the second compound contains p atoms
of A and q atoros of B. If a is the weight of an atoro of A, and b is the
weight of an atom of B, then the first compound contains ma parts of
A and nb parts of B, and one part of A combines with nb/ma parts of B.
Similarly, in the second compound, one part of A combines with qb/pa
parts of B. Hence, the weights of B combined with a fixed weight of A
are in the proportion
or np:mq.
According to the third postulate, n, m, p, and q are ali small integers;

the products np and mq are therefore also small integers and the weights
of B combining with a fixed weight of A are in the ratio of two small
integers. This treatment can be extended to any number of compounds
formed by two elements, giving the law of multiple proportions. Similar
arguments lead to the law of reciproca! proportions. Compounds of
the elements A and B are formed according to the scheme m atoros of
A and n atoms of B. Compounds of the elements A and C contain p atoros
of A and q atoms of C, and compounds of the elements B and C contain
x atoms of B and y atoros of C. If a, b, ande are the weights of the atoros A,
B, and C, respectively, then one part of A combines with nb/ma parts of
B, and qc/pa parts of C. It follows that
mq_b
-.-•
np e
In the compound formed by B and C, the proportíon by weight of B and
e is
xb:yc, or
X C
Hence,
1!.=mq_
X np
1-3) AVOGADRO'S HYPOTHESIS AND THE MOLECULE 7
Since, by the third postulate, all of the quantities in the last equation
are small integers, then y, x are the same as q, n or small integral multiples
of them. This is the law of reciprocal proportions.
The laws of chemical combination have thus been deduced from Dalton's
postulates, and the latter form the basis of an atomic theory of matter.
Dalton's theory was incomplete, however, because it provided no way of
determining even the relative weights of the atoros of the different elements.
This difficulty arose beca.use Dalton had no way of finding out how many
atoros of ea.ch element combine to forro a compound. If W 1 and W 2 are
the weights of two elements which combine to form a compound, then
W1 n1A1
W2 = n2A2'
where A 1 and A 2 are the atomic weights and n 1 and n 2 are the whole
numbers of atoros of ea.ch element which enter into corobination. When
the ratio n 1 :n 2 is known, the value of the ratio W 1 :W 2 can fix only the
ratio of the atoroic weights. To apply his theory, Dalton was forced to
make arbitrary assuroptions; he assumed, for exarople, that if only one
coropound of two elements is known, it contains one atom of each element.
Water was regarded as a coropound of one atoro of hydrogen and one
atoro of oxygen, the existence of hydrogen peroxide being unknown at
the time. Dalton's assumption was simple, but wrong, and led to many
difficulties in the application of his atomic theory to the rapidly growing
field of chemistry.
1-3 Avogadro's hypothesis and the molecule. One of the difficulties

met by the atomic hypothesis arose as a result of studies of the combining
properties of gases. Gay-Lussac showed (1805-1808) that when chemical
reactions occur between gases, there is always a simple relation between
the volumes of the interacting gases, and also of the products, if these
are gases. When the reacting gases and the products are under the same
conditions of temperature and pressure
1 volume of hydrogen + 1 volume of chlorine

- + 2 volumes of hydrogen chloride;
2 voluroes of hydrogen + 1 volume of oxygen
- + 2 volumes of steam;
3 volumes of hydrogen + 1 volume of nitrogen
- + 2 volumes of ammonia.
It follows that if elements in a gaseous state combine in simple proportions

by volume, and if the elements also combine in simple proportions by
atoros, then the numbers of atoros in equal volumes of the reacting gases
must be simply related. Dalton assumed that equal volumes of different

gases under the same physical conditions contain an equal number, say n,
of atoros. Under this assumption, when 1 volume (n atoros) of hydrogen
reacts with 1 volume (n atoros) of chlorine, then 2 volumes (2n "compound
atoros") of hydrogen chloride are forroed. Every atoro of hydrogen and
chlorine must then be split in half to forro two compound atoros of hydrogen
chloride, a result which contradicts that postula.te of Da.lton's hypothesis
according to which atoms cannot be subdivided by any chemical process.
Avogadro (1811) showed that the difficulty could be resolved if a dis-
tinction is made between elementary atoros and the small particles of a gas.
He assumed that the latter are aggregates of a definite number of atoros,
and called the aggregates mol«ules in order to distinguish them from the
elementary atoros. He then postulated that equal volumes of ali gases
under the se.me physical conditions contain the same number of molecules.
This hypothesis made it possible to interpret Gay-Lussac's law of com-
bining volumes in terms of the atomic hypothesis. Assume that each
molecule of hydrogen or chlorine consists of two elementary atoros, and
suppose that 1 volume oí hydrogen or chlorine contains n molecules.
These molecules rea.et to forro 2n molecules of hydrogen chloride, each
molecule containing one atom of hydrogen and one atom of chlorine.
Although the atoros cannot be split so that one atom of hydrogen or
chlorine enters into the composition of two molecules of hydrogen chloride,
one molecule of hydrogen and one molecule of chlorine can be divided
between two molecules of hydrogen chloride. In this way, Avogadro's
hypothesis allowed Gay-Lussac's law to be reconciled with Dalton's
atomic hypothesis. The detailed analysis of many reactions between
gases has shown that the molecules of the gaseous elements hydrogen,
oxygen, chlorine, and nitrogen contain two atoros; the inert gases helium,
neon, argon, krypton, and xenon have one atom per molecule.
1-4 Molecular and atomic weights of gaseous elements. The relative

molecular weights of gaseous substances can be determined from measure-
ments of relative densities. The relative density of a gas is defined as
the ratio of the weight of a. given volume of the gas to that oí an equal
volume of a standard gas measured at the same temperature and pressure.
In view of Avogadro's hypothesis, it is possible to write for any gaseous
substance
- - Molecular
----- weight of gas
~ - - - - - = - -Weight
- - -of-any
- -volume
- - -of~
gas
--
Molecular weight oí standard gas Weight oí equal volume of standard
Density of gas
=---------
Density of standard gas
1-4] MOLECULAR AND ATOMIC WEIGHTS OF GASEOUS ELEMENTS 9
TABLE 1-1
MOLECULAR WEIGHTS OF GASEOUS ELEMENTS
Molecular weight
Density,
g/l at M (hydrogen) M (oxygen) Molecular weight
Element N.T.P. =2 = 32 M (oxygen) = 32*
Hydrogen 0.08988 2.000 2.ol3 2.016

Oxygen 1.42904 31.816 32.000 32.000
Nitrogen 1.25055 27.828 28.000 28.016
Fluorine 1.696 37.738 37.977 38.00
Chlorine 3.214 71.52 71.97 70.914
Helium 0.17847 3.971 3.996 4.003
Neon 0.90035 20.034 20.161 20.183
Argon 1.7837 39.690 39.942 39.944
Krypton 3.708 82.510 83.03 83.7
Xenon 5.851 130.19 131.02 131.3
* Corrected for deviations from ideal gas laws.
The choice of the standard gas is arbitrary and may be, for example,
hydrogen, oxygen, or air. If hydrogen is chosen as the standard, and if
its atomic weight is arbitrarily taken to be unity, the relative molecular
weight of a gas is
Molecular weight = 2 X density of gas relative to that of hydrogen,

since a molecule of hydrogen contains two atoms. When the relative
molecular weight of a gaseous element has been determined in this way,
its relative atomic weight can be determined when the number of atoms
per molecule is known.
The results for sorne molecular weight determinations based on this
method are shown in Table 1-1. The densities of the gases are given in
grams per liter measured under standard conditions (OºC and 760 mm
mercury pressure, abbreviated as N.T.P.). The third column gives the
molecular weight relative to that of hydrogen taken as 2. Oxygen, with
its molecular weight arbitrarily taken as 32, may also be used as a standard,
and the molecular weights relative to this standard are given in the fourth
column. High-precision work on the combining volumes of ·gases has
shown that there are slight deviations from Gay-Lussac's law which ha.ve
been traced to deviations from the ideal gas laws. When corrections are
made for these deviations, the results of the fifth column are obtained for
the molecular weights relative to that of oxygen taken as 32. Of the
gaseous elements listed, hydrogen, oxygen, nitrogen, fluorine, and chlorine
are diatomic and their relative atomi~ weights are half of the molecular
weights; the relative atomic weights of the inert gases are the same as
their molecular weights.
1-5 The standard atomic weight. It has been emphasized that the
values of the molecular and atomic weights determined from density
measurements are relative, not absolute, values. For practica! purposes,
it is necessary to fix a standard for atomic weights, and this standard is
taken as oxygen with an atomic weight of 16 and a molecular weight of
32. The choice of oxygen was based on the criterion of convenience.
Oxygen is a useful standard for chemists because nearly ali the elements
form stable compounds with oxygen, and it was thought that the determi-
nation of the atomic weight of an element should be connected with the
standard as closely as possible. Hydrogen was sometimes used as a stand-
ard, but it forms very few compounds with metals which are suitable for
atomic weight determinations; consequently hydrogen was rejected in
favor of oxygen.
The choice of the value of the atomic weight of oxygen was based on
the desire for convenient values for the other atomic and molecular weights.
Experimental evidence shows that 1.008 parts by weight, say grams, of
hydrogen combine. with 8 gm of oxygen. It has also been found that
two atoms of hydrogen combine with one atom of oxygen to form one
molecule of water. Hence, 1.008 gm of hydrogen contain twice as many
atoms as 8 gm of oxygen. A possible choice for the atomic weight of oxygen
is its combining, or equivalent, weight, 8 units. If this choice were made,
then 1.008 (of the same units) would be the weight of two atoms of hydro-
gen, and the atomic weight of hydrogen would be 0.504. Now, hydrogen is
the lightest of ali the elements. If the standard is chosen so that the
atomic weight of hydrogen is equal to or greater than unity, there will be
no element with an atomic weight smaller than unity. The choice of 16 for
oxygen makes the atomic weight of hydrogen equal to 1.008, and also leads
to relative weights of heavier atoms and molecules which are not incon-
veníently Jarge. The atomic weight of oxygen is then equal to twice the
equivalent or combining weight. In general) the atomic weight of an
element is a small integral multiple of the equivalent weight, and this integer
is called the valence of the element:
Atomic weight _ V
. . h - a 1ence.
E qmva1ent weig t
The equivalent weight and the valence are useful because their values can
be determined by chemical methods and lead to the determimation of
atomic weights.
1-6] ATOMIC WEIGHTS OF NONGASEOUS ELEMENTS 11
TABLE 1-2
THE ATOMIC WEIGHT OF CARBONAS FouND
FROM THE RELATIVE DENSITIES OF CARBON CoMPOUNDS
Relative Weight of carbon

density, Molecular in one molecular
Compound Hydrogen = 1 weight weight of compound
Methane 8 16 12
Ethane 15 30 24 = (2 X 12)
Alcohol 23 46 24 = (2 X 12)
Ether 37 74 48 = (4 X 12)
Benzene 39 78 72 = (6 X 12)
Carbon monoxide 14 28 12
Carbon dioxide 12 44 12
1-6 Atomic weights of nongaseous elements. The method of relative

densities can be used to find the approximate atomic weight of a non-
volatile element which has gaseous compounds. The molecular weights
of as many volatile compounds of the element as possible are found from.
the relative densities. The weights of the particular element in the molec-
ular weights of the various compounds are found by chemical analysis.
These weights must be integral multiples of the atomic weight and if the
number of compounds used is large enough, at least one of the weights of
the element present in the molecular weights of its compounds wíll prob-
ably be the atomic weight itself. The smallest weight of the element
present in one molecular weight of all its known volatile compounds is
assumed to be the atomic weight of the element relative to oxygen = 16.
In the case of carbon, the resulta of Table 1-2 are obtained, and the atomic
weight of carbon is taken to be 12. The values given in the table are ap-
proximate; more refined measurements and corrections for deviations
from the ideal gas laws must be made in order to get accurate values of
the Btomic weight. In general, accurate values of the molecular and atomic
weights are found from careful chemical analysis of the compounds, and
the relative density measurements are used only to decide between various
· possible molecular weights.
Approximate values of the atomic weights of solid elements which have
few, or no, volatile compounds can be found with the aid of the rule of
Dulong and Petit. According to this rule, the product of the atomic
weight and the specific heat is roughly the same for many solid elements.
When the specific heat is expressed in calories per gram, the numerical
value of the product is usually between 6 and 7, with an average of about
6.4. If the specific heat of a solid element is measured, an approximate
12 CHEMICAL FOUNDATIONS OF ATOMIC THEORY [CHAP, 1
TABLE 1-3
ATOMIC WEIGHTS OF CoMMON ELEMENTS RELATIVE TO ÜXYGEN = 16
Element Atomic weight Element Atomic weight
Aluminum 26.98 Lead 207.21

Barium 137.36 Mercury 200.61
Boron 10.82 Nitrogen 14.008
Calcium 40.08 Oxygen 16.000
Carbon 12.011 Potassium 39.100
Chlorine 35.457 Silicon 28.09
Copper 63.54 Silver 107.880
Fluorine 19.00 Sulfur 32.066
Gold 197.0 Uranium 238.07
Hydrogen 1.0080 Zinc 65.38
Iron 55.85
value of the atomic weight is found. This value can then be used to
determine which of the different values represented by the combining
proportions by weight is the best choice for the atomic weight. For ex-
ample, the specific heat of copper is 0.095 cal/gm. By the rule of Dulong
and Petit,
/~5 = 67.3
is the approximate atomic weight of copper. Chemical analyses of com-

pounds of copper with oxygen show that 63.6 and 127.2 parts of this
element combine with 16 parts of oxygen. Since 67.3 is much closer to
63.6 than 127.2, there can be little doubt that 63.6 is the atomic weight of
copper. More refined chemical analysis gives the value 63.54.
Various other methods have been developed for determining accurate
values of molecular and atomic weights, anda list of the presently accepted
values of the atomic weights of sorne of the elements is given in Table 1-3.
1-7 Weights and sizes of atoms and molecules. When the standard
for atomic and molecular weights has been chosen, it is possible to calcu-
late the volume occupied at standard temperature and pressure by a gram-
molecular weight of a gas, that is, by the number of grams of gas numeri-
cally equal to the molecular weight; this volume is called the gram-molecular
volume or molar volume. Since the atomic weight of hydrogen is taken to
be 1.008, its molecular weight is 2.016, and a gram-molecular weight of
hydrogen is 2.016 gm. The density of hydrogen (Table 1-1) is very nearly
0.09 gm/liter at K.T.P., so that the gram-molecular volume is
2.016 22 1·
0.09 = .4 1ters.
1-7) WEIGHTS AND SIZES OF ATOMS AND MOLECULES 13
Experiments with a large number of gases have shown that the same
volume, 22.4 liters, is occupied by one gram-molecular weight, or mole; of
any gas at N.T.P. According to Avogadro's hypothesis, this volume con-
tains the same number of gas molecules regardless of the chemical nature of
the gas. It follows, therefore, that the gram-molecular weights of all
gases at N.T.P. contain the same number of molecules; this number is
called the Avogadro number or Avogadro' s constant. Its value has been
determined by several independent methods, with results which agree
very well; one method will be described in Chapter 2. The accepted value
at present is 6.0249 X 10 23 • This number is also the number of atoms in a
gram-atomic weight of a gaseous element, and hence of any element. The
weight in grams of a single atom or molecule can be found by dividing the
gram-atomic or gram-molecular weight by the Avogadro number. The
weight of the lightest atom, hydrogen, is 1.67 X 10- 24 gm, while that of
the heaviest naturally occurring element, uranium, is 3.95 X 10- 22 gm.
An estímate of the sizes of molecules and atoms can be made by con-
sidering water, which has a molecular weight of 18 and a density of one
gram per cubic centimeter. One gram-molecular weight of water occupies
18 cm 3 and contains 6.0 X 10 23 molecules. If the water molecules were
cubic in shape so that they packed together without any free space, the
volume of a single molecule would be 3 X 10- 23 cm 3. If the molecules are
assumed to be spherical, their volume is about one-third smaller, or
2 X 10- 23 cm 3. The radius of a water molecule is then given by the
relation
or
r = 1.7 X 10-s cm.
Values of the radii of gas molecules can also be calculated with the aid of
the kinetic theory of gases from measured values of the viscosity, the
diffusion constant, and the thermal conductivity. The radii of hydrogen
and oxygen molecules obtained in this way are 1.37 X 10-s cm and 1.81 X
10-s cm, respectively, and are quite close to the value of the radius of the
water molecule given by the above rough calculation. Similar values, in
the neighborhood of 1 X 10-s cm to 3 X 10-s cm, are obtained for the
molecular radii of most of the common gaseous elements and compounds.
Since the number of atoros per molecule of these substances is small,
10-s cm may be taken as the order of magnitude of atomic radii.
An estímate of the atomic radius of a solid element can be made from the
atomic volume, that is, the volume of a gram-atom of the element, at the
melting point. In the solid phase at the melting point, the atoms are
closely packed and the distance between· them is not much greater than the
atomic dimensions. Solid lead, at its melting point, has a density of 11
gm/cm 3 and its atomic volumé is 18.9 cm 3. The volume per atom is
3.15 X 10- 23 cm3, and if the atom is assumed cubica! (no free space), the
atomic dimension would be about 3.2 X 10-8 cm. If the lead atoro is
assumed to be spherical and the space between atoros is taken into account,
the radius is about 1.6 X 10- 8 cm. With better methods for determining
the atomic radius, the following values have been obtained for some of the
solid elements.
Carbon: 0.77 X 10-8 cm Cesium: 2.6 X 10-8 cm

Aluminum: 1.45 X 10-8 cm Tin: 1.4 X 10-8 cm
Sodium: 1.9 X 10-8 cm Bismuth: 1.5 X 10-8 cm
These values are similar to those obtained for the atomic radii of the gaseous
elements. They indicate that atoros have dimensions of the order of 10-s
cm regardless of the atomic weight, since elements of low, intermediate,
and high atomic weight are included in the list. Cesium is the atoro with
the largest radius.
1-8 The periodic system of the elements. As information about the

chemical and physical properties of the elements was accumulated, it
became apparent that there are similarities in the properties of certain
groups of elements, and attempts were made to classüy the elements
according to such similarities.. The most successful attempt was that of
Mendeléeff, who proposed bis periodic law in 1869. This law may be
summarized in the following statements made by Mendeléeff.
l. The elements, ü arranged according to their atomic weights, show a
periodicity of properties.
2. Elements which are similar as regards their chemical properties
have atomic weights which are either approximately the same (e.g., plat-
inum, iridium, osmium) or which increase regularly (e.g., potassium,
rubidium, cesium).
3. The arrangement of the elements or of groups of elements in the
order of their atomic weights corresponds with their valences.
4. The elements which are the most widely distributed in nature have
small atomic weights, and ali the elements of small atomic weights are
characterized by sharply i:lefined properties.
5. The magnitude of the atomic weight determines the character of an
element.
6. The discovery of many yet unknown elements may be expected,
for instance, elements analogous to aluminum and silicon.
7. The atomic weight of an element may sometimes be corrected with
the aid of knowledge of the weights of adjacent elements.
8. Certain characteristic properties of the elements can be predicted
from their atomic weights.
1-8] THE PERIODIC SYSTEM OF THE ELEMENTS 15
Mendeléeff's general conclusions have been verified ancl extended.

Among the properties which show the kind of periodic variation noted by
Mendeléeff are atomic volume (atomic weight/density), thermal expan-
sion, thermal and electrical conductivity, magnetic susceptibility, melting
point, refractive index, boiling point, crystalline form, compressibility,
heats of formation of oxides and chlorides, hardness, volatility, ionic
mobility, atomic heats at low temperature, valence, and others.
Mendeléeff constructed a table in which the elements were arranged
horizontally in the order of their atomic weights and vertically according
to their resemblances in properties. The early tables were imperfect,
sometimes because of errors in the atomic weights, and because of ignorance
of the inert gases. In spite of these imperfections, Mendeléeff was able
to predict the discovery of certain elements, as well as the properties of
these elements. Modern versions of the periodic table, such as that shown
in Table 1-4, have rectified these errors. The table consists of a number
of horizontal rows called periods, containing 2, 8, 8, 18, 18, 32, and 18
(incomplete) elements, respectively. In each period there is a definite and
characteristic gradation of properties from one element to the next. The
vertical columns, labeled by the Roman numerals I to VIII and by the
letter O are called groups. Groups I to VII are each divided into sub-
groups; the subgroup on the left side of the column is usually called the
"a" subgroup, that on the right side the "b" subgroup. In each subgroup,
the elements have similar properties, although there is a steady but gradual
variation with increasing atomic weight. The alkali metals constitute the
subgroup la, the alkaline earth metals 2a, oxygen-like elements 6a, halo-
gens 7a, and inert gases group O. Within a group, similarities of a minor
character exist between the members of the a subgroup and those of the
b subgroup.
For four pairs of elements, the order assigned in the periodic table
departs from that of the atomic weights, argon and potassium, cobalt and
nickel, tellurium and iodine, thorium and protoactinium. In each case
the order, as determined by the atomic weights, is reversed so that these
elements may fall into the places which correspond to their properties.
This was done for the first three pairs by Mendeléeff himself, who believed
that the accepted values of the atomic weights were inaccurate. Later
research has shown that the chemical atomic weights were not seriously
in error, and that the real property underlying the periodic classification
is the atomic number, to which the atomic weight is approximately pro-
portional.
The basis of Mendeléeff's classification of the elements, namely, the
requirement that the elements should be arranged in the order of increas-
ing atomic weight, offered no explanation of the different features of the
periodic table. So far as was known in Mendeléeff's time, atoms of dif-
TABLE 1-4. PERIODIC TABLE OF THE ELEMENTS
-
o,
Atomic weighta are baaed on the mOBt recent values adopte<! by the Intemational Union of Chemistry. (For artificially produced elementa, thc approximate atomic
weight of the mOBt stable isotope ia given in bracketa.)
Grou- I 11 III IV V VI VII Vlll o

Period Series
g
1
lH 2 He
4.003
~
~t<
1 1.0080
3 Li 4 Be 5B 6C 7N 80 9F 10 Ne
2 2 6.940 9.013 10.82 12.011 14.008 16.0000 19.00 20.183
11 Na 12Mg 13 Al 14Si 15 P 168 17 Cl 18 A óe::
3 3 22.991 24.32 26.98 28.09 30.975 32.066 36.457 39.944 :.,:
19K 20 Ca 21 Se 22Ti 23 V 24 Cr 25 Mn 26 Fe 27 Co 28 Ni ~
39.100 40.08 44.96 47.90 50.95 52.01 54.94 55.85 58.94 58.71 >-3
4
4
29 Cu 30 Zn 31 Ga 32 cie 33 Aa 34 Se 35 Dr 36 Kr
o
:,:
u,
5 63.54 65.38 69.72 72.60 74.91 78.96 79.916 83.80
37 Rb 38Sr 39 Y 40 Zr 41 Nb 42 Mo 43 Te 44 Ru 45 Rh 46 Pd
o"'l
6 85.48 87.63 88.92 91.22 92.91 95.95 199] 101.1 102.91 106.4
5 >
47 Ag 48 Cd 49 In 50 Sn 61 Sb 52 Te 531 54Xe ~
7 107.880 112.41 114.82 118.70 121.76 127.61 126.91 131.30 1(
55 Ca 56 !Ja 57-71 72 Hf 73 Ta 74W 75 Re 76 0a 77 Ir 78 Pt

ñ
190.2 >-3
6
8 132.91 137.36 Lanthanide
series•
178.50 180.95 183.86 186.22 192.2 195.09
=
o
t'l
79 Au 80 Hg 81 TI 82 Pb 831\i 84 Po 85 At 86 Rn
9 197.0 200.61 204.39 207.21 209.00 210 1210] 222 ~
89-
87 Fr 88 Ra Actinidc
7 10 1223] 226.05 serifll ..
·- n:i:
57 La 1 58Cc 1 59 Pr 1 60 Ne! 1 61 Pm 1 62Sm 1 63 Eu 1 640d 165Tb 66 Dy 67 Ho 68Er l 69Tm l 70Yb 171 Lu >
•La.nthanidc Sf'riNJ: 138.92 140.13 140.92 144.27 1147] 150.35 152.0 157.26 158.93 162.51 164.94 l 167.27 168.94 173.04 174.99 '."
89 Ac 90Tb 1 91 Pa 1 92 (l 93Np 1 94 Pu 95.\m 1 96Cm 197 llk 98 Cf 99 E 100 Fm 1101 Md 1102No ¡ 103
•• .\rtiniclf' m•ri("fl: 227 232.05 231 2311.07 12371 12421 12431 12451 12491 1249] 12531 1 12551 12561
REFERENCES 17
ferent elements had nothing in common and were unrelated. Since the
properties of an element are the properties of its atoros, it was difficult
to see why, for example, lithium, sodium, and potassium atoros should
have similar properties; it was evident that the atomic weights alone
could not explain the features of the table. The periodicity of the prop-
erties of the elements stimulated speculation about the structures of the
atoros of different elements. The gradual changes of properties from
group to group suggested that sorne unit of atomic structure is added in
successive elements until a certain portion of the structure is completed,
a condition which occurs, perhaps, in the inert gases. Beginning with the
element following an inert gas, a new portion of the structure is started,
and so on. Speculations like these introduced a new idea, that of the
structure of atoros. But before this idea could have any real physical
meaning, experimental information bearing on the possible structure of
atoros was needed. The methods and techniques of classical chemistry
could not supply this information, but new discoveries and techniques in
the field of physics opened the way to the understanding of atomic struc-
ture and the atomic number.
REFERENCES
GENERAL
J. W. MELLOR, Modern lnorganic Chemiatry, revised and edited by G. D.
Parkes. London: Longmans, Green and Co., 1951, Chapters 3, 4, 5, 6, 8.
F. T. BoNNER and M. PHILLIPS, Principlu of Phylical Science. Reading,
Mass., Addison-Wesley, 1957, Chapters 5, 6, 7, 8, 9.
G. HoLTON and D. H. D. ltoLLER, Foundationa of Modern Phyaical Science.
lteading, Mass., Addison-Wesley, 1958, Chapters 21, 22, 23, 24, 25.
T. G. CowLING, Molecules in Motion. London: Hutchinson and Co., 1950;
New York: Harper Torchbooks, 1960.
lt. D. PRESENT, Kinetic Theory of Gases. New York: McGraw-Hill, 1958.
D. C. PEASLEE, Elements of Atomic Physics. New York: Prentice-Hall, 1955,
Chapters 1, 2.
PROBLEMS
l. Show how the ti.ve oxides of nitrogen listed below illustrate the law of
multiple proportions.
Percentage composition
Compound Nitrogen Oxygen
Nitrous-oxide 63.63 36.37

Nitric oxide 46.67 53.33
Nitrous anhydride 36.~4 63.16
Nitrogen tetroxide 30.44 69.56
Nitric anhydride 25.93 74.04
2. The molecular weights of the ti.ve oxides listed in the above problem are
44.02, 30.01, 76.02, 46.01, and 108.02, respectively; calculate the atomic weight
of nitrogen from these data and the data of Problem l.
3. A certain element X forma several volatile compounds with chlorine, and
the lowest weight of the element is found in a compound which contains 22.54%
of X and 77.46% of chlorine. One liter of this compound at N.T.P. weighs 6.14 gm.
If the atomic weight of chlorine is 35.457, find (a) the equivalent weight of X,
(b) the approximate molecular weight of the compound, (e) the approximate
atomic weight of X, (d) the valence of X, (e) the exact atomic weight of X,
(f) the exact molecular weight of the compound. Which element is X? What is
the compound?
4. Chemical analysis shows that platinic chloride contains 48.8 parts of plati-
num combined with 35.46 parts of chlorine. The specific heat of platinum at
room temperature is 0.0324; what is the atomic weight of platinum?
5. A metal forms a chloride which contains 65.571 % of chlorine. The specific
heat of the metal is 0.112. Find (a) the approximate atomic weight of the metal,
(b) the equivalent weight, (e) the valence, (d) the exact atomic weight. What
is the compound?
6. The Handbook of Chemi8try and PhyBics, published by the Chemical Rubber
Company, Cleveland, Ohio, contains tables of various properties of the elements.
From information contained in the Handbook, calculate the atomic volumes of
the solid elements, and plot the values obtaíned as a function of the position of the
element in the periodic table. Plot the melting points of the solid elements as a
function of position in the periodic table. What periodic properties are observed
in the two figures?
7. In 1947 it was announced that one of the long-lived radioactive products of
the fission of U 235 was a new chemical element different from any other elements
then known. The new element formed a volatile oxide whose vapor density was
9.69 times that of oxygen gas at the same temperature and pressure; the oxide
contained 36.2% of oxygen by weight. What is the molecular weight of the oxide?
What is the' combining weight of the new element? Where does the new element
fit into the periodic system? What element would it be expected to resemble in
chemical properties? What is the approximate atomic weight of the new element?
CHAPTER 2
ATOMS, ELECTRONS, AND RADIATIONS
2-1 Faraday's laws of electrolysis and the electron. It was seen in

the first chapter that the combining properties of the elements can be
interpreted in terms of an atomic theory of matter. From the work of
Faraday on the electrolysis of aqueous solutions of chemical cornpounds,
it appeared that electricity is also atomic in nature. Faraday discovered
two laws of fundamental importance.
1. The chemical action of a current of electricity is directly proportional
to the absolute quantity of electricity whlch passes through the solution.
2. The weights of the substances deposíted by the passage of the same
quantity of electricity are proportional to their chernical equivalents.
Careful measurements have been made of the amount of electricity needed
to liberate one equivalent weight of a chemical substance, say 107.88 gm of
silver, 1.008 grn of hydrogen, or 35.46 gro of chlorine. This amount is
96,490 coul or one faraday, and is denoted by F.
Faraday inferred frorn his experimental results that one and the same
amount of electricity is associated in the pro~ess of electrolysis with one
atom of each of these substances. He thought of this charge as carried
by the atom, or in sorne cases by a group of atoros, and he called the atom,
or group of atoms, with its charge, an ion. The current in an electrolytic
process results from the movement of the ions through the solution, nega-
tive ions rnoving toward the anode and positive ions toward the cathode.
Faraday's results irnplied that there is an elementary unit, or atom, of
electricity. This idea, however, did not seern to fit in with other elec-
trical phenornena such as rnetallic conduction, and both Faraday and
Maxwell hesitated to accept the idea of the atornic nature of electricity.
In 1874 Stoney ventured the hypothesis that there is a "natural unit of
electricity," narnely, that quantity of electricity which must pass through
a solution in order to liberate, at one of the electrodes, one atom of hydrogen
or one atoro of any univalent substance. He suggested the name electron
for this quantity, and tried to calculate the rnagnitude of the electronic
charge. Stoney's calculation was based on the fact that one faraday of
electric charge liberates the number of atoros in one gram-atom of a uni-
valent element; this number is just Avogadro's number N 0 , so that
F = N 0 e, where e represents the electronic charge. From the known
value of F and the rough value of N O available at the time, Stoney obtained:
e = 0.3 X 10- 10 electrostatic unit of charge (esu). Although this is only
-h of the presently accepted value, it served to indicate the order of mag-
nitude of the electronic charge.
19
20 ATOMS, ELECTRONS, AND RADIATIONS [CHAP. 2
Faraday's experiments alone yield no certain information about the

quantity of electricity represented by one electron, but they do give pre-
cise information about the ratio of the ionic charge to the mass of the atom
with which it is associated in a given solution. For hydrogen, this ratio is
~
9
= 96490 coul = 96490 X 3 X 10 esu
mH 1.008 gm 1.008 gm
= 2.87 X 10 14 esu/gm.
2-2 The conduction of electricity in gases. Under normal conditions

a gas is a poor conductor of electricity, but if ionization can be produced
in the gas, it becomes a conductor; x-rays, discovered by Roentgen in
1895, provided a convenient means of producing ionizatíon. FoHowing
this discovery, the mechanism of gaseous conduction and the nature of
gaseous ions were investigated on an increased scale by many workers.
Theoretical and experimental studies were made of the mobility of gaseous
ions and the coefficient of diffusion of these ions. As a result of these
studies, the value of the quantity of electric charge associated with one
cubic centimeter of a gas could be calculated. This quantity can be
written as neg, where n is the number of molecules of gas per cubic centi-
meter under sorne standard conditions of temperature and pressure, and
eg is the average charge on a gaseous ion. The value of neg was found to
be 1.2 X 10 10 esu for air at 15ºC and 760 mm of mercury. The volume
occupied by one gram-molecular weight of gas under these conditions is
4
2.24 X 10 (~~) cmª = 2.36 X 104 cmª.
The total amount of electric charge associated with one gram-molecular
weight of gas is then
1.2 X 10 10 esua X 2.36 X 10 4 cmª
cm
= 2.83 X 10 14 esu.
Now, the amount of electricity carried by the univalent ions in one gram-
molecular weight of a salt in solution is one faraday:
F = 96,490 coul = 2.90 X 10 14 esu,
which is very close to the value found for the ionized gas. If the mole-
cules of the gas are assumed to be singly ionized, they seem to carry, on
the average, the same charge as that carried by univalent ions in solution.
This result also points to the existence of a unit electrical charge.
2-3 Cathode rays. The experimental proof of the independent exist-

ence of particles of negative electricity followed from the study of the
conduction of electricity in rarefied gases. When a gas is enclosed in a
2-3} CATHODE RAYS 21
Fro. 2-1. Diagram of a tube for the production of cathode rays.
glass tube with two electrodes to which a sufficiently high potential is

applied, electricity flows through the gas, which becomes strongly illumi-
nated. When the pressure of the gas is reduced below 0.1 mm of mercury
the illumination gradually clisappears. At very low pressures, of the order
of 10- 4 or 10-5 mm of mercury, a fluorescent glow appears on the wall
of the glass tube opposite the cathode. The presence of the glow was
correctly attributed by early workers to rays from the cathode, and these
rays were called cathode rays. A schematic diagram of a tube for the pro-
duction of cathode rays is shown in Fig. 2-1; C is the cathode, A is the
anode, and T is a side tube through which the gas can be exhausted.
The cathode rays were found to have severa! interesting properties.
l. The rays travel in straight lines. When screens S 1 and S 2 , pierced
with holes, are put into the tube, the glow is confined to a spot D on the
end of the tube; or, if a body is placed in the path of the rays, it casts a
shadow on the fluorescent part of the glass. From the geometrical rela-
tions it can be inferred that the rays travel in straight lines.
2. The rays can penetrate small thicknesses of matter. When a "win-
dow" of thin aluminum or gold foil is built into the end of the tube struck
by the rays, the passage of the rays through the foil is shown by the pres-
ence of luminous blue streamers in the air on the far side of the foil.
3. The rays carry a negative electric charge. When the rays are caught
in an insulated chamber connected to an electroscope, the electroscope
becomes negatively charged.
4. The rays are deflected by an electrostatic field. When two parallel
plate electrodes with theír planes parallel to the path of the rays are intro-
duced into the cathode-ray tube, the rays are deflected toward the posi-
tively charged plate.
5. The rays are deflected by a magnetic field. When an ordinary bar
magnet is brought up to the cathode-ray tube, the fluorescent spot opposite
the cathode moves in one direction or another depending on which mag-
netic pole is presented to the rays.
6. The cathode rays carry considerable amounts of kinetic energy.
When a metal obstacle is placed in the path of the rays, it can quickly
become incandescent.
\
\
- - - -1- - - - - - - ---+- --- - - -
D' ----J
Fm. 2--2. The apparatus of J. J. Thomson for measuring the ratio e/m for
cathode rays.
The only reasonable interpretation of these experimental facts is that the

cathode rays consist of negatively charged particles.
Thomson, in his fundamental experiments (1897), confirmed the view
that the cathode rays consist of negatively charged particles and determined
the ratio charge/mass = e/m for the particles. For these experiments,
the tube of Fig. 2-1 was modified so that the rays could be made to pass
through an electrostatic field. In the apparatus shown schematically in
Fig. 2-2, a potential difference of a few thousand volts is maintained
between the anode A and the cathode C. A narrow beam of rays from
the cathode passes through a slit in the anode and through a second slit
in the metal plug B. The end of the tube is coated on the inside with a
fluorescent material and the point of impact of the cathode-ray stream
appears as a bright spot. Inside the tube are two metal plates D and D'
between which a vertical electric field can be set up. By means of an
externa! electromagnet, a magnetic field can be established, perpendicular
to the plane of the diagram, within the region indicated by the dotted
circle. If the electric field is upward (D' positively charged), and if there
is no magnetic field, the cathode-ray stream is deflected downward. The
electric and magnetic deflections can be made to cancel each other by the
proper adjustment of the electric and magnetic fields.
Suppose that the cathode rays consist of negatively charged particles
of charge e and mass m, and let E be the magnitude of the electrostatic
field, and H that of the magnetic field. If the electric and magnetic fields
are so adjusted that their effects cancel, and the particle stream passes
through the tube without being deflected, then the electric force on the
particles must be equal in magnitude and opposite in direction to the
magnetic force. The electric force is given by eE, and the magnetic force
by Hev, where vis the velocity of the particles.< 0 • Then,
E
eE = Hev, or V= H' (2-1)
• Superscript numbers in parentheses refer to the references at the end of the

chapter.
2-3] CATHODE RAYS 23
[)
+ D'
Frn. 2-3. The deflection of cathode-ray particles in an electrostatic field.
and this experiment can be used to measure the velocity of the cathode-ray
particles. If E, e, and H are expressed in electromagnetic units, the velocity
has the units centimeters per second.
N ext, the magnetic field is removed and the deflection of the particles
caused by the electrostatic field is measured as shown in Fig. 2-3. In this
diagr11m, the distance P 1P 2 is the deflection caused by the electrostatic
field as observed on the fluorescent screen; D and D' are the parallel plates
as before, and l is the length of the plates. As the particles traverse the
field, they are subjected to a vertical acceleration in a direction parallel
to that of the field during the time interval l/v, so that the velocity pro-
duced in the vertical direction is (eE/m) (l/v). The horizontal velocity of
the particles is not changed by the field. On leaving the field, the velocity
of the particles has the components v in the horizontal direction and
(eE/m) (l/v) in the vertical direction. Hence, if P 2 is the deflected posi-
tion of the ray, P 1P 2 /MP 1 is the ratio of the vertical and horizontal
velocities:
P1P2 eE l
MP1 = m VZ'
or
(2-2)
In Eq. (2-2), the quantities l, MP 1 and E are known from the experimental
setup; the deflection P 1P 2 is measured, and v is known from the experi-
ment in which the effects of the electric and magnetic fields cancelled
each other.
Thomson's experiments showed that the cathode-ray particles move
with very high speeds-of the order of a tenth of the velocity of light.
For e/m, he found the value 1.7 X 10 7 emu/gm or, since 1 emu of charge
= 3 X 10 10 esu, e/m = 5.1 X 10 17 esu/gm. This value was independent
of the cathode materials which he used (aluminum, iron, and platinum),
and of the gas (air, hydrogen, and carbon dioxide) in the discharge tube.
24 ATOMS, ELECTRONS 1 AND RADIATIONS [CHAP. 2
The value of e/m for the cathode-ray particles is about 1800 times as
large as that found for hydrogen ions in electrolysis. If the cathode-ray
particles were assumed to have the same mass as the smallest atom,
hydrogen, they would have to carry a charge 1800 times as Iarge as that
on a hydrogen ion. Furthermore, since the value of e/m is independent of
the materials in the cathode tube, and materials with different ionic mas.ses
were used, it would be necessary to imagine sorne mechanism which would
make the charge on the particles directly proportional to the mass. If, as
an altemative, the charge on the particle is assumed to be the same as that
on the hydrogen ion, then the mass of the particle is about 1/1800 of that
of the hydrogen ion. The latter assumption seemed more reasonable to
Thomson, who therefore considered that the cathode-ray particles are a
new kind of negative corpuscle. Later work proved that Thomson's con-
clusion was correct.
2-4 The electron: the determination of its charge. Thomson's experi-

ments on cathode rays demonstrated the independent existence of very
small negatively charged particles, but the assumption that the charge on
one of these particles is the same as the charge on a univalent ion in solu-
tion and on a gaseous ion in a conducting gas presented a basic problem.
Although the assumption seemed logical, a proof of the equality of the
charges was needed. Tbe direct way to solve this problem would be to
measure the charge on the particles. This direct measurement could not
be made because of the great speed of the cathode-ray particles, but other
sources of similar negatively charged particles were available. Toward
the end of the 19th century it was found that when ultraviolet light falls
on certain metals, negatively charged particles with sinall velocities are
emitted (the photoelectric effect). Thomson determined the value of e/m
for these particles and found it to be practically the same as that for the
cathode-ray particles. The same result was obtained for the negatively
charged particles emitted by an incandescent filament (the thermionic
effect). In view of the constancy of e/m for the negatively charged parti-
cles produced in different ways, it was concluded that ali of these particles
are identical. There still remained, however, the problem of proving
experimentally that the charge on the different particles (monovalent
electrolytic and gaseous ions, photoelectric particles, thermionic particles)
is the same.
Severa! measurements of the charge were inade around 1900, and their
results may be summarized briefly as follows. For the average charge on
an electrolytic ion, the value of 2 X 10- 10 esu was obtained, based on an
improved but still approximate value of the Avogadro number; later work
(1905-1908) gave a value of 4.25 X 10- 10 esu. The average charge on
gaseous ions was found to lie between 3 X 10- 10 esu and 5 X 10- 10 esu,
2-4] THE ELECTRON: THE DETERMINATION OF ITS CHARGE 25
and similar values were found for the photoelectric particles. Experi-
mental evidence of this kind pointed to the identity of the various negative
electrical charges, and this view was gradually adopted. The term electron
was applied to the negative particles themselves, rather than to the mag-
nitude of the charge, and this usage has become general. The work on
cathode rays and on the photoelectric and thermionic effects also indicated
that electrons are a constituent part of all atoms, and the electron was
regarded as a fundamental physical particle. In all of the experimental
work discussed so far, either values of e/mor of the average charge of the
various particles were measured. Although it was diflicult to argue
against the conclusion that electrons exist and have a certain elementary
unit of charge, the correctness of this conclusion had not been proved
experimentally beyond any doubt. The final proof of the atomic or
particulate nature of electricity and the first really precise determination
of the value of the electronic charge was supplied by the work of Millikan
(1909--1917).
Millikan's experiments will be discussed in sorne detail because of their
fundamental place in modern physics. The earlier experiments on the
determination of the electronic charge were based on the fact that elec-
trically charged atoros (ions) act as nuclei for the condensation of water
vapor. Water vapor was allowed to condense on ions, forming a cloud,
and the total charge carried by the cloud was measured with an electrom-
eter. The number of water drops in the cloud w~s calculated by weigh-
ing the water condensed from the cloud and dividing by the average
weight of a single drop. This average weight could be found by measuring
the rate of fall of the drops through the air, because the weight of a drop
and its rate of fall were known to be related through Stokes' law of fall. <2 >
The total charge on the cloud divided by the number of drops gave the
average value of the charge per drop. This method involves several assump-
tions: (1) that the number of ions is the same as the number of drops,
(2) that Stokes' law, which had never been tested experimentally, is valid,
and (3) that the drops are all alike and fall ata uniform rate not affected
by evaporation. These assumptions were all questionable, and their
validity had to be tested in careful experiments.
The above method, used by Townsend (1897) and Thomson (1896),
was improved by Wilson (1903), who formed water clouds between two
parallel plates which could be connected to the terminals of a battery.
The rate of fall of the top surface of the cloud was first determined under
the influence of gravity. The plates were then charged so that the droplets
were forced downward by both gravity and the electric field, and the new
rate of fall was measured. When the two rates were compared, a value
of the electronic charge could be obtained. In Wilson's method, it was
not necessary to assume that the number of drops is equal to the number
of ions. Only the rate of fall of the top of the cloud was observed and,
since the more heavily charged drops were driven down more rapidly by
the field than the less heavily charged ones, the actual experiments were
made on thc least heavily charged drops. Wilson obtained the value
3.1 X 10- 10 esu, but the other assumptions remained, and there was con-
siderable uncertainty in this result.
Millikan improved Wilson's method by usirig oil drops, thus avoiding
errors caused by evaporation. He also noticed that a single drop could
be held stationary between the charged plates by adjusting the volts.ge on
them so that the weight of the drop was just balanced by the electrostatic
field between the plates. From the relationship between the forces, the
value of the charge of the electron could be obtained. In addition, while
working with balanced drops, Millikan noticed that occasionally a drop
would start moving up or down in the electric field. Such a drop had
evidently picked up an ion, either positive or negative, and it was possible
to determine the change in the charge carried by a drop without regard to
the original charge on the drop.
The essential parts of Millikan's apparatus are shown in Fig. 2-4;
P 1 and P 2 are two accurately parallel horizontal plates, 22 cm in diameter
and 1.5 cm apart, housed in a large metal box to avoid air currents. Oil
is sprayed in fine droplets from an atomizer above the upper plate and a
few of the droplets fall through a small hole in the plate. A beam of light
is directed between the plates, and a telescope is set up with its axis trans-
verse to the light beam. The oil drops, illuminated by the light beam,
look like tiny bright stars when viewed through the telescope, and move
slowly under the combined influence of their weight, the buoyant force of
the air, and the viscous force opposing their motion. The oil droplets in
the spray from the atomizer are electrically charged, probably because of
frictional effects; the charge is usually negative. Additional ions, both
positive and negative, can be produced in the space between the plates by
ionizing the air by means of x-rays. If the upper plate is positively charged
to a potential of about 1000 volts, and the lower plate negatively charged,
there is a uniform electric field between the plates, and there is an electric
force on the droplet which affects its motion.
When a spherical body falls freely in a viscous medium, it accelerates
until a terminal velocity is reached such that the net downward force on
the body equals the viscous force. The la tter is proportional to the terminal
velocity of the particle, and
mg = kv11 , (2-3)
where mis the apparent mass of the body, g the acceleration dueto gravity,
k a proportionality constant, and v11 the terminal velocity of free fall. If,
Light- -X-rays
Frn. 2-4. Millikan's oil drop experiment.
in addition, the body carries a charge q and is subjected to the action of

an electric field of intensity E, a new terminal velocity ve is reached,
given by
qE - mg = kvE, (2-4)
when the field acts counter to gravity. The ratio of the velocities is
Vg = mg (2-5)
VE qE - mg
The charge q is then
q=-
mg (Vg + VE) . (2-6)
E Vg
It is evident that q can be evaluated if m and E are known and if the ratio
(vg + ve)/vg is observed experimentally. The main difficulty lies in the
determination of m, the mass of the drop, which is too small to be meas-
ured directly.
28 ATOMS, ELECTRONS, AND RADIATIONS [CHAP, 2
It is now possible to settle a fundamental question, namely, that of

the atomic nature of electricity. If there is a real unit charge e, then
the charge q on an oil drop must be sorne integral multiple of e. Suppose
that a measurement is made, such as that leading to Eq. (2-6), and that
the charge on the drop then changes from q to t/ because another ion is
caught. Then q' is given by
t/ = ffl{1 (v,, + vá). (2_7)
E v11
Experimentally, ions supplied by the ionization of the air by x-rays can

be caught by an oíl drop while under observation, and the behavior of a
single drop carrying different charges can therefore be observed in the
same electric field E. The change in charge is
flq =q_ q = ffl{1 (v'B - VE). (2-8)

E Vg
Millikan found, in many such experiments, that liq was never smaller
than a certain Ininimum value, and that all values of liq were exact integral
multiples of this value. By taking the least common divisor of all the
values of liq for a large number of changes of charge on the same drop,
Millikan obtained the smalleat value of the change in charge, and he identified
this value with Stoney's electron. Millikan's results provide direct proof
that the elementary electric charge, the electron, is notan average charge,
but rather that all electric charges on ions have either exactly the same
value or else small, exact integral multiples of this value. If this value is
denoted by e, then q = ne, where n is a small integer. Millikan also
showed, by studying drops carrying positive ions, that the elementary
unit of positive charge has the same magnitude as that of the negative
electronic charge, and that all static charges both on conductors and
insulators are built up of units of the electronic charge.
The precise evaluation of the electronic charge depends, as has been
noted, on the mass of the oil droplet. The mass can be obtained when
the value of the proportionality factor k in Eq. (2-3) is known, and k is
given by Stokes' law of fall. From the equations of hydrodynamics,
Stokes deri"ed a theoretical expression for the rate of fa.U of a spherical
body in a viscous medium,
(2-9)
or
2
v, = -2 -ga (p - p'), (2-9a)
9 "
where a is the radius oí the body, p the density oí the body, p' the density
of the mediuro, 71 the viscosity of the medium, and v11 the terminal or
equilibrium rate of fall of the body. Equation (2-9) represents the balance
between the gravitational and viscous forces on the body_ and can be
applied to the fall of an oil drop in air. The value of v 11 is measured ex-
perimentally, and a is given by
a- [ 9 'IV
11
]1/2 (2-10)
- 2 g(p - p') '
flOthat
m = j1ra 3 (p - p') = -411" ( -9 -7JV 11)ª'2

3 2 g
G1
--
-p'
)1,2 • (2-11)
When the expression for mis inserted into Eq. (2-6) for the charge on an
oil drop, the result is
_ _ 411" (971) 3 ' 2 r 1 2

]l/ (v11 + VE) 1/2 (2-12)
q - e,. - 3 2 l_g(p - p') E v" ·
In the last equation, e,. has been written for q because the charge on the
oil drop has been provento be an integral multiple of the electronic charge.
lf many observations of the quantity (v 11 + VE) are made for changes of
charge on the drop, the greatest common divisor, denoted by (v 11 + vE)o,
should correspond to the value of the electronic charge, denoted tem-
porarily by ei, so that
_ 411" (971) 3 ' 2 r 1 ]l/

2
(V11 + VE)o 1/2
(2-13)
ei - 3 2 l_g(p - p') E v11 •
Millikan found that the value of e1 was constant for large drops, but
increased as the drops grew smaller. Careful studies then showed that
Stokes' law in the form Eq. (2-9a) does not hold for the free fall of very
small drops. It is necessary in this case to replace Eq. (2-9a) by
+ pab) ,
2
Vg = 92 -;,
ga
(p - p') (
1
where p is the pressure of the air, a is again the radius of the drop, and
b is an experimentally determined constant. The electronic charge is
then given by the expression:
e1
e = (1 + (b/pa)]ª'2,
or
_ 411" (971)
e- 3 2
3 2
' rl_g(p -1 p') ]l/
2
E[l
+ VE)o 1/2
(v11
+ (b/pa)]3/ Vi, 2
(2-14)
Millikan obtained the value e = 4.774 X 10- 10 esu, which was accepted
for sorne years. Later determinations of e by means of indirect but highly
precise methods gave results which disagreed with Millikan's value. The
discrepancy was caused by an error in the value of the viscosity of air
used by Millikan. The use of more recent data for this parameter leads to
the value e = 4.8036 ± 0.0048 X 10- 10 esu, which agrees well with
the values of e obtained by other methods.
The constants e, F, and N 0 , together with other quantities which will
be met later, are called the fundamental, constants of physics because of
their great importance. The measurement of these constants is one of the
basic problems of physics, and many highly precise methods have been
developed for measuring them. The constants are often related, for ex-
ample, F = N 0 e ¡ and statistical methods have been developed for the
analysis of the experimental results. This analysis yields best values of the
constants which are consistent with the entire body of present knowl-
edge ¡ <3 > a partial list of these values is given in Appendix II. The follow-
ing best values have been obtained for constants which have been discussed
so far:
e = (4.80286 ± 0.00009) X 10- 10 esu,
e/m = (I.75890 ± 0.00002) X 10 7 emu/gm
= (5.27305 ± 0.00007) X 10 1 7 esu/gm,
m = (9.1083 ± 0.0003) X 10- 28 gm,
F = (2.89366 ± 0.00003) X 10 14 esu/gm~quivalent weight
= (96521.9 ± 1.1) coul/gm~quivalent weight.
The Avogadro number is
NO = (6.02486 ± 0.00016) X 1023 /gm-mole.

The mass of an individual hydrogen ion, which is usually referred to as a
proton, is
M H = (1.67239 ± 0.00004) X 10- 2 • gm,
and the ratio of the mass of the proton to that of the electron is
MH!m = 1836.12 ± 0.02.
The values quoted for e/m and m are correct provided that the speeds
of the particles are very small compared with the speed of light. lt will
be seen in Chapter 6 that the mass of the electron increases with its speed;
the value given above for the mass of the electron is called the rest mass
and is often denoted by m 0 •
2-6] X-RAYS 31
2-5 Positive rays. The fact that rays consisting of negatively charged
particles are produced in a gas discharge tube suggested that rays of
positively charged particles are also fonned. Rays of this kind were
discovered by Goldstein {1886), who observed that when the cathode of
a discharge tube was pierced by narrow boles, streamers of light appeared
behind it. He assumed that the luminosity was caused by rays which
traveled in the opposite direction to the cathode rays and passed through
the holes in the cathode. These rays, which were called canal rays, were
deflected by electric and magnetic fields, and from the directions of the
deflections it was concluded that the rays consist of positively charged
particles. This result gave rise to the more generally used term positive
rays.
As in the case of the cathode rays, it was possible to measure the velocity
and charge-to-mass ratio of the positive-ray particles by means of the
deflections in electric and magnetic fields. The results were quite dif-
ferent from those for the cathode-ray particles. Both the velocities and
the e/m values were found to be much smaller than those for electrons.
The value of e/m was found to depend on the atomic weight of the gas
in the discharge tube, and to decrease with increase in atomic weight. The
largest value of e/m was obtained for hydrogen and was very close to the
value obtained in electrolysis. These results suggested that the positive
rays are streams of positive ions produced by the ionization of atoms or
molecules in the strong electric field of the discharge tube. The positive
ion is formed by the removal of one or more electrons from the neutral
atom, and the mass of the ion is practically the same as that of the atom.
Hence, any measurement of the mass of positive-ray particles should yield
direct information about the masses of atoms of elements or molecules of
compounds. Furthennore, this information should refer to the atoms or
molecules individually rather than give average values for a large number
of particles. It is for this reason that the accurate measurement of the
masses of positive-ray particles is of great importance in modern physics.
In fact, measurements of this kind resulted in the proof of the existence
of isotopes and in the determination of isotopic weights.
The subject of positive rays will be treated in greater detail in Chapter 9
in connection with the determination of isotopic masses and abundances.
2-6 X-rays. In 1895, Roentgen discovered x-rays, which are produced

when a beam of cathode rays strikes a solid target. He found that the
operation of a cathode-ray tube produced fluorescence in a screen covered
with barium platinocyanide and placed at sorne distance from the tube.
The effect was traced to radiation coming from the walls of the cathode-
ray tube. In studying the properties of this new type of radiation,
Roentgen noticed that if materials opaque to light were placed between
the tube and the screen the intensity of the fluorescence decreased but
did not disappear, showing that x-rays can penetrate substances which
are opaque to ordinary light. It was also found that the x-radiation can
blacken a photographic plate and produce ionization in any gas through
which it passes; the last property is used to measure the intensity of the
radiation. The x-rays were found to travel in straight lines from the
source and could not be deflected by electric or magnetic fields, from
which it was concluded that they are not charged particles. Later work
showed that x-rays can be reflected, refracted, and diffracted, and there
is convincing evidence that these rays are an electromagnetic radiation
like light but of much shorter wavelength.
X-rays have turned out to be a valuable tool in atomic research and a
great deal of information about atomic structure has been obtained from
studies of the scattering and absorption of x-rays by atoms, as will be
shown in Chapter 4.
2-7 Radioactivity: alpha-, beta-, and gamma-rays. In 1896, Becquerel

discovered that crystals of a uranium salt emitted rays which were similar
to x-rays in that they were highly penetrating, could affect a photographic
plate, and induced electrical conductivity in gases. Becquerel's discovery
was followed by the identification by the Curies (1898) of two other radio-
active elements, polonium and radium. The activity of radium as meas-
ured by the intensity of the emitted rays was found to be more than a
million times that of uranium.
1
Plate
F10. 2-5. The three radiations from radioactive materials and their paths
in a magnetic field perpendicular to the plane of the diagram (schematic).
2-71 RADIOACTIVITY: ALPHA·, BETA-, ANO GAMMA-RA YS 33
It can be shown, by means of experiments in magnetic fields, that there

are three kinds of radiations from naturally occurring radíoactive sub-
stances. In these experiments a collimated beam of rays is supplíed by a
small píece of radioactive material at the bottom of a long groove in a lead
block, as shown in Fíg. 2-5. A photographíc plate is placed sorne dístance
above the lead block, and the air is pumped out of the chamber. A weak
magnetic field is applíed at right angles to the plane of the díagram, and
dírected away from the reader. When the plate is developed, two dístínct
spots are found, one in the dírect line of the groove in the lead block, and
one deflected to the right. The undeviated spot is caused by neutral rays
called -Y-rays; the spot to the right is caused by rays of negatively charged
particles, called {J-rays. If the weak magnetíc field is replaced by a strong
one in the same direction, the undeviated spot is agaín found together with
a spot deflected to the left; the latter is caused by posítively charged
particles called a-rays. In the strong field, the path of the fJ-rays is bent
so much that it doesn't reach the plate. The need for a strong field
indicated that the mass of the positívely charged particles is much
greater than that of the nega.tively charged particles. The different
electrical properties of the radía.tions are shown schematically in Fig. 2-5.
It has been shown that all three radíatíons are not emitted simultane-
ously by all radioactive substances. Sorne elements emit a-rays, others
emit {J-rays, while -Y-rays sometimes accompany one and sometimes the
other. The actívity of a given material is not affected by any simple
physical or chemical process, such as change in temperature or chemical
combination with nonradioactive substances. These results, together with
Mme. Curie's demonstration that the actívity of any uranium salt is
directly proportional to the quantíty of uranium in the salt, showed that
radíoactivity is an atomic phenomenon.
The {J-rays, being negatively charged particles, could be deflected in
electric and magnetic fields, and theír velocíties and e/m values were
measured. The results, although somewhat more complícated than in
the case of cathode rays, showed that the {J-rays consíst of swíftly moving
electrons and díffer from cathode rays only in velocíty. The fJ-rays from
radium were found to have velocities up to within a few percent of that
of light, while cathode rays usually have velocitíes considerably less than
half that of light. The e/m values of the {J-particles were of the same
order of magnitude as those of the cathode-ray particles, but decreased
with increasing velocity of the particles. Even before the Millikan oil-
drop experiment, it was considered that electric charge is atomic in nature
and should not vary with velocity, and it was concluded that the mass
of a fJ-particle increases with the speed with which the particle is moving.
The value of e/m for the slower fJ-partícles was found to be almost ídentical
with that obtaíned for cathode-ray partícles. The fJ-rays are much more
34 ATOMS, ELECTRONS, AND RADIATIONS [CHAP, 2
penetrating than a-rays, being able to pass through one millimeter of

aluminum, whereas the a-rays are completely stopped by 0.006 cm of
aluminum or by a sheet of ordinary writing paper. The ionization density
produced in a gas by fJ-rays is much less intense, however, than that pro-
duced by a-rays.
Alpha-rays, because of their positive electrical charge, can be deflected
by electric and magnetic fields, and their velocities and e/m values deter-
mined in the usual way. The velocities range from about 1.4 X 10 9 to
2.2 X 10 9 cm/sec, or less than tri of the velocity of light. The value of
e/m .is close to 4820 emu/gm = 1.45 X 10 14 esu/gm, or just one-half of
the value, 2.87 X 10 14 esu/gm, for the singly charged hydrogen ion. It
is evident that if the a-particle had the same charge as the univalent
hydrogen ion, its mass would be twice that of a hydrogen ion or atom.
Ií the a-particle carried a charge twice that of the hydrogen ion, its mass
would be four times as great and would correspond closely to that of the
helium atoro, which has an atomic weight of 4 units. The exact nature
of the a-particle could not be determined until either its charge or mass
could be measured separately. Rutherford showed, in an ingenious experi-
ment, that when a radioactive substance emits a-rays, helium is produced.
A small quantity of radon, a gaseous radioactive element, was sealed in
a small glass tube with extremely thin walls, so thin that the a-rays emitted
could pass through them into an outer glass tube, while the radon was kept
in. The outer tube was provided with electrodes, and the spectrum pro-
duced on passing a discharge was observed at regular intervals. After
a few days it was found that the spectral Jines of helium appeared; control
experiments showed that ordinary helium gas could not penetrate the
thin-waJled tube, proving that the helium was produced from the a-rays.
Thus, the a-particles must be helium ions and, since their mass is 4 times
that of a hydrogen ion, the charge on an a-particle must be twice that on
the univalent hydrogen ion. The a-particle is, therefore, a doubly ionized
helium atom.
The charge on an a-particle was also determined separately by counting
the number of a-particles emitted in a known time interval from a known
weight of radioactive material, and at the same time measuring the total
positive charge emitted. It will be shown in the next section that methods
have been devised for counting individual a-particles. Two independent
experiments, in which different counting methods were used, gave
9.3 X 10- 10 esu and 9.58 X 10- 10 esu, respectívely, for the a-particle
charge, and these values are very close to twice the electronic charge.
The 'Y-rays, which are not deflected by electric or magnetic fields, have
been shown to consist of electromagnetic waves, and are therefore similar
in nature to light and x-rays. The 'Y-rays are roughly 10 to 100 times
as penetrating as the fJ-rays, but produce less ionization than do the fJ-rays.
2-8} THE DETECTION AND MEASUREMENT OF RADIATION 35
The ionization produced by a-, {3-, and 'Y-rays is roughly in the order
10,000, 100, and 1, a useful rule of thumb.
The detailed chemical study of the radioactive substances has shown
that each such substance is an element, and that its radioactivity is
caused by a spontaneous disintegration of radioactive atoms into other
atoros. This process occurs according to definite laws. The transformation
of one atoro into another atom with the emission of either an o:-particle
or a (3-particle, led to the idea that the atoms of the chemical elements
are cornplex structures built of smaller particles. The study of atomic
structure became an important branch of physics with the main object of
discovering the laws according to which atoros are built up. The dis-
covery of natural radioactivity, together with the discovery of x-rays and
the proof of the independent existence of the electron, provided new and
powerful methods for attacking the problem of atomic structure.
2-8 The detection and measurement of radiation. The study of radio-

activity and the successful use of radiations as research tools or for other
purposes depend on the quantitative detection and measurement of radia-
tion. The quantities rnost often needed are the number of particles (e.g.,
electrons, protons, o:-particles) arriving at a detector per unit time and
their energies. When a charged particle passes through matter, it causes
excitation and ionization of the rnolecules of the material. This ionization
is the basis of nearly all of the instrurnents used for the detection of such
particles and the measurement of their energies. Similar instruments can
be used for uncharged radiations (x-rays ot 'Y-rays) because these impart
energy to charged particles, which then cause ionization. The different
types of instrument differ in the material within which the ionization is
produced and in the way in which it is observed or measured. Many instru-
ments are based on the production of ion.ízation in a gas. It is necessary, in
this case, to separate and collect the positive and negative ions because, if
these remained close together, they would recombine and no electrical in-
forrnation as to their presence could be obtained. The separation and collec-
tion of the ions requires the presence of an electrostatic field, and different
instrurnents result depending on whether the field is small, large, or inter-
mediate in magnitude. The ionization may also be produced in a liquid
or in a crystal. When it is produced in a gas supersaturated with vapor
or in a photographic emulsion, the tracks of the particles can be made
visible. When particles strike certain liquid or solid materials called
phosphors, which have the property of luminescence, part of the energy
used up in molecular excitation and ionization is re-emitted as visible or
ultraviolet light. Sometimes this light can be observed by eye, or it may
be detected by means of more sensitive devices.
We shall consider sorne of the commonly used detection methods be-

cause of their great importa.nce in nuclear physics, although a deta.iled
treatment is beyond the scope of this book. Such treatments are given in
sorne of the references at the end of the chapter.
The scintiUation method. It was found, about 1900, that radiations,
especially a-partícles, can produce luminescence in zinc sulphide, barium
platino-cyanide, and diamond. This luminescence, when produced by
a-particles, is not uniform, but consista of a large number of individual
flashes, which can be seen under a magnifying glass. Careful experiments
have shown that each a-particle produces one scintillation, so that the
number of a-particles which fall on a detecting screen per unit time is
given directly by the number of scintillations counted per unit time. A
screen can be prepared by dusting small crysta.ls of zinc sulfide, con-
taining a very small amount of copper impurity, on a slip of glass. The
counting can be done with a microscope with a magnification of about
30 times, and good precision can be obta.ined, but with difficulty. This
method was especially useful for counting a-particles in the presence of
other radiations, because the zinc sulfide acreen is comparatively insensi-
tive to {j- and 'Y-rays.
The visual scintillation counter has the disadvantages that the flashes
are quite weak and can only be seen in a darkened room, and that visual
counting is limited physiologically to about 60 scintillations/min and is
very tedious. Nevertheless, the method was used in many of the basic
experimenta of nuclear physics between 1908 and about 1935 when it was
replaced by electrical methods.
Since 1947, the use of scintillation counters has been revived and greatly
extended because of the discovery of new phosphors which are also sensi-
tive to {3- and 'Y-rays, and the development of highly efficient photo-
multiplier tubes. The phosphors include: inorganic salts, primarily the
alkali halides, containing small amounts of impurities as activators for
luminescence, for example, sodium or pota.ssium iodide activated with
thallium; crystalline organic materials such as naphthalene, anthracene
and stilbene; and solutions of organic compounds such as terphenyl dis-
solved in xylene. The photomultiplier tube, which replaces the microscope
and observer, converts the scintillations from the phosphor into amplified
electrical pulses which can be counted or otherwise analysed with suitable
electronic equipment. The modern scintillation counter can detect and
record many millions of flashes per second and can be used with intense
radiations.
A schematic diagram of a photomultiplier tube is shown in Fig. 2-6.
Light from the phosphor strikes the cathode, which is usually made of
antimony and cesium, and ejects electrons {photoelectric effect). The tube
has several electrodes called dynodes to which progresssively higher
2-8] THE DETECTION AND MEASUREMENT OF RADJATION 37
+200v +400v +600v +800v +lOOOv

Paths of electrons Anode
Cathode----..
/ '' /
/
/
/
'' /
/
/
+ooov
Dynode
Fm. 2-6. Photomultiplier tube (schematic).
potentials are applied. The photoelectrons are accelerated in the electro-

static field between the cathode and the first dynode, which is at a positive
potential relative to that of the cathode, and strike the dynode. The
accelerated electrons impart enough energy to electrons in the dynode to
eject sorne of them. There may be as many as ten secondary electrons
for each electron which strikes the dynode. This process is repeated, and
the electron current is amplified as the electrons are accelerated from
dynode to dynode. The output current, or pulse, at the anode may be more
than a million times as great as the current originally emitted from the
cathode. Each particle incident on the phosphor produces a pulse, and
the pulses are fed to an electronic system where they are counted. Elec-
tronic systems have also been developed which measure the energy of the
incident particles; the resulting instrument is called a scintillation
spectrometer.
A schematic diagram of a scintillation detector is shown in Fig. 2-7.
The electroscope. It has been mentioned that many of the instruments
used for the measurement of radiation depend for theír operation on the
Light reflector
Light pipe
Photomultiplier Electronic
Radiation
tube counting system
Phosphor
High-voltage
power supply
F10. 2-7. Schematic diagram of a scintillation detector.

,,--Insulator
Case
'"'
'' '''
1'' '1
A B B' jA_'_---e-Radioactive
source
F10. 2-8. Diagram of an a-ray electroscope.
ionization of a gas. A given sample of radioactive substance emits a

definite number of a-, {J-, or "Y-rays per unit time. The number of ions
which these rays can produce in air or in another gas at a fixed tempera-
ture and pressure is directly proportional to the number of particles. Hence
the extent of the ionization is a direct measure of the amount of radio-
active substance. One of the first and simplest instruments used in work
on radioactivity is the electroscope, an example of which is shown in
Fig. 2-8. This instrument is a condenser, one of whose elements is the out-
side case; the other element ends in a pair of gold leaves. When the two
elements are connected across a battery, they become oppositely charged
and the gold leaves (A and A') separate because of the electrostatic
repulsion of their like charges. If the battery is removed and the gas in
the condenser is ionized by a radioactive substance, gaseous ions migra.te
to the oppositely charged elements of the condenser. The net charge on
the elements and the potential difference between them decrease, and the
leaves collapse to the position B, B'. The rate of collapse of the gold
leaves is a measure of the rate at which the gas in the chamber is ionized
and hence of the intensity of the ionizing radiation,
If the radioactive substance emits a-, {J-, and "Y-rays, they all contribute
to the ionization but, since the a-rays are by far the strongest ionizing
agents, the instrument is really an a-ray electroscope. A {J-ray electro-
scope is made by placing an aluminum sheet about 0.01 cm thick over the
active material to absorb ali of the a-rays, and the ionization observed is
then caused by the fJ- and "Y-rays. Since the "Y-rays are much weaker
ionizing agents than the {J-rays, the collapse of the leaves is caused almost
entirely by the {J-rays. To measure "Y-ray activity alone, the radioactive
sample must be surrounded by enough material (2 or 3 mm of lead) to
absorb completely the a- and {J-rays. Sometimes the entire electroscope
is built of, or surrounded by, lead and the sample is placed outside the
instrument. Electroscopes have been developed which are much more
2-8] THE DETECTION AND MEASUREMENT OF RADIATIO:t,¡ 39
Insulator Central electrode
V
-...
-
- -r
e R
Fm. 2-9. Scheme.tic die.gre.m for pulse opere.tion oí a ge.s-filled che.mber.
sensitive than the one just described. For example, the Lauritsen elec-
troscope has as its moving element a very fine metal-coated quartz fiber.
The movement of the fiber is observed through a microscope with a scale
in the eyepiece.
Electrical instruments: ionization chamber, proportional counter, and
Geiger-Mueller counter. Each of these three detectora is based on the pro-
duction of ionization in a gas and the separation and collection of the
ions by means of an electrostatic field. The differences in the three
systems can be explained with the aid of Fig. 2-9, which shows a cylindri-
cal conducting chamber containing a central conducting electrode located
on the axis of the chamber and insulated from it. The chamber is filled
with a gas at a pressure of one atmosphere or less. A voltage V is applied
between the wall and the central electrode through the resistance R
shunted by the capacitor C; the central electrode is at a positive poten-
tial relative to that of the chamber wall.
We suppose that sorne ionization occurs in the gas beca.use of the passage
of a charged particle. Each ion pair consista of a positive ion and an
electron and we want to know, for a given initial ionization, how many
ion pairs are collected, or how many electrons rea.ch the central electrode
as the applied voltage is varied. To make the discussion less abstract, we
assume first that 10 ion pairs are formed, corresponding to curve a of
Fig. 2-10 where curves of total ion collection are plotted as functions of
applied voltage. For convenience, the logarithm to the base 10 of the
number of ion pairs n has been used as the ordinate. The discussion is
based on that of Wilkinson (see general reference at end of chapter). If
there is no voltage across the electrodes, the ions will recombine, and no
charge will appear on the capacitor. As the voltage is increased, say to a
few volts, there is competition between the loss of ion pairs by recombina-
tion and the removal of ions by collection on the electrodes. Sorne elec-
trons, less than 10, will reach the central electrode. At a voltage V 1 of,
40 ATOMS, ELECfRONS, AND RADIATIONS (CHAP, 2
11
10
V-
/: 1
1
9 / 1
I 1
1
1
8 Proportional
7
oounter
Ionization (high primary
ehamber -ionization) J 1
1
1
Geiger-Mueller oounter 1'

1 l l .l
<1/
:
6 '
¡ 1 .,....-r~ '
---
1
=
1 5
1
1.
1
1 1) 1
1
YJl
1 l 1 1
l 1 1 1
1 1 1 1
4
1
1
1
1
1
1
1
1
1 7Y 1
1
1
1
1
1
3
:y
~,
1 1 1
1
1
1 Á 1
1
1 1
r- ----
1
2
1 ! 1 1 1
'
1 1 1 1
1
1
__,,.. 1
1
1
1
1
1
1
1
1 1-- Propcrtional oounter
1 1 (low primary ioniu.tioñf""4 1 1
t
1
t t t t
o
V1 V2 yt
3 Va v, '
V¡
-2
200 400 600 800 1000 1200 1400 1600
Voltage (V)
Fm. 2-10. Curves of total ion collection versus applied voltage to illustrate
ionbation, proportional, and Geiger•Mueller regi.ons of operation of electrical
radiation detection instrument.s.
2-8] THE DETECTION AND MEASUREMENT OF RADIATION 41
perhaps, 10 volts the loss of ions by recombination becomes negligible,

and 10 electrons reach the central electrode; log n reaches the value
unity. As V is increased, n stays constant at 10 until a voltage V 2 is
reached which may be sorne tens or hundreds of volts depending on the
conditions of the experiment. The region between V 1 and V 2 , in which
the number of ion pairs collected is independent of the applied voltage
and the curve is horizontal, is called the ionization chamber region.
When the voltage is increased above V 2 , n increases above 10 because
of a phenomenon called gas mvltiplication or gas amplification. The
electrons r_fleased in the primary ionization acquire enough energy to
produce additional ionization when they collide with gas molecules, and n
increases roughly exponentially with V. Each initial electron produces a
small "avalanche" of electrons with most of these secondary electrons
liberated close to the central electrode. If 100 ion pairs are formed ini-
tially, curve b results; it is parallel to curve a in the ionization chamber
region and is separated from it by one unit on the log n scale. The be-
havior of the two curves above V 2 is interesting. For sorne range of volt-
ages, up to V 3, each electron acts independently and gives its o·wn
avalanche, not being affected by the presence of the other electrons. The
initial 100 electrons always yield 10 times more final electrons than do the
10 initial electrons, and curves a and b continue parallel with a difference
between them of one unit on the log n scale. Between V 2 and V 3 , the
number of ion pairs collected is then proportional to the initial ionization.
This is the region of proportional counter operation.
Above V 3 , the gas multiplication effect continues to increase very
rapidly, and, as more electrons produce avalanches, the latter begin to
interact with one another; the positive-ion space charge of one avalanche
inhibits the development of the next avalanche. The discharge with 100
initial electrons is affected before the one with 10 initial electrons and
increases less rapidly than the latter; curves a and b approach each other
and eventually meet at V 4 . The region between V 3 and V 4 is the region
of limited proportionality. Above V 4 the charge collected becomes in-
dependent of the ionization initiating it, and the curves a and b' become
identical. The gas multiplication increases the total number of ions to a
value that is limited by the characteristics of the chamber and the ex-
terna! circuit. The region above V 4 is the region of Geiger-Mueller counter
operation. It ends at a voltage V 5 where the discharge tends to propagate
itself indefinitely; V 5 marks the end of the useful voltage scale, the region
above being that of the continuous disckarge.
If the initial ionization is large, say 10 5 ion pairs, curve e is obtained.
It is similar to curves a and b, and parallel to them bctween V 2 and V á;
the proportional counter region ends sooner, at V 3 instead of at V 3.
Thus, the extent of that region depends on the initial ionization.
As a. result oí the beha.vior oí the ions of the gas in the electrosta.tic

field, three detection instruments ha.ve been developed.
l. The i<mization chamber, which opemtes a.t voltages in the range V 1
to V 2 , is cha.racterized by complete collection, without gas amplification,
oí all the electrons initially liberated by the passage oí the particle. Sub-
ject to certa.in conditions, it will give a pulse proportional to the number
of these electrons. Figure 2-9 is a schematic diagram oí a chamber whicb
could be used as an ionization chamber between 10 volts and 200 volts,
approximately. The numbers in Fig. 2-9 would be appropriate for a counter
with an outer cylinder 1 cm in radius, a wire central electrode 0.01 cm
in radius, and filled with gas to a pressure of a few centimeters of Hg (e.g.,
argon at about 6 cm of Hg plus a little alcohol). The counter length would
be about 10 cm.
The electrodes of an ionization chamber may also be parallel plates.
A schematic diagram of an instrument of this type is shown in Fig. 2-11.
2. The 'J)roportional, counter, which operates in the volts.ge region V 2 to
V 3 , is characterized by a gas multiplication independent oí the number of
initial electrons. Hence, although gas multiplication is utilized, the pulse is
always proportional to the initial ionization. The use of this counter
permits both the counting and the energy determination of particles
which do not produce enough ions to yield a detectable pulse in the region
V 1 to V 2 • The proportional counter, therefore, offers advantages for
pulse-type measurements of beta radiation, an application for which
ionization chambers are not sensitive enough.
Housing High-voltage High-voltage

electrode insulator
Collecting
voltage
---
-
F10. 2-11. Schematic diagram of a para.lle! plate ionization chamber (Re-
printed by permission from W. J. Price, Nuclear Radiation Detection. MeGraw-
Hill, 1958.)
2-8) THE DETECTION AND MEASUREMENT OF RADIATION 43
Frn. 2-12. Diagram of a Geiger-Mueller counter tube.
3. The Geiger-Mueller counter, also known as Geiger, or G-M, counter,

which operates in the voltage region V 4 to V 6 , is characterized by the
spread of the discharge throughout the entire length of the counter, re-
sulting in a pulse size independent of the initial ionization. It is especially
useful for the counting of lightly ionizing particles such as ,8-particles or
'Y-rays. The G-M counter usually consists of a fine wire (e.g., tungsten)
mounted along the axis of a tube which contains a gas at a pressure of
about 2 to 10 cm Hg. The tube may be a metal such as copper, ora metal
cylinder may be supported inside a glass tube as in Fig. 2-12; a mixture
of 90% argon and 10% ethyl alcohol is suitable. A potential difference,
which may be between 800 and 2000 volts, is applied to make the tuüe
negative with respect to the wire.
Geiger-Mueller counters have been designed for the measurement of
a-, {1-, and 'Y-rays, as well as x-rays. In the form described, they are use"d
mainly for {J- and 'Y-rays, in part because of the difficulty of making tubes
with windows thin enough to be penetrated by a-rays. The principie of
the Geiger-Mueller counter was first used by Rutherford and Geiger in
1908 to count a-particles, with the object of determining their charge.
The instrument that they used, the Geiger point counter, is shown sche-
matically in Fig. 2-13, and instruments like it are still used to detect
a-particles. It has a smooth fine point P at the end of a fine wire supported
inside a tube T by an insulating plug. The tube contains dry air at atmos-
pheric pressure, and the particles enter through a thin foil window W.
To
amplifier
\
Frn. 2-13. Diagram of a Geiger point counter.

In other respects, the operation of this counter tube is similar to that of

the Geiger-Mueller counter. By comparing the activity of a radioactive
substance as measured with a Geiger point counter and with a zinc sulfide
screen, Rutherford and Geiger showed that each a-particle that strikes the
screen causes one scintillation, a fact which increases the usefulness of
the screen for quantitative measurements.
In the electrical instruments discussed, a particle causes a small amount
of ionization. The ions, either multiplied in number or not, are collected
and produce a voltage pulse which may be as small as 10 microvolts. An
electronic pulse amplifier accepts these small voltages and amplifies them
to a level usually in the range of 5 to 50 volts. The amplified voltage
pulses must then be counted in sorne way so that their rate can be measured.
For very slow rates (less than about one per second), the pulses can be
made to operate a relay-type recorder. For large rates, mechanical devices
either fail completely or miss many counts. The response of an ionization
chamber can be made sufficiently rapid so that a-particles arriving at
intervals of 10- 4 sec or less can be detected and the counting must then
be made very efficient. A scaling unit, or scaler, is an electric device which
selects precisely every mth pulse and passes it on to the recorder. The
scaling ratio is usually a power of 2 or 10 and very high counting rates
can be achieved with the aid of a ,scaler. The number of pulses indicated
on the counter per unit time multiplied by the appropriate factor of the
scaler gives the number of a-particles entering, per unit time, the space
between the electrodes in the ionization chamber. Other electronic devices
are also available for measuring the magnitude of the pulse, which gives
the energy of the incident particle. Thus, it is often desired not only to
record the occurrence of a pulse but, also, to sort pulses according to their
size (by means of an electronic discriminator circuit) or to sort them
according to the time intervals during which they arrive (by means of an
electronic timing circuit). The detector then forms part of a circuit with
appropriate electronic instrumentation. Scintillation detectors, of course,
also require electronic instrumentation.
The cloud chamber. One of the most important instruments for basic
research on radiations is the cloud chamber. This instrument, first used
in 1912, is based on the discovery by C. T. R. Wilson (1897) that ions
act as nuclei for the condensation of supersaturated water vapor. There
are two types of cloud chamber differing in the method used for obtaining
supersaturation. In the expansion chamber, <4 > the gas, saturated with
vapor, is made to expand by the quick motion of a piston. The expansion
is adiabatic and lowers the temperature. The cooling is more than suffi-
cient to overcome the effect of the increase in volume, and the air becomes
supersaturated with water vapor. If an ionizing ray enters the chamber,
the ions formed act as nuclei for the condensation of the vapor, and the
THE DETECTION AND MEASUREMENT OF RADIATION 45
F10. 2-14. A simple expansion cloud chamber.
path of the ray appears as a thin track of fog. In most expansion chambers,
the gas-vapor mixture is air with water, or argon with ethyl alcohol, at
atmospheric pressure.
A simplified diagram to indica.te the principie of the expansion chamber
is shown in Fig. 2-14. A cylinder of glass C is closed at one end by a glass
window W and at the other end by a metal piston P to form the chamber.
A small amount of water in the chamber keeps the air saturated. When
the piston is pulled down, the air becomes supersaturated as described
above and, in the presence of ionizing radiation, fog tracks are formed.
The tracks can be illuminated by light L from the side and viewed or photo-
graphed through the window. The ions can then be removed by means
of an electric field between the piston and the metal ring R. The piston
is returned to its original position and the chamber is ready for another
burst of radiation.
The Wilson cloud chamber makes it possible to study the behavior of
individual atoms, to photograph the actual paths of ionizing radiations,
and to analyze at leisure the complicated interactions which may take
place between charged particles and individual atoms. Many modifica-
tions of the original cloud chamber have been made. For example, in order
to obtain large numbers of photographs, automatic arrangements have
been made so that the expansion can be repeated rapidly and photographs
taken continuously on motion picture film. If two stereoscopic pictures are
taken simultaneously, the path of the particle in space can be reconstructed.
When a cloud chamber is placed between the pole pieces of an electro-
magnet, it is possible to distinguish between positively and negatively
charged particles. From the curvature of the path of a particle in the
magnetic field, the sign of the charge and the magnitude of the momentum
of the particle can be determined.
46 ATOMS, ELECTRO:SS, AND RADIATIONS [CHAP. 2
Fm. 2-15. Tracks of a-particles in air. The top photograph shows a com-
plete track; the bottom photograph is an enlargement of the end portion of
a track showing two deflections caused by collisions with atoms of the air.
(Reprinted by permission from Rutherford, Chadwick, and Ellis, Radiatiom from
Radioactive Substances. Cambridge University Presa (Macmillan Co.) 1930.)
Different particles produce different types of tracks. Thus, heavy,

slow particles such as a-particles produce broad, densely packed, straight
line tracks with an occasional sharp, small angle bend, especially near
the end of the track. Slow electrons produce narrow, beaded, tortuous
tracks, while fast particles, both light and heavy, produce narrow, beaded
straight tracks. Tracks of a-particles in air are shown in Fig. 2-15. The
top photograph is that of the complete track of an a-particle from radón;
the bottom photograph is an enlargement of the end portion of a track
showing two deflections caused by collisions with atoms of the air.
The diffusion cloud chamber <5 > operates by the diffusion of a con-
densable vapor from a warm region, where it is not saturated, to a cold
region, where it becomes supersaturated. The vapor is usually methyl or
ethyl alcohol. Air is satisfactory for the gas at pressures from 20 cm Hg
to 4 atmospheres, and hydrogen from 10--20 atmospheres. A schematic
diagram of a diffusion chamber is shown in Fig. 2-16. The vapor is in-
troduced at the top, the warm end, and diffuses downward continuously
through a region in which a steady vertical temperature gradient is
maintained by cooling the bottom of the chamber. The diffusion chamber
is continuously sensitive, unlike the expansion chamber, which is sensitive
only for a short time interval after the expansion. This property, together
with its relative mechanical simplicity, gives the diffusion chamber ad-
vantages over the expansion chamber. The latter, however, is still used
in a large portion of serious cloud-chamber work because of the greater
experience with it and because of certain other desirable properties which
it has.
Photographic emulsions. An ionizing particle traveling through the emul-
sion of a photographic plate leaves a track containing a number of develop-
able silver bromide grains. Special photographic emulsions, called nuclear
2-8] THE DETECTION AND MEASUREMENT OF RADIATION 47
•=" 1
\
1
1
1
Glass plate
(wann)
Trough filled
with alcohol
j:,
/
/
Light Glass
7, cylinder
l\Ietal plate
__-_-_-_-- -- - - -Dry ice
FIG. 2-16. Schematic diagram of a diffusion cloud chamber.
emulsions, have been developed. They are distinguished from optical

emulsions in their high silver bromide content, which may be as much as
four times as great as in photographic plates, in the grain size, and in the
thickness of the emulsion. Like the cloud chamber, the photographic
plate, upon development, records the path of the particle, and a variety
of information can be obtained from the study of the tracks. Counting
of the individual paths gives a measure of the number of particles entering
the plate, and study of the detailed structure of the tracks yields informa-
tion about the mass, charge, and energy of the particles. The photo-
graphic emulsion offers advantages over the cloud chamber in that the
emulsion is solid so that the tracks are short, and its sensitivity is perma-
nent rather than restricted to infrequently repeated short intervals. On
the other hand, the cloud-chamber expansions can be controlled by means
of a counter and can be made to record specific events. An emulsion is
limited in the chemical elements which it can contain.
Finally, other instruments such as the bubble counter and the Cerenkov
detector have been developed for use with particles of very high energy
(they are discussed in the book by Price listed in the references).
REFERENCES
GENERAL
J. D. STRANATHAN, The "Particles" of M odern Physics. Philadelphia: Blakiston,
1944.
J. A. CROWTHER, Ions, Electrons, and Ionizing Radiations, 8th ed. London:
Edward Amold, 1949.
R. A. MILLIKAN, Electrons (+ and -), Protons, Photons, Muotrons, and
Cosmic Raya. Chicago: University of Chicago Presa, 1947.
J. B. HoAG and S. A. KoRFF, Electron and Nuclear Phy8ic8, 3rd ed. New York:
Van Nostrand, 1948.
A. P. FRENCH, Principles of Modern Physics. New York: Wiley, 1958.
E. R. CoHEN, K. M. CROWE, and J. W. M. Du»:OND, Fundamental Con-
8tant8 of Physic8. New York: Interscience, 1957.
RADIATION DETECTION
H. H. STAUB, "Detection Methods," Experimental Nuclear Physic8, Vol. I,
Part I, E. Segre, ed. New York: Wiley, 1953.
D. H. W1LKINSON, Ionization Chambers and Counters. Cambridge: University
Press, 1950.
J. B. BIRKS, Scintillation Counters. London: Pergamon Presa, 1953.
S. A. KoRFF, Electron and NuclearCounters, 2nd ed. New York: Van Nostrand,
1955.
J. SHARPE, Nuclear Radiation Detector8. London: Methuen; New York:
Wiley, 1955.
W. J. PRICE, Nuclear Radiation Detection. New York: McGraw-Hill, 1958.
PARTICULAR
l. F. W. CoNSTANT, Theoretical Phy8ic8: ThermodynamicB, Electromagnetism,
Wavu and Particles. Reading, Mass.: Addison-Wesley, 1958, Chapter 9.
2. F. W. SEARS, MechanicB, Wave Motion, and Heat. Reading, Mass.: Addison-
Wesley, 1958, Chapter 15.
3. E. R. CoHEN, K. M. CROWE, and J. S. M. Du»:oNo, op. cit. gen. ref.,
p. 267; also E. R. CoHEN, J. W. M. Du»:oND, J. W. LAYTON, and J. S. RoLLETT,
"Analysis of Variance of ~he 1952 Data on the Atoinic Constants and a New
Adjustment," 1955. Revs. Mod. Phys. 27, 363 (1955).
4. J. G. WILSON, The Principles of Cloud-Chamber Technique. Cambridge:
University Presa, 1951; also N. N. DAS GuPTA and S. K. GHOSH, "A Report
on the Wilson Cloud Chamber and ita Applications in Physics," Revs. Mod.
Phys., 18, 225 (1946).
5. M. SNOWDEN, "The Diffusion Cloud Chamber" Progress in Nuclear
Physics, Vol. 3, O. R. Frisch, ed. London: Pergamon Presa; New York: Acadeinic
Press, 1953.
49
PRoBLEMS
l. Take 1 faraday as 96,500 coul and the valences of chromium, copper,

indium, iron, mercury, nickel, and silver as 6, 2, 3, 3, 2, 2, and 1, respectively.
Show first that 1 amp-hr corresponds to 0.03731 chemical equivalents. Then
calculate (a) the number of grams of each element deposited by 1 amp-hr of
charge, (b) the number of coulombs of charge needed to deposit 1 mg of each
element, (e) the charge in esu associated with 1 gm of each element.
2. In an attempt to measure the value of the charge-to-mass ratio of cathode
rays, a Thomson apparatus with the following dimensions was used:
distance from condenser plates to screen = 32.0 cm,
length of plates = 8.0 cm,
distance between plates = 2.75 cm.
When the potential across the plates was 1100 volts and the magnetic field
strength was 12.50 gauss, the cathode rays passed through the apparatus without
being deflected. When the magnetic field was turned off, the rays were displaced
19.5 cm, on striking the screen, from their undeflected path. What was the
speed of the rays? What value was obtained for e/m?
3. With the apparatus of Problem 2, the potential across the plates was
changed to 1350 volts. What magnetic field was needed to keep the beam
from being deflected? What displacement of the beam was observed when
the magnetic field was turned off?
4. When cathode rays with a speed v move in a direction perpendicular to
a uniform magnetic field of strength H, their path is a circle of radius R given
by the relationship Hev = mv2 /R. Show that the value of e/m can be obtained
from the equation e/m = E/H2 R, where E/H = v. In the experiment of
Problem 2, suppose that the elentric field is turned off instead of the magnetic
field. What is the radius of the path of the cathode rays?
5. Another measurement of e/m for cathode rays was made by means of
two experimenta. In the first, the cathode rays passed between condenser
plates 1 cm apart, with a potential of 730 volts across them. The effect of the
electric field was just balanced by a magnetic field of 28 gauss. In the second
experiment, the rays were bent into a path of 12.00 cm radius by a magnetic
field of 12.50 gauss. What was the speed of the rays? What was the value
obtained for e/m?
6. In experiments with cathode rays, the source often emits rays with a wide
range of speeds, but only rays with a particular speed are desired. How could
the Thomson apparatus be modified to serve as a velocity jilter, i.e., to permit
only the pa.ssa.ge of rays with a. certain desired speed? In the apparatus of
Problem 2, suppose tha.t the potential across the platea is fixed and that the
magnetic field strength can be varied; how would you obtain a monoenergetic
beam of rays with a speed of 2.0 X 109 cm/sec?
7. A uniform, radially directed, electric field of magnitude E is produced in
the space between parallel plates which are closely spaced, concentric, cylindri-
cal sections of mean ra.dius R. Show that the kinetic energy of a particle of
50 ATOMS, ELECTRONS, AXD RADIATIOXS [CHAP. 2
charge q, mass M, and speed v, movmg in the space between the plates is ½qER.
Why does such a system provide an energy filter for charged particles? If the
mean radius is 10 cm, what field strength is needed to permit a beam of cathode
rays with a speed of 1.0 X 109 cm/sec to pass through the filter?
8. In one of his experiments, Millikan observed the motion of an oil drop in
a constant electric field. As the drop captured or lost ions, the time needed
for it to traverse a given distance, 0.5222 cm, changed. Eight successive time
intervals noted were, in seconds, 12.45, 21.5, 34.7, 85.0, 34.7, 16.0, 34.7, and 21.85.
Explain, with the aid of Eq. (2-8), how these data show that the electric charge
on an ion is an exact integral multiple of an elementary charge.
9. An oil drop of radius 2.76 X 10-4 cm and density 0.9199 gm/cm 3 falls in
air at 23ºC and 76 cm Hg. Under these conditions, the viscosity of air is 1.832
X 10-4 cgs and its density is 0.0012 gm/cm 3 . Calculate the rate of fall, first
from the uncorrected Stokes formula, then from the corrected formula¡ in the
latter, the constant bis 0.000625 (cm) (cm Hg).
10. An oil-drop measurement of the charge on the electron was done under the
following conditions:
distance between the condenser plates = 1.60 cm,
potential difference across the platea = 5085 volts,
air temperature = 23ºC,
pressure = 76 cm Hg,
density of oil = 0.9199 gm/cm 3,
radius of oil drop = 2.76 X 10-4 cm.
The speed of free fall of the drop was 0.08571 cm/sec. The greatest common
+
divisor (v 0 VE)o measured for the rise of the drop through a certain fixed
distance was 0.005480 cm/sec. Calculate (a) the uncorrected value e1 of the
electron charge, (b) the corrected value e.
CHAPTER 3
THE NUCLEAR ATOM
3-1 The Thomson atom. The discovery of radioactivity, together

with Thomson's proof of the independent existence of the electron, pro-
vided a starting point for theories of atomic structure. The fact that
atoms of a radioactive element are transformed into atoms of another
element by emitting positively or negatively charged particles led to the
view that atoms are made up of positive and negative charges. If this
view is correct, the total negative charge in an atom must be an integral
multiple of the electronic charge and, since the atom is electrically neutral
under normal conditions, the positive and negative charges must be
numerically equal. The emission of electrons by atoms under widely
different conditions was convincing evidence that electrons exist as such
inside atoms. The first modern theories of atomíc structure were, there-
fore, based on the hypothesis that atoms are made up of electrons and
positive charges. No particular assumptions could be made about the
nature of the positive charges because the properties of the positive par-
ticles from radioactive substances and from gas discharge tubes did not
have the uniformity shown by the properties of the negative particles.
T'Yo important questions then arose: (I) how many electrons are there
in an atom, and (2) how are the electrons and the positive charges arranged
in an atom? lnformation about the first question was obtained experi-
mentally by studying the way in which x-rays interact with atoms, and
this problem will be treated in some detail in the next chapter. It will
suffice, for the present, to state that early experiments of this kind indi-
cated that the number of electrons per atom is of the order of the atomic
weight. It was known that the mass of an electron is about one two-
thousandth of the mass of a hydrogen atom, which has an atomic weight
very close to unity. Hence, the total mass of the electrons in an atom
is only a very small part of the mass of the atom, and it was logical to
assume that practically the entire mass of an atom is associated with the
positive chai:ge.
In the absence of information about the way in which the positive and
negative charges are distributed in an atom, Thomson proposed a simple
model. He assumed that an atom consisted of a sphere of positive elec-
tricity of uniform density, throughout which was distributed an equal
and opposite charge in the form of electrons. It was remarked that the
atom, under this assumption, was like a plum pudding, with the negative
51
52 THE NUCLEAR ATOM [CHAP. 3
electricity dispersed like currants in a dough of positive electricity. The

diameter of the sphere was supposed to be of the order of 10-s cm, the mag-
nitude found for the size of an atom. With this model, Thomson was
able to calculate theoretically how atoms should behave under certain
conditions, and the theoretical predictions could be compared with the
results of experiments. lt became clear when this comparison was made
that Thomson's theory was inadequate; but the failure of the Thomson
model in the particular case of the scattering of a-particles proved to be
most profitable because it led to the concept of the nuclear atom. This
concept is fundamental to atomic and nuclear physics, and the scattering
of a-particles will therefore be discussed in sorne detail.
When a parallel beam of rays from a radioactive substance or from a
discharge tube passes through matter, sorne of the rays are deflected, or
scattered, from their original direction. The scattering process is a result
of the interaction between the rays of the beam and the atoms of the
material, and a careful study of the process can yield information about
the rays, the atoms, or both. The scattering of a-particles was first
demonstrated by Rutherford, who found that when a beam of a-particles
passed through a narrow slit and fell on a photographic plate, the image
of the slit had sharply defined edges if the experiment was performed in
an evacuated vessel. When the apparatus contained air, the image of the
slit on the photographic plate was broadened, showing that some of the
rays had been deflected from their original path by the molecules in the
air. Alpha-particles are also scattered by a very thin film of matter such
as gold or silver foil. When a beam of particles passes through a small
circular hole and falls on a zinc sulfide screen, scintillations are seen
over a well-defined circular area equal to the cross section of the beam.
If a very thin foil is placed in the path of the beam, the a.rea over which
the scintillations occur becomes larger, and its boundary is much less
definite than in the absence of the foil, showing again that sorne of the
particles ha.ve been deflected from their original direction.
The scattering of charged particles such as a-particles can be described
qualitatively in terms of the electrostatic forces between the particles and
the charges which malee up atoms. Since atoms contain both positive
and negative charges, an a-particle is subjected to both repulsive and
attractive electrostatic forces in passing through matter. The magnitude
and direction of these forces depend on how near the particle happens to
approach to the centers of the atoms past which or through which it
moves. When a particular atomic model is postulated, the extent of the
scattering of the a-particles can be calculated quantitatively and compared
with experiment. In the case of the Thomson atom, it was shown that
the average deflection caused by a single atom should be very small. The
1
3-2] RUTHERFORD S THEORY OF THE SCATTERING OF ALPHA-PARTICLES 53
mean deflection of a particle in passing through a thin foil of thickness t

should be, accordin11: to Thomson's theory, m
'Pm =8(1rna-2t) 112 , (3-1)
where fJ is the average deflection caused by a single atom, n is the number
of atoms per cubic centimeter, a is the radius of the atom in centimeters,
and t is the thickness of the foil in centimeters. If the scatterer is a gold
foil 4 X 10-5 cm thick, and if a is assumed to be 10-s cm, then 'Pm turns
out to be about 308. The scattering of a-particles by a thin foil, according
to the Thomson theory, is the result of a relatively large number of small
deflections caused by the action of a large number of atoms of the scatter-
ing material on a single a-particle. This process is called compound, or
multiple, scattering.
In an experiment, it is convenient to count the number of a-particles
scattered through a certain angle. Rutherford showed that the number
of a-particles N • scattered through an angle equal to, or greater than,
q, should be given by 2
N • = N oe-<•l•m> , (3-2)
where N O is the number of particles corresponding to q, = O, and t/>m is the
average deflection after passing through the foil. It was found experi-
mentally by Geiger <2 > that when a gold foil 4 X 10-5 cm thick was used,
the most probable angle of deflection of a beam of a-particles was about 1º.
If this value is used for 4>m in Eq. (3-2), it can be seen that the probability
that an a-particle is scattered through a large angle becomes vanishingly
small. For example, the number of a-particles which should be scattered
through an angle of 10° or more is
N10° = Noe- 100 ""10-43No.
Geiger found that the scattering agreed with that predicted by Eq. (3-2)
for very small angles, that is, for very small values of t/>, but the number of
particles scattered through large angles was much greater< 3> than that pre-
dicted by the Thomson theory. In fact, one out of about every 8000
a-particles was scattered through an angle greater than 90°, which means
that a significant number of a-particles in a beam incident on a foil had
their directions changed to sueh an extent that they emerged again on
the side of incidence. The experimental scattering of a-particles at large
angles could not possibly be reconciled with the theoretical predictions
based on multiple scattering by a Thomson atom, and it was necessary to
look for a better model for the atom.
3-2 Rutherford's theory of the scattering of alpha-particles. Ruther-

ford <4> (1911) proposed a new theory of the scattering of a-particles by
matter; this theory was based on a new atomic model and was successful
in describing the experimental results. Rutherford suggested that the
deflection of an a-particle through a large angle could be caused by a
single encounter with an atom rather than by multiple scattering. Photo-
graphs of the tracks of a-particles in a cloud chamber showed that a-par-
ticles often traveled in a straight line for a considerable distance and then
were deflected suddenly through a large angle. Rutherford's suggestion
was in ágreement with this kind of experimental evidence. For large
angle scattering to be possible, it was necessary to suppose that there is
an intense electric field near an atom. Rutherford proposed a simple
model of the atom which could provide such a field. He assumed that
the posiiive charge of the atom, instead of being distributed uniformly
throughout a region of the size of the atom, is concentrated in a minute
center or nucleus, and that the negative charge is distributed over a sphere
of radius comparable with the atomic radius. On this model, an a-particle
can penetrate very close to the nucleus before the repulsive force on it
becomes large enough to turn it back, but the repulsive force can then
be very large, and can result in a large deflection. At the same time, when
the a-particle is near the nucleus, it is relatively far from the negative
charges, which are spread over a much larger volume, so that the attractive
forces exerted on the a-particle by the electrons can be neglected. For
purposes of calculation, Rutherford assumed that the nuclear and a-par-
ticle charges act as point charges and that the scattering is caused by the
repulsive electrostatic force between the nucleus and the a-particle. If
the magnitude of the a-particle charge is 2e and that of the nucleus is
Ze, where Z is an integer, and if r is the distance between the two charges,
the magnitude of this force is
2Ze2
F =r2 -· (3-3)
In bis first calculation, Rutherford treated the case of an atom suffi-

ciently heavy so that the nucleus could be considered to remain at rest
during the scattering process. With the above assumptions, the calcula-
tion of the orbit of the a-particle is reduced to a familiar problem of classical
mechanics, that of the motion of a highly energetic particle under a repul-
sive inverse square law óf force. <5 > The orbit is one branch of a 4yperbola
with the nucleus of the atom as the externa! focus, * as shown in Fig. 3-1,
and the formula for the deflection can be derived from geometrical and
physical relationships.
The first step in the derivation of the Rutherford scattering formula is
to write down sorne useful geometrical relationships. In Fig. 3-1, the
* This statement is proved in Appendix IV, where a second derivation of the
Rutherford scattering formula is given which contains certain points of special
interest.
3-2] RUTHERFORD'S THEORY OF THE SCA'M'ERING OF ALPHA-PARTICLES 55
y
A'
.
, .....
,,,
r Nucleus
X
e
Trajectory of
a-particle -
Asymptote of
trajectory
A
Frn. 3-1. The scattering of an a-particle by the Rutherford nuclear atom.
origin of the coordinate system is taken at the center of the hyperbola,

one of whose branches is the path of the a-particle. The initial path of
a particle, when the latter is far from the nucleus, is along AO, one of
the asymptotes of the hyperbola; the perpendicular distance from the
nucleus, situated at the external focus x = e, to the line OA is called the
impact parameter and is denoted by p. The a-particle, on nearing the
nucleus, is deflected through the angle q, and approaches the asymptote
OA'. The vertex of the orbit is at x = - a, while that of the second branch
of the hyperbola is at x = a. The equation of the hyperbola is, from
analytic geometry,
x2 y2
-~-= 1
a2 c2 - a2 '
(3-4)
where a is the major semiaxis of the hyperbola. It can also be seen from
the figure that p 2 = c2 - a2 , so that pis the minor semiaxis. The eccen-
tricity E is defined as the ratio E = e/a, and the angle between the x-axis
and the initial direction of the a-particle is denoted by O. The angle of
deviation q, is then equal to 1r - 20 radians. Now, let s denote the dis-
tance of the nucleus from the vertex of the orbit of the a-particle; the
magnitude of s is
s = e+ a = e( 1 + ~) = c(l + cos O).
Also, e = p/sin O, so that
s = p(l '. cos O) = p cot !!. . (3-5)
sm O 2
The next step is to find a relationship between the impact parameter
p and the scattering angle q,, which can be done by applying the laws of
conservation of energy and angular momentum. According to the former,

the sum of the kinetic energy and the potential energy is constant. When
the a-particle is at a great distance from the nucleus, its potential energy,
which is inversely proportional to this distance, is practically zero. If
the velocity of the a-particle is V at thís large separation, which is assumed
to be infinite, its total energy is equal to its initial kinetic energy, MV 2 /2,
where M is the mass. This energy must also be equal to the total energy
oí the a-particle when it is just at the vertex of the hyperbola. If the
velocity at this point is V 0 , then
½MV2 = ½MV~ + 2ze2.8

(H)
The second term on the right represents the potential energy of the
a-particle, at the vertex of its orbit, in the electric field of the nucleus.
If the last equation is divided through by ½MV 2 and if a new quantity b
is introduced, b = 4Ze 2 /MV 2 , the result is
vg
y =
b
2 1-
8· {3-7a)
When the expression for s from Eq. (3-5)"is inserted into the last equation,
1 _ ~
V~= sin 11
(3-7b)
V2 p {l + cos 11)
It follows from the law oí conservation of angular momentum that
MVp = MV 0 s, (3-8)
or
V0 p sin/1
v=;= 1+cos8
1
and
V~ sin 2 8 1 - cos 8
v2 = (1 + cos 11)2 = (3-9)
1 + cos 11
When this value for (V 0/V) 2 is put into Eq. (3-7), it is found that
b tan 8
p=-2-' (3-10)
and, since ef, = 1r - 28,
p = -b2 cot -,¡,2 · (3-11)
Equation {3-11) is the desired relation between the impact parameter

and the scattering angle.
3-2) RUTHERFORD'S THEORY OF THE SCATTERING OF ALPHA-PARTICLES 57
It is now possible to calculate the fraction of the a-particles scattered

through a given angle q,. Suppose that the beam of a-particles is incident
perpendicularly on a thin foil of material of thickness t, containing n
atoros per unit volume. It is assumed tbat the foil is so thin that the
particles pass through without any significant change in velocity and
that, with the exception of a few particles which are scattered through a
large angle, the beam passes perpendicularly through the foil. Then the
chance that a particle passes within a distance p of a nucleus is
(3-12)
A particle which moves so as to pass within a distance p of the nucleus is

scattered through an angle greater than q,, where q, is given by Eq. (3-11).
Hence the fraction of the total number of a-particles deflected through
an angle greater than q, is obtained by inserting for p, from Eq. (3-11),
and
q = ¼'ll"ntb 2 cot 2 P. · (3-13)
2
Similarly, the probability of deflection through an angle between q, and

q, + dq, is equal to the probability of striking between the radii p and
p + dp, and is given by
dq = 21rpnt dp.
Then,
dq = ¼1rntb 2 cot Í cosec Í dq,

2
cos Í cosec ~ dq,

3
= ¼1rntb 2
= i1rntb 2 sin q, cosec 4 ! dq,. (3-14)
In the experiments made to test the theory, the scattering was deter-
mined by counting the number of a-particles incident perpendicularly on
a constant area of a zinc sulfide screen placed at a distance R from the
foil. The fraction of the scattered a-particles falling on an element of
area of the screen at a distance R is given by
dq _ ntb 2 cosec 4 (q,/2)

(3-15)
21r R sin q, dq, -
2 16 R2
If now, Q is the total number of a-particles incident on the foil, and if

Y is the number of a-particles scattered to unit area of the zinc sulfide
58 THE NUCLEAR ATOM [CHAP, 3
screen placed at a distance R from the foil and at an angle 4> with the
original direction of the particles, then
2 4
y = Qntb cosec (4>/2) .
(3-16)
16R 2
According to Rutherford's theory, the number of a-particles falling on

a unit area of the zinc sulfide screen at a distance R from the point of
scattering should be proportional to
l. cosec 4 (4>/2), where 4> is the scattenng angle,
2. t, the thickness of the scattering material,
3. I/(MV 2 ) 2 , orto the reciprocal of the square of the initial energy of
the a-particle,
4. (Ze) 2 , the square of the nuclear positive charge.
3-3 The experimental test of the Rutherford scattering theory.

Rutherford's nuclear theory of the scattering of a-particles was tested
point by point in 1913 by Geiger and Marsden.< 6 > The dependence of
the scattering on the four quantities listed at the end of the last section
will be considered in order.
1. The dependence of the scattering on the angle of deftection. The effect
of varying the angle of deflection 4> was studied in the apparatus shown
schematically in Fig. 3-2. In the diagram, R represents a radioactive
substance which is the source of the a-particles, F is a very thin foil of
scattering material, and S is a zinc sulfide screen rigidly attached to a
microscope M. The source and foil were held fixed, while the screen
and microscope could be rotated in an airtight joint, varying the angle
of deflection. The entire apparatus was enclosed in a metal box which
could be evacuated. The number of a-particles reaching unit area of
the screen in a chosen time interval was obtained by counting the scintil-
lations. In the experiment, the angle 4> was varied while all of the other
variables in Eq. (3-16) were held constant. The number of scintilla-
tions counted, N, is proportional to Y, orto cosec 4 (4>/2); hence, the ratio
N /cosec 4 (4>/2) should be constant for a given foil under the conditions
of the experiment.
The results of two sets of experiments, one with a silver scattering foil,
the other with a gold foil, are given in Table 3-1. The first column gives
the values of the angle et, between the direction of the incident beam of
a-particles and the direction in which the scattered particles were counted;
the second column gives the corresponding values of cosec 4 (4>/2). Col-
ums III and V give the observed numbers N of scintillations for silver
and gold respectively; columns IV and VI show the value oí the ratio
N /cosec 4 (4>/2). The variation in the value of the ratio is very small
3-3) TEST OF THE RUTHERFORD SCATTERING THEORY 59
R
1 </>
1 F
Fm. 3--2. Apparatus of Geiger and Marsden for testing the angular depend-
ence of a-particle scattering. <6l
TABLE 3-1 *
THE DEPENDENCE OF THE ScATTERING OF ALPHA-PARTICLES
ON THE .ANGLE OF DEFLECTION
I II III IV V VI
Silver Gold
Angle of Number of Number of

deflection scintillations N scintillations N
cosec4 (f/¡/2) N cosec4 (f/,/2) N cosec 4 (f/¡/2)
150°
"' 1.15 22.2 19.3 33.1 28.8
135 1.38 27.4 19.8 43.0 31.2
120 1.79 33.0 18.4 51.9 29.0
105 2.53 47.3 18.7 69.5 27.5
75 7.25 136 18.8 211 29.1
60 16.0 320 20.0 477 29.8
45 46.6 989 21.2 1435 30.8
37.5 93.7 1760 18.8 3300 35.3
30 223 5260 23.6 7800 35.0
22.5 690 20,300 29.4 27,300 39.6
15 3445 105,400 30.6 132,000 38.4
30 223 5.3 0.024 3.1 0.014
22.5 690 16.6 0.024 8.4 0.012
15 3445 93.0 0.027 48.2 0.014
10 17,330 508 0.029 200 0.012
7.5 54,650 1710 0.031 607 0.011
5 276,300 ......... . .... 3320 0.012
* From Geiger and Marsden. <6 >
compared wíth that of cosec 4 (,t,/2), for angles between ,t, = 15º and
4> = 150°. For smaller angles, it was found desirable to reduce the
number of scintillations counted; the value of the ratio was practically
constant between ti> = 5° and ,t, = 30°. The results for the smaller
angles can be compared with those of the larger angles by noting that in
the case of the gold foil, the number of scintillations was reduced by
about 2500. When the results are fitted to those for the larger angles,
it is clear that the value of the ratio changes little over the entire range
of values of ,¡,, while the value of cosec 4 (q,/2) varies by a factor of 250,000.
The deviations of the ratio from constancy were thought to be within the
experimental error and it was concluded that the theory predicts the
correct dependence of the scattering on the angle of deviation.
2. The dependence of the scallering on the thickness of the scallering material.
The dependence of the scattering on the thickness of the scattering material
was tested by fixing the a.ngle of deflection and using foil» of different
thicknesses and also of different materials. The results of several experi-
ments are shown in Fig. 3-3, in which the number of particles per minute
scattered through an angle of 25° is plotted as ordinate and the thickness t
of the scattering foil is plotted as abscissa. The thickness of the foil is
expressed in terms of the equivalent length of path in air, that is, the
thickness of air which produces the same loss in energy of the a-particles
traversing it as that produced by the material ·being studied. The equiv-
alent path length in air often serves as a useful standard for comparison
160....----.----.----...----...------,
"' 120
;l
.5
...e
"'o.
"'oe
80
~·.;::
e
·¡;
.,
-...
o
Q)
.::,
40
e:,
;,;
Thirkness of foils in rentimeters oí air equivalent

F10. 3-3. The variation of the scattering of a-particles with the thickness
of the scattering material. (6>
3-3] TEST OF THE RUTHERFORD SCATTERING THEORY 61
in work with a-particles. The straight lines show that for any one element,
the number of particles scattered through a given angle per unit time is
directly proportional to the thickness of the foil, in agreement with the
theory.
3. The dependence of the scallering on the energy or the velocity of the
a-particles. According to the Rutherford scattering formula, the number
of a-particles scattered through an angle q, should be inversely propor-
tional to the square of the energy of the particles, or to the fourth power
of the velocity. In a series of experimenta, Geiger and Marsden varied
the velocity of the a-particles from a given source by placing screens of
mica between the source and the scattering foil; the thicker the screen,
the slower were the particles which reached the scattering foil. The
velocities of the particles were determined by finding how far they traveled
in air. This distance, called the range of the a-particles, could be deter-
mined in several ways, which will be discussed in Chapter 13. The range
R was known to be related to the velocity by the empirical formula,
(3-17)
where a is a known constant. When the velocity of the a-particles which

passed through a given thickness of mica screen had been obtained in this
way, the scattering through a known angle was measured by counting
the number N of scintillations. The product NV 4 should be constant
when V is varied. The resulta of a typical experiment are shown in
Table 3-2. The fourth column gives the number N of scintillations per
minute under fixed conditions, when a-particles of the ranges given in
the second column were used. The relative values of l/V 4 are given in the
TABLE 3-2*
THE VARIATION OF ALPHA-PARTICLE 8CATTERING WITH VELOCITY
Number N
Numberof Range of Relative of
sheets a-particles values of scintillations
of mica (cm) 1/V4 per minute NV 4
o 5.5 1.0 24.7 25

1 4.76 1.21 29.0 24
2 4.05 1.50 33.4 22
3 3.32 1.91 44 23
4 2.51 2.84 81 28
5 1.84 4.32 101 23
6 1.04 9.22 255 28
* Ftom Geiger and Marsden. C6>
62 THE NUCLEAR ATOM
third column. The product NV 4 given in the last column is very nearly
constant over the range of values of V studied, showing that the scattering
varíes inversely as the fourth power of the velocity, as predicted by
Rutherford's theory.
4. The dependence of the scattering on the nuclear charge. The scattering
angle, the thickness of the scatterer, and the velocity of the incident
a-particles are quantities which could be measured directly and their
effect on the scattering determined. The nuclear charge, unlike the other
parameters, could not be measured directly, and a direct comparison
between theory and experiment could not be made in this case. lt is
evident from Eq. (3-16), however, that the value of Z can be found by
counting the number of a-particles in the beam incident on the scattering
foil and the number in the scattered beam under fixed geometric condi-
tions. Sorne information about Z could also be obta.ined from experi-
ments on the scattering by different foil materials. From these two types
of experiments it was found that for elements heavier than aluminum,
the positive charge Ze on the nucleus was approximately ½Ae, that is,
Z .., A/2, where A is the atomic weight and e is the electronic charge.
These experiments were not accurate enough to provide a reliable determi-
nation of the nuclear charge Z. lt was not until 1920 that Chadwick, <7 >
using improved scattering techniques, succeeded in measuring the nuclear
e:harge with good precision. For platinum, silver, and copper foils, he
obtained
copper: Z = 29.3 ± 0.5,
silver: Z = 46.3 ± 0.7,
platinum: Z = 77.4 ± l.
These results are not precise enough to determine unique, integral values
of Z, but, as will be seen in the next chapter, they agree well with the
values 29, 47, and 78 for the three elements, obtained by an entirely inde-
pendent method. Thus, ali four tests of the Rutherford scattering theory
were met successfully and constitute the earliest if not the greatest single
piece of experimental evidence for the nuclear model of the atom.
3-4 Some charactetjstics of the atomic nucleus. The remarkably

good agreement between the predictions of Rutherford's theory and the
experimental results was interpreted as establishing the correctness of
the concept of the nuclear atom. Since 1913 the atom has therefore been
considered to consist of a minute, posítively charged nucleus around which
is distributed, in sorne way, an equal and opposite negative charge in
the form of electrons.
So far, the atomic nucleus itself is a vague concept. lt has been described
as "minute" or "very small" and has been treated mathematically as a
3-4] SOME CHARACTERISTICS OF THE ATOMIC NUCLEUS 63
point; at the same time, it is supposed to contain practically all of the

mass of the atom. It is clear that quantitative information is now needed
about the size of the nucleus. The first information of this kind was
obtained from the experiments on the scattering of a-particles; it carne
from the consideration of the distance of closest approach of an a-particle
to a nucleus, and of the range of validity of the Coulomb force law.
For any hyperbolic orbit, the distance of closest approach is s, the dis-
tance from the vertex of the hyperbola to the nucleus as given by Eq. (3-5).
The smallest value that s can have is that for a head-on collision, when
the a-particle is deflected through an angle of 180º. In such a collision,
the velocity of the a-particle at the turning point is just zero. It follows
from Eq. (3-7a), that for V 0 to vanish, s must be equal to b; furthermore,
since V O can never be negative, this is the smallest value which s may have.
Hence, the quantity b, defined by
4Ze 2
b = MV2' (3-18)
gives the closest distance which an a-particle of velocity V can approach

to a nucleus of charge Z. The roagnitude of this distance can be estimated
by calculating a typical value of b. Consider the case of a copper nucleus
bombarded by a-particles from radon. Copper has an atomic weight of
63.5; if the results of Geiger and Marsden for Z as given by scattering
experiments are used, then Z for copper is approximately half of 63.5,
or 32. An a-particle has a mass four times that of the hydrogen atoro,
or 4 X 1.67 X 10-24 gro; the velocity of an a-particle from radon is close
to 1.6 X 10 9 cm/sec. With e = 4.8 X 10- 10 esu, the result for bis
4(32)(4.8) 210-2 º -12

b = (1.67}4(10-24)(1.6)2(1018) ""' 1. 7 X 10 cm.
The above calculation depends on the assumption that the Coulomb

force law between the a-particle and the nucleus is still valid at such small
distances from the nucleus. The validity of this assumption was borne
out by the agreement between the Rutherford scattering theory and the
results of the experiments of Geiger and Marsden. By using faster
a-particles, Rutherford and others extended the experiments to see how
close to the nucleus the 1/r2 force law holds. The results showed that for
silver the Coulomb law held down to 2 X 10- 12 cm, for copper down to
1.2 X 10- 12 cm, and for gold down to 3.2 X 10- 12 cm. lt might be ex-
pected that if an a-particle approaches more closely to a nucleus, the inverse
square law would eventually break down. If this happens, the forces
between the a-particle and the nucleus should begin to change very rapidly
with the distance, and the scattering of a-particles should depart widely
from the predictions of the theory. If the nucleus is defined as the region
of deviation from the Coulomb force law, then for the elementa mentioned,
the radii oí the respective nuclei are smaller than the distances listed.
Thus, the nuclei of these elementa are about 10- 12 cm in radius, and are
indeed very small compared with an atom, with ita radius oí 10-s cm. It
will be seen in later chapters that other methods of estimating values of
nuclear radii give resulta in good agreement with that just obtained.
It is seen from Eq. (3-18) that for a-particles of a given energy the
distance of closest approach is proportional to the nuclear charge Z. On
the basis oí the finding that Z was approximately proportional to the
atomic weight, it was expected that a-particles might come closer than
10- 12 cm to light nuclei, and that it might be possible to find departures
from the Coulomb force law. Theoretical and experimental studies showed
that such departures from the inverse square law do indeed exist. es> In
the case of aluminum, the inverse square law was found to break clown at
about 6 to 8 X 10- 13 cm, with similar resulta for other light elements.
The deviations from the inverse square law scattering showed that very
close to the nucleus, the repulsion was smaller than that calculated from
the Coulomb force alone. These resulta provided the first evidence of
the existence of a nonelectrical, specifically nuclear force.
The study of the scattering of a-particles has continued to be an im-
portant source oí information concerning the atomic nucleus. <9 • 1º· rn By
1935, data had been collected on the elastic scattering of a-particles from
most of the light elements through aluminum. In each case departures
from Coulomb scattering were observed. With the aid of newer theoreti-
cal methods, these data could be used to make quantítative estimates of
nuclear radü. It was shown <9 > that the data could be interpreted in a
consistent way if the radius of the nucleus is assumed to be approximately
proportional to the cube root oí the atomic weight, that is, if
r = r0 A 113 ,
where A is the atomic weight, and r 0 = 1.4
to 1.5 X 10- 13 cm. This
property will be discussed further in Chapter 13.
PROBLEMS 65
REFERENCES
GENERAL
RUTHERFORD, CHADWICK and ELLis, Radiatiom from Radioactive Substances.
New York: Macmillan, 1930, Chapters 8, 9.
R. D. EvANs, The Atomic Nucleus. New York: McGraw-Hill, 1955, Appen-
dix B.
PARTICULAR
l. J. J. THOMSON, "On the Scattering of Rapidly Moving Electrified Particles,"
Proc. Cambridge Phil. Soc., 15, 465 (1910).
2. H. GEIGER, "The Scattering of the a-Particles by Matter," Proc. Roy.
Soc. (London), 83A, 492 (1910).
3. H. GEIGER and E. MARSDEN, "On the Diffuse Reflection of the a-Particles,"
Proc. Roy. Soc. (London), 82A, 495 (1909).
4. E. RUTHERFORD, "The Scattering of a- and ,8-Particles by Matter and the
Structure of the Atom," Phil. Mag., 21, 669 (1911).
5. K. R. SYMON, Mechanics. Reading, Ma.ss.: Addison-Wesley, 1960, Chapter 3.
6. H. GEIGER and E. MARSDEN, "The Laws of Deflection of a-Particles
Through Large Angles," Phil. Mag., 25, 604 (1913).
7. J. CHADWICK, "Charge on the Atomic Nucleus and the Law of Force;
Validity of the Inverse Square Law for the Pt Atom," Phil Mag., 40, 734 (1920).
8. RuTHERFORD, CHADWICK, and ELLIS, op. cit. gen. ref., Chapter 9.
9. E. PoLLARD, "Nuclear Potential Barriera: Experiment and Theory,"
Phys. Rev., 47, 611 (1935).
10. N. F. MoTT and H. S. W. MASSEY, The Theory of Atomic Collisions,
2nd ec,i. Oxford, England: Clarendon Press, 1950.
11. R. M. E1sBERG and C. E. PoRTER, "Scattering of Alpha Particles," Revs.
Mod. Phys., 33, 190 (1961).
PROBLEMS
l. A beam of a-particles (kinetic energy 5.30 Mev) from polonium, of in-

tensity 10,000 particles/sec, is incident normally on a gold foil of density 19.3
gm/cm 3 and 1 X 10-5 cm thick. An a-particle counter with an aperture 1 cm 2
in area is placed at a distance of 10 cm from the foil in such a way that a line
from the center of the counter to the center of the area at which the beam strikes
the foil makes an angle of iJ, degrees with the direction of the beam. Calculate
the number of counts per hour for ip = 5, 10, 15, 30, 45, and 60°.
2. Repeat the calculation of Problem 1, but with 8.00-Mev a-particles.
3. Repeat the calculation of Problem 1 with a silver foil instead of a gold foil.
The density of silver is 10.5 gm/cm3 •
4. In the setup of Problem 1 a copper foil 1 X 10-5 cm thick and of density
8.90 gm/cm 3 is used instead of the gold foil. When iJ, = 10°, the counting rate
is 820 counts/hr. Calculate the atomic number of copper from these data.
66 THE NUCLEAR ATOM [CHAP, 3
5. What is the distance oí closest a.pproach of 5.30-Mev a-particles to nuclei

of the following elements: gold, silver, copper, le&d, and uranium? How close
would 7.00-Mev a-particles come to the same nuclei?
6. Show that the fraction of a-p&rticles scattered through &n angle between
90° and 180° is given by ¼ntrb2 • What fraction of the a-particles of Problem 1
are scattered through an angle greater than 90º?
7. Wha.t is the general expression for the distance of closest approach of &
positively charged particle to a nucleus? What is the distance of closest approach
of a 2-Mev proton to a gold nucleus? How does this distance compare with
those for a deuteron andan a-particle of the same energy?
8. A beam of protons of 5-Mev kinetic energy traverses a gold foil; one
particle in 5 X 106 is scattered so as to hit a surface 0.5 cm 2 in area at a distance
10 cm from the foil and in a direction making an angle of 60º with the initial
direction of the beam. What is the thickness of the foil?
CHAPTER 4
X-RAYS AND ATOMIC STRUCTURE
The work of Geiger and Marsden on the scattering of a-particles con-

firmed Rutherford's concept of the nuclear atom, and showed that the
number of elementary positive charges on the atomic nucleus is approx-
imately equal to half of the atomic weight. This number must be the
same as the number of electrons in the atom, because the atom is electrically
neutral. lndependent evidence for the number of electrons in the atom
was supplied by the pioneer work of Barkla on the absorption and scatter-
ing of x-rays by matter. The relationship between the nuclear charge
and the position of an element in the periodic table was determined by
Moseley's work on the characteristic x-ray spectra of the elements. The
study of atomic properties by means of x-rays, with the work of Barkla
and Moseley as examples, has been one of the most fruitful sources of
information about atoms, and will be discussed in the present chapter.
In order to do so, it is necessary to review sorne of the properties of x-rays
in somewhat greater detail than was done in the second chapter.
4-1 Some properties of :x-rays. X-rays are produced when swiftly

moving electrons strike a solid target. According to classical electro-
dynamics, a moving charged particle emits electromagnetic radiation
when it is accelerated; the sudden stopping of an electron gives rise to a
pulse of radiation which takes the form of x-rays. In practice, x-rays
are sometimes produced in a low pressure, gas-filled cathode-ray tube in
which a metal anticathode is situated opposite the cathode. The anti-
cathode serves as a target for the electrons emitted from the cathode
and as the source of x-rays. In the more frequently used Coolidge tube
(Fig. 4-1), the cathode is a wire which is heated to such a temperature
that it emits thermoelectrons. The tube is evacuated until there is no
appreciable amount of gas remaining, so that all of the current through
the tube is carried by the thermoolectrons. The anticathode is usually
a metal of high atomic weight such as tungsten, because the energy carried
by the x-rays from heavy metals has been found to be greater than the
energy of the x-rays from light metals.
One of the most important properties of x-rays is, of course, their strong
penetrating power. The extent to which a beam of x-rays will penetrate
into a substance dependa on the nature of the x-rays as well as on that of
67
68 X-RAYS AND ATOMIC STRUCTURE [CHAP. 4
Fm. 4--1. Coolidge x-ray tube.
the substance, and to some extent x-rays can be characterized by their

penetrating power in a given material. A rough differentiation may be
made into soft rays, which have only small penetrating power and are
easily absorbed, and hard rays, which have relatively high penetrating
power. A distinction may also be made, on the basis of their absorption,
between homogeneous and het,erogeneous x-rays. The absorption of a homo-
geneous radiation can be described by the formula
(4--1)
where I O represents the observed intensity of the beam incident normally
on the absorbing material, I is the intensity after passing through a thick-
ness d, and µ. is a constant, called the absorption coefficient, which depends
on the absorbing material and on the radiation but is independent of
the initial intensity of the x-rays. The absorption equation (4-1) is
valid only for homogeneous x-rays, and really defines what is meant by
homogeneous x-rays. It does not hold, for example, for the radiation
from an ordinary x-ray tube. In that case, no constant value of µ. is
obtained. With increasing thickness of the absorber the value of µ. con-
tinually decreases, and finally approaches a limiting value. The variation
of µ. with thickness of the absorber indica.tes that the x-rays from the tube
are heterogeneous and only become approximately homogeneous when
the softer radiations are filtered out of the beam.
The process of absorption of x-rays is complicated and involves several
phenomena closely connected with the properties of atoros. When a
beam of primary x-rays coming from the anticathode of an x-ray tube
falls on a plate of sorne chosen element, part of the radiation goes on
through the plate, while the rest is transformed into heat, or into radia-
tion of another sort. The rays going out from the plate consist in part
of primary x-rays, the transmitted beam, and in part of rays excited by
the primary beam and called collectively the secondary radiation. Figure
4-2 is a schematic representation of what happens. The secondary radia-
4-1) SOME PROPERTIES OF X-RAYS 69
__.,,,,
Primary x-ray beam Transmitted x-ray beam
Absorbin
and emitting cJ,8
plate
---------...
Fm. 4-2. lnteraction between x-rays and rnatter.
tion contains four distinct, important types of radiation: (a) scattered

x-rays, (b) characteristic x-rays, (c) scattered /3-rays, (d) characteristic
/3-rays. The /3-rays are not of interest for the present and will be discussed
in a later chapter. The x-rays of the secondary radiation have tumed out
to be a fruitful source of information about atoros.
The scattered x-rays have very nearly the same absorption coefficient
in a given material as the primary rays; they seem to be primary rays
which have merely had their direction changed by the material through
which they pass, and their character is independent of the nature of the
absorbing and scattering material. Since the primary rays which come
from an x-ray tube are heterogeneous, the scattered radiation is also
heterogeneous provided the absorbing plate is not too thick.
The characteristic x-rays are so called because they are typical of the
material of the plate and, in addition, are homogeneous. They always
have the same properties for a given chemical element, independently of
the hardness of the primary rays; the necessary condition that the radia-
tion shall appear at all is, however, that the primary radiation shall have
a certain mínimum hardness. Barkla and his co-workers, during the
early years of the 20th century, made detailed measurements of the
absorption coefficients of the characteristic x-radiation from various
elements. They found that the elements have, in general, two types of
characteristic radiation which differ greatJy in their absorption coefficients.
These are called the K- and L-radiations; the K-radiation is the harder
and may have an absorption coefficient several hundred times smaller
than that of the L-radiation. The characteristic K- and L-radiations
of the different elements can be compared in terms of their absorption
coefficients in aluminum. The absorption coefficient for each of the
radiations decreases with increasing atomic weight; in other words, the
characteristic radiation of the elements increases in hardness with increas-

ing atomic weight.
With the above information about x-rays as background, we can now
consider sorne of the experiments with x-rays which added to the knowl-
edge of atomic structure.
4-2 The scattering of :s:-rays by atoms and the number of electrons

per atom. The scattering of the primary x-rays by a material plate was
treated by Thomson on the basis of the classical theory of electrodynamics.
According to this theory, radiation incident on a. charged particle which
can move freely exerts a force on the particle; the latter is accelerated
and emits electromagnetic radiation. When a beam of x-rays strikes
the atoros in the plate, the latter are accelerated. The electrons, because
of their small mass, are accelerated much more strongly than are the
atomic nuclei, and the radiation produced by the latter may be neglected.
The electrons seem to abstract energy from the primary beam and re-ra-
diate, or scatter it, as a seconda.ry beam. The procesa takes place in such
a way that, for elements of relatively low atomic weight and x-rays of
moderate hardness, the scattered and incident. radiations differ only in
direction. The magnitude of the energy scattered by a free electron was
calculated by Thomson, cu and can be written as
2 2
I = Sr
3 ( mc2
e ) lo. (4-2)
Here, I O is the intensity of the primary radiation and has the units of
energy per square centimeter, but I denotes the entire scattered energy.
The quotient I / I O must have the dimensiona of an area; it is often denoted
by 'Po, and is called the cla88ical sca.Uering coejficient f or a free electron.
Of the incident radiation that falls on unit area of a surface drawn per-
pendicular to the beam, a fraction 4> 0 is scattered, and the electron may
be said to scatter as much radiation as falls on an area equal to 4> 0 • For
this reason, 4> 0 is also called the classical cross section f or scattering by a free
electron. If the values e = 4.80 X 10- 10 esu, m = 0.911 X 10-27 gm,
°
and e = 3.0 X 10 1 CIJ?./sec are inserted, then
2 2
4> 0 = 38r ( mc
e ) =
2 6.65 X 10-25 cm 2 . (4-3)
If 4> 0 is regarded as the effective cro88 &ection of the electron, then the
effedive radim of the electron is given by
8)1/2 e2
a= ( mc = 4.60 X 10- 13 cm. (4-4)
3 2
4-2] THE SCATTERING OF X-RAYS BY ATOMS 71
The effective radiu& of the electron is therefore of the same order of magni-
tude 88 the radius of the lighter nuclei. The quantity e2 /mc 2 = 2.82
x 10-13 cm is usually referred to 88 the classical radiu& of the electron.
When a beam of x-rays passes through a thin sheet of material, some of
the energy is removed from the beam, and the intensity of the beam is
reduced. If there are n electrons per atom, and N atoras per unit volume,
and if it is assumed that all of the electrons scatter independently, then
the fractional diminution -al/ I of the incident beam in going a distance
llx is given by
M
- T = nN♦o ll:e. (4-5)
Since q, 0 is independent of x, both sides of this equation can be integrated

directly giving, for the intensity l of the beam after traversing a thickness
x of the material,
(4-6)
lt follows from the 188t equation that
I I
n =- --ln-·
Nf/,oX lo
(4-7)
In Eq. (4-7), N and q,0 are known, and x is the chosen thickness of the
scatterer; 10 and I can be measured experimentally, for example, with
ionization chambers. Hence, n, the number of electrons per atom, can
be found from measurementa of the scattering of x-rays.
Experimenta of this type were performed by Barkla and his collab-
orators ;<2,3> actually, other processes besides scattering were present in
these experimenta, but the necessary corrections could be made. Barkla
found that within certain limita of hardness of the x-rays, the scattering
could be represented by
nNq,0 = 0.2p, (4-8)
independently of the hardness and of the scattering material. In Eq. (4-8),

p is the density in grams per cubic centimeter.
Hence, with the known value of q, 0 ,
nN = 3.0 X 10 23p.
Now, the number of atoros per cubic centimeter of an element is
NoP
N=--:;¡-,
where N O is Avogadro's number = 6.02 X 10 23 atoms per gram atomic

weight, and A is the gram atomic weight of the element. Then,
nN = n~al! = 3.0 X 10 23p,

and
n 1
A= 2' (4-9)
that is, the number of electrons per atom is very close to half the atomic
weight.
This result must be considered to be approximate for several reasons.
First, it has been assumed that the laws of classical electromagnetics hold,
an assumption which is valid only within certain limits of hardness of the
x-rays and for scattering materials of low atomic weight. Second, the
relationship (4-8) is approximate. For these reasons, the result that
the number of electrons is half the atomic weight was a useful qualitative,
or at best semiquantitative, result. It was, however, in agreement with
the equally semiquantitative results for the charge on the nucleus obtained
from a-scattering measurements. The fact that two completely inde-
pendent methods gavé the same result was a strong argument for its
validity. The strongest evidence carne from Moseley's work on x-ray
spectra, but before that work could be done, sorne additional information
was needed on the optical properties of x-rays.
4-3 The diffraction of x-rays and Bragg's law. Early·work on x-rays

indicated that they are electromagnetic radiations with wavelengths
several thousand times smaller than those of visible light. Hence, the
methods of ordinary spectroscopy could not be used to measure the wave-
length or frequencies of x-rays. In 1912, von Laue and bis co-workers
discovered that crystals act as gratings for diffracting x-rays. Diffraction
occurs because ordinary x-rays have wavelengths between 10-8 and 10-9
cm, while the average distance between atoms in a solid is between 10-7
and 10-s cm. Furthermore, in a crystal there must be sorne atomic or
molecular unit arranged in a regular repeating order which results in the ob-
served crystal symmetty. The properties of x-rays and crystah! result in
conditions analogous to those which occur when visible light traverses an
optical grating-regularly spaced discontinuities separated by distances
several times the wavelength of the incident radiation. It is possible,
therefore, to use the diffraction of x-rays by a crystal to make quantita-
tive measurements of the wavelengths of x-rays. The way in which this
could be done was shown by Bragg.
Following Bragg,< 4 • 5> suppose that a train of monochromatic x-ray
waves strikes a crystal which consists of a regular arrangement of atoms
1
4--3} THE DIFFRACTION OF X-RAYS AND BRAGG S LA W 73
F10. 4--3. The reflection of x-rays by crystal planes, illustrating the deriva-
tion of the Bragg equation.
or molecules. The crystal acta as a series oí parallel reftecting planes, as

in Fig. 4-3. If the incident or glancing angle 8 has the appropriate rela-
tionship to the distance d between the reftecting planes and to the wave-
length X of the incident waves, the reftections from the various planes
reinforce each other and the resulting reftection is exceptionally strong.
The reflections are then said to be in phase. But if the angle 8 does not
satisfy the condition for the different reftections to be in phase, the latter
interfere with one another and the resulting beam is weak. As the angle
of incidence is changed, a series of reftections is seen which show alternate
maxima and mínima of intensity, and the diffraction caused by the reftec-
tion of x-rays from crystal pla.nes becomes apparent.
The condition for obtainfüg reftection maxima for x-rays, the Bragg
equation, can be derived by referring to Fig. 4--3. Consider a ray which
meets two successive crystal planes at 0 1 and 0 2 respectively, and let a
line drawn from 0 2 perpendicular to the planes cut the first plane at O.
Draw 0 1 P 1 and 0 2 P 2 representing rays reflected from the two planes, and
draw aOb perpendicular to 0 1P 1 and 0 2P 2 to represent a wave front of
the reflected beam. There will be reinforcement if the path 0 10 2b taken
by waves scattered at 0 2 is longer than the path 0 1a for waves scattered
at 0 1 by an integral number of wavelengths. Let Oc be perpendicular
to 0 10 2 ; then 0 1a = 0 1c and the difference in path is c0 2 b = 2d sin 8.
The condition that there be a reinforced reflected beam is then
nX = 2d sin 8, (4-10)
where n is an integer. If the difference in path length i~ equal to one

wavelength, then there will be a reflected beam at the position 81 which
satisfies the condition
X= 2d sin 81 . (4-11)
74 X-RAYS AND ATOMIC STRUCI'URF. [CHAP. 4
' X-rays- _
8 ____________
__ .1.
s2--..-1
S1
F10. 4--4. Diagram of a Bragg x-ray spectrometer.
This beam gives the reflection or spectrum of the first arder. If the differ-
ence in path length is 2, 3, ... , i wavelengths, then reflections of the
second, third, ... , ith orders will be found at the positions 91, 92 , ••• , 9¡.
These considera.tions formed the basis for the design, by Bragg, <li,II>
of an x-ra.y spectrometer. A schema.tic dia.gra.m of this instrument is
shown in Fig. 4-4. The x-ra.ys from a tube are collimated into a na.rrow
beam by the slits 8 1 and 8 2 cut in lea.el plates, a.nd then impinge on the
crystal e, which acts as a diffraction grating. The angular position of
e is rea.d by mea.ns of a vernier. After diffraction, the x-ra.ys enter a.n
ionization chamber D, filled with methyl iodide, which absorbs x-ra.ys
strongly. The electrometer E records the intensity of ionization in the
chamber.
The validity of Bragg's law, Eq. (4-10), was shown by a.n experiment
in which the gla.ncing a.ngle 9 was varied; the a.ngle between the ionization
chamber and the primary beam was kept equal to 29 in order to receive
the beam reflected from the crystal, and the ionization was measured as
a function of 9. The results of such a.n experiment are shown in Fig. 4-5.
It is seen that instead of va.rying uniformly with the gla.ncing a.ngle, the
ionization passes through peaks at certain sharply defined a.ngles; the
three peaks A 1 , B1, and C 1 represent x-ra.y spectrum lines. Secondary
peaks, A 2 , B 2 , and C2 , are seen at a.ngles whose sines are twice those of
the angles corresponding to the first order reflections. According to
Eq. (4-10), A1, B 1, and C 1 are lines of the first order spectrum, while
A 2 , B 2 , and C2 belong to the second order spectrum. Not only are the
angles at which the lines of the second order spectrum are seen just what
they should be according to Eq. (4-10), but their relative intensities also
are in the same ratio as those of the corresponding lines in the first order.
4--3] THE DIFFRACTION OF X-RAYS AND BRAGG'S LAW 75
,O 5 10 15 20 25 30 35
Glancing angle 8 in degrees
F10. 4-5. The first x-ray spectrum (Bragg, <6>).
The spectral lines are superimposed on a continuous spectrum represented

by the partially dotted line in the figure.
The x-ray spectra obtained in this way are characteristic of the target
from which the x-rays are emitted, as is shown by the fact that if the anti-
cathode in the x-ray tube is changed, an entirely different line spectrum
is obtained. On the other hand, if the crystal is changed but the target
is kept the same, the same lines appear with about the same relative
intensity, but the angles at which they appear are changed, showing that
the·grating spacing d between the layers of atoms is different for different
crystals.
When the grating spacing d is known accurately, the wavelengths of
the spectrum lines can be measured with good precision. The grating
spacing of the sodium chloride crystal can be calculated, because x-ray
studies of this crystal show that the sodium and chlorine atoms occupy
alternately the corners of the elementary cube of the lattice. Then, on
the average, to every atom there corresponda a volume d 3 with mass pd, 3 ,
where pis the density, equal to 2.165 gm/cm 3 • Also, one gram-mole of
NaCI weighs 58.454 gm and contains N O molecules or 2N O atoms, where
N O is the Avogadro number. The average mass per atom is then
58.454/2NO gm; and
a 58.454
or
d = 2(2.165)(6.025)10 23 '
and d = 2.820 X 10-s cm. The wavelengths of x-ray lines can now be
measured accurately with ruled gratings and, with the known wave-
lengths, the lattice spacing of crystals can be determined experimentally.
The best value obtained in this way for NaCI is 2.8197 X 10-s cm and is
76 X-RAYB AND ATOMIC STRUCTURE [CHAP.4
in good agreement with the calculated value. Crystals other than N aCl
can also be used and, in fact, calcite (CaCO 3) is now used as the standard
crystal; it has a lattice spacing of 3.03567 X 10-s cm.
4-4 Characteristic x-ray spectra. Moseley's law. The resulta dis-

cussed in the last section showed that the diffraction method could be ap-
plied to the analysis of the x-ray spectra of the elements, and the character-
istic radiations could be described quantitatively by the wavelengths
of the spectral linea. lt was soon found that these radiations are more
complex than was supposed from Barkla's absorption measurements,
but they could still be separated into a number of sharply defined linea.
The first systematic study of the x-ray spectra of the elements was made
by Moseley< 7> in 1913-1914. He used an x-ray spectrometer with a
crystal of potassium ferrocyanide, and detected the reflected x-rays with a
photographic plate rather than with an ionization chamber. Thirty-eight
different elements were used as targets in bis x-ray tube. Moeeley's
photographs showed that the spectral linea emitted by these elements
belonged to two different series which were identified with the K- and L-
types of characteristic radiation previously observed by Barkla. The
K-radiation was found to consist of two linea denoted by K,. and K¡¡; this
■
&;! 11
rnl■I
~■■■
~■■■
mlll
-■ m
11111~
11 ■ ~
-■~
■11~
Wavelength
Fm. 4--6. Positions oí the K,. and KfJ x-ray spectrum lines oí a sequence
oí elements as obtained by Moseley.
4--4) CHARACTERISTIC X-RAY SPECTRA. MOSELEY'S LAW 77
xesult can be seen from Fig. 4--6, which also indicates the relative positions
in which these lines appeared in the photographic plates, The lines are
those of the first sequence of elements studied by Moseley, the elements
from calcium to copper, and it is clear that the wavelengths decrease in
.a regular way as the atomic weight increases. The gap bet~een the cal-
cium and titanium lines represents the positions_of the lines of scandium,
which occurs between those two elements in the periodic system and on
which no measurements were made. ·
Moseley's quantitative analysis of sorne of his experimental data on the
K 0 -lines is shown in Table 4-1. The wavelengths of the K 0 -lines of a
sequence of elements are listed in the second column of Table 4-1. The
third column of the table gives the values of a quantity QK defined by the
relati•nship
(4-12)
where v is the frequency of the spectrum line, equal to e/X, and e is the
TABLE 4-1
MoSELEY'S MEASUREMENTS OF WAVELENGTHS OF THE Ka LINES
Wavelength, Atomic Atomic

Element QK = (11/¾110) 112
X 108 cm weight number, Z
Aluminum 8.364 12.05 26.97 13

Silicon 7.142 13.04 28.06 14
Chlorine 4.750 16.00 35.457 17
Potassium 3.759 17.98 39.096 19
Calcium 3.368 19.00 40.08 20
Titanium 2.758 20.99 47.90 22
Vanadium 2.519 21.96 50.95 23
Chromium 2.301 22.98 52.01 24
Manganese 2.111 23.99 54.93 25
!ron 1.946 24.99 55.85 26
Cobalt 1.798 26.00 58.94 27
Nickel 1.662 27.04 58.69 28
Copper 1.549 28.01 63.54 29
Zinc 1.445 29.01 65.38 30
Yttrium 0.838 38.1 88.92 39
Zirconium 0.794 39.1 91.22 40
Niobium 0.750 40.2 92.91 41
Molybdenum 0.721 41.2 95.95 42
Ruthenium 0.638 43.6 101.7 44
Palladium 0.584 45.6 106.7 46
Silver 0.560 46.6 107.88 47
velocity of light; v0 is a constant, which represents a certain fundamental

frequency, important in the study of line spectra. If the frequency is
replaced by the wave number ii (the reciprocal of the wavelength, ii = 1/>.),
v 0 is replaced by another constant R, called the Rydberg constant, for which
Moseley used the value 109,720 cm- 1 • The quantity Qx then becomes
Qx {
= {íi/fR) 112 (4-13)
The Rydberg constant will be discussed at greater length in Chapter 7 and

the basis for Moseley's choice of the constant R will be shown. For the
present purposes it is enough to note that Qx is proportional to the square
root of the frequency of the K,. line.
Moseley saw that Qx increases by a constant amount in passing from one
element to the next when the elements are in their order in the ~riodic
system. Except in the case of nickel and cobalt, this order is the same as
the order of the atomic weights of the elements listed in Table 4-1.
Although, with this exception, the atomic weights increase, they do so
in a much less regular way than Qx. Moseley concluded that there is
in the atom a fundamental quantity which increases by regular steps
from one element to thé next. He considered that this quantity could
only be the positive charge on the nucleus. It will be remembered that
the conclusion had been drawn from a-particle and x-ray scattering that
the number of unit charges on the nucleus of an atom is approximately
half the atomic weight. Moseley then noted that atomic weights increase
by approximately two units, on the average, in going from element to
element and this suggested to him that the number of charges increase&
from atom to atom by a single electronic unit. Moseley concluded that
the experiments led to the view that the number of unit charges on the
nucleus is the same as the number of the place occupied by the element
in the periodic system, and that both of these numbers could be repre-
sented by a quantity which he called the atomic number.
Moseley extended his experiments as far as gold. In order to do so,
he worked with the lines of the L series because the wavelengths of the
K series of the heavier elements became too small for precise analysis.
His results for the L,.-lines of a series of elements are shown in Table 4-2.
In this case, Moseley used the quantity QL, defined by
(4-14)
instead of Qx, for reasons which will be explained in Chapter 7. The

same conclusions can be drawn from the variation of QL as were drawn
in the case of Qx, so that experimental results were available for 38 elements
between aluminum and gold. On the basis of these results, Moseley
assigned a value of the atomic number to each element from aluminum to
4-4] CHARACTERISTIC X-RA Y SPECTRA. MOSELEY'S LA W 79
TABLE 4-2
MosELEY's MEASUREMENTS OF WAVELENGTHS OF THE L., LINES
Wa velength, ! 1
Atomic Atomic
Element QL = (11/~110) 112
X 108 cm weight number, Z
Zirconium 6.091 32.8 91.22 40

Niobium 5.749 33.8 92.91 41
Molybdenum 5.423 34.8 95.95 42
Ruthenium 4.861 36.7 101.7 44
Rhodium 4.622 37.7 102.91 45
Palladium 4.385 38.7 106.7 46
Silver 4.170 39.6 107.88 47
Tin 3.619 42.6 118.70 50
Antimony 3.458 43.6 121.76 51
Lanthanum 2.676 49.5 138.92 57
Cerium 2.567 50.6 140.13 58
Praeseodymium 2.471 51.5 140.92 59
Neodymium 2.382 52.5 144.27 60
Samarium 2.208 54.5 150.43 62
Europium 2.130 55.5 152.0 63
Gadolinium 2.057 56.5 156.9 64
Holmium 1.914 58.6 164.94 66
Erbium 1.790 60.6 167.2 68
Tantalum 1.525 65.6 180.88 73
Tungsten 1.486 66.5 183.92 74
Osmium 1.asi1 68.5 190.2 76
lridium 1.354 69.6 193.1 77
Platinum 1.316 70.6 195.23 78
Gold 1.287 71.4 197.2 79
gold. Aluminum, the first element in the periodic system studíed by

Moseley, was assigned the atomic number 13, because 12 elements were
known to precede it in the system.
The order of the atomic numbers turned out to be the same as that of
the atomic weights except where the latter were known to disagree with
the order of the chemical properties of the elements. For example, cobalt
has a greater atomic weight than nickel, but its chemical properties are
such that it should precede níckel in the periodic system. The relative
wavelengths and values of QK (Table 4-1) show that cobalt should indeed
precede nickel in the periodic system in spite of the discrepancy in atomic
weights. When elements were skipped in Moseley's work, corresponding
gaps in the values of QK or QL were found. Known elements were found
to correspond with ali the numbers between 13 and 79 except for four
80 X·RAYS AND ATOMIC STRUCTURE [CHAP. 4
numbers, 43, 61, 72 and 75, which corresponded to still undiscovered

elements. These missing elements have since been discovered and iden-
tified by their x-ray spectra. In addition, the atomic numbers of copper,
silver, and platinum turned out to be 29, 47, and 78, respectively. As
mentioned in Section 3-3, Chadwick, a few years later, obtained directly
from a-scattering measurements the values 29.3 ± 0.5, 46.3 ± O. 7, and
77.4 ± 1, respectively.
It can be seen from Tables 4-1 and 4-2 that
112
QK = (11/¾11 0 ) = Z - I, (4-15a)
QL = (11/j-,;11 0 )
112
= Z - 7.4. (4-15b)
Thus, if t.he square root of the frequency, or the wave number, for either
the Ka or La line is plotted against the atomic number, a straight line should
result. The quantity QK, which is proportional to the square root of the
frequency of the Ka line, is plotted against the atomic number in Fig. 4-7.
The points, represented by the dots in the figure, do indeed fall on a straight
line. When QK is plotted against the atomic weight, the points (repre-
sented by crosses) show much greater scatter. Similar results are obtained
for the La lines in Fig. 4-8. The figures show in a striking way that the
Atomic weight (A)

20,:::....----,;40;..___ _____;;60,=-_ _ _so,.:--_ _ _1,oo
50
~ 20 • Qx vs. z
x Qx vs. A
1º1_0_ _ _ _20.....__ _ _.,..30'-----40~---~50

Atomic number (Z)
Fm. 4-7. Moseley's relationship between the atomic number and the square
root of the frequency of the Ka line.
4-4] CHARACTERISTIC X-RAY SPECTHA. MOSELEY'S LAW 81
correlation between the frequency of the characteristic x-ray lines and the
atomic number is much more satisfactory than that between the frequency
and the atomic weight.
The frequencies may also be expressed by the empirical formulas
Ka: v = (i\ - ; 2) Vo(Z - 1)2, (4-16a)
(4-16b)
in terms of the wave numbers,
ii = _!_) R(Z -
(_!_l2 - 22 1) 2
'
(4-17a)
(4-l 7b)
These relationships will be useful later for comparison with theory.

Moseley's work has been extended to other elements, and the methods
of x-ray spectroscopy have been improved. Additional types of char-
Atomic weight (A)
80 100 120 140 160 180
10.-----.-------,-----.---....---,------,
50 60 70 80 90
Atomic number (Z)
Fm. 4-8. Moseley's relationship between the atomic number and the square
root of the frequency of the La line.
acteristic radiation have been found, such as the M and N series of lines,
which are even softer than the L-radiations. The conclusions reached
by Moseley have been confirmed, and atomic numbers have been assigned
to all of the chemical elements. The atomic number of an element is
the same as the number of unit charges on the nucleus, and is equal to
the number of electrons in the atom; it is also the same as the ordinal
number of the element in the periodic system. The identity of the latter
number with the nuclear charge number and the number of electrons had
been suggested by van den Broek<s> but could not be considered to be
proven until the work of Moseley. The fact that the position of an ele-
ment in the periodic system is more directly related to the atomic number
than to the atomic weight shows that the chemical properties are directly
related to the nuclear charge, and are not determined by the magnitude of
the atomic weight, as was thought by Mendeléeff.
REFERENCES
GENERAL
A. H. CoMPTON and S. K. ALLISON, X-Rays in Them-y and Ezperiment, 2nd
ed. New York: Van Nostrand, 1935.
M. S1EGBAHN, The Spectroscopy of X-Rays. Oxford: Oxford University Press,
1925.
M. 81EGBAHN, Spektroacopie der Rontgemtrahlen, 2nd ed. Berlin: Julius
Springer, 1931.
G. L. CLARK, Applied X-Rays. 3rd ed. New York: McGraw-Hill, 1942.
F. K. RICHTMYER, E. H. KENNARD, and T. LAuRITSEN, lntroduction to
Modern Physics, 5th ed. New York: McGraw-Hill, 1955, Chapter 8.
PARTICULAR
l. A. H. CoMPTON and S. K. ALLISON, op. cit. gen ref., Chapter III.
2. C. G. BARKLA and C. A. SADLER, "The Absorption of Roentgen Rays,"
Phil. Mag. 17, 739 (1909).
3. C. G. BARKLA, "Note on the Energy of Scattered X-Radiation," Phil.
Mag., 21, 648 (1911).
4. W. L. BRAGG, "The Diffraction of Short Electromagnetic Waves by a
Crystal," Proc. Cambridge Phil. Soc., 17, 43 (1912).
5. W. L. BRAGG, The Cry1tallíne State: Vol. I, General Survey. London:
G. BELL, 1949, Chapter 3.
6. W. H. BRAGG and W. L. BRAGG, "The Reftection of X-Rays by Crystals,"
Proc. Roy. Soc. (London), A88, 428 (1913); A89, 246 (1914).
7. H. G. J. MosELEY, "The High Frequency Spectra. of the Elements," Phil.
lifag., 26, 1024 (1913); 27,703 (1914).
8. A. VAN DEN BRoEK, "Die Radioelemente, das periodische System und die
Konstitution der Atome," Physik. Z., 14, 32 (1913).
83
PROBLEMS
l. The intensity of a beam of x-rays is reduced to 90% of its initial value

in passing through a slab of beryllium 0.30 cm thick and of density 1.84 gm/cm 3 •
How many electrons are there per atom of beryllium (a) as calculated directly
from the attenuation, (b) as calculated from Barkla's relationship, Eq. (4-8)?
(c) What is the value of the absorption coefficient?
2. A beam of x-rays is incident on a sodium chloride crystal (lattice spacing
= 2.820 X 10-8 cm; 1 X 10-s cm = 1 angstrom). The first order Bragg
reflection is observed at a grazing angle of 8° 35'. What is the wavelength of
the x-ray? At what angles would the second and third order Bragg reflections
occur?
3. The grating spacings of calcite, quartz, and mica are 3.036 A, 4.255 A, and
9.963 A, respectively. At what grazing angle would the first reflection of a
beam of 1-A x-rays be observed in rock salt, calcite, quartz, and mica, re-
spectively?
4. In his experiments, Moseley used a crystal of potassium ferrocyanide having
a grating spacing of 8.408 A. Find the positions of the first reflections of the
K,. x-rays of chromium, manganese, iron, cobalt, nickel, copper, and zinc, respec-
tively.
5, Calculate the frequencies and the wave numbers of the Ka lines listed in
Table 4-1.
6. Technecium is one of the new elements discovered among the products
of the fission of uranium. The wavelength of the K,. x-ray line from this element
has been determined by measuring the gl~cing angle as 7.0° for first-order
Bragg reflection from a cleavage face of NaCI. What is the wave length of the
x-ray? What is the atomic number of technicium?
7. The K,. line for copper has a wavelength of 1.5412 A as measured with a
ruled grating. The first-order Bragg reflection from a cleavage fa.ce of NaCl
is found in an experiment atan angle of 15° 53'. What is the lattice spacing of
NaCl? Find Avogadro's number, given that the density of NaCl is 2.164 gm/cm 3
and the molecular weight is 58.454.
8. The wavelength of the L,. x-ray lines of Ag and Pt are 4.1538 A and
1.3216 A, respectively. An unknown substance emits L,. x-rays with a wave-
length of 0.966 A. With these data, and with the known atomic numbers 47
and 78 for Ag and Pt, respectively, determine the atomic number of the unknown
substance.
CHAPTER 5
THE QUANTUM THEORY OF RADIATION
S-1 The failure of classical physics to describe atomic phenomena.

In the first four chapters, the fundamental experiments bearing on the
problem of the nature of atoms were discussed. These experiments were
interpreted in terms of certain concepts which form the basis of present
ideas of atomic structure. In 1913, an atom was supposed to consist of
a positively charged nucleus about 10- 12 cm in radius, surrounded by
electrons distributed over a volume about 10-s cm in radius. The charge
on the nucleus was known to be Ze, where Z is a positive integer and e is
the magnitude of the electronic charge; the number of electrons was known
to be equal to Z because the atom is electrically neutral under normal
conditions. Moseley's work showed that the number Z is also equal to
the atomic number, which represents the position of the element in the
periodic system. In this system, the elements are ordered according to
their properties, so that the latter are directly related to the nuclear charge
and the number of electrons in the atom. To describe in detail how these
properties depend on the nuclear charge, a more detailed model of the
atom was needed, and it would be expected that such a model would
include a description of the way in which electrons are arranged in the
atom.
Attempts were made to construct a theoretical atomic model on the
basis of classical physics, but they ali failed, and it was realized eventually
that classical physics (Newtonian mechanics, Maxwellian electromagnetics,
and tLermodynamics) could not explain or describe atomic phenomena.
This failure may be illustrated by means of two elementary examples.
The first example is that of an atom consisting of stationary positive and
negative charges. Consider an atom consisting of a nucleus with two
positive elementary charges and with two electrons somewhere outside
the nucleus. If the electrons Jire each at a distance a from the nucleus,
and 2a from each other, the repulsive electrostatic force between the
electrons is e2 /4a 2 ; each electron is attracted to the nucleus by a force
equal to 2e 2 /a 2 , eight times as great as the repulsive force. Hence, the
electrons would fall into the nucleus and there would be no mechanical
stability. This exampJe is a special case of a theorem which states that
an electric charge cannot ·be in equilibrium, at rest, under the action of
electric forces alone; and the example can be extended to any atom.
Suppose next that the electrons revolve in sorne way about the nucleus,
and consider for simplicity the case of a hydrogen atom which has one
84
5-1) FAILURE OF CLASSICAL PHYSICS 85
electron and a nucleus with one positive charge. Assurne tbat the electron
revolves about the nucleus in a circular orbit of radius a, and that the
velocity of the electron is such that the attractive force between the
electron and the nucleus provides just the centripetal force required.
The systern, nucleus and electron, should then be mechanically stable.
The electron, however, is subject to a constant acceleration toward the
nucleus, and, according to electrornagnetic theory, the electron should
radiate energy. The energy of the system should then decrease; the elec-
tron should gradually spiral in toward the nucleus, ernitting radiation of
constantly increasing frequency, and should eventually fall into the nucleus.
These predictions of classical physics are in strong disagreernent with the
experimental facts. It had long been known that when the atoms of an
element are excited, in the case of hydrogen, for example, by passing an
electric discharge through the gas, and if the resulting light (radiation)
emitted by the element is dispersed by a prism, a line spectrum character-
istic of the elernent is observed. Hydrogen always gives a set of lines with
the sarne wavelengths, helium gives another set, sodiurn still another, and
so on, in contrast to the prediction of the ernission of radiation of increas-
ing frequency. Actually, there is no atomic catastrophe, and atoms seem
to last for a long time. Theory has again failed to account for the sta-
bility of atorns. Moreover, in spite of many atternpts, classical physics
could not account for the characteristic optical spectra of the elernents,
nor could it account for x-ray spectra.
After many attempts like those cited, it became apparent that the
description of atornic phenomena required a new kind of physical theory,
one which could describe atomic phenomena in terms of new, nonclassical
concepts. Bohr's quanturn theory of atomic structure was the first suc-
cessful atternpt to fill the need, and although it eventually was found to
be unsatisfactory, it prepared the way for the modern, more successful
theory of quantum mechanics. The Bohr theory was based on the quantum
theory of heat radiation introduced by Planck in 1901 and applied to
light by Einstein in 1905. Planck's quanturn theory lies at the foundation
of present ideas about atoms and nuclei, and it would be futile to try to
continue into atomic and nuclear physics without first having sorne grasp
of the basic ideas of this theory. The quanturn theory actually arose
frorn the failure of classical physics to explain sorne of the experimental
facts of thermal radiation. In particular, classical physics could not
explain the dependence of the intensity of the radíant energy ernitted by
a blackbody on the wavelength of the radiation. To explain this phe-
nornenon, ít was necessary to develop a theory for the ernission of radiation
which was based on a concept entirely opposed to the ideas of classical
physics. The new concept, that of quanta or discrete corpuscles of energy,
must be applied to ali problerns involvíng the ernission and absorption
86 THE QUANTUM THEORY OF RADIATION [CHAP. 5
of electromagnetic radiation; but these are the basic problems of atomic

and nuclear physics, and this is the reason for the importance of quantum
theory in these fields.
The concept of quanta of energy is so different from classical physical
ideas that only a careful study of the problem of thermal radiation can
show the need for a quantum theory, and make clear what is meant by
quanta of energy. The problem of thermal radiation will therefore be
considered in sorne detail.
5-2 Tbe emission and absorption of tbennal radiation. I t is well

known that a hot body emits radiation in the form of heat. Thermal
radiation consists of electromagnetic waves and differs from visible light
and x-rays in having longer wavelengths. Thus, radiations with wave-
lengths between 7000 X 10-s cm and 4000 X 10-s cm are usually con-
sidered to be light, because the human eye can see these wavelengths.
The length 10-s cm is called the angstrom unit, and is abbreviated as A.
W aves longer than 7000 A and shorter than 0.01 cm are the infrared or
heat waves¡ wavelengths between 4000 A and about 50 A constitute the
ultraviolet radiation, and those shorter than 50 A but longer than about
0.01 A are usually classified as x-rays. Gamma-rays have still shorter
wavelengths.
At any temperature, the emitted heat energy is distributed over a
continuous spectrum of wavelengths, and this spectral distribution changes
with temperature. At low temperatures, the rate of radiation is small
and the energy is chiefly of relatively long wavelength (infrared radiation).
At temperatures between 500 and 550ºC, bodies begin to radiate visible
light, which means that the distribution of energy among the different
wavelengths has shifted so that a large enough portion of the radiant
energy has wavelengths within the visible spectrum. As the temperature
rises, the fraction of visible radiation increases until at 3000ºC, approxi-
mately the temperature of an incandescent lamp filament, the radiation
contains enough of the shorter wavelengths so that the body appears
"white hot."
The consideration of measurements by Tyndall on the radiation from
hot platinum wires led Stefan (1879) to suggest an empirical rule that
can be written in the form
(5-1)
where W is the rate of emission of radiant energy per unit area and is
expressed in ergs per square centimeter per second, and T is the absolute
temperatµre in ºK; W is called the total emusive power, or total emittance.
The quantity e is called the emissivity of the surface and has a value between
zero and unity, depending on the nature of the surface; u is a constant
ó---2] EMISSION AND ABSORPTION OF THERMAL RADIATION 87
called the Stefan-Boltzmann constant. In 1884, Boltzmann derived Eq.

(5-1) from thermodynamics, and the equation is therefore known as the
Stefan-Boltzmann law. If a body of emissivity e, at the temperature T 1 ,
is surrounded by walls at the temperature T 2 , smaller than T 1 , the net
rate of loss of energy by the body is given by
(5-2)
Equation (5-2) is familiar because of its application to practica! problems
of heat transfer. In the late 19th and early 20th centuries, physicists
were more concerned with the way in which the emissive power of a body
varies with temperature and wavelength. Their interest was concentrated
on "explaining" or "under¡,tanding" the variation in terms of the funda-
mental concepts and theories of physics. By "explaining" or "under-
standing" is meant fitting the experimental information into the scheme
of existing physical theories or, if this cannot be done, modifying the
theories so that they include the experimental data. Although this interest
may not appear to be a matter of great practica! importance, the difliculties
to which it led gave birth to the greatest revolution in physical thought
during the 20th century.
Physicists chose as the subject of their studies the thermal radiation
from a blackbody. To understand what is meant by the term blackbody
we first define the absorptivity of a body as the fraction of the radiant
energy, incident on the surface of the body, which is absorbed. A black-
body is, by definition, one with absorptivity equal to unity, i.e., a body
which absorbs all of the radiant energy falling upon it. A very simple
relation has been shown to exist between the absorptivity of a body and
its total emissive power. This relation, called Kirchhoff's law, states
that the ratio of the emissive power to the absorptivity is the same for
all bodies at the same temperature, and is equal to the emissive power
of a blackbody at this temperature; this law holds for each wavelength,
and has been confirmed by experiment. It follows from Kirchhoff's law
that no body can emit radiant energy at a greater rate than a black-
body, for the maximum value of the absorptivity, namely that for a black-
body, is unity, and any smaller value of the absorptivity necessarily implies
a smaller value of the emissive power. Hence, the blackbody, which is
the most eflicient absorber of radiant energy, is also the most efficient
emitter. The emissivity of a blackbody,-the quantity e in Eq. (5-1), is
equal to unity, and the total e:nissive power of a blackbody depends only
on the temperature and not on the nature of the body. The spectral
distribution of the energy radiated by a blackbody has, therefore, a special
interest.
A blackbody does not really exist in nature, but sorne substances such
as lampblack, flat black lacquer, rough steel plate, or asbestos board reflect
only a few percent of the incident radiation and approximate a blackbody.

It has also been shown that the radiation coming out of the small opening
of an almost completely closed, uniformly heated, hollow enclosure, or
cavity, is a good substitute for the radiation of a blackbody. As a source
of blackbody radiation, an electric furnace may be used, consisting of a
long tube, preferably with blackened walls, heated by an electric current
flowing in a wire wound around the tube. The temperature of the central
part of the interior is measured with a thermometer of some kind. A
small hole is made through the wall and the radiation coming through the
hole is observed. The spectral distribution of the radiation can be analyzed
by means of an optical spectrometer with the photographic plate replaced
by some energy-measuring device. The radiations from the blackbody
are refracted by a salt prism which may be rock salt, quartz, or fluorite.
Salt is used instead of glass because glass absorba too much infrared radia-
tion. The energy-measuring device may be a linear-type thermopile
connected to a galvanometer with high voltage sensitivity. The radiation
falling on one junction of the thermopile causes the temperature at that
junction to rise above that of the second junction, and the resulting electric
current causes a deflection of the galvanometer needle. The deflection
caused by tbe absorption of radiation in a given range of wavelength
dependa on the emissive power of the radiator in that range. Hence, the
emissive power can be measured as a function of wavelengtb. Another
useful energy-measuring device is the line bolometer, which consista of a
blackened, very thin metal strip with electrical connections. This strip
is the receiver of the radiations, and is connected as one arm of a balanced
Wheatstone bridge. Radiation falling on the strip raises its temperature,
causing a change in electrical resistance, as measured by a sensitive bridge
galvanometer. The bolometer can detect very small changes in tempera-
ture and, therefore, very small differences in radiant energy, so that it is
a sensitive and useful instrument for measuring the spectral distribution
of thermal radiation.
When the emissive power of a heated body is measured as a function of
wavelength, at a fixed temperature, a curve like that in Fig. 5-1 is ob-
tained. The abscissa is the wavelength in microns {l micron = 10-• cm).
The ordinates are relative values of the monochromatic emisswe power, or
spectral, emillance. This quantity, denoted by WA, is the emissive power
per unit range of wavelength, or the radiant energy emitted per unit area
per unit time in the range of wavelengths between >. and >. + d>.. The
monochromatic emissive power is related to the total emissive power by
the equation
w= ¡; WAd>.. {5-3)
It is seen from the figure that for very short and also for very long wave-
5-3] THE CLASSICAL THEORY OF THERMAL RADIATION 89
E
40r-----r------,---~ 140
....- - - - - - - 1 6 4 6 °
130
120
~110
30
-~ 100
'i
·a .," 90
".,
.::
> 80
.::
..s
_; 70 1449°
:'$ 20
_; >, 60
>, .,...
~
___...1259•
"'.,... s:: 50
s::
"" 10 "" 40
30
20
10
0 0._____.___...____,2'-------'3µ
1 2 3 4 5 6µ
Wavelength (microns) Wavelength (microns)
Fm. 5-1. The distribution of energy Fm. 5-2. The distribution of energy
in the spectrum of a tungsten lamp. in the spectrum of a blackbody at
(By permission from F. K. Richtmyer different temperatures.
and E. H. Kennard, Introduction to
Modern Physics, 4th ed. New York:
McGraw-Hill, 1947.)
lengths the monochromatic emissive power is very small. In the inter-

mediate wavelength range, it has a maximum value at a definite wave-
length ~m- Measurements made at different temperatures by Lummer
and Pringsheim {1899) yielded a set of curves like that of Fig. 5-2. The
curves are similar in shape, but as the temperature increases, the height
of the maximum increases and the position of the maximum is shifted in
the direction of smaller wavelengths or greater frequencies. The total
radiant energy for a given temperature is represented by the area between
the curve and the horizontal axis. This area increases as the fourth power
of the absolute temperature, according to the Stefan-Boltzmann law.
The experimental results contained in the curves of Fig. 5-2 are typical
of the phenomena which must be explained by theory.
5-3 The classical theory of thermal radiation. Wien (1893) showed

that it was possible to predict sorne features of the spectral distribution of
thermal radiation from the laws of classical physics. His results will be
discussed in sorne detail in order to understand the difficulties which arose
and which could only be resolved by the quantum theory.
In the theoretical treatment of the problem of radiation it is convenient

to use a quantíty uA, called the monochromatic energy density, rather
than the monochromatic emissive power WA. The energy density repre-
sents, not the energ.v falling on one square centimeter per second, but
the energy contained in a cubic centimeter of volume. The energy density
and the emissive power are related geometrically in a simple way. If
radiatíon is propagated in a definite direction, the amount of it which will
strike unit cross section per second is the amount contained in a prism
whose base is 1 cm 2 in arca and whose slant height along the direction of
propagation is the velocity of the radiation c. This amount is the volume
of the prism (ecos 8, if () is the angle between the direction of propagation
and the normal to the surface) multiplied by the energy of the radiation
per unit volume. Hence, if the energy density is known, the amount of
light falling on 1 cm 2 /sec can be found by integrating over ali directions
of propagation. The amount of radiation of a given wavelength falling
on 1 cm 2 of a body per second at a given temperature is equal to the
amount which 1 cm 2 of surface of the body would radia te in one second
at the same temperature (Kirchhoff's law). Since this amount is just
the monochromatic emissive power, the latter is proportional to the
energy density; the proportionality factor contains only geometrical
factors and the velocity of light, neither of which depends on wavelength
or temperature. We shall not need this relatíonshíp and shall therefore
not show it explicitly; it will suffice to remember that the monochromatic
energy density and the monochromatic emissive power can be used inter-
changeably so long as absolute values are not required. The energy
density can also be written as Up rather than UA to show that it can be
expressed in terms of the frequency rather than of the wavelength. By
definition, u,,dv is the radíant energy per unit volume in the frequency
range from II to JI + dv; uAdX is the radiant energy per unit volume in the
+
range of wavelength from X to X d>.. Frequency and wavelength are
related by the expression
11>. = e,
where e is the velocity of light. Now, u,,dv must be equal in magnitude
to 1:tAdX, since each frequency corresponds to one wavelength, and
Up = UA l!~I, (5-4a)
where the bars indicate the absolute magnitude of the quantity. The
absolute magnitude must be used because the cncrgy dcnsity is positive.
From 11>. = e, /d>./dv/ = >. 2 /e, and
(5-4b)
5-3] THE OLASSICAL THEORY OF THERMAL RADIATION 91
According to Maxwell's electromagnetic theory of light, radiation in

an enclosure exerts a pressure on the walls of the enclosure which is pro-
portional to the energy density of the radiation. Wien, therefore, treated
the radiation as a thermodynamic engine to which the first and second
laws of thermodynamics could be applied. He considered radiation of
a single wavelength and analyzed the problem of what would happen to
this radiation in an adiabatic expansion of the enclosure that it occupied.
He found that the wavelength and temperature before and after the
adiabatic change would be connected by the relationship
T Xo (5-5)
To= X'
where T 0 , X0 are the values of the temperature and wavelength of the
radiation before the adíabatic change, and T, X are the values after the
change. Equation (5-5) may be written in the more general form
XT = constant. (5-6)
This equation expresses the fact that if radiation of a particular wave-
length whose intensity corresponds to a definite temperature is changed
adiabatically to another . wavelength, then the absolute temperature
changes in the inverse ratio. Wien also obtained another important
result, namely, that the monochromatic energy density and temperature
before and after the adiabatic change are connected by the relationship
UA T5 (5-7)
UAo = T~,
where the subscript zero again refers to conditíons before the change.
Equation (5-7) may also be written in the more general form
UA
T" = constant, (5-8a)
or, with the aid of Eq. (5-6),

uAX 5 = constant. (5-8b)
In order that Eqs. (5-6) and (5-8) hold simultaneously, uAX 5 or uA/T 5
must be a function of the product XT:
{5-9)
or,
(5-10)
where C is a constant and the function f(XT) is, as yet, undetermined.

92 THE QUA?t.'TUM: THEORY OF RADIATION [CHAP. 5
TABLE 5-1 *
EXPERIMENTAL VERIFICATION OF WIEN'S DISPLACEMENT LAw
Wavelength Maximum
Observed
for maximum 1 relative X,,.T, UA.,.T- 5,
temperature,
energy density, energy density, cm-deg X 10 17
T: deg. K
X.. X 104 cm U>,,,,
621.2 4.53 2.026 0.2814 2190

723 4.08 4.28 0.2950 2166
908.5 3.28 13.66 0.2980 2208
998.5 2.96 21.50 0.2956 2166
1094.5 2.71 34.0 0.2966 2164
1259.0 2.35 68.8 0.2959 2176
1460.4 2.04 145.00 0.2979 2184
1646 1.78 270.6 0.2928 2246
* From Preston, Thwry of Heat.
Equations (5-6) and (5-7), or (5-10), constitute Wien's displacement

law. The rea.son for this name is that if the change from one temperature
to a higher one is supposed to take place by means of an adiabatic com-
pression, then, by Wien's law, the new distribution curve (cf. Fig. 5-2)
is obtained from the first one by displacing each abscissa toward the origin
in the ratio of distances, >../>.. 0 = T 0 /T, and increa.sing the ordinate in the
ratio T 5 /T~. Thus, to a ma.ximum (WA,,.)o or (u,.,,.) 0 in the first curve
corresponds the maximum ordinate uA,,. of the second; and if >..,,. is the
abscissa of a ma.ximum ordinate,
>..,,.T = constant = a, UAm! T 5 = constant = b. (5-11)
These predictions can be compared with the experimental data oí

Lummer and Pringsheim, shown in Fig. 5-2. The resulta oí the com-
parison are listed in Table 5-1; the fourth and fifth columns oí the table
give the experimental value of the producta >..,,.T and u-,.,,.T- 5 , respectively.
It is evident that there is no significant variation in either product over
a wide range of values of the different quantities, and Wien's displacement
law holds. The average value of the product obtained by Lummer and
Pringsheim wa.s 0.2940 cm-deg. More recent and more precise experi-
menta give for the best value of this quantity, the Wien displacement
constant, the value 0.2884 ± 0.0006 cm-deg. The relation (5-10) has
also been verified by the experimenta oí Lummer and Pringsheim. Since,
from Eq. (5-6), >.. = constant/T, Eq. (5-10) can be written a.s
u-,./T5 = C'J(>..T), (5-12)
5-4] FAILURE OF CLASSICAL THEORY OF THERMAL RADIATION 93
• =T=1646ºK
x =T =1449ºK
º" T=l259ºK
>,.T
F10. 5-3. The experimental verification of the displacement law for black-
body radiation. (By permission from F. K. Richtmyer and E. H. Kennard,
Introduction to Modern Phyaica, 4th ed. New York: McGraw-Hill, 1947.)
where C' is a new consta.nt. If the experimental values of the qua.ntity

u>./T 5 a.re plotted a.ga.inst the values of the product >.T, a single curve
should result in a.ccorda.nce with the functional relationship (5-12}; in
Fig. ó-3 is shown the composite curve obtained by combining the data
taken at three different temperatures. The theoretical prediction is seen
to be fulfilled, a.nd a single curve serves to represent the spectral dis-
tribution of blackbody radiation at ali tempe:ratures. The problem of
the spectral distribution of blackbody radiation has therefore been reduced
to the determination of the single function/(>.T) in Eq. (5-12)-the func-
tion that fixes the shape of the spectral distribution curve.
5-4 The failure of the classical theory of thennal radiation. The

derivation of Wien's law was be.sed on thermodynamics and did not involve
the nature of the radiator or the mechanism of emission of thermal radia-
tion. Consequently, Wien's law gives the general dependence of the
monochromatic energy density on wavelength and temperature. The
detailed, explicit form of the dependence is a.ffected by the choice of a
model for the radiator. This choice is arbitrary, a.nd any model can be
postulated which appears rea.sonable, lea.da to a distribution law of the
general form of Eq. (ó-10) or Eq. {ó-12), a.nd a.grees with experiment.
The first attempt to determine the funetion /(>.T) was by Wien (1896),
who assumed that the radiation was produced by emitters or oscillators
of molecular size, that the frequency of the radiation was proportional to

the kinetic energy of the oscillator, and that the intensity in any particular
wavelength range was proportional to the number of oscillators with the
requisite energy. With these assumptions, Wien derived the following
expression for the distribution:
(5-13)
where c1 and c2 are constants. This distribution has the required general
form and was found to fit the experimental curves quite well at short
wavelengths, say from 1 to 3 X 10-4 cm, but at longer wavelengths it
predicts values oí the energy density which are too small. The distribu-
tion formula also fails at high temperatures because uA (and WA) should
increase beyond all limits when the tempera.ture increases to an unlimited
extent, which is not the case in Eq. (5-13). The exponential function,
however, is so satisfactory at short wavelengths that any successful radia-
tion law must reduce approximately to such a function in this region of
the spectrum.
Another attempt to obtain a distribution law resulted in the Rayleigh-
Jeans formula. This formula was derived in a much more general way
than Wien's expression, and was shown to be a necessary consequence of
classical dynamics and statistics. lt was, therefore, just about the best
that classical physics could do. A brief outline of the ideas involved
follows. Consider a hollow cavity with perfectly reftecting walls and,
for convenience, assume that the cavity is rectangular. In such a cavity
there can be standing electromagnetic waves; these waves, in fact, con-
stitute the thermal radiation. If the average energy carried by each
wave is known, and if the number of standing waves with frequency
between 11 and 11 + dv can be calculated, the energy density of the radia-
tion can be obtained. The number of waves has been calculated and is
equal to
2
81r11 dv
N(11)d11 = - -·
ca (5-14)
The average energy cai:ried by each wave can also be calculated. Each
standing wave of radiation may be considered to be caused by an electric
dipole acting as a linear harmonic oscillator with a frequency 11. The
thermal radiation consists, then, of the electromagnetic waves emitted
by a large number of such oscillators. The number of these oscillators
with frequency between II and 11 + dv is given by Eq. (5-14) and each oscil-
lator has an energy E which may take on any value between O and oo.
There is, however, a certain average value E of the energy of the oscillator
that can be obtained from classical statistical mechanics. According to
FAILURE OF CLASSICAL THEORY OF THERMAL RADIATION 95
that theory, when the oscillators are at equilibrium, the value E for any
energy of the oscillator occurs with the ·relative probability e-•/kT, where
k is Boltzmann's constant. The average energy E is obtained by avera.ging
over ali values of E, with this weightfactor. Set ¡'j = 1/kT, for convenience;
then
The energy density of the radiation is, then,

81r1ikT
u~ dv = N(v)E dv = --c3- dv. (5-15)
Equation (5-15) is the Rayleigh-Jeans formula, and may also be written

in the forro
8trkT
U>.=X4· (5-16)
The Rayleigh-Jeans formula a.grees well with the experimental intensity

distribution for long wavelengths; in this region the intensity of the radia-
tion increases with the square of the frequency, as in Eq. (5-15). For
large values of the frequency, i.e., small values of >., however, the formula
fails. According to Eq. (5-16), the energy radiated by a blackbody in a
given range of wavelength increases rapidly as >. decreases, and approaches
infinity as the wavelengtb becomes very small. The experimentally
observed radiation curve is in complete disagreement with this conclusion,
since for very small wavelengths the energy density (or monochromatic
emissive power) actually becomes vanishingly small. Furthermore, the
energy carried by ali wavelengtbs would be
r
lo u,. d>.
r s1rkT
= lo X4 d>.,
and the integral is infinite for any value of T other than T = O, wbicb
would mean that the total energy radiated per unit time per unit area
is infinite at all finite temperatures. This conclusion is false, beca.use
the total energy radiated at any temperature is actually finite. Hence,
the Rayleigh-Jeans formula also fails to account for the observed depend-
ence of radiation on temperature.
5-5 Planck's quantum theory of thermal radiation. The problem of

the spectral distribution of thermal radiation was solved by Planck in
1901 by meaos of a revolutionary hypothesis. Planck postulated that a
linear harmonic oscillator like that discussed in the last section does not
have an energy that can take on any value from zero to infinity, but can
only take on values equal to O or Eo or 2E 0 or 3E 0 , ••• , or m: 0 , where Eo is a
discrete, finite amount, or quantum, of energy and n is an integer. The
average energy of an oscillator is obtained in an analogous way to that
used in deriving the Rayleigh-Jeans law except that sums are used instead
of integrals. The average value is now
_ "'"'
L..n-o nEoe -lln•o d l ~"' -fJn•o d
__ log - -I- -
E= =--ogL..Je =
1::-0 e-lln•o d{J n-o d{J I - e-fJ•o
Eo
e•olkT _ l
The energy density is then, with the aid of Eq. (5-14),

2
u,, dv = N(v)'i dv = -81r11
e3
-
Eo
----"--
e•olkT _
dv, (5-17)
1
or
81rv 2 Eo
u --- (5-18)
" - eª eEolkT - I
This formula must have the same general formas Wien's law, Eq. (5-10),
and the temperature must appear in the combination >..T, or T /v, or v/T.
Hence, Eo must be proportional to v, or
Eo = kv, (5-19)
where h is a new universal constant, called Planck's constant. Planck's

distribution law for thermal radiation is therefore
. 81rhv 3 1
u,, = ~ (el»/kT -,- 1) ' (5-20)
or, in terms of wavelength,

81rhc 1
UA = ---¡¡- (e"ctkAT _ 1)
(5-21)
It is interesting to compare Planck's radiation law with the Wien dis-

tribution Eq. (5-13) and the Rayleigh-Jeans formula Eq. (5-15). At low
5-5] PLANCK'S QUANTUM THEORY OF THERMAL RADIATION 97
frequencies, hv/kT « 1, and the exponential function in the denominator

of Eq. (5-20) can be expanded. Then
3 2
u = 81rhv l = 81rv kT + ...
• es [l + (h~/kT) + · · ·] - 1 c3
and at low frequencies (long wavelengths) the Planck formula reduces

to the Rayleigh-Jeans formula (5-15), known to be valid at long wave-
lengths. At frequencies such that hv/kT » 1, the 1 in the denominator
of Eq. (5-21) can be neglected in comparison with the exponential, and
UA = 81rhc
X5e -hc/kAT ,
which is Wien's formula (5-13), known to be valid at short wavelengths.
Thus, Planck's radiation formula reduces to forros that agree with experi-
ment at the extremes of the wavelength scale. In the intermediate wave-
length range, the Planck distribution gives a maximum value for the
energy density. The position of the maximum can be found from the
condition duA/d>. = O, with ui. given by Eq. (5-21). Omitting the details
of the calculations, we get the result
ch 1
>..mT =k 4.965 . (5-22)
In Eq. (5-22), e is the velocity of light, k is the Boltzmann constant, and h

is Planck's constant. Hence, the product >..mT is a constant, in agreement
with experiment.
lt can also be shown that Planck's radiation law leads to the Stefan-
Boltzmann law. The total energy density is given by
u= ("' uAd>..
JO
= 81rch ("'
Jo (ec
11/kA;
- 1) "
~~ ·
This integral can be evaluated, and the resulting expression for u is
u= aT4, (5-23)
the familiar fourth power law, with
8 1r:;k4
(5-24)
ª= 15 c3 h 3 ·
Equations (5-22) and (5-24) are two relations containing the three con-
stants, e, h, and k. In 1901, the value of e was known quite well; also,
the experimental value of the product >....,T was known to be 0.294 from
the experimenta of Lummer and Pringsheim (Table 5-1). The value of
the constant a was known from experimental studies of the Stefan-

Boltzmann law, a = 7.061 X 10- 15 erg/cm 3 (ºK) 4 . With the known
values of a, >-mT, and e, Planck obtained
h = 6.55 X 10-27 erg-sec,
and
k = 1.346 X 10- 15 erg/deg.
This was the first calculation of Planck's constant, and the best, up to
that time, of Boltzmann's constant. More recent and more precise ex-
perimental determinations of h and k by a variety of methods give
h = (6.62517 ± 0.00023) X 10-27 erg-sec,
k = (1.38042 ± 0.00007) X 10- 16 erg/deg.
When the above values are inserted into Eq. (5-20) or Eq. (5-21), the
shape of the spectral distribution curve is completely determined. Planck's
distribution law can then be compared directly with experiment; this is
done in Fig. 5-4 for the case of blackbody radiation at 1600°K. The points
are the experimental resulta for the energy distribution at different wave-
lengths, and the solid line is the distribution according to the Planck
radiation law; the agreement is excellent over the en tire wavelength range
1
1
1
1
1
\ Rayleigh-
\Jean's law
\
1
\
\
\
\
\
1Vx ' \
\
'\
\
'' \
\
'
o 2 3 4 5 6
Wavelength (microns)
Fm. 5-4.. A comparison of the Planck, Wien, and Rayleigh-Jeans radiation

laws with experiment at 1600°K. The ordinate represents Wi. on an arbitrary
scale, and the circles represent experimental points. (By permission from
F. K. Richtmyer and E. H. Kennard, Introduction to Modern Phy3ic&, 4th
ed. New York: McGraw-Hill, 1947.)
5-6] QUANTUM THEORY AND THE PHOTOELECTRIC EFFECT 99
of these experiments. For comparison, the theoretical predictions of the

Wien distribution (5-13) and the Rayleigh-Jeans formula (5-16) are also
included. These results are typical of those found over a wide range of
temperatures, and there is no doubt that the Planck radiation law is the
only one which satisfactorily accounts for the experiments.
The magnitude of the energy associated with a quantum of radiation
can be obtained as a function of wavelength. Equation (5-19) can also
be written
he
Eo = ~• (5-25)
Then Eo is given, in ergs, by
6.62 X 10-27 X 3.00 X 10 10
Eo (ergs ) = X
1.99 X 10- 16
=---
;\
--
where X is expressed in centimeters. For a gamma-ray with a wavelength

of 10- 10 cm, the quantum is 1.99 X 10- 5 erg. For a visible light quantum
with a wavelength of 5000 A= 5 X 10-5 cm, the energy is 4.0 X 10- 12
erg. For a short radio wave, say X = 10 cm, the quantum is 1.99 X 10- 17
erg, and so on.
The revolutionary nature of Planck's theory is contained in the postulate
that the energy of an oscillator can vary only in discrete jumps. If this
is so, then the emission and absorption of radiation must be discontinuous
processes. Emission can take place only when an oscillator makes a
discontinuous transition from a state in which it has one particular en-
ergy to another state in which it has another energy different from the
first by an amount which is an integral multiple of hv, where v is the
frequency of the emitted radiation. We are thus led to the idea that an
oscillator, or any other physica.l system capable of emitting electromagnetic
radiation, has, in general, a discrete set of possible energy values or levels;
energies intermediate between these allowed values never occur. The
idea of discrete energy levels will be seen to be a fundamental one in a.tomic
and nuclear physics, and is a consequence of Planck's postula.te of qua.nta.
of energy. This idea is not an outgrowth of classical physics, but repre-
sents a radical, empirical modification of classical ideas, needed in order
to bring theory and experiment into agreement.
5-6 Quantum theory and tbe photoelectric effect. The concept of

discrete energy states and quanta of energy was so different from the
classical ideas to which physicists were accustomed that Planck's theory
met with strong opposition. The theory was successful, however, not
only in explaining the phenomena of thermal radiation, b11t also in the

solution of other problems for which classical pbysics failed. One of
these problems was that of the photoelectric effect.
Hertz (1887) diseovered that a metallic surface can emit electricity
when light of very short wavelength falls on it, and in 1898 Thomson
showed that the e/m value of the emitted particles was the same as that
for cathode rays. It is now known, of course, that the particles are electrons.
The photoelectric effect was studied very carefully and the most im-
portant experimental results may be summarized as follows:
l. lf light of a given frequency can liberate electrons from a surface,
the electron current is proportional to tbe intensity of the light.
2. For a given metallic surface, there is a smallest value of tbe frequency
for which the incident light can liberate electrons; light of smaller frequency
cannot eject electrons no matter how long it falls on the surface or how
great is its intensity.
3. Light of frequency greater than this critica! value causes the imme-
diate emission of electrons; the time interval between the incidence of
the light on the metallic surface and the appearance of electrons is not
more than 3 X 10-9 sec.
4. The maximum kinetic energy of the emitted electrons is a linear
function of the frequency of the light which causes their emiasion, and is
independent of the intensity of the incident light.
These experimental facts could not be explained on the ha.sis of the
classical electromagnetic wave theory of light. There was no way in
which a train of light we.ves spread out over a large number of atoms
could, in a very short time interval, concentrate enough energy on one
electron to knock it out of the metal. Furthermore, the we.ve theory of
light was unable to account for the fact that the maximum energy of
tbe ejected electrons increases linearly with the frequency of the light
but is independent of the intensity. In 1905, Einstein proposed a mecha-
nism for the pbotoelectric effect, based on Planck's idea of quanta of energy,
that could account for the experiments. He assumed tha.t the energy
emitted by any radiator not only kept together in quanta as it traveled
through space, but tha.t a given source could emit and absorb radiant energy
only in units which are all exactly equal to lw. According to Einstein,
light itself consists of quanta or corpuseles of energy hv which move
through space with the velocity of light. This hypothesis is completely at
variance with the wave theory of light but, as will be seen, accounts suc-
cessfully for the photoelectric effect. With this hypothesis, Einstein
deduced bis pbotoelectric equation
½mv 2 = hv - A. (5-26)
This equation states that the maximum kinetic energy of the emitted
5-6] QUANTUM THEORY AND THE PHOTOELECTRIC EFFECT 101
electron is equal to the energy hv of the incident light quantum minus a

quantity A which represents the amount of work needed by the electron
to get free of the surface.
It is evident that Eq. (5-26) agrees qualitatively with the experimental
facts. If the frequency is so small that hv :5 A, no electrons can be emitted,
so that the frequency has a threshold value, for a given metal, below which
no photoelectric emission can occur. The energy of the electrons varies
linearly with the frequency and is independent of the intensity. If the
frequency is above the threshold value, a light quantum striking the metal
and colliding with one of its electrons can give up all of its energy to the
electron and knock it out of the metal. But before the electron emerges,
it loses a part A of this energy needed to remove it from the metal. The
number of electrons ejected is proportional to the number of incident light
quanta and therefore to the intensity of the light falling on the metal.
Although the Einstein photoelectric equation accounted for the experi-
menta qualitatively, it was not verified quantitatively until 1916 when
Millikan performed a very beautiful experiment. He studied the effect
of light of a range of frequencies on sodium, potassium, and lithium.
The energy of the ejected electrons was measured by applying to the
metallic surface a positive potential just strong enough to prevent any
of the electrons from carrying a eharge toan electrometer. If this potential
is equal to V volts, then the kinetic energy is given by
½mv 2 = Ve,
where e is the electronic eharge. Combining this relation with Eq. (5-25),
we get
Ve= hv - A,
or
V= (h/e) v - (Aje)· {5-27)
If the potential V is measured as a function of the frequency of the incident
light, the result should be a straight line with a slope whose magnitude
is h/e. The experimental resulta gave, in all cases, a good straight line,
as predicted by Einstein's theory. Since the value of e had already been
determined by Millikan, the value of h could be determined; it was found
to be in good agreement with the value obtained by Planck from the
measurements on blackbody radiation. Einstein's photoelectric law has
also been found to be valid for the electrons ejected by x-rays and 'Y-rays,
and there is no doubt of ita general validity.
A quantum of light or, more generally, of radiation is also called a
photon; the photon may be described as a "bundle" or "particle" of radia-
tion. Thus, a photon of radiation of frequency v carries an amount hv
of energy.
The quantum theory was also applied successfully by Einstein (1907)

to the problem of the specific heat of solids and polyatomic gases, where
again there was experimental information which could not be explained
by classical theory. These applications of the quantum theory by Einstein
helped establish the theory, aided its acceptance, and laid the ground-
work for the application of quantum theory to the problem of atomic
structure.
REFERENCES
GENERAL
M. PLANCK, Theory of Heat. London: Macmillan Co., 1929, Parts III, IV.
T. PRESTON, The Theory of Heat, 4th ed., J. R. Cotter, ed. London:
Macmillan Co., 1929, Chapter 6.
J. K. RoBERTs, Heat and Thermodynamics, 4th ed., revised by A. R. Miller.
Glasgow: Blackie; New York: Interscience, 1951, Chapters 20, 21.
A. SoMMERFELD, Thermodynamics and Statistical Mechanics. New York:
Academic Press, 1956, Sections 20, 35.
F. K. R1cHTMYER, E. H. KENNARD, and T. LAURITSEN, Introduction to Modern
Phyrics, 5th ed. New York: McGraw-Hill, 1955, Chapter 4.
G. P. HARNWELL and J. J. L1v1Noooo, Experimental Atomic Phyrics. New
York: McGraw-Hill, 1933, Chapter 2.
R. A. M1LLIKAN, Electrons, (+ and -), Protons, Photons, Mesotrons, and
Cosmic Rays. Chicago: University of Chica.go Press, 1947.
PROBLEMS
l. Calcula.te the total emissive power of a blackbody at the following temper-

atures: 0ºC, lOOºC, 300ºC, 500ºC, lOOOºC, 1500°C, 2000ºC, and 3000ºC. The
Stefan-Boltzmann constant u has the value 0.5669 X 10-4 erg-cm- 2-sec- 1
(ºK)-4. Give the results in terms of ergs and ca.lories.
2. Suppose that the blackbody of Problem 1 is surrounded by wa.lls kept at
a temperature of 0ºC. Calcula.te the net re.te of energy loss by the body to the
walls at ea.ch of the temperatures listed in Problem l.
· 3. Derive, from Planck's radia.tion law, the expression (5--22) for the wave-
length at which the radiation density has its maximum value. The best value
of the Wien displacement law constant is now thought to be 0.2884 cm-ºK.
Calcula.te the value of h. Use k = 1.3804 X 10- 16 erg/ºK and e = 2.99793
X 10 1º cm/sec.
4. Derive an expression for the total energy density in a blackbody enclosure
by integrating Planck's distribution law over ali possible values oí the frequency.
Note that
"'
1
3 •
X dx 11"
o ez - 1 = 15 ·
The best value for the total energy density is given by a = 7.563 X 10- 15
erg/cm 3 • Calcula.te the value of h¡ use the values of k ande given in Problem 3.
PROBLEMS 103
5. Calculate the frequency and the amount of energy associated with a

quantum of electromagnetic radiation at each of the following wavelengths:
1 km, l m, 1 cm, l mm, 10-4 cm, 5000 A, 1000 A, 1 A, 10-3 A, 10- 13 cm, 10-15
cm. Give the energy in ergs and in electron volts.
6. The photoelectric work functions for samples of sorne metals are Ag,
4.5 ev; Ba, 2.5 ev; Li, 2.3 ev; Pt, 4.1 ev; Ni, 5.0 ev. (a) What is the wavelength
at the threshold in each case? (b) What is the maximum velocity of the ejected
photoelectrons in each case, when the samples are illuminatéd with ultraviolet
light of wavelength 2000 A? (c) What is the stopping potential, in volts, for
the photoelectrons of maximum velocity?
7. The first precise determination of h was made by Millikan, who ~easured
the energy with which photoelectrons are ejected as a function of the frequency
of the incident light. He found that the potential V needed to keep the electrons
from sodium from reaching an electrometer increased with frequency by
4.124 X 10-15 voltjcycle/sec. What is the resulting value of h? Note that 300
volts == 1 esu of potential, and use e = 4.803 X 10-10 esu.
8. Show that the velocity of the fastest photoelectrons is related to the stop-
ping potential by the equation
llmax = 5.93 X 107 X vV,
where llmax is in centimeters pér second and V is in, volts.
9. Photoelectrons emitted from a photocell by light with a wavelength of
2500 A can be stopped by applying a potential of 2 volts to the collector. (a)
What is the work function (in ev) of the surface? (b) What is the maximum
wavelength that the light can have and eject any photoelectrons? (c) If an
electrometer is connected across the photocell, to what potential will it be charged
when ultraviolet light of wavelength 1000 A is used?
10. The emission of x-rays as a result of the incidence of high-speed ele'ctrons
on a metal surface, as in an x-ray tube, may be regarded as an in11erse photo-
electric ejfect. The maximum frequency of the emitted radiation is given by the
relation hvm = eV, where V is the anode voltage. Show that X... (A) = 12395/V
(volts). What are the wavelength, frequency, and energy (in ev) of the x-rays
from a Coolidge tube operated at 50,000 volts? At 105 volts? At 5 X 105 volts?
At 106 volts?
CHAPTER 6
THE SPECIAL THEORY OF RELATMTY
6-1 The role of the special theory of relativity in atomic and nuclear
physics. The second great theory of 20th century physics that is indis-
pensable to the development of atomic and nuclear physics is the special
theory of relativity proposed by Einstein in 1905. The first applications
of the relativity theory which will be met in atomic and nuclear physics
depend on two closely related ideas. One idea is that of the variation
of the mass of a particle with its velocity; the second is that of the pro-
portionality between mass and energy. The latter relationship, expressed
by the equation E= mc 2 , has achieved a certain degree of notoriety in
recent years and has even, on occasion, occupied positions of prominence
in the public press.
The mass-energy e4uation is used often in nuclear physics, and its use
should be accompanied by sorne understanding of its origin and meaning.
A real understanding of the relationship can come only from a careful
study of the relativity theory; but sorne familiarity with the ideas involved
in the mass-energy equation can be obtained from a short discussion of
the background of experiment and theory which led to relativity. Such
a discussion should dispel ~me of the aura of magic that surrounds the
relativity theory and even its simplest applications, and should allow the
student to use mass and energy as interchangeable quantities without
feeling uncomfortable.
The theory of relativity may be introduced in either of two ways. One
way is to examine very carefully what is meant by such concepta as space,
time, and simultaneity and, as a consequence of this examination, to
modify the commonly accepted meanings of these ideas. This logical
process, together with certain postulates, leads to the special theory of
relativity. The theory can then be applied to physical problems to see
how it meets the test óf comparison with experiment. The second method
is to consider certain problems which led to serious dilemmas in physical
thought at the end of the 19th century, and to see how these dilemmas
could be resolved only by the revolutionary ideas contained in the rela-
tivity theory. The second method is the one that will be adopted here
because it is more closely related to experiment and is less abstract than
the first method. Two problems will be considered¡ these are (1) the
ether and the problem of absolute velocity, (2) the problem of the in-
variant forro of physical theories.
104
6-2) THE ETHER AND THE PROBLE.M OF ABSOLUTE VELOCITY 105
6-2 The ether and the problem of absolute velocity. The interpre-
tation of the phenomena of the reflection, refraction, and diffraction of
light gave rise to the wave theory of light. The propagation of light
waves seemed to require a medium, just as water or sound waves cannot
be separated from their respective media. It was known from astro-
nomical data that light can pass through space practica.lly devoid of
matter. For these reasons, physicists falt it necessa.ry to postulate the
existence of a medium in which light waves could be propagated even
in the absence of matter. This medium, called the ether, was supposed
to be a weightless substance permeating the entire universe. Because of
the success of the wave theory of light in accounting for optical phenomena,
the ether gradually carne to be accepted as a physical reality. The as-
sumption that there is an ether raised problems in connection with the
motion of material bodies. The ether was assumed to be stationary, and
a body such as the earth was supposed to move through it without pro-
ducing any disturbance. These assumptions were not contradicted by
astronomical evidence, and it was inferred that it should be possible to
determine the absolute velocity of a body, i.e., its velocity relative to the
stationary ether. Since the ether was supposed to be the medium for
the propagation of light, it was thought that the measurement of absolute
velocity should depend on the effect of motion on some phenomena in-
volving light. Thus, if light and the earth both move with definite veloc-
ities relative to the stationary ether, it should be possible to devise an
experiment, involving the propagation of light, from which the absolute
velocity of the earth could be deduced. Many such experimenta were
proposed; the theoretical basis of the most famous one can easily be
derived.
Suppose that the earth travels through the stationary ether with a
velocity v, and that light waves travel through the ether with a velocity c.
Consider a source, fixed on the earth, from which light is emitted. If the
direction of propagation of the light is the sa.me as that of the earth's
motion, the velocity of the light relative to the earth will be e - v. If
the light travels in the direction exactly opposite to that of the earth's
motion, its velocity relative to the earth is e + v. The time taken for the
light to travel a certain distance L in the di:rection of the earth's motion
is L/(c - v), while the same distance in the opposite direction will require
+
the time L/(c v). Therefore, if a ray of light were to travel a distance
L in the direction of the earth's motion, and then be reflected back and
travel the sa.me distance in the opposite direction, the time required would
be
t-__!:____!!__2cL
11 - e + v + e - v - c2 - v2
106 THE SPECIAL THEORY OF RELATIVITY [CHAP. 6
Consider next a ray of light propagated in a direction perpendicular

to that of the earth's motion (Fig. 6-1). While this ray travels a distance
L, equal to AB, the earth has moved from A to A', and the point B has
moved to B'. The actual path of the light ray is AB'. If the time required
is t, then AB' = et. In the same time interval, A has moved to A' with
the velocity v, so that AA' = vt. In the right-angled triangle AA'B', we
have c2 t 2 = L 2 + v2t 2 , or t 2 (c 2 - v2 ) = L 2 , or t = L/..,/c 2 - v2 • The
same length of time will be needed for the return trip, i.e., to A". Hence,
the time required to travel a distance L and back in a direction perpen-
dicular to that of the earth's motion is
2L
tJ.=-;:===
..,/c2 - v2
The ratio of the times required is
(6-1)
It should, therefore, take longer for light to traverse the reflected path in
the direction parallel to that of the earth's motion than to traverse the
same distance in the direction perpendicular to that of the earth's motion.
From an experimental viewpoint, the difference between t 11 and tJ. is
important. This difference is
2
2
t 11 - tJ. = : [1 - tv 2 /c 2 ) - ..,/1 - \v2/c2)l (6- )
If we neglect terms of order higher than v2 /c 2 , we may write

1 v2
1- (v2/c2) = 1 + c2 + · · · •
2
1
--;:=::::::;::::::::::;::::==
..,/1 - (v2/c2)
= 1 + -21 c2v + ....
Hence,
t11 - t.L = 2L
e
(1 + c 1- !2 v2c + .. ·)
v2 -
2 2
L v2
= e' c2· (6-3)
The difference t11 - t.L and the ratio t11/tJ. both depend on the quantity
(v/c) 2 • The orbital velocity of the earth is 30 km/sec, so that v/c = 10-•
and (v/c) 2 = 10-8, and the effect should be difficult to detect.
6-3) THE MICHELSON-MORLEY EXPERIMENT 107
6-3 The Michelson-Morley experiment. The theoretical relations

derived in the last section were tested experimentally by Michelson and
Morley (1887) who used an ingenious instrument, the interferometer,
shown schematically in Fig. 6-2. In the Michelson-Morley experiment,
light from a source Q is divided into two perpendicular beams by a half-
silvered mirror P. One beam passes through P to a mirror S 1 and is
reflected back to P¡ the other beam is reflected by P to the mirror S2 and
is then reflected back to P. The two reflected beams pass through P to a
telescope or photographic plate F. If the distances PS 1 and PS 2 are
equal, and if one arm of the interferometer is parallel to the direction
of the earth's motion, the two light beams should arrive at F at slightly
different times, causing an interference pattern (interference fringes).
The position of the pattern should depend, according to Eq. (6-3), on the
absolute velocity of the earth and should be displaced from the position
corresponding to v = O by an amount proportional to (L/c)(v 2 /c 2 ). If the
interferometer is rotated through an angle of 90º, the arm which was
originally in the direction of the earth's motion is now perpendicular
thereto, and the interference pattern should be displaced to the opposite
side of the position corresponding to v = O. If the position of the pattern
is observed during the rotation of the interferometer, the fringes should be
shifted by an á.mount corresponding to the time interval 2(L/c)(v 2 /c 2 ).
The frequency of the light used in the experiment is 11, and its period
(inverse frequency) is T = 1/v. Then the ratio of the displacement to
the period is given by (2L/cT)(v 2/c 2). Since the wavelength >. = cT, the
ratio of displacement to wavelength is given by (2L/X)(v 2 /c 2 ).
In the experiment of Michelson and Morley, the path length L was
1.1 X 10 3 cm, as a result. of several reflections of the light beam. The
wavelength of the light used was about 5.9 X 10-5 cm, and v2/c 2 = 10-8 •
"------'------~-- V
A vt A' A"
Frn. 6-1. Basis of an experiment Frn. 6-2. The Michelson-Morley
on the velocity of the earth relative experiment: schematic diagram.
to the ether.
108 THE SPECIAL THEORY or RELATIVITY [CBAP.6
Hence, {2L/X)(v 2/c 2) = [2(1.1)10 3 X 10-8)/[5.9 X 10-111 = 0.37. Thus,

when the interferometer was rotated through 90°, the interference fringes
should ha.ve been displaced by more tha.n ½ of a. wa.velength. In the
experiment, a. displa.cement of one-hundredth of this a.mount could ha.ve
been observed, but no significa.nt displa.cement wa.s observed in the original
experiment, or in ma.ny la.ter repetitions. Tha.t is, there wa.s no significa.nt
difference between the ra.tes a.t which the light tra.veled in the two direc-
tions a.t right a.ngles to ea.ch other, a.nd the Michelson-Morley experiment
showed tha.t the ratio of the times of tra.vel in Eq. {6-1) is unity. The
simplest interpretation of this result seemed to be tha.t v, the velocity of
the ea.rth with respect to the ether, is zero, a. conclusion which led to serious
difficulties. m
The theoretical value of the ratio of displa.cement to wavelength wa.s
derived by a. simple a.nd stra.ightforwa.rd a.pplica.tion of some of the funda-
mental ideas of classica.l physics, a.nd this fa.et ma.de ít very difficult to
explain the nega.tive result of the Michelson-Morley experiment. The
simplest expla.na.tion which could be proposed was tha.t the ea.rth carries
the ether with it, in which case the velocity of the earth relative to the
ether should be zero. This expla.nation wa.s una.ttra.ctive beca.use every
moving body would presumably carry its own ether with it, a.nd the
uniqueness of "the sta.tíona.ry ether" as the medium for the propaga.tion
of light would disa.ppea.r.
There were a.lso experimental objections to the idea tha.t the ea.rth
carries the ether along with ít. Suppose that in spa.ce there is a.n observer
who is looking a.t a. light source on the earth, a.nd tha.t light waves from
this source are tra.nsmitted through the ether a.nd eventua.lly reach the
observer. If the earth, together with the ·ether, were moving towa.rd
the observer, the velocity of the light waves from the source would appear
to be greater tha.n if the earth were moving away from the observer. This
result follows from the general behavior of waves in a moving medium;
the velocity of the wa.ves relative to a stationary observer should be
equal to the sum of the velocity of the waves in tbe medium a.nd the
velocity of the medium. Thus, the velocity of light as mea.sured by the
assumcd observer would depend on the velocity of its source. This
conclusion is contradicted by all observations on the velocity of light,
especially by astronomical studies of double star systems a.nd stellar
aberration. The observed velocity of light is a.lwa.y~ the sa.me regardless
of whether or not the emitting source moves, or how it moves. The
assumption that the earth carries the ether a.long with it led, therefore, to
a serious contradiction with experiment, and had to be rejected.
Fitzgerald {1893) introduced a hypothesis to account for the negative
result of the Michelson-Morley experiment while preserving the idea of
the stationary ether. He assumed that when a b<><!y travels in a direction
6-3] THE MICHELSON-MORLEY EXPERIMENT 109
parallel to that of the earth's motion it contracts in length, becomi:ig

shorter by the factor v'l - (v 2 /c 2 ); he also assumed that no such con-
traction occurs in a direction perpendicular to that of the earth's motion.
According to this hypothesis, the arm of an interferometer that is parallel
to the direction of the earth's motion contracts in length from L to
Lv'l - (v 2/c 2), but the length of the perpendicular arm remains equal to L.
The light does not have to travel as far in the parallel direction as in the
perpendicular direction, and the difference is such as to make the theoretical
value of the shift of the interference pattern equal to zero. The contraction
in the length of one arm of the interferometer would not be observed,
because a measuring instrument placed along the arm would shrink in
the same way.
The Fitzgerald contraction seems absurd at first sight because it is
not the result of any forces acting on a body but depends only on the
fact that the body is in motion. Fitzgerald's hypothesis was purely
ad hoc, and its main object was to preserve the concept of a stationary
ether. Lorentz, who was at the same time trying to develop a consistent
electromagnetic theory of the electron, found the hypothesis useful in
bis work and incorporated it into bis theories, so that the hypothesis is
now .known as the Lorentz-Fitzgerald contraction hypothesis. It had
previously been shown that under certain assumptions, the mass of an
electron could be expressed by the formula
{6-4)
where r 0 is the radius of the electron when it is at rest, and e is its charge.
Lorentz assumed that a moving electron contracts in the direction of its
motion by the Fitzgerald factor v'I - (v 2 /c 2 ), so that the radius becomes
r 0 v'I - (v 2 /c 2 ) when the electron moves with a velocity v. Then, if the
masa of a moving electron is m, and if the mass when the electron is at
rest is m0 , it follows that
mo
m = -y'-;:1=_=:;::(v=2/:;::c=2) (6-5)
According to Eq. (6-5), the mass of an electron should increase with its
velocity, especially when the latter reaches a significant fraction of the
velocity of light. Early experiments on the value of e/m for {J-rays in-
dicated that the value of e/m decreases with increasing velocity of the
rays. Since there was no reason to suppose that the charge varies with
velocity, it was concluded that the mass of the electron does indeed in-
crease with its velocity. Thus, although the Lorentz-Fitzgerald con-
traction hypothesis seemed physically unreasonable, it led to theoretical
results consistent with experiment. But the hypothesis lacked a sound
theoretical basis, and classical physics was once again confronted by a
problem (the negative result of the Michelson-Morley experiment) that

it could not really solve.
The Michelson-Morley experiment has been repeated under different
conditions and with modifications; other tests of the possible motion of
the earth relative to the ether have also been made. No significant motion
has been detected.m In 1958, in an extremely sensitive test which in-
volved the use of molecular clocks, <2 > it was found that the maximum
velocity of the earth with respect to the ether is less than 1/1000 of the
earth's orbital velocity, or less than 0.03 km/sec. These results all show
that the hypothesis of a stationary ether through which the ea.rth moves
is not verified by experiment.
6-4 The problem of the invariant form of physical theories. The

concept of the invariant form of physical theories is a funda.mental one
in science. To see what this mea.ns, let us consider what is meant by
a physical theory. A physical theory consists, essentially, of equations
or formulas which relate measurable quantities such as distance and
time to invented quantities such as energy, momentum, pot~mtial, field
strength, and so on. Numbers can be assigned to the invented quantities,
and when they are inserted into the formulas of the theory, and when the
appropriate mathematical operations are performed, the calculated values
of the measurable quantities should agree with the values measured in
suitable experiments. The measured values can usually be expressed in
terms of positions in sorne coordina.te system such as x, y, z, and of time
intervals t. It is then highly desirable that the formulas which make up
a theory be the sa.me for different observers. For exa.mple, suppose
that there is an observer who is at rest with respect to sorne system of
reference S (Fig. 6-3). He mea.sures position in terms of the coordina.te
system (x, y, z) and mea.sures time intervals t on a clock. Suppose that
there is another observer who is at rest with respect to another system of
reference S' which moves with unüorm velocity v a.long the x-axis, the
direction of which coincides with the direction of the x'-axis. At time
y'
s S'
t'
_____!___
-----x -----x'
z'
Fm. 6-3. Two Cartesian reference systems in uniform relative translatory

motion.
6-4] THE INVARIANT FORM OF PHYSICAL THEORIES 111
t = O, the two reference frames are assumed to coincide. The second

observer uses the spatial coordinates x', y', z' and the time coordinate t'.
N ow let both observers S and S' perform an experiment which in vol ves
the motion of a body. Each observer will make measurements appropriate
to the experiment, and will describe the results by means of formulas.
In the formulas of the observer S, the quantities x, y, z, and t will appear;
in those of S', the quantities x', y', z', and t' will appear. It was found
long ago that the laws of classical mechanics as expressed by the two
observers S and S' have the same forro provided that the coordinates
used by the two observers are related by means of the transformation
equations:
x' = x - vt, y'= Y, z' = z, t' = t. (6-6)
Equations (6-6) connect the variables of the two systems S and S' in
such a way that if an experiment involving the motion of a body has
been measured by an observer moving with the system S' and has been
described in terms of the quantities x', y', z', t', the substitutions (6-6)
give a correct description of the same experiment in terms of the vari11 bles
x, y, z, t, used by an observer moving with the system S. In more formal
language, the laws of classical mechanics are said to be invariant with
respect to the transformation (6-6).
When the transformation (6-6) is applied to the theory of electro-
magnetics, the formulas of the theory are not invariant; the formulas of
an observer in the moving system S' are not the same as those of an
observer in the S system. Lorentz, who developed a theory of electrons
based on classical electromagnetics, solved the problem of what trans-
formation equations would lead to the same formulas for bis electron
theory for two observers, one of whom is moving with a uniform velocity
v with respect to the other. The system of transformation equations
that he derived is known as the Lorentz transformation, and is given by
the following relationships:
X - Vt
x' = -;====== z' = z,
v'l - (v2/c2) (6-7)
2
y'= Y, t' = t - (vx/c )
v'l - (v2/c2)
These equations are remarka.ble because they indicate that space and
time are not independent entities but are related; time as measured by an
observer in the S' system is different from that measured by an observer
in the S system, and depends on the velocity with which the S' observer
is moving with respect to the S observer as well as on the space coordinate
112 THE SPECIAL THEORY OF RELATIVITY [CHAP, 6
x. Classical physicists found it impossible to accept this relationship

between the space and time variables because it was contrary to the
fundamental classical idea of the independence of space and time. The
relation between x' and x, however, does lead to the Lorentz-Fitzgerald
contraction, as can be shown quite easily.
Suppose that a body, when at rest in S, has a length Lo in the direction
of the x-axis. Let it be set into motion relative to S at such a speed that
it is at rest in S'; that is, it moves with a velocity v relative to S, and
this velocity v is the same velocity with which S' moves with respect to S.
The length'will then be Lo as measured in S', since the length must have
the same value in any reference system in which the body is at rest.
But what value does the observer S measure for the length of the body
which is now moving with respect to him? He measures a value L given
by
where x 2 and x 1 are the coordinates of the ends of the body in the S system.
The S' observer measures the length as
If we substitute for x; and x~ the values calculated from the first of Eqs.
(6-7), we get
Lo = x2 - vt x1 - vt
v'l - (v2/c2) v'l - (v2/c2)
= =
or
L = L0v'I - (v2/c2).
According to this relation, a rigid body, when it is in uniform motion

relative to a stationary observer, appears contracted in the direction of
its relative motion by the factor ✓1 - (v 2/c 2 ), while its dimensions
perpendicular to the dírection of motion are unaffected. When this rela-
tion is applied to the interferometer arms in the Michelson-Morley
experiment, it gives just the contraction needed to account for the nega-
tive result of that experiment. Thus, the requirement that the equations
of electrodynamics be invariant under the transformation equations
(6-7) leads to the Lorentz-Fitzgerald contraction. The difficulty with
accepting these results was caused by the lack of a consistent physical
basis for them, and it became necessary to revise some of the funda-
mental ideas of physics.
6--5] THE VARIATION OF MASS WITH VELOCITY 113
6-5 The special theory of relativity; the variation of mass with velocity.
The problema raised by the Michelson-Morley experiment and the re-
quirement of invariance were solved in 1905 by Einstein, who proposed
the special, or restricted, theory of relativity. Einstein interpreted the
negative results of the Michelson-Morley experiment to mean that it is
indeed impossible to detect any absolute velocity through the ether.
The ether was as close to an "absolutely stationary" reference frame as
classical physics could come. Hence, according to Einstein, only relative
velocities can be measured, and it is impossible to ascribe any absolute
mea.ning to different velocities. It follows that the general laws of physics
must be independent of the velocity of the particular system of coordinates
used in their statement because, ü this were not so, it would be possible
to ascribe sorne absolute meaning to different velocities. But this is just
another way of stating the requirement that the equations of a physical
theory be invariant with respect to coordínate systems moving with
different velocities. In formulating the relativity theory, Einstein adopted
this requirement of invariance as one of the two basic postulates of the
theory.
In his first statement of the theory, Einstein was restricted by mathe-
matical difficulties to the consideration of reference systems moving at
a constant velocity relative to each other. He therefore took as his first
postulate:
PosTULATE l. The laws of physical phenomena are the same when
stated in terms of either of two reference systems moving at a constant
velocity relative to each other.
For his second postulate, Einstein took:
PosTULATE 2. The velocity of light in free space is the same for ali
observers, and is independent of the relative velocity of the source of light
and the observer.
The second postulate may be regarded as a result of the Michelson-
Morley experiment and other optical experimenta and of astronomical
observations, i.e., it is believed to representan experimental fact.
The term "specíal theory of relativity" refers to the restriction in the
first postulate to reference systems moving at a constant velocity relative
to each other. It was not until 1916 in the general theory of relativity
that Einstein was able to show that physical laws can be expressed in
a form that is valid for any choice of space-time coordinates. The general
theory is not needed to describe most atomic and nuclear phenomena and
therefore will not be discussed.
From his two postulates, Einstein deduced the Lorentz transformation
equations as well as other kinematical relationships. , The Michelson-
Morley experiment could be interpreted directly in terms of these rela-

tionships and there was no need for the Fitzgerald contraction hypothesis
or for the ether. The interrelationship between space and time coordi-
nates also follows from the postulates and represents a more advanced
treatment of these ideas than does the earlier classical theory. The inter-
dependence between the space and time coordinates means that time
must be treated in the same way as the space coordinates. This result
gave rise to the ideas of a four-dimensional space-time continuum as well
as to popular misconceptions about the "fourth dimension." lt must be
pointed out, however, that when v is much smaller than e, the Lorentz
transformation equations (6-7) reduce to the classical equations (6-6), so
that for ordinary mechanical phenomena, the classical or Newtonian
theory is adequate. The relativity theory does not "overthrow" the
classical theory but rather extends and modifies it. The need for the
relativity theory in atomic and nuclear physics comes in part from the
fact that fundamental particles such as electrons can travel with velocities
approaching that of light, with the result that the kinematic and dynamical
consequences of the Lorentz equations are important.
The Lorentz transformation equations can be derived quite easily from
Einstein's postulates if we consider measurements, by two observers, of
the distance traveled by a light ray. Suppose that the system S' is moving
at uniform speed v in the x-direction relative to the system S, and that
at time t = t' = O a light ray starts from the (then coincident) origina
of coordinates, x = x' = y = y' = z = z' = O. At sorne later instant,
the distance traveled by the light ray is measured by observers in the two
systems. In the S system the result is
(6-8a)
according to Einstein's postulates, the result in the S' system must be
(6-8b)
with e, the speed of light, the same in the two systems. lt is easily verified
that Eqs. (6-6) do not lead from Eq. (6-8b) to Eq. (6-8a), and another
transformation is needed. As a more general transformation we try the
equations
x' = 'Y(x - vt), y'= Y, z' = z, t' = ax + bt, (6-9)
where 'Y, a, b are constants to be determined. The first relation is chosen

because it is similar to the first of the classical equations (6-6) and should
reduce to that equation when v is small; it is also desirable that the trans-
formation equations be linear. The second and third equations are chosen
6-51 THE VARIATION OF MASS WITH VELOCITY 115
because. there is no reason to suppose that the relative motion in the x-

direction should affect the y- and z-coordinates. The fourth relation is
chosen because it is linear in x and t.
If we now insert Eqs. (6-9) into Eq. (6-8b) we get
('Y2 - a2c2) x2 + y2 + z2 = (c2b2 - 'Y2v2) t2 + 2xt (abc2 + t1'Y2). (6-10)

Since this equation is to be identical with Eq. (6-8a), we must have
abc 2 + V'Y 2 = 0. (6-11)
On setting A = ac, and fJ = v/ e, we get

-y2 - A 2 = 1, (6-12a)
(6-12b)
Ab + {J'Y 2 = O (6-12c)
Elimination of A between Eq. (6-12a} and Eq. (6-12c) gives

2 4 2
fJ
'Y2 _ __y_ = 1 = !__ (b2 _ fJ2'Y2) (6-13)
b2 b2 '
and comparison with Eq. (6-12b) yields

-y2 = b2. (6-14)
If we now substitute this result in Eq. (6-12b), we get
-y2 = b2 = __1_ = 1 . (6-15)

1 - fJ2 1 - (v2/c2) '
then
A2 = a2c2 = -y2 - 1 = ~. (6-16)
1 - fJ2
In finding 'Y, b, and A, we choose the signs of the square roots so that they
are consistent with Eq. (6-12c), and lead to the necessary results x = x'
and t = t' when v = O. This procedure yields
1
'Y = b = -v-;:i=-=(:;::v2=¡=c2=) (6-17a)
v/c 2
a = - --;==== (6-17b)
v'l - (v2/c2)
When these relations are inserted into Eqs. (6-9), we get the Lorentz
transformation equations (6-7). The latter can be solved for x and t,

with the result
x' + vt' t = t' + (v/c2)x' '
X = -y-;::l=-=(;=v2=/;=c2:=) v'l - (v2/c2) (6-18)
y= y', z = z'.
Equations (6-18) are exactly similar in form to Eqs. (6-7) except that v,
the velocity of S' relative to S, is replaced by -v, the velocity of S rela-
tive to S'.
The transformation equations for the components of the velocity of a
moving point can be found by taking the derivatives of the Lorentz trans-
formation equations with respect to t and t'. If we set
I dx' dx
u.,=dr' u.,= dt'
with similar definitions for U:,, u~, u11, and u., we have,
I dx' dt u., - V
u., = di dt' = 1 - (u.,v/c 2) (6-19a)
, dy dt v'l - (v2/c2)
(6-19b)
U11 = dt di! = 1 - (u.,v/c2) Uu,
U1 -
v'l -_o---'---'-U
....;__
(v2/c2)
(6-19c)
• - 1- (u.,v/c2) ••
dt' 1 - (vu.,/c 2)
(6-19d)
dt = v'l _ (v2/c2) ·
The inverse transformation equations are again obtained by replacing
vby -v:
u, = v'l - (v2/c2)
--~~~ u~.
1 + (u~/c )
2
(6-20)
The transformation equations for the components of the velocity lead
toan important addition theorem for velocities. In Newtonian kinematics,
relative velocities are found by simple addition or subtraction; in rela-
tivistic kinematics, the composition of velocities is more complicated.
Because of the denominator in Eqs. (6-20), the sum of two velocities
which are separately less than the velocity of light e can never exceed c.
Thus, in the first of Eqs. (6-20), if u; and v both approach e, u., also
6-5] THE VARIATION OF MASS WITH VELOCITY 117
approaches e¡ but for values of u; and v small compared with e, the de-
nominator approaches unity and u., approaches u; + v, the classical
expression for the addition of velocities. Although the relativistic equations
seem, at first to be contrary to our physical intuition, they yield correct
results when applied to such problems as the Doppler effect of a light wave
and the i.berration of starlight.
It is not surprising, in view of the interrelationships between space and
time coordinates as expressed in the Lorentz transfonnation equations,
that relativistic mechanics differs from classical, Newtonian mechanics
in certain fundamental ways. One of the most important results of the
relativity theory is the deduction that the masa of a body varies with
its velocity. The relationship between mass and velocity can be derived
by considering a simple conceptual experiment of a type first discussed
by Tolman. The derivation which follows is due to Bom, <3> and is highly
instructive because it shows how sorne of the hasic ideas of the relativity
theory are used.
Imagine an ohserver A located on the y-axis of a reference system S,
and another observer B on the y'-axis of a second system S', with S'
moving ata constant velocity v, relative to S, along the x-axis. Suppose
that A throws a hall with velocity U along the y-axis, and that B throws
a hall with velocity -U along the y'-axis. The hall thrown by A has,
y' y
Ball 2
-u
s' s -U S,S'
V u
u
º' x' o X 0,0'
Ball 1
Before collision At instant of collision
y y'
ü 2
-ü
0
- V
X X1
After collision
Fm. 6-4. Elastic collision of two identical halls, considered relativistically.
when at rest in the S system, the same mass as the hall thrown by B
when it is at rest in the S' system. Suppose, further, that the halls are
thrown so that they collide, with the line of centers at the instant of impact
in the direction of the y-axis, as shown in Fig. 6-4. Let the subscript 1
denote the hall thrown by A and the subscript 2 the hall thrown by B;
let u 1 and u 2 be the velocities of the halls before the collision, as measured
by A, and let u~ andu; be the corresponding velocities as measured by B.
We use the subscript l to denote the longitudinal component of a velocity,
that is, the component parallel to the x or x' direction, and the subscript
t for the transverse component, parallel to the y- or y'-direction. The
components of the velocity of the first hall, before the collision, as seen
by the observer in the S system are then
U}l = O; u 11 = U. (6-21a)
The components of t.he velocity of the second hall, before the collision,
as seen by the observer in the S' system, are
(6-2Ib)
The components of the velocity of the second hall relative to the S system
can be obtained by using the first and second of the velocity-transforma-
tion equations (6-20), and taking u;= u;, = -U and u; = u~ = O.
The result is
(6-21c)
We calculate next the total momentum of the two halls relative to the
S observer. lt will not be assumed, however, that the two halls appear to
that observer to have the same mass; it will, in fact, be seen that their
masses must appear different. We, therefore, denote the masses of the
two halls, before the collision, and relative to the S system, by m 1 and
m 2, respectively. We shall assume that, as in classical mechanics, the
momentum is defined as the product of mass and velocity, and that the
momentum vector has the same direction as the velocity vector. Then
the total momentum before the collision, as calculated by A (the S ob-
server), has the components
We now consider the effect of the collision. lf the halls are assumed to be
perfectly smooth and to undergo an elastic collision, they cannot exert
tangential forces on each other, the x-components of their velocities are
6-5] THE V ARIATION OF MASS WITH VELOCITY 119
not changed by the collision, and the collision must take place in a sym-
metrical way. The ohserver A must see his hall undergo the same c:hange
in motion that the observer B sees bis hall undergo. The first hall must
take on a transverse velocity -U opposite in direction to its velocity
before the collision, while the second hall must take on a velocity U,
equal in magnitude but opposite in direction to that of the first hall.
We shall not assume that the mass and velocity of either hall must be
the same after the collision as hefore hut shall see what the theory tells
us. If we label quantities after the collision with bars, we have
'u11 = o, üu = -D (6-23a)
Ü~¡ = o, ü~1 = U. (6--23b)
The components of the velocity of the second hall as calculated by A are,
from Eqs. (6--20),
(6--23c)
The components of the total momentum after the collision as calculated
by A are then,
We now require that the total momentum of the two balls be conserved in
the collision, in analogy with classical mechanics; this requirement is
fundamental in relativistic mechanics. The longitudinal and transverse
components of the momentum must be conserved separately, and com-
parison of Eqs. (6--22) and 6-24) yields
(6-25a)
(6-25b)
If the ma....qg were constant, m 1 = m 2 = m1 = m2 , the first equation
would be identically correct; but the second would lead to a contradiction
because it would require that
(U+ D) (1 - v'l - (v 2 /c 2 )) = o,
which is impossible because, by hypothesis, U and v are different from
zero and positive, and U must be positive because of the symmetry of
the collision.
If conservation of momentum is required, we must give up the idea
that the mass of either hall is constant and replace it by the assumption
that the mass is variable and depends on its velocity (since there is no
other parameter in the problem on which it can depend). More precisely,

we assume that the mass oí a body as measured in a chosen reference
system depends on the magnitude oí the body's velocity relative to that
system.
The magnitude oí a velocity u is given by
u= Vuf + u~.
The speeds of the two balls relative to the S system are then, before the
collision,
U ¡ = U, (6-26a)
(6-26b)
aíter the collision,
U¡= U, (6-26c)
(6-26d)
Now Eq. (6-25a) requires that m2 = ffl 2 ; if the mass varies only with the
velocity, then m 2 and ffl 2 can be equal only if the corresponding velocities
u 2 and u 2 are equal. This condition requires that
v
2
+ u2 ( 1 _ ~) = v
2
+ u2 ( 1 _ ~:) ,
from which it follows that U = U. Equations (6-26a) and (6-26c) then

show that u 1 = u1, that is, the speed oí the first ball relative to observer
A is not changed by the collision; hence, m 1 = m1 , and the mass of the
first ball relative to A is also not changed by the collision. Equation
(6-25) can, therefore, be written,
m 1 U - m2Uv1 - (v 2/c 2) = - m 1U + m2Uvl - (v 2/c 2),
or
It follows that
m1
m2 = --;:=====
v'l - (v2/c2)
(6-27a)
According to Eq. (6-27a), the two halls, which were assumed to have
the same mass when at rest in their respective reference systems, appear to
have different masses when one reference system moves with constant
velocity v relative to the other. When the system S' moves with constant
velocity v relative to S, the mass m2 appears larger than m 1 to an observer
stationary in S by the factor 1/v'l - (v 2 /c 2 ). If we now imagine that
the velocity U with which the halls are thrown is made smaller, then,
according to Eqs. (6-26), u 1 = O and u 2 = v in the limit U = O. The
6-5] THE VARIATION OF MASS WITH VELOCITY 121
velocity U is made to approach zero only to simplify the derivation and

is not a necessary assumption (see Problem 8). The mass m 1 , which cor-
responda to zero veloc1ty is called the rest masa and is denoted by m0 ,
while m 2 is the mass which corresponda to velocity v, and is denoted
by m. We have, finally,
mo
= --;:=::::::;== (6-27b)
m
vl - (v2/c2)
The derivation has shown that when a body moves with constant velocity
v with respect to a stationary observer, it has a mass which dependa on
the velocity and is la.rger than the rest mass m 0 characteristic of the body.
Equation (6-27b) shows that no material body can have a velocity equal
to, or greater than the velocity of light.
Equation (6-27b) had been obtained earlier for the mass of an electron;
according to the relativity theory, it holds for any moving body. The
variation in mass becomes important when the speed of the body begins to
approach that of light, as is evident from the following values.
v/c mimo v/c m/mo

0.00 1.000 0.90 2.294
0.01 1.000 0.95 3.203
0.10 1.005 0.98 5.025
0.50 1.155 0.99 7.089
0.75 1.538 0.998 15.819
0.80 1.667 0.999 22.366
For bodies moving at speeds up to about one percent that of light, the
variation in mass is negligible, but in the case of more swiftly moving
particles the correction is appreciable. For this reason, {:J-rays from radio-
active substances have been used to test the relativistic formula for the
mass.
One of the methods used to test the formula depends on the measure-
ment of the value of the specific charge, e/m, as a function of the speed
of the beta particles. According to the relativity theory, the electric charge
is a quantity which is independent of velocity and has the same value for
all observers. Consequently, the value of e/m should depend on the veloc-
ity only because of the dependence of the mass on the velocity, and
!.... = .!__ ✓l _ v2, (6-28a)

m m0 c2
or
..!!__ = e/m (6-28b)
mo VI - (v2/c2)
where e/m 0 is the value corresponding to the rest mass of the electron.
If e/m and v are measured, the measured value of e/m divided by
vl - (v 2 /c 2 ) should be a constant, which is e/m 0 , and the mass ratio
m/m 0 is also obtained.
Several experiments of this type have been made, w the most famous
of which is probably that of Bucherer in 1909. In this experiment, a small
radioactive source S of {3-rays was placed at the center between two very
closely spaced metal discs forming an electrical condenser C. The con-
denser was placed inside a cylindrical box which could be evacuated, and
a photographic film P was wrapped around the inner surface of the box.
The box was placed in a homogeneous magnetic field H parallel to the
plane of the clises, as shown schematically in Fig. 6-5(a). In the region of
the electric field between the dises the electric and magnetic deflecting
forces must balance each other exactly for a particle to escape. The con-
dition for escape is eE = evH, or v = E/H, and the velocity of an emerg-
ing {3-particle can be determined. The speed of an emerging {3-particle is a
function of the azimuthal angle IJ at which it is emitted, and only particles
with a single speed v emerge ata given angle IJ, as indicated in Fig. 6-5(b).
After emergence from between the plates, the {3-particle is deflected in the
magnetic field alone, and the result of many particles over a wide range of
p
(a) -11...
(b) -- -11.-
Fm. 6-5. Schematic view of Bucherer's apparatus for mea.suring e/m for
electrons with different speeds. (a) elevation view (b) plan view (e) traces on
photographic film. (Reprinted by permission from E. R. Cohen, K. M. Crowe,
and J. W. M. Dumond, Fundamental C=tanl8 of Phyaia. New York: lnter-
science, 1957 .)
6--5] THE VARIATION OF MASS WITH VELOCITY 123
speeds and emerging at different angles B is a continuous curve on the

photographic film. A typical trace corresponding to half the circumference
of the box is shown by one of the curves in Fig. 6--5(c); the second curve
on the film is fonned when the electric and magnetic fields are reversed
simultaneously. The value of e/m corresponding to a velocity II is obtained
from the relation
2
mv
Hev =-r
, or
ü we insert for v its value v = E/H, then

e E
m = rH2·
The field strengths E and H are known and r, the radius of the circular
path of a fj-particle of speed v, can be obtained from an analysis of the
photographic trace and geometrical relationships among the various parts
of the apparatus.
2.0
1.9
e/m e
v'f=v¼2 mo
- 1.8
1.7
1.6
1.5 m
mo
1.4
1.3
1.2
1.1
1.0
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Velocity (v/c)
FIG. 6-6. The specific charge and mass of the electron as functions of velocity.
Verification of the relativistic mass formula; e/mo is in units of 107 emu/gm.
124 THE SPECIAL THEORY OF RELATlVITY [CHAP. 6
Later experiments have used improved methods< 5 , 5> and sorne of the
resulta obtained by Bucherer and others are shown in Fig. 6-6. The solid
curves represent the theoretical expressions for m ande/moas functions of
velocity, and the points represent the experimental resulta. There is no
doubt that the theoretical predictions are confirmed for swiftly moving
~-particles. The mass formula has also been confirmed for other particles.
6-ó The relationship between mass and energy. The variation of

mass with velocity leads to modifications of our ideas about energy. In
classical physics, force is defined as the time rate of change of momentum,
and
d dv
F = dt (mv) = m dt , (6-29)
since mass is constant. This definition of force is kept in relativistic

mechanics, but now masa is also a variable and
d dv
F = -dt (mv) = m-
dt
+ v dm
- •
dt
(6-30)
The kinetic energy T of a particle is defined in terms of force and distance,

as in classical mechanics, and can be found in the usual way by calculating
the work done in setting the particle into motion:
dT = F dx. (6-31)
On insertion of Eq. (6-30) for F, there follows
dv dm
dT = m dt dx + v dt dx
= mv dv +v 2
dm. (6-32)
Differentiating both sides of Eq. (6-27b) we get
dm = mo vdv
c2 (1 - (v2/c2))ª'2
mvdv
c2 - v2 (6-33)
Substitution for mv dv from Eq. (6-33) in Eq. (6-32) yields
dT = c 2 dm. (6-34)
Thus, a change in kinetic energy dT is directly proportional to a change
in masa dm, and the proportionality factor is the square of the velocity
6-6] THE RELATIONSHIP BETWEEN MASS AND ENERGY 125
of light. The integral fonn of this result can also be obtained easily.
If we substitute for dm from Eq. (6-33) in Eq. (6-32), we get
dT = mov dv + m 0 vª/e2 dv
v'l - (v 2/c 2) [1 - (v 2/c 2 )] 312
m 0 vdv
= [1 - (v2/c2))312
Integration of both sides of the last equation gives for the kinetic energy
T = m0 c
2
[
1 - 1)
v'l - (v2/c2)
(6-35)
According to this result, the kinetic energy of a body can be expressed
in tenns of the increase in the mass of the body over the rest mass. This
relationship may be interpreted as meaning that the rest mass m0 is
associated with an amount of energy m0c2 , which may be called the rest
energy of the body. The total energy E of the body is then the sum of the
kinetic energy and the rest energy, or
E= T + m 0c2
moc 2
= mc2 = -v-;:1=-==(;:v2=;;:c2=) (6-36)
Hence, associated with a mass m, there is an amount of energy mc 2 ;

conversely, to an energy E, there corresponds a mass given by
E
m=-·
c2 (6-37)
Another important consequence of the proportionality between mass

and energy has to do with the momentum associated with the transfer of
energy. If a quantity of energy is transferred with a velocity v, then
we can write for the magnitude of the associated momentum p
Ev
p = mv = c2. {6-38)
Tbis relationship can be applied, for example, to the energy and momentum
carried by a light quantum (photon). The energy carried by a photon
is hv and ita velocity is c. The momentum is, by Eq. (6-38),
E E hv
p = c2 e= e=-¡;-· (6-39)
The rest masa oí a photon must be zero; otherwise the masa oí a photon
traveling with the velocity of light would be infinite, beca.use the de-
nominator of Eq. (6-27) becomes zero for v = c. Although the rest masa of
a photon is zero, the photon has a mass associated with its kinetic energy,
according to Eq. (6-37), and this masa is h11/c2 ; corresponding to this
masa is the momentum hv/c,
The interconversion of mass a.nd energy can illustrated by several
numerical examples which will be useful in la.ter calculations. If the
mass is in grams and the velocity is in centimeters per second, the energy
is in ergs, and
E(ergs) = m(gm) X {2.998 X 10 1º) 2 cm 2/sec 2
= m{gm) X 8.99 X 10 20 cm 2 /sec 2 . (6-40)
One gram of mass is associated with nearly 10 21 ergs of energy, or about

8.5 X 10 10 Btu, since 1 Btu = 1.055 X 10 10 ergs. This is a very large
amount oí energy indeed, but the practical conversion of mass to energy
is a difficult problem. In an ordinary chemical procesa, mass is always
converted into energy. For example, if 100 gm of a. chemical substance
take part in a reaction, with the result that 10 6 calories of heat are liberated,
the mass-energy balance is (1 cal = 4.184 X 107 ergs)
ti.E 4.184 X 107 X 106

t..m = c2- = 8.99 X 1020
= 4.65 X 10-8 gm.
The loss of masa equivalent to the energy involved in the reaction is

4.65 X 10- 10 gm/gm of substance, a decrease which ca.nnot be detected
by even the most sensitive chemical balances. This is the rea.son why the
mass-energy effect has not yet been observed in chemical reactions.
lt will be seen in la.ter chapters that the mass-energy conversion can
be detected in nuclear phenomena. beca.use the amounts of energy involved
per atom are very much gres.ter tha.n those involved in chemical reactions.
lt is useful to obtain now sorne of the conversion factora which will be
needed in la.ter work. In atomic a.nd nuclear calculations, the erg is an
inconvenient unit of energy because the amounts of energy involved in
single events usually are only small fractions oí a.n erg, say 10- 12 to 10-6
erg. lt is therefore the practice to expresa energies in units oí electron volt&,
abbreviated as ev. The electron volt is the energy acquired by any
charged particle carrying a unit electronic charge when it falls through
a. potential oí one volt; it is equivalent to 1.602 X 10- 12 erg. For con-
6--7] THE COMPTON EFFECT 127
venience, two other units are also used; one, equal to 10 3 electron volts, is
represented by kev and the other, equal to 10 6 electron volts, is a.bbreviated
by Mev. One kev is therefore equal to 1.602 X 10-9 erg, and one Mev
is equal to 1.602 X 10-5 erg. The mass-energy relation becomes
8.99 X 10 20 cm 2/sec 2
E (ev) = m(gm) X 1.602 X 10-12 erg/ev
32
= m(gm) X 5.61 X 10 , (6-41a)
29
E (kev) = m(gm) X 5.61 X 10 , (6--41b)
E (Ml;lv) = m(gm) X 5.61 X 10 26 . (6-41c)
The rest masa of an electron is 9.108 X 10-28 gm; by Eq. (6--4lc), the rest
energy is 0.511 Mev. For a hydrogen atoro, the rest mass is 1.674 X 10- 24
gm and the corresponding energy is 939 Mev. It· is often convenient to
use the energy corresponding to one atomic mass unit. This mass is
1.66 X 10-24 gm, or 931.141 Mev.
E (Mev) = m(atomic mass units) X 931.141. (6--42)
The values of the kinetic energy which will be used most often range
from electron volts to millions of electron volts. For example, an a-
particle may have a velocity of 2 X 109 cm/sec, or about one-fifteenth
of the velocity of light. The relativistic mass correction for this velocity
is sufficiently sma.ll so that it may be safely neglected. The rest mass
m.. of an a-particle is very closely four times that of a hydrogen atom.
Hence, the kinetic energy of this a-particle is
9 2
½m.. v2 = ½X 4 X 1.674 X 10- 24 X (2 X 10 ) erg
= 13.39 X 10-5 erg
13.39 X 10-6
= l.602 X I0- 6 Mev = 8.36 Mev.
• 6-7 The Compton effect. Sorne of the ideas treated in Chapter 5 and
in the present chapter can be illustrated by discussing an effect discovered
by A. H. Compton in 1923 during the course of his studies of the scattering
of x-rays by matter. The theoretical treatment of the Compton effect
involves Einstein's quantum theory of light as well as sorne of the ideas of
the relativity theory. The effect is also of practica! importance in the
absorption of x-rays by matter and will be treated in further detail from
that viewpoint in Chapter 15. Finally, sorne of the ideas used by Compton
in bis explanation of the efJect play an important part in the modern

quantum theory of atomic structure and will be needed in Chapter 7.
For all of these reasons, it is convenient to discuss the Compton effect now.
In the treatment of the scattering of x-rays by electrons in Chapter 4,
it was assumed that the wavelength of the scattered radiation was the
same as that of the incident radiation. This kind of scattering is called
"Thomson" scattering; its effect is to change the direction of the incident
radiation. With improved techniques, Compton {1923) was able to show
that when a beam of monochromatic x-rays was scattered by a light element
such as carbon, the scattered radiation consisted of two components, one
of the same wavelength as that of the incident beam, the second of slightly
longer wavelength. The wavelengths of the two lines were measured
with a Bragg x-ray spectrometer. The difference in wavelength .iA
between the two scattered radiations was found to vary with the scattering
angle and to increase rapidly at large angles of scattering. When the angle
between the incident and scattered radiation was 90°, the difference in
wavelength was found to be 0.0236 X 10-s cm, independent of the wave-
length of the primary beam and of the nature of the scattering material.
To account for the presence of the shifted component and for the amount
of the shift, Compton assumed that the scattering procesa could be treated
as an elastic collision between a photon and a free electron, and that in
this collision energy and momentum are conserved. According to Ein-
stein's quantum theory of the photoelectric efJect, the primary x-rays,
being electromagnetic radiations like light, are propagated as quanta
with energy hv. Along with their energy, they carry momentum hv/c in
accordance with the ideas developed in the Iast section. The scattered
quantum or photon moves in a different direction from that of the primary
photon and carries a different momentum. In order that momentum
be conserved, the electron which scatters the photon must recoil with
a momentum equal to the vector difference between that of the inci-
dent and that oí the scattered photon, as in Fig. 6-7. The energy of
this recoiling electron is taken from that of the primary photon, leaving
a scattered photon which has lesa energy-and hence a lower frequency,
or longer wavelength-than that of the primary photon.
The requirement that energy be conserved gives
hvo = hv + m c 2 [ v'l -
0
1 - 1] , {6-43)
(v2/c2)
where 110 is the frequency of the incident x-ray, v that of the ray scattered
by the electron, the recoil velocity of the electron is v, and the relativistic
kinetic energy of the electron is used because the velocity of the electron
may be great enough for relativistic effects to be significant. The require-
6-7] THE COMPTON EFFECT 129
Scattered photon
Momentum = hv/c
Incident photon
Momentum = hvo/c
F10. 6-7. The Compton effect.
ment that momentum be conserved gives two equations, one for the
x-component of the momentum, and one for the y-component
hv 0 hv mov
x-component: - = - cos et,+---;:=:=:::;== cos 0, (6-44)
e e yl _ (v2/c2)
. mov
y-component: O= -hve sm et, - --;:==;::=== sin O.
yl - (v2/c2)
(6-45)
To solve these equations, it is convenient to set

fj = v/c, (6-46)
and rewrite the last three equations. This gives
hvo = hv + m 0c2 [1~

- fj2
- 1], (6-47)
hvo = hv cos et, + mofjc cos 0, (6-48)

e e yl _ fj2
hv . 0 m fjc .
O = -smct, - -~--sm 0. (6-49)
e ~
For a particular angle of scattering et,, the last three equations contain
three unknowns, v, fj, and 8, and expressions for these quantities can be
found by a straightforward solution of the equations. For comparison
with experiment, however, we are more interested in the shift in wave-
length. We therefore introduce the wavelengths
>-o = .E.. and >. = ~
Vo V
of the incident and scattered radiations, respectively. Then Eqs. (6-48)

and (6-49) become:
h h m {Jc
-
>-o
- - cos q,
X
= ---;:=====
0
v'l - fJ2
cos 9
'
h . m 0 tk .
X sm 4> = Vl - 112 sm 9.
Squaring these two equations and adding, we have

h2 2
2h 2 cos 4> m~{J 2c 2 m~ 2
~ + hx2 - >-o>- = 1 _ 112 = 1 _ 112 -
2 2
moc . {6-50)
Similarly, Eq. {6-47) can be written
->-oh - -h>. + moc = -

~
moe
--
On squaring, we get
2 2 2 2
h
>,g + hx2 - 2h
>-o>- + 2moeh >-o - X
( 1 1)
+ m¡¡c2 2 = m~
1 - 112 • {6-51)
Subtracting Eq. (6-50) from Eq. {6-51), we get

2
~>.
2
h (cos 4> - 1) + 2moch (_!_
~
- !)>. = O.
Therefore,
h
..:1>. = X- >-o = -moe (1 - cos 4>). {6-52)
If the values h = 6.624 X 10-21 , m = 0.9107 X 10-21 , and e = 2.998

X 10 10 are inserted, then
'1>. = >. - >. 0 = 0.0242 X 10-8 (1 - cos 4>) cm
= 0.0242(1 - cos 4>) A. {6-53)
Equation {6-53) states than when an incident x-ray of wavelength

>-o is scattered through an angle 4> by a free electron, the wavelength >. of
the scattered x-ray should be greater than that of the incident x-ray by
the amount 0.0242(1 - cos 4>) A. For a given value of the scattering
angle 4>, the shüt in wavelength is independent of the wavelength of the
incident radiation. For 4> = 90°, ..:1>. = 0.0242 A, which agrees very
well with the observed value of 0.0236 A. The predicted dependence
of ..:1>. on the angle q, was also verified by experiment.
REFERENCES 131
The kinetic energy T of the recoil electron can be calculated without

much difficulty, and it is found that
T = hv (1 - cos </>)a , (6-54)

o 1 + a(l - cos ,t,)
where a = hv 0 /m 0c 2 •
The recoil electrons were sought and found, and their observed energies
agreed with the values predicted by the theory.
The theory just developed explains the observed shift in wavelength,
but <loes not account for the presence of the unshifted line. It will be
recalled that it was assumed that the electrons are free. Actually, elec-
trons are bound to atoms more or less tightly, and a certain amount of
energy is needed to shake an electron loose from an atom. If the amount
of energy given to the electron by the photon is much larger than the work
needed to detach it from the atom, the electron acts like a free electron,
and Eq. (6-53) is valid. If the collision is such that the electron is not
detached from the atom, but remains bound, the rest mass m0 of the
electron in Eq. (6-53) must be replaced by the mass of the atom, which is
severa! thousand times greater. The calculated value of t::.>.. then be-
comes much too small to be detected. A photon which collides with
a bound electron, therefore, <loes not have its wavelength changed, and
this accounts for the presence of the unshifted spectral line.
It will be noted that in the Compton formula, Eq. (6-52), the only
trace of the relativity theory that appears is the zero subscript denoting
the rest mass of the electron. It would seem as though the same effect
would have been obtained if the problem had been set up without taking
into account the relativistic effects, a.nd this is actually the case so far
as the shift in wavelength is concerned. The more refined details of the
Compton effect, however, demand a rigorous relativistic treatment.
Hence, even the more elementary problems are usually set up relativisti-
cally.
REFERENCES
GENERAL
A. EINSTEIN, Relativity: The Special and General Theory, 15th ed. London:
Methuen, 1954; New York: Crown Publishers, 1961.
A. EINSTEIN and L. lNFELD, The Evolution of Physics. New York: Simon and
Schuster, 1938, Part III.
A. EINSTEIN, The .Meaning of Relativity, 5th ed. Princeton: Princeton Univer-
sity Press, 1956.
C. V. DuRELL, Readable Relativity. London: G. Bel! and Sons, 1926; New
York: Harper and Brothers (Harper Torch Books/The Science Library), 1960.
A. EDDINGTON, Space, Time and Gravitation. Cambridge: University Press,

1920¡ New York: Harper and Brothers (Harper Torch Books/The Science
Library), 1959.
H. DINOLE, The · Special Thwry of Relativity. London: Methuen and Co.,
1940¡ New York: Wiley.
W. H. McCREA, Relativity PhyBics, 4th ed. London: Methuen and Co.,
1954¡ New York: Wiley.
P. G. BEROMANN, Introduction to the Theory of Relativity. New York: Prentice-
Hall, 1942.
C. MottER, Thwry of Relativity. Oxford University Press, 1952.
W. PAULI, Thwry of Relativity. London: Pergamon Press, 1958.
R. C. TouuN, Relativity, Thermodynamics, and Cosmol-Ogy. Oxford University
Press, 1934.
The Principle of Relativity, A Collection of Memoirs on The Special and General
Thwry of Relativity, by H. A. Lorentz, A. Einstein, H. Minkowski, and H. Weyl.
London: Methuen and Co., 1923¡ New York: Dover Publishing Co.
W. K. H. PANOFSKY and M. PHILLIPS, ClasBical Electricity and Magnetism.
Reading, Mass.: Addison-Wesley, 1955, Chapters 14, 15, 16.
A. H. CoMPTON and S. K. ALLISON, X-Rays in Theory and Experiment, 2nd ed.
New York: Van Nostrand, 1935, Chapter 3.
PARTICULAR
l. W. K. H. PANOFSKY and M. PHILLIPS, op. cit. gen. ref., Chapter 14. "The
experimental basis for the theory of special relativity.")
2. CEDARHOLM, BLAND, HAVENs, and ToWNES, "New Experimental Test of
Special Relativity," Phys. Rev. Letters, l, 342 (1958).
3. M. BoRN, Die Relativitatstheorie Einsteins und ihre Physilcali.8che Grund-
lagen. Dritte Aufiage. Berlin: Springer, 1922, Chapter 7, Section 7.
4. P. S. FARAOO and L. JANOSSY, "Review of the Experimental Evidence for
the Law of Variatíon of the Electron Mass wíth Velocity", N uovo cimento, Vol. V,
No. 6, 1411 (1957).
5. CoHEN, CRoWE, and DUMOND, Fundamental Constants of PhyBÍcB. New
York: Interscience, 1957, pp, 13(}-142.
6. RoGERs, McREYNOLDs, and RoGERS, "A Determination of the Masses
and Velocities of Three Radíum B /3-particles¡ The Relativistic Mass of the
Electron," Phys. Rev. 57, 379 (1940).
PROBLEMS
·1. Suppose that a clock at rest in the reference system S emits signals at
time intervals At = t2 - ti. At what intervals will the signals appear to an
observer at rest in the system S' moving along the x-axis with uniform velocity
v with respect to the S system? Suppose next that there is a source of homo-
geneous x-rays at rest in the S' system, and that the wavelength of the x-rays as
mea.sured by the S' and S observers is 'A' and 'A, respectively. What is the
relationship between 'A' and 'A?
PROBLEMS 133
2. Derive an expression for the quantity u! 2 = u~ 2 + +

t4 2 u~ 2 in terms of
uz, u., and Uz. Show that if v < e, and u = e, then u' is also equal to c.
3. Suppose that two electrons, ejected by a heatcd filament stationary in S,
move off with equal speeds of magnitude 0.9c, one moving parallel to the +x-
direction, the second parallel to the -x-direction. What is their speed relative
to each other in the S system? Suppose that the S' system moves at a speed
v = 0.9c in the - x-direction, so that it keeps up with the second electron.
What is the velocity of the second electron relative to the first, as measured in
the S' system?
4. Calcule.te the ratio of the mass of a particle to its rest mass when the
particle moves with speeds that are the following fractions of the speed of light:
0.1, 0.2, 0.5, 0.7, 0.9, 0.95, 0.99, 0.995, 0.999, 0.9999.
5. Calculate the kinetic energy, in ergs and in Mev, of the particle of Problem
5 at each speed; assume first that the particle is an electron, then that it is a
proton.
6. Show, by expanding the expression (1 - v2/c2)-ll2 in powers of v/c, that
the kinetic energy can be written
4
1
T= 2 mov
2
+ 83 mo~+···.
v
Find the values of v/c for which the second term of the expression is 1%, 10%,
and 50%, respectively, of the first term.
7. Charged particles can be accelerated to high energies by appropriate
machines (Chapter 21). A constant frequency cyclotron can accelerate protons
to kinetic energies of about 20 Mev; synchrocyclotrons have been built which
accelerate protons to kinetic energies of about 400 Mev, and the Cosmotron at
the Brookhaven National Laboratory has produced 2.5 Bev protons. What
is the mass of the protons relative to the rest mass in each of the three cases
cited? What is the velocity relative to that of light? What is the total energy?
8. In connection with the derivation of the formula for the variation of mass
with velocity, show that Eq. (6-27a),
is va.lid generally, for U '7" O.

9. Derive the following useful relationships concerning the relativistic proper-
ties of moving particles (the symbols are defined in Section 6-6).
(b) p = -1e yT 2 + 2mocT
(e) mo = .!..
c2
yE2 _ p2c2 (d) T = Y(moc 2) 2 + (pc) 2 - moc
2
10. Considera body with rest mass Mina given frame of reference; the body
is composed of two parts with rest masses M 1 and M 2. Suppose that the body
breaks up spontaneously into its two partil with velocities v1 and v2, respectively.
134 THE BPECIAL THEORY OF RELATIVITY [cruP. 6
(a) Write an equation for the conservation of total energy. (b} Show that the
total energies of the two parta are given by
EI
_
- e
2 M
2
+ M1
2
2M
- M2
2
11. A particle of rest mass m1 and velocity II collides with a particle of mass
m2 at rest, after which the two particles coalesce. Show that the mass M and
velocity V of the composite particle are given by
M2 = 2
m1
+ m22 + -;=::::::::====::;::
2m1m2
; V=----;:::====
m111
VI - (112/c2) m1 + m2Vl - (11 2/c2)
12. Show that the following relations hold in a Compton scattering process.
(a) The maximum energy transfer to the electron is given by
T mr,x = 1 +""º
(I/2a)
and
(b) For very large energy of the incident photon, a» 1, the energy of the
backsca.ttered photon a.pproaches ½moc2 •
(e) The angles 8 and •• which the directions of the recoil electron and scattered
photon make with respect to that of the incident neutron, obey the relation
cot 8 = (l + a) 1-cosl/,
. 1/,
sin
= (1 + a) tan -ti,2 ·
(d) Show also that the Compton wavelength, h/moc, is the wavelength of
radiation having a quantum energy equivalent to the rest energy of an electron.
13. What are the greatest wavelength and energy a photon can have and be
able to transfer half its energy to the recoil electron during a Compton colli-
sion? What are the direction and ma.gnitude of the velocity of the recoil electron
in this case? How important is the relativistic variation of mass with velocity
in this problem?
14. A beam of homogeneous x-rays with a wavelength of 0.0900 A is incident
on a carbon sca.tterer. The scattered rays are observed atan angle of 54° with
the direction of the incident beam. Find (a.) the wa.velength of the scattered
rays, (b) the energies of the incident and scattered photons, (e} the momenta
of the incident a.nd scattered photons, (d) the energy, momentum, and velocity
of the recoil electron, (e) the angle a.t which the electron recoils.
CHAPTER 7
ATOMIC SPECTRA AND ATOMIC STRUCTURE
7-1 Atomic spectra. The first problem that must be solved by a

theory of atomic structure is that of the line spectra of the elements. This
problem was discussed briefly in the fifth chapter; it will now be treated
in greater detail. A line, or emission, spectrum is produced when light
from a gas through which an electric discharge is passing, or from a flame
into which a volatile salt has been put, is dispersed by a prism or grating
spectrometer. Instead of a continuous band of colors, as in a rainbow
oras in the spectrum from an incandescent solid, only a few colors appear,
in the forro of bright, isolated, parallel lines. Each line is an image of
the spectrometer slit, deviated through an angle which depends on the
frequency of the light forming the image. The wavelengths of the lines
are characteristic of the element emitting the light, and eacb element has
its own particular line spectrum, which is a property of the atoms of that
element. For this reason, line spectra are also called atomic spectra. The
lighter atoms such as hydrogen and helium yield fairly simple spectra with
a relatively small number of lines, but for sorne of the heavier atoms the
spectra may consist of hundreds of lines. The lines may be in the ultra-
violet or infrared as well as in the visible range of wavelengths.
The spectrum of atomic hydrogen is especially interesting for historical
and theoretical reasons. In the visible and near ultraviolet regions, it
consista of a series of lines whose relative positions are shown in Fig. 7-1.
There is an apparent regularity in the spectrum, 'the lines coming closer
together as the wavelength decreases until the limit of the series is reached
at 3646 A. In 1885, Balmer showed that tbe wavelengths of the nine
lines then known in the spectrum of bydrogen could be expressed by tbe
very simple formula
n2
>. = B n2 - 4' (7-1)
where B is a constant equal to 3646 A, and n is a variable integer which

takes on the values 3, 4, 5, ... , for the Ha, H11, H-r, ... lines, respectively.
Rydberg, a few years later, found that the lines of the visible and near
ultraviolet hydrogen spectrum, now called the Balmer series, can be
described in a more useful way in terms of the wave number, or reciproeal
of the wavelength. The wave number is the number of waves in one
centimeter of light path in vacuum and is expressed in reciprocal centi-
meters, cm - 1 • For example, the wavelength of the Ha line is 6562.8 X 10-s
135
136 ATOMIC SPECTRA AND ATOMIC STRUCTURE [CHAP. 7
n=3 4 5 6 7 8
1 >.
~1
= 6563.I
Red
H~
.,¡ ~1 1111111111~1
4861.3 4340.5 4IQl.7
Blue Violet Ultraviolet
3646 A
FIG. 7-1. Diagram of the lines of the Balmer series of atomic hydrogen.
cm, and the wave number ¡¡ = 1/>. = 15237.4 cm- 1 • According to

Rydberg, the wave numbers of the Balmer lines a.re given by
¡¡ = Ru (_!_
22
- _!_)
n2
, (7-2)
where Ru is a constant with the dimension cm- 1 , called the Rydberg con-
stant f or hydrogen, and n takes on the values 3, 4, 5, . . . . The value of
the Rydberg constant for hydrogen has been determined empirically
from spectroscopic measurements and is
Ru = 109,677.576 ± 0.012 cm- 1 ; (7-3)
with this value of Ru, the wa.ve numbers and thus the wavelengths of the
linea of the Balmer series are given with great accuracy, as can be seen
TABLE 7-1
COMPARISON OF ÜBSERVED AND CALCULATED WAVELENGTHS OF THE
LINES OF THE BALMER SERIES IN THE HYDROGEN SPECTRUM
Wavelength Measured wavelength Wavelength

Line n measured in air, corrected to calculated from
>.: A vacuum, >.: A the Rydberg formula
H., 3 6562.85 6564.66 6564.70

H,, 4 4861.33 4862.68 4862.74
H,. 5 4340.47 4341.69 4341.73
H, 6 4101.74 4102.89 4102.93
H. 7 3970-07 3971.19 3971.23
Hr 8 3889.06 3890.16 3890.19
~ 9 3835.40 3836.48 3836.51
H, 10 3797.91 3798.98 3799.01
H, 11 3770.63 3771.85 3771.74
H. 12 3750.25 3751.31 3751.26
H1 13 3734.37 3735.43 3735.46
H,. 14 3721.95 3723.00 3723.03
H. 15 3711.98 3713.03 3713.06
7-1] ATOMIC SPECTRA 137
from Table 7-1. The wavelength measured in air has been corrected to
vacuum by means of the relation Xvac = µ.Xair, where µ, the index of refrac-
tion of air, is a known, tabulated function of wavelength. The wavelength
in vacuum is then compared with the value calculated from the Rydberg
formula.
The results illustrate the accuracy with which spectroscopic measure-
ments can be made as well as the remarkable agreement between the
values of the wavelengths measured experimentally and _those given by the
simple Rydberg formula with the empirically determined value of the
Rydberg constant for hydrogen. They also show that a theory of atomic
structure, in order to be acceptable, must predict the values of the wave-
lengths and the Rydberg constant w1th great accuracy.
In addition to the Balmer series, the hydrogen spectrum contains four
other series, one of which is in the ultraviolet region, and three in the
infrared. Thus, for hydrogen, the following spectral series, named after
their discoverers, are now known.
ii= RH ( _!_
12
- _!_),
n2
n = 2, 3, 4, ... (Lyman series, ultraviolet)
(Balmer series, visible and

ii= RH ( _!_
22
- _!_)
n2
, n = 3, 4, 5, ...
near ultraviolet)
ii= RH ( _!_
32
- _!_)
n2
, n = 4, 5, 6, ... (Paschen series, infrared)
ii= RH ( _!_ _ _!_), n

42 n2 = 5, 6, 7, ... (Brackett. series, infrared)
ii= Ra ( _!_
52
- _!_)
n2
, n = 6, 7, 8, ... (Pfund series, infrared)
Rydberg noted that for each line in the Balmer series (the only series
known at the time) the wave number could be expressed as the difference
between two terms, one fixed and one varying. He looked for similar regu-
larities in the spectra of elements other than hydrogen, and found that
although the spectra of the heavier elements are more complex than that of
hydrogen, a series of spectral lines can always be represented by the general
formula
ii=-,-----
R R
(7-4)
(m + a)2 (n + b)2
In this equation, R is a constant for the element, a and b are characteristic
constants for a particular series, m is an integer which is fixed for a given
series, and n is a varying integer whose values give the different lines of the
series. The líne spectrum of any .element consists of several series of lines,
and each line can be expressed as the difference between two terms.
The emission spectrum of an alkali metal is an example of a spectrum of
this type; the following four series are found,
R R
principal series: ii=
(1 + 8)2 - (n + p)2 ' n = 2,3,4, ... ,
R R
sbarp series: ¡; = (2 + p)2 - (n + 8)2 ' n = 2, 3, 4, ... ,
R R
diffuse series: ii=
(2 + p)2 - (n + d)2 ' n = 3, 4, 5, ... ,
R R
Bergmann series: ¡; = (3 + d)2 - (n + /)2' n = 4,5, 6, ....
In these series, R is the Rydberg constant for the element, and 8, p, d, f,

are certain constants which are characteristic of the alkali atoros. The
principal series is so called because it contains the brightest and most
persistent lines; the lines of the sbarp series appear relatively sharp in a
spectrograph picture, while the lines of the diffuse series are relatively
broad. In each case, as n becomes very large the lines of the series con-
verge to a limit given by the constant term.
The lines of the spectra of other elements also can be represented as
differences between terms, and the formula for the wave number can be
written generally as
¡; = T 1 - T2. (7-5)
For hydrogen the terms are given by
T = RH, n = 1, 2, 3, .... (7-6)

" n2
For other elements, the terms usually have a more complicated form, and
the first and second members of the formula are obtained from different
series. Ritz (1908) extended Rydberg's formulation of wave numbers by
means of his combination principie, which states that, with certain limita-
tions, the difference between any two terms gives the wave number of a
spectral line of the atom. For example, the difference between T 3 and T 8
for hydrogen gives the fifth line of the Paschen series. As a result of the
discovery of this combination principie, one of the chie{ problems of
spectroscopy became the representation of the lines of a spectrum as differ-
ences between terms, with as few terms as possible. The clarification of
the meaning of the terms then became a major theoretical problem.
In addition to emission spectra, the elements also have characteristic
absorption spectra. The latter are obtained when light with a continuous
7-21 THE BOHR THEORY OF ATOMIC SPECTRA 139
spectrum (as that from a filament larop) is passed through an absorbing

layer of the element and then analyzed with a spectrograph. In a photo-
graph of an absorption spectrum, light lines (absorption lines) appear •on
a dark continuous background. The absorption lines coincide exactly in
wavelength with corresponding emission lines and, in principie, every
emission line can also occur as an absorption line.
The field of atomic spectra is one oí the great branches of physics, and
the spectra of the elements have been analyzed in great detail. Only
sorne of the simplest features of atomic spectra have been presented here;
e.g., the multiplet structure of spectral lines has not been mentioned. In
roany eleroents, each line in the spectrum is single, but more commonly
the lines form doublets, triplets, or groups of even more components.
In such cases, a separate Rydberg formula must be written for each com-
ponent line; in the spectra of the alkali metals, for example, doublets
occur, so that six formulas instead of three are required for a complete
representation of the series in the visible region. There are also other
effects, such as the changes in the spectrum of an element when the atoros
radiating the light are in a magnetic or electric field. When a light source
is brought into a magnetic field, each emitted spectral line is split into a
number of components (the Zeeman effect). The extent of the splitting
dependa on the strength of the field and on the nature of the spectral
linea, whether they are singlets or multiplets. A splitting of the spectral
linea is also observed if the radiating atoros are in an electric field. But
in spite of the va.st amount of information which has been collected about
atomic spectra, the simple way in whieh spectral series can be expressed
points to the existence within atoros of a relatively simple and universal
method by which the characteristic atomic radiations are emitted.
7-2 The Bohr theory of atomic spectra and atomic structure. In a

series of epoch-making papera written between 1913 and 1915, Bohr
developed a theory of the constitution of atoros which accounted for many
of the properties of atomic spectra and laid the foundations for later
research on theoretical and experimental atomic physics. Bohr applied
the quantum theory of radiation as developed by Planck and Einstein
to the Rutherford nuclear atoro. His theory is based on the following
postula tes.
l. An atomic system possesses a number of states in which no emission
of radiation takes place, even if the particles are in motion relative to
each other, although such an emission is to be expected according to
ordinary electrodynamics. These states are called the stationary states
of the system.
2. Any emission or absorption of radiation will correspond to a transi-
tion between two stationary states. The radiation emitted or absorbed in
a transition is homogeneous, and its frequency v is determined by the

relation
(7-7)
where h is Planck's constant and W 1 and W 2 are the energies of the system
in the two stationary states.
3. The dynamical equilibrium of the system in the stationary states
is governed by the ordinary laws of mechanícs, but these laws do not hold
for the transition from one state to another.
4. The different possible stationary states of a system consisting of
an electron rotating about a positive nucleus are those for which the
orbits are circles determined by the relation
p = n (h/21r), (7-8)
where p is the angular rooroentum of the electron, h is Planck's constant,
and nis a positive integer, usually called the quarúum number.
These postulates are a combination of sorne ideas taken over from classi-
cal physics together with others in direct contradiction to classical physics.
Bohr solved the problem of the stability of a system of moving electric
charges simply by postulating that the cause of the instability, the emis-
sion of radiation, did not exist so long as the electron remained in one
of ita allowed, or stationary orbits. The erníssion of radiation was as-
sociated with a jump of the system from one stationary state (energy level)
to another, in accordance with the ideas of Planck's quantum theory.
Also, the laws of classical mechanics were presumed to be va.lid for an
atomic system in a stationary state, but not during a transition from one
such state to another. Finally, the fÓurth postulate, which separates the
allowed orbits of the electron from the forbidden ones by means of the
quantum condition, Eq. (7-8), has been shown to be in harmony with
Planck's condition for the energy of an oscillator. The angular momentum
is said to be quantized in these orbits. Thus, Bohr's theory is a hybrid,
containing sorne classical ideas and sorne quantal ideas. The justification
for the postulates was the astonishing success of the theory in accounting
for many of the experimental facts about atomic spectra. But, as will
be seen later, the hybrid nature of the theory eventually led to serious
difficul ties.
Starting with Bohr's postulates, we can calculate the values of the
energy that correspond to the stationary states of an atoro consístíng of
a nucleus and a single electron. Consider an atom with nuclear charge
Ze, where Z is the atomic nurober and e is the magnitude of the electronic
charge in electrostatic units. If Z = I, the atoro is the neutral hydrogen
atoro; for Z = 2, the system is the singly ionized helium atoro He+; for
Z = 3, the system is the doubly ionized lithium atom 1¡++
7-2] THE BOHR THEORY OF ATOMIC SPECTRA 141
In accordance with the fourth postulate, the electron revolves in a

circle about the nucleus and can have the angular momentum p, given by
h
p=mvr=n-, (7-9)
271"
where m is the mass of the electron, v is the linear velocity of the electron,
r is the radius of the orbit, and nis a positive integer. By the third postu-
late, the motion of the electron in an allowed orbit must satisfy the classical
mechanical laws. Hence, the centripetal force on the electron is provided
by the attractive electrostatic force on the electron:
mv 2 Ze 2
-r- = r2 . (7-10)
Equations (7-9) and (7-10) can be solved for the radius r and the velocity
v, and it is found that
(7-lla)
and
(7-llb)
The radii of the allowed orbits are proportional to n 2 , and increase in the
proportions 1, 4, 9, 16, ... , from orbit to orbit. For hydrogen, Z = I,
and if we insert the known values of m (rest mass), e, and h, we get, for
the smallest orbit in hydrogen (n = 1),
r = 5.29 X 10-9 cm,
a value that is in good agreement with estimates of atomic radii obtained

by other methods. The velocity of the electron in the smallest orbit is
about 2.2 X 10 8 cm/sec, and decreases as the radius of the orbit increases,
so that m may be taken as the rest mass of the electron.
The energy of the electron is partly kinetic and partly potential. If
the state of zero energy of the electron is arbitrarily chosen as that state
in which the electron is at rest ata very great distance ("infinitely far")
from the nucleus, then the potential energy of the electron is, by the usual
electrostatic formula,
Ze 2
U=--· (7-12)
r
The kinetic energy of the electron is
l 2 1 Ze 2
T=-mv = - - , (7-13)
2 2 r
142 ATOMIC SPECTRA. AND ATOMIC STRUCTURE [CHAP. 7
by Eq. (7-10), and T = -½U, a relation typical of motion under an

inverse-square law of force. The total energy is W = T + U or, for the
stationary state defined by the quantum number n,
1 Ze 2
Wn= -2---;¡:-= (7-14)
this is the energy of the atom when the electron is in its nth stationary or
quantum state. The quantum number n, it will be remembered, may
take on any integral value; n = 1, 2, 3, ....
From Eq. (7-14) it is seen that the larger the value of n, the smaller
numerically, but the larger in algebraic value, is the energy of the system.
The lowest value of W n is that corresponding to the first orbit. This is
known as the normal or gr<YUnd state of the atom, since it should be the
most atable state and the one ordinarily occupied by the electron.
According to Bohr's second postulate, a hydrogen atom radia.tes energy
when the electron jumps from one stationary state to another state of
lower energy. The difference in the two energies is emitted as a single
quantum of radiant energy whose frequency is given by the condition
(7-15)
If we insert for Wn 1 and Wn 3 the values given by Eq. (7-14) for the states
with quantum numbers n 1 and n 2 , we get for the frequency of the line
emitted when the electron jumps from state n 1 to state n 2
JI =
2
2r me
4
z2 (_!_ _ _!_) . (7-16)
h3 n~ nf
In terma of wave numbers this becomes
(7-17)
where we have defined the constant R.., as
(7-18)
In the case of hydrogen, Z = l and Eq. {'l-17) becomes
(7-19)
7-2] THE BOHR THEORY OF ATOMIC SPECTRA 143
Since n 2 < ni, the wave number is positive. If we set n 2 = 2,
n1 = 3, 4, 5, .... (7-20)
This equation has exactly the same forro as the empirical equation (7-2),
which was found to represent the lines of the Balmer series of the hydrogen
spectrum. Bohr's theory also gives the value of the constant coefficient
R.,, in terms of fundamental constants. Hence, R,,, can be calculated
theoretically, anda crucial test of Bohr's theory is the agreement or dis-
agreement between the theoretical value of the coefficient and the experi-
mental value of the Rydberg constant.
Before this comparison can be made, it is necessary to correct for an
approximation which has been made in the treatment so far. It has
been tacitly assumed in the derivation of the formula for the wave number
that the nucleus does not move, that it is infinitely heavy in comparison
with the electron. The subscript oo has been used to denote the value
of the coefficient under this assumption. Actually, for hydrogen, the
nucleus of the atoro is known to be about 1840 times as heavy as the electron
rather than infinitely heavy, and it does have a slight motion. The pre-
cision of spectroscopic measurements is so great that in testing the theory,
the motion of the nucleus cannot be neglected even though the nucleus
is so much heavier than the electron. The effect of the motion of the
nucleus is to replace the mass m of the electron by the reduced ma88,
+
mM/(M m), where Mis the mass of the nucleus.
The formula for the reduced mass is one which is important in many
problems involving the motion of two particles, and it can be derived
quite easily. If neither of the particles can be assumed to be at rest,
both particles will move about the center of mass of the two particles.
In the case of a one-electron atom, if the electron revolves about the
nucleus, its path in space will be a circle about the center of mass of the
combined system. At the same time, the nucleus revolves about the
center of mass in a smaller circle. Both particles must have the same
angular velocity w, since the center of mass must always lie on the line
between the two particles.
Let r be the distance of the electron from the nucleus, and r - x its
distance from the center of mass; x is then the distance of the nucleus
from the center of mass. Taking moments about the center of mass, we
have
Mx = m(r - x),
or
Mr
r-x=
M+m ·
The total angular momentum of the system about the center of ma.ss is
2 2
M mM
(
m r - )2
x w + 2
x w + M) 2 + (mMm
= [
(m + M) 2] r 2w = ( m mM
+ M ) r 2w.
The system acts as though a ma.ss m' = mM/(m + M) is revolving in a
circle of radius r, with an angular velocity w about the center of ma.ss.
The ma.ss m of the electron must therefore be replaced by m', that is, m
must be multiplied by the factor M/(m + M), and the reduced mass
of the electron is mM/(m +
M). Since M varies from atom to atom,
the factor 1/[1 + (m/M)l varies slightly, and thís accounts for the small
variation in the Rydberg constant for different elements.
The constant R.,, in Eqs. (7-17) through (7-20) should be replaced
therefore by the new constant
211"2 me4 1
Ra = ~ [1 + (m/MH)l (7-21)
1
= R.,, [l + (m/MH)]'
where MH is the ma.ss of the nucleus. If we insert into Eq. (7-21) the
presently accepted values of the constants e, m (rest ma.ss), e, h, and
m/MH, we get RH (theor) = 109681 as compared with the experimental
value of 109677.58. The two values agree to within 3 parts in 100,000,
which is considerably smaller than the uncertainty in the theoretical
value. This agreement representa a remarkable feat of theoretical physics
and is one of the great triumphs of the Bohr theory.
When the Bohr theory was proposed, only the Balmer and Paschen
series for hydrogen were known. The theory [Eq. (7-19)] suggested that
for values of n 2 different from 2 and 3, additional series should exist. The
search for these series yielded the Lyman series (1916), the Brackett series
(1922), and the Pfund series (1924), and in each series the wave numbers
of the lines were found to be those predicted by the theory.
Sorne of the circular orbits of the electron in the hydrogen atom are
shown on a relative scale in Fig. 7-2 together with the transitions that
give rise to typical lines of the different series in the hydrogen spectrum.
The spectrum of singly ionized helium, He+, was also successfully
explained by the Bohr theory. The experimental value of the Rydberg
constant for helium is Rae = 109722.269, and is in good agreement with
the theoretical value. A very important result is obtained when the
experimental values of the Rydberg constants for hydrogen and ionized
helium are compared. For helium
R R.,, R.,, (7-22)
He= [l + (m/MHe)] = [l + (m/4MH)]'
7-3] THE STATIONARY STATES OF AN ATOM 145
n=5
o 2 4 6 sx10-8 cm
FIG. 7-2. Scheme of tr.ansitions between eta.ble Bohr orbits in the hydrogen
atom (approximately to scale).
since the mass of the helium nucleus is very nearly equal to 4 times that
of the hydrogen nucleus. From Eqs. (7-21) and (7-22) it can be shown
that
(7-23)
This result is very close to that obtained in Section 2-4 by other methods,
and representa another triumph for the Bohr theory.
7-3 The stationary states of an atom. The Bohr theory, in addition

to accounting for the spectra of one-electron atoms, provides a physical
interpretation of the spectral terms. In the frequency condition, Eq. (7-7)
or Eq. (7-15), on substituting the wave number Ji = 11/c for the frequency,
we get
(7-24)
Comparison with the Rydberg series formula, Eq. (7-5), shows that a term
is equal to an energy state of a particular electron orbit divided by he. The
experimental determination of the spectrai terms of an element, therefore,

gives the values of the energy for the stationary states of the atom. It
will be remembered that the energy scale was chosen in such a way that
the electron has the energy W = O when it is completely removed from
the nucleus. Hence, the work that must be done in order to remove the
electron from its nth orbit to infinity is - W ,.. This amount of work is
called the separation energy and is a positive quantity. Apart from the
factor he, the terms are equal to the separation energies of the electron
in the given states. For the Iowest state (n = 1) of the atom, the ground
state, the separation energy is called the ionization energy, or the ionization
potential, which is, accordingly, equal to the largest term value of the atom.
The states which correspond to quantum numbers greater than unity are
called excited states of the atom because their energies are greater than
those of the ground state. A transition from the ground state to an
excited state can occur only when an atom receives an amount of energy
equal to the excitation energy for the excited state, for example, by electron
bombardment in a gas discharge tube.
The connection between t.erm values and energies was demonstrated
experimentally (1914) by the work of Franck and Hertz< 2 > on collisions
between slow electrons and mercury atoros. They observed that electrons
could not transfer energy to the mercury atoros unless the former had a
kinetic energy of at least 4.9 ev. This should mean, according to the
theory, that the mercury atom has an excited stationary state 4.9 ev
a.hove the ground state. If there is such a state, the energy of the excited
atom should be emitted as a light quantum whose frequency or wave-
length is determined by the Bohr frequency condition, Eq. (7-7). Now,
an energy difference of 1 ev is equivalent to a certain wave number obtained
by dividing 1 ev = 1.602 X 10- 12 erg by he; the result is 8066.1 cm- 1.
An excitation energy of 4.9 volts should correspond therefore to a wave
number i1 = 4.9 X 8066.1 = 39523 cm- 1, ora wavelength of >. = 1/ii
= 2530 A, which is a line in the ultraviolet region. There is a line of
wavelength 2535 A in the spectrum of mercury and the experimental
result can therefore be interpreted in accordance with the Bohr theory.
Thus, by electron impact, an excitation energy of 4.9 ev was transferred to
the mercury atom, and when an electron jumped back to the ground state
this excitation energy was emitted as a photon with the theoretically pre-
dicted wavelength. Many similar resulta have since been obtained with
other lines and gases, and Bohr's relationship between the excited states
and the term values of the spectral Iines has been completely verified.
In this way, the vs.lidity of Bohr's assumption about stationary states
has been verified by direct experimenta, and the concept of the stationary
states of an atoro has been strongly confirmed.
7-3] THE STATIONARY STATES OF AN ATOM 147
Energy leve!
(ev)
r--------
n
--------- --------
Wave nurnber
(cm- 1)
o o
-0.544
-0.850
-
10,000
-1.511
e,:,~""' .:: en
00C)Ol0 c8 "'C
~!!ta>
~ i -~e
1 -
20,000
-3.399 . !J :i j _
1
2 . .+f+......................____
o
30,000
40,000
50,000
60,000
70,000
80,000
90,000
100,000
- 13.597 110,000
F1G. 7-3. Energy level diagram of the hydrogen atom.

148 ATOMIC SPECTRA AND ATOMIC STRUCTURE [CHAP, 7
The stationary states, the terms, and the spectral lines of hydrogen can
be represented graphically in a very useful way by means of an energy
level diagram, as shown in Fig. 7-3. In this figure, the ordinates give
the energy, and the energy levels or terms R/n 2 are drawn as horizontal
lines. At the right, the wave number or term scale is given in units of
cm- 1, increasing from top to bottom; the value zero corresponds to the
complete separation of the nucleus and electron (n = oo). At the left
is a scale in electron volts beginning with th1> ground state as zero. The
electron-volt scale can be used directly to ot..tain the ionization potential
of a given level. Thus, to remove the electron from the ground state
(n = 1) to infinity (n = oo), "i.e., to ionize the atom in its ground state,
13.597 ev are needed.
A spectral line results from the transition of the atom from one energy
level to another and is represented in the energy diagram by a vertical
line joining the two levels. The length of the line connecting the two
levels is directly proportional to the wave number of the spectral line
(right-hand scale). A rough indication of the intensity of the spectral
lines may also be given by means of their thickness. Toward the top of
the diagram, the separation of the horizontal lines, which represent the
energy levels or terms, decreases and converges to the value zero as n -+ oo.
A continuous spectrum joins the term series here, indicated by cross-
hatching. The theoretical explanation for the continuous spectrum will
be given later. Spectra of other elements can be represented by similar,
but more complicated, diagrams.
The Bohr model can account for the lines of absorption spectra as well
as for those of emission spectra if it is assumed that an orbital electron
can absorb a light quantum only if the energy so absorbed raises the
electron into another allowed orbit. The absorption of light is then the
exact inverse of emission and every absorption line should correspond
exactly to an emission line, in agreement with experiment. Every emis-
sion line, however, is not found to correspond to an absorption line. For
example, the absorption spectrum of hydrogen gas under ordinary condi-
tions shows only the lines of the Lyman series. The Bohr theory can
account for this experimental result in the following way. Under normal
conditions, the atoms in hydrogen gas are in the most stable state (ground
state), with n = l. The energy differences between that level and all
other levels correspond to lines in the Lyman series; photons with less
energy, as all those corresponding to lines in the Balmer series, cannot be
absorbed because they cannot make the electron go even from the inner-
most (n = 1) orbit to the second orbit. According to the theory, there-
fore, only the lines of the Lyman series should appear in the absorption
spectrum of normal hydrogen. Only if the gas is initially excited, and
there is an appreciable number of electrons in orbits with n = 2, should
7-4) EXTENSION OF THE BOHR THEORY: ELLIPTIC ORBITS 149
absorption linea corresponding to the Balmer series appear. Balmer linea

are found in the absorption spectra of sorne stars in which the tem-
perature of the stellar atmospheres is so high that an appreciable fraction
of the hydrogen atoms is in the first excited state (n = 2).
7-4 Extension of the Bohr theory : elliptic orbits. The Bohr theory
in its simplest form predicted with great accuracy the positions of the
spectral lines of the neutral hydrogen atom and the singly ionized helium
atom. Refined spectroscopic analysis showed, however, that these lines
are not simple but have what is called fine structure, i.e., they consist of a
number of component lines lying close together. In terms of energy
levels, the existence of fine structure means that instead of a simple energy
level corresponding to a given value of the quantum number n, there are
actually a number of energy levels lying very close to one another. Som-
merfeld succeeded in part in accounting for these levels and for the result-
ing fine structure of the spectral lines of hydrogen and He+ by postulating
the existence of elliptic orbits as well as circular orbits, and by taking into
account the relativi:stic variation of the electron mass with velocity.
When it is assumed that an electron can travel in an elliptic orbit, there
are two coordinates which vary. In polar coordinates these are r, the
distance of the electron from the nucleus, which is at the focus of the ellipse,
and fP, the azimuthal angle. The two coordinates represent two degrees
of freedom to be quantized. The quantum condition that has previously
been discussed is not enough to fix unambiguously both axes of the ellipse,
and two quantum conditions are needed to fix the orbit. Sommerfeld and
Wilson introduced a new and more general postulate than the original
one of Bohr, namely, for the stationary states the so-called action integral,
extended over one period of the motion, must be an integral multiple of h.
This condition is expressed mathematically by the relation
f p¡ dq¡ = n¡h, (7-25)
where p¡ is the momentum associated with the coordinate q¡. The integral
has the units of action, i.e., erg-seconds, as does Planck's constant. When
q¡ = 'P, the angle of revolution, p¡ is the orbital angular momentum of
the system, denoted by P♦ to distinguish it from the radial momentum Pr•
According to classical mechanics, the angular momentum of any isolated
system is a constant. Then,
fP♦ d'P = P♦ t" d'P = 27rp.,

and
(7-26)
This condition is the same as that used by Bohr for bis circular coordinates.
The quantity n., called the azimuthal quantum nvmber, has been used
instead of n for reasons which will soon become apparent.
The radial momentum Pr is not constant and the integral in the second
quantum condition,
JPrdr = n,.h, (7-27)
can be calculated only by more complicated methods. The result obtained

is
(7-28)
In this equation, E is the eccentricity of the ellipse with semimajor and

semiminor axes a and b, respectively. The quantity nr is called the radial
quantum number. From Eqs. (7-28) and (7-26), it follows that
(7-29)
The eccentricity of the allowed ellipses is thus determined by the ratio

of the quantum numbers n. and nr. The quantity 1 - E2 can take on
values only between O and 1 because of the geometrical properties of the
ellipse. For 1 - E2 = 1, E = O, b = a, and the ellipse becomes a circle.
For 1 - E2 = O, n. = O, b = O, and the ellipse becomes a straight line.
Physically, the last result means that the electron would oscillate on a
straight line through the nucleus. This case was excluded in the Bohr
theory as having no physical meaning, so that n. was not allowed to be
equal to zero. Hence, n. may have the values 1, 2, 3, ... , and nr may
have the values O, 1, 2, 3,.... The sum nr + n. may have the values
1, 2, 3, ... ; it is therefore set equal to another integer n,
n = n. + nr; n = 1, 2, 3, ... , (7-30)
and n is called the principal quantum number. For any value of n, n.

can take on the values i, 2, 3, ... n, and nr is given by n - n •.
The discussion so far has clarified the relations between the quantum
numbers needed for the elliptical orbits and has defined the eccentricity
of the allowed orbits, but has not told anything about the size of the
ellipses, that is, a and b, or about the energies in the orbits. The latter
quantities are detennined, as in the case of the circular orbits, by com-
bining the quantum conditions with the expression for the centripetal
force provided by the attractive electrostatic force on the electron. The
7-4] EXTENSION OF THE BOHR THEORY: ELLIPTIC ORBITS 151
®
n =¡
n=2
n =3
F10. 7-4. Relative positions and dimensions of the elliptical orbits corre-
sponding to the first four values of the quantum number n. (Sommerfeld's
modification for the -hydrogen atom.)
analysis is more complicated than in the circular case and the details are
omitted. The results are
(7-31a)
ª = 41r m'e Z '
2 2
b = a n~, (7-31b)
n
21r 2m'e 4Z 2
Wn= - - - --
n2h2
(7-32)
where m' is the reduced mass of the electron. The values of the semi-
major axis a and the energy W n are determined by the principal quantum
number n, and are the same as the values for the circular orbit of radius
a and quantum number n. The shape of the orbit is determined by the
ratio n~/n, and the number of possible orbit shapes of the same energy
is increased over the case in which only circular orbits are considered.
Figure 7-4 shows the circular and elliptic orbits (drawn to sea.le) for
hydrogen with various n values, for n~ = 1, 2, and 3. According to this
model of the atom, a given spectral line may be produced by a number
of different but equivalent transitions between which no distinction can
so far be made. Although the system has two degrees of freedom, one
quantum number suffi.ces to determine the permitted energy values.
Such a system is said to be degenerate.
Sommerfeld showed that in the case of a one-electron atom, the degen-
eracy could be removed. That is, the previously equivalent states could
take on different energies when the special theory of relativity was applied
to the motion of the electron. The velocity of an electron in an elliptic
orbit varies (in a circle it is constant), being greater when the electron
is near the nucleus, and smaller when it is relatively far away. In accórd-
ance with the special theory of relativity, the mass of the electron varíes
in the orbit. If this relativistic effect is taken into account, the energy
n•
:• •- -
2 3 4 5
/ O 1 2 3
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
90,000
100,000
F10. 7-5. -Energy level diagram of the hydrogen atom, including fine structure.
7-4) EXTENSION OF THE BOHR THEORY: ELLIPTIC ORBITS 153
of the electron is found to depend on both n and n+. The mathematical

analysis is again quite complicated; the result is Sommerfeld's formula
wn,n+ = - 2,r2m'e4 z2 [1
n2h2
+ a2z2
n
(l.n+ - --ª--)]'
4n
(7-33)
where a = 2,re 2 /he = 1/137 is the fine structure constant. The dependence
on n+ is small because of the smallness of the fine-structure constant, and
the relativistic treatment of the elliptic orbits has the effect of replacing
a single energy leve} by several levels lying close together. Hence, the
frequency of the radiation emitted in the transition of an electron from
a level given by n 2 to the level n 1 will be slightly different according to
the various values of n+ whieh are possible. The small variations in
energy will result in small energy and frequency differences, with the
result that in the hydrogen spectrum, a group of slightly separated lines
appears, rather than a single line.
When the predictions of Sommerfeld's theory were compared with the
experimental resulta on the resolution of the Balmer lines and of several
He+ lines, it was found that the theory predicted too many components
for the fine structure. Agreement between theory and experiment was
obtained by the introduction of a selection rule which limited the number
of transitions. In the present case, the selection rule stated that only
those transitions are possible for which the quantum number n+ changes
by+ 1 or - 1:
.:.\n+ = ± l. (7-34)
Rules Iíke this one were found to be highly important in the theory of
atomio spectra. They were not an intrinsic part of the Bohr theory but
could be obtained by means of Bohr's correspondence principle. <3 > It
was known that for very large quantum numbers n, the behavior of the
atoro tends to that which would be expected on the basis of classical
electrodynamics. For example (cf. Chapter 5), for radiation of very
small frequency or very long wavelength, classical radiation theory was
satisfactory. But classical theory made possible the calculation of allowed
transitions, while the Bohr atomic theory did not. According to the
correspondence principie, the selection rules calculated classically, and
valid for very large quantum numbers, were also assumed to be correct
for small quantum numbers. This method for deriving the selection
rules worked, although it had the disadvantage of not being an integral
part of the Bohr theory.
The fine structure of the lines of the hydrogen spectrum can be illus-
trated in an energy level diagram somewhat more complicated than that
of Fig. 7-3. In Fig. 7-5, the levels with different values of n+ are now
drawn side by side at the same height, while states with equal values of
n,. and dífferent values of n are drawn above one another. For a given
value of n, ea.ch level has n sublevels corresponding to the n values which
n,. can have. In place of the vertical lines of Fig. 7-3 to represent transi-
tions from one value of n to another, there are oblique lines corresponding
to changes in n,. of ±1. The wave number for a transition is still given
by the vertical distance between two levels. To show the fine structure
lines, the energy and wave number sea.les would have to be very large,
beca.use of the very sroall spacing between the sublevels.
In addition to circular a.nd elliptic orbits, it is possible for electrons to
have hyperbolic orbits; these are associated with the continuous spectrum.
The energy of a one-electron atoro has been defined in such a way that
it has the value zero when the electron is at rest completely removed from
the nucleus. If the free electron has kinetic energy, its total energy is
positive and, by classical mechanics, its orbit is a hyperbola. According
to the Bohr theory, only the circular and elliptic orbits are quantized,
although an electron does not radia.te in any orbit. Hence, ali positive
values of the energy are possible and, extending from the limit of the dis-
crete energy Ievels, there is a continuous region of pQSitive energy values.
Radiation is emitted when a.n electron with positive energy jumps to a
lower state of positive energy or to a quantized state with nega.tive energy,
and the series of discrete lines in the spectrum is followed by a continuous
spectrum.
7-5 Failure of the Bohr theory: wave mechanics and the correct
quantum numbers. Although the use of rela.tivistic, elliptíc orbits in
the Bohr atom helped expla.in the fine structure of the spectra of hydrogen
and singly ionized helium, serious difficulties were met in the further
a.pplication of the Bohr theory. For two-electron atoros such as neutral
helium and singly ionized lithium, there were serious discrepancies between
theory and experiment. To try to a.ccount for the splitting of spectral
lines in a magnetic field, it was necessary to introduce a third quantum
number, the magnetic quantum number, m. As a result, three-dimensional,
or spa.tia.l, quantiza.tion wa.s necessary; the new qua.ntum number wa.s
rela.ted to the orientation of the elliptic orbits in space. In the presence
of a magnetic field, there is a coroponent of the angular rooroentum in the
dírection of the field. This coroponent, whose direction roa.y be taken
to be pa.ra.llel to the z-a.xis of a three-dimensional coordina.te system, wa.s
qua.ntized a.ccording to the condition
h
P• = m 21r' (7-35)
where m could have the va.lues - n., - n. +

1, ... O, 1, ... , n,., or
2n,. +
1 va.Iues in all. The mathema.tical treatment of the problem of the
7-5] FAILURE OF THE BOHR THEORY 155
stationary states again led to too many energy levels and the selection
rule,
Am = O, ±1, (7-36)
had to be introduced. The success of the theory in dealing with the effect
of a magnetic field on spectral lines was, however, still only partial.
Additional difficulties arose in connection with the multiplet structure
of the spectral lines of many elements. Multiplets are different in char-
acter from the fine-structure lines in that the multiplets can be quite
widely separated. The well-known sodium lines with wavelengths of
5895.93 A and 5889.96 A, respectively, are an example of a doublet, as are
the other members of the series to which the D lines belong. In the
magnesium spectrum, triplets are found, and the lines become more com-
plexas the elements move to the right in the periodic table. In the absence
of a magnetic field, multiplets could not be accounted for on the basis of
the quantum numbers n and n 9 • ·
The difficulties discussed above (and others not mentioned) could not
be resolved by further changes in the Bohr theory. It gradually became
apparent that the difficulties of the theory were caused by an essential
failure of the atomic model that was being used. The solution to the
problem carne in 1925 and the years following in the form of wave mechanics
or quantum mechanics. Sorne of the basic ideas of the newer theory will
be discussed later in this chapter. For the present it will suffice to state
that wave mechanics provides a theory which does not depend on the
classical picture of electrons revolving in planetary orbits around the
nucleus. The new theory yields all of the correct ideas and results of
the Bohr theory, such as stationary states with the correct energy values,
the explanation of the emission and absorption of radiation in terms of
transitions between stationary states, and the Rydberg-Ritz combination
principle for spectral lines. Moreover, wave mechanics solves many of the
problems which could not be handled by the Bohr theory. One of the
most important results of the new theory has to do with the quantum
numbers. It is found that in the case of the one-electron atom treated
three-dimensionally, three quantum numbers appear in a direct way in
the mathematical solution of the problem. These are the principal quan-
tum number n, the orbital quantum number l, which takes the place of
n♦, and the magnetic quantum number m 1, which replaces m. The values
which the new numbers can take on are
n = 1, 2, 3, ... ,
l = O, 1, 2, 3, ... , n - 1, (7-37)
mz = -l, -l + 1, ... , O, 1, ... , l - 1, l.

156 ATOMIC SPECTRA ANO ATOMIC STRUCTURE [CHAP, 7
The necessary selection rules also turned out to be a direct result of the
mathematical treatment and did not have to be introduced in an artificial
way.
Wave mechanics has two disadvantages. First, it does not allow the
structure of the atoro to be depicted in a familiar way. The concepts
from which the theory is built are more abstmet than those of the Bohr
theory and it is quite difficult to get an intuitive feeling for the problem
of atomic structure. Second, the mathematical methods are considerably
more advanced than those of the older theory. Hence, Bohr's mode of
description, from which the eBSential· features of a.tomic phenomena. can
readily be made out, is still often used. lt is possible to go quite a long
way with the Bohr model without coming into disagreement with facts,
provided that the quantum numbers of the wave mechanical theory are
used and that corresponding corrections are made in the conclusions re-
sulting from the Bohr theory. In this sense, the Bohr theory may be
considered to be a first approximation to wave mechanics, although both
its conceptual basis a.nd its mathematical methods have been repla.ced by
those of the newer theory.
To account for multiplet spectra and for the so-called "anomalous
Zeeman effect," wave mechanics was supplemented by the electron spin
hypothesis of Uhlenbeck and Goudsmit {1926) according to which the
electron was assumed to have a spin of magnitude ½h/2r and a magnetic
moment equal to (e/2mc) (h/2,r). The spin angular momentum of the
electron was represented by a quantum number that could have the
values +½ or -½, and it was necessary to introduce the new quantum
condition,
spin angular momentum = m. (h/21r) , (7-38)
where m. can have the values ½, -½. Thus, four quantum numbers are
needed to a.ccount for atomic spectra. The electron spin was, at first,
an arbitrary addition to wave mechanics, and notan intrinsic part of the
theory, but this difficulty was removed in 1928 by Dime. lt had been
known that the Schroedinger equation, the fundamental equation of wave
mechanics, was not relativistically invariant. Dime succeeded in obtain-
ing an equa.tion which satisfied the requirements of the special theory of
relativity, and showed that all four quantuni numbers followed from his
theory without the introduction of any arbitrary assumptions. The result
was a complete, consistent theory of atomic spectra and of many other
atomic properties.
7-6 Atomic theory and the periodic table. One of the most important
applications of modern atomic theory is to the interpretation of the struc-
ture of the periodic table in tenns of the physical ideas that have been
7-6] ATOMIC THEORY AND THE PERIODIC TABLE 157
developed about atomic structure. As a result of the wave mechanical

theory of the atom, it is possible to assign to any electron in an atom a
value for each of the four quantum numbers n, l, m1, and m., and an
electron can be identified by the particular set of values that it has. The
possible quantum numbers that an electron can have are limited by a
fundamental rule, the Pav1i exclusion principle, which states that no two
electrons in an atom can have the same four quantum numbers. When
this principie is applied to the electrons in an atom, two déductions can
be made. First, there can be no more than 2n 2 electrons with the same
quantum number n. Second, there can be only 2(2Z + 1) electrons with
a given value of l, These two rules can be shown to account completely
for the form of the periodic table.
Consider first the case n = 1. The only possible value of l is zero,
and m 1 must also be zero. The quantum number m, may be either +½
or -½, so that there can be at most two electrons with n = 1. All the
electrons which have the saroe total quantum number n are said to belong
to a shell, the shells being denoted by K, L, M, N, ... , for n = 1, 2, 3,
4, ... , respectively. The first or K-shell is full when it has two electrons.
For the second or L-shell, n = 2 and the shell can contain up to 8 electrons
(2n 2 = 8). It follows from the second rule that this shell can have no
more than two electrons with l = O, and no more than 6 electrons with
l = 1. As a result of the second rule, 8Ubgroups can forro within each
shell, all of the electrons in a subgroup having the saroe values of n and l.
The subgroups are designated by letters, as in Table 7-2.
An electron with n = 3 and l = 2 is called a 3d electron, and so on, the
number giving the value of n, the letter indicating the value of l. All
the electrons belonging to the same subgroup are called equivalent elec-
trons, and it is usual to denote the nurober of such electrons in an atoro
by an index. Thus, two Is electrons are written as ls 2, four 2p electrons
as 2p4, seven 4d electrons as 4d 7, and so on.
The structure of the periodic table can now be described in terms of
the possible numbers of electrons in the subgroups. The basic idea under-
lying this description is that the number of electrons in an atoro is equal
to the atomic number. The chemical properties of an atom are coropletely
determined by the number of outer electrons, and this number is in turn
fixed by the nuclear charge. Any atomic electron structure is formed by
TABLE 7-2
Orbital angular momentum

quantum number (1) : o 1 2 3 4 5 6
Electron designation: 8 p d f g h i
Atomic state: s p D F G H I
158 ATOMIC SPECTRA AND ATOMIC STRUCTURE [CHAP, 7
TABLE 7-3
ELECTRON SUBGROUPS FOR V ALUES OF THE PRINCIPAL QUANTUlol
NUMBER UP TO n = 4
Shell K L M N
n 1 2 3 4
Possible electrons 1s 2s 2p 3s 3p 3d 4s 4p 4d 4f
Maximum number
of electrons 2 2 6 2 6 10 2 6 10 14
Tot&l, 2n2 2 8 18 32
the addition of a single electron to the electronic structure characteristic

of the preceding atom in the periodic table. The quantum numbers
chosen for the added electron are such as to place this electron into the
most tightly bound state possible. This state is the one that, for the
given number of electrons, leads to the smallest value for the total potential
energy of the atom. The values of the potential energy can be calculated
from the wave mechanical theory of the atom.
The subgroups of electrons which are possible are shown in Table 7-3
for values of n up to 4.
In the case of hydrogen, which has one electron, it is clear from Table 7-3
that this electron in the ground state occupies a Is orbit; if the atom is
excited, the electron can occupy one of the higher orbits (states).
Helium has two electrons, both of which can go into Is orbits, making
the potential energy a mínimum. The configuration of the normal helium
atoro is then 1s 2 , and the K-shell is closed. The third electron needed for
the formation of the lithium atom goes into a 2s orbit (to keep the potential
energy a minimum); the electron configuration for this atoro is ls22s.
Beryllium, with four electrons, has the configuration ls 2 2s 2 ; the K-shell
is full, as is the s-electron subgroup of the L-shell. The procesa of adding
electrons continues until, at neon, the entire L-shell is filled. The procesa
can be continued until the entire periodic table is accounted for; the
electron configurations of the elements are listed in Table 7-4.
The filling of the K- and L-shells at helium and neon, respectively, closes
the first two periods of the periodic table (cf. Table 1-4). At argon the
first two subgroups of the M-shell are filled; at krypton the first two sub-
groups of the N-shell are filled, and at xenon the first two subgroups of
the 0-shell are filled. Thus, the rare gases are associated with the com-
pletion of either a shell or a subgroup. After each rare gas a new sub-
group begins, as required by the chemical properties of the elements.
Not only are the chemical properties of the atoms accounted for, but the
arrangement of subgroups explains the observed spectra of the elements.
7-7] ATOMIC THEORY AND CHARACTERISTIC X-RAY SPECTRA 159
For example, Table 7-4 shows that each alkali metal has one or more
closed shells and subgroups and a single outer electron. The electron
occupies the 2s, 3s, 4s orbits for Li, Na, K, respectively. All of the alkalis
have the same chemical valence and similar chemical properties, and all
have similar doublet spectra, as expected from the electron configuration.
The alkali earths have two outer electrons, and so on through the columns
of the periodic table. The last column in Table 7-4 gives the experimental
values of the ionization potential or separation energy; these values offer
clear evidence for the structure of the electron shells. The second electron
in the helium atom has an ionization energy of 24.5 ev as compared with
13.6 ev for hydrogen, and the helium electrons must therefore be much
more closely bound than the single electron in the hydrogen atom. The
third electron which is built into the lithium atom has the small ioniza-
tion potential of 5.4 ev; it must be at a considerably greater distance from
the nucleus and is, consequently, loosely bound. With this electron, the
construction of the second, or L-shell, begins. This shell is filled in the
neon atom with an ionization potential of 21.5 ev, and the M-shell starts
with sodium, with a potential of 5.1 ev. The trend of the experimental
values of the ionization potential is exactly that predicted by the theory.
It can now be said that modero atomic theory has solved the problem of the
interpretation, in basic physical terms, of the chemical properties of the
elements.
7-7 Atomic theory and characteristic x-ray spectra. The importance

of Moseley's work on the characteristic x-ray spectra of the elements was
discussed at sorne length in the fourth chapter. His results can be inter-
preted in terms of modero atomic theory and, in fact, the Bohr theory
can be used to give a very neat picture of the production of the character-
istic x-ray lines. It was seen in Section 4-4 that the wave numbers of
the K,. and L,. lines of the x-ray spectrum are given by the formulas
K,.: ji = R(Z - 1) 2 (i\ - J 2) ,

(7-39)
L,.: ji= R(Z - 7.4) 2 ( ;2 - ;

2
) · (7-40)
Equation (7-39) suggests that a K,. line is the result of the jump of an
electron from the L-shell (n = 2) to the K-shell (n = 1). Such a jump
can occur only ü one of the two electrons in the K-shell is missing. A
vacancy in the K-shell can occur if one of the two electrons in that shell
is struck by an energetic cathode ray and knocked either out of the atom or
into an unfilled outer shell. Subsequently, an electron from the L-shell
may drop into the vacancy in the K-shell with the emission of a quantum
TABLE 7-4
ELECTRON ÜONFIGURATIONS AND foNIZATION POTENTIALS OF ATOMS
-
8
z Element
K L M N o p Q Ionization
Is 2s 2p 3s 3p 3d 4s 4p 4d 4f 5s 5p 5d 6s 6p 6d 7s potential, ev
1
2
H
He
1
2
13.595
24.580 ~
a::
3 Li 2 1 5.390 ñ
4 Be 2 2 9.320 rr,
5
6
7
8
B
e
N
o
2
2
2
2
2
2
2
2
1
2
3
4
8.296
11.264
14.54
13.614
i
>
zi::,
9 F 2 2 5 17.418
10
11
Ne
Na
2
2
2
2
6
6 1
21.559
5.138 ~
a::
12 Mg 2 2 6 2 7.644 ñ
13 Al 2 2 6 2 1 5.984
14
15
Si
p
2
2
2
2
6
6
2
2
2
3
8.149
10.55
!
16
17
18
s
CI
Ar
2
2
2
2
2
2
6
6
6
2
2
2
4
5
6
10.357
13.01
15.755
ª
::o
l'!I
19 K 2 2 6 2 6 1 4.339
20 Ca 2 2 6 2 6 2 6.111
21 Se 2 2 6 2 6 1 2 6.56 g
22 Ti 2 2 6 2 6 2 2 6.83 ~
23 V 2 2 6 2 6 3 2 6.74 ....¡
24
25
Cr
Mn
1
2
2
2
2
6
6
2
2
6
6
5
5
1
2
6.764
7.432 J
26 Fe 2 2 6 2 6 6 2 7.90
27 Co 2 2 6 2 6 7 2 7.86
28
29
Ni
Cu
2
2
2
2
6
6
2
2
6
6
8
10
2
1
7.633
7.724
~
9.391
!5
n
30 Zn 2 2 6 2 6 10 2
31 Ga 2 2 6 2 6 10 2 1 6.00
32
33
34
35
Ge
As
Se
Br
2
2
2
2
2
2
2
2
6
6
6
6
2
2
2
2
6
6
6
6
10
10
10
10
2
2
2
2
2
3
4
5
7.88
9.81
9.75
11.84
1
~i::,
36 Kr 2 2 6 2 6 10 2 6 13.996 n
Rb
~
37 2 2 6 2 6 10 2 6 1 4.176
38 Sr 2 2 6 2 6 10 2 6 2 5.692 >
39 y 2 2 6 2 6 10 2 6 1 2 6.377 ~
40 Zr 2 2 6 2 6 10 2 6 2 2 6.835 ~
41 Nb 2 2 6 2 6 10 2 6 4 1 6.881 ~
42 Mo 2 2 6 2 6 10 2 6 5 1 7.131 ()
43
44
45
Te
Ru
Rh
2
2
2
2
2
2
6
6
6
2
2
2
6
6
6
10
10
10
2
2
2
6
6
6
(5)
7
8
(2)?
1
1
7.23
7.365
7.461
;
-<
46 Pd 2 2 6 2 6 10 2 6 10 8.33 ~
47 Ag 2 2 6 2 6 10 2 6 10 1 7.574 l'l
48 Cd 2 2 6 2 6 10 2 6 10 2 8.991 ~
::a
49 In 2 2 6 2 6 10 2 6 10 2 1 5.785 >
50 Sn 2 2 6 2 6 10 2 6 10 2 2 7.332
.....
(Continued) .....
O>
TABLE 7-4 (Continued) ~
K L M N o p Q Ionization
z Element
ls 2s 2p 3s 3p 3d 48 4p 4d 4f 5s 5p 5d 6s 6p 6d 7s potentia.l, ev
51 Sb 2 2 6 2 6 10 2 6 10 2 3 8.639
52
53
54
Te
1
Xe
2
2
2
2
2
2
6
6
6
2
2
2
6
6
6
10
10
10
2 6 10
2 6 10
2 6 10
2
~
2
4
5
6
9.01
10.44
12.127
i
5
55 Ca 2 2 6 2 6 10 2 6 10 2 6 l 3.893 ~
56 Be. 2 2 6 2 6 10 2 6 10 2 6 2 5.210 l:".I
57
58
La.
Ce
2
2
2
2
6
6
2
2
6
6
10
10
2 6 10
2 6 10 1
2
2
6
6
1
l
2
2
5.61
~
59 Pr 2 2 6 2 6 10 2 6 10 3 2 6 2 >
:t.
t,
60 Nd 2 2 6 2 6 10 2 6 10 4 2 6 2 6.3
61
62
Pm
Sm
2
2
2
2
6
6
2
2
6
6
10
10
2 6 10 5
2. 6 10 6
2
2
6
6
2
2 5.6 ~
63 Eu 2 2 6 2 6 10 2 6 10 7 2 6 2 5.67 ~
64 Gd 2 2 6 2 6 10 2 6 10 7 2 6 1 2 6.16
1
65 Tb 2 2 6 2 6 10 2 6 10 8 2 6 1 2
66 Dy 2 2 6 2 6 10 2 6 10 9 2 6 1 2 ?
67 Ho 2 2 6 2 6 10 2 6 10 10 2 6 1 2 ?
68 Er 2 2 6 2 6 10 2 6 10 11 2 6 1 2 7 ~
69 Tm 2 2 6 2 6 10 2 6 10 13 2 6 2
70 Yb 2 2 6 2 6 10 2 6 10 14 2 6 2 6.22
71 Lu 2 2 6 2 6 10 2 6 10 14 2 6 1 2 6.15
72
73
74
Hf
Ta.
w
2
2
2
2
2
2
6
6
6
2
2
2
6
6
6
10
10
10
2 6 10 14
2 6 10 14
261014
2
2
2
6
6
6
2
3
4
2
2
2
5.5
7.7
7.98
f
..¡
75 Re 2 2 6 2 6 10 2 6 10 14 2 6 5 2 7.87 ¡t
76 Os 2 2 6 2 6 10 2 6 10 14 2 6 6 2 8.7 .::!
77 Ir 2 2 6 2 6 10 2 6 1014 2 6 7 2 9.2
78 Pt 2 2 6 2 6 10 2 6 10 14 2 6 9 1 ? 9.0
79 Au 2 2 6 2 6 10 2 6 10 14 2 6 10 1 9.22 ~
80 Hg 2 2 6 2 6 10 2 6 10 14 2 6 10 2 10.434 ~
81 TI 2 2 6 2 6 10 2 6 10 14 2 6 10 2 1 6.106 ñ
82 Pb 2 2 6 2 6 10 2 6 10 14 2 6 10 2 2 7.415 -3
83 Bi 2 2 6 2 6 10 2 6 10 14 2 6 10 2 3 7.287 =
[,j
o
84 Po 2 2 6 2 6 10 2 6 10 14 2 6 10 2 4 8.43 ~
85 At 2 2 6 2 6 10 2 6 10 14 2 6 10 2 5
86 Rn 2 2 6 2 6 10 2 6 10 14 2 6 10 2 6 10.745 ~
t::I
87 Fr 2 2 6 2 6 10 2 6 10 14 2 6 10 2 6 1 (")
~
88 Ra 2 2 6 2 6 10 2 6 10 14 2 6 10 2 6 2 5.277
89 Ac 2 2 6 2 6 10 2 6 10 14 2 6 10 2 6 1 2?
2 10 2 6 2 2?
~
90 Th 2 2 6 2 6 10 2 6 10 14 6
91 Pa 2 2 6 2 6 10 2 6 10 14 2 6 10 2 2 6 1 2?
92 u 2 2 6 2 6 10 2 6 10 14 2 6 10 3 2 6 1 2 4 ~...
93 Np 2 2 6 2 6 10 2 6 10 14 2 6 10 4 2 6 1 2? (")
94
95
96
Pu
Am
Cm
2
2
2
2
2
2
6
6
6
2
2
2
6
6
6
19
10
10
2
2
2
6
6
6
10 14
10 14
10 14
2
2
2
6
6
6
10
10
10
5
7
7
2
2
2
6
6
6
1
1
2?
2
2?
;><
97 Bk 2 2 6 2 6 10 2 6 10 14 2 6 10 8 2 6 1 2? ~
98 Cf 2 2 6 2 6 10 2 6 10 14 2 6 10 9 2 6 1 2? [,j
~
~
....
~
o
N 1
Na
M 1
M., M,
L
LaL,L,
Fm. 7-6. Simple term scheme of x-ray levels showing transitions corre-
sponding to x-ray lines.
of ra.diation whose frequency is that of the Ka line. The electron which

goes from the L-shell to the K-shell moves in the electric field of the positive
nuclear charge and the negative charge of the electron remaining in the
K-shell. This electric field is equivalent to that of a positive charge of
magnitude (Z - l)e. The single K-shell electron tends to screen the
nucleus and makes its effective nuclear charge about one unit less than
the actual charge.
The explanation of the experimental results for the La line is, in prin-
cipie, the same as that for the Ka line. This time the vacancy is in the
L-shell and the ra.diation is emitted when an electron from the M-shell
(n = 3) drops into the vacancy. An M-shell electron is screened from
the nucleus by the other 7 L-shell electrons and the K-electrons, and the
nuclear charge is effectively reduced by about 7 units.
J ust as for the optical spectra, the stationary energy levels and the
spectral lines can be represented graphically in a term diagram, and a
simple example of such a diagram is shown in Fig. 7-6.
7-8 The basic ideas of wave mechanics. Sorne of the basic ideas of
wave mechanics are applied often in nuclear physics anda brief discussion
of these ideas is in order. The fundamental idea of wave mechanics is
that of the wave nature of matter. This concept arose as an extension of
the treatment of light sometimes as waves and sometimes as particles.
Light and x-rays have been treated as waves during the discussion of
experiments involving reflection, refraction, and diffraction. The design
and operation of optical and x-ray spectrometers as well as the interpreta-
tion of the results obtained with these instruments depend on properties
7-8) THE BASIC IDEAS OF W AVE MECHANICS 165
typical of waves. On the other hand, in the photoelectric effect, light

must be assumed to consist of particles (quanta, or photons). In the
Compton effect, the interaction between an x-ray and an electron is treated
as an elastic collision between two particles, and equations are written for
the balance of energy and momentum, typical particle properties. The
results of the calculation are expressed in terms of the shift in wavelength
because this is the quantity that is measured experimentally. It is clear
that sorne experiments involving radiation can only be interpreted on
the hypothesis that radiation is wave-like in nature, while for other experi-
ments radiation must be regarded as particle-like. This duality is shown
in the expression for the momentum of a photon
p = -hve = -·
h
>.
(7-41)
The momentum (a particle property) is related directly to the wavelength

(a wave property).
In 1924, de Broglie< 4 > suggested that matter also has a dual (particle-
like and wave-like) character. He assumed that the relation (7-41) holds
for electrons and other particles as well as for radiation. Although this
assumption was a bold and surprising one to make, it was amenable to
experimental tests. Suppose that the particles are electrons which have
been accelerated to a velocity v under the action of a potential difference V.
If v is small compared with the velocity of light, then
1 2 eV
2mv = 300' (7-42)
where m, v, and e are in cgs units and V is in volts. The wavelength

associated with the electron is, when Eqs. (7-41) and (7-42) are combined,
h
>. = ----:==== (7-43)
ymeV/150
If V is 100 volts, the value of >. should be 1.2 X 10-s cm, of the order of
magnitude of the distances between atomic planes in crystals. This fact
suggested that the existence of electron waves might be shown by using
crystals as diffraction gratings, as is done for x-rays, and experiments
based on this idea were done in 1927 by Davisson and Germer. <5 l On
bombarding a single crystal of nickel with an electron beam, they found
that the reflected beam showed maxima and mínima which could be
explained in terms of the diffraction of electron waves. The analysis of
the experiments was similar to that used for x-ray diffraction by crystals,
and the wavelength was determined from the Bragg equatíon. The
166 ATOKIC SPECl'RA AND ATOKIC STRUCTURE [CHAP. 7
quantitative agreement between the predictions of the de Broglie hypoth-

esis a.nd the experimental resulta was rem&rkably good.
Diffraction patterns were also obtained< 6> when electrons passed
through thin films of matter such as metal foil or mica, and the patterns
were exactly a.nalogous to thoee obtained in experimenta on the diffraction
of x-rays. In Fig. 7-7, the similarity between the effect of x-rays and
that of electrons is so striking as to prove conclusively the wave properties
of electrons. Particles of atomic size, such as helium atoms and helium
molecules, have also been found to show diffraction effecta, a.nd there is
no doubt about the correctness of de Broglie's hypothesis.
The dual wave and particle nature of matter can be expresaed mathe-
matically by mea.ns of a wave equation first derived by Schroedinger. <7>
This equation can be derived quite easily for the case of a particle, or a
beam of particles ali with the same energy, moving in a.n electrostatic
field of force. The usual three-dimensional wave equation in rectangular
Cartesian coordinates is
a2+ a2• a2• 1 a2+

clz2 + cly2 + clz2 = u2 dt2 ' (7-44)
where 'fr (x, y, z, t) is the amplitude of the wave associated with the particle
a.nd u is-ita velocity. If we desire a solution that will represent sta.nding
waves (such as thoee on a string fastened at both ends), we may write
'fr in the form
(7-45)
where t/, is a function of x, y, a.nd z, but not of the time t, and II is the
frequency. If the expression for 'fr is inserted into Eq. (7-44), we get
a2.¡, a2.¡, a2.¡, 41"21'2 (2r)2
ax2 + cly2 + clz2 - - u2 t/, - - A t/,, (7-46)
where A = u/v·is the wavelength. In view of the wave nature of particles,

we may write for the energy W and the momentnm p of a particle
W= lw, (7-47a)
1,,
p = ¡• (7-47b)
a.nd the wavelength is then
(7-47c)
We assume now that the speed v of the particles is much less than that of
7-8] THE BABIC IDEAS OF WAVE MECHANICB 167
(a) (b)
Fm. 7-7. (a) Diffraction pattem obtained from a gold crystal with x-rays
of wavelength 0.709 A. (Courtesy of Mr. O. F. Kammerer, Brookhaven.National
Laboratory.) (b) Diffraction pattem obtained from a gold crystal with electrons
of momentum corresponding to a wavelength of 0.055 A. (Courtesy of Dr. S.
Bauer, Comell University.)
light. The kinetic energy of the particle is the difference between the total
energy W and the potentiaJ energy U, and we can write
½mv2 = W- U. (7-48)
Then
p = mv = v'2m(W - U)
and, from Eq. (7-47c),
>. = h (7-49)
v'2m(W - U)
On substituting this expression for A in Eq. (7-46), we get
a2i/, + a2,t_ + a2i/, + 8r:.im (W - U)if, = O. (7-50)

ax2 ay2 az2 h2
Equation (7-50) is the Schroedinger equation, which gives the wave
function as a function of the coordinates; the potential energy is a.lso a
function of the coordinates.
Fxo. 7-8. Representation of a one-dimensional standing electron wave in a

Bohr orbit.
The mathematical problem which must be solved consista of inserting

the appropriate expression for the potential energy into the Schroedinger
equation and solving for the wave function "1. The on~lectron atom is
treated by setting U = -Ze 2 /r, where r = v
x 2 + y'J. + z2; actually, the
Schroedinger equation would be written in spherical polar coordinates in
this case. It is found that physic:i.lly significant solutions for "1 for bound
states of the atom exist only when W has the values
2r2 me 4 2
W,. = - n2h2 z' (7-51)
where nis an integer. These values are the same as those found for the
stationary states in the Bohr theory, but the result is obtained from the
mathematical formulation of the experimentally proven idea of the wave
nature of matter rather than from a hybrid set of classical-quantal
assumptions.
Bohr's quantum conditions also follow from the idea of the wave nature
of matter. As a simple example, consider the electron in the hydrogen
atom as a standing wave extending in a circle around the nucleus. In
order that the electron wave may just fill the circumference of a circle,
as in Fig. 7-8, the circle must contain an integral number of wavelengths,
or
2rr = n~, (7-52)
where r is the radius of the circle and n is an integer. By de Broglie's

hypothesis, ~ = h/p = h/mv, and Eq. (7-52) becomes
h h
21rT = n -,
mv
or mvr=n-·
2r
(7-53)
But mvr is the angular momentum of the electron regarded as a particle,

7-8) THE BASIC IDEAS OF W AVE MECHANICS 169
and we see that the wave mechanical picture leads naturally to Bohr's
postulate that the angular momentum is equal to an integral multiple of
h/21r. This type of proof can be extended to the other quantum condi-
tions, and the latter appear as a consequence of the condition for a stand-
ing wave rather than as arbitrary assumptions. The concept of the
electron as a standing wave rather than as a particle revolving in an orbit
also removes the difficulty that in the Bohr theory the electron travels in
its orbit without radiating.
It can also be shown that in wave mechanics the emission of a photon
of frequency " is a logical consequence of an energy change of the atom ¡
the mechanism of emission does not have to be introduced as a separate
postulate, as in the Bohr theory. Thus, the difficulties inherent in Bohr's
postulates are all satisfactorily removed by wave mechanics.
The concept of the dual particle and wave nature of matter leads to
another important result. According to classical mechanics, a particle
has a position and momentum which can, in principie, be determined with
any desired accuracy. But a wave is extended throughout a region of
space, and the location of an electron, considered as a wave, seems to
present sorne difficulty. The question arises: If an electron must be
described sometimes as a wave and sometimes as a particle, is it possible
to know exactly where an electron is in space at sorne given instant? The
answer to this question is given by Heisenberg's uncertainty principle, <S>
which states that the position and momentum of a particle cannot be
determined simultaneously with any arbitrary desired accuracy. These
quantities can be determined only with accuracies limited by the relation
(7-54)
where ax is the error in the determination of the position and t:.p is the
error in the momentum. The product of the two errors is at least of the
order of magnitude of Planck's constant. A similar relation holds for
the energy of a particle and the time,
t:.tt:.E ~ h. (7-55)
The condition of Eq. (7-54) means that the more precisely the position
of a particle is determined, the less precisely ,can the momentum be meas-
ured, and conversely. The last condition means that the more accurately
the time at which an atomic event occurs is determined, the less accurately
can the energy change of the atomic system be determined, and conversely.
The meaning of the uncertainty principie can be illustrated by an
idealized experiment. Consider the following attempt to measure simul-
taneously the position and momentum of an electron with an imaginary
microscope of very high resolving power. It has been shown in wave
170 ATOMIC SPECTRA ANO ATOMIC STRUC'I'URE (CHAP, 7
optics that the resolving power of a microscope is given approximately by
(7-56)
where Ax is the distance between two points which can just be resolved by
the microscope, >. is the wavelength of the light used, and A is the half-
angle of the leos used, as in Fig. 7-9. Then Ax is the uncertainty in the
determination of the position of the electron; to make Ax as small as pos-
sible the wavelength must be very small and either x-rays or 'Y-raye must
be used. N ow the position of an electron can be measured with a microscope
only ü at the very least one photon is deflected from its initial direction
into the microscope. But when this photon is scattered it gives to the
electron, because of the Compton effect, a momentum of the order of hv/c.
The change, Ap, in the momentum
of the electron cannot be determined
exactly because the scattered photon
can enter the microscope anywhere
within the angle A. The uncer-
Lens
tainty in the momentum is of the
order
hv . A
Ap ~ -;-sm . (7-57)
From Eqs. (7-56) and (7-57), it

follows that F10. 7-9. Imaginary experiment in
Ap Ax~ h,
connection with the uncertainty prin-
cipie.
which is the Heisenberg relationship.
This discussion shows that one of the causes of the uncertainty relation-
ship is the fe.et that the procesa of measurement itself perturbe the particle
whose properties e.re being measured. From an operational viewpoint,
the exact location of an electron has therefore no real meaning, and there
is no real point in thinking of an atom in the same way that we think about
a macroscopic planetary system. This result gives a clue to the failure
of Newtonian mechanics when it is applied to atomic physics.
The interpretation of the Schroedinger wave function .¡, presented a
problem until Bom (1926) showed that logical consistency and agreement
between theory and experiment could be obtained provided that .¡, is
regarded as a measure of the probability of finding a particle in a given
region. Thus, ltl 2 dx dy dz, where the vertical lines denote the absolute
magnitude of 'f, is the probability that the electron will be found in the
volume element dx dy dz about the point (z, y, z). Other physical resulte
can be expressed in terma of ezpectation valuea; these e.re mathematical
7-9] THE SOLUTION OF THE SCHROEDINGER EQUATION 171
expectations, in the sense of probability theory, for the results of a single

measurement, or for the average of a large number of experiments. This
interpretation is in accord with the statistical nature of the results of
experimenta on atomic systems.
The success of wave mechanics is not limited to the examples that have
been discussed here. The quantitative treatment of many problems in
nuclear physics depends on the application of wave mechanics. Molecular
physics, the physics of the solid and liquid states, statistical mechanics,
the theory of chemical reaction rates, and many electric and magnetic
phenomena can also be treated by means of the ideas and techniques of
wave mechanics. Many examples of these applications will be found
in the books cited in the general references at the end of this chapter.
7-9 The solution of the Schroedinger equation: some useful examples

and resulta. l. Particle in a box. The solution of the Schroedinger equa-
tion for significant physical problems usually involves mathematical
methods beyond the scope of this book. There are, however, a few simple
problems whoee solutions show some of the basic differences between
classical mechanics and quantum mechanics. One of these problems is
that of a free particle in a "one-dimensional box." We considera particle
of masa m which can move freely over the range O < x < a, but is
reflected when it strikes either end of the range. These conditions can
be represented in Schroedinger's equation by assuming that the potential
energy U is zero for O < x < a, and that the wave function 1/, vanishes
at x = O and at x = a. The Schroedinger equation for the particle is
then
(7-58)
If we set
(7-59)
the general solution is
1/,(x) = A sin kx + B cos kx. (7---00)
The boundary condition 1/, = O at x = O requires B = O, and the con-

dition at x = a requires
"1(a) = A sin ka == O. (7-61)
We cannot take A = O because there would then be no solution. Hence,

we must have sin ka = O, or
n = 1, 2, 3, .... (7-62)
The only allowed solutions of the wave equation are the functions
·'·
'l'n (X ) = A s1·n n1rx, (7-63)
a
with n related to k by Eq. (7-62). That relation and Eq. (7-59) require
that the energy W can have only the values
(7-64)
An energy level W,. corresponds to a value of n, and to a wave function

,¡,,,. given by Eq. (7-63); each value of W,. is called an eigenvalue or proper
value, and the corresponding wave function ,J;,. is called an eigenfunction
or proper function. The interpretation of the wave function requires that
r
Ío
l,J;n(x)l 2 dx = A2 fo
O sin 2 n1rx dx
a
= 1, (7-65)
or that A = v'2/a, and

,J;n(x) =
i . sm - ·
a
n1rx
(7-66)
The wave functions are said to be normalized when they satisfy Eq.
(7-65). The wave functions have another important property, that of
orthogonality: for n = m, it is easily verified that
(7-67)
while for n "F m,
Equation (7-64) is a quantization condition which introduces the quan-

tum number n; it comes from the physical boundary conditions and gives
rise to the quantized values W n of the energy. The momentum is also
quantized; for each energy value W n, there is a momentum Pn given by
p! = 2mWn,
or
nh
p
n
=· -2a = -a1r n -21r
h
= -1r
a
nli. (7-68)
The first three wave functions are plotted in Fig. 7-10.

We can calculate the probability that the particle, in a given state of
energy, will be located within any element dx of the one-dimensional box.
·an::,
j
íc:
l\f<..---'---'--+----'\-----1'--~~---+----
.,3
<.>
e:
.E
1;
::=:"
Fm. 7-10. The first three wave functions for a particle in a one-dimensional
box.
Quantum
i{lt!f\7\
~ z 1 1
¡;=¡¡ 4
1
3
1
2
2
:i
3
4
5
¡¡
z
¡;=
1 -x
Frn. 7-11. The probability distribution for finding a particle of given energy
at points in a one-dimensional box (case n = 3).
According to quantum mechanics, this probability is
liJ,(x)I 2 dx = -2a sm
. 2 n1rx
-
a
dx, (7-69)
which is a wave-like function; it is shown graphically in Fig. 7-11 for the

state defined by n = 3, and W 3 = 9h 2 /8ma 2 • There are actually nodes,
at x/a = ! and x/a = i in this case, where the probability is zero.
Thus, the quantum mechanical result is very different from the cla.ssical
one for which the probability is the same for any element dx.
If the particle is confined within a three-dimensional box with edges of
lengths a, b, e, parallel to the x-, y-, and z-axes, respectively, the normalized
wave functions for the stationary states and the energy levels are found
to be
(7-70)
(7-71)
where n 1 , n 2 , n 3 denote any set of three positive integers.

2. The hydrogen atom: angular momentum; -parity. The hydrogen atom

is treated in quantum mechanics by putting U = - e2 /r into the Schroe-
dinger equation (7-50). The solution of the equation is complicated;
one of the results obtained has already been mentioned, namely, that the
energy has the eigenvalues given by Eq. (7-51). Another very important
result involves the angular momentum a.nd its quantization. It is found
that the angular roomentum L (vector quantities are indicated by bold-
face letters), has the magnitude, L, given by
L = vl(l + 1) 2~, (7-72)
where l is the orbital quantum number, which may take on the values
given in the second of Eqs. (7-37). The angular momentum is a vector
quantity and its direction must be specified, as well as its magnitude.
The magnetic quantum number m 1, whose possible values are given in
the third of Eqs. (7-37) defines a limited number of orientations which
L may take up with respect to a specified axis. These orientations are
shown, for l = 3, in Fig. 7-12; in this case, L = yl2h/2'11". The orbital
angular momentum is said to be spatially quantized in terms of the
quantum numbers m,. The angular momentum vector L can never lie
entirely along a chosen axis but only at a minimum inclination given by
cos- 1 (l/l + 1) 112 , which a.pproaches zero in the liroit of la.rge values of l.
The projection of L along a specified axis, usually taken as the z-axis, has
magnitudes which a.re given by
L. = m1 (2h11) . (7-73)
According to quantum mechanics, an electron also has a spin angular

momentum S characterized by a spin quantum number which can take
on the values ±½h/21r. More precisely, the magnitude of the spin
angular momentum is
S = v's(s + 1) (;1r), 8
1
= -2'. (7-74)
the component of the spin angular momentum along a specified axis s.

is quantized according to the rule
(7-75)
The orbital and spin motions of the electron are combined, or coupled,
m=
+1 l =3
-1
-2
-3
-vTI
F10. 7-12. Spatial quantize.tion of the orbital angular momentum.
according to the rules of vector addition to give a resultant angular

momentum J characterized by a quantum number j;
J = L+ S; (7-76)
for a given value of l, j can take on only the values l ± ½-

When changes in atomic states are produced by the action of two or
more electrons, the value of the total angular momentum of these electrons,
denoted by J, depends on the coupling between the orbital and angular
momenta of the individual electrons. The study of atomic spectra has
shown that the predominant type of coupling is that in which all the orbital
angular momentum vectors L; of the electrons combine to form a resultant
L and, independently, all the spin angular momentum vectors S; combine
to forma resultant S. The total angular momentum is then given by the
relation
J= L+S = ¿L; + ¿S;. (7-77)
i i
This type of coupling is called L-S, or Russell-Saunders coupling after its

discoverers. The total angular momentum is quantized and there are
appropriate rules for the possible values which may be taken on by the
quantum numbers corresponding to L, S, and J.
We have seen that the quantum-mechanical study of the hydrogen-
like atom has led to certain quantum numbers which apply to physical
quantities, such as energy and angular momentum, well known in classical
mechanics. There is another quantum number, the parity of a wave
function ,/t, which is related to the symmetry properties of ,/t and has no
relation to a classical physical quantity. The parity of a wave function

refers to its behavior under a simultaneous reflection of the space co-
ordinates, x --+ -x, y --+ -y, z - -z. We may distinguish between two
ideal cases
t/1(-x, -y, -z) = +
t/l(x, y, z), (7-78a)
t/1(-x, -y, -z) = - t/l(x, y, z); (7-78b)
the wave function may be invariant under the reflection, or it may change
sign. Equations (7-78) can be combined in the forro
1/1(-x, -y, -z) = Pi/t(x, y, z), (7-79)
where P = ±1. We can treat Pasa quantum number, and the property
it defines is called the parity of the system. The parity is then quantized
and has the quantum number P; when P = I, the system is said to have
even parity; when P = -I, the system is said to have odd parity. As
for other quantum numbers, there are selection rules involving the change
of parity in a transition.
In the case of a hydrogen-like atoro, it is found that the parity is re-
la.ted to the orbital quantum number l,
P= (-1) 1• (7-80)
Since the parity is defined in terma of the Schroedinger wave function 1/t,
which is a purely quantum-mechanical concept, it can have no classical
analog.
REFERENCES
GENERAL
F. K. R1cHTKYER, E. H. KENNARD, and T. LAURITSEN, Introduction to
Modero Physia, 5th ed. New York: McGraw-Hill, 1955, Chapters 5, 6, 7.
A. P. FRENCH, Principlu o/ Modero Physia. New York: Wiley, 1958, Chapters
5, 7, 8.
R. B. LEIGHTON, Principle8 of Modero Physia. New York: McGraw-Hill,
1959, Chapters H.
EXPERIMENTAL TECHNIQUES
G. P. HARNWELL and J. J. L1v1Noooo, Experimental Atomic Phy8ics. New
York: McGraw-Hill, 1933, Chapters 5, 7, 8.
J. B. HoAG and S. A. KoRFF, Electron and Nuclear Phy8ics, 3rd ed. New York:
Van Nostrand, 1948, Chapters 3, 7.
G. R. HARRISON, R. C. LoRD, and J. R. LooFBOUROW, Practical Spectro8Copy.
New York: Prentice-Hall, 1948.
REFERENCES 177
ATOMIC SPECTRA
L. PAULING and S. GounsMrT, The Structure of Atomic Spectra. New York:
McGraw-Hill, 1930.
R. BACHER and S. GounsMIT, Atomic Energy States. New York: McGraw-
Hill, 1932.
11. E. WHITE, Introduction to Atomic Spectra. New York: McGraw-Hill, 1934.
G. HERZBERG, Atomic Spectra and Atomic Structure. New York: Dover, 1946.
G. W. SERIES, The Spectrum of Atomic Hydrogen. Oxford: Clarendon Press,
1957.
AT0MIC THE0RY AND ITS APPLICATIONS

N. BoHR, Atomic Theory and the Description of Nature. Cambridge: Univer-
sity Press, 1934
M. BoRN, Atomic Physics, 6th ed. London and Glasgow: Blackie; New York:
Hafner, 1957. ·
U. FANO and L. FANO, Basic Physics of Atoms and Molecules. New York:
Wiley, 1959.
H. SEMAT, Introduction to Atomic and Nuclear Physics, New York: Rinehart,
1954.
D. C. PEASLEE, Elements of Atomic Physics. New York: Prentice-Hall, 1955.
G.P. HARNWELL and W. E. STEPHENS, Atomic Physics. New York: McGraw-
Hill, 1955.
J. C. SLATER, Quantum Theory of Matter. New York: McGraw-Hill, 1951.
J. C. SLATER, Quantum Theory of Atomic Structure. New York: McGraw-Hill
Vol. I, 1960; Vol. II, 1960.
W. KAUZMANN, Quantum Chemistry. New York: Academic Press, 1957.
WAVE MECHANICS (QUANTUM MECHANICS)

W. HEJTLER, Elementary Wave Mechanics, 2nd ed. Oxford: Clarendon Press,
1956.
N. F. MoTT, Elements of Wave Mechanics. Cambridge: University Press, 1958.
R. H. D1cKE and J. P. WITTKE, Introduction to Quantum Mechanics. Reading,
Mass.: Addison-Wesley, 1960.
J. L. PowELL and B. CRASEMANN, Quantum Mechanics. Reading, Mass.:
Addison-Wesley, 1961.
D. BoHM, Quantum Theory. New York: Prentice-Hall, 1951.
N. F. MoTT and I. N. SNEDDON, Wave Mechanics and lts Applications. Oxford:
Clarendon Press, 1948.
L. I. ScmFF, Quantum Mechanics, 2nd ed. New York: McGraw-Hill, 1955.
F. MANDL, Quantum Mechanics, 2nd ed. London: Butterworth; New York:
Academic Press, 1958.
P. A. M. DIRAC, The Princi-pl,es of Quantum Mechanics, 4th ed. Oxford:
Clarendon Press, 1958.
L. D. LANDAU and E. M. LusHITZ, Quantum Mechanica: Non-Relativiatic
Theory. London: Pergamon Press; Reading, Mass.: Addison-Wesley, 1958.
PARTICULAR
l. N. BoHR, "On the Constitution of Atoms and Molecules," Phil. Mag. 26, 1,
476, 857 (1913); "On the Quantum Theory of Radia.tion and the Structure of
the Atom," ibid., 30, 394 (1915).
2. J. FRANCK and G. HERTZ, G. phys. Ges., V erhandlungen, 16, 457, 512 (1914);
also A. P. FRENCH, op. cit. gen. ref., Chapter 12.
3. N. BOHR, On the Application of the Quantum Theory to Atomic Structure.
Part l. The Fundamental Postulates. Cambridge: University Press, 1924; also
Z. Physik 13, 117 (1923).
4. L. DE BROGLIE, in L. DE BROGLIE and L. BRILLOUIN, Selected Papers on W ave
Mechanics. London: Blackie, 1929.
5. C. DAv1ssoN and L. H. GERMER, "Diffraction of Electrons by a Crystal of
Nickel," Phys. Rev. 30, 705 (1927); Proc. National Acad. Sci., 14, 217, 619
(1928).
6. G. P. THOMSON, "Experiment on the Diffraction of Ca.thode Rays," Proc.
Roy. Soc. (London), All7, 600 (1928); All9, 663 (1928).
7. E. ScHROEDINGER, Collected Papers on Wave Mechanics. London: Blackie,
1929.
8. W. HEISENBERG, The Physical Principles of the Quantum Theory. Chicago:
University of Chicago Press, 1929; New York: Dover.
PROBLEMS
l. Three lines in the Balmer series of hydrogen ha.ve the wavelengths 3682.82 A,
3669.43 A, and 3662.24 A, respectively, mea.sured in a.ir. Wha.t is the value
of n for each line? Take RH as 109,677.576, and the index of refraction of air
as 1.0002837.
2. The hydrogen isotope oí ma.ss 2, deuterium, has a spectrum in which the
Balmer lines are displaced slightly from the lines of ordinary hydrogen (mass 1).
Find the value of the Rydberg constant for deuterium, then calculate the
difference between the wavelengths of the first three Balmer lines of hydrogen
and deuterium.
3. Show tha.t the frequency of revolution of an electron in a Bohr orbit with
quantum number n is
llorb =
Show next that if an electron makes a transition from a state for which n = ni to
a state for which n = n2, with n1 - n2 = 1, then the frequency II of the einitted
radiation is intermediate between the frequencies of orbital revolution in the
initia.l a.nd final states. What is the rela.tionship between the einitted a.nd orbital
frequencies when n1 a.nd n2 are both very large?
4. Ca.lculate the energy of each of the five lowest stationary states (energy
levels) of the hydrogen a.tom, in wa.ve-number units, ergs, electron volts, and
calories per gram.
5. The best values of RH and RHe are now considered to be RH = 109677.576
cm- 1 a.nd RHe = 109722.267 cm- 1 • Calcula.te a.ccurately the value of the ratio
PROBLEMS 179
of the mass of the hydrogen atom to that of the electron. Use the result to
calculate the value of e/m for the electron.
6. In an experiment of the Franck-Hertz type, hydrogen was bombarded
by electrons with known kinetic energy. The temperature of the hydrogen
was high enough so that the gas was dissociated into atomic hydrogen. It was
observed that r&diation was emitted when the electrons h&d the following
discrete energies: 10.19 ev, 12.08 ev, 12.75 ev, 13.05 ev, 13.12 ev, 13.31 ev.
What is the wavelength of the radi&tion corresponding to each electron energy?
To which spectral series do the lines belong? What are the quantum numbers
of the initial and final states in the transition represented by each spectral line?
7. From the constancy of angular momentum at kh/21r and the constancy
of total energy at -2r 2m' e4Z 2/n 2h2, where m' is the reduced mass of the electron,
find the semiaxes of the Sommerfeld elliptic orbits in terms of k and n. Prove
that b/a = k/n. (See Born, Atomic Physics, Appendix XIV.)
8. The noble gases Ne, A, Kr, Xe, and Rn have electron configurations in
which the s and p orbits of the outer shell are filled. Prepare a table showing
the number of electrons in each shell and each orbit for these elements, and
give the value of n and l associated with each shell and each orbit. Add to
this table the electron configurations attributed to the alkalis, Na, K, Rb, Cs,
and Fr, and to the halogens, F, CI, Br, I, and At. How is the occurrence of
15 rare earth elements with similar chemical properties explained?
9. Calculate the wavelengths of the K. x-ray linee of Al, K, Fe, Co, Ni,
Cu, Mo, and Ag, and compare the values with those listed in Table 4-1. Use
R = 109737.31 cm- 1 •
10. Show that the wavelength of an electron can be written in the form
X = ~ VI - (v2/c2).
moc (v/c)
The quantity h/moc is called the Compton wavelength of the electron. What is its
magnitude? Compute the wavelength of electrons for each of the speeds listed
in Problem 5, Chapter 6. What is the Compton wavelength of a proton?
What is the wavelength of protons at each of the above speeds?
11. Show that the wavelength of electrons moving at a velocity very small
compared with that of light and with a kinetic energy of V electron volts can
be written
.f\,. = 12.268
v112
X 10-s
cm.
What is the wavelength of electrons with energies of 1, 10, 100, 1000 volts?
12. If V exceeds a few thousand electron volts, the effect of the relativistic
increase of mass with velocity must be considered. Show that the wavelength
is given by
.,. _ 12.268(m/mo)-l/ 2 -s
f\ - y112 X 10 cm,
where m/mo is the ratio of the moving mase of the electron to its rest mass.
What is the wavelength of electrons accelerated through !', potential of 104 volts?
105, 106, 10 7 volts?
13. What is the de Broglie wavelength of (a) a hydrogen molecule moving at

a speed of 2.4 X 10 5 cm/sec? (b) a rifle bullet weighing 2.0 gm, and moving
ata speed of 400 m/sec?
14. What relationship exists between the emission spectrum of singly ionized
helium and that of atomic hydrogen? What are the wavelengths of the first four
lines of the series which includes transitions to the state n = 2 of singly ionized
helium? How does the spectrum of doubly ionized lithium compare with that
of hydrogen? What is the radius of the first Bohr orbit of doubly ionized lithium?
15. It is known that a particle exists, called the positron, with the same mass
as that of the electron and with a charge equal in magnitude but opposite in
sign. A positron and an electron can form a system, called positronium, analogous
to hydrogen, with a very short, but measurable, half-life. (a) Show that the
energy levels of positronium are analogous to those of hydrogen but that their
magnitudes are smaller by the factor ½[1 + (m/Mu)). (b) Show that the radii
of the Bohr orbits of positronium are larger than thoSP. of hydrogen by a factor
which is the reciproca! of the factor of part (a). (c) What is the ground state
energy of positronium in electron volts? What is the radius of the first Bohr
orbit?
16. A particle called a negative r-meson has been shown to exist; its charge is
the same as that of the electron and its mass is 273 times that of the electron.
A proton and a negative r-meson can form short-lived stationary states analo-
gous to those of the hydrogen atom. (a) How do the mesonic orbits compare
with those of the corresponding electronic orbits? What is the radius of the
mesonic orbit for n = 1? (b) What is the wavelength of the electromagnetic
radiation emitted when a negative r-meson, initially at rest infinitely far from
the proton, drops into the smallest orbit permitted by quantum theory?
Part II
The Nucleus
CHAPTER 8
THE CONSTITUTION OF THE NUCLEUS
In the last chapter, it was shown that the application of quantum theory
to the nuclear model of the atom led to the development of a satisfactory
theory for those properties of atoms that depend on the extranuclear
electrons. The development of a theory of the nucleus is a more difficult
problem. The density of matter in the nucleus offers a clue to the source
of the difficulty. The work of Rutherford and bis colleagues on the
scattering of a-particles showed that the atomic nucleus has a radius of
10- 12 to 10- 13 cm, so that the volume of the nucleus is of the order of
10-36 cm 3 or less. Now, the mass of one of the lighter atoms is about
10-24 gm, and it is almost ali concentrated in the nucleus, with the result
that the density of the nucleus is at least 10 12 gm/cm 3 • A density of
this magnitude is inconceivably large, and it is clear that in the atomic
nucleus, matter is put together in a way which may not be amenable to
ordinary experimental and theoretical methods of analysis. Consequently,
tbe interpretation of the nuclear properties of atoms in terms of a theory
of nuclear structure presents great problems.
There is now available a large amount of experimental information
about nuclei, derived from work in several fields: (I) the precise measure-
ment of the masses of atoms; (2) radioactivity, natural and artificial;
(3) the artificial transmutation of nuclei by bombardment with particles
from radioactive substances or with higb-speed particles produced by
laboratory metbods; (4) optical spectroscopy in the visible and ultra-
violet regions; (5) the direct measurement of certain nuclear properties,
such as spin and magnetic moment. The main problems of nuclear
pbysics are the collection and correlation of the experimental facts and
their interpretation in terms of a theory or model of the nucleus. In this
chapter, the problem of the constitution of the nucleus will be discussed,
and sorne ideas will be developed which will be useful in the interpretation
of experimental data in the following chapters. In addition, sorne new
and unfamiliar properties of the nucleus will be considered.
8-1 The proton-electron hypothesis of the constitution of the nucleus.

The fact that certain radioactive atoms emit a- and {J-rays, both of which
are corpuscular in nature, led to the idea that atoms are built up of ele-
mentary constituents. As early as 1816, on the basis of the small number
of atomic weights then known, Prout suggested that ali atomic weights
are whole numbers, that they might be integral multiples of the atomic
183
184 THE CONSTITUTION OF THE NUCLEUS [CHAP. 8
weight of hydrogen, and that all elernents rnight be built up of hydrogen.

Prout's hypothesis was discarded when it was found that the atornic
weights of sorne elernents are fractional, as for exarnple, those of chlorine
(35.46) and copper {63.54). Nevertheless, so ma.ny elernents have atomic
weights which are very close to whole numbers that there seerned to be
sorne basis for Prout's hypothesis. The idea that all elernents are built
up frorn one basic substance received new support during the early years
of the 20th century when the study of the radioactive elernents led to
the discovery of isotopes. It was found that there are atornic species which
have different masses in spite of the fact that they belong to the same
elernent and have the sarne atomic number and chemical properties; the
different species belonging to the same elernent are called isotopes. For
exarnple, four radioactive isotopes of lead were discovered with atomic
weights of 214, 212, 211, and 210, as well as three nonradioactive isotopes
with atomic weights of 206, 207, and 208. The nuclei of these isotopes,
varying in ma.ss from 206 to 214, show a wide range of stability as measured
by the extent of their radioactivity, although all have the same charge.
The proof of the existence of isotopes in the radioactive elements led
to experiments to test whether sorne of the ordinary elernents also consist
of a mixture of isotopes. It was found that this is indeed the case. Most
elernents are mixtures of isotopes, and the atomic ma.sses of the isotopes
are very close to whole nurnbers. Chlorine, as found in nature, has two
isotopes with atomic weights of 34.98 and 36.98, respectively; 75.4% of
the chlorine atoras have the smaller ma.ss, while 24.6% have the greater
mass, and this distribution explains the atomic weight 35.46 of chlorine.
Analogous results were obtained for copper. The different isotopes of an
elernent have the same number and arrangement of extranuclear electrons,
and consequently their spectra have the same general structure; they are
distinguished from one another by their different atornic masses.
The fact that the atomic masses of the isotopes of an element are close
to whole numbers led Aston to formulate his whole number rule. Accord-
ing to this rule, which is really a rnodified form of Prout's hypothesis, all
atornic weights are very close to integers, and the fractional atornic weights
determined by chernical methods are caused by the presence of two or
more isotopes each of which has a nearly integral atomic weight. Much
of the experimental work on isotopes involved the analysis of the positive
rays from different substances; and in all the work of this kind the lightest
positively charged particle that was ever found had the sarne mass as the
hydrogen atom, and carried one positive charge equal in rnagnitude to
the electronic charge, but of opposite sign. This particle is evidently
the nucleus of a hydrogen atom and, as shown in Chapter 1, has a mass
very close to one atomic mass unit. The combination of the whole number
rule and the special properties of the hydrogen nucleus led to the ~urnp-
8-2] THE ANGULAR MOMENTUM OF THE NUCLEUS 185
tion that atomic nuclei are built up of hydrogen nuclei, and the hydrogen
nucleus was given the name proton to indicate its importance as a funda-
mental constituent of ali atoros.
The whole number rule is a.ctually an approximation, holding to an
accuracy of about 1 part in 1000. The most precise experiments show
that there are small but systematic departures from this rule over the
whole range of elements. It will be shown in the next chapter that these
variations are of great importance in adding to our knowledge of the
structure of nuclei.
To account for the mass of a nucleus whose atomic weight is very close
to the integer A, it was necessary to assume that the nucleus contained
A protons. But if this were the case, the charge on the nucleus would
be equal to A, nearly the same as the atomic weight and not equal to the
atomic number Z, which is half, or less, of the atomic weight. To get
around this difficulty, it was assumed that in addition to the protons,
atomic nuclei contained A - Z electrons; these would con tribute a
negligible amount to the mass of the nucleus, but would make the charge
equal to + Z, as required. It was thus possible to consider the atoro as
consisting of a nucleus of A protons and A - Z electrons surrounded by Z
extranuclear electrons. The number A is called the mass number and is
the integer closest to the atomic weight.
The proton-electron hypothesis of the nucleus seemed to be consistent
with the emission of a- and ,8-particles by the atoros of radioactive elements.
The interpretation of certain generalizations about radioactivity in terms
of the nuclear atom showed that both the a- and ,8-particles were ejected
from the nuclei of the atoms undergoing transformation; and the presence
of electrons in the nucleus ma.de it seem reasonable that under the ap-
propriate conditions one of them might be ejected. It was also reasonable
to assume that a-particles could be formed in the nucleus by the combina-
tion of four protons and two electrons. The a-particles could exist as
such, or they might be formed at the instant of emission.
S-2 The angular momentum of the nucleus ¡ failure of the proton-

electron hypothesis. Although the hypothesis that nuclei are built up of
\>rotons and electrons had sorne satisfactory aspects, it eventually led to
contradictions and had to be abandoned. One of the failures of the hy-
pothesis was associated with a hitherto unknown property of the nucleus,
the angular momentum. The discovery that the atomic nucleus has an
angular momentum, or spin, with which is associated a magnetic moment,
was the result of the detailed study of spectral lines. When individual
components of multiplet lines were examined with spectral apparatus of
the highest possible resolution, it was found that each of these components
is split into a number of lines lying extremely close together; this further
splitting is called hyperfine structure. The total splitting, in units of wave

number, is only about 2 cm- 1 or less. The hyperfine structure could not
be accounted for in terms of the extranuclear electrons, and it was neces-
sary to assume, as Pauli did in 1924, that it is related to properties of the
atomic nucleus. The properties associated with the hyperfine structure
are the mass and angular momentum of the nucleus.
It was shown in Chapter 7 that because of the simultaneous motion of
nucleus and electron around the common center of gravity, the Rydberg
constant of an element depends on the mass of the atomic nucleus. Hence,
if an element has more than one isotope, each isotope has a slightly dif-
ferent value of the Rydberg constant, and corresponding spectral lines of
different isotopes have slightly different wave numbers. This effect has
been found experimentally, and the theoretical predictions have been
confirmed. In many cases, however, the isotope effect is not enough to
explain the hyperfine structure becauE>e the number of components is
greater than the number of isotopes. Elements which have only one
isotope, such as bismuth, also show hyperfine structure. In these cases,
the hyperfine structures can be accounted for quantitatively if it is as-
sumed, as for the extranuclear electrons, that the nucleus has an angular
momentum.
A magnetic moment is associated with the angular momentum of the
nucleus justas one is associated with the angular momentum of an elec-
tron. The two magnetic moments interact, and the interaction energy
perturbs the total energy of the electrons; there is, therefore, a splitting
of the atomic levels, which gives rise to the hyperfine structure of the
lines of the atomic spectrum. The multiplicity and the relative spacings
of the lines can be derived theoretically and depend on the magnitudes of
the nuclear angular momentum and magnetic moment. The nuclear
angular momentum can then be deduced from the experimentally deter-
mined multiplicity and relative spacings. Newer methods, which involve
radiofrequency spectroscopy, microwave spectroscopy, or the deflection
of molecular beams in magnetic fields, have been developed for measuring
the nuclear magnetic moment, and a large body of experimental informa-
tion has been built up concerning these nuclear properties.
The nuclear angul~r momentum has quantum mechanical properties
analogous to those of the angular momentum of the electron. It is a
vector, I, of magnitude v /(/ + 1) h/21r, where / is the quantum number
which defines the greatest possible component of I along a specified axis,
according to the rule
h
lz = I -
27r·
(8-1)
The value of I has been found experimentally to depend on the mass

8-2) THE ANGULAR MOMENTUM OF THE NUCLEUS 187
number A of the nucleus; if A is even, I is an integer or zero; if A is odd,

J has an odd half-integral value (half of an odd integer). In other words,
if the mass number of the nucleus is even, I may have one of the values:
O, 1, 2, 3, ... ; if A is odd, I may have one of the values ½, i, i, ....
The rules just cited lead to one of the failures of the proton-electron
hypothesis for the constitution of the nucleus. Nitrogen has an atomic
number of 7 and a mass number of 14, and its nucleus would have 14
protons and 7 electrons under this hypothesis. The contribution of the
protons to the angular momentum should be an integral multiple of
h/21r, whether the angular momentum of a proton is an integral or odd
half-integral multiple of h/21r. An electron has spin ½(h/21r) so that its
total angular momentum is always an odd half-integral multiple of h/21r
(see Section 7-9). The contribution of 7 electrons is, therefore, an odd
half-integral multiple of h/21r, and the total angular momentum of the
nitrogen nucleus should be an odd half-integral multiple of h/21r. But
the angular momentum of the nitrogen nucleus has been found experi-
mentally to be I = 1, an integer, in contradiction to the value predicted
by the hypothesis. Experiments show that the proton, like the electron,
has an intrinsic spin of ½(h/21r), and it is possible with this additional
information to predict the angular momenta of many other nuclei on the
basis of the proton-electron hypothesis. Thus, the isotopes of cadmium
(Z = 48), mercury (Z = 80), and lead (Z = 82) with odd mass numbers
should each have an odd number of electrons and an even number of
particles ali together. The angular momenta of these nuclei should,
therefore, be zero or integral; they have been found experimentally to be
odd half-integers, and the hypothesis again predicts values which disagree
with the experimental results.
The proton-electron hypothesis also fails to account for the order of
magnitude of nuclear magnetic moments. Measurements of the magnetic
moments of many nuclei have given values which are only about 1/1000
of the value of the magnetic moment of the electron. The magnitude of
the latter is
eh
IJ.B = 41rmc ' (8-2)
where m is the mass of the electron; this quantity is called the Bohr
magneton and has the value 0.92 X 10- 20 erg/gauss. All measured nuclear
magnetic moments are of the order 10- 23 erg/gauss, and their values can
be expressed appropriately in terms of the quantity
/J.N = -4 1reh
M =
HC
0.505 X 10-23 erg/ gauss, (8-3)
in which the electron mass has been replaced by the proton mass; the
quantity µ.N is called the nuclear magneton. Measured values of the nuclear
magnetic moment vary from zero to about 5 nuclear magnetons, and the
proton has a magnetic moment of 2. 7926 ± 0.0001 nuclear magnetons.
If electrons were present in the nucleus, we should expect to find nuclear
magnetic moments of the order of magnitude of the Bohr magneton, at
least in those nuclei for which A - Z is odd (a.nd which would, on the
proton-electron hypothesis, contain an odd number of electrons). The
fact that nuclear magnetic moments are only of the order of magnitude of
the nuclear magneton is, therefore, another strong argument against the
existence of electrons inside the nucleus.
There is also a wave-mechanical argument against the existence of
free electrons in the nucleus. According to the uncertainty principie
(cf. Section 7-8),
llxll.p ~ h, (8-4)
where /lx is the uncertainty in the position of a particle and lip is the
uncertainty in its momentum. Suppose that the principie is applied to
an electron in the nucleus. The uncertainty /lx in the position of an
electron is roughly the same as the diameter of the nucleus, which is
assumed here to be 2 X 10- 12 cm. Then
h 6.6 X 10-27
lip ~ /lx = 2 X 10_ 12 = 3.3 X 10 -16 erg-sec/ cm.
From the uncertainty in the momentum, it is possible to get a rough
estímate of the energy of an electron in the nucleus. Relativistic formulas
must be used because the electron moves very rapidly in the nucleus, as
will be seen. From Eqs. {6-36) and (6-38), the total energy of a particle
can be expressed in terms of the momentum,
(8-5)
where m 0 is the rest mass of the electron, and e is the velocity of light.
If it is now assumed that the momentum p of the electron is no larger
than the values just found for the uncertainty lip, then p =
3.3 X 10- 16,
and p 2c2 = (3.3 X 10- 16) 2 (3 X 10 1 º) 2 = 10-8 erg. This value is much
greater than the term m~c• = (9 X 10-28 ) 2 (3 X 10 1 º) 3 ~ 10- 12, which
can consequently be neglected. Then E 2 ~ 10-8 and
E = 10-• erg = 1.6 X10-•10_ = 6 X 10 ev = 60 Mev.

12
7
According to this result, a free electron confined within a space as small

as the nucleus would have to have a kinetic energy of the order of 60 Mev,
anda velocity greater than 0.999c. Experimentally, however, the electrons
8-3) NUCLEAR TRANSMUTATION 189
emitted by radioactive nuclei have never been found to have kinetic

energies greater than about 4 Mev, or at least an order of magnitude
smaller than that calculated from the uncertainty principle. Although
the calculation is a rough one, similar resulta are obtained from more
rigorous calculations, and in view of the large discrepancy it seems im-
probable that nuclei can contain free electrons.
The above argument does not apply to a proton in the nucleus because
the proton mass (I.67 X 10-24 gm) is nearly 2000 times as great as the
electron rest mass. For the proton in the nucleus, /u is also about 2 X
10- 12 cm, and il.p ~ 3.3 X 10-n erg-sec/cm. If it is assumed, as in the
case of the electron, that the momentum is of the order of the uncertainty
in the momentum, then from Eq. (8-5),
E 2 ~ 10-s + {I.67 X 10-24) 2 (3 X 1016) 4
= 10-s + 2.3 X 10- 6
,
and now the first term may be neglected in comparison with the second.
Then
~ 1.5 X 10-3 erg = :.: : {~1 2 = 9.4 X 108 ev =
3
E 940 Mev.
This value is only slightly greater than the rest energy of the proton, which
is 1.008 X 931 = 938 Mev. Hence, the kinetic energy of a proton in
the nucleus is of the order of a few Mev, and it should be possible for a
free proton to be contained in the nucleus.
8-3 Nuclear transmutation and the discovery of the neutron. The

failure of the proton-electron hypothesis of the nucleus was related to
the properties of the free electron. It was proposed, therefore, that the
electrons are bound to the positively charged particles and have no inde-
pendent existence in the nucleus. One possibility, which had been sug-
gested by Rutherford as early as 1920, was that an electron and a proton
might be so closely combined as to forro a neutral particle, and this hypo-
thetic&l particle was given the name netdron. N ow, ali of the methods
that have been discussed so far for detecting particles of nuclear size
depend on effects of the particle's electric charge, as deftection in magnetic
or electric fielde and ionization. The preeence of a neutron, which has
no charge, would be very hard to detect, and many uneucceBBful attempts
were made to find neutrons. Finally, in 1932, as one of the resulta of
reeearch on the disintegration or transmutation of nuclei by a-particles,
Chadwick demonstrated the exietence of neutrons. This discovery opened
up a vast field for further experimental work and led to the preeently
accepted idea of the constitution of the nucleus: that it is built of protone
and neutrons.
190 THE CONSTITUTION OF THE NUCLEUS [CHAP, 8
It has been mentioned that the fact that the masses of the atoms of
all elements, or their isotopes, are very close to whole numbers led to
the hypothesis that the atomic nucleus is a composite structure made up
of an integral number of elementary particles with masses close to unity.
If so, it should be possible by suitable means to break down the nucleus
and to change one element into another. Radioactivity is a naturally
occurring process of this kind. Rutherford and bis colleagues used energetic
a-particles from radioactive substances as projectiles with which to
bombard a.tomic nuclei in the a.ttempt to produce artificial nuclear dis-
integrations. They found that when nitrogen was bombarded with a-
particles, energetic protons were obtained. The protons were identified
by ma.gnetic deflection measurements, a.nd their energies were consider-
ably greater tha.n those of the bombarding a-pa.rticles. They must have
been shot out from the struck nucleus in such a way that pa.rt, at least, of
their energy was derived from the internal energy of the nucleus. Cloud-
chamber studies of the process showed that the a-particle was captured
by the nitrogen nucleus, and then the proton was ejected. This explana-
tion corresponds to the forma.tion from nitrogen (a.tomic number 7, a.tomic
weight 14) of an atom of a.tomic number 8 and atomic weight 17, i.e., the
oxygen isotope of ma.ss 17. The bombardment with a-particles was found
by Rutherford to cause emission of protons from all the elements of atomic
number up to 19 with the exception of hydrogen, helium, lithium, carbon,
and beryllium. The most ma.rked transmutation effects occurred with
boron, nitrogen, a.nd a.luminum.
Closer investiga.tion of the bomba.rdment of boron and beryllium by
a-pa.rticles gave sorne a.dditional a.nd unexpected results. Bothe a.nd
Becker (1930) discovered that these elements emitted a highly penetrating
radia.tion when so bombarded. It was thought that this radia.tion might
be a form of 'Y-ray of very high energy. Curie a.nd Joliot (1932) found
that when the radia.tion was allowed to fall on substances containing
hydrogen, it caused the production of highly energetic protons. Chadwick
{1931) was a.lso able to show that the rays emitted from bombarded
beryllium gave rise to rapidly moving atoms when a.llowed to fall on other
substances, for exa.mple, He, Li, Be, C, O, and N. These results could
not be explained under the a.ssumption that the new radia.tion consisted of
high-energy 'Y-rays. Chadwick finally (1932) proved that the energies
of the protons ejected from hydrogenous materia.Is, and of the other
ra.pidly moving a.toms, could only be explained on the view tha.t the
"rays" from bombarded beryllium actua.lly consisted of particles with a
ma.ss close to that of the proton. These pa.rticles, unlike protons, produce
no tracks in the cloud chamber a.nd no ioniza.tion in the ioniza.tion chamber.
These facts, together with the extremely high penetrating power of the
particles, show that the charge of the latter must be zero. Since the new
8-4] THE PROTON-NEUTRON HYPOTHESIS 191
particle was found to be neutral and to ha.ve a mass close to unity, it was
identified with Rutherford's neutron. La.ter measurements have shown
that the mass of the neutron is 1.00898 amu, so that it is slightly heavier
than the proton, with a mass of 1.00758 amu.
8-4 The proton-neutron hypothesis. The discovery of a particle,

the neutron, with an atomic weight very close to unity and without electric
charge, led to the assumption that every atomic nucleus consists of protons
and neutrons. This hypothesis was used for the first time as the basis of
a detailed theory of the nucleus by Heisenberg in 1932. Under the proton-
neutron hypothesis, the total number of elementary particles in the
nucleus, protons and neutrons together, is equal to the mass number A of
the nucleus; the atomic weight is therefore very close to a whole number.
The number of protons is given by the nuclear charge Z, and the number
of neutrons is A - z.
The new nuclear model avoids the failures of the proton-electron hy-
pothesis. The empirical rule connecting mass number and nuclear angular
momentum can be interpreted as showing that the neutron, as well as the
proton, has a half-integral spin; the evidence is now convincing that the
spin of the neutron is indeed ½h/2,r. If both proton and neutron ha.ve
spin ½then, according to quantum theory, the resultant of the spins of A
elementary particles, neutrons and protons, will be an integral or half-
integral multiple of h/2,r according to whether A is even or odd. This con-
clusion is in accord with all the existing observations of nuclear angular
momenta. The value of the magnetic moment of the neutron is close to
-2 nuclear magnetons; it is opposite in sign to that of the proton, but
not very different in magnitude. The values for both the proton and
neutron are consistent with those measured for many different nuclei.
Finally, since the mass of the neutron is very close to that of the proton,
the argument showing that protons can be contained within the nucleus
is also va.lid for neutrons.
The neutron-proton hypothesis is consistent with the phenomena of
radioactivity. Since there are several reasons why electrons cannot be
present in the nucleus, it must be concluded that in /3-radioactivity, the
electron is created in the act of emission. This event is regarded as the
result of the change of a neutron within the nucleus into a proton, an
electron, and a new particle called a neutrino, a.nd both experimental
and theoretical evidence offer strong support for this view. In /3-ra.dio-
activity, then, the nucleus is tra.nsformed into a different one with one
proton more a.nd one neutron less, a.nd an electron is emitted. An a-
particle can be formed by the combination of two protons and two neutrons.
It may exist as such in the nucleus, or it may be formed at the instant of
emission; the latter possibility is now regarded as more likely.
It must be emphasized, however, that in considering the nuclei of

different elements as being built of protons and neutrons, the neutron is
not regarded as a composite system formed by a proton and an electron.
The neutron is a fundamental particle in the same sense that the proton is.
The two are sometimes called nucleons in order to indicate their function
as the building blocks of nuclei.
One of the main problems of nuclear physics is that of understanding
the nature of the forces holding the protons and neutrons together. Much
of the research in nuclear physics is aimed at the clarification of the laws
of interaction between nuclear particles. This problem will not be treated
in a detailed, quantitative way in this book because of its complexity, but
it will be discussed in a la.ter chapter. The emphasis in the following
chapters will be rather on the facts about nuclei, and on the transforma-
tion of one nuclear species into another by the rearrangement of its con-
stituent nucleons. The information accumulated will be considered and
interpreted in terms of the proton-neutron hypothesis. This aspect of
nuclear physics is analogous to the application of atomic physics to the
chemical properties of the elements, and is sometimes referred to among
physicists as nuclear chemistry.
For the present, it will suffice to point out sorne of the qualitative
properties of the forces between the protons and neutrons in the nucleus.
Beca.use of the positive charge ·of the proton, there must be repulsive,
electrostatic forces between the protons tending to push the nucleus apart.
It is apparent from the small size and great density of the nucleus that
these forces must be very large in comparison with the forces between
the nucleus and the extra.nuclear electrons. Hence, if stable complex
nuclei are to exist, there must be attractive forces in the nucleus strong
enougb to overcome the repulsive forces. These attractive forces are the
specifically nuclear forces between a proton and a neutron, between two
neutrons, or between two protons. They seem to be more complex than
the gravitational or electromagnetic forces of classical physics. The
nuclear attractive forces must be very strong at distances of the order of
the nuclear radius, i.e., they are slwrt-range forces. Outside the nucleus,
they decrease very rapidly, and the Coulomb repulsive forces responsible
for the scattering of- a-particles predominate. The magnitude of the
nuclear forces is such that the work required to divide a nucleus into its
constituent particles (the binding energy of the nucleus) is very much
greater than the work needed to separate an extranuclear electron from
an atom. Whereas the latter is of the order of electron volts, energy
changes in the nucleus are of the order of millions of electron volts. It is
the magnitude of the energies associated with nuclear transformations
that is responsible for the large-scale applications of nuclear physics, and
for the rise of the new field of nuclear engineering.
8-5] MAGNETIC AND ELECTRIC PROPERTIES OF THE NUCLEUS 193
8-5 Magnetic and electric properties of the nucleus. I t will be con-

venient to include in this chapter a few more remarks about the angular
momentum of the nucleus, as well as a brief discussion of sorne magnetic
and electric properties of the nucleus. These properties are important
in the interpretation of many nuclear phenomena and in the theory of
the nucleus, and they occupy a prominent place in nuclear physics. They
are, however, among the less elementary aspects of the subject and will
not be treated in detail in this book. Nevertheless, sorne familiarity with
the ideas and terminology of these matters will be helpful to the reader.
Each proton and each neutron in the nucleus has an angular momentum
which may be pictured as being caused by the particle's spinning motion
about an axis through its center of mass. The magnitude of this spin
angular momentum is ½h/21r. The wave-mechanical properties of an
angular momentum of this kind are such that its orientation in space
can be described by only two states: the spin axis is either "parallel" or
"antiparallel" to any given direction. The component of the spin along
a given direction, say the z-axis, is either ½h/21r or -½h/21r. In addition,
each nucleon may be pictured as having an angular momentum associated
with orbital motion within the nucleus. According to quantum theory,
the orbital angular momentum is a vector whose greatest possible com-
ponent in any given direction is an integral multiple of h/21r. Each
nucleon has a total angular momentum i about a given direction, with
i = l± B, (8-6)
where l is the orbital angular momentum and s is the spin angular mo-
mentum. The spin of any single nucleon can add or subtract ½h/21r
depending on its orientation with respect to the axis of reference, and i is
therefore half-integral. For nuclei containing more than one particle, it
is customary to write corresponding relationships between the momenta
in capitals; the resultant total angular momentum of the nucleus is then
I = L ± S, (8-7)
where L is the total orbital angular momentum, and S is the total spin
angular momentum. The total angular momentum is actually a vector,
denoted by I, and the scalar quantity I is defined as the maximum possible
component of I in any given direction. The orbital angular momentum
L is an integral multiple of h/21r; S is an even half-integral multiple of
h/21r if the number of nuclear particles is even, and an odd half-integral
multiple if the number of particles is odd. Hence, I is an integral multiple
of h/21r when A is even, and an odd half-integral multiple when A is
odd, in agreement with experimental resulta.
There are two possible sources of confusion which arise from careless
usage. The term "spin" is often used for the total angular momentum
of a nucleus rather than for the spin S alone. This incorrect usage was
introduced before the problem of the internal structure of nuclei had
attained its present importance and has been continued. In addition, the
total orbital angular momentum of a nucleus is often denoted by l rather
than by L. The meaning of the term "spin" and the symbol l can usually
be inferred without difficulty from the context in which they appear.
The magnetic moment of a nucleus can be represented as
h
p,¡ = 'Y¡ 211" I = UIP.NI, (8-8)
where 'Yr and gr are defined by Eqs. (8-3) and (8-8) and are called the
nuclear gyromagnetic ratio and nuclear g-factor, respectively; P.N is the
nuclear magneton defined by Eq. (8-3), and has the numerical value
5.04929 X 10-u erg/gauss. The quantity which measures the magnitude
of p,r, and is called the nuclear magnetic moment µ. 1, is
h
P.r = 'Yr 2 1r l. (8-9)
The details of the methods for measuring nuclear spins and magnetic
moments will not be discussed in this book. The interested reader is
referred to the works by Ramsey and Kopfermann listed at the end of
the chapter, for a treatment of the methods and results. The spins and
magnetic moments of many nuclei have been measured, particularly for
the ground state of the nucleus, and certain pattems have been observed
in the experimental results. Significant conclusions as to the structure
of the nucleus may be obtained from data on nuclear spin and magnetic
moment, just as knowledge of the arrangement of the extranuclear elec-
trons was obtained from information about their angular momenta. It
has been found, for example, that / = Ofor nuclei containing even numbers
of protons and neutrons. It follows from Eq. (8-9) that a so-ealled
even-even nucleus should have no magnetic moment, and this has been
found experimentally to be the case. This generalization, and others,
have proved very useful in the study of nuclear structure and other
properties.
Another property which is highly important in connection with the
aha-pe of the nucleus is the electric quadrupole moment. This quantity,
which cannot be discussed in a simple way, is a measure of the deviation
of a nucleus from spherical symmetry. H a nucleus is imagined to be an
ellipsoid of revolution whose diameter is 2b along the symmetry axis and
2a at right angles to this axis, and ü the electric charge density is assumed
8-6] ADDITIONAL PROPERTIES OF ATOMIC NUCLEI 195
uniform throughout the ellipsoid, the quadrupole moment Q is given by

(8-10)
Its magnitude depends on the size of the nucleus, the extent of the deviation
from spherical symmetry, and the magnitude of the charge; the sign may
be positive or negative. Many nuclei have been found to have quadrupole
moments; thus, deuterium, a nucleus with one proton and one neutron, has
a Q-value of +o.00274 X 10-u cm 2 , while an isotope of lutecium con-
taining 176 nucleons has a Q-value of 7 X 10-24 cm 2 • The investigation
of quadrupole moments along with spins and magnetic moments has led to
important developments in the theory of nuclear structure, as will be
shown in Chapter 17.
8-6 Additional properties of atomic nuclei. In addition to their

electric and magnetic properties, nuclei have certain properties which are
not obviously physical in nature. Although these properties will not be
treated in detail, the reader should at least know of the existence and
usefulness of the concepta involved. The properties which will be dis-
cussed very briefly are the statistics to which nuclei are subject, and the
parity.
The concept of statistics in physics is related to the behavior of large
numbers of particles. Thus, the distribution of energies or velocities
among the molecules of a gas can be described by the classical Maxwell-
Boltzmann statistics, as can many other macroscopic properties of gases.
The properties of assemblies of photons, electrons, protons, neutrons, and
atomic nuclei cannot, in general, be described on the basis of classical
statistics, and two new forros of statistics have been devised, based on
quantum mechanics rather than on classical mechanics. These are the
Bose-Einstein statistics, and the Fermi-Dirac statistics. The question of
which form of quantum statistics applies to a system of particles of a
given kind is related to a particular property of the wave function which
describes the system. This property has to do with the effect on the
wave function of interchanging all of the coordinates of two identical
particles, say of two protons in a nucleus. A nucleon is described by
a function of its three space coordinates and the value of its spin, whether
it is ½h/21r or -½h/21r. The Fermi-Dirac statistics apply to systems of
particles for which the wave function of the system is antisymmetrical,
i.e., it changes sign when all of the coordinates (three spatial and one
spin) of two identical particles are interchanged. It follows from this
property of the wave function that each completely specified quantum
state can be occupied by only one particle; that is, the Pauli exclusion
principie applies to particles obeying the Fermi-Dirac statistics. It has
been deduced from experiments that electrons, protons, and neutrons obey
196 THE CONSTITUTION OF' THE NUCLEUS [CHAP. 8
the Fermi-Dirac statistics, as do ali nuclei of odd mass number A. In the

Bose-Einstein statistics, the wave function is symmetrical, i.e., it does
not change sign when ali the coordinates of any pair of identical particles
are interchanged. Two or more particles may be in the same quantum
state. Ali nuclei having even mass number A obey the Bose-Einstein
statistics. There is a direct correlation between the total angular mo-
mentum of a nucleus and its statistics: Fermi-Dirac particles (odd A)
have total nuclear angular momenta which are odd half-integral multiples
of h/27r, while Bose-Einstein particles (even A) have momenta which are
integral multiples of h/27í.
The last property that will be mentioned is the parity. To a good ap-
proximation, the wave function of a nucleus may be expressed as the
product of a function of the space coordinates and a function depending
only on the spin orientation. The motion of a nucleus is said to have
even parity if the spatial part of its wave function is unchanged when
the space coordinates (x,y,z) are replaced by (-x, -y, -z). This trans-
formation of coordinates is equivalent to a reflection of the nucleus'
position about the origin of the x, y, z system of axes. When reflection
changes the sign of the spatial part of the wave function, the motion of
the nucleus is said to have odd parity. It has been shown that the parity
of a nucleus in a given state is related to the value of the orbital angular
momentum L; if Lis even, the parity is even; ü Lis odd, the parity is odd.
A system of particles will have even parity when the sum of the numerical
values for L for ali its particles is even, and odd parity when the sum is odd.
Although the parity seems to be an abstract sort of property, the selection
rules for many nuclear transitions involve conditions on the parity, as
well as on the total angular momentum.
REFERENCES
GENERAL
R. D. EvANS, The Atomic Nucleu,. New York: McGraw-Hill, 1955, Chap-
ters 2, 4, 5.
E. U. CoNDON and H. ÜDISHAW, eds., Handbook of Phym,. New York:
McGraw-Hill, 1958, Part 7, Chapters 1-6; Part 9, Chapter 3; by various authors.
N. F. RAMSEY, "Nuclear Moments and Statistics," Experimental Nuclear
Phy8ÍC8, E. Segre, ed., New York: Wiley, 1953, Vol. I, Part III.
N. F. RAMsEY, Molecular Bea1118. Oxford: Clarendon Press, 1956.
K. F. 8MITH, "Nuclear Moments and Spins," Progru, in Nuclear Phy8icll,
O. R. Frisch, ed. London: Pergamon Press, Vol. 6, p. 52, 1957.
H. KoPFERMANN, Nuclear Momenlll (English version prepared from the second
German edition by E. E. Schneider). New York: Academic Press, 1958.
C. H. TowNES and A. L. ScaAwww, Microwave Spectro,copy. New York:
McGraw-Hill, 1955.
CHAPTER 9
ISOTOPES
9-1 Natural radioactivity and isotopes. The disco~ry of isotopes was

one of the results of work on the radioactive elements. It was found
that generally the product of radioactive decay is itself radioactive, and
that each of the decay products behaves chemically in a different manner
from its immediate parent and its daughter product. Radium, a metal
with an atomic weight of 226, was shown to be the product of the a-dis-
integration of an element that was given the name ionium. Radium
in turn emits an a-particle, forming the rare gas radon, which also is radio-
active. The chemical properties of ionium turned out to be similar to
those of thorium and, in fact, the two elements when mixed could not be
separated chemically even with extremely sensitive methods. Moreover,
ionium and thorium were found to be spectroscopically identical. The
systematic study of the radioactive elements showed, however, that the
atomic weight of ionium must be 230, while that of thorium was known
to be 232. Other pairs of elements were also discovered which were
chemically identical but differed in atomic weight, and Soddy, in 1913,
suggested as a name for them the word isotopes, meaning the same place
in the periodic table.
The way in which isotopes arise in the radioactive elements can be
understood in terms of the effect of radioactive decay on- the atomic number
and atomic weight. Each time an a-particle is emitted, the charge of the
nucleus of the radioactive atom decreases by two units, since the a-particle
carries a positive charge of two units. At the same time, the atomic
mass decreases by four units. Each time a ,B-particle is emitted, the
nuclear charge increases by one unit because one negative charge is
removed, but the mass is practically unchanged. Thus, the emission of
an a-particle causes a decrease of two in the atomic number, i.e., a shift
of two places to the left in the periodic table. The emission of a ~-particle
causes an increase of one in the atomic number, i.e., a shift of one place
to the right in the periodic table. This law, first deduced by Soddy and
Fajans in 1913, is known as the displacement law o/ radioactivity. Con-
sider what happens as a result of the radioactive decay of the element
called uranium I (UI) with an atomic number of 92 and an atomic weight
(or mass number) of 238 units. This element emits an a-particle, and
according to the displacement law, the product, which was called uranium
X 1 (UX 1 ), has an atomic number of 90 anda mass number of 234. Uranium
X 1 emits a ~-particle giving the product uranium X 2 (UX 2 ) with an atomic
197
198 ISOTOPES [CHAP, 9
number of 91 anda mass number of 234; UX 2 , in turn, emits a ,9-particle

giving the element uranium 11 (UII), with an atomic number of 92 and
a mass number of 234. Hence, UI and UII have the same atomic number
92. It follows that these two elements must be entirely identical in ali
their chemical properties, and in fact it is impossible to separate these two
elements from each other by chemical means; they are isotopes. From
a chemical point of view, the mixture constitutes one element. Now,
UII emits an a-particle giving the product ionium, with an atomic number
of 90 anda mass number of 230. Naturally occurring thorium also has
the atomic number 90, and has the mass number 232. Hence, thorium,
ionium, and UX 1 are also isotopes. There are many other groups of
isotopes among the radioactive elements, and they will be discussed in
greater detall in the next chapter. The elements UX 1, UX 2 , and UII,
which are isotopes of thorium (Z = 90), protoactinium (Z = 91), and
uranium (Z = 92), respectively, ali have masa numbers of 234. They
are chemically different elements with the same atomic weight or mass
number, and are called i&obars.
9-2 Positive-ray analysis and the e:nstence of isotopes. After the

existence of isotopes had been proven for radioactive elements, Thomson
(1913), by deflection experiments on positive rays, proved that isotopes
also occur among ordinary elements. Thomson's work was extended and
improved by Aston, who developed a new type of positive-ray apparatus
which he called a mass spectrograph. Aston showed that most of the
elements are mixtures, having in sorne cases as many as nine or ten isotopes.
He was also able to obtain fairly accurate measurements of the relative
abundances of the isotopes of the different elements. The most important
uses to which the mass spectrograph has been put by Aston and others is
the identification of i.Mtopes and the precise determination of their masses.
The isotopic mass is one cf the directly measurable quantities essential to
an understanding of the atomic nucleus, and the importance of highly
precise measurements of this quantity will become evident.
One form of apparatus used by J. J. Thomson m for the analysis of
positive rays is shown schematically in Fig. 9-1. The discharge by which
the positive rays are produced takes place in a large tube A. The anode
is an aluminum rod D; the cathode K has a very small hole through which
the positive ions pass. The cathode is kept cool during the discharge by
means of the water jacket J. The gas to be studied is allowed to leak in
through a fine glass capillary tube L, and after circulating through the
apparatus is pumped off at F. The pressure in the tube is adjusted so
that the discharge potential is 30,000 to 50,000 volts.
The positive ions emerge from the end of the cathode as a narrow beam
of ions whose velocities depend on their charge and mass and on the dis-
9-2) POSITIVE-RAY ANALYSIS AND THE EXISTENCE OF ISOTOPES 199
{a)
Ion
(b)
F10. 9-1. Thomson's method for the analysis of positive raya. (a) Diagram
of the apparatUB. (b) Formation of Thomson's parabolas.
cha.rge potential. The beam is analyzed by causing it to pass between

the pieces of soft iron P 1 and P 2 , which are placed between the polea M
and N of an electromagnet; P 1 and P 2 are the pole pieces of the magnet
but are electrically insulated from M and N by insulating strips I. Thus,
P 1 and P 2 can be maintained at any desired potential difference, so tha.t
in the space between them there can be pe.ralle} magnetic and electric
fields. A positive ion moving from left to right in this space undergoes a
defl.ection in the plane of the paper because of the electric field, and per-
pendicular to tbe plane of the paper because of tbe magnetic field. Mter
leaving the space between P 1 and P 2 , the deflected particle moves in the
200 ISOTOPES [CBAP. 9
field-free, evacuated "camera" C and falls on the fluorescent screen or ·

photographic plate S. A soft iron shield W is placed between the magnet
and the discharge tube to prevent the stray magnetic field from inter-
fering with the discharge in the tube A.
Thomson found that the positive ions, after passing through the fields,
formed a pattem of parabolas on a photographic plate. The equation
of one of these parabolas can easily be derived, and it involves the charge-
to-mass ratio of the ions in the beam. Suppose that an ion of mass M,
charge q, and velocity v enters the space P 1P 2 along a path coincident with
the x-axis of a system of rectangular coordinates with origin in the space
P 1P 2 • Let E and H be the electric and magnetic field strengths, respec-
tively, let L be the length of the path through the fields, and assume
that the latter are constant and terminate sharply. The electrostatic
force on the ion is qE, so that the ion is subjected to an acceleration qE/ M
in the direction of the field. The time taken to pass through the field is
L/v. From the usual formula for the distance traveled by a particle under
a constant acceleration, the deflection y' from the x-axis is
y' =! qE
2 M
(f)2 .
v
The magnetic field exerts a force Hqv on the particle, which consequently
suffers a deflection z' in the z-direction, given by
z' =!
2 M
Hqv (f)2.
v
After passing through the fields, the particle moves in a straight line. If
the distance of the photographic plate from the fields is large compared
with L, the point where the particle strikes the plate will have coordinates
y and z which are proportional to y' and z', respectively. The relation
between z and y will therefore be the same as that between z' and y'; The
latter relation is obtained by eliminating v from Eqs. (9-1) and (9-2),
2 q H2
z = eM EY,
where C is a constant which depends on the dimensions of the apparatus.

Equation (9-3) represents a parabola. For constant values of the fields
E and H, ions with different values of the velocity v but with the same
value of q/M should make a parabolic trace on the photographic plate,
as Thomson found. lons with different values of q/M produce different
parabolas. If there are no fields, the beam of ions will strike the plate
at a point in line with the hole through the cathode called the undeftected
spot.
9-2) POSITIVE-RAY ANALYSIS AND THE EXISTENCE OF ISOTOPES 201
coNollNe"
coNoll Ne"
---1----4-----Y 1
1
~KI:
1
' e
1
1
1
t D
(a) (b)
F10. 9-2. Positive-ray parabolas obtained by the Thomson method. (a)

Schematic diagram of the parabolas. (b) Actual parabolas obtained by Thomson
(1913). Ne• is Ne 20 ; Ne" is Ne 22 • The lower picture had a longer exposure
than the upper, and brought out the weak Ne 22 trace that is barely visible in the
upper picture.
If the ions in a positive-ray beam have the same value of q/M but travel
with different velocities, a parabola like that denoted by A in Fig. 9-2(a)
is observed. If ions are present with the same value of q, but with a mass
M' greater than M, a different parabola B will be seen. This parabola
will be lower than A because the magnetic displacement is smaller for the
heavier ions. If the mass M of the ions giving the parabola A is known
accurately, then M' can be determined in the following way. The mag-
netic field is reversed after half the time interval of the exposure of the
plate, and a pair of curves C and D is obtained which are the mirror
images of A and B; the mid-line Oy is also obtained in this way. From
Eq. (9-3), it follows that M' /M = {ad/bc) 2, and the latter ratio can be
measured on the plate, giving the mass M' in terma of the standard atomic
mass M. By a systematic analysis of the relative positions of the various
parabolic traces appearing on a series of platea, Thomson was able to
determine the origin of the traces. Parabolas corresponding to H+, H-f¡,
o+, ot co+, and other ions with known maeses were identified. These
ions could then be used as standards from which the maeses of ions pro-
ducing other traces could be determined. If the positive rays consist of
a mixture of different ions, the relative intensities of the individual parab-
202 ISOTOPES
olas correspond to the relative abundances of the various ions in the

mixture.
In 1912, Thomson< 2 > analyzed the positive-ray parabolas fonned when
neon was used as the source of the ions [Fig. 9-2(b)]. With the apparatus
that he used, the parabolas corresponding to masses differing by 10%
could be clearly distinguisbed. The atomic weight of neon had been
detennined as 20.20, and neon was the lightest element whose atomic
weight differed significantly from a whole number. An intense parabola
was found, as expected, at a position corresponding to the atomic weight 20.
A much lesa intense trace was also found corresponding to the atomic
weight 22. The same two curves, with the same relative intensities, were
obtained with neon samples of different purity and for different conditions
of discharge and gas pressure. No element of atomic weight 22 was known;
the trace could not be identified with any known molecular ion. The
experimental facts suggested that neon could exist in two forma which
could not be distinguisbed chemically, but with different atomic weights.
The chemical atomic weight 20.20 would result if there were present nine
times as many neon atoms with atomic weight 20 as with atomic weight 22,
since
(9 X 20) +
10
(1 X 22) = 20 2
..
The suggestion that neon has two isotopes was so striking that Aston
looked for further evidence that might bear on the problem. It was well
known that a light gas diffuses through a porous partition more rapidly
than a heavier gas, and Aston used this property of gases in an attempt
to achieve a partial separation of the two constituents of neon. He passed
a sample of neon through a pipe-clay tube, collected the portion of the
gas which diffused through and allowed this to diffuse once more, and so
on. After repeated diffusion and rediffusion, he obtained from 100 cm 3 of
ordinary neon gas two extreme fractions of 2 to 3 cm 3 with atomic weights,
calculated from their densities, of 20.15 and 20.28, respectively. The
fraction with the smaller atomic weight was supposed to contain more
neon of atomic weight 20 than ordinary neon; the fraction with the greater
atomic weight was supposed to contain more neon of atomic weight 22
than ordinary neon. The changes in the proportions of the two constit-
uents were large enough to produce appreciable changes in the relative
intensity of the two traces (for the masses 20 and 22) in the positive-ray
photograph. The changes in intensity corresponded approximately to
the degree of enrichment or depletion of the masa 20 constituent. Although
the proof was not complete, it was difficult to avoid the conclusion that
tbere are two isotopes of neon, one with atomic weight 20, the otber with
atomic weight 22.
9--3) ISOTOPIC MASSES AND ABUNDANCES 203
9-3 lsotopic masses and abundances : the mass spectrograph ,and msss
spectrometer. The Thomson parabola method of analyzing positive rays
was adequate for a general survey of masses and velocities, but it could
not yield precise values of isotopic masses and abundances. The further
quantitative study of the constitution of the elements required the deter-
mination of isotopic masses with a precision of at least one part in a thou-
sand. To meet this need, Aston< 3 > {1919) designed the mass spectro-
graph. Aston's method of analysis was an improvement on that of
Thomson in that greater dispersion was achieved (i.e., greater separation
of ions of different masses) and all ions with a given value of q/M were
brought to a focus instead of being spread out in a parabola. By these
mea.ns, greater sensitivity and precision were attained.
A schematic diagram of Aston's first mass spectrograph is shown in
Fig. 9-3. The positive rays from a discharge tube pass through two very
narrow parallel slits 8 1 and 8 2 and enter the space between the metal
plates P 1 and P 2 • An electric field between these platea causes a deflec-
tion of the ions toward P 2 , the amount of the deflection being greater the
smaller the velocity of the ions. The narrow beam contains particles
with a wide range of velocities, with the result that it is broadened as it
passes through the field. A group of these particles is selected by mea.ns
of the relatively wide diaphragm D. Mter passing through D, this diverg-
ing stream of ions enters a ma.gnetic field, indicated by the circle at O,
perpendicular to the plane of the paper. The ma.gnetic field causes deflec-
tions, as shown in the diagram, the more slowly.moving ions being deflected
more than the faster ones. The paths of the slow-moving ions, therefore,
intersect those of the faster-moving ions at sorne point F. If the instru-
ment is properly designed, ions having the same value of q/M but slightly
different energies can be brought to a single focus on a photographic plate.
Other ions with the same range of energies but a different value of q/M
are brought to a focus at a different point on the photographic plate. The
46
'.1
1
1
S¡ S2
X ·----%. ·--
P2 1
1
z
Fm. 9--3.
~
Diagram of Aston's first mass spectrograph (1920).
204 ISOTOPES [CHAP. 9
focus for a particular value of q/M is actually a line, and the result of an
analysis is a series of lines reminiscent of an optical line spectrum. Because
of this similarity, the series of lines is called a mass spectrum and the
apparatus a mass spectrograph. Sorne typical mass spectra are shown in
Fig. 9-4.
The isotopic masses can be determined quantitatively in severa) different
ways. In one method, for example, the positions of the lines caused by
the masses in question are compared with the positions of the lines caused
by standard substances whose masses are accurately known. With this
new instrument, Aston was able to prove beyond a doubt that neon has
two isotopes with masses very close to the integers 20 and 22, respec-
tively. <4> He also showed that in the mass spectrum of chlorine there
was no line corresponding to the chemical atomic weight (35.46) of chlorine.
Instead, two lines were seen, corresponding very closely to the masses
35.0 and 37.0, the former being the more intense line. This result showed
that chlorine has two isotopes of nearly integral atomic mass, and ordinary
chlorine is a mixture of these two kinds of atoms in such proportions that
the chemical atomic weight is 35.46.
Aston improved the design of the mass spectrograph, and in bis second
instrument< 5 > isotopic masses could be determined with an accuracy of
1 part in 10,000. With this instrument, Aston determined the masses
of the isotopes of a large number of elements, as well as their abundances,
and showed that the masses of atoms are very nearly, but not quite, integers,
when the mass of oxygen is taken as 16. For example, the isotopes of
chlorine were found to have the masses 34.983 and 36.980 rather than
35.0 and 37.0, respectively. Later instruments designed and built by
Aston, Dempster, Bainbridge, Jordon, Mattauch, and others have yielded
isotopic masses with accuracies approaching 1 part in 100,000.
Modem mass spectroscopic measurements are based on the mass-
doublet technique in which the quantity actually determined is the differ-
ence in mass between two ions of the same mass number but having
slightly different masses. The newer spectrometers yield high dispersion,
\!!'!1':::'!! ¡;¡ ~ ~~~::,~:¡~ ~ ~

, e_ 1 1 ' 1 J I b' ' ' 1 i ' ' 22.aaz::..."::_ ~
:ll1 11!1
11 Cl
lllm:al
VI ■
'
Fm. 9-4. Mass spectra obtained by Aston with his first spectrograph (1920).
9-3) ISOTOPIC MASSES AND ABUNDANCES 205
that is, the distance between the two lines of a doublet are relatively
large, and differences in mass can be measured with high precision. The
mass of one of the members of the doublet must be known accurately.
Hydrocarbon compounds (which give molecular ions) are used as sources
of reference masses because of the relative ease with which fragments of
almost any mass number can be obtained for comparison with other ions
of the same mass number. The masses of C 12, the carbon isotope of mass
number 12, and H1, the hydrogen isotope of mass number 1, are used as
secondary standards, with 0 16 the primary standard.
The masses of C 12 and H 1 relative to that of 0 16 can be determined in a
number of different ways. One method is to measure the following mass-
doublet differences:
(016h - sª2 = a,
cc12>• - sª2O16= b,
c12cH1>• - o16 = c.
The three equations are solved simultaneously, with 0 16 = 16 atomic

mass units (exactly), and the result is
S 32 = 32 - a,
c12 = 12 +b - a,
4
H1 = 1 +a - b +e_
16
More complicated cycles involving a larger number of atoms can also be

used, and highly precise results can be obtained; values of mass-doublet
differences have been compiled, <5 •7 > and values of atomic masses deter-
mined by this method.<s- 17>
It will be seen in Chapter 11 that information from nuclear reactions
can also be used to determine atomic masses with precision comparable
to that obtained with mass spectroscopic methods. Authoritative com-
pilations of atomic masses usually combine results obtained with both
methods. os- 2 o
Other techniques have also been developed <22 > which will only be
mentioned here. The chronotron of Goudsmit< 23- 25 > measures the time
needed for ions to describe a number of revolutions in a uniform
magnetic field. This time is proportional to the mass of the ion, with the
result that the precision is practically constant for all masses, whereR.S
the precision which can be obtained with the mass spectrograph decreases
with increasing mass. Another device which depends on the angular
A----+
p
FIG. 9-5. Diagram of Dempst.er's mass spectromet.er.
Volts
935 911 888 5 8665 846
10 • 2
J \ /\
o - µa.o-'
38 39
' -....... ••••JV
40
Atomic weight
41
\
42
FIG. 9-6. Mass spectrum of potassium (Dempster, 1922).
motion of ions in a known magnetic induction is the maaa synchrometer, <25>

which has given good results. Microwave spectroscopy< 27> has also been
u.sed successfully for the measurement of atomic masses.
In the systematic study of isotopes, it is essential to know not only the
isotopic masses, but also the relative numbers of atoros of each isotope of
an element. The mass spectrograph can be u.sed for making abundance
measurements, and indeed many of the isotopes now known were first
discovered and their abundances measured by Aston by means of this
instrument. When u.sed for abundance measurements, however, a mass
spectrograph is inconvenient because a photographic plate is u.sed to
record the different isotopic ions, and the procedure of determining abun-
9-3] ISOTOPIC MASSES AND ABUNDANCES 207
dances from the plate traces is both more tedious and less reliable than the
direct measurements made with a somewhat simpler instrument, the mass
spectrometer.
About the time of the development of the mass spectrograph by Aston,
Dempster< 2 s> built an instrument which was basically simpler and which
was well suited for making abundance measurements, although it could
not be used for making accurate mass measurements. It was called a
mass spectrometer because the ion current was measured electrically
rather than recorded on a photographic plate. A schematic diagram
of one of Dempster's spectrometer models is shown in Fig. 9-5. Ions
of the element to be analyzed are formed by heating a salt of the element,
or by bombarding it with electrons. Upon emerging from the source A,
the ions are accelerated through a potential difference V of about 1000
volts by an electric field maintained in the region between A and C. A slit
S 1 in the plate C allows a narrow bundle of ions to pass into the region
of the magnetic field H. In passing from A to C, the positive ions carrying
a charge q acquire energy equal to qV. This energy may also be repre-
sented by ½Mv 2 , where Mis the mass of the positive ion and v its velocity
on emerging from the slit in C; consequently,
qV = ½Mv 2 • (9-4)
If the magnetic field is perpendicular to the plane of the paper, the ions
180
~ 160
.2l
"s 140
·= 120
"""
"t:l
i 100
"
~
., 80
"...2:!::,
<,)
60 -
e
.9 I
40
-~"
·¡¡;
o
p.. 20 J
o It \j l.,; IJ d ~
207 206 205 204 203 202 201 200 199 198 197 196 195 194
Atomic mass units
Frn. 9-7. Mass spectrum of mercury (Nier, 1937).

will be forced to move along a circular path defined by the relationship
Mv 2
Hqv = R' (9-5)
where R is the radius of the circle. Since the radius of the circle must
have a certain definite value in order that the ions enter the slit S2 and
be detected by an electrometer at P, it is clear that ions with only one
particular value of q/M, say q/M', will be received for a given combination
of accelerating potential and magnetic field. By varying the potential
difference V, ions with different values of q/M are, in turn, made to pass
through the second slit S 2 to the collector plate P. The current recorded
by the electrometer is proportional to the number of positive ions reaching
it per unit time and, since each accelerating potentíal corresponds to a
definite mass of particle reaching the electrometer, the current can be
plotted against the atomic weight. A typical curve obtained by Dempster
Phillips
ion gauge
Gas
inlet==-lli~P==.,r n
rce
'Magnet
Mercury
diffusion
pump
Preamplifier
Fm. 9-8. Schematic diagram of a Níer mass spectrometcr (Xíer< 30l).
9-4] THE STABLE ISOTOPES OF THE ELEMEXTS 209
for potassium with a slightly modified apparatus is shown in Fig. 9-6.

There are two isotopes with atomic weights 39 and 41 units, the former
being much more abundant; for the ratio of the abundances, Dempster
obtained 18: l.
During recent years, owing largely to improved vacuum techniques and
the development of new methods of electrical measurement, mass spectrom-
etry has advanced rapidly. One of the most useful instruments was
that of Nier, c29 > which had extremely high resolving power and sensitivity
and was especially designed for searching for rare isotopes and measuring
relative abundances. It was a modification of the Dempster apparatus
and gave results like those shown in Fig. 9-7. The figure shows a
"mass spectrogram" of mercury, and the isotope abundances are closely
proportional to the magnitude of the peaks of the positive ion current.
A more recent form of spectrometer designed by Nierc 3 o> is shown in
Fig. 9-8. Ions are produced by the electron bombardment of the gas
under investigation, and are accelerated by a potential drop of about
1000 volts. The beam of ions passes into a wedge-shaped magnetic field
in which they suffer a deflection of 60°, rather than 180° as in the Dempster
spectrometer. This deflection makes it possible to obtain high resolution
with a simple magnet. In the diagram the ion beam is shown broken into
two parts which fall on separate collectors and are measured by separate
amplifiers. This instrument has been used to get highly precise isotopic
abundances.
9-4 The stable isotopes of the elements and their percentage abun-
dances. The mass spectra of the elements have been investigated in detail
and the isotopic composition of the elements has been determined. c3 i. 32 , 33 >
The results obtained for the isotopes of 83 elements are collected in
Table 9-1. Before discussing these results, sorne remarks about termin-
ology are in order. In recent years, the term nuclide has been widely
accepted for a species of atom characterized by the constitution of its
nucleus, i.e., by the numbers of protons and neutrons it contains. Thus,
the atomic species listed in Table 9-1 may be referred to as the naturally
occurring stable nuclides. Similarly, every radioactive species is a radio-
active nuclide or radionuclide. An isotope is then one of a group of two
or more nuclides having the same number of protons or, in other w'ords,
having the same atomic number. An element like beryllium or aluminum,
of which only one species exists in nature, is said to form a single stable
nuclide, rather than a single stable isotope, since the word isotope implies
more than one species occupying the same place in the periodic system.
A nuclide is usually indicated by the chemical symbol with a subscript
at the lower left giving the atomic number, and a superscript at the upper
right giving the mass number; in the symbol zSA, Z is the atomic number,
TABLE 9-1
THE STABLE KUCLIDES AND THEIR PERCENTAGE ABUNDANCE
Atomic Mass Rela.tive Atomic Mass Relative

Sym- Sym-
number, number, abundance, number, number, abundance,
bol bol
z A % z A %
H 1 1 99.9849- Al 13 27 100
99.9861
Si 14 28 92.21
D 1 2 0.0139- 29 4.70
0.0151 30 3.09
He 2 3 ~10-Ll0-• p 15 31 100
4 ~100
s 16 32 95.0
Li 3 6 7.52 33 0.760
7 92.48 34 4.22
36 0.014
Be 4 9 100
Cl 17 35 75.529
B 5 10 18.45-- 37 24.471
19.64
11 80.36- A 18 36 0.337
81.55 38 0.063
40 99.600
e 6 12 98.892
13 1.108 K 19 39 93.10
40 0.012
99.634 41 6.88
N 7 14
15 0.366
Ca 20 40 96.97
42 0.64
o 8 16 99.759
43 0.145
17 0.037
44 2.06
18 0.204
46 0.003
48 0.185
F 9 19 100
Se 21 45 100
Ne 10 20 90.92
21 0.257 Ti 22 46 7.93
22 8.82 47 7.28
48 73.94
Na 11 23 100 49 5.38
50 5.34
Mg 12 24 78.70
25 10.13 V 23 50 0.24
26 11.17 51 99.76
9-4) THE STABLE ISOTOPES OF THE ELEMENTS 211
TABLE 9-1 (Continued)
Atomic Mass Relative Atomic Mass Relative

Sym- Sym-
bol
bol
z A % z A %
Cr 24 50 4.31 Br 35 79 50.54
52 83.76 81 49.46
53 9.55
54 2.38 Kr 36 78 0.354
80 2.27
Mn 25 55 100 82 11.56
83 11.55
Fe 26 54 5.82
84 56.90
56 91.66
86 17.37
57 2.19
58 0.33
Rb 37 85 72.15
Co 27 59 100 87 27.85
Ni 28 58 67.88 Sr 38 84 0.56
60 26.23 86 9.86
61 1.19 87 7.02
62 3.66 88 82.56
64 1.08
y 39 89 100
Cu 29 63 69.1
65 30.9 Zr 40 90 51.46
Zn 30 64 48.89 91 11.23
66 27.81 92 17.11
67 4.11 94 17.40
68 18.56 96 2.80
70 0.62
Nb 41 93 100
Ga 31 69 60.4
71 39.6 Mo 42 92 15.84
94 9.04
Ge 32 70 20.52 95 15.72
72 27.43 96 16.53
73 7.76 97 9.46
74 36.54 98 23.78
76 7.76 100 9.63
As 33 75 100
Ru 44 96 5.51
Se 34 74 0.87 98 1.87
76 9.02 99 12.72
77 7.58 100 12.62
78 23.52 101 17.07
80 49.82 102 31.61
82 9.19 104 18.58
(Continued)
Atomic Masa Relative Atomic Masa Relative

Sym- Sym-
number, number, abundan ce, number, number, &bundance,
bol bol
z A % z A %
Rh 45 103 100 Te 52 126 18.71

128 31.79
Pd 46 102 0.96 130 34.49
104 10.97
105 22.23 I 53 127 100
106 27.33
108 26.71
Xe 54 124 0.096
110 11.81
126 0.090
128 1.919
Ag 47 107 51.35
129 26.44
109 48.65
130 4.08
131 2l.l8
Cd 48 106 1.215
132 26.89
108 0.875
134 10.44
110 12.39
136 8.87
111 12.75
112 24.07
113 12.26 Ca 55 133 100
114 28.86
116 7.58 Ba 56 130 0.101
132 0.097
In 49 113 4.28 134 2.42
115 95.72 135 6.59
136 7.81
Sn 50 112 0.96 137 11.32
114 0.66 138 71.66
115 0.35
116 14.30 La 57 138 0.089
117 7.61 139 99.911
118 24.03
119 8.58
Ce 58 136 0.193
120 32.85
138 0.250
122 4.72
140 88.48
124 5.94
· 142 11.07
Sb 51 121 57.25
123 42.75 Pr 59 141 100
Te 52 120 0.089 Nd 60 142 27.11

122 2.46 143 12.17
123 0.87 144 23.85
124 4.61 145 8.30
125 6.99 146 17.22
9-4] THE STABLE ISOTOPES OF THE ELEMENTS 213
Atomic M888 Relative Atomic M888 Relative

Sym- Sym-
bol bol
z A % z A %
Sm 62 144 3.09 Yb 70 172 21.82

147 14.97 173 16.13
148 11.24 174 31.84
149 13.83 176 12.73
150 7.44
152 26.72 Lu 71 175 97.40
154 22.71 176 2.60
Eu 63 151 47.82 Hf 72 174 0.18

153 52.18 176 5.20
177 18.50
Gd 64 152 0.20 178 27.14
154 2.15 179 13.75
155 14.73 180 35.24
156 20.47
157 15.68 Ta 73 181 100
158 24.87
160 21.90 w 74 180 0.135
182 26.41
Tb 65 159 100 183 14.40
184 30.64
Dy 66 156 0.0524 186 28.41
158 0.0902
160 2.294 Re 75 185 37.07
161 18.88 187 62.93
162 25.53
163 24.97 Os 76 184 0.018
164 28.18 186 1.59
187 1.64
Ho 67 165 100 188 13.3
189 16.1
Er 68 162 0.136 190 26.4
164 1.56 192 41.0
166 33.41
167 22.94 Ir 77 191 37.3
168 27.07 193 62.7
170 14.88
Pt 78 190 0.0127
Tm 69 169 100 192 o:78
194 32.9
Yb 70 168 0.135 195 33.8
170 3.03 196 25.3
171 14.31 198 7.21
(Continued)
TABLE 9-1 (Concluded)
Atomic Mase Relative Atomic Mase Relative

Sym- Sym-
bol bol
z A % z A %
Au 79 197 100 Pb 82 204 1.48

206 23.6
Hg 80 196 0.146 207 22.6
198 10.02 208 52.3
199 16.84
200 23.13 Bi 83 209 100
201 13.22
202 29.80 Th 90 232 100
204 6.85
u 92 234 0.0056
TI 81 203 29.50 235 0.7205
205 70.50 238 99.2739
A is the mass number, and S is the chemical symbol. Thus, the symbol
120
50Sn represents the nuclide with 50 protons and a mass number of
120; it is one of the ten isotopes of tin. The number of neutrons is, of
course, equal to A - Z. Sometimes it is not necessary to show the number
of protons explicitly, and the symbol for a nuclide is then shortened to s..t.
Not ali of the nuclides listed in Table 9-1 are actually stable. Thorium
and uranium are radioactive, but they occur in sufficient amounts and
with sufficiently weak activity so that they can be handled in the same
way as the stable elements. At least nine naturally occurring isotopes
of "atable" elements show feeble radioactivity· K 4 º Rb 87 In 115 La 138
Nd 144 , Sm 147 , Lu 176 , Re 187 and Pt 190 • The~ nuclides ar~ disti~ct fro~
the families or chains of the heavy naturally occurring radionuclides and
are much feebler in activity. It is therefore more convenient to include
them with the stable elements than with the radioactive ones.
With the exceptions noted, Table 9-1 contains 284 stable nuclides
divided among 83 elements. Twenty elements, about one-fourth in ali,
are single species; ali the others consist of two or more isotopes. Hydrogen
has two isotopes, the one with mass number 2 having a relative abundance
of only about 0.015%. This rare isotope, however, has a mass about
double that of the common isotope, so that the difference in mass is as
great as the mass of the ordinary hydrogen atom itself. This relation-
ship between the masses is an exceptional one and, as a result, the differ-
ences between the properties of the two isotopes are more marked than in
any other pair of isotopes. The hydrogen isotope of mass 2 has therefore
been given its own name, deuterium, with the occasionally used symbol D.
TABLE 9-2
SoME lsoToPE 8TATIST1cs
Number of Average number

Odd A Even A Total
elements of isotopes
OddZ 40 53 8 61 1.5
EvenZ 43 57 166 223 5.2
Total 83 llO 174 284 3.4
Carbon and nitrogen also have two isotopes, while oxygen has three, two
of which are rare. Tin has the greatest number of isotopes, ten, while
xenon has nine, cadmium and tellurium have eight each, and several
elements have seven.
There are sorne striking regularities in Table 9-1. Nuclides of even Z
are much more numerous than those of odd Z, and nuclides of even A are
much more numerous than those of odd A. Nearly all nuclides with even
2 6 10
A have even Z, the only comroon exceptions being 1 H , 3 Li , 5B , and
40 176
14
7 N . The nuclides 19K and 71 Lu have odd Z and even A but are
weakly radioactive, while 23V 50 and 57 La 138 are very rare. The nurobers
of nuclides with the various corobinations of even and odd atomic and
mass numbers are listed in Table 9-2. Of the 20 elements which have
only a single nuclide, only beryllium has an even value of Z, while the
other 19 have odd z. Nineteen elements with odd Z have two isotopes
apiece, and each of these nuclides has odd A. One element with odd Z,
potassium, has three isotopes; two of these have odd mass numbers, and
the only isotope with an even mass number is the weakly radioactive K 40.
The four common elements which have odd values of Z and A, hydrogen
lithium, boron, and nitrogen, have equal numbers of protons and neutrons.
The elements which have more than two isotopes (apart from potassiuro)
all have even values of Z.
An examination of the values of Z and A in Table 9-1 shows that in the
stable nuclei, with the exception of H 1 and He 3 , the nui;nber of neutrons
is always greater than or equal to the number of protons. There is always
at least one neutron for each proton. This property of the stable nuclides
is shown in Fig. 9-9, in which the number of neutrons A - Z is plotted
against the number of protons. The number of neutrons which can be
included in a stable nucleus with a given number of protons is limited.
For example, tin with an atomic nurober of 50 has neutron numbers from
62 to 74, and the mass numbers of the tin isotopes lie between 112 and 124.
Apparently the tin nucleus cannot contain less than 62 neutrons nor more
than 74 neutrons and still remain stable. For the other elements (except
216 ISOTOPES
[CHAP, 9
140
uo
120 ~
.l
110 .r
•
..
¡
100
90 .f:
.....r
1·
.. ...
;J
.f. .
.. •
40
.,,r .il'=
. lf.•
'
l,r-'
20
Sta.bte nuclei
10
,7
• 10 20 ílO 40 50 60 70 80 90
Proton number (Z)
FIG. 9-9. Neutron-proton plot of the atable

nuclides.
xenon) this range is smaller and the limits for the existence of stable nuclei
are narrower.
The regularities that have just been noted are closely connected with
the problem of nuclear stability and must eventually be accounted for in
terms of the proton-neutron theory of the nucleus, in particular, in terms
of the forces between nucleons. From this standpoint, the hydrogen
isotope of mass two (deuterium) is especially important because its nucleus
consists of one proton and one neutron. The properties of this nucleus
yield information about the force between a proton and a neutron, and
deuterium has a prominent place in theoretical nuclear physics.
In the preceding discussion it was assumed that the relative abundance
of an isotope is constant in nature, that is, independent of the source of
the sample that is measured. This assumption is true, in general. There
are, however, sorne exceptions which, although interesting, do not affect
any of the conclusions previously drawn from the consideration of abun-
dances. The relative abundances of H 1 and H 2 depend somewhat on the
source, and the spread in the values listed in Table 9-1 shows that the
variation is small. The ratio of hydrogen to deuterium abundance has
been determined as 6700 ± 50 for samples of tap water from London,
Osaka, and various cities in the United States; this corresponds to a value
of 0.0149% for the relative abundance of H 2 in tap water. The range
0.0139 to 0.0151 % includes the variation over a wide range of substances
such as water, snow, ice, organic compounds, and animal and mineral
materials from many ·sources. There are wide variations in the ratio of
He 3 and He 4 abundances; for example, the abundance of He 3 is approxi-
mately ten times as great in atmospheric helium as in well helium. Helium
is formed in radioactive minerals because of alpha-decay, and such helium
is all He 4 • In nonradioactive ores, the He 3 content varies widely, and it
has been suggested that He 3 can be formed as a result of the transfor-
mation of various nuclides in the air and on the ground by cosmic-ray
bombardment. Cosmic rays are highly penetrating radiations which
originate outside the earth a~d consist of protons, electrons, neutrons,
photons, and other particles. The range of abundances of the boron
isotopes shown in the table is equivalent to a variation in the B 11 /B 1 º ratio
from 4.27 to 4.42 or about 3%. Although this variation seems small,
it affects the use of certain boron standards in nuclear physics; it is neces-
sary, therefore, to cite the abundance ratios when measurements involving
this standard are reported. The abundance of the carbon isotopes cited in
Table 9-1 are those found in limestone and correspond to a C 12 /C 13 ratio
of 89.2; in coa! the ratio is 91.8. In general, the relative content of C 12
seems to be somewhat greater in plant material than in limestone.
The relative abundances of the oxygen isotopes also vary, and the value
of the 0 16/0 18 ratio in nature has a spread of about 4%. The values of
218 ISOTOPF.S [CHAP, 9
the abundances quoted in the table are those for atmospheric oxygen and
correspond to a value of the 0 16/0 18 ratio of 489.2 ± 0.7. This value
is also correct for the oxygen from limestone, but for oxygen from water
or iron ores the ratio may be 4% higher. The values used for the abun-
dances of the oxygen isotopes affect the value of the factor for converting
atomic weights from the physical scale to the chemical scale (see Section
9-5). The variation in the value of the conversion factor may be large
enough to affect the precision of atomic weight determinations and, if
the latter are to be made to more than five significant figures, the isotopic
composition of the oxygen used as a reference must be specified.
Wide variations in the abundances of the lead isotopes are also found,
and these are usually associated with the radioactive sources from which
the different lead samples are derived. Even for common lead, it is not
possible to give exact isotopic abundances without specifying the source
of the material; the values in the table are for Great Bear Lake galena.
The abundance variations which have been discussed are the most
important ones known and, for the remaining elements, either there are
no significant variations, or else they do not seriously affect further work
in nuclear physics.
9-5 Atomic masses: packing fractions and binding energies. Some
atomic masses' 18 • 19 •36> are given in Table 9-3. The standard of mass
used here is slightly different from that used for the chemical atomic
weights. It is seen from Table 9-1 that oxygen actually has three isotopes,
the most abundant of which has the mass number 16. The other two
isotopes together constitute only about 0.2% of the oxygen atoms. In
the determination of isotopic weights by the mass spectrograph it is the
practice to take as the standard the value of 16.00000 for the weight of
the common isotope of oxygen. The weights so obtained differ slightly
from those based on the ordinary chemical atomic weight scale. In the
latter case, the number 16.00000 is associated with ordinary atmospheric
oxygen, which is a mixture of isotopes, whereas on the mass spectrographic,
or physical, atomic weight scale, this is taken as the isotopic weight of
the single, most abundant isotope. The relationship between the chemical
and physical atomic weight scales may be determined in the following way.
Atmospheric oxygen consists of 99.759% of the isotope of mass 16.00000,
together with 0.037% of the isotope of mass 17.004529, and 0.204% of
the isotope of mass 18.004840. The weighted mean of these values is
16.004462, and this is the atomic weight of atmospheric oxygen on the
physical sea.le, as compared with the postulated value of 16.00000 on the
chemical sea.le. Then
Physical atomic weight _ 16.004462
1.000279.
Chemical atomic weight - 16.000000 -
9-5) ATOMIC MASSES 219
TABLE 9-3
ATOMIC MASSES, PACKING FRACTIONS, AND BINDING ENERGIES

OF SoME OF THE 8TABLE NucLIDES
Binding energy
Number Number Packing
of of M888,
fraction, per
Nuclide amu total,
protons, neutrons, nucleon,
x10•
z A -Z Mev
Mev
nl o 1 1.0089830 (*l. 7) 89.8

Hl 1 o 1.0081437 ( •1.8) 81.4
H2 1 1 2.0147361 ( •2.9) 73.7 2.225 1.113
He• 2 2 4.0038727 ( •2.1) 9.7 28.29 7.07
Li7 3 4 7.018222 (•6) 26.0 39.24 5.61
Be• 4 5 9.015041 (•5) 16.7 58.15 6.46
Bll 5 6 11.012795 (•5) 11.6 76.19 6.93
cu 6 6 12.0038065 ( •3.9) 3.2 92.14 7.68
cu 6 7 13.0074754 ( •4.1) 5.8 97.09 7.47
Nu 7 7 14.0075179 ( •3.0) 5.4 104.63 7.47
ou 8 8 16.0000000 o 127.58 7.97
Q17 8 9 17.0045293 ( •3.9) 2.7 131.73 7.75
011 8 10 18.004840 (•9) 2.7 139.80 7.77
F•• 9 10 19.004447 (•7) 2.3 147.75 7.78
Ne 20 10 10 19.998765 (•10) -0.6 160.62 8.03
A121 13 14 26.990080 (•14) -3.7 224.92 8.33
Si 28 14 14 27.985777 (*16) -5.1 236.51 8.45
pu 15 16 30.983563 (•18) -5.3 262.88 8.48
su 16 16 31.982190 (•20) -5.6 271.74 8.49
CIU 17 18 34.979906 (•80) -5.7 298.13 8.52
Ci37 17 20 36.9776573 ( • 16) -6.0 317.00 8.57
Au 18 22 39.975088 (*4) -6.2 343.71 8.59
Ca• 0 20 20 39.975330 (*30) -6.2 341.92 8.55
Fe 51 26 30 55.952722 (*6) -8.4 492.11 8.79
euu 29 34 62.949607 (*11) -8.0 551.22 8.75
As?a 33 42 74.945510 (*100) -7.3 652.23 8.70
Sr 81 38 50 87.933680 (•300) -7.5 768.13 8.73
Mo 91 42 56 97.937240 (•350) -6.3 845.73 8.63
Sn 118 50 66 115.938850 (•300) -5.3 988.14 8.52
Snuo 50 70 119.940330 (*140) -5.0 1020.2 8.50
Xeuo 54 76 129.944810 (•30) -4.3 1096.6 8.44
Xe 131 54 82 135.950420 (•25) -3.7 1141.5 8.39
NdUO 60 90 149.968490 (•70) -2.1 1237.1 8.25
Hf•1• 72 104 175.99650 (*800) -0.2 1419.1 8.06
wu• 74 110 184.008300 (•600) 0.4 1473.5 8.01
Auu1 79 118 197.028000 (•1000) 1.4 1560.0 7.92
Pb•oa 82 124 206.037900 (•500) 1.8 1623.7 7.88
Th212 90 142 232.109800 (•500) 4.7 1768.0 7.62
un• 92 146 238.124300 (•500) 5.2 1803.1 7.58
Hence, isotopic weights obtained by means of the mass spectrograph must

be divided by 1.000279 in order to convert the results to the chemical
atomic weight scale. The conversion factor is used only when it is neces-
sary to compare mass spectrographic results with weights obtained by
chemical methods; in nuclear work the atomic masses on the physical
scale are used.
Estimates of the errors of the mass spectrographic measurements have
been included in order to indicate the high precision with which atomic
masses can now be determined. The error, given in parentheses, is ex-
pressed in units of 10- 6 amu. This precision is essential in the study of
nuclear reactions and transformations, as will be seen in later chapters.
It is also important in the precise calculation of packing fractions and
binding energies, as will now be shown.
It is seen from Table 9-3 that the isotopic masses are indeed very close
to whole numbers. It seemed clear to the early workers in this field that
the systematic study of the divergences of the masses of nuclides from
whole numbers was an important problem. Aston expressed these díver-
gences in the form of a quantity called the packing fraction defined by
p k" f t" Atomic mass - Mass number

ac mg rae 10n = Mass number
Mz,A - A
A
where Mz,A is the actual weight of a nuclide on the physical atomic weight
scale, and A is the mass number. If the packing fraction is denoted by f,
then
Mz,A = A(I + f). (9-7)
Sample values of the packing fraction are listed in Table 9-3, and the varia-
tion of the packing fraction with mass number for a larger number of
nuclides is shown graphically in Fig. 9-10. The packing fractions, with
the exception of those for He4, C 12 , and 0 16, fall on ornear the solid curve.
The values a.re high for elements of low mass number, a.pa.rt from the
nuclides mentioned. For 0 16, the va.lue is zero, by definition. As A
increa.ses, the packing fraction becomes negative, passes through a rather
tla.t minimum a.nd then rises gradually, becoming positive again at values
of A of a.bout 180. The packing fraction was very useful in the study of
isotopic masses, but it <loes not ha.ve a precise physical meaning. The
explanation for its usefulness will appear from the discussion of the binding
energies of nuclei.
The a.tomic mass of a nuclide can be understood in terms of the masses
of its constituent particles and a quantity called the binding energy. It
9-5] ATOMIC MASSES 221
100
80
60
f
X
.j 40
~
.:::
bl)
:é
~
~
20
-10
o 20 40 60 80 100 120 140 160 180 200 220 240

M81!8 number (A)
Fm. 9-10. Packing fraction as a function of mass number.
might at first be supposed that the mass of an atom should be the sum of
the masses of its constituent particles. A survey of the atomic masses
shows, however, that the atomic mass is less than the sum of its constituent
particles in the free state. To account for this difference in mass, the
principie of the equivalence of mass and energy, derived from the special
theory of relativity, is used. If 11M is the decrease in mass when a number
of protons, neutrons, and electrons combine to form an atom, then the
above principie states that an amount of energy equal to
11E = c
2
11M (9--8)
is released in the process. The difference in mass, 11M, is called the masa
defect; it is the amount of mass which would be converted to energy if a
particular atom were to be assembled from the requisite numbers of pro-
tons, neutrons, and electrons. The same amount of energy would be
222 ISOTOPES [CHAP, 9
9
20 40 60 80 100 120 140 160 180 200 220 240

Mass number
Fm. 9-11. Binding energy per nucleon as a function of masa number.
needed to break the atom into its constituent particles, and the energy
equivalent of the mass defect is therefore a measure of the binding energy
of the nucleus. The mass of the constituent particles is the sum of Z
proton masses, Z electrons, and A - Z neutrons. The proton and electron
masses can be combined into the mass of Z hydrogen atoms because the
minute change in mass which may accompany the formation of a hydrogen
atom from a proton and electron is negligible. The mass defect can then
be written
llM = Zma + (A - Z)m,. - Mz,A, (9-9)
where ma, the mass of the hydrogen atom is 1.0081437 mass units and m,.,
the mass of the neutron, is 1.0089830 mass units. Then
ti.M = 1.0081437Z + 1.0089830(A - Z) - Mz,,1.. (9-10)
Since one atomic mass unit is equivalent to 931.145 Mev, the binding
energy of the nucleus is given by
MJ(Mev) = 931.145[1.0081437Z + l.0089830(A - Z) - Mz,,1.]. (9-11)
The average binding energy per nucleon is obtained by dividing the
total binding energy of the nucleus by the mass number A. Sorne values of
ATOMIC MASSES 223
the binding energy obtained by the method just outlined are listed in the
last two columns of Table 9-3, and a graph of the binding energy per
nucleon as a function of the mass number is shown in Fig. 9-11. With
the exception of He', C 12 , and 0 16 , the values of the binding energy per
particle lie on or close to a single curve. The binding energies of sorne
of the very light nuclides, such as H2 , are very small. The binding energies
of He", C 12, and 0 16 are considerably greater than those of their neighbors,
as is shown clearly by the values in the last column of Table 9-3: The
binding energy per particle rises sharply, and reaches a maximum value of
about 8.8 Mev in the neighborhood of A = 50. The maximum is quite
flat, and the binding energy is still 8.4 Mev at about A = 140. For higher
mass numbers, the value decreases to about 7.6 Mev at uranium.
The magnitude of the binding energy is enormous, as can be shown by
converting from Mev per nucleon to more familiar units such as Btu per
pound. One Mev is equivalent to 1.519 X 10- 16 Btu. One pound of
an atomic species contains 453.6/M gram atomic weights (M is the atomic
weight), or
453.6 X 6.023 X 10 23 X A/M nucleons.
The mass number A and the atomic weight M are practically the same,
so that one pound contains 2.73 X 10 26 nucleons. The unit "1 Mev/
nucleon" is then practically the same as 4.15 X 10 10 Btu/pound. In the
neighborhood of the maximum of the binding energy curve, the binding
energy is a.bout 8.8 Mev /nucleon or 350 billion Btu/pound. This enormous
value of the energy that would be needed to dissociate a nucleus into its
constituent protons and neutrons is another indication of the magnitude
of the nuclear forces.
The application of the principie of the equiva.lence of mass and energy
and the introduction of the concept of binding energy have, thus far in
our treatment, only a theoretieal basis. In recent years, however, many
nuclear transmutations ha.ve been accura.tely studied and ca.reful mea.sure-
ments ha.ve been ma.de of the changes of mass and energy in these reactions.
The va.lidity of the rela.tivistic mass-energy relationship has been proven,
as well as that of its application to the problems of nuclear physics. The
binding energy of a nucleus is, therefore, a quantity with real physical
meaning, and the ma.sses and binding energies of nuclei yield useful informa-
tion about the constitution and stability of nuclei. In a.ddition, the ana.lysis
of nuclear transforma.tions has provided an independent mea.ns of determin-
ing atomic ma.sses which is a.t present even more powerful than the mass
spectrographic method. These matters will be discussed in detall in later
chapters.
The subject of a.tomic masses should not be left without a brief dis-
cussion of a proposa1<H> to replace the mass standard, 0 16, by C 12, i.e.,
to assign to C 12 the atomic mass 12.0000000 units. The use of C 12 as

the standard would have several advantages in mass spectrometry.
Carbon forms many more chemical compounds which can provide mo-
lecular ions for use in a spectrometer than does oxygen. Doubly, triply,
and quadruply charged ions of C 12 occur at integral mass numbers and
can be paired in doublets with ions of mass number 6, 4, and 3, respectively,
so that C 12 would be a convenient standard for atoms of low mass number.
No other element besides carbon forms molecular ions containing as many
atoms of one kind, up to 10 or more; this property would permit many
more doublet comparisons to be made directly with the reference nuclide
than are now possible, and would yield masses with increased precision
at intermediate and large values of A. Thus, carbon forms many com-
pounds with hydrogen, providing easy reference lines for doublets with
masses up to 120 units; for values of A between 120 and 240, doubly
charged ions of heavier elements could be compared directly with singly
charged ions of the type (C 12 ),. or (C 12),. (H 1 ),,..
A table of atomic masses has been preparedc 35 > based on C 12 as the
standard and the following results have been obtained.
Nuclide 0 16 Standard C 12 Standard
n 1.0089861 1.0086654
Hl 1.0081456 1.0078252
H2 2.0147425 2.0141022
c12 12.0038150 12.0000000
Ql6 16.0000000 15.9941949
The relation between the atomic mass units on the two scales is
1 amu (C 12) = 1.00031792 amu (0 16).
The conversion factor from mass to energy is
1 amu (C 12 ) = 931.441 Mev,
as compared with 1 a~u (0 16) = 931.145 Mev. The values of binding

energies are unchanged; thus, the binding energy of the deuteron is 2.2247
Mev when either mass standard is used.
Throughout this book, we shall use 0 16 as the mass standard and the
atomic masses given in Table 9-3, but the reader should be aware of the
possibility that the standard may be changed within the next few years.
225
REFERENCES
GENERAL
F. W. AsToN, Mass Spectra and Isotopes, 2nd ed. London: Edward Arnold,
1942.
Sm J. J. THOMSON, Rays of Positive Electricity and Their Application to Chemi-
cal Analyses, 1st ed. London: Longmans, Green and Co., 1913; 2nd ed., 1921.
M. G. INGHRAM, "Modern Mass Spectroscopy," Advances in Electronics, Vol. l.
New York: Academic Press, 1948.
J. D. STRANATHAN, The "Particles" of Modern Phy8ics. Philadephia: Blakis-
ton, 1944, Chapter 5.
K. T. BAINBRIDGE, "Charged Particle Dynamics and Optics, Relative Isotope
Abundances of the Elements, Atomic Masses," Experimental Nuclear Physics,
E. Segre, ed., Vol. l. New York: Wiley, 1953.
1959 Nuclear Data Tables. Nuclear Data Project, National Academy of
Sciences-National Research Council. Washington, D.C.: United States Atomic
Ene1gy Commission. Table IV: Relative Isotopic Abundances; Table VI: Mass
Differences and Ratios.
PARTICULAR
l. J. J. THOMSON, Papers on positive rays and Isotopes, Phil. M ag., 13, 561
(1907); 16, 657 (1908); 18,821 (1909); 20,752 (1910); 21, 225 (1911); 24, 209,
669 (1912).
2. F. W. AsToN, Mass Spectra and lsotopes, op. cit. gen. ref., Chapter 4.
3. F. W. AsToN, "A Positive Ray Spectrograph," Phil. Jtfag., 38, 707 (1919).
4. F. W. AsToN, "The Constitution of Atmospheric Neon," Phil. Jtfag., 39,
449 (1920).
5. F. W. AsToN, "A New Mass Spectrograph and the Whole Number Rule,"
Proc. Roy. Soc. (London), AllS, 487 (1927).
6. DucKW0RTH, Hooo, and PENNINGTON, "Mass Spectroscopic Atomic Mass
Differences," Revs. Mod. Phys., 26, 463 (1954).
7. H. E. DucKWORTH, "Mass Spectroscopic Atomic Mass Differences II,"
Revs. Jlfod. Phys., 29, 767 (1957).
8. QUISENBERRY, ScOLMAN, and NIER, "Atomic Masses of H 1 , D 2 , C 12 , and
32
8 ," Phys. Rev. 102, 1071 (1956).
9. ScoLMAN, Qu1sENBERRY, and N1ER, "Atomic Masses of the Stable Isotopes,
10 ::;; A ::;; 30," Phys. Rev. 102, 1076 (1956).
10. M. E. KETTNER, "Atomic Masses from C 12 to Ne 22," Phys. Rev .. 102,
1065 (1956}.
11. QmsENBERRY, G1EsE, and BENSON, "Atomic Masses of H 1, C 12 , and
32
8 ," Phys. Rev. 107, 1664 (1957).
12. C. F. G1ESE and J. L. BENSON, "Atomic Masses from Phosphorus Through
Manganese," Phys. Rev. 110, 712 (1958).
13. CoLLINs, NIER, and JoHNSON, "Atomic Masses in the Region about Mass
40 and from Titanium through Zinc," Phys. Rev., 84,717 (1951); 86,408 (1952).
14. R. E. HALSTEAD, "Atomic Masses from Palladium through Xenon," Phys.
Rev., 88, 666 (1952).
15. H. E. DucKWORTH and R. S. PRESTON, "Sorne New Atornic Mass Measure-

rnents and Rernarks on the Mass Evidence for Magic Nurnbers," Phys. Rev.,
82, 468 (1951).
16. DucKWORTH, KEGLEY, ÜLSoN, and STAMFORD, "Sorne New Values of
Atornic Masses, Principally in the Region of 82 Neutrons," Phys. Rev., 83,
1114 (1951).
17. STAMFOBD, DucKW0BTH, Hooo, and GEIGER, "Masses of Pb 2º8 , Th 232,
U 23 4, and U 238 ," Phys. Rev., 86, 617 (1952).
18. J. MATTAUCH and F. EvERLING, "Masses of Atoms of A < 40," Progrua
in Nuclear Physics, O. R. Frisch, ed., Vol. 6, p. 233. London: Pergarnon Press,
1957.
19. H. E. DucKWORTH, "Masses of Atoms of A > 40," Progre,, in Nuclear
Physica, O. R. Frisch, ed., Vol. 6, p. 138. London: Pergarnon Press, 1957.
20. A. H. WAPSTRA, "Atornic Masses of Nuclides," Handbuch der Physik,
Vol. 38, Part l, p. l. Berlin: Springer Verlag, 1958.
21. American Instilute of Physics Handbook, p. 8-6. New York: McGraw-Hill,
1957.
22. H. HINTENBERGER, ed., Nuclear Ma88e8 and their Determination. London:
Pergarnon Press (1957).
23. S. GounsMIT, "A Tirne-of-Flight Mass Spectrorneter," Phya. Rev., 74,
622 (1948).
24. HAYS, RICHARDS, and Gouos111T, "Mass Measurernents with a Magnetic
Tirne-of-Flight Mass Spectrorneter," Phya. Rev., 84, 824 (1951); 85, 1065 (1952).
25. R1cHARDS, HAYS, and Gouos11JT, "Masses of Lead and Bismuth," Phy,.
Rev., 85, 630 (1952).
26. L. G. S111TH, "Measurernents of Light Masses with the Mass Synchrom-
eter," Phys. Rev., 111, 1606 (1958).
27. S. GEscHWIND, "Determination of Atomic Masses by Microwave Meth-
ods," Handbuch der Phyaik, Vol. 38, Part 1, p. 38. Berlin: Springer Verlag,
1958.
28. A. J. DEMPSTER, "A New Method of Positive-Ray Analysis," Phys. RetJ.,
11, 316 (1918); "Positive-Ray Analysis of Potassium, Calciurn and Zinc," Phy,.
Rev., 20, 631 (1922).
29. A. O. NIER, "Mass Spectrograph Studies of the lsotopes of Various Ele-
ments," Phys. Rev., 50, 1041 (1936); 52, 933 (1937).
30. A. O. N1ER, "A Mass Spectrometer for Isotope and Gas Analysis," Ret,.
Sci. Instr., 18, 398 (1947).
31. K. T. BAINBRIDGE and A. O. NlER, Relotive Iaoúrpic Abundancea of the
Elementa, Preliminary Report No. 9, Nuclear Science Series, Division of Mathe-
matical and Physical Sciences of the N ational Research Council, Washington,
D.C. (1950).
32. HoLLANDER, PERLMAN, and SEABORG, "Table of Isotopes," Revs. Mod.
Phya., 25, 469 (1953).
33. K. WAY, ed., 1969 Nuclear Dota Tablea. National Academy of Sciences-
National Research Council. Washington, D.C.: U.S. Government Printing
Office, April 1959, Table V, pp. 66-88.
PROBLEMS 227
34. KoHMAN, MATrAUCH, and WAPSTRA, "New Reference Nuclide," Science

127, 1431 (1958).
35. EvERLING, KoNIG, MATrAUCH, and WAPSTRA, "Atomic Masses of Nu-
clides, A ~ 10," Nuclear Phyaics; 15, 342 (1960); 18, 529 (1960).
36. BHANOT, JoHNSON, and N1ER, "Atomic Masses in the Heavy M&SS
Region," Phys. Rev., 120, 235 (1960).
PROBLEMS
l. In a m&SS spectrometer, a singly charged positive ion (q = 1.602 X 10-20

emu) is accelerated through a potenti&l difference of 1000 volts. lt then travels
through a uniform magnetic field for which H = 1000 gauss, and is deftected
into a circular path 18.2 cm in radius. What is (a) the speed of the ion? (b)
the m&SS of the ion, in grams and atomic m&SS units? (c) the m&SS number of
the ion?
2. Show that for a singly charged ion in a m&SS spectrometer, the following
relation holds:
MV = 4.826 X 10- 5H 2 R 2 ,
where R is expressed in centimeters, H in gauss, V in volts, and lit is in atomic

mass units relative to 0 16 = 16.0000. Use the formula to check the result of
Problem l.
3. Suppose that singly charged ions with maeses close to 12 and 14 amu are
accelerated through a potential diflerence of 1000 volts and then travel through
a magnetic field of 900 gauss. Where should collector plates for the two different
ions be located?
4. Plot the number of stable isotopes per element against the atomic number
Z. At what values of Z are maxima observed?
5. Plot the number of stable nuclides against the neutron number A - Z.
At what values of A - Z are maxima observed?
6. From the values of the atomic maeses listed in Table 9-3, calculate the
binding energy of the last proton in C 12, Nª, F 19, Ne 20, and Si 28 • Compare
the results with the average binding energy per nucleon.
7. Suppose that C 13 , N 13, N 14, and 0 16 were formed by combining the ap-
propriate atoms with C 12 ; how much energy would be liberated in each case?
(For the m&SS of N 13 , see Table 11-1.)
8. The following values of mass spectroscopic doublets were obtained, with
016 = 16.000000.
(H 1)2 - H 2 = 0.0015483 amu
(H 2 )a - t C 12 == 0.042298 amu
C 12 (H 1)4 - 0 16 = 0.036390 amu
Calculate the atomic masses of H1, H 2 and C 12 •

9. Suppose that, in Problem 8, C 12 = 12.000000 were taken as the mase
standard. What would be the doublet values and the masses of H1, H 2 , and
Ol6?
10. The following doublet values were obtained with 0 16 = 16.000000.
2(0 16) - S 32 = 0.017762

4c12 - sa2016 = 0.033027
What values are obtained for the masses of C 12 and S32 ?

11. Starting with Eq. (9-9), show that the average binding energy per nucleon
for all but the lightest elements is close to 8 Mev.
[Hint: Note that Eq. (9-9) may be written
flM MzA
A = m,. - A - AZ (m,. - mH).]
12. The best value of the faraday is considered to be 96521.9 coul/gm-

molecular weight, on the physícal scale. What is the corresponding value on
the chemical scale?
CHAPTER 10
NATURAL RADIOACTMTY AND THE LAWS OF

RADIOACTIVE TRANSFORMATION
Many of the ideas and techniques of atomic and nuclear physics are
based on the properties of the radioactive elements and their radiations,
and the study and use of radioactivity are essential to nuclear physics.
It has been seen that the emission of a- and ¡'j-particles by certain atoms
gave rise to the idea that atoms are built up of smaller units, and to the
concept of atomic structure. The investigation of the scattering of a-par-
ticles by atoms led to the idea of the nuclear atom, which is fundamental
to all of atomic theory. The analysis of the chemical relationships between
the various radioactive elements resulted in the discovery of isotopes.
The bombardment of atoros with swift a-particles from radioactive sub-
stances was found to cause the disintegration of atomic nuclei, and this
led in turn to the discovery of the neutron and to the current theory of the
composition of the nucleus. It will be shown in a later chapter that the
transmuted atoros resulting from this kind of bombardment are often
radioactive. This discovery of artificial, or induced, radioactivity by
Joliot and Curie, in 1934, started a new line of research, and hundreds of
radioactive nuclides have now been made by various methods. The
investigation of the radiations from the natural and artificial radionuclides
has shown that the nucleus has energy levels analogous to the atomic energy
level~ discussed in Chapter 7. Nuclear spectroscopy, which deals with the
identification and classification of these levels, is an important source of
information about the structure of the nucleus. Thus, radioactivity has
been intimately connected with the development of nuclear physics, and
it is impossible to conceive of nuclear physics as something separate from
radioactivity.
The importance of radioactivity depends to a large extent on the ability
to measure radioactive changes with high precision, and to describe them
quantitatively by means of a straightforward theory. The laws of radio-
active change were developed from information about the natural radio-
elements, but they are also valid for the artificial radionuclides. They
can be applied, therefore, to any radioactive transformation, and are
fundamental to a large part of the work to be discussed in the remaining
chapters of this book.
10-1 The basis of the theory of radioactive disintegration. The first

problem which will be considered is that of the quantitative description
of radioactive growth and decay. A clue to the way in which one radio-
229
230 NATURAL RADIOACTIVITY [CHAP. 10
100
\ J.>-;-
T,!} V
' \ V
/
~,
~
~
1\
r '\...
""'
'-
,. --.......:!'hli
¡--¡-.
o 12 16 20 24 28
4 8
Time (days)
Fm. 10-1. The decay of thorium X activity and the recovery of thorium
activity.
100
_J..,,.
1g 80
\ !!- ~
,/
'o 60
'\ /
~ •'I'-. I
~ 40
/ .....
1 20
/
"-
'- -UX
~ 1....... -
o 20 40 60 80 100 120 140
Time (days)
Fm. 10-2. The decay of uranium X activity and the recovery of uranium
activity.
active substance is produced from another was provided by early experi-

menta of Crookes, Becquerel, and Rutherford and Soddy. m Crookes
(1900) found that if a uranium salt was precipitated from solution by the
addition of ammonium carbonate, and then redissolved in excess of the
reagent, a small residue was left. This residue, wben removed from the
solution, was found to be highly radioactive. The product obtained
by evaporating the solution, which contained practically ali the uranium,
had very little activity. It appeared, therefore, that most of the observed
activity of compounds of uranium was not caused by that element, but
by another substance which could be separated from the uranium. The
active substance was contained in the residue and was given the name
10-IJ BASIS OF THE THEORY OF RADIOACTIVE DISINTEGRATION 231
uranium X (UX) to distinguish it from uranium. Becquerel then found

that if the uranium X and uranium fractions were allowed to stand sep-
arately for sorne time, the activity of the UX decreased, while that of the
uranium fraction increased. Rutherford and Soddy (1902) <2 > obtained
similar results with thorium salts; an active material which was called
thorium X was separated, and the main body of the thorium was pi:ac-
tically inactive. After a few days, it was noticed that the thorium X
was losing its activity, while the thorium, which had been freed from
thorium X, was recovering its activity.
Rutherford and Soddy studied quantitatively the rate of decay of the
ThX activity and the rate of recovery of the thorium activity, and obtained
the curves shown in Fig. 10-1. The experimental decay curve for the
ThX was exponential in nature, i.e., the activity could be expressed as a
function of time by the equation
(10-1)
where A., 0 is the initial activity oí the ThX, A.,(t) is the activity after a
time t, and >. is a constant, called the dísintegratíon constant. The recovery
curve for the thorium was found to fit the formula
in which the constant >. has the same value as in Eq. (10--1); the decay
and recovery curves are therefore symmetrical. The results for UX and
U are shown in Fig. 10--2. The curves are similar to those for the thorium
bodies except that the time scale is different. Thorium X loses half of
its activity in about 3.5 days, while UX loses halí oí its activity in 24 days,
so that the value of >. is greater for ThX than for UX.
These experimental observations enabled Rutherford and Soddy to
formulate a theory of radioactive change. They suggested that the atoms
of radioactive elements undergo spontaneous disintegration with the
emission oí a- or ~-particles and the formation of atoms of a new element.
Then the intensity of the radioactivity, which has been called the activity,
is proportional to the number of atoms which disintegrate per unit time.
The activity, A, measured by one of the methods discussed in Chapter 2,
may then be replaced by the number of atoros N, and Eq. (10-1) may be
written
(10--3)
The notation may be simplified by dropping the subscript x, which was

used only to distinguish between the X body and its parent substance.
Then
(10-4)
is the equation which represents the change with time of the number of
atoms of a single decaying radioactive substance. Differentiation of both
sides of Eq. (lo-4) gives
dN
---).N
dt - ' (10-5)
where N(t) has been abbreviated as N. According to Eq. (10-5), the

decrease per unit time in the number of atoms of a radioactive element
because of disintegration is proportional to the number of atoms which have
not yet disintegrated. The proportionality factor is the disintegration
constant, which is characteristic of a particular radioactive species.
Equation (10-5) is the fundamental equation of radioactive decay.
With this equation, and with two assumptions, it was possible to account
for the growth of activity in the thorium or uranium fractions from which
the ThX or UX had been removed. The assumptions are (1) that there
is a constant production of a new radioactive substance (say UX) by the
radioactive element (uranium), and (2) that the new substance (UX)
itself disintegrates according to tbe law of Eq. (10-5). Suppose that Q
atoms of UX are produced per second by a given mass of uranium, and
let N be tbe number of atoms of UX present at time t alter the complete
removal of the initial amount of UX. Then the net rate of increase of
UX atoms in the uranium fraction is
dN
dt=Q- ).N. (HH>)
The first term on the right side of Eq. (lo--6) gives the rate of formation
of UX atoms from U atoras; tbe second term gives the rate of disappearance
of UX atoms beca.use of their radioactive disintegration. To integra.te
Eq. {lo--6), write it in the form
dN
dt + ).N = Q,
and multiply through by eA1• Then
eA1 dN
dt + .ruYe
,_ u =
A1 QeA1
, or
The last equation can be integrated directly to give
or
C is an integration constant determined by the condition that N = O

10-1] BASIS OF THE THEORY OF RADIOACTIVE DISINTEGRATION 233
when t = O. This condition gives C = -Q/>.., and
N = q (1 - e->- 1) = N 0 (1 - e->- 1), (10-7)

>.
with N 0 = Q/>...
Equation (10-7) is the same as the recovery equation (10-2) so that
the theory gives the correct result for the growth of activity in the uranium
or thorium after the removal of the X body. Equation (10-7) also shows
that the number of UX atoms in the mass of uranium approaches an
equilibrium value for large values of t given by the ratio
q= Number of atoms of UX produced from U per second
X Fractions of atoras of UX which decay per second
The exponential law of decay was deduced by E. von Schweidler (1905)
without any special hypothesis about the structure of the radioactive atoms
or about the mechanism of disintegration. He assumed only that the
disintegration of an atom of a radioactive element is subject to the laws
of chance, and that the probability p for an atom to disintegrate in a time
interval llt is independent of the past history of the atom and is the same
for all atoms of the same type. The probability of disintegration then
depends only on the length of the time interval and, for sufficiently short
intervals, is proportional to llt. Then p = >..llt, where >. is the disintegra-
tion constant characteristic of the particular radioactive substance. The
probability that the given atom will not disintegrate during the short
interval lit is 1 - p = 1 - Mt. If the atom has survived this interval,
then the probability that it will not disintegrate in a second time interval llt
is again 1 - Mt. The probability that the given atom will survive both
the first and the second intervals is (1 - >..llt) 2 ¡ for n such intervals, the
probability of survival is (1 - >.lit)". If the total time nllt is set equal
to t, the probability of survival is (1 - >.(t/n)]". The probability that
the atom will remain unchanged after time t is the limit of this quantity
as llt becomes vanishingly small, oras n becomes very large. Now, one of
the definitions of the exponential functions is
e-s = lim (1 - ~)" ,

n-+oo n
from which it follows that
l. (1 , nt)" =e
lffi
ft-+OIO
-"- ->.t .
The statistical interpretation of this result is that if there are initially a

large number N O of radioactive atoms, then the fraction remaining un-
changed after a time t is N /N O = e-u, where N is the number of un-
changed atoms at time t.
234 NATURAL RADIOACTIVITY (CHAP. 10
The law of radioactive decay is thus a statistical law and is the result
of a very large number of events subject to the laws of probability. The
number of atoms which disintegrate in one second is, on the average, >.N,
but the number which break up in any second shows ftuctuations around
this value. The magnitude of these ftuctuations can be calculated with
the aid of the theory of probability, and the statistical considerations
involved are important in the design and interpretation of experiments
having to do with the measurement of radioactivity.< 3>
The number of radioactive atoms N and the activity A have been used
interchangeably so far on the grounds that the latter is proportional to
the former. For a given radioactive substance, the two quantities are
actually connected by the relationship
A= c'JI.N. (10-8a)
The proportionality factor e, which is sometimes called the detection coeffi-

cient, depends on the nature and efficiency of the detection instrument
and may vary considerably from one radioactive substance to another.
For any one substance, the quantity in which we are usually interested
is the ratio of the number of atoms at two values of the time. But, from
Eq. {10-8a), it follows that
(I0-8b)
and the detection coefficient cancels out. Hence, the use of N and A as
equivalent quantities usually leads to no confusion in the case of a single
substance. When two different substances are considered, the measured
activities are A 1 = c 1'11. 1N 1 and A 2 = c2 '11.2N 2 , respectively. If the number
of atoms which disintegrate per unit time is the same for both substances,
"Jl. 1N 1 = X~ 2 , but the measured activities, say in counts per minute, are
not necessarily equal. They are equal only if the detection coefficients
are equal. It will be assumed, with occasional exceptions, that the detec-
tion coefficients are ali equal to unity, i.e., that each disintegration is
detected. This condition usually cannot be achieved in practice, but it
is adopted here to simplify the discussion. The activity will then be
equal to the number of atoms disintegrating per unit time, A = >i.N, unless
otherwise noted.
1~2 The disintegration constant, the balf-life, and the mean life. A
radioactive nuclide may be characterized by the rate at which it disinte-
grates, and any one of three quantities, the disintegration constant, the
half-life, or the mean life, may be used for this purpose. The disintegra-
10--2] DISINTEGRATION CONSTANT 1 HALF-LIFE 1 AND MEAN LIFE 236
104
.'\.
'-
'l\.
...,"'-
"
" "'
" '\.
I'..
"" ~
102
'
60 120 180 240
Time (min)
Fro. 10--3. Gra.phical method for determining the value of the disintegra.tion
constant.
tion constant >. can be determined experimentally, in many cases, with

the help of Eq. {lQ-4), which may be written
In N(t) = - >.t (lo-9)

No '
where the symbol "In" represents the natural logarithm. The latter can
be transformed to the ordinary logarithm, denoted by "log," and Eq. {10-9)
becomes
N(t)
log N = -0.4343>.t, (10-10)
O
since the logarithm to the base e is equal to 2.3026 times the logarithm to
the base 10. The number of atoros N(t) is proportional to the measured
activity A(t), so that N(t)/N 0 = A(t)/A 0 , and Eq. (10-10) may be
written
log A(t) = log A 0 - 0.4343>.t. (10-11)
Hence, if the logarithm of the measured activity is plotted against the
time, a straight line should result whose slope is equal to -0.4343>.. An
example of this method of determining >. is shown in Fig. 10-3¡ for con-
venience, the plot is made on semilog paper. In the example shown, the
slope is -0.00808, with the time expressed in minutes¡ >. is then 0.0186
min-1, or 3.10 X 10-4 sec- 1•
1.00 1.0
1\.
0.9
0.8 " ~
0.90 0.7
0.6
"\. w
o:,
0.80
0.5 -- \ 1\
0.4
(1/e)
0.3
---·...!·-\
1 .
0.70 1 1~
-----+-..!..- ~
0.2
! ! 1
l~ 0.60 l 1
1 1
! 1
1
\
~
~ 1 .
i----+--1--+---
1
·;:
ºi
.,"
.::
.s.,
0.50
~
·;: 0.1
~
.,.
0.09
0.08
0.07
...
! 1 1
'
!
1
'
1
.
Í\
1
""
"
1t<
1
0.40 . .:.. .L
il: (1/e)
0.06
1 i 1
_J.. -
1\. ~
0.05 t:::I
1 1 1 1\ ~
;
1
,_ ___ J __¡_¡_ _¡_1

0.04 1
0.03
1
_J.._
1
.l \
1 1
0.02
1
1
1 j
1
i
i
1
1
1
1
!
1
1
1
1
!\
1 \
O.IO 1----+-r-¡---+---+---+
1 , 1 1 1
--- 1
1 1 1 I 1 1 1
0.01 ! i ! ! ! ! !
T , 2T 3T 4T 57' 67' T , 2T 3T 4T 5T 6T
Elapsed time (in units oí halí-life, T) -;:;-
=
~
Fm. lo--4. Re.dioactive dece.y: the half-life. The curve on the left is plotted on linear graph
paper, while the curve on the right is plotted on semilog paper.
-
o
10-2) DISINTEGRATION CONSTA.NT, HA.LF·LIFE, AND MEAN LIFE 237
Another quantity which is used to characterize a radionuclide is the

half-life, T, the time needed for half of the radioactive atoros to disinte-
grate. After one half-life, N(T)/N O = 0.5, and it follows from Eq. (10-11)
that log 0.5 = -0.4343>.T or log 2 = 0.4343>.T. Since log 2 = 0.3010,
half-life = T = 0.693/>.. (10-12)

The relationship between the activity and the half-life is illustrated in
Fig. 10-4. After n half-lives (t = nT), the fraction of the activity remain-
ing is (½)n. This fraction never reaches zero, but it becomes very small;
after seven half-lives the activity is 1/128, or leas than one percent of the
initial activity. After ten half-lives, the activity has fallen to 1/1024 or
about 0.1 % of the original amount, and is usually negligible in comparison
with the initial value.
lt is also possible to determine the mean lije, or average life expectancy,
of the atoros of a radioactive species. The mean life, usually denoted by -r,
is given by the sum of the times of existence of all the atoms, divided by
the initial number. Mathematically, it is found in the following way.
The number of atoros which decay between t and t dt is +
dN = >.Ndt;
but the number of atoros still existing at time t is
so that
Since the decay process is a statistical one, any single atom may have a
life from O to oo. Hence, the mean life is given by
-r = _!_ ["' N 0>.te-At dt = >. ["' te-At dt =!, (10-13)

NoÍo Ío >.
and is simply the reciprocal of the disintegration constant. From Eqs.
(10-12) and (10-13), it follows that the half-life and the mean life are
proportional quantities:
T = 0.693-r. (10-14)
If the half-life of a single radioactive species has a value in the range from
several seconds to several years, it can be detennined experimentally by
measuring the activity as a function of the time, as in the case of the dis-
integration constant. When the activity is plotted against the time on
semilog paper, a straight line is obtained, and the half-life can be read off
directly.
It often happens that two or more radioactive species are mixed together,
in which case the observed activity is the sum of the separate activities.
If the activities are independent, i.e., one component of the mixture does
not give rise to another, the various activities can sometimes be distin-
guished, and the separate half-lives determined. When the total activity
is plotted against the time on semilog paper, a curve like the solid one in
Fig. 10-5 is obtained. The curve is concave upward because the shorter-
lived components decay relatively rapidly, eventually leaving the long-
lived components. After a sufficiently long time, only the longest-Iived
activity will remain, and the value of its half-life can be read from the
late portion of the decay curve, which will be a straight Iine. If this
straight-line portion is extrapolated back to t = O, and if the values of
the activity given by the line are subtracted from the total activity, the
curve that remains will represent the decay of ali the components of the
mixture except the longest-lived. The example in Fig. 10-5 is a mixture
of two activities, one with a half-life of 0.8 hour, the other with a half-
life of 8 hours. The curve for the total activity is the sum of the two
straight lines which represent the individual activities. Although, in
principie, any complex decay curve can be analyzed into its components,
practica! difficulties may limit the usefulness of the method to three com-
300
\
200
\\
\ ~
--
100
,....._ ~ f
-- -
-
.;- 50
1l ' e
:, 30
_g \ r-- r--_
<: 20
~
~
' \
:¡ 10
<
5
3 '·
2
\
l
o 2 4 6 8 10 12 14
Time (hours)
Fm. 10-5. Analysis of a composite decay curve. (a) Composite curve.

(b) Longer-lived component (T½ = 8.0 hr). (c) Shorter-lived component
(T½ = 0.8 hr). (Reprinted by permission from G. Friedlander and J. W.
Kennedy, lntroduction to Radiochemutry. New York: Wiley, 1949.)
10-3] SUCCESSIVE RADIOACTIVE TRANSFORMATIONS 239
ponents, and even a two--component mixture may be hard to resolve if the

half-lives differ by less than a factor of about two.
If the half-life of a radionuclide is either very long or very short, methods
different from those discussed so far must be used. When the half-life
is very long, i.e., X is very small, it may not be possible to detect a change
in activity during the course of the measurement. The experimental
activity, A, is equal to c>.N, where e gives the fraction of the disintegrating
atoros detected by the measuring device. Then X can be found from the
relation
dN A
XN=--=-,
dt e
provided that N, the number of atoros of the nuclide in the sample, is
known, and that e is known as a result of an appropriate calibration. This
method has been used successfully for long-lived a-emitters, and values
of half-lives up to the order of 10 1 º years have been determined. For very
short half-lives, other methods <4 > must be used, the discussion of which
is beyond the scope of this book; half-lives as short as 10-9 sec have been
determined.
10-3 Successive radioactive transformations. It was found experi-

mentally that the naturally occurring radioactive nuclides forro three
series. In each series, the parent nuclide decays into a daughter nuclide,
which decays in turn, and so on, until finally a stable end product is reached.
In the study of radioactive series, it is important to know the number of
atoros of each member of the series as a function of time. The answer to
a problem of this kind can be obtained by solving a system of differential
equations. The procedure may be illustrated by treating the case of a
radioactive nuclide, denoted by the subscript 1, which decays into another
radioactive nuclide (subscript 2); the latter, in turn, decays into a stable
end product (subscript 3). The numbers of atoros of the three kinds at
any time tare denoted by N 1 , N 2 , and N 3 , respectively, and the disintegra-
tion constants are >. 1 , >. 2 , and >. 3 • The system is described by the three
equations
dN 1 _
dt - -X1N1, (I0-15a)
dN2 _
dt - X1N1 - X2N2, (10-15b)
dN 3 _
dt - X2N2. (I0-15c)
These equations express the following facts: the parent nuclide decays ac-
cording to the basic law Eq .. (10-5); atoros of the sec·ond kind are formed
at the rate X1N 1 because of the decay of parent atoms, and disappear at
the rate X2 N 2 ; atoms of the stable end product appear at the rate X2 N 2
as a result of the decay of atoms of the second kind.
lt is instructive to solve this system of equations in detail because the
procedure is one which is often used. The number of atoms N 1 can be
written down immediately,
(10-16)
where N/. is the number of atoros of the first kind present at the time t = O.
This expression for N 1 is inserted into Eq. (10-15b), 11,nd gives
dN2 _ },. Nº -A¡t _ },. N

dt - 1 1e 2 2,
or
d: + 2
X2N2 = X1N~e-A
11
• (10-17)
Multiply Eq. (10-17) through by e>·21 ; this gives
or
The last equation can be integrated directly to give
where C is a constant of integration. Multiplying through by e-A,t gives
(10-18)
The value of the integration constant is determined by noting that when

t = O, the number of atoms of the second kind has sorne constant value,
or N 2 = N~, with N~ equal to a constant. Then
o X1 o
C = N2 - X
2 -
X Ni,
1
Inserting this value into Eq. (10-18) and rearranging, we obtain the
solution for N 2 as a function of time:
N2 = X X1
_ Xi Nº(
1 e
-A¡t
- e-A,1) + Nº2e-A,1 • (10-19)
2
10-3] SUCCESSIVE RADIOACTIVE TRANSFORMATIONS 241
The number of atoms of the third kind is found by inserting this expres-
sion for N 2 into Eq. (10-15c) and integrating, which gives
N3 = ( , >-1 , Nº1 _ Nº)

2 e-A2I _ , >-2 , Nº1e-A¡t + D, (10-20)
"2 - "l "2 - "l
where D is an integration constant, determined by the condition N 3 = N~

at t = O. This condition gives
D = Ng + Ng + N~.
When this expression for D is inserted into Eq. (10-20), the result is
Na= N8 + Ng(l - e-x,t) + N~ (1 + >-1 e-x,1 - X2 e-A11).

>-2 - >-1 >-2 -- >-1
(10-21)
Equations (10-16), (10-19), and (10-21) represent the solution of the
problem.
One of the cases met most often in practice is that in which only radio-
active atoms of the first kind are present initially. In this case, the con-
stants N~ and N~ are both equal to zero, and the solutions for N 2 and
N 3 reduce to
N2 = , >-1 , Nº(
1 e
-A¡t -e -A2!) , (10-22)
"2 - "l
-
N a- Nº1 (1 + >-2 >-1- >-1 e -A2! - >-2
>-2 - >-1 e
-A¡I) · (10-23)
The curves of Fig. lo-6 show what happens in this case if it is assumed
that the half-lives of the active species are T 1 = 1 hour and T.2 = 5 hours,
respectively; the corresponding values of the disintegration constants are
X1 = 0.693 hr- 1 , and X2 = 0.1386 hr-1, respectively. The ordinates of
the curves represent the relative numbers of the substances 1, 2, and 3 as
functions of the time when the initial number of atoms of the substance 1 is
taken as N~ = 100. The number of atoms, N 1, of substance 1, decreases
exponentially according to Eq. (10-16) with a half-life of 1 hour; N 2 is
initially zero, increases, and passes through a maximum after about three
hours, and then decreases gradually. The number of atoms N 3 of the
stable end product increases steadily with time, although slowly at first;
when t becomes very large, N 3 approaches 100, since eventually ali the
atoms of the substance 1 will M converted to atoms of the stable end
product.
The treatment just discussed can be extended to a chain of any number
of radioactive products, and the solution of this problem is often useful.
100
i..-- ...-
-i----
J.,-/' ~
.,-Ni
y V
V"'\ i\(N2
/
/ '\ /
-8 50
'"' ""~
§
z J
I\ V
~
/ ""'- ['-...._ ...__
10
V
o 2
"
4
'-- 6 10 12 14 16 18
--
20
r---.
22 24 26
Time (hours)
Fm. lG-6. A radioactive series with three members: only the parent (T½ =
1 hr) is present initially; the daughter has a half-life of 5 hr, and the third member
is atable.
The procedure is similar to that of the special case already considered

except that the mathematics becomes more tedious and the expressions for
the numbers of atoms become more complicated as the length of the chain
increases. The differential equations of the system are
dN1
ae= ->..1N1,
dN2
ae= >..1N1 - >..2N2,
dNa (10-24)
X2N2 - XaNa,
dt
dN,.
>,.,._ 1N,._ 1 - >..,.N,..
dt
The solution of this system of equations under the assumption that at
10-3] SUCCESSIVE RADIOACTIVE TRANSFORMATIO:--S 243
LO No--..---.....--...---r----,----.---,
~ 0.6 ,'lio11-+--µ,,...-+---+-...,,._-+---+---+---t
5::,
: 0.4 Nott--t--~---i---i-----+---+---1
>
-~
~ 0.2 No H--'1-+-+--+"'oe""'""""---+----+----i
O m ~ 00 AA 100 lW 1~
Time (min)
F10. 10-7. The deeay of Radium A. [Reprinted by permission from Ruther-

ford, Chadwiek, and Ellis, Radiationa from Radiooctive Substances, Cambridge
University Press (Maemillan Co.), 1930.]
t = O only the parent substance is present was derived by Ba.teman. <5>

The initial conditions are
t = O; (10-25)
The number of atoros of the nth member of the chain is given by

N,.{t) (10-26)
with
(10-27)
An example of the application of the Bateman equations, Eqs. (10-24) to

(10-27), is shown in Fig. 10-7, taken from Rutherford, Chadwick, a.nd
Ellis' book Ra.diations from Radioactive Substances. The curves were
obtained under the conditions which follow. A test body was exposed
for a few seconds to radon (Em 222 ), and a certain number of atoms of
the decay product of radon, RaA or Po 218, with a half-life of 3.05 min
were deposited on the test body. The RaA decays into RaB (Pb 214 ) with
a half-life of 26.8 min, whích decays, in turn, into RaC (Bi 214 ) with a
half-life of 19.7 min. Finally, the end product RaD or Pb 210 , with a
half-life of 22 years, is formed. The last half-life is sufficiently long so
that the number of radium D atoms which disintegrate may safely be

neglected. The number of RaA atoms decreases exponentially. The
nurober of RaB atoros is initially zero, passes through a maxirouro about
10 min later, and then decreases with time. The number of RaC atoms
passes through a maximum after about 35 min. The number of RaD
atoros increases, reaching a maxiroum when the RaA and RaB have dis-
appeared. Eventually the RaD would decay exponentially with a half-life
of 22 years. The suro of ali the atoms present at any time is N 0 , the initial
number of atoros of RaA.
Solutions can be obtained for other problems with different initial con-
ditions, and the reader is referred to the book by Rutherford, Chadwick,
and Ellis for additional examples of the application of the theory.
10-4 Radioactive equilibrium. The term equilibrium is usually used

to express the condition that the derivative of a function with respect to
the time is equal to zero. When this condition is applied to the merobers
of a radioactive cha.in described by Eqs. (10-24), it means that the deriva-
tives dN i/dt, dN 2 / dt, ... , dN,,/dt are all equal to zero, or that the number
of atoms of any member of the chain is not changing. The conElitions for
equilibrium are then
dN1
dt= -'A.1N1 = O,
>-1N1 = >-2N2,
>-2N2 = 'A.aNa, (10-28)
>-n-tNn-1 = >..,,N,,,
These conditions cannot be satisfied rigorously if the parent substance is
a radioactive substance beca.use the first of Eqs. (10-28) implies that
>.. 1 = O, which is a contradiction. lt is possible to achieve a state very
close to equilibrium, however, if the parent substance decays much more
slowly than any of the other members of the cha.in; in other words, if the
parent has a half-life very long compared with that of any of its decay
products. This condition is satisfied by the naturally occurring radio-
active chains. Uraniuro I has a half-life of 4.5 X 10 9 years and the fraction
of UI atoms transformed during the life of an experimenter is indeed
negligíble. In such a case, the number of atoms N 1 can be taken to be
constant, and the value of >.. 1 is very much smaller than that of any of the
other X's in the chain. The first of Eqs. (10-28) is then a very good ap-
proxiroation, and the rest of the conditions are rigorously valid. This
10-4) RADIOACTIVE EQUILIBRIUM 245
type of equilibrium is called secul,ar equilibrium, and satisfies the condition

(10-29)
or, in terms of half-lives,

N1 N2 Na Nn-1 Nn (10-30)
Ti - T2 - Ta - - Tn-1 - Tn
The relationships (10-29) and (10-30) may be applied whenever severa!

short-lived products arise from successive decays beginning with a rela-
tively long-lived parent. It is only necessary to be sure that the material
has been undisturbed, that is, that no decay products have been removed or
allowed to escape for a long enough time for secular equilibrium to be
established. Secular equilibrium can also be attained when a radioactive
substance is produced at a steady rate by sorne artificial method, such
as a nuclear reaction in a cyclotron or chain-reacting pile. The term
>.. 1N 1 in the second of Eqs. (10-24) is then constant, as in the case of a
very long-lived parent, and the condition {10-29} is satisfied.
The relationships {10-30} can be used to find the half-life of a radio-
nuclide whose half-life is very long. For example, uranium minerals in
which secular equilibrium has been established have been shown to contain
one atoro of radium for every 2.8 X 10 6 atoms of uranium I. If uranium I
is denoted by the subscript 1 and radium by the subscript 2, then at equi-
librium N i/N 2 = 2.8 X 10 6 • The half-life of radium is known from direct
measurements to be 1620 years. Consequently, the half-life of uranium I
is
T1 =N:N 6 9
T2 = 2.8 X 10 X 1620 = 4.5 X 10 years.
We consider next an example of the approach to secular equilibrium.

The case is that of a long-lived parent (T "" oo) anda short-lived daughter.
It is assumed that the daughter has been separated from the parent, so
that the latter is initially pure. The mathematical expressions for the
number of atoms of parent and daughter may be obtained from Eqs.
(10-16) and (10-22) if it is noted that >.. 1 "" O, and >.. 1 « >. 2. Then
e-xit "" 1, and
N1 ""N?, (10-31)
N2 >..1
"" >..
Nº(l
1 - e-A2I) . (10-32)
2
Equation (10-32} may be rewritten

21
X2N2 "" >.. 1N?(l - e-A ). (10-33)
30( )
21M )
:,<
100
/ - l!
\ I
50 j
I \
2()
/{I \
1e ~f
H)
\
::,
7.
5
'
4
3
'
2 \
'~.
1 \ 1
o 2 -l 6 7 8 IO 12
Time (hours)
Frn. 10-8. Secular equilibrium. (a) Daughter activity growing in freshly

purified parent fraction. (b) Activity of parent (T½ = oc). (e) Total activity
of an initially pure parent fraction. (d) Decay of freshly isolated daughter
fraction (T½ = O.SO hr). (Reprinted by permission from G. Friedlander and
J. W. Kennedy, Introduction to Radiochemistry. New York: Wiley, 1949.)
The last equation gives the activity of the daughter as a function of the
time, in terms of the (constant) activity of the parent. The total activity
is given by
(10-34)
Equation (10-33) shows that as t increases the activity of the daughter

increases, and after severa} half-lives, X2 N 2 approaches X1N~, satisfying
the condition for secular equilibrium. These relationships are shown graph-
ically in curves a and b of Fig. 10-8; curve e gives the total activity. For
comparison, curve d shows how a freshly isolated daughter fraction would
decay.
A somewhat different state of affairs, called transient equilibrium, results
if the parent is longer-lived than the daughter (X 1 < >. 2 ), but the half-
life of the parent is not very long. In this case, the approximation >. 1 = O
may not be made. If the parent and daughter are separated so that the
parent can be assumed to be initially pure, the numbers of atoms are
again given by Eqs. (10-16) and (10-22). After t becomes sufficiently
large, e-A 21 becomes negligible compared with e-A,,, and the number of
10-4) RADIOACTIVE EQUILIBRIUM 247
3()0
200
- - .f...
~
100 ,:"
~>-e
,,, ~
.¡- 50 b -....,. ....

1i., / 1\ i - i;::::, .....
to) I \ r- e:::::
.s 20
,d
! a
>, \
:] 10
.
~
5 \
.
3
\
2
1
~.
\
o 2 4 6 8 10 12 14
Time (hours)
FIG. 10-9. Transient equilibrium. (a) Daughter activity growing in freshly

purified parent fraction. (b) Activity of parent (T½ = 8.0 hr). (e) Total
activity of an initially pure parent fraction. (d) Decay of freshly isolated
daughter fraction (Ti = 0.80 hr). (e) Total daughter activity in parent-plus-
daughter fractions. (Reprinted by permission from G. Friedlander and J. W.
Kennedy, lntroduction to Radiochemistry. New York: Wiley, 1949.)
atoms of the daughter becomes
(10-35)
Thus, the daughter eventually decays with the same half-life as the parent.
Since N~e->- 11 = N 1, it follows from Eq. (10-35) that
(10-36)
The ratio of the measured activities at equilibrium is

A1 >.1N1 >.2 - >.1 (10-37)
A2 = >.2N2 = >.2
and the daughter activity is greater than that of the parent by the factor
>.2/(>.2 - >.1), The above results, which are characteristic of transient
equilibrium between parent and daughter atoms, are shown graphically
in Fig. 10-9.
When the parent has a shorter half-life than the daughter {A 1 > A2 ),
no state of equilibrium is attained. If the parent and daughter are sep-
arated initally, then as the parent decays, the number of daughter atoms
will increase, pass through a maximum, and eventually decay with the
half-life of the daughter.
10-5 The natural radioactive series. A1l a result of physical and chemi-
cal research on the naturally occurring radioactive element.s, it was proved
that each radioactive nuclide is a member of one of three long chains, or
radioactive series, stretching through the last part of the periodic system.
These series are named the uranium, actinium, and thorium series, respec-
tively, a.fter element.s at, ornear, the head of the series. In the uranium
series, the mass number of each member can be expressed in the forro
(4n + 2), where n is an integer, and the uranium series is sometimes
called the "4n + 2" series. In the actinium and thorium series, the mass
numbers are given by the expressions 4n +
3 and 4n, respectively. There
TABLE 10-1
THE URANIUM SEBIF.8
Diainte-
Type of Particle
Radioactive gration
Nuclide diainte- Half-life energy,
epecies
Ura.nium I (UI) .,un•

gration
a 4.50 X 10• y
--·
collltant,
4.88 X 10-11 4.20

Mev
Ura.nium X, (UX,) ,oTh2H fJ 24.1 d 3.33 X 10-7 0.19

Uranium X2 (UX2) uPatH fJ 1.18m 9.77 X 10-• 2.32
Ura.nium Z (UZ) uPatH fJ 6.7 h 2.88 X 10-• 1.13
Uranium 11 (UII) .,u2u a 2.50 X 105 y 8.80 X 10-1 • 4.768
Ionium (lo) ,oThUº a 8.0 X 10' y 2.75 X 10-11 4.68 m
Radium (Ra) 88 Ra2H a 1620 y 1.36 X 10-11 4.777 m
Ra Emanation (Rn) aaEmtU a 3.82d 2.10 X 10-• 5.486
Radium A (RaA) uPotu a, fJ 3.05m 3.78 X 10-1 a:5.998
fJ:?
Radium B (RaB) uPb 2 " fJ 26.8m 4.31 X 10-• 0.7
Astatine-218 (Atª 11) aaAttu a l.&-2.0 s 0.4 6.63
Radium C (RaC) aaBi 1 " a, fJ 19.7m 5.86 X 10-• a:5.51 m
fJ:3.17
Radium C' (RaC') uP02u a 1.64 X 10-, s 4.23 X 101 7.683
Radium C" (RaC") a,T1210 fJ 1.32m 8.75 X 10-• 1.9
Radium D (RaD) uPb210 fJ 19.4y 1.13 X 10-• 0.017
Radium E (RaE) aaBi210 11 5.0 d 1.60 X 10-e 1.155
Radium F (RaF) uP02to a 138.3 d 5.80 X 10-• 5.300
Thallium-206 (Tl2oe) 81
n2os fJ 4.2m 2.75 X 10-a 1.51
Radium G (RaG) uPb 10• Stable
10-5) THE NATURAL RADIOACTIVE SERIES 249
is no natural radioactive series of nuclides whose mass numbers are

represented by 4n + l.
The members of the uranium series are listed in Table 10-1 together
with the mode of disintegration, the half-life, <5> the disintegration con-
stant, and the maximum energy of the emitted particles. <5 > The changes
in atomic number and mass number are shown in Fig. 10-10. Table 10-2
and Fig. 10-11 give the corresponding information about the actinium
series, while Table 10-3 and Fig. 10-12 are for the thorium series. The
first column of each table gives the old-fashioned, historical names of the
radionuclides, while the second column gives the modern symbol. Both
are included because both are found in the literature of physics, and it is
helpful to have a code for translation readily available. The older name
usually indicates the series to which a radioactive substance belongs and
gives sorne idea of the relative position in the series; the modero symbol
gives the atomic and mass numbers. The mode of decay, the value of
the half-life (or disintegration constant), and the energy of the emitted
particle characterize the radioactivity. The half-life may be in years,
days, hours, minutes, or seconds, abbreviated as y, d, h, m, s, respectively.
TABLE 10-2
THE ACTINIUM SERIES
Dfainte-
Typeof Particle
Radioactive gra.tion
Nuclide disinte- Ha.lf-life energy,
species constant,
gration ~c-1 Mev
Aetinouranium (AeU) uu2aa a 7.10 X 10 8 y 3.09 X 10-11 4.559m

Uranium Y (UY) eoTh23t fJ 25.6h 7.51 X 10-s 0.30
Protoactinium (Pa) uPa2a1 a 3.43 X 104 y 6.40 X 1O-1a 5.O46m
Aetinium (Ac) seAc221 a, fJ 21.6y 1.02 X 10-9 a:4.94
{J:O.O46
Radioactinium(RdAe) eoTh227 a 18.17 d 4.41 X 10-1 6.O3m
Actinium K (AcK) 87Fr22a a, fJ 22m 5.25 X 10-• {J: 1.2
a:5.34
Aetinium X (AcX) ..Ra223 a 11.68 d 6.87 X 10-1 5.864
Astatine-219 ssAt219 a, fJ O.9m 1.26 X 10-2 a:6.27
Ac Emanation (An) 88Em21e a 3.92 s 0.177 6.810m
Bismuth-215 saBi 215 a, fJ 8m 1.44 X 10-a ?
Actinium A (AcA) uP02u a, fJ 1.83 X 10- 3 s 3.79 X 10 2 a:7.37
Actinium B (AcB) 82Pb211 fJ 36.1 m 3.20 X 10-• 1.39
Astatine-215 ssAt215 a 10-• s 7X 10 3 8.00
Actinium C (AcC) saBi 211 a, fJ 2.15m 5.28 X 10- 3 a:6.617m
Actinium C' (AcC') s,Po211 a O.52s 1.33 7.442m
Actinium C" (AcC") 81Tl201 fJ 4.79m 2.41 X 10-a 1.44
Actinium D (AcD) s2Pb201 Stable
238 1 IAc,
234
U~UX2"1j. 011
Jz¿
230 lo... ......
V
/
.. 226
Ra"" .....
~ /',,
e Rn"~
El 222 V
/
! ,r-
:E 218 RaA..L.At
/ ..... ~
214
......
...
Ra~Ra~RaC'
210 RaC~Ran.ÍR&~RaF
/
RaG""~
206
TI
81
~ lli fu M • fi Ra ~ Th ~ u
82 83 84 85 86 87 88 89 90 91 92
Atomic number
Frn. 10-10. The uranium (4n + 2) series.
TABLE 10-3
THE THORIUM SERIES
Type of Disinte-
Radioactive Pa.rticle
gration
Nuclide disinte- Half-life energy,
species constant,
gration Mev
sec-•
Thoriwn (Th) e0Th2•2 a 1.39 X 1O 18 y 1.58 X 10- 19 4.007

Mesothoriuml(MsThl) 88
Ra22s /j 6.7 y 3.28 X 1O-e 0.04
Mesothorium2(MsTh2) uAc•2s /j 6.13h 3.14 X 10-• 2.18
Radiothoriwn (RdTh) eoTh 228 a 1.910 y 1.15 X 1O-s 5.423 m
Thoriwn X (ThX) saRa224 a 3.64d 2.20 X 10-e 5.681 m
Th Emanation (Tn) seEm220 a 51.5 s 1.34 X 10- 2 6.280
Thoriwn A (ThA) uPo2le a, /j O.16s 4.33 6.774
Thoriwn B (ThB) uPb212 /j 1O.6h 1.82 X 10-• 0.58
Astatine-216 (At 216 ) uAt2u a 3 X 10-• s 2.3 X 10 3 7.79
Thoriwn C (ThC) saB¡212 a, /j 60.5m 1.91 X 10-• a:6.086m
/j:2.25
Thoriwn C' (ThC') uP0212 a 3.0 X 10- 7 s 2.31 X 10 6 8.780
Thoriwn C" (ThC") uTl2° 8 /j 3.10m 3.73 X 1O-a 1.79
Thoriwn D (ThD) uPb2os Stable
10-5) THE NATURAL RADIOACTIVE SERIES 251
235 AcU
/
.....
UY~ ~Pa
231
.,.,.Ac- ~RdAc
,11
.. /
227
_/ /
/
AcK~ AcX_..,,,
_/
.
/
At,219Á
.
-,;;-- ► An_..
_/
,11
215 Bi21°? ,AcA- ~At215
~
... /
211 AcB- ~Act ►., AcC'

.
207 AcC" •Ac~.,.
/ /
TI Pb B1 Po A.t Em Fr Ra Ac Th Pa U
81 82 83 84 85 86 87 88 89 90 91 92
Atomic number
F10. 10-11. Tbe aetinium (4n + 3) series.
232 ,Th
/
,,,.-
228 ,sTht-'. • MsTh:z+ RaTh
/
.,.,....-
.,.,.,.
... ...
·hx
__....
To ...
216 ThA.,. ~At21&

.. /"'
_/ __....
,,, ...
¡.ThC'
212 ThB- ~ThC
_/ __....
208 The1/' ~Th~.,.

1
TI Pb Bi Po At Em J,'r Ra Ac Th Ra U
81 82 83 84 85 86 87 88 89 90 91 92
Atomic number
FIG. 10-12. The thorium (4n) series.

252 NATURAL RADIOACTIVITY [CHAP, 10
In the case of P-particles, the energy value listed is that of the most ener-
getic particles. For sorne radionuclides, the a-pa.rticles are monoenergetic;
for others, an a-particle may have one of severa! energy values, in which
case the greatest value is listed and denoted by "m." Figures 10--10
through 10--12 illustrate Soddy's displacement law discussed in Section
9-1. It will be remembered that the emission of an a-particle decreases
the charge of a nucleus by two units and the mass number by 4 units,
while the emission of a P-particle increases the charge of the nucleus by
one unit and leaves the masa number unchanged. In the figures, there-
fore, an a-disintegration is represented by an arrow sloping downward
and to the left, and a P-transition by a horizontal arrow pointing to the
right. All of the elements in any one column have the same atomic number
and must occupy the sa.me place in the periodic table. For example,
RaA, RaC', RaF, AcA, AcC', ThA, and ThC' all have the atomic number
84 and are isotopes of polonium. Similarly, RaB, RaD, RaG, AcB, AcD,
ThB, and ThD all have the atomic number 82 and are isotopes of lead.
Series schemes like those of Figs. 10--10 through 10--12 led Soddy (1913)
to the discovery of the existence of isotopes.
In most of the disintegration processes that make up a radioactive
series, each of the radionuclides breaks up in a definite way, giving one
a- or P-particle and one atom of the product nuclide. In sorne cases,
however, the atoms break up in two different ways, giving rise to two
products with different properties. This kind of disintegration is called
bTanching decay and is illustrated by the decay of Po 218 (RaA), Bi 214 (RaC),
and Bi 21 º (RaE) in the uranium series; by Ac 227 , Po 215 (AcA), Bi 211
(AcC), 87 Fr 223 (AcK), and 85 At 219 in the actinium series; and by
Po 216 (ThA) and Bi 212 (ThC) in the thorium series. Each of these nu-
clides can decay either by a-emission or by /Hlmission. The probability of
disintegration is the sum of the separate probabilities and X = X.+ X11;
the half-life is T = 0.693/X = 0.693/(X. + x,,). In most cases, one
mode of decay is much more probable than the other. Radium A
and AcA decay almost entirely by a-emission, with only a small fraction
of one percent of the atoms disintegrating by /Hlmission. 0n the otber
band, RaC, RaE, Ac, and AcC decay almost entirely by P-decay, with one
percent or less of the atoms decaying by a-emission. The case of ThC
(Bi 212) is especially interesting because 66.3% of the disintegrations are
by p-emission, and 33.7% by a-emission, in contrast to the other cases
mentioned. In each branched decay, the product atoms decay in turn to
give the same nuclide. For example, ThC emits a tJ-particle to form ThC'
(Po 212 ) or an a-particle to form ThC" (Tl 268 ); ThC' then emits an a-
particle to give stable Pb 268 (ThD), while ThC" emits a P-particle and
also forms stable Pb 208 •
UNITS OF RADIOACTIVITY 253
lo-6 Units of radioactivity. The intensity of radioactivity has so far

been considered in terms of the number of atoms which disintegrate per
unit time, or in terms of the number of emitted particles counted per unit
time by a detector. As in other branches of physics, it is useful to have a
standard quantitative unit with an appropriate name. In radioactivity,
the standard unit is the curie, at present defined as that quantity of any
radioactive material giving 3.70 X 10 10 disintegrations/sec. The milli-
curie, which is one-thousandth of a curie, and the microcurie, equal to
one-millionth of a curie, are also useful units; these would correspond to
amounts of active materials giving 3.7 X 107 and 3.7 X W 4 disintegra-
tions/sec, respectively.
For various reasons, mainly historical, there has been sorne confusion
about the use of the curie as the standard unit. Consequently, a new
absolute unit of radioactive disintegration rate has been recommended,
namely, the rutherford (rd), defined as that amount of a radioactive sub-
stance which gives 10 6 disintegrations/sec. The millirutherford (mrd)
and microrutherford (µrd) correspond to 10 3 disintegrations/sec and 1
disintegration/sec, respectively.
To get an idea of sorne orders of magnitude, consider the problem of
calculating the weight in grams of 1 curie and 1 rd of RaB (Pb 214 ), from
its half-life of 26.8 min. The disintegration constant is
If W is the unknown weight, then
-dn/dt = XN = 4.31 X 10-4 X W /214 X 6.02 X 1023
= 1.21W X 10 18 disintegrations/sec.
If -dN/dt = 3.70 X 10 1º disintegrations/sec (1 curie), then
W 3.70 X 1010 31 10-8

= 1.21 X 10 18 = . X gm.
For 1 rd,
6
10 8 -13
W = 1.21 X 1018 = .3 X 10 gm.
Thus, for a substance with a short, but not very short, half-life very little
material is needed to provide a curie of activity. When the half-life is a
small fraction of a second (Po 214, Po 215, At 215 , At 216, or Po 212 ), the
amount of material which gives a curie of activity is almost unimaginably
small. In the case of a nuclide with a very long half-life, such as U 238
(T = 4.50 X 109 y),>. = 4.9 X 10- 18 sec- 1 . Then
-d: = W= 4.9 X 10- 18 X ~ X 6.02 X 10 23
= 1.24 X lO'W disintegrations/see.

For 1 curie of activity,
3.70 X 10 10 11
W = 1.14 X lO' = 3.2 X 10 gm,
and more than three metric tons &re needed.
REFERENCES
GENERAL
RUTHERFORD, CHADWIClt, and ELL1s, Radiationa from ~ 8ubsttmcu.
New York: Macmillan, 1930. (Reprinted with corrections, 1951.)
G. HEVEBY and F. A. PANETH, A Manual of Radioactwity, 2nd ed. Oxford
University Press, 1938.
G. Fa1EDLANDER and J. W. KENNEDY, Nuclear and Radioc/&emWry. New
York: Wiley, 1955.
R. D. EvANS, The Atomic Nucleu,. New York: McGraw-Hill, 1955, Chaptera
15, 25, 26.
E. SEG.RE, "Radioactive Decay," Experimental Nuclear Phyaia, E. Segre, ed.,
Vol. III, Part IX. New York: Wiley, 1959.
J. L. PUTNAK, "Measurement of Disintegration Rate," Beta,- and Gamma-
Ray 8r,ectr08C0'/IY, K. Siegbahn, ed. Amsterdam: North Holland Publishing Co.;
New York: Interscience, 1955, Chapter 26.
G. J. BINE and G. L. BROWNELL, eds., Radiation Do&imetry. New York: Aca-
demic Presa, 1956.
PARTICULAR
l. G. E. M. JAUNCEY, "Early Years of Radioactivity," Am. J. Phy,. 14, 226
(1946).
2. E. RUTHERFORD and F. SoDDY, "The Cause and Nature of Radioactivity,"
J. Chem. Soc. 81, 321, 837 (1902); Phil. Mag. 4, 370, 569 (1902); ibid 5, 576
(1903).
3. G. FRIEDLANDER and J. W. KENNEDY, op. cit. gen. ref., Chapter 9.
4. S. RoWLANDB, "Methods of Measuring Very Long and Very Short Half-
Lives," Nvd,eonic, 3, No. 3, 2 (Sept. 1948).
5. H. BATEIIAN, "The Solution of a System of Differential Equations Occur-
ring in the Theory of Radio-active Transformations," Proc. C a ~ Phil. Soc.
16, 423 (1910).
6. BTROKINGER, HoLLANDEK, and SEABORG, "Table of Iaotopes," Rt:v,. Mod.
Phy,. 30, 585 (1958).
7. E. 81:ou, op. cit. gen. ref., pp. S-16.
255
PROBLEMS
l. Find the values of the disintegration constant and half-life of a radio-

active substance for which the following counting rates were obtained at different
times.
Counting Counting
Time, Time,
rate, rate,
hr hr
counts/min counts/min
o.o 6.0 1800
0.5 9535 7.0 1330
1.0 8190 8.0 980
1.5 7040 9.0 720
2.0 6050 10.0 530
3.0 4465 11.0 395
4.0 3295 12.0 290
5.0 2430
What would have been the counting rate at t = O?
2. The counting ratea listed below were obtained when the activity of a
certain radioactive sample was measured at different times. Plot the decay
curve on semilog paper and determine the half-lives and initial activities of the
component activities.
Counting Counting
Time, Time,
rate, rate,
hr hr
counts/min counts/min
o.o 19100 4.0 3500
0.5 14500 5.0 2520
1.0 11410 6.0 1835
1.5 9080 8.0 985
2.0 7345 10.0 530
3.0 4985 12.0 290
3. The half-life of radon is 3.82 days. What fraction of a freshly separated
sample of this nuclide will disintegrate in one day? In 2, 3, 4, 5, 10 days? If
the sa.mple contains, initially, one microgram of radon, how many atoms will
disintegrate during the first day? During the fifth day? During the tenth day?
4. The half-life of UX1 is 24.1 days. How long after a sample of UX1 has
been isolated will it take for 90% of it to change to UX2? 95%? 99%? 63.2%?
5. What are the weights of (a) 1 curie and 1 rd of Ra? (b) 1 curie and 1 rd
of Rn? (c) 1 microcurie and 1 rd of RaA? (d) 1 microcurie and 1 rd of RaC'?
6. A sample of ionium (Th 23 º) weighing 0.100 mg was found to undergo
4.32 X 106 disintegrations/min. What is the half-life of this nuclide? Ionium
is formed by the a-decay of u2u. How many rutherfords of U 234 would be
needed to produce a 0.100-mg sample of Th 230 ?
7. A freshly separated sample of RaF (Po 21 º) contains 1.00 X 10- 6 gm of
that nuclide. How many disintegrations per second would the sample un-
dergo immediately after separation? How many after 10, 30, 50, 70, 100, 300
days? How many curies of activity would the sample contain at these times?
How many rutherfords?
8. Consider a radioactive series whose first two members have half-lives of
5 hours and 12 hours, respectively, while the third member is atable. Assume
that there are initially 106 atoms of the first member, and none of the second
and third members. Plot, as functions of time, the numbers of atoms of the
three kinds. After how many hours will the number of atoms of the second
member reach its maximum value?
9. A freshly purified sample of RaE contains 2.00 X 10- 10 gm of that nuclide
at time t = O. If the sample is allowed to stand, what will be the greatest
weight of RaF that it will ever contain? At what time will that amount be
present? At that time, what will be the a-activity in disintegrations per second?
The tl-activity? How many rutherfords of RaF will be present? Plot the a-
and tl-activities as functions of time.
10. The nuclide Bi 212 (ThC) decays both by a-emission and fl-emission.
(a) What are the partial disintegration constants for a-decay and tl-decay?
(b) What are the half-lives for a-decay alone and for P-decay alone? (e) At
what rates will a freshly separated sample containing 10-7 gm of Bi 212 emit
a-particles? tl-particles? (d) What will be the rates after three hours?
11. Natural samarium has been found to emit a-particles at a rate of 135
particles/gm/sec. The isotope Sm 147 (abundance 15.0%) is responsible for the
activity; what is its half-life in years?
12. The nuclide U 235 has a half-life of 7.10 X 108 years; its daughter Th 231
has a half-life of 24.6 hours. Suppose that the Th 231 is separated chemically
from a sample of U 235 • Plot, as functions of time, the relative activities oí the
initially pure U235 fraction, the isolated Th231 fraction, the Th231 growing in the
freshly purified U 235 fraction, and of the U 235 itself. What kind oí equilibrium
is reached, ií any?
13. The nuclide Th 234 (UX1) is normally present in uranium compounds in
secular equilibrium with its parent U 238, which has a half-liíe oí 4.50 X 109 years.
If the formation of UZ is neglected, the UX1 decays with a half-life of 24.1 days
to Pa 234 (UX2), which decays with a half-life of 1.18 min into U 234 (UII); the
latter has a half-life of 2.50 X 105 years. Suppose that at time t = O, all the
Th 234 from 100 gm of uranium is removed. Plot the activities (in rutherfords)
of Th 234 and Pa 234 as íunctions oí time for (a) a 10-min period, (b) a 10-day
period. Discuss the kinds of equilibrium involved, if any.
14. Natural uranium contains 0.72% by weight oí U 235 and 99.28% of U 238 •
· In normal uranium ores, ali radioactive nuclides produced from U 238 and U 235
are in secular equilibrium with these parent nuclides. Find the atomic concen-
trations, in parts per billion, of each daughter nuclide of uranium, with a half-
life of one day or more. Neglect the less frequent type of decay when branching
occurs.
15. The Pb 2 º6 (RaG) content of a uranium-containing mineral can be used
to determine the age of the mineral. In the transition from U 238 to Pb 2º6, the
U 238 loses eight a-particles,
u2as-+ 8He + Pb2o6.

PROBLEMS 257
Show that if the presence of U 235 is neglected (and this has been shown to be
valid), and if the decrea.se in the amount of U 238 is neglected, then the age can
be represented approximately by the formula
Age = (Pb 2º6/U) X 7 .5 X 109 years,
where the quantity in parentheses is the weight in grams of Pb 2 º6 per gram of

ura.nium. What is the a.ge of a. uranium-bearing rock which conta.ins (a)
1.33 X 10- 2 gm of Pb 2 º6 per gra.m of uranium? (b) 0.1 gm of Pb 2 º6 per gram
of ura.nium? If the decrease in the a.mount of ura.nium because of disintegra.tion
is ta.ken into account, the la.st formula. me.y be replaced by. the better approxi-
ma.tion
º
Age = (Pb 2 6/U + º
0.58Pb 2 6) X 7 .5 X 109 years.
How much difference does this correction ma.ke in the ages obta.ined in (a) and
(b)?
16. Show that the disintegra.tion of one gra.m of uranium per year, under the
assumptions of the la.st problem, leads to 2.07 X 10- 11 gm of. helium, or 1.16
X 10-7 cm 3 of helium a.t N.T.P. Most of the helium produced in a uranium-
bearing mineral is trapped and retained as gaseous helium. When a. sámple
of the mineral is dissolved or fused, the helium can be collected and its volume
mea.sured. Show that the a.ge of the mineral is given, approximately, by the
formula
Age = (He/U) X 8.6 X 10 6 years,
where the quantity in bra.ckets represents the quantity of helium in cubic

centimeters a.t N.T.P. per gra.m of ura.nium. Show tha.t the a.ge may a.lso be
written
Age = 3.0(He/Ra) yea.rs.
A given sa.mple of a. ura.nium-bea.ring mineral wa.s found to conta.in 5.1 X 10- 5

cm3 of helium per gra.m of mineral, a.nd 1.05 X 10- 13 gm of radium per gra.m of
mineral. Wha.t is the a.pproxima.te a.ge of the sa.mple?
CHAPTER 11
ARTIFICIAL NUCLEAR DISINTEGRATION
11-1 Transmutation by alpha-particles: alpha-proton reactions. The

fact that certain atoms undergo spontaneous disintegration led to specu-
lation about the possibility of causing the disintegration of the ordinary
inactive nuclides. It seemed possible that if atoms were bombarded with
energetic particles, one of the latter might penetrate into a nucleus and
cause a disruption. The radiations from the natural radionuclides could
be used as projectiles, and a-particles seemed most likely to be effective
because of their relatively great energy and momentum. Since most of
the bombarding a-particles would probably be scattered, it was apparent
that the probability of causing a nuclear disintegration would be small.
The extent of the scattering could, however, be reduced and the probability
of disintegration increased by using sorne of the lighter atoms as targets,
thereby reducing the magnitude of the repulsive Coulomb forces between
the target nuclei and the a-particles.
The first disintegration based on these ideas was made by Rutherford
(1919),m who showed tbat the nuclei of nitrogen atoros emit swift
protons when bombarded with a-particles from radium C. The apparatus
that Rutherford used was simple but sensitive, and is shown schematically
in Fig. 11-1. In one end E of a box B was cut an opening which was
covered by a silver foil F. A zinc sulfide acreen was placed at S, just
outside the opening, and scintillations on the acreen were observed by
meaos of a microscope M. The source of the a-particles was radium C
placed on a small disc D, whose distance from S could be varied. The
silver foil F was thick enough to absorb the a-particles from the source.
Different gases could be introduced into the box and removed through the
side tubes T. When the box was filled '\\ith oxygen or carbon dioxide
F10. 11-1. Diagram of Rutherford's apparatus for the disintegration of

nitrogen nuclei by a-particles. [Rcprinted by permission from Rutherford,
Chadwick, and Ellis, Radiatiom from Radioactive Substancu, Cambridge Univer-
sity Press (Macmillan Co.), 1930.]
258
11-1] TRANSMUTATION BY ALPHA-PARTICLES 259
at atmospheric pressure, no scintillations were seen on the screen with

the source 7 cm or more away. This thickness x of gas was great enough
to absorb all of the a-particles from radium C even without the silver foil.
When the gas in the box was nitrogen, however, scintillations were ob-
served on the screen when the source of the a-particles was as much as
40 cm away. Since it was known that a-particles from radium C could
not penetrate 40 cm of air, Rutherford concluded that the scintillations
were caused by particles ejected from the nitrogen nucleus by the impact
of an a-particle. Measurement of the magnetic deflection of the particles
suggested that they were protons, and this surmise was confirmed by
more precise work.
Rutherford ruled out, by means of careful experiments, the possibility
that the protons carne from hydrogen present in the nitrogen asan impurity,
and concluded that artificial disintegration of nitrogen atoms had taken
place. The disintegration was caused by the a-particles from the radium
C, and one result was the emission of a highly energetic proton by the
nitrogen nucleus. The experimental results also showed that the prob-
ability of disintegration was very small; one proton was produced for
about one million a-particles passing through the gas. Rutherford and
Chadwick< 2 > extended the work on nitrogen to other elements and found
evidence of the disintegration of all of the light elements, from boron to
potassium, with the exception of carbon and oxygen. They also found
that in sorne cases the energy of the ejected protons was greater than that
of the bombarding a-particles. This result provided additional evidence
that the protons were emitted as the result of a disintegration process, the
extra energy being acquired in the accompanying nuclear rearrangement.
Two hypotheses were suggested as to the nature of the nuclear process
leading to the emission of the proton. They were: (a) The nucleus of the
bombarded atom simply loses a proton as the result of a collision with
a swift a-particle. (b) The a-particle is captured by the nucleus of the
atoro it hits, and the new, or compound, nucleus emits a proton. These
two hypotheses for the disintegration process could be subjected to
experimental test because in case (b) the a-particle should disappear,
while in case (a) it should still exist after the collision. The choice be-
tween the two possibilities was settled in 1925 when Blackett <3 > studied
the tracks produced by a-particles passing through nitrogen in a cloud
chamber. He showed, as can be seen in Fig. 11-2, that as a rcsult of a dis-
intcgration, the only tracks which could be secn wcre those of the inci-
dent a-particle, a proton, anda recoil nucleus. The absencc of a track cor-
responding to an a-particle aftcr the collision provcd that the a-particle
disappeared completely. If the disintegration process had been the result
simply of a disruption leading to the emission of a proton from the nitrogen
nucleus, there should have been four tracks rather than the three actually
260 ARTIFICIAL NUCLEAR DISI:-;TEGRATION [CHAP. 11
FIG. 11-2. A cloud chamber ph.otograph showing the disintegration of a X 14

nucleus by an a-particle with the formation of O 17 and a proton. Thc long
track of the proton and the short track of the recoiling oxygen nuclcus can be
seen. (Blackett. <3 >)
seen. It was concluded, therefore, that the a-particlc entered the nuclcus
of the nitrogcn atom with the formation of an unstable systcm which
immediately expelled a proton.
The disintegration of a nitrogen atom by an a-particle may be rcprc-
sented by an equation analogous to those uscd for chcmical reactions,
1N
14
+ 2He 4 -> (9F 18 )-+ s0 17 + 1H 1 •
In this equation, the symbols on thc lcft stand for thc rcacting nuclidcs.
The symbol in brackets stands for the unstable nuclcus formcd as thc
result of the capture of the a-particle by thc nitrogen nucleus; this kind
of nuclcus is oftcn callcd a compound nucleus. Thc emitted proton and
the final nucleus, the products of the reaction, are on thc right sidc of
the equation. Thc charge and mass numbers must be the samc on the
two sidcs of the cquation, so that the nitrogen nucleus (Z = 7) must be
transformcd into the nucleus of an isotope of oxygcn. In othcr words,
an atom of nitrogcn has becn transformcd, or transmutcd, into an atom
of oxygm. This transmutation may also be rcprcscntcd by the abbre-
viatcd notation N 14 (a,p)O 17 . Thc transmutation of othcr nudides hy
11-2] BALANCE OF MASS AND ENERGY IN NUCLEAR REACTIOXS 261
a-particles from radioactive substances may be represented by equations

similar to that for nitrogen.
:;Blº + 2He4 - [1Nl4] - 6cl3 + 1Hl
23 27 26 1
11Na + 2He 4 - [i 3Al ] - 12Mg + 1H
27 31 1
1aAl + 2He4 - [1sP ] - uSi
30
+ 1H
168
32
+ 2He 4 - 36
ltsA ] - 11CJ
35
+ 1H
1
39 42 1
19K + 2He 4 - 43
b1Se ] - 20Ca + 1H
In each case, the charge of the nucleus is increased by one unit and the
mass is increased by three units. The alpha-proton reaction may, there-
fore, be written in the form
zXA + 2He 4 - [z+2CnA+4] - z+1yA+a + 1H1, (11-1)
where X, Y, and Cn represent the target, product, and compound nuclei,
respectively.
It is often convenient to refer to a nuclear reaction in terms of the
incident and emitted particles, apart from the nuclei involved. Thus,
the reactions mentioned so far are examples of (a,p) reactions.
11-2 The balance of mass and energy in nuclear reactions. A nuclear

reaction such as that represented by Eq. (11-1) can be analyzed quantita-
tively in terms of the masses and energies of the nuclei and particles
involved. The analysis of nuclear reactions is one of the main sources
of information about nuclear properties and will therefore be discussed
in sorne detail before more reactions are considered. The analysis is
similar to that used for chemical reactions except that the relativistic
relation between mass and energy must be taken into account. Consider
a nuclear reaction represented by the equation
X+ x- Y+ Y, (11-2)
where X is the target nucleus, x the bombarding particle, Y the product

nucleus, and y the product particle. In the only type of reaction considered
so far, x is an a-particle, and y is a proton. In other reactions, to be dis-
cussed in later sections of this chapter, other bombarding and product
particlcs will be met. It will be assumed that the target nucleus X is
initially at rest so that it has no kinetic energy. Since the total energy of a
particle or atom is the sum of the rest energy and the kinetic energy, the
statement that the total energy is conserved in the nuclear reaction
means that
(11-3)
262 ARTIFICIAL NUCLEAR DISINTEGRATION [CHAP. 11
In Eq. (11-3), ms, Mx, m 11, and M y represent the masses of the incident
particle, target nucleus, product particle, and product nucleus, respectivcly;
the E's represent kinetic energies. We now introduce the quantíty Q,
which represents the difference between the kinetic energy of the products
of the reaction and that of the incident particle,
(11-4)
The quantity on the right side of Eq. (11-4) can be expressed in terms of
the masses because of the relationship (11-3),
Ey +E 11 - Ez = (Mx + ms - My - m 11)c 2 • (11-5)
Hence, from Eqs. (11-4) and (11-5),
Q = Ey +E 11 - Ez = (Mx + m,, - My - m 11)c 2 • (11--6)
The quantity Q is called the energy balance of the reaction or, more com-
monly, the Q-value; it can be determined either from the energy differ-
ence or from the mass difference in Eq. (11--6).
If the value of Q is positive, the kinetic energy of the products is greater
than that of the reactants; the reaction is then said to be exothermic
or exoergic. The total mass of the reactants is greater than that of the
products in this case. If the value of Q is negative, the reaction is endo-
thermic or endoergic. It is apparent from Eq. (11-6) that the analysis of
nuclear reactions involves information about nuclear masses and particle
energies. Nuclear reactions can, therefore, be used to obtain information
about the masses of nuclei, about particle energies, or about Q-values,
depending on wha.t information is available, and which quantities can be
measured.
The term Ey in Eq. (11--6) represents the recoil (kinetic) energy of
the product nucleus. It is usually small and hard to measure, but it can
FIG. 11-3. Conservation of momentum in nuclear rcactions. Thc outcoming

particles are viewed at an angle of 90° with thc dircction of the incidcnt particlcs.
11-2) BALANCE OF MASS AND ENERGY IN NUCLEAR REACTIONS 263
be eliminated by taking into account the conservation of momentum.

Consider, for simplicity, the special case in which the outcoming particles
are observed at an angle of 90° with the direction of the beam of pro-
jectiles, taken as the x-axis, as in Fig. 11-3. Before the collision, the
momentum vector is directed along the x-axis. Since the resultant mo-
mentum after the collision must also be directed along the x-axis, it follows
that
m.,v., = MyVy cos q,, m,¡;11 = M y V y sin q,, (11-7)
where v.,, v11 , and V y are the velocities of the incident particle, the ejected
particle, and the recoil nucleus, respectively, and q, is the angle between
the x-axis and the direction of recoil of the product nucleus. When the
last two equations are squared and added, the result is
(m.,v.,)2 + (m,1111)2 = (MyVy)2.
Setting E., = ½m.,vt E 11 = ½m,11!, Ey = ½M y VJ, and solving for Ey,

we get
(11-8)
Insertion of this expression for Ey into Eq. (11--6) gives
Q = E 11 ( 1 + ;~) - E., ( 1 - ;~) · (11-9)
The value of Q can be determined, therefore, if the energies of the incident

and ejected particles are measured and if the mass number of the product
nucleus is known.
In the general case of a reaction in which the outcoming particle is
observed at an angle fJ with the direction of the projectile beam, it can
be shown that (see Problem 19 at the end of the chapter),
Q = E(1 + ;;;)- E.,(1 - ;~)- iy

11 (E.,E 11m.,m11 )
112
cosfJ.
(11-10)
The importance of the last term decreases as the mass of the target nucleus
increases. If the masses are not known accurately, a good approximation
to the value of Q can be obtained by using the mass numbers instead of
the actual masses in Eqs. (11-9) and (11-10). Equation (11-10) reduces,
of course, to Eq. (11-9) when fJ = 90°.
In an endoergic reaction, the energy -Q is needed to excite the com-
pound nucleus sufficiently so that it will break up. This energy must be
supplied in the form of kinetic energy of the incoming particle. But not
ali of that kinetic energy is available for excitation because sorne is used to
264 ARTIFICIAL NUCLEAR DISINTEGRATION (CHAP. 11
impart momentum to the compound nucleus; this momentum is then dis-

tributed among the products of the reaction. Consequently, for -Q to
be available for excitation of the compound nucleus, we must supply
sorne energy in addition to -Q. The amount of energy needed for an
endoergic reaction is called the thruhol.d. energy and can be calculated
easily.
H we let Me and Ve denote the masa and velocity of the compound
nucleus, conservation of momentum requires that
m.,
m,.v,. = McVc, or V e= Me v.,.
The part of the kinetic energy of the incident particle needed for excita-
tion of the compound nucleus is
-Q = ½m,.v! - ½Me vi = ½m.,v! ( 1 - ;;) •
But, Me = Mx + m.,, and
( -Q) = ½m.,v! (MxM.:. m.,) .

The threshold energy is then
Eth = ½m.,v! = (-Q) ( 1 + ~) ; (11-11)
it can be determined experimentally and the result used to find the value
of Q from Eq. (11-11). For a reaction induced by 'Y-rays, m., = O, and the
threshold energy is just -Q.
Tbe masses which appear in the equations of this section are nuclear
masses. In actual calculations, however, they may be replaced by the
masses of the neutral atoros. The electrons which must be added to the
nuclei to form the neutral atoros cancel in the equation for a nuclear
reaction because the number of electrons is the same on the two sides
of the equation. As an example of the use of Eq. (11--6) consider the
N 14 (a,p)O 17 reaction. The following values of the atomic masses are
obtained from Table 11-1, Section 11-9.
Mx = M(N 14) = 14.007518 amu
m., = M(He 4 ) = 4.003873 amu
My = M(0 17) = 17.004529 amu
mw = m(H 1) = 1.008144 amu

11-3] THE NEUTRON: ALPHA-NEUTRON REACTIONS 265
The value of Q as calculated from the masses is Q = -0.001282 amu

= -1.19 Mev and the reaction is endoergic. A Q-value of --1.16 ± 0.04
Mev has been obtained experimentally from energy measurements and
agrees very well with the value calculated from the masses; according to
Eq. (11-11), the threshold energy is 1.49 Mev.
The equation for a nuclear reaction sometimes includes Q; for example,
X + x -+ [Cn] -+ Y + y + Q.
Although this is not necessary, it is a convenient way of showing the
energy balance.
11-3 The neutron : alpha-neutron reactions. The discovery of the

neutron has been discussed briefly in Section 8-3; it will now be treated
in somewhat greater detail. The capture of an a-particle by a nucleus
does not always result in the emission of a proton by the compound nucleus.
When beryllium was bombarded by a-particles, one of the products of
the reaction seemed to be a very penetrating k.ind of radiation, and no
protons were observed. It was assumed that the radiation consisted of
'Y-ray photons produced by the reaction
(11-12)
This assumption led, however, to difficulties. It was shown from absorp-
tion measurements that the energy of the "photons" should be about 7 Mev,
but these radiations were able to_ knock protons out of hydrogenous
materials and to impart to those protons energies of about 5 Mev. On the
assumption that the protons were liberated by elastic collisions with
'Y-ray photons, calculations showed that each photon must ha.ve had an
energy of about 50 Mev, a value much greater than that deduced from the
absorption measurements. There was, therefore, a serious contradiction
bctween the values of the photon energy given by the two methods. The
energy that can be attained by the assumed 'Y-radiation can also be com-
puted from known masses and energies. It is the sum of the Q-value
for the reaction (11-12) anci the kinetic energy of the incident a-particle
minus the recoil energy of the product nucleus. The Q-value is given by
the formula
since m 11 , the mass of the photon, is zero. Inserting the values :Mx =
9.01504, m,, = 4.00387, and M y = 13.00748, we get Q = 10.5 Mev. If
the kinetic energy of the a-particle is assumed to be 5 Mev, the total
energy available for the reaction is 15.5 Mev, part of which is retained
as recoil energy of the product nucleus. The greatest possible energy of
the assumed 'Y-radiation is, therefore, lesa than 15.5 Mev, which is again
much smaller than 50 Mev.
The problem of the identity of the producta of the bombardment of
beryllium was solved by Chadwick, <4 > as was mentioned in Chapter 8.
In a series of experimenta on the recoil of nuclei which were struck by
the rays coming from the bombarded beryllium, he showed that if these
rays were assumed to be 'Y-rays then the resulta of the experimenta led to
values for their energy which depended on the nature of the recoil nucleus.
For example, protons ejected from paraffin had energies which led to a
photon energy of 55 Mev, while recoiling nitrogen nuclei had energies which
led to a photon energy of about 90 Mev. If the energy of the recoil nuclei
was assumed to be produced by elastic collisions with photons, it turned
out that the energy that had to be attributed to the photon increased with
the mass of the recoil atom-a requirement that is contrary to the prin-
cipies of conservation of energy and momentum in elastic collisions.
Chadwick showed that ali of the difficulties disappeared when the hypoth-
esis was made that the radiation coming from beryllium bombarded with
a-particles consista of particles (neutrons) with a masa very nearly equal
to that of the proton, but having no charge. The nuclear reaction which
produces these neutrons is
{11-13)
where 0 n 1 is the symbol for the neutron. The penetrating nature of the
neutrons follows from the absence of a charge, and the energies of the
recoil atoms in Chadwick's experimenta could be completely accounted
for on the basis of elastic collisions with an energetic particle of unit masa.
Chadwick also proved that the masa of the neutron is approximately
equal to that of the proton, by meaos of the experimenta on the collisions
of neutrons with protons and nitrogen nuclei. Since the neutron is not
a charged particle, ita mass could not be determined from deffection
measurementa in electric and magnetic fields, and indirect methods had
to be used. Chadwick's method was based on the fact that in a collision
between a moving and a stationary particle, the velocity imparted to
the latter is greatest in a head-on collision, one in which the struck particle
moves in exactly the same direction as that in which the incident particle
approached it. An expression for the maximum velocity can be derived
from the equations of conservation of kinetic energy and momentum in
a head-on collision. Suppose that a particle with masa m 1 and velocity v
collides with a stationary particle of masa m3 , and let the velocity of the
incident particle after the collision be v1 and that of the struck particle
be v3 • The equation of conservation of energy is
(11-14)
11-3] THE NEUTRON: ALPHA-NEUTRON REACTIONS 267
and the equation of conservation of momentum is

(11-15)
The velocity v1 can be eliminated from the last two equations, and a
relationship is obtained between v2 and v,
2m1 (11-16)
If the incident particle is a neutron with known velocity v, its mass m 1

can be calculated from Eq. (11-16) if the mass m2 and the maximum speed
v2 of the recoiling nucleus are known. Chadwick, however, did not have
reliable information about the speed of the neutrons and had to use a
less direct method. If a neutron with the same velocity v collides with a
third particle of mass m 3 , the maximum velocity imparted to the latter
is given by
va= mi+ ms v, (11-17)
where m 1 and vare the same as in Eq. (11-16). Upon dividing Eq. (11-16)
by Eq. (11-17), we get
V2 mi+ ma (11-18)
-=
va mi + m2
If m 2 , m 3 , v2 , and v3 are known, the mass mi of the neutron can be found.
The maximum velocity imparted to hydrogen nuclei ejected from pa.raffin
wax by the neutrons emitted by beryllium bombarded with polonium
a-particles was found to be 3.3 X 109 cm/sec. The recoils of nitrogen
nuclei after being struck by the same neutrons were observed in cloud-
chamber experimenta an.d their maximum velocity was found to be
4. 7 X 108 cm/sec. Taking m 2 and m 3 to be the masses of the proton and
nitrogen nuclei, equal to 1 amu and 14 amu respectively, and with the
corresponding values v2 = 3.3 X 10 9 and v3 = 4.7 X 108 , we get from
Eq. (11-18)
mi + 14 v2 3.3 X 109
mi+ 1 = v3 = 4.7 X 108 '
and m 1 = 1.15. This result, although approximate because of errors in

the determination of the maximum recoil velocities, showed that the mass
of the neutron is roughly the same as that of the proton.
Chadwick obtained a somewhat better value of the neutron mass by
considering the reaction
aBú + 2He 4 - . 1N 14 + oni.
The Q-value was found from the kinetic energies of the incident a-particles,
of the nitrogen nuclei (measured in cloud-chamber experiments), and of
the neutron (estimated from the maximum recoil energy of a proton after
being struck by a neutron). From the Q-value and the known atomic
masses of B 11 , He 4, and N 14, the mass of the neutron can be obtained with
the aid of Eq. (11-6). In this way, Chadwick found a value between
1.005 and 1.008 atomic mass units. The best methods now available for
determining the neutron mass give 1.008983 amu.
The proof of the existence of the neutron and the development of
methods for analyzing nuclear reactions in which one of the products is a
neutron led to the discovery of other reactions similar to that represented
by Eq. (11-13). Among these are
10
aLi 7 + 2He 4 --+ [sB 11 ]--+ sB + 0n1,
17
1N + 2He 4 --+
14 (9F 18)--+ 9F + on1.
23 27 26
uNa + 2He 4 --+ liaAl )--+ 1aAl + on1,
27 31 30
1aAI + 2He --+ [1sP ] - 15P + on1,
4
1sA
40
+ 2He 4 --+ [20Ca 44]--+ 20Ca 43 + 0 n 1.
When these reactions are compared with those mentioned in Section 11-1,
it is seen thll,t sorne atoms, such as N 14, Na 23, and A 27, emit either a proton
or a neutron when bombarded with a-particles. The compound nucleus,
e.g., [9F 18], [13Al 27 ], [1sP 31 ], may disintegrate by means of either procesa,
giving a different product in the two cases. It will be seen in the next
chapter that the products of these nuclear reactions are sometimes
radioactive.
11-4 The acceleration of charged particles. The early artificial dis-

integrations were produced by bombarding materials with a-particles
from natural radioactive substances. The neutrons from the bombard-
ment of beryllíum with a-particles were also used as projectiles and, as
will be seen later, have produced many artificial disintegrations. The
development of the field of nuclear transmutation was limited, however,
because a-particles could be obtained only in beams of low intensity and
with energies not greater than 7.68 Mev (RaC'), and transmutations by
these particles were possible only with the lighter elements. Furthermore,
it seemed possible that other particles, protons, deuterons, and even 'Y-rays,
might yield interesting results when used as projectiles. Protons and
deuterons with a single positive charge would experience smaller repulsive
Coulomb forces than a-particles in the field of the nucleus, and it was
11-5] TRANSMUTATION BY PROTONS 269
thought that this effect might make disintegrations possible with low-
energy protons. It became important, therefore, to develop laboratory
methods of accelerating charged particles to high energies, and several
successful methods were invented about 1930.
Among the earliest particle accelerators were the voltage multiplier of
Cockcroft and Walton, the electrostatic generator of Van de Graaff, and
the cyclotron of Lawrence and Livingston. By the suitable choice of a
machine and its operating conditions, protons, deuterons, or a-particles
with energies as high as several Mev could be obtained by 1936. Thc
further development of these machines and the design of new accelerators
such as the frequency-modulated cyclotron and the synchrotron made it
possible to obtain particles with energies up to hundreds of Mev. The
"Cosmotron" at the Brookhaven National Laboratory has produced pro-
tons with energies of 2.5 billion electron volts, and recently built machines
have raised the energies to 25 Bev. Electrons can be accelerated to ener-
gies up to 300 Mev in a betatron, orina linear accelerator. When these
electrons strike a metal target, highly energetic x-rays are produced,
which can also be used as projectiles. Finally, the discovery of nuclear
fission and the development of chain-reacting piles, or nuclear reactors
(cf. Chapter 20), have supplied highly intense sources of neutrons for
nuclear bombardment.
With the available particles and accelerators, a large variety of nuclear
reactions can be made to occur. The rest of this chapter will be limited
to the commonest reactions which take place with particle energies up to
about 20 Mev. These reactions will serve to show the most important
features of artificial disintegration, and the ideas involved can then be
extended to the more complex reactions which occur with particle energies
of the order of 100 Mev. The particle accelerators used to produce pro-
jectiles for nuclear reactions will be discussed in Chapter 21. For the
purposes of the present chapter, it is enough to realize that various bom-
barding particles of almost any desired energy can be obtained with the
help of suitable devices, and that these projectiles can cause many different
kinds of nuclear reactions.
11-5 Transmutation by protons. The first case of a nuclear disin-

tegration brought about entirely by artificial means was one for which
protons were used as the projectiles. Cockcroft and Walton< 5 > bombarded
lithium with protons accelerated to energies of 0.1 to 0.7 Mev. Scintilla-
tions caused by particles ejected from the lithium were observed on a
zinc sulfide screen placed a short distance away. The particles were
proved to be a-particles by photographing their tracks in a cloud cham-
ber, <5 > and the lithium isotope that was disintegrated was proved to be
Li7. The cloud-chamber pictures showed that two a-particles lea.ve the
point of disintegration and proceed with equal energies in opposite diree-

tions. The reaction may be represented by the equation
aLi 7 + 1H 1 --+ (4Be 8] --+ 2He 4 + 2He 4.
This reaction has a certain historical interest because it provided one
of the earliest quantitative proofs< 7> of the validity of the Einstein mass-
energy relationship. It was a good reaction for this purpose because the
energies of the products could be measured precisely, and the masses were
known. The proof will be repeated here, with newer values of the various
quantities. The values of the atomic masses are, from Table 11-1,
Mx = M(Li 1 ) = 7.018222 amu,
m., = m(H 1) = 1.008144,
My = m 11 = m(He 4) = 4.003873.
The value of Q obtained from the masses is Q,,. = 0.01860 amu =

17.34 Mev. The best experimental value of Q, obtained from the energies
of the incident protons and the emergent a-particles, is 17.33 Mev.
This agreement shows clearly that the theoretical expression for Q,
based on the mass-energy relationship, agrees with experiment, and that
there was a genuine release of energy from the lithium atom at the
expense of its mass. Since 1932, many nuclear transformations have been
studied in detall and the results invariably agree with the relationships
deduced from the Einstein equation.
The disintegration that has just been discussed is an example of the
general type
zXA + 1H 1 --+ [z+1CnA+I]--+ z-1yA- 3 2He 4. +(11-19)
Other examples are
3 Li + 1H 1 --+ (4Be 7] --+ 2He 3 + 2He 4,

6
4Be 9 + 1H 1 --+ [sB 10] --+ aLi 6 + 2He 4,

9F 19 + 1H 1 --+ [ioNe 2°] --+ s0 16 + 2He4,
1
1aAI 27 + 1H --+ [uSi 28 ] --+ 12Mg 24 + 2He 4.
An interesting reaction occurs when the target is B 11 ,
5B
11
+ 1H 1 --+ [eC 12] --+ 4Be8 2He4; +
Be 8 is highly unstable and breaks up into two more a-particles,
4Be 8 --+ 2He 4 +
4
2He .
11-5] TRANSMUTATION BY PROTONS 271
The final result of the bombardment is three a-particles, and the reaction
may be regarded as a case of multiple particle production.
Proton bombardment can also result in nuclear reactions in which one
of the products is a neutron. The reaction is of the type
1
zXA + 1H 1 - lz+1CnA+I] - z+1YA + on . (11-20)
The effect of the transmutation is to increase the charge on the nucleus

by one unit, moving it one place to the right in the periodic table; the
mass number is not changed. Examples of this reaction are
sB + 1H 1 -
11 12
(5C ] - 6C+ on 1,
11
8018 + 1H1 - (9F19] - 9F1s + on1,

58 1 59 58 1
2sNi + 1H - [29Cu ] - 20Cu + on .
In the (p,n) reactions, the mass change is usually negative, so that the
reactions are endoergic. The reaction
65 1 66 65 1
20Cu + 1H - [aoZn ] - aoZn + on
is an example of a case for which the threshold energy has been meas-
ured. The source of the protons was an electrostatic generator, and
the relative number of neutrons produced was determined as a function
of the energy of the bombarding protons. The experimental results in
the neighborhood of the threshold energy were then analyzed, and the
value of the threshold energy was found to be Eth = 2.164 ± 0.01 Mev.
The Q-value is obtained from Eq. (11-11), in which the masses Mx and
m,, may be replaced by the mass numbers 65 and 1, respectively. Then
Q = -Eth X ff = -2.164 X ff = -2.13 Mev.
In sorne cases the bombarding proton is simply captured by a nucleus.

The compound nucleus which is formed is again unstable, but becomes
less unstable by emitting a 'Y-ray photon rather than a neutron or an
a-particle. The reaction is of the type
zXA + 1H 1 - lz+1CnA+ 1] - z+1YA+i + 'Y, (11-21)

and sorne examples are
7 1 8 8
aLi + 1H - (4Be J- + 'Y,
4Be
12
6c + 1H 1 - 1
l1N ªJ - 1N ª + 'Y,
1
1 20 20
0F + 1H - [10Ne ] - 10Ne + 'Y,
19
27 1 28 28
1aA1 +
1H - (14Si ] - 14Si + 'Y.
The "Y-ray photons emitted in these reactions are often very energetic and
can, in turn, be used to produee nuclear disintegrations. The bombard-
ment of lithium yields photons with an energy of 17.2 Mev, far more
energetic than the 2.6-Mev photons which are the most energetic available
from the natural radioactive nuclides.
Another type of disintegration caused by protons is that in which
deuterons are produced. Examples of this reaction are
4Be 9 + 1H 1 -+ [sB 10 ) -+ 4Be 8 + 1H 2,
aLi 7 + 1H 1 -+ (4Be 8 ] -+ aLi 6 + 1H 2 .
11-6 Transmutation by deuterons. A great many nuclear reactions
have been observed with high energy deuterons as the bombarding par-
ticles. In most of these cases, the deuterons have been accelerated up to
energies of several Mev in a cyclotron or in an electrostatic generator.
One of the first deuteron-induced reactions studied was again that on a
lithium atom, <9>
aLi 6 + 1H 2 -+ [4Be 8] -+ 2He 4 + 2He 4.
Other examples of (d,a) reactions are
16 + H 2 -+ (9F 18] -+ 1N 14 + 2He4,
80 1
1aAl + 1H 2 -+ [uSi 29] -+ 12Mg 25 + 2He 4.
27
The mass change is usually positive so that the Q-values are positive, and
the reactions are exoergic. The general reaction of this type is
(11-22)
As in the case of proton bombardment, disintegrations produced by

deuterons do not always yield a-particles. Deuteron-proton reactions are
often found, such as
aC 12 + 1H 2 -+ [¡N 14 ] -+ oC 13 + 1H1,
uNa 23 + 1H2 -+ [12Mg 25} -+ 11Na24 + 1H1,
1sP31 + 1H 2 -+ !io833 }-+ 1sP32 + 1H 1.
The result of these transformations is to increase the mass of the nucleus
by one unit, leaving the charge unchanged. The general (d,p) reaction
IS
(11-23)
The Q-values for these reactions are usually positive, so that the reactions
are exoergic.
11-7] TRANSMUTATION BY NEUTRONS 273
Neutrons are often produced as a result of deuteron bombardment, as

in the following reactions:
6C + 1H 2 --+ [1N 14]--+ 1N 13 + on1,

12
10
4Be 9 + 1H 2 --+ [sB 11 ]--+ sB + on1,
1
aLi 7 + 1H 2 --+ (4Be 9 )--+ 4Be 8 + on .
The general (d,n) reaction is represented by the equation
zXA + 1H 2 --+ [z+1CnA+2]--+ z+1YA+I + on 1 . (11-24)
One of the most interesting cases of deuteron bombardment is that in

which the target contains deuterons. Deuterium targets have been made
by freezing deuterium oxide (D 20-"heavy water") onto a surface kept
cold by liquid air. Both the (d,p) and the (d,n) reactions have been
observed,
2 +
2 4 3 +
1H 1H --+ [2He ] --+ 1H 1H1,
1
1H 2 1H 2 --+ bHe 4] --+ 2He3
+ +
on .
The excited compound nucleus [2He 4] can disintegrate in two ways. In

the first, a proton and a new isotope of hydrogen are formed; in the second,
a neutron and an isotope of helium are formed. The new hydrogen isotope
has a mass very nearly equal to 3 atomic mass units, and has been given
the name tritium. lt is unstable and has a haH-life of about 12 years.
The helium isotope of mass 3 is stable and is found in nature.
11-7 Transmutation by neutrons. Neutrons have proved to be espe-

cially effective in producing nuclear transformations. Since they have no
electric charge, they are not subject to repulsive electrostatic forces in the
neighborhood of a positively charged nucleus, and are therefore more
likely to penetrate nuclei than are protons, deuterons, or a-particles. N ot
only are highly energetic neutrons capable of causing nuclear reactions,
but slowly moving neutrons are also extremely effective. Because of
these properties of the neutron many more disintegrations have been
produced with neutrons than with any other particle. Before the develop-
ment of the chain-reacting pile, or nuclear reactor, the main sources of
neutrons were reactions such as H 2(d,n)He 3, Be 9 (d,n)B 1º and Be 9 (a,n)C 12 .
The fast neutrons produced were slowed down by allowing them to pass
through sorne hydrogen-containing substance such as water or paraffin.
A neutron gives upa large fraction of its energy in a collision with a hydro-
gen nucleus and, after many collisions, the average energy of the neutrons
is reduced to a few hundredths of an electron volt. These slow neutrons
have been found to be especially useful from a practical standpoint, and

their interactions with nuclei are particularly interesting.
The reaction between a neutron and a nucleus gives rise in most cases
to an a-particle, a proton, a 'Y-ray photon, or to two neutrons. The (n,a)
reactionuo> is represented by the general equation
zXA + on 1 -+ [zCn.t+i¡-+ z_zyA-3 + 2He 4. (11-25)
Among the most interesting examples of this reaction are
3Li
6
+ 0n 1 -+ (3Li 7] -+ 1H3 + 2He4,
5B
10
+ 0n 1 -+ lsB 11 ]-+ aLi7 + 2He4,
1aA1
27
+ on 1 -+ l1aA1 28]-+ 11Na2• + 2He 4
•
The first two reactions have a relatively high yield and are therefore often
used to detect neutrons. In one method, an iotlization chamber is lined
with boron, usually in the form of a compound. The capture of a neutron
by an atom of the B 1 º isotope causes the liberation of an a-particle, which
is detected by the ionization it produces in the chamber.
In sorne cases, the compound nucleus formed by the capture of a neutron
emits a proton. The (n,p) reaction is described by the equation
(11-26)
The effect of this reaction is to replace a proton in the nucleus by a neutron;

the mass number is not changed, but the charge is decreased by one unit
and the atom is moved one place to the left in the periodic table. Sorne
examples of this reaction are
1N 14 + on -+ [1N ]-+ 6C +
1 15 14
1H1,
1aA1 27 + on 1 -+ [1aA1 28]-+ 12Mg27 + 1H 1 .

The first of these reactions can be induced by slow neutrons; with heavier
nuclei, more energetic neutrons must be used, as in the case of Al 27 and
in the reaction
aoZn 64 + .on 1 -+ [aoZn ] -+ 29Cu + 1 H •
65 64 1
In another type of reaction, one neutron is captured by the nucleus

and two neutrons are emitted. The (n,2n) reaction leaves the charge
of the nucleus unchanged and decreases the mass number by one unit.
The result is an isotope of the target nucleus with a mass number one
unit smaller,
(11-27)
11-8] TRANSMUTATION BY PHOTONS 275
An example of this reaction is

13Al21 + on1-+ (13Al2s]-+ iaAi26 + on1 + on1.
The mass change in the (n,2n) reaction is always negative; the Q-value
is negative and fast neutrons are needed to bring about this reaction.
The commonest process which results from neutron capture is radiative
capture, crn represented by
(11-28)
The compound nucleus emits one or more 'Y-ray photons, and the final
nucleus is an isotope of the target nucleus with a mass number one unit
greater. The (n,"Y) process has been observed in nearly all of the elements.
The Q-values are always positive, the excess energy being carried away
by the "Y-rays. The simplest (n,"Y) reaction with slow neutrons occurs
with hydrogen as the target nucleus,
1H 1 +
on 1 -+ [1H 2] -+ 1H
2 "Y. +
The product of the reaction is deuterium. When deuterium is bombarded
with slow neutrons, tritium is formed,
1H 2 + on 1 -+ [1H 3]-+ 1H 3 + "Y.

Other typical (n,'Y) reaetions are
28 28
+
1aAl 27 on 1 -+ [1aAl )-+ 1aAl + "Y,
+ on 1 -+ (49ln ]-+ 49ln + "Y,
116 116
49ln 115
o2U2as + on i -+ [92U2ªºJ -+ 92U2ao + 'Y.
The radiative capture of slow neutrons often results in product nuclei
which are radioactive, and this reaction is one of the most important
sources of artificial radioactive nuclides.
11-8 Transmutation by photons. Atomic nuclei can also be disin-

tegrated by bombardment with high-energy photons, a process which is
usually called photodisintegration. Since the photon has no mass, it can
supply only its kinetic energy to a nuclear reaction. This energy must be
at least as great as the binding energy of a nuclear particle before such a
particle can be ejected from a nucleus. Photodisintegration reactions are,
therefore, endoergic and usually have threshold energies of the order of 10
Mev. With only two exceptions, photodisintegration does not occur with
"Y-rays from natural radioactive substances. These exceptions are the
deuteron, which has a bindingenergy of only 2.2 Mev, and the nuclide 4Be 9,
276 ARTIFICIAL NUCLEAR DISI!l1TEGRATION [CHAP. 11
in which one neutron is loosely bound. In the case of the deuteron,' 12¡
the reaction is
and is the reverse of the radiative capture of a neutron by a proton. In

beryllium, the reaction is
9
4Be +'Y--+ [4Be9] --+ 4Be
8
+ 1
0n ,
with a Q-value of -1.67 Mev.

It will be recalled from Section 11-5 that when Li is bombarded with
protons, 17-Mev 'Y-rays are produced. These have been used successfully
to cause the photodisintegration of other nudides, as in the ('Y,n) reaction
1sP
31
+'Y--+ [u;P 31 ] --+ 1sP
30
+ on 1 •
The ('Y,p) reaction requires still higher energies, and has been observed
with high-energy photons from a betatron.
11-9 Nuclear chemistry: nuclear masses. The transmutations which

have been discussed are the commonest examples of a large variety of
nuclear reactions brought about by charged particles, neutrons, and pho-
tons. Many other reactions have been discovered and analyzed. Sorne
have not been considered here because they are relatively rare; others
have been omitted because they take place only at very high energies, and
their importance and interest lie outside the limits of elementary nuclear
physics.
It has been seen that nuclides can be transmuted in a number of different
ways. A given nuclide can be transformed to any one of severa! other
nuclides, the nature of the newly formed substance depending on the
particle used to produce the transmutation and upon the particle given
off during the reaction. For example, the nuclide 13 Al 27 has been made
to undergo the following transformations.
(1) 13Al +
27
2He --+ [1sP 31 ]--+ 1sP30 + on 1
4
(2) 1aA1
27
+ 2He --+ [1sP ] --+ uSi
4 31 30
+ 1H 1
(3) 13Al
27
+ 1H
2
--+ [uSi 29 ] --+ 12Mg25 + 2He4
(4) 1aA1
27
+ 1H
2
--+ [uSi 29] --+ 1aA1 28 + 1H 1
(5) 13Al
27
+ 1H
2
--+ [14Si 29]--+ 14Si 28 + 0 n 1
(6) 13Al
27
+ 1H
2
--+ [uSi 29] --+ 11Na24 + 1 H 1 + 2 He
4
(7) 13Al
27
+ 1H
1
--+ [uSi 28] --+ 12Mg24 + 2He 4
11-9) NUCLEAR CHEMISTRY: NUCLEAR MASSES 277
(8) 1aA1 27 + 1H 1 - [14Si2s] - 14Si21 + on1

(9) 13Al 27 + 1H 1 - (14Si2s] - 14Si2s + 'Y
(10) 1aA1 27 + 1H 1 - [14Si 28 ] - 11Na 24 + 3(1H 1) + on 1
(11) 1aA1 27 + on 1 - [1aAl 28] - 11Na 24 + 2He 4
(12) 1aA1 27 + on 1 - [1aA1 28 ) - 12Mg 27 + 1H 1
(13) 1aAl27 + on1 - [1aAl 28 ] - 13Al 28 + 'Y
(14) 13Al21 + 'Y - [1aA1 27 ) - 1aAl 26 + 0 n 1
(15) 1aA1 27 + 'Y - [1aA1 27 ) - 11Na 25 + 1H 1 + 1H 1
(16) 1aA1 27 + 'Y - [1aA1 27 ) - 11Na 24 + 1H 1 + 1H 1 + 0 n 1
The reactions numbered (10), (15), and (16) are included to give sorne
idea of transmutations which are possible at higher energies. The protons
in (10) and the 'Y-rays in (15) and (16) have energies of about 50 Mev.
At still higher energies, still more particles are given off and the term
atom smashing is really appropriate.
In the 16 reactions listed, 11 different nuclides are formed. One nuclide,
11 Na 24, is formed from 13Al 27 in 4 different ways, in reactions (6), (10),
(11), and (16). lt can also be formed in 5 more reactions.
11Na 23 + 1H 2 - f12Mg 25] - 11Na 24 + 1H 1

11Na 23 + on 1 - [11Na 24 ) - 11Na 24 + 'Y
12Mg 24 + on 1 - [12Mg 25 ) - 11Na + 1H 1
24
12Mg 25 + 'Y - li2Mg 25 ) - 11Na 24 + 1H 1

12Mg 26 + 1H 2 - [1aA1 28 ) - 11Na 24 + 2He4
Thus, the nuclide 11 Na 24 can be formed in at least 9 different ways. These
examples show the tremendous variety of nuclear reactions which can be
induced with the nuclides, particles, accelerators, and other sources of
projectiles now available. The study of these reactions is one of the great
branches of nuclear physics and because of the problems involved in
separating and identifying the products, this field is aptly called "nuclear
chemistry."
One of the important applications of nuclear reactions is in the deter-
mination of nuclear, or atomic, masses, a problem which has a direct
analog in chemistry. The determination of atomic masses by means
of the mass spectrograph has been discussed in Chapter 9 and the results
obtained by that method can now be combined with values obtained from
nuclear reactions. According to Eq. (11-6), a Q-value can be expressed
TABLE 11-1*
~
THE MASSES OF LIGHT ATOMS UP TO CHLORISE-37
Nuclide Atomic mass Nuclide Atomic mass Nuclide Atomic mass
ni 1.0089830 (±1.7) C14 14.0076845 (±3.1) Mg23 23.001409 (±15)

Hi 1.0081437 (±1.8) N1a 13.0098617 (±4.8) Mg24 23.992638 (±13) >
H2 Ni4 Mg25
2.0147361 (±2.9) 14.0075179 (±3.0) 24.993747 (±14) ~
Hª 3.0169980 (±4.5) Ni5 15.0048627 (±4.5) Mg26 25.990796 (±15) :;_;
He 3 3.0169786 (±4.5) Ni6 16.01100 (±130) Mg21 26.992866 (±15) ;:;
He 4 4.0038727 (±2.1) Ql4 14.013052 (±42) A12s 24.99835 (±60) >
t"
He5 5.013880 (±31) Ql5 15.007768 (±7) AI21 26.990080 (±14) z
q
He 6 6.020818 (±29) Ql6 16.0000000 (±0) A12s 27.990768 (±15)
~
Li 5 5.01392 (±70) 011 17.0045293 (±3.9) 8¡21 26.995233 (±17) ~
Li 6 6.0170281 (±4.7) Ql8 18.004840 (±9) s¡2s 27.985777 (±16) :,:i
Li7 7.018222 (±6) Ql9 19.00931 (±300) s¡29 28.985661 (±16) t,
Li 8 8.025020 (±7) FI1 17.0074989 (±4.1) Si 30 5l

Be 7 7.019149 (±6) F1s 18.006635 (±10) s¡a1
29.983252
30.985153
(±17)
(±18)
z8
t"1
Be8 8.0078473 (±4.2) F19 19.004447 (±7) p28 28.00055 (±300) o
Be 9 9.015041 (±5) Fzo 20.006341 (±10) p30 29.987885 (±36) ~
8
p31
Belº 10.016706 (±7) Nel9 19.007944 (±9)
p32
30.983563 (±18) o
z
B9 9.016190 (±6) Ne20 19.998765 (±10) 31.984025 (±20)
n10 10.016109 (±6) Ne21 21.000494 (±10) sa2 31.982190 (±20)
Bil 11.012795 (±5) Ne22 21.998346 (±16) Sªª 32.981887 (±23)
w2 12.018166 (±7) Ne23 23.001755 (±18) S34 33.97920 (± 170)
cio 10.02014 (±100) Na22 22.001396 (±17) sas 34.98014 (±80) -¡:;-
:i:
c11 11.014924 (±6) Na23 22.997047 (± 11) c¡as 34.97996 (±80)
c12 12.0038065 (±3.9) Na24 23.998560 (±12) Cl 36 35.97974 (±90) ~
Cl3 13.0074754 (±4.1) Na25 24.99772 (±310) c¡a1 36.997540 (±45)
*From M&tta.uch &nd Ji:verling.

REFERENCES 279
as the difference between the masses of the reactants and products of a

nuclear reaction. Many Q-values have been measured with high precision
in recent years, 03 • 14 > and it is now possible to obtain the masses of light
nuclei directly in terms of the mass of 0 16 without the use of any mass-
spectroscopic results.° 5 • 16> The most recent compilations of atomic
masses combine mass-spectroscopic and Q-value data. Nuclear reactions
are, of course, an essential source of information concerning the masses of
nuclides which do not occur naturally and can be obtained only in nuclear
reactions. The atomic masses of nuclides 07> up to Cl 37 are listed in Table
11-1. Both naturally occurring and artificially produced nuclides are
included in the table. The uncertainty, given in parentheses, is expressed
in units of 10- 5 amu.
REFERENCES
GENERAL
RuTHERFORD, CHADWICK, and ELLIS, Radiatúms from Radioactive Substances.
New York: Macmillan. 1930, Chapter 10 (reprinted with corrections 1951).
M.S. LIVINGSTON and H. A. BETHE, "Nuclear Physics, C. Nuclear Dynamics,
Experimental," Revs. Mod. Phys. 9, 245--390 (July 1937).
J. D. STRANATHAN, The "Particles" of Modern Physics. Philadelphia: Blakis-
ton, 1944, Chapters 10, 11.
R. T. BEYER, ed., Foundations of Nuclear Physics; Facsimiles of Thirteen
Fundamental Studies as they were originally reported in the Scientific Journals,
with a Bibliography. New York: Dover, 1947.
R. D. EvANS, The Atomic Nucleus. New York: McGraw-Hill, 1955, Chapters
12, 13.
PARTICULAR
l. E. RuTHERFORD, "Collision of a-Particles with Light Atoms, Part IV."
An Anomalous Effect in Nitrogen, Phil. Mag. 37, 581 (1919). (In Beyer.)
2. E. RuTHERFORD and J. CHADWICK, "Artificial Disintegration of Light
Elements," Phil. M ag. 42, 809 (1921); 44, 417 (1922); Proc. Phys. Soc. (London)
36, 417 (1924).
3. P. M. S. IlLACKETT, "Photography of Artificial Disintegration Collisions
and Accuracy of Angle Determinations," Proc. Roy. Soc. (London) Al34, 658
(1932).
4. J. CHADWICK, "The Existence of a Neutron," Proc. Roy. Soc. (London)
Al36, 696 (1932). (In Beyer.)
5. J. D. CoCKCROFT and E. T. S. WALTON, "Experiments with High Velocity
Positive Ions." II. "The Disintegration of Elements by High Velocity Protons,"
Proc. Roy. Soc. (London) Al37, 229 (1933).
6. P. l. DEE and E. T. S. WALTON, "Transmutation of Lithium and Boron,"
7. K. T. BAINBRIDGE, "The Equivalence of Mass and Energy," Phys. Rai.

44, 123 (1933).
8. SHOUPP, JENNINGS, and JoNEs, "Threshold for the Proton-Neutron Reac-
tion in Copper," Phys. Rev. 73, 421 (1948).
9. LEwrs, L1v1NGSTON, and LAWRENCE, "The Emission of a-Particles from
Various Targets Bombarded by Deuterons of High Speed," Phys. Rev. 44, 55
(1933).
10. N. FEATHER, "Collisions of Neutrons with Nitrogen Nuclei," Proc. Roy.
Soc. (London) A136, 709 (1932).
11. FERMI, AMALDI, D' AoosT1No, RASETTI, and SEGRE, "Artificial Radio-
activity Produced by Neutron Bombardment," Proc. Roy. Soc. (London) A14ó,
483 (1934).
12. J. CHADWICK and M. GoLDHABER, "Nuclear Photoelectric Effect,"
Nature 134,237 (1934); Proc. Roy. Soc. (London) AlSl, 479 (1935).
13. W. W. BuECHNER, "The Determination of Nuclear Reaction Energies by
Deflection Methods," Progress in Nuclear Physics, O. R. Frisch, ed., Vol. 5,
p. 1, 1956.
14. D.M. VAN PATTER and W. WHALING, "Nuclear Disintegration Energies."
I, Revs. Mod. Phys., 26, 402 (1954); II, ibid., 29, 757 (1957).
15. L1, WHALING, FowLER, and LAURITSEN, "Masses of Light Nuclei from
Nuclear Disintegration Energies," Phys. Rev. 83, 512 (1951).
16. C. W. L1, "Nuclear Mass Determinations from Disintegration Energies:
Oxygen to Sulphur," Phys. Rev. 88, 1038 (1952).
17. J. MATTAUCH and F. EvERLING, "Masses of Atoms of A < 40," Progress
in Nuclear Physics, O. R. Frisch, ed., Vol. 6, p. 233 (1957).
281
PROBJ,EMS
l. Complete the reactions listed below. Rewrite them in a way which shows
the balance of atomic and mass numbers, and indicate the compound nucleus
in each case: H 1(n,'Y); H 2(n,'Y); Li 7(p,n); Li 7 (p,a); Be9(p,d); Be 9(d,p); Be 9 (p,a);
B 11 (d,a); C 12 (d,n); N 14 (a,p); N 15 (p,a); O16 (d,n).
2. Calculate the Q-values of each of the reactions of Problem 1. Which
reactions are exoergic? Which are endoergic? Use the masses listed in Table
11-1.
3. In the four reactions which follow, the mass of the target nucleus Al is
given as 26.99007 amu, and the Q-value of each reaction is given. Calculate
the mass of the product nucleus.
(a) Al 27 (n,'Y); Q = 7. 722 Mev

(b) AI 27(d,p); Q = 5.497 Mev
(e) Al 27 (p,a); Q = 1.594 Mev
(d) Al 27 (d,a); Q = 6.693 Mev
4. One of the reactions which occurs when boron is bombarded with 1.510-
Mev deuterons is B 11 (d,a)Be9. The a-particles coming off at an angle of 90°
with the direction of the deuteron beam have an energy of 6.370 Mev. What is
the Q-value of the reaction?
5. Under the conditions of Problem 4, the reaction B 11 (d,p)B 12 was also
found to occur, and to have a Q-value of 1.136 Mev. What is the energy of the
protons observed at 90º?
6. The reactions Be9(p,n)B 9, C 13 (p,n)N 13 , and O18 (p,n)F 18 were found to
have the threshold energies 2.059 Mev, 3.236 Mev, and 2.590 Mev, respectively.
What are the Q-values for the reactions?
7. Calculate, from the atomic masses, the threshold energies for the reactions
I3ll(p,n)C11, O1s(p,n)F1s, and Na2ª(p,n)Mg23.
8. Write ten nuclear reactions which might be expected to take place with
C 12 as the target. Which of these reactions would be expected to be exoergic?
Endoergic?
9. Write ten nuclear reactions which should give the product nucleus C 12 .
Which reactions should be exoergic? Endoergic?
10. A sample of silicon is bombarded with 1.8-Mev deuterons. Show that
the following reactions can produce charged particles with energies greater than
4 Mev: Si 28 (d,p)Si 29 ; Si 29 (d,p)Si 3º; Si 29 (d,a)Al 27 ; Si3º(d,p)Si 31 ; Si3º(d,a)Al 28 .
ll. The reactions listed below have been found to have the in_dicated Q-
values
(a) Ql6(d,a) Nl4; Q = 3.112 Mev
(b) Nl4(d,p) Nl5; Q = 8.615 Mev
(e) Nl5(d,a) cia; Q = 7.681 Mev
(d) C 13 (d,a) B 11; Q = 5.160 Mev
(e) B 11 (p,a) Be8 ; Q = 8.567 Mev
(f) Be 8 (a)a; Q = 0.089 Mev
Show that these data lead to a value of 4.003869 amu for the mass of He 4 ü
the mass of deuterium is taken as 2.014723 amu.
12. Suppose that, in separate experiments, C 12 is bombarded with neutrons,
protons, deuterons, and a-particles, each having kinetic energy of 10.00 Mev.
(a) What would be the excitation energy of the compound nucleus in each case
ü it is assumed that the recoil energy of the compound nucleus can be neglected?
(b) What is the recoil energy of the compound nucleus in each case when the
recoil energy is not neglected?
13. With the aid of the data of Table 11-1, calculate the binding energy of
the "last neutron" in each of the following sets of isotopes:
(a) c12, Cl3, Cl4; (b) Nl4, Nl5, Nl6;
(e) O1s, 011, ois; (d) Ne20, Ne21, Ne22,
14. With the aid of the data of Table 11-1, calculate the binding energy of
the last proton in each of the following sets of isotopes:
(a) cu, cu, cia; (b) N1a, N14, N15;
(c) 01•, ois, O1s, 011; (d) Ne1v, Ne20, Ne21.
15. What hypotheses would you propose to account for those values of the
binding energy in Problems 13 and 14 which are particularly small, i.e., smaller
than 5 Mev?
16. Use the values of the mass differences: neutron-proton, and Be7-Li 7 to
determine the mínimum proton energy needed to observe the reaction Li 7
(p,n)Be 7•
17. Calculate the difference between the masses of AI28 and Si 28 from the
following data: AI27 (d,p)AI28, Q = 5.48 Mev; Si28 (d,p) Si29, Q = 6.25 Mev;
Si 29(d,a)Al 27 , Q = 5.99 Mev; 2H 1 - H 2 = 1.584 millimass units; 2H 2 -
He4 = 25.600 millimass units.
18. From "the nuclear masses involved, calculate the range of energy of the
neutrons produced when Be is bombarded with 5.30 Mev a-particles from Po210 .
19. Derive Eq. (11-10).
CHAPTER 12
ARTIFICIAL RADIOACTMTY
12-1 The discovery of artificial radioactivity. The discovery by Curie

and Joliot, in 1934, that the products of sorne induced nuclear trans-
mutations are radioactive opened a new era in nuclear physics. The study
of the radioactivity of the artificially made atomic nuclei has done much
to clarify concepts of the constitution and stability of nuclei. Many of
the nuclear reactions discussed in the last chapter yield radioactive nuclides
and about 1200 radioactive species have been made up to l!l58. These
nuclides decay spontaneously according to the same laws that govern the
disintegration of the naturally occurring radioactive elements, and the
term artificial radioactivity refers to the way in which the new radio-
nuclides are produced rather than to their decay.
Artificial radioactivity was discovered by Curie and Joliot while they
were studying the effects of a-particles on the nuclei of light elements.
When boron, magnesium, and aluminum were bombarded with a-particles
from polonium, protons and neutrons were observed, as expected from the
known (a,p) and (a,n) reactions. In addition to these particles, positive elec-
trons or positrons were also observed. m The positive electron is a funda-
mental particle whose rest mass is the same as that of the electron, and
whose charge has the same magnitude but opposite sign from that of the
electron. It was discovered by Anderson< 2 > in 1932 in the course of a study
of photographs of cosmic-ray tracks in a Wilson cloud chamber. In the
presence of a magnetic field, sorne tracks were seen which could have been
caused only by particles of electronic mass and charge, but from the direc-
tion of the curvature of these tracks it was evident that the particles pro-
ducing them must have been positively charged. Anderson called these
particles positrons. Throughout the rest of this book the word electron
will be used to denote the negative electron and the word positron will be
used for the positive electron. This usage is common and more convenient
than the use of the terms "negative electron" and "positive electron."
The word "negatron" which has been suggested for the negative electron
has not been generally adopted.
The source óf the positrons observed by Curie and Joliot was not clear
at first and several hypotheses were suggested which proved to be incorrect.
For example, it was suggested that the positrons might result directly from
· the disintegration by the a-particles, with the formation of a neutron and
! ~ instea.d of the usual proton. Subsequent experiments by Curie
.aad Joli~e:howed, however, that the light-element targets continued to
283
284 ARTIFICIAL RADIOACTIVITY [CHAP. 12
emit positrons after the source of the a-particles had been removed; the
rate of emission of the positrons gradually decreased with time, finally
approaching zero. When the rate was plotted against the time after the
removal of the a-particle source, an exponential curve was obtained for
each target similar to those obtained in the ,8-decay of natural radio-
nuclides. The activity could be expressed accurately by the equation
A = A 0 e-u, and characteristic half-lives were found, namely, 14 min,
2½ min, and 3¼ min for boron, magnesium, and aluminum targets, re-
spectively. The explanation offered by Curie and Joliot was that the
product nucleus formed in the (a,n) reaction in each case was an unstable
nuclide, which then disintegrated with the emission of a positron. The
nuclear reactions when boron and aluminum were the targets were postu-
Jated to be
T = 14min,
T = 3¼min.
The symbol 1e 0 or Heº is used to denote the positron, since its charge is
the same as that of the proton, and its mass number is zero. The half-life
of N 13 was later found to be 10 min and that of Pªº 2½ min, rather than
14 min and 3¼ min, respectively, as first reported by Curie and Joliot. The
nuclides N 13 and P 30 are not stable, as can be verified from Table 9-1, but
the nuclides which remain after the emission of the positron are stable.
The explanation offered by Curie and Joliot was tested chemicalJy< 4 > by
separating the target element from the product element; radioactivity was
found only in the latter. In the boron reaction, the target, boron nitride
BN, was irradiated with a-particles for several minutes and then heated
with caustic soda, liberating ali the nitrogen as gaseous ammonia. The
ammonia was found to have ali the radioactivity. The half-life was the
same as that found with other boron targets, and was not found with
nitrogen targets. Hence, the boron must have been changed into a radio-
active isotope of nitrogen. An aluminum target, after being irradiated
with a-particles, was dissolved in hydrochloric acid, liberating any phospho-
rus present as the gas phosphine (PH 3), while the aluminum remained
behind in the solution. The solutior, was then evaporated to dryness and
the residue tes~d for positron activity. No activity was found. The gas,
which contained the phosphorus, showed the characteristic positron ac-
tivity, and the activity was therefore associated with an isotope of phos-
phorus. These tests provided the first definite chemical evidence for the
artificial transformation of one element to another. The che_~J:ii1ide~t~·
12-2] THE ARTIFICIAL RADIONUCLIDES 285
fication of the product nucleus, together with the determination of the

nature of the ejected particles by measurements of ionization in a cloud
chamber and of the deflection in a magnetic field, proved the correctness
of the equations proposed for the disintegrations.
After the discovery that the bombardment of light nuclides with a-
particles can lead to radioactive products, it was found that nuclear
reactions induced by protons, deuterons, neutrons, and photons can also
result in radioactive products. As in the case of the natural radionuclides,
an artificial radionuclide can be characterized by its half-life and by the
radiation it emits. These properties have helped greatly in determining
the chemical nature of the products of nuclear reactions. When the
products are radioactive, they can be traced in chemical reactions by means
of their characteristic half-lives or disintegration products. Otherwise
they could not be traced because of the very srnall arnounts in which they
are present, and their chemical nature could only be inferred frorn indirect
physical evidence, such as rnass and energy balances. The special branch
of chemistry that deals with these and similar problems is called radio-
chemistry, and has become an important part of nuclear science.
By radiochemical methods, the active products of nuclear reactions
can be separated in forms suitable for the study of their properties. It
is possible to determine the element with which a particular product is
isotopic, and to assign a mass number. The latter assignment often
depends on cross-bombardment, that is, the particular nuclide is obtained
by several different nuclear reactions, from which the mass number of the
given active nuclide can be inferred. When the nuclide has been obtained
in suitable form, its decay rate can be determined, as well as the propertie~
of its radiations. About 1200 artificially radioactive nuclides have been
identified and their properties determined by combined chernical and
physical techniques.
12-2 The artificial radionuclides. Electron and positron emiss1on.

Orbital electron capture. The constitution of the known nuclides, both
natural and artificial, can be shown by means of a graph in which the
number of neutrons is plotted against the number of protons, as in Fig.
12-1. The stable nuclides are indicated by black squares (as in Fig. 9-9).
Alpha-emitters are indicated by open triangles and squares; the latter are
8-stable, while the former may also undergo t3-decay. Electron-emitters
are indicated by open circles and positron emitters by crosses. A cross in
an open circle indicates that the nuclide decays by ¡3--emssion, ¡3+-emis-
sion, and electron capture. The diagonal lines are lines of constant rnass
number, A, with A = Z + N; nuclides which lie on the same diagonal
line are called isobars because they have the same mass number and,
therefore, very nearly the same atomic weight. For the most part, the
286 ARTIFIC IAL RADIOA CTIVITY [CHAP, 12
•
o {J- em1
+ (J+ emit electron capture)
140
• (J+, {J- (or electron capture)
,. }a-emitte {J-, {J+, or electron ca
o a-emitte or {J-stable)
§:
.
.,
.o
90
8
g 80
e
..,e
.,
z" 70
60
50
40
30
20
40 50 60 70 80 90 100
Proton number (Z)
Frn. 12-1. Chart of the known nuclei. (Reprinted by permission from
Liverhant, Elementary Introductionto Nuclear Reactor PhyBia. New York: S. E.
Wiley,
1960)
12-2] THE ARTIFICIAL RADIONUCLIDES 287
artíficially radioactive nuclídes lie above or below the stable nuclides;

they have either too many neutrons (not enough protons) or too few
neutrons (too many protons) to be stable. A nucleus with too few pro-
tons tends to become stable by increasing the nuclear charge, while a
nucleus with too many protons tends to become stable by decreasing the
nuclear charge. One way by which the nuclear charge can be increased is
by electron emission, as was seen during the treatment of natural radio-
activity. The nuclear charge can be decreased by emíssion of a positron.
Consequently, positron activity is usually seen in the isotopes of an ele-
ment having mass numbers smaller than those of the stable isotopes, while
electron activity usually occurs in isotopes with mass numbers greater
than those of the stable isotopes.
The charge on the nucleus is also reduced if the nucleus captures an
orbital electron. The extranuclear electrons, in the course of their motions,
often approach close to the nucleus and, according to wave mechanics, may
even penetrate it. The electrons which are most likely to do this are the
K-electrons in atoms with a large value of Z. These electrons may be
thought of as available for capture by the nucleus, especially if the latter
is proton-rich. If an electron from the K-shell is captured, the process is
called K-capture; less often an L-electron is captured. The vacancy in the
K-shell or L-shell is usually filled by an electron from an outer shell with
the emission of a K or L x-ray, respectively, characteristic of the product
nucleus. In this process, no charged particle is emitted and the process
can be observed only because of the x-ray emission. Orbital electron
capture is sometimes accompanied by the emission of electrons from the
extranuclear structure. These electrons, called A uger electrons after their
discoverer, result from what might be described as an internal photo- -
electric effect. For example, the emission of a K x-ray may be replaced by
the ejection of an L-electron with a kinetic energy equal to the difference
between the K x-ray energy and the binding energy of the L-electron.
The three types of disintegration are isobaric transformations because they
involve changes in nuclear charge but not in mass number.
As in the case of natural radioactivity, sorne artificially radioactive
nuclides emit 'Y-rays as well as electrons or positrons. About one-third of
the active nuclides are simple in the sense that they give off only particles;
the remainder also emit 'Y-rays.
The question of whether an artificial nuclide will decay by electron
emission, positron emission, or orbital electron capture can be discussed in
terms of the energy available for the disintegration. Consider first the
case of electron emission, in which the reaction is
(12-1)
The energy balance for this reaction is given by the expression
(12-2)
where the M n's represent the nuclear masses oí the artificial nuclide and its
decay product, and m is the electron mass. The last equation can be
written in terms oí the atomic masses ií it is noted that
where M a is the atomic mass; the binding energy oí the electrons is then
included but leads to no difficulty because it always cancels. Then Eq.
(12-2) becomes
Q/c 2 = M 0 (zXA) - Zm - Ma(z+iYA) + (Z + l)m - m

(12-3)
The condition that electron emission be energetically possible is that the

value of Q be positive, or
(12-4)
i.e., the atomic mass of the artificial nuclide must be greater than that of its
isobar with nuclear charge one unit greater.
In the case of positron emission, the reaction is
(12-5)
and the Q-value is given by
Q/c 2 = M,.(zXA) - M,.(z- 1YA) - m,

or,
Q/c 2 = Ma(zXA) - Zm - Ma(z-1YA) + (Z - l)m - m
= Ma(zXA) - Ma(z-1YA) - 2m. (12--6)
For positron emission to be energetically possible, the atomic mass of the

artificial nuclide must be greater than the atomic mass of its isobar with
nuclear charge one unit smaller by at least 2 electron masses, or 1.02 Mev,
(12-7)
The reaction in the case of orbital electron capture is
(12-8)
12-2) THE ARTIFICIAL RADIONUCLIDES 289
and the energy available is
(12-9)
The condition that orbital electron capture be energetically possible is

therefore
(12-10)
i.e., the mass of the artificial nuclide must be greater than that of its isobar
with nuclear charge one unit smaller.
Comparison of Eqs. (12-7) and (12-10) shows that the capture of an
electron by a nucleus is an energetically more favorable process than the
emission of a positron. But the capture of even a K-electron depends on
the small probability that the electron be very close to, or "within," the
nucleus. The result is that as soon as the energy available exceeds 2mc 2
positron emission tends to happen more often than orbital electron capture.
Hence, K-capture has been observed only when the condition (12-10) is
satisfied, but it is not always observed when that condition is fulfilled.
There are many examples of the above rules. Thus, the nuclides C 11 ,
N 13 , 0 15 , F 17 , Ne 1 9 , Na 2 1, Mg 23 , and Al 25 all decay by positron emission.
Each has a mass number smaller than those of the stable isotopes of the
element, and is consequently proton-rich. In each case, the mass of the
radioactive nuclide is greater than that of the decay product by more than
2 electron masses. For example, according to Table 11-1, C 11 has an
r1,tomic mass of 11.014924 amu, and B 11 has a mass of 11.012795 amu; the
difference is 0.002129 amu, considerably greater than 2 electron masses,
which are equal to 0.001098 amu. The nuclides C 14 , N 16 , F 2 º, Ne 23 , Na 24,
Mg 27, and Al 28 decay by electron emission. Each has a mass number
greater than those of the stable isotopes of the element and is therefore
neutron-rich. It can be seen from Table 11-1 that, in each case, the mass
of the active nuclide is greater than that of the product. For example, the
mass of C 14 is 14.0076845 amu, while that of N 14 is 14.0075179 amu.
The isotopes of an element of intermediate atomic weight, such as iodine,
show all three types of decay. lodine has only one stable species, 53 1127 ,
containing 53 protons and 74 neutrons. There are, however, 17 artificially
radioactive isotopes, containing from 68 to 86 neutrons. Of these, 1121 and
1 122 , with the smallest number of neutrons, decay by positron emission.
The 1 123 nucleus captures an orbital electron, while 1 124 sometimes cap-
tures an orbital electron and sometimes emits a positron; l 125 captures an
orbital electron; l 126 sometimes captures an orbital electron and sorne-
times emits a negative electron; 1 127 is stable; 1 128 decays by emission of
an electron 95% of the time and either by orbital electron capture or
positron emission 5% of the time. The eleven remaining neutron-rich
iodine isotopes, from 1129 to 1139 inclusive, decay by the emission of an

electron. The decay products oí sorne of the iodine isotopes are also
unstable and decay in turn; 1 133 decays into Xe 133 , which is unstable and
decays by electron emission into stable Cs 133 .
There is sorne correlation between the radioactivity of an artificial
nuclide and the transformation by meaos of which it is produced. Elec-
tron emission is common for activities produced by (n,'Y), (n,p), (n,a), and
(d,p) reactions, since these reactions decrease the charge-to-mass ratio.
Examples are
49ln 115 (n, 'Y)49ln 116 i 49ln 116 ----+ soSn 116 + -1e0, T = 13 sec,
1N 14 (n, p)oC 14; oC 14
----+ 1N 14 + -1eº, T = 5568y,
27
13Al (n, a)11Na ;
24 11Na 24 ----+ 12Mg 24 + -1e 0, T = 15.Oh,
1sP 31 (d, PhsP 32 ; u;P 32 ----+ 108
32
+ -1eº, T = 14.3d.
Positron emission is common for activities produced by (p,'Y), (p,n),

(a,n), (d,n), and ('Y,n) reactions, which increase the charge-to-mass ratio
of the nucleus. Examples are
13
oC1ª(p, 'Y)¡N13; 1N 13 ----+ + 1e0 ,
oC T = lOmin,
58 58
28Ni (p, n)2 9Cu ; · 29Cu58 ----+ 2sNi
58 + 1e0 , T = 2.6 sec,
7N14(a, n)9F17; 9F11 ----+ s011 + ieº, T = 70sec,
14 15 sÜ 15 ----+ 1N 15 + 1e 0, =
7N (d, n)s0 ; T 2.lmin,
15p3I('Y, n)¡sPªº; 11iP 30 ----+ uS 30 + 1eº, T = 2.5min.
The (a,p), (p,a), and (d,a) reactions usually lead to stable products. The
charge and mass changes are such that the target and product nuclei are
generally both within the band oí stable nuclides, as in the examples
N14(a,p)011, A121(a,p)Siªº;
and
12-3 The transuranium elements. The most íruitful source of artificial

radionuclides is the radiative capture of neutrons, <si the (n,'Y) reaction,
which usually yields an electron-emitting product. In 1934, Fermi< 6 > sug-
gested the possibility that the bombardment of uranium with neutrons
might result in the production of elements with nuclear charge greater
12-3] THE TRANBURANIU!lf ELEMENTB 291
than 92. If U 238 were to capture a neutron, the following reaction should
occur:
if the U 239 were then to decay by electron emission, the result would be
a nuclide with Z = 93, an isotope of a· hitherto unknown element. Early
experiments showed that when uranium was bombarded with neutrons,
four different ,B-activities were detected, one of which was presumably
associated with 92 U 239 . It was also considered possible that the new ele-
ment with Z = 93 might decay by electron emission to form an isotope
of another new element with Z = 94. Any elements which might have
values of Z greater than 92'were called transuranium elements because they
would lie beyond uranium in the periodic table. Between 19;34 and 1939,
many attempts were made to make and identify transuranium elements,
but although sorne doubtless were produced, there was difficulty with
their identification. The research on these elements led to the discovery ··
of nuclear fission, and the work which followed this discovery resulted,
lllll.ong other things, in the systematic production and study of the trans-
uranium elements.
In 1940, it was proved that the bombardment of uranium with slow
D11Utnms does indeed produce a new isotope U 239 which has a half-life of
23 aún and is an electron emitter. It decays into an isotope of a new
......t with Z = 93 which, in turn, emits an electron and has a half-life
.eftÍdays. <7> The element with Z = 93 was given the name neptunium .
.•~\ 1'e •w
239
nuclide, Np , results from the decay .
. \'}¡.. 92U239 -+ 93Np239 + -1eº, T = 23 mm.
~ · of Np 239 yields an isotope of a new element with Z = 94, called
.~(8,9).,,
\ 93Np 239 -+ 94Pu 239 + -1eº, T = 2.3 d .
.~:::;- This ~ of plutonium is manufactured in large amounts m nuclear
"-:·,..
=:\.,It
reaet411'11 and can be used for the production of nuclear power or atomic
is an a-emitter with a half-life of 24,400 years. Soon after the
'()f 93Np 239 , two other isotopes of neptunium were made. When
•-~ =1'·.·· º. xide is bombarded with fast deuterons from a cyclotron, the
f4111 .. · Jd,2n) reaction takes place:
¡< . - 1_
t .f u2U 238 + 1H 2 -+ 93Np 238 + on 1 + on 1•
f
"""~·"t.The,product Np 238 emits an electron to form Pu 238 , and has a half-life of

,,., .... _ {
2.0 days. When uranium is exposed to fast neutrons, an (n,2n) reaction

occurs,
The isotope 92 U 237 is an electron emitter with a half-life of 6.8 days, and
decays into Np 237 . The latter is an a-emitter with a half-life of 2.2 X 10 6
years, and is the longest-lived of the known transuranium nuclides. Iso-
topes of neptunium can also be obtained by bombarding uranium with
deuterons with energies of about 100 Mev; reactions occur in which 4, 6,
or even 9 neutrons are ejected. Alpha-particles can also be used. Sorne of
the isotopes and the reactions by which they are made are
Np 231 : U 233 (d,4n); U 235 (d,6n); U 238 (d,9n),

Np2aa: u2as(d,n); u2as(a,p2n); u2as(d,4n),
Np 238 : U 238 (d,2n); U 238 (a,p3n).
In all, 10 isotopes of neptunium are now known, with mass numbers fl'Olll
231 to 240.
Isotopes of elements 95 (americium) and 96 (curium) can be made by
bombarding Pu 239 with neutrons. The reactions are
Pu239 + n -+ Pu240 + 'Y,
Pu240 + n -+ Pu2u + 'Y,
94Pu 241 ~ usAm 241 .
At the same time, 96Cm 242 is formed,

Am241 + n -+ Am242 + 'Y
,,-
usAm 242 -+ 115Cm 242 -
The nuclide Am 241 , on bombardment with a-particles, gives an i8otope

of element 97 (berkelium),
241
11sAm + 2He 4 -+ 97Bk 243 + 2n.
These reactions were used in the discovery of berkelium, 00> and aimilar
reactions were used in the first preparation and identification 411 an ~pe
of californium, 98 Cf 244 . (l ll The new nuclides are identified, i.e., &heir
Z- and A-values are determined, by separating them chemically and ·
studying their radioactive properties together with those of their decay
products. Thus, 117 Bk 243 decays by orbital electron capture to 116CJ\n 243
which decays to Pu 239 by a-emission; Bk 243 also emits an a-particle (, vith
12-3) THE TRANSURANIUM ELEMENTS 293
Fm254
100
E253
§: 99 Cf250
....
~"' 98
s:::, 97
A
Cm244
·e.s" 96 Cm249
< 95
Pu239
94 Pu243
240 242 244 246 248 250 252 254

Mass number (A)
F10. 12-2. Nuclear reaction sequences for production of heavy nuclides by

intense slow neutron irradiation of Pu 239 • (From Seaborg, The Transuranium
Elements, gen. ref.).
1.0
s
.s
,,,°' 0.8 ·--·-:;
~
p...
Pu239
-¡;
:10.6
....
"'P.
UJ
..,§ 0.4
't5
....
'a"'
:::,
0.2
z
o
(l 2 3 4 5 &
Years
239 at a
Frn. 12-3. Production of sorne heavy nuclides by irradiation of Pu
flux of 3 X 10 14 neutrons/cm 2 /sec (from Seaborg, The Transuranium Ele-
ments, gen. ref.).
a branching ratio of less than one percent) to give 95 Am 239 which captures
an orbital electron to become Pu 239 . Since the radioactive properties of
Pu 239 are known and it can be shown how the Pu 239 is formed, the identity
of 97 Bk 243 can be established.
The bombardment of Pu 239 with neutrons can be carried much further
than has been indicated, and many transuranium nuclides which are ,8-
stable or decay by negative ,8-emission can be built up through the absorp-
tion of successive neutrons. If the Pu 239 is irradiated in a nuclear reactor
with a high neutron flux, i.e., 3 X 10 14 neutrons/cm 2/sec (see Section
18-4), it is possible to build up, in the course of severa! years, appreciable
yields of nuclides which correspond to the absorption of as many as 15
neutrons. The sequences of neutron capture reactions and ,8-decays leading
to nuclides as heavy as 100 Fm 254 are shown in Fig. 12-2; the horizontal
arrows indicate neutron capture, and the vertical arrows ,8-decay. Still
more extensive reactions have been found to occur in a thermonuclear
explosion and elements 99 and 100 were actually discovered in debris from
an explosion in November 1952. Many isotopes of elements 95 to 100
have been identified as products of Pu 239 irradiated either in a nuclear
reactor or an explosion. 02- 20>
The rate of buildup of higher nuclides can be analyzed mathematically
by means of equations analogous to the Bateman equations of natural
radioactivity.< 21> Sorne results for the irradiation of Pu 239 in a neutron
flux of 3 X 10 14 neutrons/cm 2/sec are shown in Fig. 12-3.
It is also possible to produce isotopes of transuranium elements by
bombarding urEtnium with highly ionized heavy nuclei such as C 12 , N 14 , or
0 16. It has been shown,< 22 ,23 > for example, that C 12 (6+) or N 14 (6+)
accelerated to energies in the neighborhood of 100 Mev, can induce reac-
tions in which the atomic number is increased by 6 or more units. Isotopes
of californium, einsteinium and fermium have been produced by the
bombardment of uranium with carbon, nitrogen and oxygen ions, respec-
tively. <24 •25 ·26 > Among the reactions which occur are
92U 238 + sC 12 ---. 9sCf 244 + 6n,

92U2as + 7N1~ _. ;;¡;E246 + 6n,
92U2as + ~01& -'-+ 100Fm2so + 4n.
An isotope of element 101 (mendelevium) has been prepared by bom-

barding a sample of 99 E 253 with a-particles; the product was identified
by means of its decay products as 101 Mv 256 . <27 > An isotope of element
102 (as yet unnamed) has been made by bombarding a mixture of Cm 244
and Cm 246 with C 12 atoros at energies of 60-100 Mev; analysis of the
products indicates that 102 254 was produced. <28 •29 • 30> The production
12-4) THE ARTIFICIAL RADIONUCLIDES: ALPHA-EMITTERS 295
of a second isotope of element 102 by bombardment of a mixture of curium

isotopes with C 13 nuclei has also been reported. ca o
Work on the transuranium nuclides has so far resulted in the identifica-
tion of 10 isotopes of neptunium (Z = 93), 15 isotopes of plutonium
(Z = 94), 10 isotopes of americium (Z = 95), 13 isotopes of curium
(Z = 96), 8 isotopes of berkelium (Z = 97), 11 isotopes of californium
(Z = 98), 10 isotopes of einsteinium (Z = 99), 8 isotopes of fermium
(Z = 100), one species of mendelevium (Z = 101) and one (or perhaps 2)
species of element 102.
12-4 The artificial radionuclides : alpha-emitters. It was mentioned

in the last section that sorne of the isotopes of the transuranium elements
decay by emission of an a-particle. In 1939, only 24 a-emitters were
known and these were members of the naturally occurring radioactive
series. At the present time, about 150 a-decaying nuclear species are
known. <32 > About 50 of these are among the transuranium elements;
others forro part of a new radioactive series, the 4n + 1, or neptunium,
series. There are several so-called coUateral series, and additional a-
emitters are found among the neutron-deficient isotopes of bismuth
(Z = 83), polonium (Z = 84), astatine (Z = 85), emanation (Z = 86),
and francium (Z = 87).
33 34
The members of the neptunium series< • > and their properties are
listed in Table 12-1, and the relationships between the members are
241 _/ v:; ~
V,.,,
237 .L. t,,-
_/
J!.
.--
233
_/
~
] 229 .,,,
,,,...,,,
§ ~
V
e: 225 ., ~
B v .... .--
~ 221
_/
217 .,v
,,,, --
.--
213
)----: '?"
209 .I!!. :::. .e: Vf--
Stable
TI Pb Bi Po At Em Fr Ra Ac Th Pa U Np Pu Am
fil ~ ~ ~ ~ ~ m ~ ~ 00 ITT ~ W M %
Atomic number
Frn. 12-4. The neptunium (4n + 1) series.
296 ARTIFICIAL RADIOACTIVITY [CHAP, 12
TABLE 12-1
THE N EPI'UNIUll SERIES
disintegration particle
Type of
Nuclide Half-life constant, energy,
disintegration -1
sec Mev
94
p U 241 fl, a 13.Oy 1.69 X 10-9 a: 4.893m
fl: 0.021
95A m241 a 458y 4.80X 10-11 5.535m
u231 6.75d 1.19 X 10-6 0.248
92 fJ
N 2a1
93 P a
6
2.20X 10 y 9.99X 10- 16 4.870
p 233 27.Od 2.97 X 10- 7 O.568m
91 a fJ
u233 5 1.36 X 10-13
92 a 1.62 X 10 y 4.816
90Th229 a 7340y 3.OX 10- 12 5.02m
88
Ra225
fJ 14.8 d 5.42 X 10-7 0.32
89 Ae226 a 1O.Od 8.02X 10- 7 5.818 m
87 F r 221 a 4.8m 2.41 X 10-3 6.33m
85At211 0.018 s 1
a 3.85 X 10 7.05
83 B.21a
1 fl, a 47m 2.46 X 10-" a: 5.86
fl: 1.39
p 213 4.2X 10-6 8 5
84 O a 1.65 X 10 8.35
9 2.2m 5.25X 10-3
81Tfº fJ 1.99
209 5.84X 10-5
82Pb fJ 3.30h 0.635
B.209 Stable
83 1
shown graphically in Fig. 12-4. Like the naturally occurring radioactive

series, the neptunium series shows branched disintegration near the end.
The stable end product is the ordinary bismuth species of mass number 209,
rather than an isotope _of lead as in the uranium, thorium, and actinium
series. The half-life of Np 237, the longest-lived member of the series, is
2.2 X 106 years. lt is generally considered that the earth is about 5 X 109
years old. If it is assumed, as is probable, that neptunium was formed at
the same time as the earth, then many half-lives have elapsed for this
nuclide and the amounts still present would be so minute as to be beyond
the possibility of detection. The absence of a naturally occurring 4n + 1
series can therefore be understood; even ü such a series did exist at one
time, its members would long since have decayed to Bi 2 º9
12-5) ISOTOPE TABLES AND NUCLIDE CHARTS 297
In addition to the neptunium series, artificial a-emitters have been

made which form chains collateral to the heavy radioactive series. <35>
These chains include series collateral to each of the natural radioactive
series, the uranium (4n + 2), thorium (4n), and actinium (4n + 3), as
well as the artificial neptunium (4n + 1) series. Asan example of such a
series, consider the decay of 91 Pa 227 , formed by bombarding thorium with
80-Mev deuterons¡ the reaction is Th 232 (d,7n)Pa 227 . The nuclide Pa 227
decays in the following way:
91Pa221 ~ s9Ac22a ~ ~ ssAt215 ~
s1Fr219
T-38.3 m 2.2 m 0.025 • 10-• s
211 (AcC) ~ 207 207 (stable).
83 Bi 81 Tl (AcC") _!_. Pb
2.16m 4.79m
If this series is compared with the actinium (4n + 3) series in Fig. 10-11,
it is seen that the new series runs parallel to part of the actinium series, and
this is the reason for the name collateral series. There is a series starting
with U227 which is also collateral to the actinium series; this chain decays
by successive a-emissions as follows:
u221 ___. Th22a ___. Ra219 ___. Em21s ___. p 0 211 ___. Pb201_
The other collateral series are

Pa 226 ---. Ac 222 ---. Fr 218 ---. At 214 (4n + 2 mass type),
Pa22s ___. Ac224 ___. Fr220 ___. At216(4n mass type),
u22s ___. Th224 ___. Ra220 ___. Em216(4n mass type),
u229 ___. Th22a-> Ra221 ___. Em211(4n + 1 mass type).
12-5 Isotope tables and nuclide charts. The vast amount of informa-
tion which has been collected about the diffrrent nuclear species, both
stable and unstable, makes it necessary to have sorne convenient methods
for recording nuclear data. Tables of isotopes are published periodically,
such as those cited in the general references at the end of this chapter, and
compilations of nuclear data are available Iike that prepared by the
National Bureau of Standards. The most convenient method for quick
reference is that of the nuclide chart, a portion of which is shown in Fig.
12-5. In this type of chart the atomic number Z is plotted against the
number of neutrons, A - Z. Ea.ch nuclide occupies a square, and stable,
naturally radioactive, and artificial nuclides are differentiated by the
color or shading of the square. The symbol and mass number are shown
in each case, as well as the abundance of the isotope if it is stable, the
half-Iife, the type or types of decay, and the energy of the emitted particles
P 28 P32' P 33 1,,:)
0.28, c:o
151 ~+11,~1...
14.5 d 25 d 1 00
-+ .25
-y 1.11, 2.6-7.6
"" ....
"º
SI 26 SI 27 SI 31 SI 32
1.7. 4.4.
.
2.62 h ~300y
141 ~ ~u
..... /r.1
•T
.,
Al 24 Al 28 Al 29 17 18
2.1, 2.30n,
13( ,+<u,1.1
6.6"' >
Tl.Jl-7.2
¡r 2.17
l' 1:,1
/r:U,1.A
Yl.21,2-'3
~
...
i.~21 "" U.65 111 Dlspla-=-ts caused by
artlflclal disintegratlons
...
"'l
n
MI 23 Mg 27 Mg 28 s:t"
12, 9.5'" 21.3 h
ª• 3n a,2n a,n
12 I ~,.o
u.o
/rl:75, 1.57
T.M.1.112,.11
,-_,.. T.G22, ~t:,
1.35,.95,M
~
11.12
p,n p, 'Y a,np

Na 22 Na 24 Na25 15 16 d,n ~
...
2.6y 15.0 h 601
11 I ~~•••. ,I /rl.39, .••
:i
z
')' 1.21
UM
'Y 2.75', 1.361,-
15.31
-- '-0, 2.6-3.A
T.99,.51..31,
uo l<I.O f
z
'Y, n
n. 2n
Original
nudeus
n. 'Y
d, P
~
11 13 14
A-Z-
~ .. 1
'Y, p 1 n,p
n. a
ñ
=
A-z- ~
Frn. 12-5. Portion of the nuclide chart. .,.,...
REFERENCES 299
and 'Y-rays. In sorne charts, the atomic mass and other data are also in-
cluded, such as the nuclear spin and magnetic moment. Sometimes all
nuclides on the same horizontal line have the same atomic number, and
nuclei with the same mass number (isohars) lie on a 45° diagonal line
running from upper left to lower right. The example of Fig. 12-5 is of this
type. A complete nuclide chart is inserted at the back of the book. In
other charts, the isotopes lie along a diagonal line and isobars lie in the
same vertical column.
The nuclide chart can be used to obtain quickly information about the
products of artificial nuclear disintegrations. The displacements in charge
and mass caused by various disintegrations are usually shown as in the
lower right-hand corner of Fig. 12-5. It can be seen immediately that the
product of an (a,n) reaction on Al 27 is P 30, which is a positron emitter with
a half-life of 2.5 min. The chart gives the energy of the swiftest posi-
trons emitted, which is, in this case, 3.3 Mev; no 'Y-rays are emitted.
Similarly, the effect of a (d,p) reaction on Al 27 is seen immediately to be
Al 28, which has a half-life of 2.3 min, and emits electrons with a maximum
energy of 2.87 Mev, along with 1.78-Mev 'Y-rays.
The information shown in the squares of Fig. 12-5 has been limited to
keep the figure clear. Even with the information shown, it should be
apparent that the nuclide chart is very useful to anyone who needs easily
available data about atomic nuclei.
The discussion of artificial radioactivity in this chapter is far from
complete. Only the most common modes of decay have been discussed.
One reason is that further discussion of the artificial nuclides and the infor-
mation they give about nuclear stahility and nuclear structure will require
knowledge of the properties of the radiations from radioactive substances.
These properties will be treated in the next three chapters.
REFERENCES
GENERAL
M.S. LIVINGSTON and H. A. BETHE, "Nuclear Physics, C. Nuclear Dynamics,
Experimenta.!," Revs. Mod. Phys. 9, 245--390 (July 1937).
STROMINGER, HoLLANDER, and SEABORG, "Table of Isotopes," Revs. Mod.
Phys. 30, 585 (1958).
G. T. SEABORG, The Tramuranium Elements. New Haven, Conn.: Yale Univer-
sity Press; Reading, Mass.: Addison-Wesley, 1958.
E. K. HYDE and G. T. SEABORG, "The Transuranium Elements," Handbuch
der Physik, Vol. 43, p. 205, Berlin: Springer Verlag (1958).
PARTICULAR
l. l. Cua1E and F. JOLIOT, "Electrons Produced by Artificial Disintegration,"
Compt. rend. 196, 1885 (1933); 198, 254 (1934); J. phys. et radium 4, 494 {1933).
2. C. D. ANDERSON, "The Positive Electron," Science 76, 238 (1932); Phy,.

Rev. 43, 491 (1933).
3. l. CURIE and F. JoLIOT, "A New Type of Radioactivity," Compt. rend. 198,
254 (1934); Nature 133, 201 (1934); J. phy,. et radium 5, 153 (1934).
4. l. CURIE and F. JoLIOT, "Chemical Separation of New Radioactive Ele-
ments," Compt. rend. 198, 559 (1934).
5. FERMI, AMALDI, D'AaosTINO, RAsETTI, and SEGRE, "Artificial Radio-
activity Produced by Neutron Bombardment," Proc. Roy. Soc. (London) Al46,
483 (1934).
6. E. FERMI, "Possible Production of Elements of Atomic N umber Higher
than 92," Nature 133, 898 (1934).
7. E.M. McMILLAN and P. H. ABELSON, "Radioactive Element 93," Phya.
Rev. 57, 1185 (1940).
8. SEABORG, McMILLAN, KENNEDY, and WAHL, "Radioactive Element 94
from Deuterons on Uranium," Phya. Rev. 69, 366 (1946).
9. KENNEDY, SEABORG, SEGRE, and WAHL, "Properties of 94 (239)," Phya.
Rev. 70, 555 (1946).
10. THOMPSON, GmoRsO, and SEABORG, "The New Element Berkelium
(Atomic Number 97)," Phya. Rev. 80, 781 (1950).
11. TH0MPSON, STREET, ÜHIORso, and SEABORG, "The New Element Cali-
fornium (Atomic Number 98)," Phys. Rev. 80, 790 (1950).
12. THOMPSON, GHIORso, HARVEY, and CHOPPIN, "Transcurium Isotopes Pro-
duced in the Neutron Irradiation of Plutonium," Phys. Rev. 93, 908 (1954).
13. DIAMOND, MAGNUSSON, MECH, 8TEVENS, FRIEDMAN, 8TUDIER, FIELDS,
and Hu1zENGA, "ldentification of Californium Isotopes 249, 250, 251, and 252
from Pile-Irradiated Plutonium," Phys. Rev. 94, 1083 (1954).
14. CHOPPIN, THOMPSON, GHIORso, and HARVEY, "Nuclear Properties of
sorne Isotopes of Californium, Elements 99 and 100," Phys. Rev. 94, 1080
(1954).
15. GmoRso, THOMPSON, CHOPPIN, and HARVEY, "New Isotopes of Ameri-
cium, Berkelium, and Californium," Phys. Rev. 94, 1081 (1954).
16. STUDIER, FIELDS, DIAMOND, MECH, FRIEDMAN, SELLERS, PYLE, STEVENS,
MAGNussoN, and HUIZENGA, "Elements 99 and 100 from Pile-Irradiated Plu-
tonium," Phys. Rev. 93, 1428 (1954).
17. GHIORSO, THOMPSON, HIGGINS, 8EABORG, 8TUDIER, FIELDS, FRIED,
DIAMOND, MEcH, PYLE, HUIZENGA, HIRSCH, MANNING, SMITH, and SPENCE,
"New Elements 99 and 100," Phys. Rev. 99, 1048 (1955).
18. JoNES, ScHUMAN, BuTLER, CowPER, EAsTwooo, and JACKSON, "Isotopes
of Einsteinium and Fermium Produced by Neutron Irradiation of Plutonium,"
Phys. Rev. 102, 203 (1956).
19. EASTWOOD, BuTLER, CABELL, JACKSON, SCHUMAN, RoURKE and COLLINS,
"lsotopes of Berkelium and Californium Produced by Neutron Irradiation of
Plutonium," Phys. Rev. 107, 1635 (1957).
20. FIELDS, 8TUDIER, DIAMOND, MECH, INGHRAM, PYLE, 8TEVENS, FRJED,
MANNING, GmoRSo, THOMPSON, HIGGINS, and SEABORG, "Transplutonium
Elements in Thermonuclear Test Debris," Phys. Rev. 102, 180 (1956).
PROBLEMS 301
21. W. RuBINSON, "The Equa.tions of Ra.dioa.ctive Tra.nsforma.tion in a.

Neutron Flux," J. Chem. Phys. 17, 542 (1949).
22. M1LLER, HA!IIILTON, PUTNAM, HAYMOND, a.nd Ross1, "Acceleration of
Stripped C 12 a.nd C 13 Nuclei in the Cyclotron," Phys. Rev. 80, 486 (1950).
23. Ross1, JoNES, HoLLANDER, a.nd HA!IIILTON, "The Acceleration of Nitro-
gen-14 (6-t) Ions in a 60-inch Cyclotron," Phys. Rev. 93, 256 (1954).
24. GHIORSO, THOMPSON, STREET, and SEABORG, "Californium Isotopes from
Bombardment of Ura.nium with Carbon Atoros," Phys. Rev. 81, 154 (1951).
25. GHroRSO, Ross1, HARVEY, and THOMPSON, "Reactions of U 238 with
Cyclotron-Produced Nitrogen lona," Phys. Rev. 93, 257 (1954).
26. ATTERLING, FoRsLING, HoLM, MELANDER, and AsTROlll, "Element 100
Produced by Mea.ns of Cyclotron-Accelerated Oxygen Ions," Phys. Rev. 95,
585 (1954}.
27. GHIORSO, HARVEY, CHOPPIN, THOMPSON, and SEABORG, "New Element
Mendelevium, Atomic Number 101," Phys. Rev. 98, 1518 (1955).
28. GHIORso, S1KKELAND, WALTON, and SEABORG, "Element No. 102,"
Phys. Rev. Letters. 1, 18 (1958).
29. A. GHIORSO and T. SIKKELAND, "Hea.vy Ion Rea.ctions with Heavy Ele-
ments," Proceedinga of the Second International Conference on the Peaceful Uses
of Atomic Energy, Geneva, 1958, P /2440, 14, 158 (1958).
30. G. N. FLEROV, "Heavy Ion Reactions," see Ref. 29, P/2299, 14, 151
(1958).
31. FIELDS, FRIED!IIAN, MII.STED, ATTERLING, FoRSLING, HoLM and AsTRoM,
"On the Production of Element 102," Phys. Rev. 107, 1460 (1957); Arki11. f.
Physik. 15, 225 (1959).
32. STROMINGER, HoLLANDER, and SEABORG, "Table of lsotopes," Revs.
Mod. Phys. 30, 585 (1958).
33. HAGEMANN, KATZIN, STUDIER, GHIORso, and SEABORG, "The (4n 1)+
Ra.dioactive Series: The Decay Product of U 233 ," Phys. Rev. 72, 252 (1947).
34. ENGLISH, CRANSHAW, DEMERS, HARVEY, HINCKS, JELLEY, and MAY,
"The (4n + 1) Radioactive Series," Phys. Rev. 72, 253 (1947).
35. MEINKE, GHIORSO, and SEABORG, "Artificial Chains Collateral to the
Heavy Ra.dioactive Families," Phya. Rev. 81, 782 (1951); Phys. Rev. 85, 429
(1952).
PROBLEMS
l. Deduce, from the data in the table of stable nuclides (Table 9-1), which
of the following nuclides are electron emitters and which are positron emitters:
Ga.73 1 Nb96 1 c81211 Ir 197, Aul9S 1 Br7s, y4s, Sc47, Aguo, xe1a1, xe12a, zn63.
Check your results by comparison with a nuclide chart.
2. Show, from the masses involved (Ta.ble 11-1), which of the following
nuclides are stable, which are electron emitters, and which are positron emitters:
Na.23 , p 32, Si31 , Mg 27, Na.22, F 18, Be 10, He 6. Check your results by comparison
with a nuclide chart.
302 ARTIFICIAL RADIOACTIVITY [CHAJ'. 12
3. Calculate the Q-value in Mev for each of the following nuclear changes:
He6~-)Li6; c14~-)Nl4; Nl3~+)Cl3; Fl8~+)Ql8; F20~-)Ne20; Na22~+)
Ne22; Na24~-)Mg24; Af28(~-)Si2s.
4. Derive a condition for the occurrence of a-decay analogous to those for
electron emission and positron emission.
5. A sample of manganese is bombarded for 20 hr with deuterons in a cyclo-
tron under conditions such that 5 X 108 atoros of Mn 56 are formed per second
as a result of the Mn 55 (d,p)Mn 56 reaction. The Mn 56 is an electron emitter
with a half-Iife of 2.58 hr. Plot the number of Mn 611 atoros present in the sample
as a function of time, from the time the cyclotron is turned on until 40 hr Iater.
What would be the number of atoms of Mn 56 if secular equilibrium were attained?
What fraction of this number is actually reached?
6. A sample of gold is exposed to a neutron beam with an intensity such
that 10 10 neutrons are absorbed per second because of the reaction Au 197 (n,'Y)
Au 19S. The nuclide Au 198 decays by electron emission with a half-life of 2.70
days. How many atoros of Au 198 will be present after 100 hr? After 10 days?
How many atoms of Hg 198 will be present at these times if it is assumed that
the neutron beam does not affect this nuclide?
7. Under the eonditions of Problem 6, how long would the irradiation have
to be continued until the number of Au 198 atoros reaches 95% of the value
at secular equilibrium? Suppose that the sample is exposed for this length
of time and then removed from the neutron beam. Plot the value of the number
of atoms of Au 198 during the period of the irradiation, and for 15 days there-
after.
8. A sample of iron is bombarded with deuterons in a cyclotron, and a radio-
nuclide with a half-life of 46 days is obtained. Chemieal analysis shows that
the nuclide is an isotope of iron. When a eobalt target is bombarded with
neutrons of moderate energy, the 46-day activity is observed again. Which
isotope of iron is the nuclide?
9. When a zinc target is bombarded with a-particles, a new nuclide is formed
as the resul,t of an (a,n) reaction. This nuclide is a positron emitter with a
half-life of 1.65 days. The same activity is found as a result of a (d,2n) re-
action on gallium, and as a result of an (n,2n) reaction on germanium. The
(d,p) reaction on the germanium isotope of mass number 70 gives a nuclide
with a half-life of 11.4 days. Identify the unknown nuclide.
10. The following atomic masses are given.
Fluorine Neon Sodium A
20.006341 19.998765 20.015236 20

21.006840 21.000494 21.004281 21
Which of these nuclides should decay by ~--emission? By ~+-emission? By

orbital electron capture? Explain your answers and compare them with data
in the "Table of lsotopes" of Strominger, Hollander, and Seaborg.
PROBLEMS 303
11. The following atomic masses are given.
Sulfur Chlorine Argon A

--
34.98014 34.97996 34.98572 35
35.97844 35.97974 35.97892 36
Which of these nuclides should decay by ,g--emission? By ,g+-emission? By

orbital electron capture? Explain your answers, and compare them with data in
the "Table of Isotopes" of Strominger, Hollander, and Seaborg.
12. The following atomic masses are given: A40 : 39.975050; K 4 º: 39.976653;
Ca 40 : 39.975230; Sc 40 : 39.990250. Which of these nuclides would you expect
to be radioactive and how would you expect'them to decay?
13. The nuclide Cu 64 with a half-life of 12.8 hr decays by electron emission
(39%), positron emission (19%) and orbital electron capture (42%). Calculate
the partial disintegration constant and half-life for each mode of decay.
14. One gram of potassium emita 29 ,g--particles/sec owing to the decay of
K 40 , which has an abundance ratio of 0.012 atom percent. Gamma-raya are
also emitted, and the ratio of gammas to betas is 0.12. The 'Y-rays follow electron
capture in K 40 , and one photon is emitted for each orbital capture. What is the
half-life of K 4 º?
15. Explain, with the aid of Q-values, why He5, Li 5 , and Be8 are not found
in nature.
16. Show that Li 6 must be stable by considering ali the possible ways into
which it might be imagined to split.
CHAPTER 13
ALPHA-DECAY
The properties of the radiations from radioactive substances have been

studied diligently since the discovery of radioactivity. The first interest
in these radiations was in connection with the series of transformations of
uranium, thorium, and actinium. This interest has been extended to the
information that the radiations give about the nucleus and the energy
changes involved in its transformations. Accurate measurements of the
energies of the radiations emitted by the natural radionuclides led to the
idea of nuclear energy states analogous to atomic energy states, and the
study of the radiations emitted by both the natural and artificial radio-
nuclides has resulted in the accumulation of a large amount of information
about nuclear levels. Theories of the emission of a-, {J-, and 'Y-rays have
been developed, and the combination of the experimental and theoretical
knowledge of these processes forms one of the great branches of nuclear
physics.
13-1 The velocity and energy of alpha-particles. The determination

of the velocity and energy of a-particles will be discussed in sorne detail for
several reasons. First, the accurate measurement of a-particle energies
made it possible to determine energies which differ only by small amounts,
and this led to the discovery that sorne radionuclides actually emit a
spectrum of a-particles. Second, knowledge of the energies of the com-
ponents of a-spectra makes it possible to assign certain nuclear energy
levels with confidence. Third, the methods for determining the energies
of a-particles are also used for protons and deuterons. These three charged
particles are involved in many artificial disintegrations, and the accurate
measurement of their energies yields accurate Q-values; from the Q-values,
it is possible to determine nuclear masses and nuclear energy levels.
Fourth, accurate values of a-particle energies are needed in the develop-
ment and use of the theory of a-decay.
The method that gives the most precise results for the velocity and
energy of a-particles depends on the measurement of the deflection of the
paths of the particles in a magnetic field. <1 • 2 • 3 > Wben a charged particle
moves in a magnetic field, its orbit is a circle whose radius is determined
by the relation
Mv 2
Hqv = --, (13-1)
r
304
13-1] THE VELOCITY AND ENERGY OF ALPHA-PARTICLES 305
Frn. 13-1. Schernatic diagrarn of the deflection charnber of a rnagnetic

spectrograph for a-particles.
where H is the field strength, q and M are the charge and mass of the
particle, respectively, and r is the radius of the orbit. Equation (13-1)
may also be written
q
v = M Hr. (13-2)
The velocity can be determined if the strength of the magnetic field is

known and if the radius of the orbit is measured, since the value of the
charge-to-mass ratio is well known.
A schematic diagram of an apparatus based on this principie is shown in
Fig. 13-1. The a-particles from a radioactive source emerge in a narrow
beam through a defining slit. A magnetic field of known strength, acting
in a direction perpendicular to the plane of the diagram, bends the a-
particles through an angle of 180°. The chamber slits help to reduce the
scattering of a-particles from the top, bottom, and walls of the chamber.
Particles with the same velocity have semicircular paths of the same
radius; they may be detected with a photographic plate or with counters,
and the radius of the path measured. In the figure, the paths of two groups
of particles with different velocities are shown. In most instruments of
this type, the a-particles can be bent into a semicircle of 40 or 50 cm
maximum radius. The instrument is called a magnetic spectrograph and
is designed on the principie of semicircular magnetic focusing.
The velocity v is obtained in centimeters per second when H is expressed
in gauss, r in centimeters, and q/ M in emu per gram. The charge q is, in
emu,
= 2 X 4.8029 X 10- 10 esu = _ 20
q 2.9979 X 10 10 cm/sec 3 ·2043 X 10 emu.
The mass of the a-particle is equal to the atomic mass of the helium atom
306 ALPHA-DECAY [CHAP. 13
TABLE 13-1
VELOCITIES AND ENERGIES OF ALPHA-PARTICLES
ÜBTAINED BY THE MAGNETIC DEFLECTION METHOD( 2 b)
Magnetic deftection, Velocity, Energy,

Radionuclide
Hr: gauss-cm X 10-5 cm/sec X 10-9 Mev
AcC(Bi 211 ) 3.7067 ± 0.003 1.7846 6.620 ± 0.013

ThC(Bi 212) 3.54232 ± 0.0008 1.7056 6.0466 ± 0.0027
RaF(Po 21 º) 3.31649 ± 0.0008 1.5972 5.3007 ± 0.0026
ThC'(Po 212 ) 4.26934 ± 0.0009 2.0514 8.7801 ± 0.004
Tn(Em220, 3.6108 ± 0.0003 1.7373 6.2823 ± 0.0013
Rn(Em 222 ) 3.37401 ± 0.00020 1.6247 5.4861 ± 0.0007
ThX(Ra224) 3.4336 ± 0.0003 1.6533 5.6814 ± 0.0011
Ra(Ra226) 3.1490 ± 0.0016 1.5167 4.779 ± 0.005
RdTh(Th 228) 3.3544 ± 0.0010 1.6154 5.4226 ± 0.003
lo(Th2ao) 3.118 ± 0.003 1.5018 4.685 ± 0.010
less two electron masses, or
M = 4.003873 - 0.001098 = 4.00278 amu

= 4.00278 X 1.6596 X 10- 24 = 6.6430 X 10- 24 gm.
Then
if = 4823.5 emu/gm,
and
v (cm/sec) = 4823.5 Hr. (13-3)
For a-particles from the natural radionuclides, the magnetic field strength
used is generally of the order of 10,000 gauss, and the values of Hr are in
the range 300,000 to 500,000 gauss-cm. The velocities, which can be
conveniently measured by this method, vary from about 1.6 X 10 9 cm/sec
to about 2.2 X 10 9 cm/sec. At these velocities, the relativistic mass
correction is small and may often be neglected. In this case, the kinetic
energy E may be taken equal to ½Mv 2 and is given by
E= ½Mv 2 = ½(6.643)10- 24 v2 erg

= 2.074 X 10- 18v2 Mev, (13--4)
where v is given by Eq. (13-3). When the relativity correction must be

taken into account, as in highly accurate work, Eqs. (13-3) and (13--4) are
13-2] THE ABSORPTION OF ALPHA-PARTICLES 307
replaced by the more complicated formulas
v = Hr- qR2
Mo
1 - -2,
c
(13-5)
and
T = M oc 2 [ ------;:==~====
vl - 1(v2/c 2) - 1] , (13--6)
where T represents the relativistic kinetic energy.

When the value of Hr is known, the velocity is obtained from Eq. (13-5);
with the known value of v, the kinetic energy is then given by Eq. (13--6).
Sorne typical results for natural a-emitters, from a critica! survey, <2b> are
listed in Table 13-1. The uncertainty in the measured value of the magnetic
deflection is small enough so that the relativistic correction must be made.
The uncertainty in the velocity is one unit in the fourth decimal place.
If the relativistic correction were not made, the effect would be to reduce
the energy by ¼to ½%, or 0.015 Mev in the case of ThX and 0.039 Mev
in the case of ThC'.
Although the magnetic deflection method gives highly accurate results,
it requires samples with a relatively large amount of activity, and cannot
be used successfully with substances su~h as U 238 , U 235 , or Th 232 . It is
relatively expensive, and more convenient (if less accurate) methods of
determining a-particle energies have also been developed.
13-2 The absorption of alpha-particles : range, ionization, and stopping

power. The energies of charged particles, including a-particles, can be
determined from measurements of their absorption by matter. <4 •5 > Before
the methods and the results are discussed, sorne of the features of the
absorption of a-particles will be treated.
Alpha-particles are easily absorbed; those emitted in radioactive dis-
integrations can generally be absorbed by a sheet of paper, by an aluminum
foil 0.004 cm thick, or by severa! centimeters of air. If the particles
emitted by a source in airare counted by counting the number of scintilla-
tions on a zinc sulfide screen, it is found that their number stays prac-
tically constant up to a certain distance R from the source, and then drops
rapidly to zero. This distance R is called the range of the particles, and is
related to the initial energy of the particles. If measured ranges are
plotted against energies determined by magnetic deflection methods, the
resulting range-energy curve can be used to find unknown energies from
measured ranges. It is usually easier and cheaper to measure ranges
than energies, and this method is used often.
Precise measurements of ranges in air can be made with an apparatus< 5 >
like that shown in Fig. 13-2. The source is placed on a movable block
308 ALPHA-DECA Y [CHAP. 13
~----_.,.,___co l!imated
source on
11 ad¡ustable
Sneen-walled-+
ionization 1¡ block
chamber
,j
Fm. 13-2. An apparatus for precise measurements of the range of a-particles

(Holloway and Livingston<5>).
1.0
0.8
,::
o 0.6
.
·,;;;
·a"
.s
-~"' 0.4
~"
0.2
3.90 4.00
Fm. 13-3. Range curves for the a-particles from Po 212 (Holloway and
Livingston< 5 >). Curve A: Number-distance curve with the extrapolated range
R, = 3.897 cm. Curve B: Differential range curve, with the mean range
~ = 3.842 cm. Curve C: End of a specific ionizatíon curve with the ionization
extrapolated range R; = 3.870 cm. Curve D: End of a specific ionization
curve for a single particle of mean range.
whose distance from a detector can be varied. A narrow beam of particles

emerges through a collimating slit, passes through a known thickness of air,
and reaches the detector. The latter is a thin, screen-walled ionization
chamber 1 to 2 mm deep. Ions are formed in pulses in the chamber when
individual a-particles pass through it. The voltage pulses induced on the
chamber electrode are amplified electronically and counted. The counting
rate is then determined as a function of the distance between source and
Fm. 13-4. A cloud chamber photograph showing the straggling of a-particles

from Thorium C'. The last three centimeters of range are shown. [Reprinted by
permission from Rutherford, Chadwick, and Ellis, Radiations from Radioactive
Substances. New York: Macmillan, 1930.]
detector. The results of an experiment with a-particles from polonium are

shown in Fig. 13-3. The ordinate is the relative number of particles, and
the abscissa is the distance from the source. Curve A shows the fraction
of particles in the beam detected at various distances from the source;
only the results near the end of the path are shown. All of the particles
which pass throúgh the slit are counted until the detector is about 3.75 cm
from the source; the fraction detected falls rapidly to 0.-2 at 3.88 cm, and
then decreases somewhat more slowly to zero. A quantity called the
extrapol,ated range, R., is obtained by drawing the tangent to the curve at its
inflection point and noting where the tangent crosses the distance axis.
The dotted line in the figure is the tangent and gives the value 3.897 cm
for the extrapolated range.
If the derivative of the number-distance curve (curve A) is computed at
different distances from the source and then plotted against the distance,
310 ALPHA-DECA Y [CHAP. 13
a differential range cun•e (curve B) is obtained. This curve gives the rela-
tive number of particles stopping ata given distance as a function of the
distance from the source; the unit of the ordinate is so chosen that the area
under the differential range curve is unity and ali the particles are accounted
for. The maximum ordinate of the differential range curve occurs at a
value of the abscissa which is called the mean range, 1l, defined so that half
the particle tra.ck lengths exceed it, while half are shorter. In the case
considered, the mean range is 3.842 cm. The results show that the track
lengths of the particles in the beam are not ali the same, but vary around
an average value. This effect is called straggling, and is illustrated in
Fig. 13-4, which is a cloud chamber photograph of a-particles from ThC'.
Alpha-particles lose a large fraction of their energy by causing ionization
along their paths. The extent of the ionization caused by an a-particle
depends on the number of molecules it bits along its path and on the way in
which it bits them. Sorne particles hit more and others hit less than the
average number of molecules in passing through a centimeter of air.
Hence, the actual distance from the source at which their energy is com-
pletely used up is somewhat different for different particles, giving rise to
straggling. Because of straggling, the actual range of an a-particle is not
definite, and to avoid this indefiniteness either the extrapolated or the
mean range is used. The values of the ranges are usually given for air at
15°C and 760 mm Hg.
The ionization caused by a beam of a-particles can be measured, and is
related to the energy and range. An electron and the positive ion which
results from its removal from an atom forman ion pair, and the intensity of
the ionization caused by the particles is expressed by the specific ionization,
defined as the number of ion pairs formed per millimeter of beam path.
The apparatus of Fig. 13-2 can be used to measure the relative specific
ionization produced by a beam of a-particles at different distances from the
source. The amplifier can be designed so that the voltage height of its
output pulse is very nearly proportional to the number of ion pairs formed
in the chamber. Specific ionization-distance curves are shown in Fig. 13-5
for the particles from RaF (Po 21 º) and RaC' (Po 214 ). The same type of
curve is obtained for any gas and for any group of monoenergetic particles
even though the magnitudes of the ionization and of the range vary con-
siderably. As the distance of the a-particles from the source increases, the
relative specific ionization increases, at first quite slowly and then more
rapidly, reaches a maximum, and then drops sharply to zero. The effect
of straggling is seen near the end of the range, and is shown clearly in
curve C of Fig. 13-3. The ionization extrapolated range Ri is defined as the
value of the abscissa at which the tangent to the curve at its inflection
point crosses the horizontal axis; the value of Ri for Po 21 º was found to be
3.870 cm.
~1.1.----,---r----, ---...---.------,---- ,r--,

·a 1.01--+---+--+--..:---r::L --:-:----':,:-=--im:,-:;;::rlr ---t
::s
t: 0.8 t----+---+------,t--1 t--+---t---+--Hi----f
f! Polonium (Po21
] 0.6 t---+-----~--t--++ ---~--+--+-i----1
a
';;' 0.4 ~:::::¡...~,,J::,==!== =:tl=--1--!-_JU--1
:3
::g 0.21---+-----t----t- --t+---t----+---1i-- --1
] QL----'---'---'---'" '---L------'------'..........
O 2 3 4 5 6 7 7.5
Distance in cm in air at 760 mm and 15ºC
Fm. 13-5. Specific ionization of a beam of a-particles as a function of

distance from the source. [Reprinted by permission from Rutherford, Chadwick,
and Ellis, Radiations Jrom Radioactive Subatances. New York: Macmillan, 1930.]
Distance of the detector from the source (cm)

1.5 2.0 2.5 3.0 3.5 4.0
10 1 1
1 1
' / '1-
/"
8
• Ve
/i X
',,
--
a__,.-- --
'
~ IX
1
2
X
o
"
V
2.5 2.0 1.5 1.0 0.5 o

Distance from the end of the track (cm)
Fm. 13-6. Specific íonization of a single particle of mean range as a function

of distance from the source and from the end of the track (Holloway and Living-
ston<5>).
lt is possible to correct the relative ionization-distance curve of a beam

of a-particles for the effect of straggling, and to obtain a specific ionization
curve for a single particle of mean range; a curve of this kind is shown in
Fig. 13-6. The specific ionization is usually measured from the end of the
track, and the bottom abscissa scale gives the distance from the end
of the track, while the top scale gives the distance of the detector from
the source. A specific ionization curve for a single a-particle can be
obtained directly by studying the photographic density of tracks in a

cloud chamber;< 6> the results for the a-particles from Po 21 º are plotted as
in Fig. 13-6.
The total number of ions produced in a gas by the complete absorption
of an a-particle of known energy can be estimated. An a-particle from
RaC' (Po 214 ) whose energy is 7.68 Mev and whose ionization extrapolated
range is 6.95 cm produces a total of 2.2 X 10 5 ion pairs in air at 15°C
and 760 mm of Hg before being stopped. It follows that the particle
loses about 35 ev, on the average, for each ion pair formed. The ionization
per millimeter of path for the average a-particle from RaC' has been
deduced from the ionization curve of Fig. 13-5. When the particle is
just starting it produces about 2200 ion pairs/mm. The ionization in-
creases very slowly at first as the a-particle loses energy, and is 2700 ion
pairs/mm 3.0 cm from the end of the range. lt then increases more and
more rapidly and reaches a maximum of about 7000 ion pairs/mm when
the particle is 4 or 5 mm from the end of its range. The specific ioniza-
tion finally decreases very rapidly in the last few millimeters of its range.
The shape of the ionization curve depends on the change in speed of an
a-particle as it traverses its path. In producing ion pairs, the pe.rticle
loses energy and its speed decreases. When it moves more slowly it spends
more time in the neighborhood of the air molecules it encounters, and the
probability of producing ion pairs increases. This effect accounts for the
increase in specific ionize.tion as the pe.rticle moves fa.rther from the source.
RdTh, 5.42 Mev

lOOOr-+--=T...,hX=-,- - - - - - - - - - - - - - - - - - - - - - ,
900 5.68
Mev
al 800
8 Tn,
ThA,
11 700 6.28
6.77
" Mev
Mev
~ 600
i ThC,
.s 500 6.05
.& Mev
§ 400
8
.¡ 300
~ 200
100
oL.i_.1...__L~_..L___L~lr>Aad.a-'l.a.ab""""ku,o.~t:.....l.--:~Mb>oaabCl..-.J
7 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80
Channel number
Frn. 13-7. Analysis of a-emitters in a multichannel pulse analyzer: radio-
thorium and its descendants. (Reprinted by permission of the United St&tes
Atomic Energy Commission.< 4 >)
13-2) THE ABSORPTION OF ALPHA-PARTICLES 313
Finally, electrons are captured by the a-particle anda neutral helium atom
is formed which can no longer cause ionization.
The total ionization caused by an a-particle can also be used to determine
the energy of the particle. If the entire path of an a-particle is contained
in an ion chamber and the total number of ion pairs formed is measured,
the energy of the particle can be determined, since the average number of
electron volts per ion pair is known for various gases. <7 ,s, 11 > Instead of
measuring the amount of charge collected as a result of the ionization
caused by each a-particle, the voltage pulses formed are analyzed by
means of ingeniously designed electronic circuits called multichannel,
differential pulse height selectors. <9 , im The output of the ionization chamber
is amplified and then fed into a number of different pulse-selector circuits.
Each circuit is set to select pulses of a different height, so that pulses of
a number of different sizes are recorded simultaneously. If there are
enough of these channels, the whole range of pulse heights may be recorded
at once. A plot of energy against channel number is made by using samples
of a-emitters whose energies are known from magnetic deflection methods,
and this plot serves as a calibration. An instrument of this kihd, <Io>
with 48 channels, has been very useful in a great deal of the work on
artificial a-emitting radionuclides discussed in the last chapter. An example
of the analysis with this instrument of the ionization produced in argon
by a sample containing RdTh (Th 228 ) and its descendants is shown in
Fig. 13-7. The channel number representa the size of the pulse (which
dependa on the energy); the curve was taken in two stages, RdTh to ThA
in the first stage and ThC' in the second stage. The number of a-particles
emitted by each nuclide is proportional to the area under the peak.
The discussion of the absorption of a-particles has so far been limited to
absorption by air. Other gases can be used and, in fact, studies of the
ionization produced in argon have recently< 11 > helped to clarify sorne
problems. Photographic emulsions have always been useful in the study
of the radiations from radioactive substances. Until the development of
electrical counting instruments, the emulsion was one of the most important
tools in the field of radioactivity. Rapid advances in the design of elec-
tronic instruments pushed photographic emulsions into a secondary posi-
tion, but recent improvements in the preparation of emulsions have made
them once again important tools in radioactive measurements.<1 2 > Meas-
urements of track lengths in emulsions can be correlated with particle
energies, giving range-energy curves for different charged particles.° 3 >
Because of their usefulness in nuclear physics, these emulsions are called
nuclear emulsions.
Range measurements can also be made with solid foils as absorbers.
Thin, uniform foils are placed over an a-emitting sample, and the thickness
needed to absorb the a-particles completely is a measure of the range.
Mica, aluminum, or gold foils are used commonly because they ciin be
prepared easily in different thicknesses with reasonable uniformity. The
range of a-particles in such materials is very small; the particles from RaC'
which have an extrapolated range of 6.953 cm in air have extrapolated
ranges of 0.0036 cm in mica, 0.00406 cm in aluminum, and 0.00140 cm in
gold.
Another quantity of great value in the treatment of the absorption of
charged particles by matter is the stopping power, defined as the energy lost
by the particle per unit path in the substance,
dE
S(E) = - - , (13-7)
dx
where E is the classical kinetic energy. The stopping power varies with
the energy of the particle, and the range of the particle is given by
(13-8)
where E 0 is the initial kinetic energy.

The stopping power S(E) of a substance can be determined experi-
mentally by measuring (e.g., by magnetic deflection) the energy of the
particles which have gone through a certain thickness of the substance.
When the energy loss is found in this way for different initial velocities,
the range in the substance can be deduced from Eq. (13-8) as a function of
the initial energy. If the range is known as a function of energy, the
stopping power can be obtained from the relation
(13-9)
One of the reasons for the importance of the stopping power is that it
can be calculated theoretically both from classical mechanics and quantum
mechanics.° 4 • 15> The theory of the stopping power depends on knowl-
edge of the behavior of electrons in atoms. It makes predictions which are
in good agreement with the results of experiments on the ionization pro-
duced by fast charged particles, on the energy loss, and on the range. The
theory provides an understanding of the empirical information concerning
the absorption of a-particles by matter, and there is a sound theoretical,
as well as empirical, basis for the determination of charged particle energies
from range and ionization measurements.
Although a detailed treatment oí the stopping power is beyond the scope
of this book, the usefulness of the concept can be illustrated by discussing
one of its applications. The energy lost by a nonrelativistic charged
13-2) THE ABSORPTION OF ALPHA-PARTICLES 315
particle per unit length of its path in a given substanoe can be expressed
in the following form derived from theory,
S = _ dE = 4n2 e4N z In (2m,v2) , (13-10)

d:x m.,v2 1
where ze and vare the charge and the speed of the particle, respectively,
m. is the mass of an electron,
Z is the atomic number of tbe substance,
N is the number of atoms per cubic centimeter of the substanoe, and
J is a quantity, called the average excitation potential of an atom of the
substance, whieh must be obtained experimentally.
For an a-particle, z = 2, and Eq. (13-10) may be rewritten
S' = - meV2 dE = z In (2m,v2) . (13-11)

16'ire 4N d:x I
The quantity on the left, called S' for convenience, is proportional to the
stopping power; it can be calculated and measured for different va.lues of
the atomic number Z of the absorbing substance. The quantity I depends,
aroong other things, on Z, so that a graph of S' against Z should give a
curve whieh is not quite a straight line. The a.greement between theory
and experiment is very good, as can be seen from Fig. 13-8; the solid
curve representa the theoretical prediction, the circles represent experi•
mental results. Calculations of ranges from Eqs. (13--8) and (13-10) give
value~ in good a.greement with experiment.
o 10 20 40
z
F10. 13-8. The va.ria.t.ion of the stopping power with atomic number.
316 ALPHA-DECAY [cu..u>. 13
Since the stopping power varíes with the velocity of the particle, it is
convenient to consider the ratio of the stopping power to that of a standard
substance, a ratio a.pproximately independent of the velocity. Air at
15ºC and 760 mm Hg is usually ta.ken as the standard. The relatwe stop-
ping pO'IJJer of a substa.nce, defined as the ratio of the range in a.ir to the
range in the substance, is often used,
R I t" 5 to . Range of a-particle in air ( )
ª
e ive ppmg power = Range of a-particle in substance I 3- l 2
For sorne purposes it is useful to express the stopping power of a substanee
in an altemative fonn, the equivalent thickne88 in units of mg/cm2 defined
by the relationship
Equivalent thickness in mg/cm 2 = Range X Densíty X 1000. (13-13)
This quantity gíves the mass per unit a.rea, or the thickness, of material
needed to absorb the a-particles. The thickness that is equivalent in
stopping power to one centimeter of a.ir is obtained by dividing the equiva-
lent thickness, defined by Eq. (13-13), by the range of the a--particles in
air. Then, from Eqs. (13-10) and (13-11),
Thickness in mg/em 2 equivalent to 1 cm of air =
Density X 1000
(13-14)
Relative stopping power
Values of the quantities discussed are listed in Table 13-2 for sorne of the
most frequently used foil materia.Is. The values gíven are for a-particles
from RaC', which have an extrapolated range of 6.953 cm in air at 15ºC
and 760 mm Hg. The extrapolated range is used here beca.use this is the
quantity for which experimental values are gíven in the table.
TABLE 13-2
RANGE AND 8ToPPING POWER FOR RaC' ALPHA-PARTICLES
IN VARJOUS SuBSTANCES
Thickness
Rela.tive Equivalent
Extra.polated Density, in mg/em2
Substance stopping thickness,
range, cm gm/cm 3 equiv&lent to
power mg/em 2
l cm of a.ir
Mica 0.0036 1930 2.8 10.1 1.45

Aluminum 0.00406 1700 2.70 n.o 1.57
Copper 0.00183 3800 8.93 16.3 2.35
Gold 0.00140 4950 19.33 27.l 3.89
13-3] RANGE-ENERGY CURVES 317
13-3 Range-energy curves. The ranges in air of the a-particles from

sorne natural a-emitters are listed in Table 13-3 together with the energies
measured by the magnetic deflection method. The ranges are quoted from
Reference 5 and the energies from Reference 2b. The values show clearly
the differences between the mean and extrapolated ranges as well as the
general relation between range and energy. The range increases monotoni-
cally with the energy, and for mean ranges between 3 and 7 cm, the em-
pirical equation 7l = 0.318E312 , with 71, in centimeters of air at 15ºC and
760 mm Hg, and E in Mev, holds fairly well. The extrapolated range, as
obtained from the number-distance curve, is greater than the mean range;
the difference depends on the magnitude of the mean range and varies
from about 0.05 cm for particles with a mean range of 4 cm to about 0.15 cm
for particles with a mean range of 11 cm. lt is easier to determine the
extrapolated range experimentally, and the mean range can then be cal-
culated quite well 06> from the extrapolated range. The mean range has
the advantage of being somewhat less dependent on the particular experi-
TABLE 13-3
RANGES AND ENERGIES oF SoME NATURAL ALPHA-EMITTERS
Range from
extrapolated Rangefrom
Mean range
number- extrapolated Energy,
Radionuclide in air,
distance ionization, Mev
cm
distribution, cm
cm
uEm 219 (An) 5.240 = 0.015 5.312 5.272 6.542 = 0.006

5.692 = 0.015 5.769 5.727 6.807 = 0.006
uPo 216 (AcA) 6.457 = 0.008 6.542 6.496 7.383 = 0.012
aaBi 211 (AcC) 4.984 = 0.015 5.053 5.015 6.273 = 0.013
5.429 = 0.015 5.503 5.462 6.620 = 0.013
uPo 211 (AcC') 6.555 = 0.015 6.641 6.595 7.442 = 0.015
uEm•• 0 (Tn) 5.004 = 0.018 5.073 5.035 6.2823 = 0.0013
uPo 21 ª (ThA} .5,638 = 0.008 5.714 5.672 6.7746 = 0.0013
uBi 212 (ThC) 4. 730 * 0.008 4.796 4.778 6.0466 = 0.0027
uPo 212 (ThC') 8.570 =0.007 8.676 8.616 8.7801 = 0.004
9.724 * 0.008 9.841 9.780 9.4923 = 0.004
11.580 = 0.008 11.713 11.643 10.5432 = 0.004
uEm 222 (Rn) 4.051 = 0.008 4.109 4.076 5.4861 = 0.007
uPo•t• (RaA) 4.657 = 0.008 4.722 4.685 5.9982 = 0.0008
a.Po• u (RaC') 6.907 = 0.006 6.997 6.953 7.6804 = 0.0009
7.793 = 0.015 7.891 7.839 8.2771 = 0.0036
9.04 * 0.02 9.15 9.09 9.0649 = 0.0040
11.51 = 0.02 11.64 11.57 =
10.5058 0.0045
uP0•10 (RaF) 3.842 * 0.006 3.897 3.870 5.3007 = 0.0026
4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5

3.0 s.°s.o
Energy (Mev} - Curve II
7.5
2.5 7.0
6.5
2.0 ....., 6.0
~ l 5.5
(.)
"
(.)
1 1
1.5 5.0
6~ 6~
i i 4.5
1.0
! ! 4.0
3.5
0.5 3.0
2.5
Energy (Mev} - Curve I
0 ~ ~ - ~ - ~ - ~ ~ - _ _ . __ _.___.2.o
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Fro. 13-9. Range-energy curves for a-particles (Bethe< 16>).
mental setup, and is now generally used in range tables and for range-
energy curves.
When the mean range is plotted against the energy, the resulting range-
energy curve can be used to find the energy of a--particles whose range in
air is measured. The data of Table 13-3 would give a range-energy
curve down toan energy of only about 5 Mev, which would limit the useful-
ness of the curve. Information at lower energies can be obtained either
by studying the range and energy of particles which have been partially
slowed down, or of particles produced in nuclear reactions. Two reactions
producing a-particles have been found to be valuable for the range-energy
relation. The first of these is
Li
6
+n - He
4
+ H 3 + Q,
with a Q-value of 4.788 ± 0.023 Mev. The energy of the a--particle is
2.057 ± 0.010 Mev and the range is 1.04 ± 0.02 cm. The second reaction
is
B 10 n - Li 7 He 4 Q,+ + +
13-4) ALPHA-PARTICLE SPECTRA 319
with a Q-value of 2.316 ± 0.006 Mev. The energy of the a-particle is

1.474 ± 0.004 Mev and its ralllge is 0.720 ± 0.015 cm. With the points
provided by these nuclear reactions, and wíth recent improvements in the
knowledge of the relationship between ionization and energy loss, it has
been possible to prepare more 'extensive and more accurate range-energy
curves.< 16 • 17¡ A curve for a-particles is shown in Fig. 13-9; similar curves
are available for protons and deuterons.
The use of the range-energy curve roa.y be illustrated by sorne simple
examples. A recent measurement of the mean range of a-particles from
U 238 (UI) gave a value of 2.70 ± 0.02 cm. According to curve II of
Fig. 13-9, this range corresponds toan energy of 4.25 Mev. The energy
as determined from total ioni~tion measurements is 4.180 ± 0.015 Mev,
and the agreement is not bad. For Th 232, the range was found to be
2.49 ± 0.02 cm, and the energy from the curve is 4.02 Mev; the energy
determined from the total ionization is 3.99 Mev, and the agreement is
very good.
13-4 Alpha-particle spectra. Long-range particles and fine structure.

In the treatment of the energy and range of a-particles, it has been im-
plied that all a-particles from a given active nuclide have the same initial
velocity and energy. Careful experiments ha.ve shown, however, that a
given nuclide often emits particles with a number of different energies.
Radium C' and ThC' emit a few long-range particf,es having energies con-
siderably greater than that of the main group of particles. The existence
of long-range a-particles was first observed by Rutherford and Wood, os>
who studied the absorption of the a-rays from a sample of ThC. Thorium C
emits a-particles with a mean range of 4.73 cm; it also decays by ~-emission
to ThC' which, in turn, emits a-particles wíth a mean range of 8.57 cm.
A sample of ThC contains ThC' and acts as a source of a-particles wíth
both of the ranges cited. Rutherford and W ood found that a few particles
from such a source were able to pass through an absorbing screen suf-
ficiently thick to stop all particles with a ra.nge of 8.6 cm. When more
absorbing material was used the penetrating particles were found to have a
range of about 11 cm in a.ir, and it was clear that they were different from
the normal particles emitted by ThC and ThC'.
La.ter studies with the cloud chamber, ionization chamber, and magnetic
spectrograph indicated that ThC' emits two groups of long-range particles
with energies of 9.492 Mev and 10.543 Mev, respectively, as compared
wíth an energy of 8. 780 Mev for the pa.rticles in the roa.in group. The most
recent magnetic spectrograph measurements show the presence of a tbird
group with an energy of 10.422 Mev. For ea.ch million particles of the
roa.in group there are about 40 particles in the 9.492 Mev group, 20 in the
10.422 Mev group, and 170 in the 10.543 Mev group.
FIG. 13-10. A cloud chamber photograph of a-particles from RaC' showing

two long-range particles. (Nimmo and Feather, Proc. Roy. Soc. (London) Al22,
668 (1929).)
TABLE 13-4
LoNG-RANGE ALPHA-PARTICLES FROM RaC'
Difference
Alpha- Alpha- of alpha-
Mean Relative
particle disintegration disintegration
range, energy, number of
energy, energy from
cm particles
Mev Mev that of main
group, Mev
6.870 7.680 10º 7.827 0.000

7.755 8.277 0.43 8.435 0.608
.... 8.938 0.45 9.108 1.281
9.00 9.065 22 9.238 1.411
.... 9.313 0.38 9.490 1.663
.... 9.489 1.35 9.669 1.842
.... 9.657 0.35 9.841 2.014
.... 9.779 1.06 9.965 2.138
.... 9.905 0.36 10.093 2.266
.... 10.074 1.67 10.265 2.438
.... 10.146 0.38 10.339 2.512
.... 10.326 1.12 10.523 2.696
11.47 10.506 0.23 10.706 2.879
13-4) ALPHA-PARTICLE SPECTRA 321
Measurements with the magnetic spectrograph have also shown that

RaC', whose main group of a-particles has an energy of 7.680 Mev, emits
12 groups of long-range particles with energies ranging from 8.277 Mev to
10.506 Mev.° 9 > Altogether, about 30 long-range particles are emitted per
million of normal range. A cloud-chamber picture of typical long-range
particles from RaC' is shown in Fig. 13-10. The long-range a-particles
from RaC' are lísted in Table 13-4. Another quantity, the alpha dis-
integration energy, is also listed; it will be discussed in Section 13-5.
Until 1930 it was thought that with the exception of the rare long-range
particles, the a-particles emitted by an active nuclide all had the same
initial energy. In that year it was shown by Rosenblum< 20 > in very careful
magnetic deflection experiments that the normal a-particles emitted by
sorne active nuclides fall into several closely spaced velocity groups. The
velocities and energies of the different groups differ so little that the
ranges of ali the particles lie within the region of straggling. For this
reason the different groups could be separated only in a magnetic spectro-
graph of high resolving power. The discrete, closely spaced components
of the a-rays are said to form a spectrum, or to show fine structure. A great
deal of research has been done on a-particle spectra since Rosenblum's
discovery and the spectra of many natural and artificial a-emitters are
now known. Thus, Th 232 has two groups, with energies 3.994 Mev(76%)
and 3.936 Mev(24%). lonium (Th 23 º) has at least four groups, with
energies of 4.682 Mev(76.3%), 4.615 Mev(23.4%), 4.471 Mev(0.2%) and
4.436 Mev(0.07%); it may have several other groups with very small
abundances. Radium (Ra 226) has three groups with energies of 4.777
Mev,94.3%), 4.593 Mev(5.7%), and4.34 Mev(0.01%), while RaAc(Th 227 )
has 14 groups with energies between 6.030 Mev and 5.661 Mev and rela-
tive abundances varying from 25 to 0.1%.
TABLE 13-5
THE ALPHA-PARTICLE SPECTRUM OF ThC
Alpha.- Difference
Alphe.- Rela.tive disinte- of a.lpha.-
Group pa.rticle number of gration disintegra.tion
energy, pa.rticles, energy, energy from that
Mev % Mev of ao group, Mev
6.086 27.2 6.203 0.000

ªºa1 6.047 69.9 6.163 0.040
a2 5.765 1.7 5.874 0.329
5.622 0.15 5.730 0.473
a4
ªª 5.603 1.1 5.711 0.492
as 5.478 0.016 5.584 0.619
The a-particle spectrum of ThC (Bi 212 ) is given in detail in Table 13-5.
The values of the energy used are averages of two sets of measurements
given in the isotope table of Hollander, Perlman, and Seaborg.
The information now available about a-particle spectra leads to the con-
clusion that these spectra may be divided into three groups:
l. Spectra consisting of a single group or "line," for example, Rn,
RaA, RaF.
2. Spectra consisting of two or more discrete, closely spaced (in velocity
or energy) components with intensities of the same, or of only a slightly
different, order of magnitude, for example, ThC, An, AcX, ThX, Pa,
RdAc.
3. Spectra consisting of a main group and groups of much higher energy
(long-range) particles, the latter containing, bowever, only a very small
fraction (10- 4 to 10-7 ) of the number of particles in the main group. The
third kind of spectrum occurs only in two extremely short-lived nuclides
RaC' and ThC'.
13-5 Nuclear energy levels. The discovery of a-spectra gave rise

to the problem of why a-particles should be emitted by a given nucleus
only with certain discrete energies. In view of the successful interpreta-
tion of atomic spectra in terms of discrete electronic energy levels, as in
the Bohr theory of the atom and in wave mechanics, it seemed reasonable
to try to account for discrete a-spectra in terms of nuclear energy levels.
It is supposed that there is a number of discrete energy Ievels in the nucleus,
and that a nucleus is normally in its lowest energy state, but under certain
conditions it may exist for short times in excited states, i.e., in configura- -
tions having more than the normal amount of energy. A nucleus in such an ·
excited state would be expected to give up its energy by sorne emission
process, for example, by the emission of radiation. To apply these ideas,
it must be possible to determine energy levels in the nucleus experi-
mentally. As in the case of atoms, nuclei must be excited to emit radiation
so that the energy levels in the nucleus may be deduced from the fre-
quencies of this radiation. It was known that those <Hlmitters that have
discrete a-particle spectra also emit 'Y-rays. Careful measurements of the
a-particle energies and the 'Y-ray energies led to the conclusion that the
'Y-rays are emitted by the product nucleus that has been left in an excited
state after the emission of an a-particle. Thus, the process of a-emission
sometimes leaves excited nuclei which then emit radiation, and the way in
which the energies of the 'Y-rays and the a-particles can be correlated in
terms of nuclear energy levels was shown by Gamow. <2 u
Before proceeding further with the discussion, it is necessary to dis-
tinguish between the kinetic energy of an a-particle and the total energy
change in an a-decay process; the latter quantity has been called the
13-5] NUCLEAR ENERGY LEVELS 323
a-disintegration energy. When an a-particle is emitted, the product,

or residual, nucleus recoils, carrying with it a certain amount of energy.
The a-disintegration energy is the sum oí the kinetic energies of the
a-particle and the product nucleus, and is found as follows. The principie
of conservation of momentum requires that
Mv = M,.v,, (13-15)
where M, is the mass of the product nucleus, v, is its velocity, and the
quantities without subscripts refer to the a-particle. Where relativistic
effects are neglected, the a-disintegration energy E,. is given by
E,. = ½Mv 2 + ½M,v~.
From Eq. (13-15), v, = (M/M,)v, and
E,. = ½Mv 2 ( l + :) . (13-16)
The a-disintegration energy is obtained, therefore, simply by multiplying

the kinetic energy of the a-particle by the quantity
1
+ mass of a-particle .
mass of product nucleus
The a-disintegration energy is listed in the fifth column of Tables 13-4 and
13-5 for the a-particle groups from RaC' and ThC, respectively.
The interpretation of complex a-spectra in terms of nuclear energy
evels can be illustrated by the case of ThC(Bi 212 ), which emits six groups
::ir a-particles with energies Iisted in Table 13-5. The most energetic
particles have a kinetic energy of 6.086 Mev corresponding to an a-
disintegration energy of 6.203 Mev. It is supposed that a ThC nucleus
always releases 6.203 Mev of energy when it decays by a-emission to forro
a nucleus of ThC"(Tl 2 º8 ), and that this amount of energy is associated
with one of the most energetic a-particles. The emission of one of these
particles is assumed to leave the product nucleus in its lowest energy state,
or ground state. All of the disintegration energy has gone into the kinetic
energy of the a-particle and the recoil energy of the ThC" nucleus. Sup-
pose now that a ThC nucleus emits an a-particle of the a 1-group for which
the disintegration energy is 6.163 Mev or 0.040 Mev less than the total
available a-disintegration energy. The ThCH nucleus, which retains the
0.040 Mev of energy, should be left in an excited state; it might be ex-
pected to undergo a transition to the ground state by emitting electro-
magnetic radiation (in analogy with the transitions undergone by atoros
which have electrons in excited states). If this is the case, the radiation
should appear in the form of a 'Y-ray with an energy of 0.040 Mev. Gamma
Energy above ground
state of ThC" (Mev)
ThC"(TI208) ThC (Bi212 )
0.619
1'(0.144):
0.492
0.473
0.329
1'(0.288)
·----'}(0.472)
0.040
1' (0.040)
o
FIG. 13-11. Decay scheme for the a-decay of ThC to ThC".
rays of this energy are, in fact, found experimentally when ThC under-
goes a-decay. By extending this analysis we can postulate the existence of
at least five excited states of the ThC" nucleus and we would expect to
find additional transitions from more excited states to less excited states.
These transitions should involve the emission of -Y-rays. Experimentally,
eight -Y-rays have so far been identified, with energies of 0.040, 0.144,
0.164, 0.288, 0.328, 0.432, 0.452, and 0.472 Mev, respectively. Each of
these -Y-rays can be matched (within the uncertainties of the measure-
ments) toan energy difference between two excited states, and the energies
of the a-particle groups and the -Y-rays forro a consistent scheme.
As a result of the correlation between the energies of the a-particles and
the -Y-rays, it is possible to construct nuclear energy leve! diagrams analo-
gous to the atomic energy leve! diagrams shown in Chapter 7. The leve!
diagram for the daughter nucleus is usually shown together with the decay
data for the parent nucleus and the result is called a decay scheme. The
scheme for the decay of ThC to ThC" is shown in Fig. 13-11. The diagonal
lines represent the different a-particle groups observed in the decay of
ThC; the disintegration energies are also shown in this example. The
13-5] NUCLEAR ENERGY LEVELS 325
fj, 0.08 Mev (5%)
fj, 0.45 Mev (8.5%)
fj, 0.67 Mev (6%)
fj, 0.93 Mev (7.5%)
fj, 1.55 Mev (10%)
fj, 2.27 Mev (63%)

Energy above ground
leve! of ThC' (Mev)
--11.20 - - - - - - - Tn'"'-------2.19
--10.75 - - - - - - -
--10.55 - - - - - - -
"------+-+-
_,,,__,_.,...,..._.,...~1.6 1.8
-- 10.29 - - - - - - - - , ......,..._........+--+-+- 1.34

-- 9.67•-------
ThD (Pb208)
Fra. 13-12. Energy level diagram for the ThC-+ ThC'-+ ThD decays,
showing the origin of the long-range a-particles of ThC'.
vertical lines represent the 'Y-rays found experimentally. Each 'Y-ray

corresponds to the energy difference between two states, although there
is sorne am biguity in the case of the 0.144-Mev 'Y-ray which may correspond,
within the limits of experimental error, to either of two possible tran-
sitions. It is evident from the diagram that 'Y-rays are not observed for all
transitions which seem to be possible. The absence of these 'Y-rays may
be caused either by their small intensity, which may make them hard to
detect, or by nuclear selection rules which may make the probability of
these transitions very small. There is a theory which accounts in a satis~
factory way for the radiations from excited nuclei and the selection rules
involve the angular momentum and parity of the initial and final states.
The a-spectra of a large number of natural and artificial a~mitters have
now been correlated with the 'Y-ray spectra of the decay products, and
energy level schemes have been worked out. This procedure represents a
326 ALPHA-DECAY [CHAP, 13
start on the general problem of nuclear energy levels because it provides

one of the methods by which levels can be determined. Other methods
will be discussed in later chapters.
The origin of the long-range particles emitted by RaC' and ThC' can
also be interpreted in terms of energy levels. lt is possible for a nucleus,
before an a-disintegration, to be in an excited state; the excitation may
result from a previous disintegration. It has been found in the study of the
emission of {J- and 'Y-rays from nuclei that the emission of a nuclear {J-ray
sometimes leaves the newly formed nucleus in an excited state. In most
cases the excited nucleus goes to its ground state by emitting a 'Y-ray of
the proper energy. This process will be discussed in greater detail in the
next chapter. In sorne cases, however, the newly formed excited nucleus
is an a-emitter, and gets rid of its excess energy by emitting a-particles
with greater energy than that of the normal particles. The existence of
long range a-particles is thus explained as being caused by the decay of an
excited nucleus, and the extra energy of the a-particle measures the excita-
tion energy of the initial nucleus. An energy level diagram for the ThC-
ThC' decays showing the origin of the long range a-particles of ThC' is
given in Fig. 13-12.
13-6 The theory of alpha-decay. When the range, energy, half-life,

and disintegration constant of the natural a-emitters are compared, certain
remarkable facts stand out, which are illustrated by the data listed in
Table 13-6. With the exception of the rare long-range particles, the ranges
lie between 2.5 cm and 8.6 cm, and the a-disintegration energies vary from
4.05 Mev to 8.95 Mev. The ratio of the longest to the shortest range is
about 3.5 and the ratio of the greatest to the least energy is between 2 and
2.5. But the half-lives vary from 1.39 X 10 10 years for the longest-lived
nuclide to 3.0 X 10-7 sec for the shortest, and the disintegration con-
stants vary from 1.58 X 10- 18 sec- 1 to 2.31 X 10 6 sec- 1• A factor of two
or three in the energy corresponds to a factor of 10 24 in the half-life and
disintegration constant. In other words, the disintegration constant
varies extremely rapidly with small changes in energy. The longest-lived
nuclides emit the least energetic a-particles, while the shortest-lived
nuclides emit the most energetic particles. These empirical facts were
first correlated in the Geiger-Nuttauc 22 i rule, which stated that when the
logarithm of the disintegration constant (in sec- 1) was plotted against the
logarithm of the range (in cm of standard air) for the a-emitting members
in the same radioactive series, an approximately straight line was obtained.
The lines for the three natural series turned out to be practically parallel.
The rule may be represented by the formula
log X = A log R + B. (13-17)

13-6) THE THEORY OF ALPHA-DECAY 327
TABLE 13-6
RANGE, ENERGY, HALF-LIFE, AND DISINTEGRATION
CoNSTANT oF SoME ALPHA-EMITTERS
Mean ra.nge, Alpha.- Disintegra.tion

Nuclide cmof disintegra.tion Ha.lf-life consta.nt,
standard a.ir energy, Mev sec- 1
Th2a2 2.49 4.06 1.39 X 10 1º y 1.58 X 10-18

Ra,226 (Ra.) 3.30 4.86 1.62 X 103 y 1.36 X 10-11
Th 228 (RdTh) 3.98 5.52 1.9 y 1.16 X 10-8
Em 222 (Rn) 4.05 5.59 3.83 d 2.10 X 10-6
Po21s (Ra.A) 4.66 6.11 3.05 m 3.78 X 10-3
Po 216 (ThA) 5.64 6.90 0.16 s 4.33
Po214 (Ra.C') 6.91 7.83 1.64 X 10-4 s 4.23 X 103
Po 212 (ThC') 8.57 8.95 3.0 X 10-7 s 2.31 X 10 6
The consta.nt A is the slope of the line and has nea.rly the sa.me value for
the three series, but B has different values. The Geiger-Nutta.ll rule is
illustra.ted in Fig. 13-13. In Section 13-3 it wa.s noted tha.t the mean
range was found empirica.lly to be proportional to E 312, where E is the
kinetic energy of the a-particle. The Geiger-Nutta.ll rule ma.y then be
written in the form
log ~ = A' log E + B', {13-18)
which expresses the rela.tionship between the logarithm of the disintegra-

tion constant and that of the energy.
The facts of a-pa.rticle emission presented a difficult theoretical problem
which is brought out sha.rply when the sponta.neous emission of a-particles
by a nucleus is compared with the scattering of a-particles by the sa.me
nucleus. Consider the case of emission and sca.ttering by U 238. In ex-
perimenta on the scattering of a-particles, it wa.s found tha.t the fa.stest
natural a-particles availa.ble, those of ThC' with energies of nearly 9 Mev,
are una ble to penetra.te close enough to the nucleus to show departures from
the Coulomb la.w. lt follows tha.t at lea.st up to a distance of about
3 X 10- 12 cm from the nucleus, the potential energy of the a-particles
in the field of the nucleus is still expressed by the Coulomb formula
2Ze2
U(r) = -r-·
For r = 3 X 10- 12 cm, the potential energy is about 9 Mev, and increases
at least up to sorne maximum value for smaller values of r, until the
22
Th
.+
o
20
18 ""¾: A;eEm
.. 16
e
~ 14
00
e
o
"e 12
(Z..
Th
·-3
t 10
~ Pa
-~
:§,
8
..,
.s 6
0.4 0.5 0.6 0.7 0.8

Log range
Fm. 13-13. The Geiger-Nuttall rule.
Coulomb law breaks down. This repulsive potential prevents the incident
a-particle from entering the nucleus, and may be considered to form a
potential barrier.
Since radioactive uranium nuclei emit a-particles, the latter may be
supposed to exist, at least for a short time before emission, within these
nuclei. The interactions between a radioactive nucleus and an a-particle
inside and outside the nucleus may then be represented by a potential
energy curve such as that in Fig. 13-14. The rising portion of the curve
from r 2 to r 1 indica tes increasing repulsion of an a-particle as it approaches
the nucleus. Close to the nucleus and inside it, the shape of the potential
energy curve is not known with certainty, but the Coulomb potential must
break down and be replaced by an attractive potential. The interaction
between the nucleus and the a-particle in the region of uncertainty may be
represented by a constant attractive potential U 0 , exerted over a distance
To called the effective ,-adius of the nucleus. This type of potential is spoken
of as a potentíal well of depth U O and width or range r 0 • Under classical
mechanics, the a-particle could exist inside the potential well, with a
kinetic energy equal to E,. - U. (The horizontal line E.,. in the figure
represents the a-disintegration energy when the particle is far from the
13-61 THE THEORY OF ALPHA-DECAY 329
U(r) = 2ze2
,,-.-'
r
'
1
••
1
1
1
'
' 1
1
''
1
1
1
1
: Ea
-~------------------
! 1
ro r¡
FIG. 13-14. Alpha-particle emission: potential energy of a nucleus and an

a-particle as a function of their separation.
nucleus.) To escape from the nucleus, the particle would have to havé a
kinetic energy at least as great as the value of the energy at the maximum
of the potential energy curve. Similarly, an a-particle approaching the
nucleus from outside could penetrate the nucleus only if it had enough
kinetic energy to get over the potential energy barrier.
Thus, the maximum value of the potential energy curve corresponds to
the value of the kinetic energy which an a-particle must have according to
classical mechanics to get into the nucleus from outside, or to escape from
inside the nucleus. This maximum value must be greater than about
9 Mev, since a-particles from ThC' are scattered by uranium. The
uranium nucleus, however, emits an a-particle with an energy of about
4 Mev, and it is very difficult to understand how the particles contained in
the inside of the nucleus can go over a potential barrier which is more than
twice as high as their total energy. Classical physics provides no solution
to this problem, nor can it account for the extremely large variations in
half-life corresponding to small changes in energy, i.e., the Geiger-Nuttall
rule.
The paradox was resolved when the problem of a potential barrier such
as that involved in a-emission or a-scattering was analyzed by the methods
of wave mechanics. The mathematical treatment will not be given, but
the results and their implications will be discussed. If the motion of a

particle in the neighborhood of a potentia.l barrier is treated wave-mechan-
ically, i.e., if the Schroedinger equation for an a-particle is solved, it is
found that there is a fisite probability that the particle can leak through
the barrier even though its kinetic energy is less than the potential energy
represented by the height of the barrier. In other words, a particle im-
pinging on the barrier will not necessarily be reflected, but may pass
through the barrier and continue its forward motion. Gamow< 23 > and,
independently, Condon and Gurney, <24 > in 1928, showed how this comes
about. The probability that an a-particle can leak through the barrier
(the "tunnel effect ") can be calculated; the permeability of a simple barrier
is given mathematically by an expression of the type
2v'2M
p = exp [ - -1'&- J (U - ]
Ea) 112 dr , (13-19)
where I'& = h/21r, M is the mass of the a-particle, and the integral is taken
over the entire region in which U(r) > Ea. The wave-mechanical treat-
ment of a-decay yields an expression for the disintegration constant
which depends on the permeability or transparency of the barrier. An
approximate form of the result is
2
Va [ S(Z - 2)e ( • )]
X = ro exp - l'&va a0 - sm a 0 cos a 0 , (13-20)
where
Mv2r ]112
ªº = are cos [ 4e2(Z ª_!! 2) i
r 0 is the radius of the nucleus or, more strictly, the "effective radius of
the nucleus for a-decay"; Va is the velocity of the a-particle relative to
the nucleus, and is equal to v (1 +
(M/Mr)], where v is the measured
velocity of the a-particle and M r is the mass of the recoil nucleus. The
quantity Ea is the alpha-disintegration energy given by
Ea = E (1 + M/Mr)J,
and (Z - 2) is the charge on the product nucleus.
The expression for the disintegration constant may be considered to con-
sist of two factors. One, va/ro, gives in a rough way the number of times
per second an a-particle inside the nucleus collides with the potential bar-
rier; the exponential factor gives the probability that the a-particle will
leak through the barrier. Equation (13-20) has sorne interesting proper-
ties. Since Xis proportional to an exponential function, a relatively small
13-6] THE THEORY OF ALPHA-DECAY 331
change in the argument of the exponential may cause a large change in the
value of >... The argument of the exponential depends on the velocity of the
a-particle and a small change in the velocity affects the value of >.. strongly.
This effect may be seen more directly when sorne numerical values are
inserted. Consider the factor 8e 2 (Z - 2)/hv,. in the exponential (the func-
tion a 0 - sin a 0 cos a 0 is of the order unity and varies slowly with v,.).
The value of e is 4.80 X 10- 10 esu, so that e2 = 2.304 x 10- 19 , lí =
1.054 X 10-21 . For Z - 2 = 90, the factor is
8(2.304) 10- 19 (90) 157.5 X 10 9
(1.054)10-27v., v..
Values for the velocities of a-particles range from 1.4 to 2.2 X 10 9 cm/sec,
so that the value of the factor is of the order of magnitude 100. A small
change in the value of v., or v will, therefore, have a large effect on the value
of the disintegration constant. lt is also seen that the greater the velocity
of the a-particle, the greater is the value of the disintegration constant,
and the theoretical expression agrees in a general way with the observed
facts. Equation (13-20) represents a highly important result of the applica-
tion of wave mechanics to nuclear physics.
The theory also offers a clue to the meaning of the Geiger-Nuttall rule.
Equation (13-20) becomes, when the common logarithm of both sides is
taken,
8e 2 (Z - 2) ( .
l og >.. = 1og
v..
ro -
2.303/ív., ªº - sm ªº cos ao); (13-21)
the factor 2.303 enters because of the use of common logarithms. lt can
be shown that
f
a 0 - sin a 0 cos a 0 ""' + terms of higher order,
and, to a good approximation,

2
1
og >.. = og ro
1 v..
- 41re2.303/ív.,
(Z - 2) 8e [
+ 2.303h (Z - 2)roM]
1/2
+. . . . (13-22)
The variation of log v.,/r 0 is small because both r 0 and v., vary relatively
little, and the higher order terms in Eq. (13-22) are small. All of these
terms may be lumped into one term which is practically a constant. Then
(z - 2)
log >.. ""' a - b - - ~ , (13-23)
v..
where a and b are constants. According to this equation, if the logarithm

of the disintegration constant is plotted against the reciproca! of the
.., o
o:
!l
.,
o: -2
o
"o: -4
o
~...
-6
1o:
·¡¡¡
-8
st,j)
.s -10
-12
-14
-16
-18
0.45 0.55 0.65 0.75
Reciproca! alpha velocity
Fm. 13-15. Logarithm of the disintegration constant vs. reciproca! of the
velocity, for constant Z. Disintegration constant >. in sec- 1 ; velocity in units of
10 9 cm/sec< 25 >.
a-velocity for the a-emitting isotopes of a radioactive element (Z constant),

the result should be a straight line. The equation may be regarded as a
theoretical form of the Geiger-Nuttall rule and may conveniently be com-
pared wjth experiment. A comparison has been made by Kaplan, <25 >
who showed that a-emitters with even values of A and Z give results which
are in good agreement with the theory. It can be seen from Fig. 13-15
that when log >. is plotted against 1/va, for even Z and A, straíght lines are
indeed obtained; the slopes of the lines are in good agreement wíth the
values predicted by the theory. The results can also be shown in a graph
like that of Fig. 13-16 in which the logarithm of the half-life is plotted
against the "effective total decay energy," Qerr' 26 >; the latter is just the
a-disintegration energy with a small correctíon for the screening effect of
the atomic electrons. A square-root scale is used for Q.rr; the points are
experimental, and the last digit in the mass number of the a-emitter is
given beside each point. The figure contains information more recent
than that of Fig. 13-15 so that there are more experimental data; again,
very good straíght lines are obtained. The relationships between the
13-6] THE THEORY OF ALPHA-DECAY 333
18
16
14
12
~
! 8
""
',
e;:' 6
o
.,;¡; 4
.Q
Q,tr(Mev)
Frn. 13-16. Half-life for alpha-decay as a function of the square root of the
effective total decay energy, Qe11• The energy scale is linear in Q.n- 1' 2 ; Qe11
includes the electron screening correction. The numbers give the last digit of the
mass number of the alpha-emitter; for example, 4 on the uranium scale refers to
u2a4. <26)
logarithm of the half-life and the a-disintegration energy of the elements

with even values of Z are shown in Fig. 13-17. There are analogous curves
for even-Z-odd-A nuclides, and for odd-Z nuclides.
The above discussion has been limited to the case of the even-even
a-emitting nuclides in order to avoid complications which are beyond the
scope of this book. The application of the theory to a-emitters with odd
Z and even A, even Z and odd A, and odd Z and odd A is less straight-
forward, in part because the probability of a-emission depends on the
difference in the angular momenta of the parent and daughter nuclei, and
most of the values of the momenta are not yet known when either A or Z
is odd. For this reason and others, the theory of a-decay cannot yet be
considered a closed subject, although it has achieved an important degree
of success.
If the energy and the disintegration constant are known for a given
nucleus, Eq. (13-20) can be used to calculate the value of r 0 , the effective
radius of the nucleus for a-emission. It must be pointed out that the
TABLE 13-7
NUCLEAR RAou FROM ALPHA-DECAY: EvEN-EVEN ALPHA-EMtTrERS
Radius of
Parent Product product Ro = roA- 113,
nucleus nucleus nucleus, cm X 1013
cm X 10 13
cm242 Pu2as 9.32 1.50

Cm240 Pu236 9.29 1.50
Pu240 u236 9.40 1.52
Pu238 u234 9.30 1.51
Pu236 u2a2 9.30 1.51
u2as Th234 9.54 1.55
u2a6 Th2a2 9.36 1.52
u2a4 Th2ao 9.39 1.53
u2a2 Th22s 9.33 1.53
u2ao Th226 9.33 1.53
Th2a2 Ra228 9.48 1.55
Th2ao Ra226 9.34 1.53
Th22s Ra224 9.33 1.54
Th226 Ra222 9.34 1.54
Ra226 Em222 9.35 1.54
Ra224 Em220 9.33 1.55
Ra222 Em21s 9.30 1.55
Em222 p 0 21s 9.34 1.55
Em220 p 0 216 9.36 1.57
Em21s p 0 214 9.48 1.58
p 0 21s Pb214 9.23 1.54
p 0 216 Pb212 9.19 1.54
Po214 Pb210 9.18 1.55
p 0 212 Pb2os 9.05 1.53
radius of the nucleus is a quantity which cannot be measured directly and

is obtained by calculation from sorne formula which also contains measured
quantities. The particular formula used depends on the experimental
process under consideration, and the value of the radius obtained must be
regarded as an effective value for the given process. Fortunately, the
values obtaíned from different processes agree quite well, and lie between
10- 13 and 10- 12 cm. The heavy a-emitting nuclei have radii close to
10- 12 cm, as can be seen from Table 13-7. The values of the radii given
are those of the product nuclei, and were obtained from a more rigorous
form of the theory. The radii are found to conform to the formula
ro = RoA 113 X 10- 13 cm,
13-6) THE THEORY OF ALPHA-DECAY 335
12 Th 232
12 10 V, U""
JO .i'/ -JO'ºy Th230
Puuo
8 J0 8y 0 23, u2u
p 0 23s
G JOºy
Pu 236
4 JO'y
2 IOOy
Cm2o12
.,;:: O ly Ha ...--Cm.,º
i
{
-2 ::::.~
-4 t hr
Em...
.,_. J>u234 Th22&
u22•
na2H Po218 ¿/Y"" Ra220
'"a -G 1 min Em22
u2ao
o 1 sec
~ -8 100 ms
- -10 !Oms
IOO¡,.s 1 ms
-12 10¡,.s
-14 l¡,.s
0.1 ¡,.s p 0 212
4.0 -l.5 5.0 55 O.O 0.5 7.0 7.5 8.0 8.i\ 9.0
..\lpha disintegration energy (Mev)
Fm. 13-17. Relationship between half-life and a-disintegration energy for

sorne even-even a-emitters (Perlman, Ghiorso, and Seaborg<27 >).
where Ro is a constant closely equal to 1.53 and A is the atomic weight.

The radius turns out to be proportional to the cube root of the atomic
weight, a fact which will be useful in later work.
There is, finally, a considerable amount of work, which has only been
touched upon, on the regularities among the properties of a-emitters.II
This work, sometimes referred to as the "systematics of a-emitters has
for its object the correlation of the available experimental information
about a-emission. It is well summarized in the general references by Perl-
man and Rasmussen and by Hanna.
REFERENCES
GENERAL
RuTHERFORD, CHADWICK, and ELLIS, Radiations from Radioactive Substances.
New York: Macmillan, 1930, Chapters 2, 3, 4, 5, 7.
16, 22.
l. PERLMAN and J. O. RASMUSSEN, "Alpha Radioactivity," Handbuch der
Physik, Vol. 42. Berlin: Springer Verlag, 1957, pp. 109-204.
G. C. HANNA, "Alpha Radioactivity," Experimental Nuclear Physics, ed. by
E. Segre. New York: Wiley, 1959, Vol. 111, Part IX.
F. AsARO and l. PERLMAN, "Table of Alpha-Disintegration Energies of the
Heavy Elements," Revs. Mod. Phys. 26, 456, 1954; 29, 831, 1957.
H. A. BETHE and J. AsHKIN, "Passage of Radiation Through Matter,"
Experimental Nuclear Physics, E. Segre, ed. New York: Wiley, 1953, Vol.
I, Part 11.
D. STROMINGER, J. M. HoLLANDER and G. T. SEABORG, "Table of Isotopes,"
Revs. Mod. Phys. 30, 586-904, 1958.
PARTICULAR
l. RuTHERFORD, WYNN-WILLIAMS, LEw1s, and BoWDEN, "Analysis of a-Rays
by an Annular Magnetic Field," Proc. Roy. Soc. (London) Al39, 617 (1933).
2. (a) G. H. BRIGGs, "A Determination of the Absolute Velocity of the a-
Particles from Radium C'," Proc. Roy. Soc. (London) Al57, 183 (1936); (b)
"Energies of Natural a-Particles," Revs. Mod. Phys. 26, 1 (1954).
3. S. RosENBLUM and G. DuPOUY, "Mésure Absolue des Vitesses des Princi-
paux Groupes des Rayons a," J. phys. et radium 4, 262 (1933).
4. A. H. JAFFEY, "Radiochemical Assay by a- and Fission Measurements,"
Manhattan District Declassified Document MDDC-1336, Feb. 6, 1947; also
in National Nuclear Energy Series, Plutonium Project Record, Vol. 14A, The
Actinide Elements, G. T. Seaborg and J. J. Katz, eds. New York: McGraw-Hill,
1954, Chapter 16.
5. M. G. HoLLOWAY and M. S. L1v1NGSTON, "Range and Specific Ionization
of Alpha-Particles," Phys. Rev. 54, 18 (1938).
6. N. FEATHER and R. R. NIMMO, "The Ionization Curve of an Average a-
Particle," Proc. Cambridge Phil. Soc. 24, 139 (1928).
7. RuTHERFORD, WARD, and W YNN-W ILLIAMS, "The Ranges of the a-Particles
from the Radioactive Emanations and 'A' Products and from Polonium,"
8. l. CuRIE and S. T. Ts1EN, "Parcours des Rayons a de I'Ionium", J. phys. et
radium 6, 162 (1945).
9. W. E. GLENN, JR., "Pulse Height Distribution Analyzer," Nucleonics 4,
50 (June 1949).
10. GmoRso, JAFFEY, RoBINSON, and ,vE1SSBORD, "A 48-Channel Pulse-
Height Analyzer for Alpha-Energy Measurements," The Transuraníum Ele-
ments, ed. by Seaborg, Katz, and Manning. National Nuclear Energy Series,
Vol. 14B, Book 2, p. 1226. New York: McGraw-Hill, 1949.
REFERENCES 337
11. JEssE, FoRSTAT, and SADAUSKIS, "The Ionization in Argon and in Air by
Single a-Particles as a Function of their Energy," Phy8. Rev. 77, 782 (1950).
12. H. YAGODA, Radioactive Measurements with Nuclear Emulsions. New York:
Wiley, 1949.
13. GAILAR, SEIDLITZ, BLEULER, and TENDAM, "Range-Energy Relations for
Alpha-Particles and Deuterons in the Kodak NTB Emulsion," Rev. Sci. lnstr.
24, 126 (1953).
14. F. BLOCH, "The Theory of Stopping Power," in Lecture Series in Nuclear
Physics, Manhattan District Declassified Document MDDC-1175, Lectures
11, 12.
15. M. S. LIVINGSTON and H. A. BETHE, "Nuclear Physics, Part C. Nuclear
Dynamics, Experimental," Revs. Mod. Phys. 9, 245 (1937), Section XVI.
16. H. A. BETHE, "The Range-Energy Relations for Slow Alpha-Particles and
Protons in Air," Revs. Mod. Phys. 22, 213 (1950).
17. W. P. JEssE and J. SADAUSKIS, "Range-Energy Curves for Alpha-Particles
and Protons," Phys. Rev. 78, 1 (1950).
18. E. RuTHERFORD and A. B. Wooo, "Long-Range Alpha-Particles from
Thorium," Phi!. Mag. 31, 379 (1916).
19. W. B. LEWIS and B. V. BowDEN, "An Analysis of the Fine Structure of
the a-Particle Groups from ThC and of the Long-Range Groups from ThC',"
20. S. RosENBLUM, "Progres Récents dans l'Etude du Spectre Magnétique
des Rayons a," J. phys. et radium 1, 438 (1930).
21. G. GAMOW, "Fine Structure of a-Rays," Nature 126, 397 (1930).
22. H. GEIGER and J. M. NuTTALL, "The Ranges of the a-Particles from
Various Radioactive Substances and a Relation Between Range and Period of
Transformation," Phi!. Mag. 22, 613 (1911); 23 1 439 (1912); 24, 647 (1912).
23, G. GAMOW, "Zur Quantentheorie des Atomkernes," Z. Phys. SI, 204 (1928).
24. R. W. GuRNEY and E. U. CoNDON, "Quantum Mechanics and Radio-
active Disintegration," Nature 122 1 439 (1928); Phys. Rev. 33, 127 (1929).
25. l. KAPLAN, "The Systematics of Even-Even Alpha-Emitters," Phys.
Rev. 81, 962 (1951).
26. C. J. GALLAGHER, JR., and J. O. RAsMUSSEN, "Alpha-Decay Hindrance
Factor Calculations," J. lnorg. Nucl. Chem. 3, 333 (1957).
27. PERLMAN, GHioRso, and SEABORG, "Systematics of Alpha-Radioactivity,"
Phys. Rev. 77, 26 (1950).
PROBLEMS
l. Calculate the mass, relative to that of the rest mass, the kinetic energy
in ergs and Mev, and the value of Hr for a-particles moving at each of the
following speeds (in cm/sec): 1 X 108, 5 X 108 , 7.5 X 108 , 1 X 109, 1.25 X 109 ,
1.50 X 109 , 1.75 X 109 , 2.0 X 109 , 2.5 X 109 , 3.0 X 109 , 5.0 X 109, 1.0 X 10 10,
and 2.0 X 10 10 . Plot the energy in Mev against the value of Hr in gauss-cm.
2. Find the speeds of a-particles with energies of 1 Mev, 2 Mev, 4 Mev,
5 Mev, 6 Mev, 8 Mev, 10 Mev, respectively.
3. Find the speed and Hr-value of the a-particles emitted by each of the
isotopes of polonium listed in Table 13-3.
4. In an investigation of the Al 27 (d,a)Mg 25 reaction, an aluminum target
was bombarded with 2.10-Mev deuterons. The a-particles coming off at 90°
were analyzed in a 180° magnetic spectrograph, and 10 different groups of a-
partides, each with a different Hr-value, were observed. The Hr-values were, in
kilogauss-cm: 393, 379, 369, 354, 345, 329, 325, 323, 305, and 290. Find the
Q-value corresponding to each group of a-particles. (The interpretation of
the occurrence of the different a-particle groups is treated in Chapter 16.)
5. Plot the values of the kinetic energy listed in Table 13-3 against those
of the mean range in air. Use the resulting curve and those of Fig. 13-9 to
find the mean ranges in air of the a-particles of Problema 1 and 2, which fall
within the energy limits covered by the curves.
6. lf the range in air is known, the range in another substance, denoted by
the subscript x, may be obtained, to a first approximation, from the empirical
formula
Rz = RairX
Nair [ Zz
N,. X 0.563 (Z,.+ 10)112
]-l '
where the N's .represent the number of atoms per cubic centimeter and Zz is the
atomic number of tbe substance x. The density of air may be taken as 0.001226
gm/cm3 and the "atomic weight of air" as 14.4. Calculate (a) the range in
aluminum of the a-particles from each isotope of polonium listed in Table 13-3
(use the mean range in air), (b) the relative stopping power in aluminum, (e)
the equivalent thickness in mg/cm 2, (d) the thickness in mg/cm 2 equivalent
to one cm of air. Compare the results with the experimental values given in
Table 13-2.
7. The nuclide Rn 211 emits three groups of a-particles, with kinetic energies of
5.847 Mev, 5.779 Mev, and 5.613 Mev, respectively. Associated with the a-
particles are 'Y-rays with energies of 0.0687 Mev, 0.169 Mev, and 0.238 Mev.
Construct a decay scheme based on these data. Compare it with the scheme
given by Strominger, Hollander, and Seaborg (gen. ref.).
8. The nuclide U 233 emits six groups of a-particles, with kinetic energies of
4.816 Mev, 4.773 Mev, 4.717 Mev, 4.655 Mev, 4.582 Mev, and 4.489 Mev,
respectively. Gamma-rays with energies 0.0428 Mev, 0.0561 Mev, and 0.099
Mev have also been reported. Construct a decay scheme based on these data.
9. The nuclide Am 241 emits six groups of a-particles, with kinetic energies of
5.534 Mev, 5.500 Mev, 5.477 Mev, 5.435 Mev, 5.378 Mev, and 5.311 Mev,
PROBLEMS 339
respectively. Gamma-rays are found, with energies of 0.0264 Mev, 0.0332 Mev,
0.0435 Mev, 0.0555 Mev, 0.0596 Mev, 0.103 Mev, 0.159 Mev. Construct a
decay scheme based on these data.
10. The height U of the Coulomb barrier around a nucleus of charge Z1e
and radius r for a particle of positive charge Z2 may be found to a first approxi-
mation by calculating the energy of Coulomb repulsion at a distance equal to
the radius of the nucleus,
r Assume that the nuclear radius is given by the formular = 1.5 X 10- 13 A 113 cm,
where A is the mass number of the nucleus. Calculate the height of the barrier
fora-particles and the nuclei Ne 20, Ca40, Zn 66, Sn 112, Yb 174, and Th 232 • Repeat
the calculation for protons and deuterons.
11. The artificially produced radioactive nuclide Pa 230 may be considered
the parent of a collateral (4n + 2) series. Find its radioactive properties from
the nuclide chart or the "Table of Isotopes" of Strominger, Hollander, and
Seaborg. Trace the series by means of a graph of A against Z (cf. Fig. 10-10),
_and show its relationship to the uranium series.
12. It has been shown <25> that the a-emission of the thorium isotopes can
be described by the relation
log A = 56.13 - 105.07 /va.
The a-energy in the case of Th 224 is 7.33 Mev; what is the half-life of Th224 as
estimated from the above relation?
13. What is the Q-value of the reaction Po 21 º -+ Pb 2 º 6+ He4 ? If the atomic
mass of Pb 2 º6 is 206.0386, what is the atomic mass of Po 21 º?
CHAPTER 14
BETA-DECAY
The study of the properties of the ,S-particles emitted by the natural

and artificial radioactive nuclides has added greatly to our knowledge of
the structure and properties of atomic nuclei. As in the case of a-particles,
the early interest was centered on the series of transformations of uranium,
thorium, and actinium. At present, the main interest is in the information
that ,S-particle emission gives about nuclear energy levels and decay
schemes, and in the theory of ,S-decay and its relation to other funda-
mental nuclear problems. Many more nuclides decay by electron emis-
sion, positron emission, and orbital electron capture than by a-particle
emission, and the P-decay processes yield information about hundreds oí
nuclear species not limited to those with large masses. Consequently, infor-
mation about the energy levels and decay schemes of light and intermediate
weight nuclides as well as those in the region of the natural radioactive
elements can be obtained by studying their f:3-radiations. This study is
an important part of nuclear spectroscopy.
The emission of P-particles differs from that of a-particles in respect
to the spectrum of the energies of the emitted particles. The most char-
acteristic feature oí the spontaneous ,S-disintegration of a nucleus is the
continuous distribution in energy of the emitted electrons, which is in
sharp contrast to the line spectra observed for a-particles. The continuous
energy distribution led, as will be seen, to serious theoretical problems.
These problems have been treated with considerable success, but only by
means of a theory involving new and radical ideas. Beta-decay is impor-
tant, therefore, not only because of its relationship to the practica! problems
of nuclear physics, but also because of the conceptual problems involved.
14-1 The velocity and energy of beta-particles. The velocity, or

momentum, of ,S-particles can be measured by means of the deflection of
the path of the particles in a magnetic field. One kind of instrument
which is often used is the semicircular-focusing, or 180º, magnetic spectro-
graph similar in principie to the instrument used for a-particles (Fig.
13-1). The details of the design of the two instruments are different
because of the differences between the properties of a- and ,S-particles.
For example, the value of the charge-to-mass ratio is much greater for
electrons than for a-particles and much smaller magnetic fields may there-
fore be used to deflect f:3-particles. Insteaci of fields of about 10,000 gauss
340
14-1] THE VELOCITY AND ENERGY OF BETA-PARTICLES 341
as in the case of the a-particle spectrograph, fields of 1000 gauss or less

are strong enough. In the earlier instruments, the magnetic field was
usually uniform, m but in the more recent designs the focusing of the
particles has been improved by using a suitably inhomogeneous, or
"shaped," field. <2 > The velocities, or momenta, of the ¡3-particles emitted
by a radioactive source can be measured with high precision with these
instruments, and when the velocities are known, the energies can be
computed.
It was found in early experiments that ¡3-particles emitted by naturally
radioactive nuclides may have velocities up to about 0.99 that of light.
In general, the energies of the ¡3-particles, both positive and negative, are
smaller than those of the a-particles emitted by radioactive nuclides.
Most of the ¡3-particles have energies smaller than 4 Mev, while nearly
all of the a-particles have energies greater than 4 Mev. At the same
kinetic energy, the ¡3-particle, because of its much smaller mass, travels
much faster than the a-particle. An a-particle with an energy of 4 Mev
has a velocity about 1/20 that of light, but a 4-Mev electron would have
a velocity close to 0.995 that of light.
The ¡3-particles must be treated relativistically because of their large
velocities. The formulas for the velocity and kinetic energy, analogous
to Eqs. (13-5) and (13-6) for a-particles, are
v = Hr-
mo
eR2 1 - -,
c2
(14-1)
T= m 0 c2 [ R 1
1 - -
2 -
(14-2)
c2
. where m0 is the rest mass of the electron and e is the magnitude of the
electronic charge. It is useful to express the quantities Hr and T as
functions of the ratio v/c and Eq. (14-1) may be written
V e r:--;;z
e= Hr moc '\J l - c2. (14-3)
The specific charge of the electron, in electromagnetic units, is e/m 0 =

10
1.75888 X 10 7 emu/gm and the velocity of light is 2.99793 X 10 cm/sec.
Then
Hr (gauss-cm) -- moc _v
e e
(1 -
_v2)-112
c2
(14-4)
V ( V2)-1/2
= 1104.5 e 1 - c2 .
100,000 ~
t.:>
~
~
"
vv
¡ 10,000
,,
/
~ /
ti: ~
~
:a
·¡;¡,
.,,,, .,,, ti,
l".!
~
·.: V ¿,
.,"
·.;:i
5'
'1
1000
~
/ s
l".!
.,,,
....,
V
.,,..,,,.
.,,.
100
0.001 0.01 0.1 10 ñ
Kinetic energy T (Mev)
~
Fm. 14-1. Kine.tic energy of P-particles as a function of thc magnctic rigidity Hr. -""
TABLE 14-1
MAGNETIC RIGIDITY (Hr) AND KINETIC ENERGY OF
BETA-PARTICLES AS FuNCTIONS OF VELOCITY
Velocity, Magnetic rigidity, Kinetic energy,

v/c Hr: gauss-cm T:Mev
o o o
0.10 171.3 0.00258
0.15 258.6 0.00585
0.20 347.9 0.01054
0.25 440.1 0.01676
0.30 536.0 0.02468
0.35 636.9 0.03450
0.40 743.9 0.04655
0.45 858.9 0.06121
0.50 984.1 0.07905
0.55 1122 0.10085
0.60 1278 0.1278
0.65 1458 0.1614
0.70 1671 0.2045
0.75 1933 0.2616
0.80 2273 0.3407
0.85 2750 0.4590
0.90 3519 0.6613
0.92 4001 0.7928
0.94 4696 0.9868
0.96 5844 1.314
0.98 8394 2.057
0.99 11,960 3.111
0.995 16,980 4.605
0.996 19,000 5.208
0.997 21,960 6.901
0.998 26,910 7.573
0.999 38,085 10.92
The quantity m0 c2 is the energy associated with the rest mass of the elec-
tron and is 0.5110 Mev. Hence, the kinetic energy is
1
= 0.511 [ 1] · (14-5)
T v -
~2
'\}1-c2
Values of Hr and T are listed in Table 14-1 for different values of v/c,
and the kinetic energy T is plotted as a function of Hr in Fig. 14-1.
344 BETA-DECAY [CHAP. 14
9
0
o 0.1 0.2 o.3 0.4 o.s o.6 0.1 o.s . . .1 1.2
Kinetic energy of the beta-particles, T (Mev)
F10. 14-2. The P-pa.rticle spectrum of Ra.E (Nea.ry< 9>).
Hr (gauss-cm)
F10. 14-3. Beta.-pa.rticle spectrum of Au 198, showing the line spectrum.

[Fa.n, Phys. Rev. 87, 258 (1952).]
The results obtain~d when the P-particles (electrons) from a radio-

active nuclide are studied with a magnetic spectrograph are more complex
than those obtained with a-particles. lf the P-particles, after being
deflected, are allowed to fall on a photographic plate, a fogging of the
plate is observed which extends for a definite distance and then ceases
at a point characteristic of the ,8-emitting nuclide. This fogging is con-
tinuous, but of nonuniform intensity along the plate. In addition to
this continuous ¡,rimary spectrum, several sharp lines are sometimes seen
on the plate at positions characteristic of the emitting substance. These
lines form the so-called secondary, or line spectrum, observed for many,
but not ali, ¡3-emitters. It is difficult, because of the continuous spectrum,
to determine the numbers of ¡3-particles with different energies from the
darkening of the photographic plate. A velocity distribution curve (or
an energy distribution curve) can be obtained by using a Geiger couriter
as the ¡3-particle detector instead of a photographic plate; when this is
done, the instrument is more properly called a spectrometer. The counter
is placed in a fixed position (i.e., the value of r is fixed), the magnetic
field is varied, and the number of ¡3-particles reaching the counter per
unit time is obtained for different values of H. Since each value of H
corresponds to a different value of Hr and hence to a different value of
the kinetic energy, the numbers of particles corresponding to different
energy values are obtained. The continuous spectrum of ¡3-particle
energies found for RaE is shown in Fig. 14-2. In this case no line spectrum
is found and the upper limit or maximum energy is at 1.17 Mev. When
a line spectrum is also present, the lines appear as distinct peaks super-
imposed upon the continuous distribution curve, as in Fig. 14-3.
Beta-spectra will be discussed in greater detail in Section 14-4. For
the present it will be noted that the spectrum of RaE shows the most
important features of continuous spectra. Every continuous ¡3-spectrum
has a definite maximum, the height and position of which depend on the
nucleus emitting the particles. There is also a definite upper limit of
energy for the particles emitted by a nuclide; this upper limit, or endpoint,
of the continuous spectrum is different for different nuclides. The shape
of the curve at the lower energies is not known with certainty because
it is hard to make accurate measurement on the low-energy particles.
These properties of continuous spectra are observed for both natural and
artificial ¡3-emitters, and when the emitted particles are electrons or
positrons.
The sernicircular-focusing magnetic spectrometer is just one of several
types of instruments for the analysis of beta-rays. Another type, often
used, is the solenoidal magnetic spectrometer which is a variety of magnetic
lens spectrometer; an example< 3 , 18 > is shown in Fig. 14-4. It consists of
a long solenoid with the source of ¡3-radiation on the axis near one end and a
thin-walled Geiger-Mueller counter at the other end, about 90 cm from
the source. The solenoid produces a homogeneous magnetic field whose
direction is parallel to the axis of the cylinder. The field forces the particles
to move along helical paths defined by a series of disks and rings, which
act as bafHes. By adjusting the current in the solenoid, and hence the
magnetic field, ¡3-particles from the source can be focused so that they
make a single turn of the helix and just reach the counter. For a given
value of the magnetic field only those particles which have a particular
velocity will reach the counter. The velocity can be calculated frorn the
Baffle system
WJJJJ
Shutter control
10 cm
Fm. 14--4. Schematic diagram of a solenoidal magnetic spectrometer.< 3 - 18>
magnetic field and the dimensiona of the apparatus. In practice, however,

instrumenta of this type are usually calibrated by means of electrons of
known speed, and the velocity can then be determined directly from the
current through the solenoid. In this way, the relative numbers of /j-
particles with different velocities can be determined.
The instrument just described has a relatively long, uniform magnetic
field and is called a úmg l,ens magnetic spectrometer. Because of the
difficulty of producing a really uniform magnetic field, the so-called short
lens magnetic spectrometer is often used. In this case, the magnetic field
is not uniform, but may be stronger in the region around the axis than at
the axis, or may vary along the axis. The analysis of the motion of the
/j-particles is somewhat more complicated than in the case of the uniform
field, but velocity and energy distributions can again be obtained with the
help of suitable calibration. For further details about these and other
instrumenta, the reader is referred to a recent review. <4 >
14-2 The absorption of beta-particles. Range, ionization, and energy

loss. The energies of tJ-particles can be determined from measurements
of their absorption in matter,< 5 • 5> as was the case with a-particles. Beta-
particles are much more penetrating than a-particles and there are marked
differences in the methods used for measuring the absorption in the two
cases. An a-particle with a kinetic energy of 3 Mev has a range in standard
air of about 2.8 cm and produces about 4000 ion paira/mm of path. A
3-Mev IJ-particle has a range in air of over 1000 cm and produces only
about 4 ion paira/mm of path. Although the use of airas an absorbing
medium for tJ-particles is therefore impractical, the particles are penetrat-
ing enough so that solid absorbers can be used.
Aluminum foils are the most frequently used absorbers, though gold and
mica are often used. In absorption experimenta, the absorber foils are
14-2] THE ABSORPTION OF BETA-PARTICLES 347
Amplifier and
sca.ling circuít
·· ·:: _Beta-particle counter

Absorber foils
Al shelf
arrangement
Support for
source
Fm. 14-5. Diagram of an experimental arrangement for measuring the

absorption of ¡3-particles (Glendenin< 5>).
placed between the active source and a thin-windowed detector, usually

a Geiger counter oran ionizatíon chamber, and the counting rate is deter-
mined as a function of the absorber thickness. A conveníent experi-
mental arrangement is shown schematically in Fig. 14-5. A typical
absorption curve is shown in Fig. 14-6 for the partícles which form the
continuous ¡3-spectrum. The counting rate, plotted on a logarithmic
scale, decreases linearly• or very nearly so, over a large fraction of the
absorber thickness. In other words, the number of ¡3-particles decreases
exponentially with absorber thickness, to a good approximation, and the
absorption may be represented by the formula
(14-6)
where A 0 is the counting rate, or activity, without absorber, A(x) is the
activity observed through a thickness x, and µ. is the absorption coefficient.
The actívíty does not decrease to zero as the absorber becomes very thíck
but becomes practically constant at a value which represents the so-called
"background." There is always sorne radiatíon present which contributes
to the countíng rate even though ít does not represent ¡3-partícles from the
source. Near íts end, the absorptíon curve deviates from the exponential
form, and the point at which it meets the background is called the range
R/J of the ¡3-particles.
The exponential form of the curve is accidental, since it also includes
the effects of the continuous energy distributíon of the '3-particles, and
of the scattering of the partícles by the absorber. The range Rfl is the
distance traversed by the most energetíc partícles emitted, and corresponds
to the energy at the endpoínt of the continuous spectrum. It is possible
to obtain range-energy curves for electrons in different absorbers, and the
1000~----------,
E> 0.2 :l.lev
.\bsorber thickness (mg, cmt)

(linear scale) Total ahsorber (mg/rm 2)
Frn. 14-6. A typical absorption Frn. 14-7. A typical absorption
curve for /3-particles in aluminum. The curve for monoenergetic /3-particles
J;l-particles are accompanied by 'Y-rays (Glendenin< 5 >).
which form part of the background.
endpoint energies of /3-spectra can then be determined by absorptlon

measurements. The absorption method does not give results comparable
in accuracy to those obtained with the magnetic spectrometer, but it
gives results of fair accuracy and has the advantages of simplicity and
speed.
Thicknesses in absorption measurements are often given, as índicated
in Fig. 14-6, in units of grams per square centimeter or milligrams per
square centimeter of absorber. The actual thickness is multiplied by the
density, and the resulting quantity, which may be called the su:perjicial
density, is used as a measure of the thiclrness. This usage has certain
practica! advantages. It is often inconvenient to measure the thickness
of very thin foils, but their weight and cross section can be measured easily,
and the weight divided by the cross section is equal to the product of
the density and the thickness. Furthermore, it turns out that if the amount
of absorber is expressed as the product of the density and thickness, the
range is nearly independent of the nature of the absorber. This result
follows from the fact that the range of a given /3-particle depends not only
on its initial energy, but also on the number of electrons with which it
collides in passing through the absorber. The latter number depends on
the density oí electrons in the absorber or on the number of electrons per
unit mass. The ability of an element to stop /3-particles depends, there-
fore, on the ratio of the atomic number to the mass number, i.e., on Z/ A.
Although this ratio decreases from the light to the heavy elements, the
effect of this variation on the thickness needed to stop /3-particles is not
large. In the case of aluminum, with Z/ A = 13/27 = 0.48, the range
14-2] THE ABSORPTION OF BETA-PARTICLES 349
of a 1-Mev fJ-particle is about 400 mg/cm 2 ; in gold, with Z/A. 79/197

= 0.40, the range is about 500 mg/cm 2 , which is not very different from
the range in aluminum.
The range may be obtained from the absorption curve simply by inspect-
ing the curve, a method that is called the v1ºsual or inspection method.
To find where the curve meets the background many accurate experi-
mental points are needed close to the range thickness. Although this
method is the simplest way of analyzing the absorption curve, it is prob-
ably the least reliable. In any case in which the curve approaches the
background very slowly it is difficult to obtain good accuracy, and this
kind of absorption curve is met often. To avoid the difficulties of the
visual method, comparison methods have been developed< 0 > in which
the range of the fl-particles from a given nuclide is measured in terms of
the range of a standard emitter. The standard emitter is one with an
absorption curve that is particularly favorable for the visual method, and
the absorption curve of the nuclide whose range is sought is compared in
a detailed way with the curve of the standard substance. Still other
methods of analyzing absorption curves have been developed;< 0 > these
depend on fitting the experimental absorption curve with an empirical
formula which includes either the range or the endpoint energy as a
parameter.
The discussion so far may be applied to both fJ--particles (electrons) and
¡3+-particles {positrons). In the case of ¡3+-radiation the shape of the
absorption curve is much the same as for ¡3--particles with the exception of
certain differences in the background. It is possible to correct for these
differences, and ranges of ¡3+-particles can be determined with the methods
used for ¡3--particles.
The absorption of monoenergetic fJ-particles is somewhat different from
that of the particles which constitute the continuous spectrum. The parti-
cles which form a line in the discrete line spectrum often seen in fJ-decay
are monoenergetic fJ--particles and their absorption has practical impor-
tance. The absorption of monoenergetic electrons with energies greater
than about 0.2 Mev is practically linear, as in Fig. 14-7. If the count-
ing rate is plotted against the absorber thickness on ordinary graph paper,
the curve closely approximates a straight line over most of the absorber
thickness, with a small tail at the end of the curve. The linear portion of
the curve can be extrapolated to cut the thickness axis ata value called the
extrapolated or effective range. For monoenergetic electrons with energies
less than 0.2 Mev, the absorption curve deviates more and more from a
straight line as the energy decreases, and it becomes increasingly difficult
to determine the effective range by absorption methods.
Quantitative results for ranges of fJ-particles in aluminum will be given
in Section 14-3 in the course of the treatment of the range-energy curve
for electrons.
The details of the passa.ge of ,8-particles through matter are more com-
plicated than is the case with a-particles. The di.fferences are caused,
among otber things, by the much smaller mass of the ,8-particle and its
greater speed. A ¡'j-particle may lose a large fraction of its energy in a
single collision with an atomic electron, with the result that straggling is
much more marked in the case of electrons than with heavier particles.
Beta-particles are also scattered much more easily by nuclei than are
a-particles, so that their paths are usually not straight. Consequently,
even if a beam of ,8-particles is initially monoenergetic, the straggling and
scattering make possible widely di.fferent path lengths for particles passing
through the same thicknesses of absorber. The range is, therefore, a
less precisely defined quantity for electrons than for a-particles or protons.
The energy lost by an electron per unit path length can be calculated
from theory, and a formula has been derived for the energy lost by the
particles because of ionization of absorber atoms. It is more complicated
than the corresponding formula given for a-particles, Eq. (13-10), because
relativistic effects must be taken into account. The theory predicts in
a satisfactory way the dependence of the energy loss on the energy of
the ,8-particle and on the atomic number of the absorbing element.
For high energy ,8-particles, an additional mechanism for losing energy
must be taken into account. When an electron passes through the electric
(Coulomb) field of a nucleus, it loses energy by radiation. This energy
appears as a continuous x-ray spectrum called bremastrahlung or braking
radiation. The energy loss per unit path length because of radiation may
be denoted by (dT /dx)rad as distinct from (dT/dx);onia, the energy loss
by ionization. The ratio of the two losses is given approximately by
(dT/dx),ad TZ 1
(dT /dx)ioni• = 800
where T is the ,8-particle energy in Mev and Z is the atomic number of~·
the absorber. In heavy elements such as lead (Z = 82), the energy loss
by radiation is significant even for a 1-Mev particle; in aluminum (Z = 13)
the radiation loss amounts to only a few percent for the energies available
from ,8-emitters.
In general, it can be said that even though the interaction between
ti-particles and matter is considerably more complicated than is the case
for a-particles, the processes are well understood and there is good agree-
ment between tbeory and experiment.
14-3 Rang~nergy relations for beta-particles. When the ranges of

¡'j-particles are known from absorption experiments, a ran~nergy rela-
tion can be used to obtain the maximum energies of ,8-spectra. The
range-energy relation for aluminum has been determined and can be repre-
14-3) RANGE-ENERGY RELATIONS FOR BETA-PARTICLES 351·
TABLE 14-2*
ENERGIES AND RANGES OF BETA-PARTICLES
Endpoint
Nuclide Range, mg/ cm 2
energy, Mev
R 8 22s 0.053 6
Rbs1 0.13 20
Nb9s 0.146 30
Lul76 0.22 48
coso 0.31 81
Zr9s 0.400 122
BelO 0.555 181
J131 0.600 213
Sbl24 0.65 254
Mnss 0.73 277
Aul98 0.97 399
cu 0.98 447
Bal40 1.022 426
Mnss 1.05 462
Cdll5 1.13 527
Bi210 (RaE) 1.17 508
Nl3 1.24 557
Na24 1.39 601
Na24 1.39 621
Sr89 1.50 741
p32 1.71 810
Tel29 1.80 812
Mg27 1.80 821
Mg21 1.80 885
Homogeneous rays 2.00 966
y9o 2.18 1065
m212 (ThC) 2.25 1023
Rhl06 2.30 1080
p 8 234 2.32 1105
Sbl24 2.37 1220
As76 2.56 1384
Rhl04 2.6 1198
Mn56 2.86 1440
Cu62 2.92 1440
Homogeneous rays 3.00 1540
Prl44 3.07 1575
As76 3.12 1454
Rhl06 3.55 1770
*Katz a.nd Penfold: Revs. llfodern Phys. 24, 28 (1952).

sented either by a graph or by empirical equations. <6> The results of

accurate determinations oí sorne energies and ranges in aluminum are
listed in Table 14-2, and the corresponding range-energy curve is shown
in Fig. 14-8. The differently marked points represent ranges obtained
by different workers who used various methods, and it is clear that there
is good agreement among them. The values oí the energy were deter-
mined with magnetic spectrometers and are taken from N.B.S. circular
499: Nuclear Data. The relation between range and energy can also be
expressed by the empirical formulas
R = 4 l2Tk265-0.0954 ln To; To< 2.5Mev, (14-7)
where To is the endpoint energy, and
R = 53OTo - 106; T0 > 2.5 Mev. (14-8)
In Eqs. (14-7) and (14-8), T 0 is in Mev and R in mg/cm 2 • As noted,

Eq. (14-7) represents a good fit to the experimental results for endpoint
energies less than 2.5 Mev, and Eq. (14-8) represents a good fit above
2.5 Mev. When a range is determined by an absorption measurement,
the energy can be obtained either from the graph or from the appropri-
ate empirical formula.
14-4 Beta-particle spectra. The continuous spectrum. The problems

set by ~particle spectra are among the most important in nuclear physics,
and the main íeatures oí these spectra will therefore be discussed in sorne
detail. The discussion will be limited to the continuous spectrum because
the ,B-particles which form this spectrum are the primary disintegration
particles. The line spectrum, or secondary spectrum, is now known to
consist of extranuclear electrons ejected froro the atoro by a process called
internal conversian. A nucleus in an excited state can pass spontaneously
to a state oí the same nucleus, but of lower energy, either by emitting
a 'Y-ray with an energy h11 equal to the difference between the energies
oí the two nuclear states, or by giving the energy to an electron in the
K-, L-, . .. , shell oí the same atoro. The electron is ejected with kinetic
energy h11 - Ex, hv - EL, ... , where Ex, EL, ... are the binding
energies oí the electron in the K-, L-, . .. , shells, respectively. This
process is called internal conversion by analogy with the ordinary photo-
electric conversion oí the energy of the 'Y-ray in an atom other than the one
from which the 'Y-ray was emitted. It was thought for sorne time that
the energy of the nuclear transition was first emitted as 'Y-radiation, the
'Y-ray then ejecting a photoelectron justas it would from any atoro through
which it might pass. It has been shown that this explanation is incorrect
and that the process of internal conversion is caused by the direct ínter-
14-4) BETA-PARTICLE SPECTRA. THE CONTINUOUS SPECTRUM 353
10~-------,------,--------,------,----..,--,
l.O
>
6"'
>,
i!
"',::
;:.¡
0.1
-Plot of R=41ZT 0n n=l.265-0.094 In To

1 ' 1
--Plot of R=530T0 .-106, T 0 >Z.5
O.O!-~---~-~--------~-----------
('.! l.O 10 100 1000 10,000
Range (mg/cm 2)
Frn. 14-8. Range-energy curve for ,8-particles (Katz and Penfoid< 6 >).
action of the nucleus with its surrounding electrons. In any case, the
nuclear charge does not change in interna! conversion, so that this process
is not a form of ,8-decay. The electron line spectrum is, therefore, related
to the emission of 'Y-radiation rather than to the process of ,8-decay and
will be treated more fully in the next chapter.
In view of the above discussion, the terms ",8-spectrum" and "con-
tinuous ¡3-spectrum" are synonymous, and the monoenergetic electrons
which make up the lines found with the magnetic spectrograph will be
disregarded in the rest of this chapter. Sorne typical ,8-spectra are shown
in Fig. 14-9. In each case, the ordinate is proportional to the number
of ,8-particles detected per unit range of Hr, and the abscissa is Hr. Since
Hr is proportional to the momentum, the curves are momentum-dis-
tribution curves. They can be converted to energy dístribution curves
by converting the values of Hr to energy values with the aíd of Fig. 14-1.
The momentum-distribution curve has the same properties as the energy
distribution curve and is used often because it represents the actual experi-
mental results more directly. The relationship between the two distribu-
tion curves may be seen by comparing Fig. 14-9(e) for RaE with Fig.
14-2. In each case, the curve passes through a maximum and then de-
creases to zero at a value of the abscissa which represents the endpoint
energy.
In sorne cases of ¡3-emission, a single continuous spectrum is found, and
therefore a single endpoint energy; such a spectrum is called simple and
BETA-DECA Y [CHAP. 14
2000 4000
Hr (ga1188-Cm)
(a) Na24
01L...-----,2 000,-,l------- ,-4000~--.... ::!a"",6000~ ::-

H r (ga1188-Cm)
(b) N1ª
.004
o,L-_..___20_
00_.....__4_000..___~---,6~000-~
Hr (gauss-cm)
(C) p32
Fm. 14-9. (continued on nezt page)

14-4] BETA-PARTICLE SPECTRA. THE CONTINUOUS SPECTRUM 355
300~----------------,
]~
---~
.::::X:
~~ 200
c. ...-
o::i::
t 'o
s
-""'
~
aJ.~
e
bl)
100
·E§
j~
1000 9000
Hr (gauss-cm)
(d) Jn114
~,-¡:- 6
.:;:X:
~~ 5
c. ...-
º::i::
... o
4
j !ili 3
S e
¡¡ :! 2
_;·E
... ::, 1
.!l ...
~~º o 1000 2000 3000 4000 5000
Hr (gauss-cm)
(e) RaE
FIG. 14-9. Continuous fj-particle spectra. (a) Na 24 [Siegbahn, Phys. Rev. 70,
127 (1946)). (b) N 13 [Siegbahn and Slatis, Ark. Ast. Math. Fysik 32A, No. 9
(1945)]. (e) P 32 [Warshaw, Chen, and Appleton, Phys. Rev. 80, 288 (1950)].
(d) In 114 (Lawson and CorkCl 7>.) (e) RaE (NearyC 9>.)
is analogous to the emission of a single group of a-particles by an a-emitter.

AH of the spectra shown in Fig. 14-9 are simple. The emission of the
,8-particles may or may not be associated with the emission of "Y-rays.
Thus Na 2 4, Fig. 14-9(a), emits "Y-rays as well as fj-particles, while the
other nuclides do not emit 'Y-rays. The nuclide N 13 , Fig. 14-9(b), is
a positron emitter and, as seen from the figure, its continuous spectrum
is similar in its general features to the other spectra. Radium E, Fig.
14-9(e), is a naturally occurring radiooctive nuclide, while the other
nuclides are produced artificially. Phosphorus 32, Fig. 14-9(c), is espe-
cially interesting because of its use as a biological tracer material, and
Inu 4 , Fig. 14-9(d), is interesting for historical reasons which will be
·1~ MOl----+...,,O'r--'V----+---+---+----+--- -+----+--------<

o!,_
'o
] 60!---++-----'l-++-----+---+---+----+----+----+---j
s::,
e
·=~. 4111---+-+----,---+---+---+-""a::--;-----;---- -+---+-----t
2 4 6 8 14 16 18
Hr (gauss-cm) X 10-:i
Fm. 14-10. The P-pa.rticle spectrum of Cl 38 [Langer, Phys. Rev. 77, 50

(1950)).
mentioned later. The endpoint energies for Na 24, Na 13 , P 32 , In 11 4, and

RaE are 1.39, 1.20, 1.70, 1.98, and 1.17 Mev, respectively.
Many P-emitters have cmnplex spectra, which means that there are
two or more spectra with different endpoint energies and intensities.
These spectra can be separated by analyzing the distribution curve
obtained with the magnetic spectrometer. The spectrum of Cl 38 is shown
in Fig. 14-10, together with the three simple spectra of which it is com- ·-
posed. The curves are again momentum-distribution curves rather than
energy-distribution curves. The endpoint energies are 4.81 Mev, 2.77
Mev, and 1.11 Mev, respectively, and the relative intensities are 53%,
16%, and 31%, respectively. Two 'Y-rays have been observed along with
the P-particles, one with an energy of 2.15 Mev, the other with an energy
of 1.60 Mev. One of the isotopes of antimony (Sb 1 2t) has an even more
complex spectrum made up of five simple spectra with endpoints of 2.291
Mev (21%), 1.69 Mev (7%), 0.95 Mev (7%), 0.68 Mev {26%), and 0.50
Mev (39%); at least six different 'Y-rays are observed.
The average energy of the P-particles which form the continuous spec-
trum can be calculated from the energy distribution curve. The total
energy is found by integrating the product of the number of particles
with a given energy and that energy. The number of particles is deter-
14-4] BETA-PARTICLE SPECTRA. THE CONTINUOUS SPECTRUM 357
mined by integrating the number of particles with a given energy with

respect to the energy; this number is just the area under the distribution
curve. Expressed mathematically, the average energy is
_ J 0
Tº N(T)T dT
T=-'--"------, (14-9)
f'~
N(T) dT
where T 0 is the endpoint energy, T is the energy, and N(T) dT is the

number of particles with energies between T and T + dT. For RaE,
with an endpoint energy of 1.17 Mev, the average energy found in this
way is 0.34 Mev. The average energy associated with a disintegration
can be determined directly by placing a ¡9-emitter inside a calorimeter
designed to absorb all the known disintegration products, and by meas-
uring the total heat energy produced by a known number of disintegra-
tions. Experiments of this kind have been performed with RaE, and
three independent observers< 7 ,s. 9> have obtained values of 0.35 ± 0.04
Mev, 0.337 ± 0.020 Mev, and 0.340 Mev, respectively. These calori-
metric results are in excellent agreement with the value 0.34 Mev obtained
from the distribution curve, so that the average energy is only about one-
third of the maximum energy, 1.17 Mev.
It can be shown that when a ,8-transformation takes place, the energy
of the nucleus decreases· by an amount just equal to the maximum or
endpoint energy of the emitted spectrum. Consider the decay of ThC
to the ground state of ThD. This process can take place in either of
two ways, as will now be shown. Thorium C (Bi 212 ) can decay by electron
emission (maximum energy, 2.250 Mev), to ThC', which decays by a-
emission (a-disintegration energy, 8.946 Mev) to the ground state of
ThD(Pb 2º8). The total energy change is 2.250 + 8.946 = 11.196 Mev.
On the other hand, ThC can decay to ThC" by a-emission (a-disintegra-
tion energy, 6.203 Mev); ThC" can then decay by ,8-emission (maximum
energy, 1.792 Mev) to an excited state of ThD. The latter then decays
by two successive 'Y-ray emissions with a total energy of 3.20 Mev to the
ground state of ThD. The total energy change in the second mode of
decay is 6.203 + 1.792 + 3.20 = 11.195 Mev. The total energy change
must be the same in the two cases, and it is, provided that the maximum
,8-decay energy is used in the calculations. If the average energy were
used, there would be no such agreement.
Another example is the decay of C 14 to N 14 by electron emission. Ac-
cording to the result just obtained, the difference between the masses of C 14
and N 14 should be just equal to the mass corresponding to the endpoint
energy of the ,8-spectrum of C 14, since no 'Y-ray is emitted. The latter
energy is 0.155 Mev or 0.155/931.15 = 0.000166 amu. The difference in
mass between C 14 and N 14 can be obtained independently by considering

the threshold reaction
C 14 + p -+ N 14 + n + Q,
for which the value oí Q has been found experimentally to be -0.628
± 0.004 Mev. From Eq. (11--6), the m8.8s difference is given by
M(C 14) - M(N 14 ) = m (neutron) - m (proton) - 0.628/931.15
= 1.008983 - 1.008144 - 0.000674 = 0.000165 amu,
which is just the same as the mass difference given by the maximum
energy oí the ¡S-spectrum. Many other similar cases have been studied,
and in each case the same result is obtained. There is, therefore, no
doubt that the maximum energy of the spectrum is the one which enters
into the balance of mass and energy in nuclear reactions.
The above results lead to the following paradox: a parent nucleus in
a definite state of energy emits an electron and leaves a product nucleus
which is also in a definite energy state. The energy of the emitted elec-
tron, however, is not equal to the difference between the energy of the
nucleus before and after emission, but may have any value between zero
and the maximum energy of the continuous spectrum; the average energy
of the emitted particle is only about one-third of the maximum energy. <10>
Part of the energy, approximately two-thirds in fact, seems to have dis-
appeared, and the principie of conservation of energy seems to be violated.
In view of the fact that this principie is basic to atomic and nuclear physics,
as well as to all other branches of physics and chemistry, the possibility
that it might not hold for ¡S-decay presented a serious challenge to physical
theory.
14-5 The theory of beta-decay. Basis of the theory. In 1934, there
seemed to be only two ways of accounting for the difliculties raised by · -
the continuous ¡S-spectrum. One way was to give up the idea that energy
is conserved during a P-decay process. This solution was not regarded
with favor because the principie of conservation of energy had been uni-
formly successful in ali of its previous applications and it was diflicult
to accept the idea that it should fail in ¡S-decay. Furthermore, the con-
cept of conservation is practically an article of faith to physical scientists
and almost any other way out of a difliculty would be preferable to giving
up this concept. The second solution, which was applied successfully by
Fermi in bis theory of ¡S-decay, is based on the neutrino hypothesis first
suggested by Pauli. lt was postulated that an additional particle, the
neutrino, is produced in p-decay and carries away the missing energy.
Neutrinos had not been detected experimentally, and their properties
must be such as to make them very hard to detect. lt was, therefore,
14-5] THE THEORY OF BETA-DECAY. BABIS OF THE THEORY 359
postulated further that the neutrino is electrically neutral and has a very
small mass, i.e., very small compared with that of the electron; the neu-
trino mass may even be zero. The absence of charge is in accord with
the requirement of conservation of charge during ,8-decay, so that this
property is not an arbitrary one.
U nder the neutrino hypothesis, each ,8-decay process is accompanied
by the release of an amount of energy given by the endpoint of the spec-
trum. This disintegration energy is shared among the ,8-particle, the
neutrino, and the recoil nucleus. It has been proved in analytical mechanics
that when energy is divided among three particles, the principies of con-
servation of energy and momentum do not determine uniquely the energy
and momentum of each particle, as they do in two-body problems. When
this result is applied to ¡3-decay it means that different nuclei of the same
species distribute their equal available energies among the product par-
ticles in a continuous range of different ways. Thus, the hypothesis that
a neutrino is also emitted makes it possible to account in a general way
for the continuous spectrum of ,8-particle energies. To be acceptable,
however, a theory of ,8-decay must accomplish more than this; it must
account for the shape of the distribution curve, the existence of an end-
point energy, and the existence of a maximum in the curve. The theory
must also give the correct relationship between the average energy and
the endpoint energy. Finally, it must account for certain correlations,
which will be discussed later in this section, between the endpoint energies
and the decay constants in ,8-disintegrations. The Fermi theory of /j-
decay and its extensions, ha.sed on the neutrino hypothesis, have sué-
ceeded in accounting for all of the features of ,8-decay just listed. Although
it may, at first, seem strange and arbitrary to postulate the existence
of an undetected particle, this hypothesis has led to a highly successful
theory of ,8-decay. The success of the theory may be regarded as pre-
sumptive evidence for the existence of the neutrino, and has led to a great
deal of experimental and theoretical work on attempts to detect neutrinos
directly. This work has finally succeeded, and the successful experiments
will be discussed in Section 14-8.
The Fermi theory is based on the following ideas. When a nucleus
emits a ,8-particle its charge changes by one unit, while its mass is prac-
tically unchanged. When the ejected ,8-particle is an electron, the number
of protons in the nucleus is increased by one, and the number of neutrons
is decreased by one. In positron emission the number of protons decreases
by one and the number of neutrons increases by one. Beta-transforma-
tions may then be represented by the following processes:
,8--emission: 0 n 1 - 1H
1
+ - 1e0 + v, (14-10)
¡3+-emission: 1 H 1
- on 1
+ 1e
0
+ v, (14-11)
where " represents the neutrino. The neutron is not to be regarded as

composed of a proton, an electron, and a neutrino, but is considered to be
transformed into those three particles at the instant oí /Hlmission. Simi-
larly, the proton is transformed at the instant of jHlmission. The neutron
or proton that is transformed is not, oí course, a free particle but is bound
in the nucleus by the nuclear forces. lt has been found 0 0 that the free
neutron undergoes ¡3-decay according to Eq. {14-10), and has a half-
life of about 12 min, and this phenomenon will be discussed in more detall
later on. In a more complex radioactive nucleus, the neutrons cannot
act as free particles; their decay is a property of the nucleus as a whole
and is not to be confused with the decay of the free neutron.
Another property of the neutrino can be deduced from Eq. {14-10) or
Eq. (14-11); this property is the intrinsic angular momentum or spin.
lt was seen in Chapter 8 that the measured values of the total angular
momentum of different nuclei can be accounted for on the hypothesis
that nuclei contain only protons and neutrons. A difficulty with the
angular momentum arises, however, when one nucleus changes to another
by emitting a {3-particle. Suppose that a nucleus A emits a {3-particle fJ,
leaving a product nucleus B, and that no neutrino is emitted. The angular
momentum of the system (B + fJ) consists of three parts: the angular
momentum of B, the rotational angular momentum of B and fJ about their
common center of mass, and the spin of {3. For the total angular momen-
tum to be conserved, the resultant of these three contributions must be
equal to the angular momentum of the initial nucleus A. Since the mass
number of the nucleus is not changed by the decay process, the first con-
tribution (angular momentum of B) must be equal to the angular momen-
tum of A, or differ from it by an integral multiple of h/2r. It is shown
in wave mechanics that the second contribution, which is a rotational
angular momentum, can only be zero, or an integral multiple of h/21r.
Finally, the spin of the {3-particle {electron or positron) is known to be
½(h/2,r). According to wave mechanics, when two or more angular
momenta are added, the resultant is either equal to the sum of the com-
ponents, or is less than the sum by an integral multiple of h/2-r. When
the three contributions are added in accordance with this rule, it is found
that the total angular momentum of the system (B + fJ) must differ
from that of A by an odd multiple of ½(h/211"). The total angular momen-
tum of a system of even mass number would be an integral multiple of h/21r
for the parent nucleus and a half-integral multiple of h/2r for the produ~t
nucleus and {3-particle together. For a system of odd mass number, the
total angular momentum would be half-integral for the parent nucleus
and integral for the product nucleus and {3-particle together. If, however,
it is assumed that a neutrino with spin ½(h/2r) is emitted, the resultant
of the spins of the electron and the neutrino is integral (1 or O, in units
14-5] THE THEORY OF BETA-DECAY. BASIS OF THE THEORY 361
of h/21r). The resultant angular momentum of aU particles left after

the decay, product nucleus, electron, and neutrino, is then integral or
half-integral according to whether that of the parent nucleus is integral or
hall-integral, and the total angular momentum of the system can be
conserved.
An analogous argument can be made for the statistics of a system under-
going {j-decay. Without the neutrino, the number of particles in the
system differs by one unit before and after the ~-transformation. This
difference implies that the statistics would change, either from Fermi-
Dirac to Bose-Einstein or vice versa. This kind of nonconservation is
also unpleasant for the physicist to contemplate, and is avoided by assum-
ing that the neutrino obeys the Fermi-Dirac statistics. The number of
particles then changes by 2 units and, since neutron, proton, electron and
neutrino are Fermi-Dirac particles, the initial and final nuclei obey the
same kind of statistics.
The properties assumed for the neutrino may then by summarized as
follows: no charge; very small mass, possibly zero, at least very small
compared with the electron mass; spin equal to ½(h/2,r); Fermi-Dirac
statistics.
The neutrino hypothesis and the theory developed by Fermi apply
also to the process of orbital electron capture, which may be represented
by the equation
(14-12)
It was seen in Section 12-2 that orbital electron capture can occur pro-
vided that the mass of the initial atom is greater than that of the final
atom. The energy corresponding to this mass difference is released.
In the process, the atomic number of the nucleus is changed from Z to
Z - I; the outer electrons rearrange themselves, and x-rays are emitted
which are characteristic of the product nucleus. The total energy released,
liE, must be the sum of the binding energy, E,, of the orbital electron,
the recoil energy, Er, of the nucleus, and the kinetic energy of the neu
trino. But liE as measured may be of the order of Mev, while the bind-
ing energy of the electron and the recoil energy of the nucleus are very
much smaller. Nearly ali of the available energy must be carried off
by the neutrino, since no /j-particle or 'Y-ray is observed; only in this
way can energy be conserved. The neutrinos emitted during orbital elec-
tron capture ali have the same energy, which is equal to liE - E. - E,
The Iast result follows from the fact that the energy t:.E - Ee is divided
between only two particles, whose energy and momentum are uniquely
determined from simple kinematics. It also follows from Eq. (14-12)
that the total angular momentum and statistics are conserved only if
the spin of the neutrino is ½(h/2,r) and ü the neutrino obeys the Fermi-
Dirac statistics.
362 BETA-DECAY [CHAP. l<t
1:~c
7
6
•AcB
5 T
,<
o 4 B Th2
~
~
+ 3 RaE
CD
7.0
Log10 energy (ev)
F10. 14-11. Sargent diagram for natural P-emitters. The log of the decay
constant is plotted against the log of the maximum energy of the P,.particles
(Se.rgentC1 2l).
Before the Fermí theory is díscussed, it will help to consider certain

correle.tions which have been observed between the endpoint energy
and the half-life or disintegration constant. These correle.tions have an
important ple.ce in the systematics of p..decay. Sargentc 12 > found that
when the logarithm of the disintegration constant was plotted against
the logarithm of the endpoint energy for the naturally occurring ,B-emitters,
most of the points fell on ornear two straight lines, as shown in Fig. 14-
11. These lines, called Sargent curves, represent empirical rules which are
analogous to the Geiger-Nuttall rule in a-decay. In contrast with the
results for a-decay, in P-decay the disíntegration constant does not vary
so rapidly with the energy, and the two curves do not correspond each to
one radioactive series. For a given value of the endpoint energy, the
upper curve in Fig. 14-11 gives a value of A which is about 100 times as
great as the corresponding value on the lower curve. For a given value
of the energy, a transformation on the lower curve is, therefore, about
100 times less likely than one on the upper curve. The upper curve is
said to represent allowed transitions, and the lower curve is said to represent
f orbidden transitions. The terms allowed and f orbidden are to be considered
in a rele.tive sense, and stand for different degrees of proba.bility of spon-
taneous disintegration.
Sargent's idea of plotting log A a.gainst log T O has been extended to
artificially made P-emitters and a le.rge number of Sargent curves has
been obtained.° 3 > It is necessary to consider separately the curves
for nuclides of small, intermediate, and large atomic numbers. In each
14-6) THE THEORY OF BETA-DECAY 363
case several curves are obtained, one of which is considered to represent

allowed transitions; the others represent transitions with different degrees
of forbiddenness. These regularities in the experimental results mean that
the theory of {3-decay must account for the different relative probabilities
of ,S~mission as well as for the properties of the energy spectrum.
14-6 The theory of beta-decay. Results and comparison with experi-

ment. The problem of the theory of {3-decay is to calculate the probability
of the processes of Eqs. (14-10), (14-11), and (14-12). To make the cal-
culations it is necessary to introduce a force which can produce the changes
expressed by those equations, i.e., convert a neutron into a proton, or
vice versa, and at the same time produce an electron or positron and a
neutrino. Fermi introduced such a force and treated its operation math-
ematically by methods analogous to those used in the quantum mechanical
treatment of the emission of electromagnetic radiation. The Fermi
theori 14 · 15> yields a formula for the energy distribution of the emitted
¡3-particles, i.e., for the fraction of the nuclei which disintegrate per unit
time by emitting a ¡'.i-particle with kinetic energy between T and T + dT.
This fraction is
P(T) dT = G2 jMj 2F(Z,T) (T + m 0c 2 ) (T 2 + 2m 0c2 T} 112 (T 0 - T) 2 dT.

(14-13)
Equation (14-13) will be discussed in sorne detail, and its predictions

compared with experimental results. The quantity G is a natural con-
stant which represents the strength of the Fermi force; its actual magni-
tude will not be important in the following discussion. F(Z,T) is a com-
plicated function which describes the effect of the Coulomb field of the
nucleus on the emitted ¡3-particle. For Z = O, F is unity; it changes
very slowly as Z increases U:p to about 20, after which it varíes more
rapidly. Electron and positron emission are treated by taking Z posi-
tive in the formula for F when electrons are emitted, and negative when
positrons are emitted. The factor IMl 2 is the square of the absolute
magnitude of a wave mechanical quantity M, called a matrix element,
which is a measure of the relative probability that the nucleus will undergo
a ¡3-transformation. In most cases, M cannot be calculated exactly,
but general considerations about it make possible theoretical distinctions
between allowed and forbidden transitions. The other quantities on
the right side of Eq. (14-13) form a statistical factor which describes
how the energy T O is shared among the electron, the neutrino, and the
product nucleus. The statistical factor is responsible for the general
properties of the continuous ,8-spectrum; T O is the energy available for
the disintegration and m0 is the rest mass of the electron.
The structure of Eq. (14-13) may then be described as follows. The

factor G2 IMl 2 is a measure of the relative probability that the nucleus
may emit a (j-particle and a neutrino; F(Z,T) shows the effect of the
Coulomb field of the nucleus on the emission of the fj-particle, and the
statistical factor shows what fraction of the available disintegration
energy is carried off by the fj-particle. All of the factora together give
the required probability that a fj-particle will be emitted with the appro-
priate energy.
For transitions of the allowed type in nuclides with low values of Z,
Eq. (14-13) may be simplified; F may be taken equal to unity and M
may be taken to be independent of the energy. The distinction between
allowed and forbidden transitions will be made clearer later in this section.
Until then, an allowed transition should be regarded simply as one with
a relatively high probability, i.e., relatively short half-life and large disin-
tegration constant. With the above approximation, the theoretical decay
probability may be written
where G' is a new constant. Equation (14-14) holds for such ,g...emitters
as the neutron, H 3 , He 6 , C 11 , N 1 3, 0 16, and F 17 , and it is easier to discuss
than Eq. (14-13). It shows that for T = O and T = T 0 the probability
of ,8-decay vanishes, while for values of T between O and T 0 , P(T) is
positive. When P(T) is plotted against T, the value is zero when T = O,
increases as T increases, passes through a maximum, and then decreases
to zero at T = T 0 • For a given value of T 0 , the average value of the
energy of the emitted ,8-particle can be found by integration and is in
good agreement with measured values of the average energy.<1°> The
theoretical expression for the spectrum in an allowed transition thus gives
correctly the most important features of experimental ,8-spectra.
The theoretical values of the disintegration constant >. and the mean
life T are obtained by íntegrating Eq. (14-13) or Eq. (14-14) over the
possible values of the kinetic energy:
>. = !T = ¡
Ta
O
P(T) dT. (14-15)
In the case of the simple allowed spectra represented by Eq. (14-14),

the íntegration can be done analytically and >. can be expressed directly
in terms of the endpoint energy T 0 • It is found that, to a first approxi-
mation, the disintegration constant is proportional to the fifth power of
the endpoint energy
14-6] THE THEORY OF BETA-DECAY 365
where k is a constant. The last equation may be written

log X "" log k + 5 log T 0,
and should represent, at least roughly, the Sargent curves. It is found

that this is the case, since the slopes of the Sargent lines are approximately
5. When the effects of F(Z,T) and M must be taken into account, as
in allowed transitions of nuclei with greater values of Z, or in forbidden
transitions, the comparison is less direct and will be discussed later.
In practice, the quantitative comparison between the theoretical and
experimental spectra is made in an ingenious way involving the so-called
Kurie plot, or Fermi plot. It is convenient to rewrite Eq. (14-13) before
discussing this method. It will be assumed that M is independent of
the energy and can be regarded as a constant. The expression (T + m 0c 2 )
is just the total energy E of the emitted electron, since moe 2 ( = 0.511 Mev)
is the rest energy of the electron. It can be shown that in relativistic
mechanics, the expression ( T 2 + 2m 0 c2 T) 112 is equal to the "momentum"
pe. The difference T 0 - T is equal to E 0 - E. The fraction P(T) dT is
proportional to the number N(T) dT of ~particles with energies between
T and T + dT. Equation (14-13) may then be replaced by
N(T) ~ FEp(T 0 - T) 2 = FEp(E 0 - E) 2 , (14-16)
or
[N(T)/FEp]1 12 ~ T0 - T = Eo - E. (14-17)
The Fermi theory predicts, therefore, that if the quantity [N(T)/FEp]1 12
is plotted against T or against E, the result should be a straight line which
intersects the energy axis at T = T 0 , or at E = E 0 • If N(T) is taken
to be the number of ~-particles observed experimentally, then the plot
should be a test of the Fermi theory. This method of comparing theory
and experiment, suggested by Kurie, Richardson, and Paxton, <16> has
the advantage that it is independent of the actual measurement of T 0 •
If a straight line is obtained, it is good evidence for the validity of the
theory, and the Kurie plot then provides a straight-line extrapolation
method for finding the endpoint energy.
In practice, the momentum distribution of the emitted ~-particles is
often used. A function P(p) can be defined so that P(p) dp is the fraction
of disintegrations that give ~-particles with momenta bet,ween p and
p +dp. To each energy interval dT there corresponds a momentum
interval dp. The number of processes is the same, by definition, for cor-
responding intervals, so that P(T) dT = P(p) dp, or N(T) dT = N(p) dp.
It is then found that the energy distribution, Eq. (14-13), is replaced by
the momentum distribution
(14-18)
VN(p)/p2 4
00!,--'---:::2-!,,oo,,......--1.._..,.-1±00,,..-......_-,600=--'---,~t-'
Kinetic energy in kilovolts
(a)
''
20
Kinetic energy (kev)
(h)
FIG. 14-12. (continued on next page)

1.736
o 0.5 1.0 1.5 2.0

Kinetic energy (l\lev)
(e)
2 3 4 5
...
.. Total Pnergy E (units mc2)
o 0.4 0.8 1.2 1.6 2.2

Kinetir energy (Mev)
(d)
Frn. 14-12. Kurie, or Fermi, plots for allowed ~-transformations. (a) The
neutron (Robson< 11>). (b) Tritium: H 3 (Langer and Moffat< 27 >). (e) 0 15
[Kistner et al., Phys. Re11. 1051 1339 (1957)]. (d) In 114 (Lawson and Cork0 7>).
In the Kurie plot, the function [N(p)/Fp 2 ]1 12 can be plotted against T,

where N(p) is the number of tl-particles counted which have a momentum p.
The result should again be a straight line.
Kurie plots are shown in Fig. 14-12 for sorne ti-emitters of special
interest, the neutron, tritium (H 3 ), 0 15, and In 11 4. A straight line is
obtained in each case, as well asan accurate value of the endpoint energy.
These results, together with many others like them, provide strong evi-
dence for the validity of the Fermi theory of allowed spectra. · The Kurie
plot for the neutron, Fig. 14-12(a), yields an endpoint energy of 0.782
± 0.013 Mev. This value is just the one predicted from the difference
between the masses of the neutron and the proton, which is 0.00084 amu
or 0.782 Mev. Tritium is the simplest electron-emitting nuclide, apart
from the neutron; its Kurie plot, Fig. 14-12(b), gives an endpoint energy
of 17.95 ± 0.10 kev. This value and the plot shown are important be-
cause, as will be discussed later, they make possible an estimate of the
mass of the neutrino. The plot for 0 15, Fig. 14-12(c), is an example of
the spectrum of an allowed positron transformation; the endpoint energy
is 1.736 ± 0.010 Mev. Finally, the Kurie plot for In 11 4, Fig. 14-12(d),
was the first caseº 7 > for which unambiguous confirmation of the Fermi
theory of allowed transformations was obtained.
For certain highly forbidden transformations, for which the half-life is
relatively long and the disintegration constant small, the shape of the
spectrum departs considerably from that predicted by the simpler forros
of the theory. These deviations have been accounted for in a satisfactory
way as a result of further development of the theory and refinements in
the experimental methods. lt tums out that in these cases, the matrix
element M also depends on the energy and affects the shape of the spec-
trum. When this additional energy dependence is taken into account,
good agreement is obtained between theory and experiment.<1 8 • 19 • 2 º>
The effect of jMj 2 on the disintegration constant and the half-life, and
hence its relation to the problem of allowed and forbidden transforma-
tions, will be discussed next. Although the magnitude of IMl 2 cannot be
calculated accurately in most cases, part of its effect on >. may be estimated
without detailed calculations. The theory makes it possible to expresa
jMj 2 asan infinite series, and to estimate in a rough way the relative mag-
nitudes of the successive terms of the series. lt is found that the second
term is smaller than the first by a factor of about 100, the third term is
smaller than the second by about the same factor, and so on. The mag-
nitude of IMI 2 is, therefore, practically the same as that of the first non-
vanishing term. If the first term is not zero, the succeeding terms may
be neglected; IMl 2 then has a certain value, as do P(T) and >., and the
transformation is called allowed. If the first term of the series vanishes,
and the second term does not, the third and succeeding terms may be
neglected. The value of IMl 2 is given by the second term of the series,
and is smaller than in the allowed case by a factor of about 100. The
values of P(T) and >. are also correspondingly smaller than in the allowed
case, and the transformation is called first forbidden or once forbidden.
Similarly, if the third term of the series is the first nonvanishing one,
jMj 2 , P, and >. are still smaller, by another factor of about 100, and the
transfonnation is called second f orbidden or twice f orbidden. Higher degrees
of forbiddenness may be arrived at by continuing this argument.
The question immediately arises as to why the first term, or the first
and second, or any number of leading terma in the expansion for jMj 2
should vanish. The theory again provides an answer. In a nuclear trans-
formation, the magnitude I of the total angular momentum of the nucleus
may change by an integral multiple of h/21r, with the difference Al ap-

pearing, for example, as orbital angular momentum of the electron and
the neutrino. As in the case of atomic transitions (cf. Sections 7-3, 4),
there is a selection rule which defines the values that Al may have. If a
particular transformation obeys the selection rule, the first term in the
series for \MI 2 does not vanish, and the transition is allowed. This result
follows from the way in which the angular momenta of the initial and final
nuclei are worked into the expression for j.Mj 2 by the theory. If the selec-
tion rule is not obeyed, the first term in the series for IMl 2 vanishes and
the transition can be, at most, first forbidden. There is also a selection
rule for first forbidden transitions. If it is obeyed, the transition actually
is a first forbidden one; if not, the transition can be, at most, second for-
bidden. Depending on which selection rule is obeyed, a ,8-transformation
may be an allowed one, or may have a particular degree of forbiddenness.
The selection rules derived from the theory depend on the way in which
the special interaction introduced into the theory is defined. This inter-
action can be defined in severa} different ways, and there is no a priori
physical reason for choosing one definition over another. There is, there-
fore, sorne arbitrariness in the theory which stíll remains to be removed.
The first, and simplest, definition of the interaction was suggested by
Fenni and led to the selection rule 1:11 = O for allowed transformations.
Inconsistencies were found between theory and experiment, and Gamow
and Teller used a different definition of the interaction, which led to the
selection rule Al = O, ± l. The most recent work in this branch of the
field of ,8-decay indicates that the actual interaction must be a combin-
ation of the Fermi and Gamow-Teller types. The problem of the exact
definition of the interaction that leads to ,8-decay has not yet been com-
pletely solved, but nearly all of the experimental facts of ,8-decay can be
explained by the theory, and the problem mentioned seems to be well on
the way to solution. <20 >
In spite of the fact that the matrix element /.M/ 2 cannot be calculated
accurately in most cases, it is possible to treat differences between allowed
and forbidden transitions quantitatively, or at least semiquantitatively.
The mean life of an allowed transition, for which /.Ml 2 is independent of
the energy, is, from Eqs. (14-13) and (14-15),
The integral in the last equation is a function only of Z and T 0 • Its value
may be computed by analytical or numerical integration depending on
whether or not approxirnate expressions for F may be used. In either
case, the integral may be represented by a function f(Z,T 0 ), and Eq.

(14-19) may be written
1/T = G2 /M/ 2f(Z,To), (14-20)
or
fT = 1;c 2 /M/ 2 • (14-21)
The product fT is a quantity whose numerical value depends only on the

atomic number and the measured values of the endpoint energy and the
mean life. If the dependence of /M/ 2 on energy is neglected in the case
of forbidden transitions, the last equation may be applied also to those
transitions. Although sorne error is involved because /M/ 2 for forbidden
transitions <loes depend to sorne extent on the energy, the product jT may
be used as an indication of the degree of forbiddenness. Since the value
of /M/ 2 for a first forbidden transition is smaller by a factor of about 100
than the value for an allowed transition, it would be expected that the
value of fT would be about 100 times as large for a first forbidden trans-
formation as for an allowed one.
In practice, the product of the function f and the half-life is usually
calculated. The half-life has been denoted by T 112 , so that the values of
the product JT 112 should be considered, with fT 112 = 0.693fT. The half-
life, however, is often denoted by t, and the product ft is called the com-
parative half-life, or the ft-value. The ft-values of severa} hundred {J-
emitters have now been determined< 20 and analyzed. <20 •22 > They range
from about 1000 sec to about 10 18 sec and, for convenience, the log-
arithm of the ft-value is usually considered rather than the ft-value itself.
The smallest values of the comparative half-life are found for a group of
light nuclei for which log ft = 2.1 to 3.7. This group includes the lightest
,3-emitters, s11ch as the neutron (log ft = 3.21), H 3 (log ft = 3.06), He 11
(logft = 2.74). It also includes the positron emitters: 5C 11 , 7N 13 , 80 15 ,
gF 17 , 10 Ne 19 , 11Na 21 , 12Mg 23 , 1aAI 25, uSi 27 , 15P 29 , 1&8 31 , 11Cl 33 , 1sA35,
37 , Ca 39 , Se 41 , and Ti 43 . In the latter nuclides, the mass number
19K 20 21 22
A is equal to 2Z - l, and the number of protons exceeds the number of
neutrons by one. The emission of a positron makes the product nucleus
have one more neutron than proton. Thus, in the reaction
11 11
6C + 1eº + "·
-+ sB
11
the parent nucleus 6C has six protons and five neutrons, while the product
nucleus 5B 11 has five protons and six neutrons. In these cases, the initial
and product nuclides are called mirror nuclides, because of the symmetrical
relationships between the numbers of protons and neutrons. The emission
of a positron has a particularly high probability in the mirror nuclides and
the ft-values are especially low. The transformations of this group with
logft between 2.7 and 3.7 are described asfavored allowed transitions.
14-7) ENERGY LEVELS AND DECA Y SCHEMES 371
Another group of nuclides is found for which the values of log Jt lie
between 4 and 5.8. The transformations are described as normal allowed
12 35 64 114
transitions, and include such cases as B 12 , N , S , Cu , and In .
The nuclide Cu is 64 particularly interesting beca use it undergoes electron
emission, positron emission, and K-electron capture; the value of log jt
is 5.29 for electron emission and 4.94 for positron emission, and both
transitions are allowed.
Values of log jt in the range 6-9 are characteristic of first forbidden or
97 115 140
once-forbidden transitions. Nuclides such as Zr , Cd , Ba , and
W 187
are included in this group; RaE is a borderline case which is probably
first forbidden. There is a relatively small number of second forbidden
9
transitions, for which the value of logft is in the range 12.2-13.5; Cl 36 , Te 9 ,
137
135
Cs , and Cs are examples of this kind of transformation. Finally,
Be 1°, with a value of logft equal to 13.7, and K 40 with log ft = 17.6, are
cases of nuclides whose ¡3-transformations are even more highly forbidden.
The degree of forbiddenness in nearly ali of the above cases can be corre-
lated in a satisfactory way with the shape of the spectrum and with the
known or assumed angular momenta of parent and daughter nuclei.
14-7 Energy levels and decay schemes. Beta-transformations often

yield information about the energy levels of the product nuclei and about
decay schemes. These transformations are sometimes accompanied by
'Y-rays, and the presence of the 'Y-radiation means that the product nucleus
is formed in an excited state and passes to its ground state by emitting
one or more 'Y-rays. If no 'Y-ray is emitted, the ¡3-transition is directly
15
to the ground state of the product nucleus. In the case of 0 a positron
is emitted, and no 'Y-ray is observed. The endpoint energy is 1.73 Mev,
and the ¡3-disintegration energy is 1.73 Mev +
1.02 Mev = 2.75 Mev.
The latter result follows from the fact, discussed in Section 12-2, that
two extra electron masses are needed for positron emission to be ener-
getically possible. It follows that the difference between the ground
state energies of 0 15 and N 15 is 2.75 Mev, which is also the energy equiva-
lent to the difference between the masses of these two nuclides. The
nuclide F 20 (Fig. 14-13a) emits electrons with an endpoint energy of 5.41
Mev, and also 1.63-Mev 'Y-rays. The total disintegration energy is 7.04
Mev, and the product nucleus Ne 20 has an excited state 1.63 Mev above
the ground state.
The beta-spectra of 0 15 and F 2 º are simple and the decay schemes are
also simple. Sometimes, however, two or more groups of ¡3-particles are
emitted. The spectrum is then complex; it can be broken down into two
14
or more simple spectra, and 'Y-rays are observed. In the case of 0 , in
more than 99% of the disintegrations, positrons are emitted with an end-
point energy of 1.84 Mev; 2.30-Mev 'Y-rays are also observed. The total
872 BETA-D ECAY
[CHAP. 14
~. 2.77 Mev
(16%)
_,..... _l.63
'Y (1.63 Mev)
-'!'!"'1 1-0
(a)
(d)
-, 1.15 Mev and lower

tJ-, 1.67 Mev (48%)
(14%)
-, 1.34 Mev (30%)
------a.oo
'\..--- --...+ -2.52
. _ 1-_.....,.........+-+-2.4 1
,•-""""1~..........
(b)
------1.60
'Y 'Y
-2.08
_ _Ce_U_II.....,_ 0
(e)
{e) {f)
Fm. 14-13. Decay sehemes for /3-emitters. (a) F 20• (b) 14
(d) CIª 8 • (e) La 1411• (0 Cu64 • {Strominger, Hollander, and0 • (e) Mg •
27
Mod. Phy,. 30, 585 (1958).] Seaborg, Rev,.

U-7] ENERGY LEVELS AND DECAY SCHEMES 373
disintegration energy is 1.84 Mev + 1.02 Mev + 2.30 Mev = 5.16 Mev,
of which 2.86 Mev is the difference in energy between the ground state of
0 14 and the excited state of the product nucleus N 14 . The N 14 nucleus
passes to its ground state by emitting a 2.30-Mev 'Y-ray. In about 0.6%
of the disintegrations 0 14 undergoes a transition directly to the ground
state of N 14 by emitting 4.1-Mev positrons. The decay scheme is shown
in Fig. 14-13(b). The electron-decay of Mg 27 is more complicated: about
70% of the disintegrations correspond to an endpoint energy of l. 78 Mev
and about 30% toan endpoint energy of 1.59 Mev; 'Y-rays are observed
with energies of 0.834 Mev and 1.015 Mev, respectively, and in less than
one percent of the disintegrations, a 'Y-ray with an energy of 0.18 Mev is
observ."rl. Coincidence experiments show that the 1.78-Mev /j-ray and
the 0.834-Mev 'Y-ray belong to the same transition, (i.e., they are co-
incident), and that the 1.59-Mev /j-ray and the 1.015-Mev 'Y-ray belong to
the same transition. A decay scheme consistent with áll of these data is
shown in Fig. 14-13(c). The direct transition from the ground state of
Mg 27 to the ground state of Al 27 by electron--emission is evidently highly
forbidden. The electron decay of Cl 38 has three groups of electrons, as
discussed in Section 14-4; two 'Y-rays have been observed, and the decay
scheme is shown in Fig. 14-13(d). A still more complicated scheme is that
of 57 La 140 , shown in Fig. 14-13(e); at least four groups of electrons a.re
involved and at least nine 'Y-rays have been observed.
The nuclide Cu 64 is a particularly interesting case of /j-decay beca use
it emits both electrons and positrons and also undergoes orbital electron
capture. In 39% of the disintegrations, an electron is emitted; the /j--
spectrum is simple and the endpoint energy is 0.57 Mev. The product
nucleus Zn 64 is formed in its ground state. In 19% of the disintegrations,
a positron is emitted with an endpoint energy of 0.66 Mev; the product
nucleus Ni 64 is formed in its ground state. In 42% of the disintegrations,
a _K--electron is captured. In nearly all of the captures, the product nucleus
Ni 6 " ís formed in its ground state, but in a small fraction of the X-captures,
a 'Y-ray is observed with an energy of 1.34 Mev. There is, therefore, an
excited level of Ni 64 1.34 Mev above the ground state. It has been shown
that the 'Y-ray is observed only in coincidence with the orbital electron
capture, and is not associated with the emission of either the electron or
the positron. The decay sche~e of Cu 64 is shown in Fig. 14--13(f).
The examples of decay schemes that have just been discussed represent
a small fraction of the schemes that have been analyzed. Several hundred
decay schemes have now been determined, varying in complexity and in
the detail with which they have been worked out. Compilations of these
schemes are useful in many phases of nuclear physics, and the reader is
referred to that by Strominger et al., listed among the general references
at the end of the chapter.
374 BETA-DECAY
14-8 The neutrino. The success of the Fermi theory o{ (Hlecay in

accounting for the properties of ,8-transformations provides indirect evi-
dence of the existence of neutrinos. This success encouraged attempts to
show experimentally that neutrinos exist. <23 > But the properties which
the neutrino must have to make it fit the requirements of the theory, in
particular the zero charge and very small mass, imply that the interaction
between neutrinos and matter is extremely small. The smallness of this
interaction is emphasized by the fact that the discovery of the neutron
with a mass very much greater than that assumed for the neutrino pre-
sented sorne difficulty, and it is not surprising that attempts to detect
ionization caused by neutrinos failed. Additional indirect evidence for
the existence of neutrinos has, therefore, been sought from studi.,., oí the
momentum and energy balances in ,8-decay. It has also been possible to
deduce, from nuclear data, an upper limit-for the rest mass of the neutrino.
Finally, in 1956, a nuclear reaction induced by neutrinos was observed,
an inverse ,8-decay, and the experimental search for the neutrino has ended
successfully.
We consider first the problem of the mass of the neutrino. Estimaf.el.·
of this mass can be obtained from two types of experiments, comparison
of observed maximum energies of ,8-ray spectra with known availabll
decay energies and the shape of ,8-ray spectra near the endpoint.
The greatest kinetic energy with which a ,8-particle can be emitted is,
for electrons,
(14-22a)
where l:!.M is the difference in mass between parent and daughter atoms
and µ. is the rest mass of the neutrino. For positron emission the relation '
To = (l:!.M - 2m 0 - µ.)c 2 , (14-22bJ
where m 0 is the rest mass of the electron or positron. Compari"')!l of the

observed endpoints, T 0 , with atomic mass differences l:!.M, determinecl
from Q-values for nuclear reactions or from atomic masses, can be used
to estímate the neutrino mass. The most direct comparison is that with
a reaction which is just the reverse of a ,8-decay: a (p,n) reaction
is combined with a ,a--emission, oran (n,p) reaction is combined with a
,a+-emission. For example, the reaction H 3 (p,n)He 3 with Q = - 0.764 ±
0.001 Mev can be combined with the ,a--decay of H 3 with T 0 =
0.0181 ± 0.0002 Mev. We have
Q = [M(H 3 ) - M(He 3 ) - M,. + M 11 )c 2,
µ.c 2 = T0 - [M(H 3 ) - M(He 3 )]c 2 ¡
elimination of the difference M(H 3 ) - M(He 3 ) between the two equa-
THE NEUTRINO 375
tions and insertion of the values of Q, T 0 and the neutron-proton mass

difference gives µc 2 = O ± 0.0015 Mev, or µ = O ± 0.003m 0 • In this
case the upper limit for the rest mass of the neutrino is 0.003m 0 or about
1.5 kev, and comes from the uncertainty in the Q-value. About a dozen
similar examples involving light nuclei can be used, ali of which yield
1imits below µ = O.Olmo, with the Iimits depending on the uncertainty
in the Q-values and the endpoint energies.
In Fermi's formulation of the theory of /3-decay, the only condition on
the rest mass of the neutrino was that it be small compared to the rest
mass of the electron. The precise theoretical shape of an allowed fJ-
spectrum in the neighborhood of the endpoint depends on the actual
valoo of the rest mass of the neutrino, and comparison of theoretical and
experimental spectra near the endpoint can be used to obtain estimates of
the rest mass of the neutrino. Analyses of this kind <24 > yield µ :::; 5 X
10- 4 m 0 , or a rest energy of 250 ev or less.
There are a number of reactions in which indirect evidence of the
nresence of the neutrino has been obtained from the momentum given to
t recoiling nucleus. One of the objects of recoil experiments has been to
test the hypothesis that a single neutrino is emitted in a /3-decay process.
iThe simplest reaction in which to study the nuclear recoil is orbital electron
capture. With the assumption that a single neutrino is emitted the
reaction is
(14--23)
If the initial nucleus is at rest, conservation of momentum requires that
p(Y) = MV = V2ME. = p., (14--24)
where M, v, and E, are the mass, velocity, and kinetic energy of the recoil-
ing nucleus, respectively; p represents momentum, p, being that of the
neutrino. The velocity with which the nucleus recoils is small because
of its enormous mass relative to that of the neutrino, and its kinetic
energy isjust ½Mv 2 • The neutrino, however, must be treated relativistically
because its very small mass must be associated with high speed to give
kinetic energies of the order of Mev; its total energy E, can be· then
written [see Problem 9(d) of Chapter 6]
(14-25)
By Eq. (14-24), P• = v'2ME, so that E, = [(µc 2) 2 + 2c 2ME,] 112, and

the kinetic energy of the recoil nucleus is given by
(14-26)
376 BETA-DECAY
Conservation of energy in orbital electron capture requires that
E. = [M(X) - M{Y)]c 2 - E, - BK,L, {14-27)
where BK,L is the binding energy of the electron in the K- or L-shell.

In most neutrino experiments, E.» E,+ BK,L so that, to a good
approximation,
E,, = M(X) - M(Y). (14-28)
The mass difierence can be obtained from Q-values for nuclear reactions
or mass spectroscopic data, and E,, is then known; µ,c 2 is generally mu"h
smaller than E,,, and Eq. (14-26) becomes
(14-29)
It is convenient to expresa E,, in units of moe 2 = 0.511 Mev, and the last
equation may be written
E - moc 2 (-E• -)2. (14-30)

' - 2(Mc 2 /moe 2 ) moe 2
If we now express M in atomic mass units, and note that m0 =
5.488 X 10-4 amu, Eq. (14-28) becomes
_ 140.2(E,,/m0c 2 ) 2
Er - M ev, (14-31)
where Mis the atomic mass of the recoiling nucleus in amu.

If the assumption that a single neutrino is emitted is correct, the re-
coiling atom should have a single energy, since the entire energy oí the
transformation is carried off by the nucleus and the neutrino. If, on the
other hand, two or more neutrinos are emitted, there should be a continuous
distribution of recoil energies. A number of orbital electron capture
experiments have been made in which the energy spectrum of the recoil-
ing nucleus has been m~ured;< 25 • 26 J the nuclides studied were Be 7 , A 37
and Cd 107 . The simplest of these cases is that of A 37, which decays ex-
clusively to the ground state of Cl 37 . The energy E,, is given by the
A 37 - Cl 37 mass difierence; a value of 0.816 ± 0.004 Mev for this mass
difierence has been obtained from the Q-value of the Cl 37 {p,n)A37 reac-
tion and the neutron-proton mass difierence. Equation (14-31) then gives
E,= 9.67 ± 0.08 ev. The recoil energy has been 1l'easured< 24 > in several
independent experiments with results of 9.7 ± 0.8 ev, <27 >9.6 ± 0.2 ev< 2 s>
and 9.65 ± 0.05 ev, <29> respectively. These results show clearly that
:-;.
14-8) THE NEUTRINO 377
monoenergetic single neutrinos are emitted. The recoil experiments with

Be 7 and Cd 1 º7 are complicated by the emission of 'Y-rays; e.g., Li 7 is
left in an excited state in about 10% of the disintegrations and a 0.478-
Mev 'Y-ray is emitted. The Be 7 - Li 7 mass difference is 0.864 ± 0.003
Mev; in 90% of the decays the recoils should be monoenergetic with
57.3 ± 0.5 ev energy if a single neutrino is emitted. The 10% branch
going to the excited state of Li 7 should show a continuous recoil spectrum
extending from nearly zero to 57.3 ev and resulting from the nearly
simultaneous emission of a neutrino and a 'Y-ray. Experimentally, recoil
energy spectra were found with maxima of 56.6 ± 1.0 ev< 3 o. 3 u and
55.9 ± 1.0 ev, <32> in good agreement with 57.3 ± 0.5 ev. The interpre-
tation of the continuous spectrum, however, was complicated by experi-
mental difficulties; although it did not prove conclusively that only one
neutrino is emitted in orbital capture in Be 7 , the results were consistent
with that hypothesis.
In ¡'1-particle emission the momentum conditions are more complicated
than in electron capture, because the recoil momentum is determined by
the proportion in which the decay energy is shared between the electron
and the neutrino and by the angle between the two. It is possible to meas-
ure the angle between the ¡9-particle and the recoil nucleus and, from
conservation relations, to deduce the angle between electron and neu-
trino. These electron-neutrino angular correlation experiments yield
information concerning the forro of the interaction involved in ¡9-decay,
i.e., the force which converts the neutron into a proton and at the same
time produces an electron or positron and neutrino. ' 33 > This problem is
beyond the scope of this book and will not be discussed further.
The successful detection of neutrinos was made possible by the avail-
ability of a source of enormous numbers of neutrinos (a nuclear reactor)
and of extremely sensitive radiation detectors. The details of the experi-
men t to be described involve sorne properties of neutrons, positrons, and
electrons which have not yet been discussed and it will be necessary to use
sorne information from later chapters.
The nuclear reactions which appeared most likely to permit the detec-
tion of neutrinos are the inverse ¡9-decay processes. The inverse of the
¡9--emission process, Eq. (14-10), is obtained from Eq. (14-10) by trans-
posing the electron (_ 1e 0 ) to the left side of equation, noting that
- (_ 1e ) = 1e , and reversing the direction of the reaction; the result is
0 0
(14-32a)
The inverse of the ¡9+-emission process, Eq. (14-11), can be written
(14-32b)
and the inverse of orbital electron capture can be written
or (14-32c)
Before proceeding with the discussion of the neutrino detection experi-

ments we must consider an additional complication. It was suggested,
in accord with relativistic quantum mechanics, that the neutrinos in-
volved in tr-decay are different from those involved in p+-decay and
orbital electron capture. Recent theoretical and experimenta1< 34 , 35 •35 •3 n
work which will be discussed in the next section has shown that a distinc-
tion must indeed be made depending on the relatioo between the spin and
momeotum of the neutrino. The neutrino (11) is now defined as a particle
with its spin vector parallel to its momentum vector; and the a11tineutrino
(11*) as a particle with spin opposite in direction, or antiparallel, to its
momentum. The spin and momentum together define the sense of a screw,
the neutrino acting like a right-handed screw and the antineutrino like a
left-handed screw. The choice of which particle is called neutrino or anti-
neutrino is an arbitrary one. The choice shown in the following equations
for the direct P-processes is the one which has been generally adopted;
the equations are written in simpler notation than that of Eqs. (14-10),
(14-11), and (14-32)
p--emission: n - + p-+ ,,•,
p (14-33a)
p+-emission : p-n + p+ + ", (14-33b)
orbital electron capture: p + rr-n + "· (14-33c)
The antineutrino is emitted in p--decay; the neutrino is emitted in t3+-
decay and orbital electron capture. The inverse reactions are
inverse /3-: 11* +p - n + t3+, {14-34a)
inverse t3+: (14-34b)
inverse orbital electron capture: (14-34c)
In a nuclear reactor, nuclei of a fissionable material, for example, U 235 ,
are split into two fragments of intermediate atomic mass (Chapter 19).
The fragments are highly unstable, containing too many neutrons for the
number of protons. They therefore decay, by several successive /3--
emissions, to stable or long-lived fission products, liberating antineutrinos
in the process. If these antineutrinos are permitted to bombard a hydrog-
enous material, reactions of the form of Eq. (14-34a) should occur, and
J
14-8] THE NEUTRINO 379
the problem is to detect the simultaneous production of a neutron and a

positron. This detection is possible because of certain properties of neu-
trons and positrons. When a positron approaches an electron, the two
interact in a very special way, they disappear and two 'Y-rays (photons)
appear in their place (Chapter 15). This mutual annihilation takes place
in a very short time in matter and prompt 'Y-rays are formed, which can be
observed with a scintillation detector. A neutron produced in a hydrog-
enous material undergoes elastic collisions with protons and loses energy
(Chapter 18). When it is slowed down it can be captured in an (n,'Y)
reaction, especially if a relatively heavy material is present. Cadmium is
such a heavy material with a very high probability for capturing slow
neutrons in (n,'Y) reactions. The experiments of Reines, Cowan, and co-
workers138·39> for the detection of antineutrinos were based on the use of
a large, liquid, hydrogenous scintillation counter (terphenyl dissolved
in triethylbenzine) containing cadmium. When an antineutrino interacts
with a proton, the annihilation of the positron produced is signalled by a
prompt 'Y-ray pulse. A delayed pulse should appear several microseconds
after the prompt pulse because of the capture of the neutron in the cadmium
after being slowed down. The time required for the slowing down of the
neutron can be calculated and the pulses due to the prompt disappearance
of the positron and the delayed capture of the neutron are counted by a
delayed coincidence method. In the later experiment< 39> the detector con-
sisted of a multiple-layered arrangement of three scintillation counters
and two target tanks with one of the latter consisting of a polyethylene
box containing a water solution of cadmium chloride. A schematic diagram
of the detector is shown in Fig. 14-14. The liquid scintillator was sur-
rounded with 110 photomultiplier tubes connected in parallel in each
tank. A typical event is shown in the figure. An antineutrino induces an
inverse ~--decay at point l. A prompt pulse appears in both the top and
central scintillation counters because of the two 'Y-rays from the annihila-
tion of the positron at point 2, which is close to point l. The neutron,
after undergoing severa} collisions, is captured in cadmium at point 3
with the appearance of several capture 'Y-rays and a delayed pulse. To
show that the prompt and delayed pulses were due to a reaction induced
by antineutrinos it was necessary to relate the pulse rate to the power
level of the reactor and to the number of protons in a target box; to show
that the prompt pulses were due only to the annihilation of positrons,
and that the delayed pulses were due only to the capture of neutrons in
cadmium. The effect of fast neutrons from the reactor had to be ruled
out along with other possible spurious effects. The final result was a
maximum neutrino signa} rate of 2.88 ± 0.22 counts/hr with a signal-
to-background ratio of three to one; this result has been accepted as
convincing evidence of a nuclear reaction induced by antineutrinos.
380 BETA-DECAY [CRAP.14
..
Scintillation
oounter
Target taPk
with cadmium
Scintillation
oounter
Target taPk
with cadmium
Scintillation
oounter
F10. 14-14. Schematic diagram of the antineutrino detector used in the

experiment of Ref. 38.
An experiment complementary to that on the detection of antineu-

trinos has been in progresa for sorne time 140 ·•u
in which an attempt has
been made to detect the reaction
v• + Cl 37 - A37 + fr. (14-35)
Comparison with Eq. (14-34c) shows that this reaction would be the in-
verse of orbital electron capture in A 37 if the neutrino and antineutrino
were identical. According to theory, however, these particles are different
and the antineutrinos from a nuclear reactor should not induce the reac-
tion (14-35). In the most recent version of the experiment, 3000 gallons
of carbon tetrachloride, free of argon, were exposed to antineutrinos from
the same reactor used in the experiment of Reines and Cowan. At the
end of the irradiation the CC1 4 was swept with 12,000 liters of helium
gas to remove argon. A small measured volume of argon gas was intro-
duced into the tanks before sweeping to serve as a carrier and to measure
the recovery of argon. The argon was then separated from the helium in a
charcoal trap cooled with liquid nitrogen, and its /j-activity was counted
in a small Geiger counter. The results showed that the induction of in-
verse ,8-decay in Cl 37 by antineutrinos, the reaction (14-35), is much less
probable than the reaction v*(p,n),8+. The quantitative measure of nuclear
reactions is the cross section (see Section 16-3) which has values of the order
of 10-24 cm 2 for most nuclear reactions. The value foundcrn for the
14-9) SYMMETRY LAWS AND NONCONSERVATION OF PARITY 381
11*(p,n){3+ reaction is (11 ± 2.5) X 10- 44 cm 2 ; for the reaction (14-35), the
value of the cross section was found to ha.ve an upper limit of 0.2 X 10- 45
cm 2 • These resulta show that the probability of antineutrino-induced
inverse orbital electron capture is much smaller than that of the absorp-
tion of antineutrinos by protons. They show also that the probability
for antineutrino-induced reactions is about twenty orders of magnitude
smaller than that for ordinary nuclear reactions.
14-9 Symmetry laws and the nonconservation of parity in /1-decay.

The problem of the conservation of momentum and energy in ti-decay
led, as we have seen, to the neutrino hypothesis and eventually to the
detection of the neutrino. In recent years it has been discovered that
an important symmetry law, that of right-left symmetry, breaks down in
certain reactions, the so-called weak interactions, of which ¡i-decay is an
example. To see how this comes about it is necessary that we consider the
relationship between symmetry or invariance principles and conservation
laws. We shall follow the discussion by Yang.< 43 >
During the first half of the 20th century it was recognized< 44 > that,
in general, a symmetry or invariance principie (see Sections 6-4 and 6-5)
genera.tes a conservation law. For example, the invariance of physical
laws under a displacement in space has as a consequence the law of con-
servation of momentum; the invariance under a rotation in space has as a
consequence the law of conservation of angular momentum. In the theory
of relativity the concept of invariance was extended to invariance under
the Lorentz transformation, and the requirement that the equations of a
physical theory satisfy this type of symmetry principle became a funda-
mental condition on a physical theory. This new type of invariance is
related to the impossibility of detecting absolute motion, to the variation
of mass with velocity, and to the relationship between mass and energy.
The development of quantum mechanics extended the use of symmetry
principies still further, into the very details and language of atomic
physics, and the consequences of these principles appear explicitly in the
methods and results of quantum mechanics. <45 > Thus, the quantum
numbers which designa.te the states of an atomic system are often closely
related to the numbers which represent the symmetries of the system.
One example is that of the total orbital angular momentum quantum
number L of the electrons in an atom. The atora is symmetrical with
respect to a three-dimensional rotation about the nucleus, and orbital
angular momentum is conserved; the Schroedinger equation for the atom
must be invariant under such rotations and the detailed solutions of that
equation, which embody the conservation of angular momentum, depend
explicitly on L. A second example involves a kind of symmetry typical
of quantum mechanics. Since one electron cannot be distinguished from
another, an atom is symmetrical with respect to the interchange of elec-

trons. The Schroedinger equation which describes an n-electron atom
must therefore be invariant under a transformation which involves any
permutation of its n electrons. This invariance turns out to be related to
the multiplicity of the spectral terms of the atom. In more formal terms,
"the Schroedinger equation for an n-electron atom is invariant under the
symmetric group of order n," and the degree or order n of this invariance
is closely related to the multiplicity of the spectral lines. In fact, the general
structure of the periodic system (because of the way it is built up in terms
of quantum numbers) may be said to be a consequence of the spatial
symmetry of Coulomb's law, the law of force between the nucleus and the
electrons. The combination of relativity and quantum mechanics leads
to an especially striking example of the fundamental role of symmetry
laws. The existence of the positron was anticipated theoretically by the
requirement that the quantum mechanical equation for the electroo, the
Dirac equation, must be invariant under the Lorentz transformation. <46 >
The existence of other antiparticles,<H. 4 S> e.g., antiproton, antineutron,
antineutrino, has a similar theoretical interpretation, and their discovery
testifies to the power and beauty of the invariance principies.
Another basic type of symmetry is that between right and left, or sym-
metry under a reflection. The principie of invariance involved may be
stated in the following way: Any process which occurs in nature can also
occur as it is seen reflected in a mirror; the mirror image of any object is
also a possible object in nature; the motion of any object as seen in a mirror
is also a motion which would be permitted by the laws of nature; an
experiment made in a laboratory can also be made in the way it appears
as seen in a mirror, and any resulting effect will be the mirror image of the
actual effect. More precisely, we expect that the laws of nature are in-
variant under reflection, and experíence seems to support this idea.
In quantum mechanics it has been shown( 49 > that the consequence of
the ríght-left symmetry (invariance under reflection) of the electro-
magnetic forces in the atom appears as the conservation of parity. It
will be recalled from Section 7-9 that parity is a quantum number which
takes on the value +1 or -1 depending on whether the Schroedinger
wave function is unchanged or changes sign under a reflection through
the origin of the spatial coordinates. Experimentally (in the study of
atomic spectra) it has been found<so> that in an atomic transition with
the emission of a photon the parity of the initial state is equal to the total
parity of the final state, i.e., to the product of the parities of the final
atomic state and the photon emitted. Since right-left symmetry was un-
questioned in interactions other than electromagnetic, the concept of
parity and its conservation was taken over into nuclear physics, in par-
ticular into nuclear reactions and ,8-decay, and into the analysis of reactions
14-9] 8YMM.ETRY LA.WS AND NONCONSERVATION OF PARITY 383
involving new particles such as mesons and hyperons (whicb we do not

treat in this hook). In these developments, the concept of parity and
the law of conservation of parity proved to be very fruitful, and the
success accompanying their use was taken as support for the validity of
right-left symmetry.
It has been mentioned that fj-decay is an example of a weak interaction
and we must define what is meant by this description. To do so, we must
compare the forces involved in p.decay with those ínvolved in other reac-
tions. The interactions with wbieh we are most familiar so far are those in
which electromagnetic radiation is emitted (see Sections 7-2 and 7-8).
The force involved is that between charged particles and is proportional
to e2• lt is convenient to use a dimensionless quantity to compare the
magnitudes of forces, and the one which enters naturally into atomic
theory is e2 /lic = 1/137.04, (where lí, = h/21r), or approximately 10- 2 •
This quantity is spoken of as the "coupling constant for electromagnetic
interactions." The forces binding neutrons and protons together in the
nucleus are much stronger than the electromagnetic forces and also much
more eomplicated (they will be treated in Chapter 17). The nuclear
forces may be characterized by a coupling constant /2/líc "'-' 1, where f is
a constant analogous to e; these forces are referred to as strong int,eractions.
The forces in p-decay, those linking the neutron, proton, electron, and
neutrino (or antineutrino), may be characterized by the coupling constant
g 2 /lic ""' 10- 13, where gis a constant related to the constant G of Section
14-6. They are very small comparad to the nuclear and electromagnetic
forces, and it is for this rea.son that they are called weak interactions.
For comparison, the analogous coupling constant of Newtonian gravita-
tional forces is of the order 10-39.
Weak intera.ctions occur not only in P-decay but also in the deeay of
certain particles found in cosmic rays a.nd nuclear reactions at high energies
(several hundred Mev and higher). Among these particles are the T and
6 mesons which have been discovered in recent years, and whose proper-
ties have been the subject of intensive study. Results were obtained which
seemed paradoxical in that ali the physical properties of the parent particle
in T-decay and 8-decay seemed to be identical, but detailed analyses
indicated that the final states had opposite parities. Lee and Yanl 34)
showed that the results would fall into a consistent scheme only if parity
were not conserved in these decay processes, i.e., only if these processes
were not invariant under reflection. This surprising possibility led tbem
to ínvestigate the experimental information conceming parity conserva-
tion. They found impressive evidence for the validity of this law in atomic
physics, the domain of electromagnetic interactions, and in nuclear physics
when strong interactions were concemed. There were, however, no experi-
mental data relevant to the question of the conservation of parity in
~ 1
E:) 1
1
1 1
1 1
1 1
♦ Mirror ♦
~ Counter
E1J
,s.!auoO
Fm. 14-15. Schematic diagram of an experiment to test the conservation

of parity in ,8-decay (Yang< 43l).
weak interactions; no experiment had ever been designed specificially to

test invariance under reflection in the ,8-decay of nuclei or elementary
particles.
Lee and Yang proposed sorne experiments, one of which involved the
,8-decay of Co 60, to test invariance under reflection. The basic idea of
this experiment is to build two sets of experimental arrangements which
are mirror images of each other, and which contain weak interactions, as
shown in Fig. 14-15. The experiment then consists in determining whether
the two arrangements always give the same results in terms of the read-
ings of their meters (or counters). The practica! feasibility of such an
experiment depends on the following facts. The cobalt nucleus has a
spin angular momentum; it rotates with a well-defined angular momentum
when it is in its normal state. In a piece of Co 60 under normal conditions,
the nuclear spins are oriented in ali directions because of the thermal
motion and the electrone are emitted in ali directions. Under certain con-
ditions, however, the nuclei can be oriented in a particular direction; all
the nuclei can be forced to align their axes of rotation parallel to a given
direction so that they all rotate in the same sense. This can be done because
the nuclear spin gives rise to a magnetic moment which can be acted
upon by a magnetic field externa! to the nucleus. At very low tempera-
tures, less than 0.IºK, the externa} field can force the spins into a given
direction,< 50> and the nuclei are said to be polarized. The low temperature
14-9] SYMMETRY LAWS AND NONCONSERVATION OF PARITY 385
(a) (a') (b)

Mirror J\firror
FIG. 14-16. Cobalt,S-decay: possible electron emission patterns and their

mirror images.
is needed to reduce the thermal motion of the nuclei to a minimum and

make the orderly alignment possible. The experiment proposed by Lee
and Yang was simply to line up the spins of the electron-emitting nuclei
of Co 60 along the same axis, and then see whether the ,S-particles were
emitted preferentially in one direction or the other along the axis. If
there is preferential emission in one direction, the mirror image of the ex-
periment would give preferential emission in the opposite direction; if
there is no preferential emission of ,S-particles, the mirror image would
give the same result. Possible emission patterns are shown in Fig. 14-16.
The circles represent samples of Co 00 and the arrows represent electrons,
with the arrowheads showing the direction of emission. The nuclei in
a given sample are aligned so that they spin in the sense indicated by the
curved arrow. In case (a) there is no preferred direction of emission;
the electrons are emitted isotropically. In the image in the horizontal
mirror, the sense of rotation is the same and the electron distribution is
Lucite rod
Pumping tube for
vacuum space
41.S·cm
Re-entrant
vacuum space
Specimen
Housing of
CeMg nitrate
Fm. 14-17. The experimental arrangement used by Wu, Ambler, Hayward,

Hoppes, and Hudson in their demonstration of the asymmetric ~-part.icle dis-
tribution from polarized cobalt nuclei. <36>
again isotropic. In (a'), the direction in which the nuclei are aligned is
rotated through 180º; electron emission is again isotropic and the mirror
image of the experiment is identical with the experiment itself. The image
in the vertical mirror looks just like the original turned upside down.
Cases (a) and (a') show the results expected with right-left symmetry and
conservation of parity. In case (b), more electrons are emitted in the
direction opposi,te to that in which the nuclei are aligned. The electron
emission is anisotropic and the mirror image shows more electrons being
emitted in the direction .of the nuclear spins. The mirror image experiment
would then give results different from those of the original experiment.
This kind of result would be expected if parity is not conserved.
The Co 60 decay experiment was made by Wu and co-workers. <36>
Cobalt-60 emits an electron of energy 0.312 Mev in over 99% of its decays
and two other electrons in the remaining decays; severa! "Y-rays are also
emitted. It was known<rn that Co 60 nuclei can be polarized in cerium
magnesium (cobalt) nitrate and the degree of polarization determined by
14-91 SYMMETRY LAWS AND NONCONSERVATION OF PARITY 387
1.3
Gamma anistropy
1.2 (a) Equatorial counter
(b) Polar counter
Gamma. anistropy calculated from(a) and (b)

wm-W(O)
• w(J)
(At pulse height 10 v)

Exchange gas in
¡
••
0.70 O
2 4 6 8 10 12 14 16 18
Time (min)
Fm. 14-18. The results of the experiment of Wu et al. on the asymmetric
distribution of the ,6-particles from polarized cobalt nuclei.< 36>
measuring the anisotropy of the 'Y-rays. In the ,6-decay experiment,

additional difficulties were introduced by the need to put a ,6-particle
counter into the cryostat in which the sample was to be cooled, and to
locate the nuclei to be polarized in a thin surface layer. The cooling is
accomplished by a special method of low-temperature physics called adia-
batic demagnetization. A schematic diagram of the part of the cryostat
of interest is shown in Fig. 14-17.
To detect the ,6-particles, a thin anthracene crystal was located inside
the vacuum chamber near the cobalt source. The scintillations were
transmitted to a photomultiplier located at the top of the cryostat. To
measure the extent of the polarization of the Co 00 nuclei, two Nal 'Y-ray
scintillation counters were used, one in the equatorial plane and one near
the polar position. The observed 'Y-ray anisotropy provided a measure of

the polarization and hence of the temperature of the sample. Samples
were made by taking good single crystals of cerium magnesium nitrate
and growing on the upper surface an additional crystalline layer about
0.002 in thick, containing a few microcuries of Co 60 . The cerium mag-
nesium sulfate is needed because the field produced by an external
magnet cannot, by itself, align the nuclear magnetic moments due to the
spin; the external field is used to line up atoros (provided by the cerium
magnesium sulfate), wíth specíal properties, and the magnetic field pro-
duced by these atoms in turn, lines up the nuclei. After the material in
the cryostat was cooled to about 0.01 ºK, the magnet used in the cooling
process was turned off; a vertical solenoid was raised around the lower part
of the cryostat within 20 sec after the demagnetization, and the counting
was started.
The results of the experimentare shown in Fig. 14-18. The time scale
is actually a temperature scale measuring time from the instant when
cooling is stopped. The curve labelled "gamma-anisotropy" is a measure of
the extent of polarizatíon; after about 8 min the nuclei have become
sufficiently warm so that their thermal motion causes the alignment to
become random, and the 'Y-anisotropy disappears. The curve labelled
"/3 asymmetry" is the significant one; it indicates the number of electrons
emergíng in the direction of the magnetic field (the direction in which the
nuclei are aligned), and the number emerging in the opposite direction.
It is evident from the figure that more electrons are emitted in the direc-
tion opposite that of the magnetic field, i.e., in the direction opposite to
that in which the nuclei are aligned. In other words, the experimental
result corresponds to case (b) of Fig. 14-16; the electrons are emitted in
a preferred direction, the principie of right-left symmetry is violated, and
parity is not conserved. The size of the effect is remarkable; the electron
intensity in one direction along the axis of rotation was found to be 40%
greater than in the opposite direction, and there was no doubt about the
result.
Many experiments have now been made on {3-decay and on the decay
of mesons and other fundamental particles< 37 > with results similar to that
of the Co 60 experiment. The breakdown of the law of conservation of
parity has led to a modified theory of the neutrino< 35 • 52 > and to extensive
research into sorne of the basic concepts of physics, <49 > especially into the
concepts of symmetry and invariance and their relation to conservation
laws.
389
REFERENCES
GENERAL
RuTHERFORD, CHADWICK, and ELLIS, Radiations from Radfoactive Sub-
stances. New York: The Macrriillan Co., 1930, Chapters 12, 13, 14.
R. D. EvANS, The Atomic Nucleus. New York: McGraw-Hill, Hl55, Chapters
6, 17, 18, 19, 20, 21.
K. SrEGBAHN, ed., Beta- and Gamma-Ray Spectroscopy. New York: Inter-
science Publishers Inc.; Amsterdam: North-Holland Publishing Co., 1955.
M. DEUTSCH and O. KoFOED-HANSEN, "Beta-Rays," Experimental Nuclear
Physics, E. Segre, ed. New York: Wiley, 1959, Vol. III, Part XL
J. S. ALLEN, The Neutrino. Princeton, New Jersey: Princeton University Press,
1958.
C. S. Wu, "The Neutrino", Theoretical Physics in the Twentieth Century. A
Memorial Volume to Wolfgang Pauli, M. Fierz and V. F. Weisskopt, eds. New
York: Interscience, 1960.
R. W. Krno, "Table of Total Beta-Disintegration Energies,'' Revs. Mod.
Phys. 26, 327 (1954).
L. J. LmoFSKY, "Table of Total Beta-Disintegration Energies," Revs. Mod.
Phys. 29, 773 (1957).
STROMINGER, HoLLANDER, and SEABORG, "Table of Isotopes," Revs. Jfod.
Phys. 30, 585 (1958).
PARTICULAR
l. J. L. LAWSON and A. W. TYLER, "The Design of a Magnetic Beta-Ray
Spectrometer," Rev. Sci. lnstr. 11, 6 (1940).
2. L. M. LANGER and C. S. CooK, "A High Resolution Nuclear Spectrom-
eter," Rev. Sci. lnstr. 19, 257 (1948).
3. C. M. WITCHER, "An Electron Lens Type of Beta-Ray Spectrometer,"
Phys. Rev. 60, 32 (1941).
4. R. W. HAYWARD, "Beta-Ray Spectrometers," Advances -in Electronics,
Vol. 5, p. 97. New York: Academic Press (1953).
5. L. E. GLENDENIN, "Determination of the Energy of Beta-Particles and
Photons by Absorption," Nucleonics 2, 12 (Jan. 1948).
6. L. KATZ and A. S. PENFOLD, "Range-Energy Relations for Electrons and
the Determination of Beta-Ray End-Point Energies by Absorption," Revs . .llfod.
Phys. 24, 28 (1952).
7. C. D. ELLIS and W. A. WoosTER, "The Average Energy of Disintegration
of Radium E," Proc. Roy. Soc. (London) All7, 109 (1927).
8. L. MEITNER and W. ÜRTHMANN, "Uber eine Absolute Bestimmung der
Energie der primaren /:l-Strahlen von Radium E," Z. Physik. 60, 143 (1930).
9. G. J. NEARY, "The /:l-Ray Spectrum of Radium E," Proc. Roy. Soc. (Lon-
don) Al75, 71 (1940).
10. MARINELLI, BRINCKERHOFF, and HINE, "Average Energy of Beta-Rays
Emitted by Radioactive Isotopes," Revs. Mod. Phys. 19, 25 (1947).
11. J. l\1. RoBSON, "The Radioactive Decay of the Neutron," Phys. Rev.
83, 349 (1951).
12. B. W. SARGENT, "The Maximum Energy of the ¡'3-Rays from Uranium X

and other Bodies," Proc. Roy. Soc. (London) Al39, 659 (1933).
13. N. FEATHER, Nuclear Stability Rules. Cambridge: University Press, 1952,
Chapter 3.
14. E. FERMI, "Versuch einer Theorie der ¡'3-Strahlen," Z. Phyaik. 88, 161
(1934).
15. E. J. KoNOPINSKI, "Beta-Decay," Revs. Mod. Phys. 15, 209 (1943).
16. KuruE, RrcHARDSON, and PAXTON, "The Radiations Ernitted from
Artificially Produced Radioactive Substances. l. The Upper Lirnit and Shapes
of the ¡'3-Ray Spectra from Several Elements," Phys. Rev. 49, 368 (1936).
17. A. W. LAwsoN and J. M. CoRK, "The Radioactive Isotopes of lndium,"
Phys. Rev. 57, 982 (1940).
18. C. S. Wu, "Recent Investigation of the Shapes of ¡'3-Ray Spectra," Revs.
Mod. Phys. 22, 386 (1950).
19. T. H. R. SKYRME, "Theory of Beta-Decay," Progresa in Nuclear Phyaics.
Vol. 1, p. 115. New York: Academic Press, 1950.
20. E. J. KoNOPINSKI and L. M. LANGER, "The Experimental Clarification of
the theory of Beta-Decay," Ann. Rev. of Nuc. Sci., VQI. 2. Stanford: Annual
Reviews, Inc., 1953; see also, C. S. Wu, "The Neutrino," op. cit. gen. ref.
21. A. M. FErNGOLD, "Table of /t-Values in Beta-Decay," Revs. Mod. Phys.
23, 10 (1951)
22. E. FEENBERG and G. TRIGO, "The Interpretation of Comparative Half-
Lives in the Ferrni Theory of Beta-Decay," Revs. Mod. Phys. 22, 399 (1950).
23. H. R. CRANE, "The Energy and Momentum Relations in Beta-Decay
and the search for the Neutrino," Revs. Mod. Phya. ZO, 278 (1948).
24. J. S. ALLEN, The Neutrino, see gen. ref., Chapter 2.
25. J. S. ALLEN, The Neutrino, see gen. ref., Chapter 3.
26. O. KoFOED-HANSEN, "Neutrino Recoil Experimenta," Beta-and Gamma-
Ray Spectroscopy, K. Siegbahn, ed., see gen. ref., p. 357.
27. G. W. RoDEBACK and J. S. ALLEN, "Neutrino Recoils following the
Capture of Orbital Electrons in A37," Phys. Rev. 86, 446 (1952).
28. O. KoFOED-HANSEN, "Neutrino Recoil Spectrometer. lnvestigation of
A37 ," Phya. Rev. 96, 1045 (1954).
29. A. H. SNELL and F. PLEASONTON, "Spectrometry.of Recoils from Neutrino
Emission in Argon-37," Phys. Rev. 97, 246 (1955); 100, 1396 (1955).
30. J. S. ALLEN, "Experimental Evidence for the Existence of a Neutrino,"
Phys. Rev. 61, 692 (1942).
31. P. B. SMITH and J. S. ALLEN, "Nuclear Recoils Resulting from the Decay
of Be7," Phys. Rev. 81, 381 (1951).
32. R. DAvrs, JR., "Nuclear Recoil Following Neutrino Emission from Beryl-
lium 7," Phya. Rev. 86, 967 (1952).
33. J. S. ALLEN, The Neutrino, see gen. ref., Chapters 4, 5.
34. T. D. LEE and C. N. YANG, "Question of Parity Conservation in Weak
Interactions," Phys. Rev. 104, 254 (1956).
35. T. D. LEE and C. N. YANG, "Parity Nonconservation anda Two-Com-
ponent Theory of the Neutrino," Phys. Rev. 105, 1671 (1957).
REFERENCES 391
36. Wu, AMBLER, HAYWARD, HoPPEs, and HuDsoN, "Experimental Test of

Parity Conservation in Beta Decay," Phys. Rev. 105, 1413 (1957).
37. C. S. Wu, "Parity Experiments in Beta Decays," Revs. Mod. Phys. 31,
783 (1959).
38. CowAN, REINES, HARRISON, KRUSE, and McGurnE, "Detection of the
Free Neutrino; a Confirmation," Science 124, 103 (1956).
39. REINES, CowAN, HARRISON, McGUIRE, and KRUSE, "Detection of the
Free Antineutrino," Phys. Rev. 117, 159 (1960).
40. R. DA vis, "An Attempt to Observe the· Capture of Reactor Neutrinos in
Cl 37 ," (a) Phys. Rev. 97, 766 (1955); (b) Radioisotopes in Scientific Research,
Vol. 1 (Proc. First UNESCO Int. Conf. Paris, 1957). London: Pergamon Press,
1958, p. 728.
41. R. DAvrs, JR. and D. S. HARMER, "An attempt to Opserve the Cl 37
(v*,e-)A 37 Reaction Induced by Reactor Antineutrinos," unpublished.
42. F. REINES and C. L. CowAN, JR., The Free Antineutrino Absorption
Cross Section. Part l. Measurement of the Cross Section by Protons. Pro-
ceedings of the Second International Conference on the Peaceful Uses of Atomic
Energy. Geneva: United Nations, 1958. Vol. 30, p. 253.
43. C. N. YANG, "Law of Parity Conservation and other Symmetry Laws,"
Science 127, 565 (1958).
44. E. P. WrnNER, "Invariance in Physical Theory," Proc. 11m. Phil. Soc.
93, 521 (1949).
45. E. P. WrnNER, Group Theory and its Application to the Quantum Mechanics
of Atomic Spectra. New York: Academic Press, 1959, Chapters 17, 18.
46. P. A. M. DIRAC, The Principles of Quantum Mechanics, 4th ed. Oxford:
Clarendon Press, 1958, Chapter 12.
47. E. SEGRE, "Antinucleons," Ann. Rev. of Nuc. Sci. 81 127 (1958).
48. A. M. SHAPIRo, "Table of Properties of the 'Elementary' Particles,"
Revs. Mod. Phys. 28, 164 (1956).
49. G. C. WICK, "Invariance Principies of Nuclear Physics," Ann. Rev. of
Nuc. Sci. 8, 1 (1958).
50. BLIN-SToYLE, GRAcE, and HALBAN, "Oriented Nuclear Systems," Prog.
Nuc. Phys., O. R. Frisch, ed. London: Pergamon Press, 1953. Vol. 3, p. 63.
51. AMBLER, GRAcE, HALBAN, KuRTI, DURAND, and JoHNSON, "Nuclear
Polarization of Co 6º," Phil. Mag. (7) 44, 216 (1953).
52. T. D. LEE, "Weak lnteractions and Nonconservation of Parity," Science
127, 569 (1958).
PROBLEMS
l. Show that the magnetic rigidity for electrons may be written in the form
Hr = A(T2 + 1.022T) 1' 2 ,

where A = 3335.8 gauss-cm/Mev, and T is the kinetic energy of the electron.
(Hint: Use the result of Problem 12 of Chapter 6.)
2. Show that the kinetic energy may be writ~n in the form
T = [a 2 + b(Hr) 2)1 12 - a,
where a = 0.511 Mev, and b = 8.989 X 10-8
Mev 2/(gauss-cm) 2 • [Hint: Use
the result of Problem 11 of Chapter 6.)
3. An absorption curve of a sample emitting fj--particles was taken with
aluminum as the absorber and the following data were obtained.
Absorber thickness, Activity, Absorber thickness, Activity,

mg/cm 2 counts/min mg/cm 2 counts/min
0.375 16000 4.875 36

0.750 6100 7.375 13
1.75 2100 10.00 11
2.75 520 13.00 10
3.75 170
Find, by inspection, the range of the fj--particles. What is the maximum energy
of the ¡3-particles? What is the half-thickness of Al for these particles?
4. The absorption of the /j-rays from P 32 was measured with aluminum and
the following results were obtained.
Absorber thickness, Relative Absorber thickness, Relative

mg/cm 2 intensity mg/cm 2 intensity
o 1000 550 8
100 600 600 3.5
200 375 650 1.5
250 250 700 0.75
300 165 750 0.50
350 110 800 0.40
400 65 850 0.35
450 37 900 0.33
500 18 950 0.32
Find the range by inspection; then find the maximum energy from (a) Eq.
(14--7), (b) Feather's empirical formula R = 543T - 160, where R is in mg/cm 2
and T is in Mev.
PROBLEMS 393
5. In a spectrometer study of the t:1-spectrum of P 32 , the following data were

obtained.
Hr, Intensity, Hr, Intensity,

gauss-cm relative units gauss-cm relative units
500 14 4000 91
1000 27 4500 82
1500 42 5000 68
2000 58 5500 49
2500 76 6000 31
3000 91 6500 12
3500 94 7000 3
The intensity is proportional to the number of emitted electrons per unit H.r
interval. Plot (a) the momentum distribution of the electrons, (b) the energy
distribution of the electrons, (e) a Kurie plot of the momentum distribution,
(d) a Kurie plot of the energy distribution. In (e) and (d), assume that the
Coulomb factor is constant. What is the extrapolated endpoint energy in each
of the graphs?
6. In the following examples of t:1-decay, the endpoint energy of the transi-
tion is given together with the mase of the initial or final atom. Calculate the
unknown mase.
Mass of Mass of
Endpoint
Transition initial atoro, final atoro,
energy, Mev
amu amu
Cl4(¡3)Nl4 0.156 14.007685 !'

Fl7(t:l)Ol7 1.748 17.007499 ?
p29(t:l)Si29 3.94 ? 28.985661
S35 (t:l)Cl35 0.167 ? 34.97996
7. The reactions P 31 (d,a)Si 29, Si 29 (d,p)Si 30, and Si3 º(d,p) Si31 have the
Q-values 8.158, 8.388, and 4.364 Mev, respectively. Calculate the energy avail-
able for the t:1-decay of Si31 •
8. Find the difference in mase (in Mev) between a neutron and a hydrogen
atom from (a) the masses, as given in Table 11-1 ¡ (b) the endpoint of the t:1-
spectrum of the neutron; (e) the reactions 0 18 (p,n)F 18 (Q = - 2.447 Mev),
and F 18 (¡3)0 18 with an endpoint energy of 0.645 Mev; (d) the reactions 0 16
(d,p)0 17 (Q = 1.917 Mev), 0 16 (d,n)F 17 (Q = - 1.624 Mev), and F 17 (t:1)0 17
with endpoint energy of 1.748 Mev.
9. The atomic masses of Ni64 , Cu 64 , Zn64 , and Ga64 are 63.94813, 63.94994,
63.94932, and 63.95710 amu, respectively. (a) Which of these nuclides are
stable? Which are radioactive? (b) For those nuclides which can decay by
¡3-emission, what are the endpoint energies of the spectra and the Q-values of
the decay reactions?
10. Show that the ratio of the average energy to the maximum energy of
a weak ¡3-emitter (To« moc 2) is ¼, when the energy distribution is given by
the Fermi formula of Eq. (14--14).
11. A sample of hydrogen containing 2.57 cm3 of tritium at N.T.P. was
found to produce 0.1909 cal/hr of heat. The half-life of tritium is 12.46 years.
Find (a) the disintegration rate, (b) the average energy of the ¡3-particles emitted,
(e) the ratio of average energy to maximum energy.
12. A sample of RaE contains 4.00 mg. If the half-life is 5.0 days and the
average energy of the ¡3-partícles emitted is 0.34 Mev, at what rate in watts
does the sample emit energy?
13. In the nuclear fission of U 235, about 9% of the total ene:-gy liberated
comes from the ¡3-decay of fission products. How much power is carried away by
neutrinos from a nuclear power plant whlch liberates heat at the rate of 100,000
kw?
14. Plot log X against log To for the positron emitters listed below. The
result is a Sargent curve for allowed, favored ,8-decay. From the curve, get
the maximum energy of the particles emitted by AI 25 (half-Jife of 7.3 sec) and
Cl 33 (half-life of 2.0 sec).
Endpoint Endpoint
Half-life, Half-life,
Nuclide energy, Nuclide energy,
sec Mev sec
Mev
cu 1200 0.98 Mg23 11.9 2.99

Nl3 606 1.24 8 ¡21 4.9 3.48
Ql5 122 1.68 p29 4.6 3.6
Fl7 70 1.72 s¡a1 3.2 3.86
Nel9 18.4 2.18 Aªs 1.86 4.40
Na21 22.8 2.52 Sc41 0.87 4.94
15. The nuclide Sm 153 emita four groups of ,B-rays, with endpoint energies of
0.83, 0.72, 0.65, and 0.13 Mev, respectively; 'Y-raya with energies of 0.1032,
0.0697, 0.172, 0.545, and 0.615 Mev, respectively, are also emitted. Construct
a decay acherne to fit these data. (Note that the 'Y-ray energies are known with
hlgher precision than the ,8:.ray energies.)
16. The nuclide As76 einits four groups of ¡3-rays with endpoint energies of
2.97, 2.41, 1.76, and 0.36 Mev, respectively, and 'Y-raya with energies 0.561,
0.643, 1.200, 1.40, and 2.05 Mev. Devise a decay acherne to fit these data.
17. The nuclide Tc94 decays by orbital electron capture and ,B+-emission;
'Y-raya are found, with energies 3.27, 2.73, 1.85, and 0.874 Mev, respectively,
as well as positrons with an endpoint energy of 2.41 Mev. Devise a decay acherne
to fit these data. Compare your acherne with that given in the "Table of
Isotopes" of Strominger, Hollander, and Seaborg (gen. ref.).
CHAPTER 15
GAMMA-RAYS AND GAMMA-DECAY
The importance of 'Y-rays as a. source of information about nuclear energy

levels has been shown in Chapters 13 and 14 in connection with a-decay
and fj-decay. In those chapters, the emission of 'Y-rays was regarded
simply as a means by which a nucleus can pass from an excited state to a
less excited one. Gamma-decay is in itself, however, a subject of great
theoretical and practica! importance and will therefore be discussed in
greater detail in this chapter. The study of 'Y-decay depends on the ability
to measure the energy of 'Y-rays with high precision, but the problem of
measuring these energies is different from that presented by charged
particles. Since 'Y-rays are electromagnetic radiations and have no electric
charge, they cannot be deflected by magnetic or electric 6.elds. Con-
sequently, direct measurements of their energies with a magnetic spectro-
meter are not possible. The mechanism of the absorption of 'Y-rays by
matter is also different from that of charged particles, as is indicated by
the very much greater penetrating power of 'Y-rays. Because of these
properties of 'Y-rays, it will be necessary to examine first the interactions
between 'Y-rays and matter. The methods used for measuring 'Y-ray energies
will then be treated because they are based on these interactions, and then
the subject of 'Y-decay will be taken up.
15-1 The absorption of gamma-rays by matter: experimental data. The

basic property of the absorption of 'Y-rays is the exponential decrease in the
intensity of radiation as a homogeneous beam of 'Y-rays passes through a
thin slab of matter. When a beam of 'Y-rays of intensity I is incident on a
slab of thickness .:1x, the change in intensity of the beam as it passes through
the slab is proportional to the thickness and to the incident intensity
.:1/ = -µ[ .:U, (15-1)
where the proportionality constant µ is called the absorption coeffici,ent.

If the 'Y-ray photons all have the same energy, µ is independent of x, and
the integration of Eq. (15-1) yields
(15-2)
Equation (15-2) gives the intensity of radiation / after a beam of initial

395
396 GAMMA-RAYS AND GAMMA-DECAY [CHAP. 15
- - P b 125cm
- -Pb
Cpunter
circwts ,~ ~Counter
Frn. 15-1. Schematic plan view of the apparatus used for the measurement
of total 'Y-ray absorption coefficients (Davisson and EvansOl).
intensity I O has traversed a thickness x of a given material. The intensity

may be written
I = Bhv, (15-3)
where B is the number of photons crossing unit area in unit time, and hv is
the energy per photon; B is often called the flux, defined as the number of
photons per square centimeter per second, and I is the corresponding
energy flux. Equation (15-2) may then be written
(15-4)
The equations which describe the absorption of 'Y-rays are the same as those
for x-rays as discussed in Section 4-1. This result is not surprising because
both kinds of rays are electromagnetic radiations. In fact, it is not possible
to differentiate between 'Y-rays and x-rays on the grounds of differences in
their properties, since no differences have been found. The terms 'Y-rays
and x-rays are used now chiefly to distinguish between the sources of the
rays; 'Y-rays come from nuclei, while x-rays are the high energy radiations
resulting from jumps of extranuclear electrons in atoms, or produced by
artificial sources such as a Coolidge tube or a betatron. The discussion of
absorption in this section applies to both 'Y-rays and x-rays.
It is necessary to define precisely the conditions under which Equations
(15-1), (15-2), and (15-4) are valid. These conditions are (1) the 'Y-rays
15--1) THE ABSORPTION OF GAMMA-RAYS BY MATl'ER 397
0.5
e: 0.2
·i 10-1
·e 5
~ 2
E-< 10-2
5
2 .___...._J...._.____.__~_.____.___,
o 2345678
Thickness of lead (cm}
Frn. 15--2. Experimental verification of the narrow-beam attenuation for-

mula for 'Y-rays, and measurement of the absorption coefficient in lead. Source,
1.14 Mev 'Y-rays from Zn 65 •
are monoenergetic, i.e., the beam is homogeneous; (2) the beam is colli-
mated and of small solid angle; (3) the absorber is "thin." An experi-
mental arrangement for studying the decrease in intensity, or attenuation,
of a 'Y-ray beam under these conditions is shown in Fig. 15-1. This arrange-
ment provides a narrow collimated beam and is spoken of as "good
geometry." The thick lead shielding blocks, separated by a narrow space,
act as the collimator. The counter used to detect the 'Y-rays is surrounded
by a lead shield to absorb extraneous 'Y-rays and electrons which might
otherwise be scattered into it and counted. In a "good geometry" experi-
ment, photons are removed from the beam either by absorption or by
scattering. A photon which undergoes a scattering collision with an atom
in the absorber slab may be deflected enough so that it does not reach the
detector and is lost from the beam just as though it had actually been
absorbed. The decrease in the intensity of the beam in passing through
the slab is a measure of the combined effects of outright absorption and
deflection. The term absorption is often used as synonymous with attenua-
tion, but this practice causes less confusion than might be expected because
the meaning is usually clear from the context.
The validity of Eqs. (15-2) and (15-3) can be shown by a "good geom-
etry" experiment. The results of such an experiment are given in Fig.
15-2; the absorber is lead, and the source is Zn 65 , which cmits 'Y-rays with
an cnergy of 1.14 Mev. When the fractional transmission is plotted on a
logarithmic scale against the absorber thickness on a linear séale, the result
is a straight line in agreement with the equations. The slopc of the linc
gives thc value µ (cm- 1) = 0.7068 ± 0.0051. Many experimental <leter-
minations of absorption coefficients havc been made in this way and it has
been found that the value of the absorption coefficient <lepends on the
nature of the absorber and on the initial encrgy of the 'Y-rays. Sorne
values obtained by Davisson and Evanscii are listed in the third column
of Table 15-1. It is evident that for a given element, the absorption
coefficient decreases as the energy of the 'Y-rays increases. The value of µ.
also varíes from element to element and is, in general, greater for heavy
elements than for light elements. Theoretical values are listed for future
reference.
The thickness x may be expressed in cm or in gm/cm 2 , as was done in
Chapter 14 in the discussion of the absorption of ~particles. Since the
product µ.x must be dimensionless, µ. may be correspondingly expressed in
cm- 1 or cm 2 /gm. It is sometimes convenient to express the absorber
thickness in atoms/cm 2 or electrons/cm 2 rather than in gm/cm 2 ; the
absorption coefficiei:it then has the units cm 2 /atom or cm 2 /electron. The
absorption coefficient is usually denoted byµ. when the units are cm- 1 , by
µ/p for cm 2/gm, by .µ. for cm 2/electron, and by 4 µ. for cm 2/atom. The
subscripts e and a are placed at the lower left in order to avoid confusion
later on. In terms of .µ. and aµ., the other coefficients are
aµ.= Z.µ., (15-5)
·: = N (!) .µ. = 1aJl, (15-6)
µ = pN A (z) .µ =
pN
A aµ, (15-7)
where Z is the atomic number, A is the atomic weight, N is Avogadro's

number, and p is the density in gm/cm 3 • In Eq. (15-7), the product
pN/ A • Z is the number of electrons/cm 3 of absorber and, since .µ. has
the units cm 2 /electron, µ. comes out in cm- 1• The coefficient 4 µ. has the
units cm 2/atom, and when it is multiplied by pN/ A, the number of atoms/
cm 3 , the result is again µ. in cm- 1 • The coefficients .µ. and 4 µ. are often
called the "cross section per electron" and the "cross section per atom,"
respectively, because of their units.
The attenuation of a 'Y-ray beam may also be expressed in terms of a
quantity called the half-thickness, i.e., the thickness of absorber needed
to reduce the intensity to half its initial value. Equation (15-2) may be
written in terms of the common logarithm
log (l / I 0 ) = -0.4343µ.x. (15-8)

When I/1 0 = ½,
log (½) = -0.4343µ.x1¡2,
and
0.693
µ. = - -,
X1¡2
(15-9)
15--1] THE ABSORPTION OF GAMMA-RAYS BY MATTER 399
TABLE 15--1
MEASURED VALUES OF SoME ÜAMMA-RAY ABSORPTION CoEFFICIENTs
Experimental Theoretical
absorption absorption
Source Absorber coefficient,
coefficient,
µ: cm- 1 µ: cm- 1
Mn54 Al 0.1823 ± 0.0003 0.1820

(0.835 Mev) Cu 0.5782 ± 0.0013 0.5718
Sn 0.4683 ± 0.0014 0.4628
Ta 1.210 ± 0.004 1.228
Pb 0.9368 ± 0.0041 0.9256
zn65 Al 0.1571 ± 0.0022 0.1559

(1.14 Mev) Cu 0.4862 ± 0.0070 0.4914
Sn 0.3923 ± 0.0054 0.3858
Ta 0.9127 ± 0.0100 0.9536
Pb 0.7068 ± 0.0051 0.7057
Na24 Al 0.0956 ± 0.0026 0.1001

(2.76 Mev) Cu 0.3164 ± 0.0080 0.3273
Sn 0.2668 ± 0.0045 0.2692
Ta 0.6433 ± 0.0055 0.6467
Pb 0.4776 ± 0.0045 0.4644
where x 1 ¡ 2 is the half-thickness. In the example of Fig. 15-2, the half-

thickness of lead for 1.14 Mev 'Y-rays is 0.98 cm.
A useful rule of thumb may be deduced from the absorption coefficients.
In Eqs. {15-6) and {15-7), the ratioZ/ A changes very slowly asZ increases
and, as will be shown later,'.µ is approximately the same for all elements
in a certain energy region. Hence, the absorption coefficient shows the
smallest variation from element to element when expressed as .µ or µ/ p.
From Eq. {15-9),
{15-10)
and, since µ/p varies slowly with Z, the product {x 112 )p also varies slowly
from element to element. It follows from this result that the greater the
density the smaller the thickness needed of a given material to decrease
the 'Y-ray intensity to a specified extent. For this reason, heavy metals
such as iron, and especially lead, are used for shielding against 'Y-rays and
x-rays. The approximate constancy of µ/p a.nd (x 112 )p means that the
weights of different materials needed to decrease the intensity of the
radiation by a certain fraction are very nearly the same. But for sub-
stances of higher density, the volume, and hence the thickness, will be Iess
than for materials of lower density.
15-2 The interaction of gamma-rays with matter. The interaction

of 'Y-rays with matter is markedly different from that of charged particles
such as a- or (3-particles. The difference is apparent in the much greater
penetrating power of 'Y-rays and in the absorption laws. Gamma-rays
and x-rays, which are both electromagnetic radiations, show a character-
istic exponential absorption in matter, and have no definite range such as is
found for charged particles. Charged particles, especially heavy ones,
lose their energy during the course of a large number of collisions with
atomic electrons. The energy loss occurs in many small steps and the
particle gradually slows down until it is stopped altogether and ab30rbed.
When a beam of 'Y-ray photons is incident on a thin absorber, however,
each photon that is removed from the beam is removed individually in a
single event. The event may be an actual absorption process, in which case
the photon disappears, or the photon may be scattered out of the beam.
The one shot nature of the removal process is responsible for the expo-
nential absorption. For, the number of photons that can be removed in
passing through a thickness '1x of absorber is proportional to t.x and to the
number of photons reaching '1x; this kind of dependence leads directly to
the exponential absorption law, as shown by Eqs. (15-1) and (15-2).
Three processes are mainly responsible for the absorption of 'Y-rays.
These are (1) photoelectric absorption, (2) Compton scattering by the
electrons in the atoms, and (3) production of electron-positron pairs as a
result of the interaction between 'Y-rays and the electric fields of atomic
nuclei. Quantum mechanics makes it possible to derive formulas for the
probability of each process and the probability can be expressed as an
absorption coefficient or as a cross section. The total absorption coeffi-
cient, which appears in Eqs. (15-1) and (15-2), is the sum of the absorp-
tion coefficients for the three different processes. The problem is further
complicated by the fact that the absorption coefficient depends on the
energy of the incident 'Y-rays as well as on the nature of the absorbing
material. Consequently, the absorption of 'Y-rays cannot be described by
a single formula, or by a range-energy curve. Each partial cross section
or absorption coefficient must be evaluated as a function of energy for a
given material, and tables or sets of curves can then be prepared from
which values of the total absorption coefficient of a given material can be
obtained. Thus,
µ(E) = T(E) + u(E) + K(E), (15-11)
where T, u, and K denote the photoelectric, Compton, and pair formation

coefficients, respectively, and each absorption coefficient is related to the
15-2] INTERACTION OF GAMMA-RAYS WITH MATI'ER 401
corresponding cross section by Eq. (15--7). In the following sections,

formulas for the cross sections for the three processes will be written down,
discussed, and illustrated by means of curves and tables. Before the
detailed treatment is begun, the processes will be discussed in a general way.
In the photoelectric process, all the energy hv of the incident photon is
transferred to a bound electron which is ejected from the atom with a
kinetic energy T = hv - I, where I is the ionization potential of the
electron. The electron may be ejected from the absorber or, if the absorber
is not too thin, is more likely to be reabsorbed almost immediately becauEe
of the short range of an electron in a solid. At low photon energies, below
50 kev for aluminum and 500 kev for lead, the photoelectric effect gives the
chief contribution to the absorption coefficient.
As the energy of the radiation increases, Compton scattering replaces
the photoelectric effect as the chief means of removing photons from the
initial beam. It was seen in See;tion 6-7 that in Compton scattering the
incident photon is scattered by one of the atomic electrons and the latter is
separated from its ato:ni. The photon moves off at an angle with its
originaldirection and with less energy than it had initially. The change of
direction serves to remove the photon from the incident 'Y-ray beam.
Compton scattering gives the main contribution to the absorption coeffi-
cient between 50 kev (0.050 Mev) and 15 Mev for aluminum, and between
0.5 and 5 Mev for lead. In these energy ranges, the energy of the incident
'Y-ray is much greater than the binding energy of the atomic electrons.
The process may therefore be regarded as the scattering of a photon by a
free electron initially at rest and, since each electron in the atom scatters
ú1dependently, the Compton absorption coefficient per atom is proportional
" the atomic number Z.
t\.t high enough energies, both the photoelectric absorption and the
npton scattering become unimportant compared with pair formation.
;he latter process, in the Coulomb field of an atomic nucleus, a 'Y-ray
1 sufficient energy disappears andan electron and positron are created.
J total energy of the pair is equal to the energy hv of the incident 'Y-ray;
1';inetic energy T of the pair is
(15-12)
if the small recoil energy of the nucleus is neglected. For pair production to
occur, hv must be greater than 2m 0 c2 or 1.02 Mev. Pair production
cannot occur for h1I < 2m 0c2 because this amount of energy is needed to
supply the rest energy of the two particles. A pair of particles, with
opposite charge, must be formed if charge is to be conserved. At photon
energies greater than 5 Mev for lead and 15 Mev for aluminum, the prob-
abilíty of pair production is greater than that for Compton scattering and
continues to increase with increasing energy.
402 GAMMA.-RAYS AND GAMMA-DECAY [CHAP. 15
In addition to the three effects discussed, there are sorne minor effects
which may contribute to the attenuation oí a 'Y-ra.y beam. The most
important oí these is coherent scattering by whole atoros or molecules,
which may add severa} percent to the absorption coefficients of high-Z
materials at low 'Y-ra.y energies. Among the other effects are (1) the
nuclear photoelectric effect, in which high-energy photons eject neutrons
from the nuclei of high-Z materials, and (2) Thomson and Compton
scattering by nuclei ra.ther than by electrons. For most pra.ctical pur-
poses, all oí these effects may be neglected.
The resulta of the theory of the intera.ction between 'Y-rays and matter
ha.ve been collected and compared with experiment in the excellent review
article< 2 > by C. M. Davisson and R. D. Evans. Many tables and curves
are given there which present much of the presently available information
in useful forro.
15-3 Photoelectric absorption. Formulas for the probability that a

photon of energy h11 will undergo photoelectric absorption ha.ve been
derived by quantum mechanical methods. Severa.l different formulas must
be used ü the probability, expressed asan absorption coefficient, is to be
calculated over the range of photon energies from 0.1 Mev to 5 or 10 Mev.
The reason for this is that the theory ir- complicated and different assump-
tions are made in different energy ranges in order to ease the mathematical
difficulties. The main features oí the dependence of the absorption
coefficient on the atomic number of the absorber and on the energy of the
photon are shown by the simplest of the formulas. If the energy oí the
photon is sufficiently small so that relativistic effects are not important
but large enough so that the binding energy of the electrons in the K-she
may be neglected, the cross section per atoro ,..,. for photoelectric absorpti
is
,..,. = 4>oZ
5 ( 1)t 4v'2 (moe
137 1w
2
)
7 2
'
' (15-
r
where
'Po = \11" (::2 = 6.651 X 10-

25
cm 2. (15-
In Eq. (15-13), hv is the energy oí the incident photon, m0 c2 is the res,

energy oí the electron, and Z is the atomic number of the absorbing ma-
terial; q,0 is a convenient unit for measuring the cross section and repre-
senta the cross section for the scattering of low energy photons by a free
electron at rest. This kind of scattering, called Tlwmson scattering, was
discussed in Section 4-2. The simple equation (15-13) applies only to the
ejection of electrons from the K-shell of the atom, which accounts for
about 80% of the photoelectric effect.
15-3] PHOTOELECTRIC ABSORPTION 403
The most important property of aT is the strong dependence on the

atomic number and on the energy of the incident photon; aT is propor-
tional to Z 5 and inversely proportional to (hv) 712 . The Z 5 dependence
means that, for a given photon energy, the process of photoelectric absorp-
tion is much more important in heavy metals such as lead than in light
ones such as aluminum. The energy dependence shows that for a given
element, the effect is much greater at small photon energies than at large
ones. It follows that the photoelectric effect is especially important in the
absorption of low energy photons by heavy elements.
In the more rigorous formulas, the dependence of aT on Z and hv is not
quite the same as that shown in Eq. (15-13). Thus, the variation of aT
with Z does not follow the simple fifth power law, although for practica!
purposes the deviation from a power law is small. The exponent givini
the best agreement with theory increases with "Y-ray energy and is between
4 and 5 for energies above 0.35 Mev, being approximately 4.5 at 1.13 Mev
and 4.6 at 2.62 Mev. The rigorous dependence of aT on the energy of the
"Y-rays also cannot be described by a simple power law. The cross section
decreases rapidly with increasing energy, approximately as (hv)- 3 for
energies less than 0.5 Mev, andas (hv)- 1 for energies greater than 0.5 Mev.
Values obtained from the more rigorous theory are given in Table 15-2.
For purposes of comparison and ease of use, the values of the quantity
TABLE 15-2
THE PHOTOELECTRIC EFFECT*
Energy of "Y-raya z =O Z .;, 13 Z = 26 z = 38 Z = 50 z = 65 Z = 82

n = moc 2 /hP Mev Al Fe Sn Pb
=====-.=----------,--=-==== ===== ===== ==-=== =====
o 0.353 0.275 0.228 0.203 0.188 0.166 0.159
0.1 5.108 0.436 0.343 0.289 0.255 0.230 0.206 0.182
0.125 4.086 0.453 0.362 0.306 0.272 0.247 0.216 0.188
0.194 2.633 0.524 0.412 0.347 0.302 0.270 0.237 0.205
0.25 2.043 0.581 0.447 0.375 0.331 0.290 0.259 0.225
0.375 1.362 0.740 0.556 0.462 0.403 0.366 0.319 0.278
0.50 1.022 0.956 0.694 0.575 0.500 0.440 0.397 0.3-11
0.75 0.6711 1.528 1.162 0.947 0.781 0.690 0.600 0.522
1.0 0.5108 2.375 1.862 1.506 1.237 1.038 0.878 0.747
1.25 0.4086 3.578 2.756 2.188 1.815 1.512 1.262 1.031
1.5 0.3405 5.3 3.834 3.062 2.528 2.116 1.750 1.362
2.0 0.2554 9.48 7.46 5.34 4.24 3.55 2.96 2.17
3.0 0.1703 24.3 17.61 12.44 8.81 7.88 6.25 4.18
4.0 0.1277 46.8 33.1 22.8 14.8 14.2 10".5 6.6-1
5.0 0.1022 79.9 53.4 34.6 21.8 21.6 15.6 9.61
*Numerical values of aT/Z 5n in units of 10-32 cm 2.

404 GAMMA-RAYS AND GA.MMA-DECAY [CHAP. 15
100
\
\
'
\
10
\
..
'' ,P,
\
\
i
!.
.. \ . fl
1
!
O\
-~ '
IS '\
~
c.> \ \ \
,=
o
'.;:l \ \
e-jJ 1 Fe
\
~
c.>
·.5c.>
0.1
\
\
. \
jo '
'
.e:
11. ' '' "
\ \ ---
\ \ \
\ ~ ~
Al
\
\ \.
""
\
.01
. ...
'\
' '-
...
\
'
\
\
\
\ \; '
\ \
\
0.1
' 100
Energy (hv/mcc2)
Fm. 15-3. The absorption coefficient per centimeter as a function oí the

photon energy for photoelectric absorption in aluminum, iron, tin, and lead.
15-4] COMPTON SCATTERING 405
aT/Z 5 nare listed, where nis the ratio m 0 c2 /hv. Increasing values of n
correspond to decreasing 'Y-ray energies. For a particular absorber, say
lead, the values of aT are obtained for different 'Y-ray energies by multiply-
ing the values listed in the table first by Z 5 = (82) 5 , and then by the
appropriate value of n. Curves obtained in this way are shown in Fig.
15-3 for aluminum (Z = 13), iron (Z = 26), tin (Z = 50), and lead
(Z = 82). It will be noted that the values of aT/Z 5n listed in the table
are in units of 10-32 cm 2 . Values of aT for energies and atomic numbers
other than those listed can be obtained by plotting the appropriate curves
and interpolating. The values for Z = O are included to allow interpola-
tion for elements with Z < 13.
15-4 Compton scattering. The main features of Compton scattering

were discussed in Section 6-7, but it will be useful to repeat sorne of the
results obtained there in somewhat different form. The energy lost by a
photon in a single Compton scattering procesa can be obtained from Eq.
(6-52), which gives the change in wavelength of the photon,
h
X- >-o = -moc (1 - cos ,p). (15-15)
In Eq. (15-15), >.. 0 is the wavelength of the photon before collision, >.. is the
wavelength after the collision, and q, is the angle between the initial and
final directions of the photon as shown in Fig. 6-7. Since X = c/11, Eq.
(15-15) may be written
-e - -e = - h (1 - cos ,p)
11 "º moe '
or
-1 = -1 h
+ -- (1 - cos q,).
" "º moc2
Inversion of both sides of the last equation gives
1
" = ---------
__!_ + _h_ (1 - cos q,)
"º moc
2
If the top and bottom of the right side are multiplied by 11 0 , and if both
sides of the resulting equation are multiplied by h, the result .is
hvo
hv = --....-----='-----
hvo (15-16)
1 + -m-c2 (1
0
- cos q,)
Equation (15-16) gives the energy of the scattered photon in terms of the
initial energy and the scattering angle. The recoiling electron has a
kinetic energy given by
T = hv 0 - hv
(1 - cos q,) hvo

moc2
(15-17)
hvo
1 + m 0c- - 2 (1 - cos q,)
The kinetic energy of the electron has its maximum value when
cos q, = -1 or q, = 180º, and the photon is scattered directly backward.
The electron energy, in this case, is
T - hvo (15-18)
max - 1 + (moc 2/2hvo)
The electron receives the least energy in a grazing collision in which the
photon continues with its initial frequency in the forward direction, and the
electron is ejected with very nearly zero velocity in a direction perpendicu-
lar to that of the photon path.
The discussion so far refers only to a single Compton scattering process.
In order to treat the contribution of the Compton effect to the attenuation
of a beam of 'Y-rays in matter, it is necessary to calculate the probability
that such a scattering process will occur. This probability was calculated •
on the basis of relativistic quantum mechanics by Klein and Nishina. •
Although the details of the theory are complicated, the results may be
expressed in straightforward formulas with which calculations can be
made quite easily. A formula was obtained for the cross section per
electron .u for the removal of photons from the incident beam by scattering:
.U= i4'o
!.1n (1 + 2a)] +_!._In (1 + 2a) 1 + 3a 2} '
X {1 +a
a2 1++2aa)
[2(1 -
a 2a
-
(1+ 2a)
(15-19)
where a has been defined as
(15-20)
and q,0 is given by Eq. (15-14). When .u is multiplied by pN(Z/ A), the
result is the Compton absorption coefficient u (cm- 1),
(15-21) .
15-4) COMPTON SCATTERING 407
The coefficient u is a measure of the probability that a photon is scattered

out of the beam per centimeter of absorber, and since the beam is initially
homogeneous, u is also a measure of the total amount of energy removed
from the beam per centimeter of absorber. It is sometimes necessary to
know the amount of energy carried off by the scattered photons or the
amount of energy absorbed by the recoil electrons. These amounts may be
expressed in terms of two quantities denoted by ,u, and ,<Ta. The former,
,u., may be called the "Compton scattering cross section per electron for
the energy of the scattered photon." It is given by
2 2
3 [ 1 2(1 +
a)(2a - 2a - 1) 8a ]
,<T, = g <Po aª ln (l + 2 ª) + a2(1 + 2a)2 + 3(1 + 2a) 3 •
(15-22)
The quantity ,<Ta is the "Compton cross section per electron for the energy
absorbed." The coefficient .u is simply the sum of ,<T• and ,ua:
(15-23)
so that when ,u and ,u, are known, ,<Ta is obtained by subtraction. In

analogy with Eq. (15-21),
z
<T, = pN A .u.,
(15-24)
z
<Ta = pN A e<Ta.
Finally, the sum

<T = <T, + <Ta, (15-25)
shows how the total Compton coefficient is broken down into a coefficient
for the energy scattered out of the beam and a coefficient for the energy
removed by the recoil electrons. The electrons, because of their very
short range, are often absorbed and their energy may appear in the form
of heat. In sorne practica! problems it is necessary to know how much heat
is produced as a result of the attenuation of a beam of 'Y-rays in passing
through a given absorber. The contribution of Compton scattering to
this heat can be obtained when ,<Ta is known for the particular Y-ray energy.
It is convenient to tabulate the values of ,u, ,u., and ,<Ta, i.e., the values
in cm 2 /electron, because these are independent of the properties of the
absorber. The values for a particular absorber are then obtained by using
Eqs. (15-21) and (15-24). Table 15-3 gives the cross sections for the
Compton effect .u, ,u,, and ,<Ta in units of 10-25 cm 2/electron. The values
of the absorption coefficients per centimeter, <T8 , <Ta, and u, for the special
case of lead (Z = 82) are plotted in Fig. 15-4.
1.75
'"r-.r-., ~
1.50 'i-.. i'-11.
' ~\
1.25
\ '\ ,, - ,,, + "ª'total Compton coefficient
!"il 1\,1\~ \
,,,, Compton scattering coefficient
"ª' Compton absorption coefficient
1.00
i "~
e
"! 0.75
\
r,...
', 1
1~
j
< '\ \ ~ t:I
0.50
' '
~"- "' ,
'\
!
~
--
......
0.25
..,::-- ~- -
-ot.::::: ..........
.....
.
~
- ...... ..........
. =:::::::::::-
.01
~/
.,.- ...---
.,. ...
0.1 1.0
............ ........
10
---= -- 100 'ñ'
Photon energy (h•!mcc2)
~
F10.15--4. The absorption coefficients per centimeter as functions of the photon energy for the Compton
effect in lead; u is the total coefficient, u, is the Compton scattering coefficient, and <Ta is the Compton absorp-
-
e,,
tion coefficient. ·• • ' · ·

15-4) COMPTON SCATI'ERING 409
TABLE 15-3
CROSS 8ECTIONS FOR THE COMPTON EFFECT*
Initial energy
e<T e<T, e<T a
a = h11/moc 2 Mev
0.025 0.0128 6.31 6.31 0.00

0.05 0.0256 6.07 5.79 0.28
0.075 0.0383 5.83
0.10 0.0511 5.599 5.138 0.461
0.15 0.0767 5.243
0.20 0.102 4.900 4.217 0.683
0.25 0.128 4.636
0.30 0.153 4.410 3.597 0.813
0.40 0.204 4.032 3.152 0.880
0.50 0.255 3.744 2.818 0.928
0.60 0.307 3.507 2.551 0.956
0.70 0.358 3.309 2.337 0.972
0.80 0.409 3.140 2.158 0.982
0.90 0.460 2.994 2.008 0.986
1.0 0.511 2.866 1.879 0.987
1.5 0.767 2.397 1.432 0.965
2.0 1.022 2.090 1.164 0.926
2.5 1.278 1.868 0.983 0.884
3.0 1.533 1.696 0.852 0.844
3.5 1.789 1.559 0.753 0.806
4.0 2.044 1.446 0.674 0.772
4.5 2.300 1.351 0.611 0.739
5.0 2.555 1.269 0.559 0.710
6.0 3.066 1.136 0.477 0.659
7.0 3.577 1.031 0.417 0.614
8.0 4.088 0.9465 0.370 0.5765
9.0 4.599 0.876 0.333 0.543
10.0 5.108 0.8168 0.3023 0.5145
12.0 6.132 0.722 0.256 0.466
20.0 10.22 0.502 0.158 0.344
30.0 15.33 0.371 0.107 0.264
50.0 25.55 0.250 0.066 0.184
70.0 35.77 0.191 0.047 0.144
100.0 51.10 0.143 0.033 0.110
*In units of 10-25 cm 2 /electron.

Equations (15-19) and (15-22) show that the Compton scattering per
electron is independent of Z, so that the scattering per atom is proportional
to Z. The mass scattering coefficient 11'/ p is given by
11' z
P =NA .u, (15-26)
and varies only slowly with Z. For the light elements Z/ A is closely
equal to ½, so that for a given photon energy 11'/ p is practically constant for
these elements. The total scattering coefficient per electron, .u, decreases
with increasing photon energy, as can be seen from the values in Table 15-3.
The decrease is quite slow at low values of the energy, and for energies
above 0.5 Mev, .u is roughly proportional to (hv)- 1 . Thus Compton
scattering decreases much more slowly with increasing energy than does
photoelectric absorption; even in heavy elements, it is the most important
process in the energy range from about 0.6 to 2.5 Mev.
In the treatment so far, it has been assumed that a singly-scattered
photon was removed from the 'Y-ray beam. This assumption may be con-
sidered to define a thin absorber. In a thick absorber sorne of the singly-
scattered photons which have been considered lost will be scattered and
rescattered and may eventually reach the detector. The problem of
repeated, or multiple, Compton scattering of 'Y-rays is important in con-
nection with the shielding of personnel against the 'Y-radiation emitted by
accelerators or nuclear reactors. Considerable progress has been made
in the theoretical and experimental treatment of this problem, and the
reader is referred to the book by H. Goldstein listed in the general ref-
erences at the end of the chapter.
15-5 Electron-positron pair formation. The third mechanism by

which electromagnetic radiation can be absorbed is the production of
electron-positron pairs. This process, which has no analogy in classical
physics, may be accepted as a strictly experimental phenomenon. The
discovery of the positron and of pair formation, however, represents one
of the great triumphs of modern physical theory, and for this reason a
short digression seems worth while. It was mentioned very briefly at the
end of Section 7-5 that the Schroedinger equation, the fundamental
equation of wave mechanics, did not satisfy the requirement of invariance
postulated by the thoery of special relativity. When the Bohr theory of
the atom was replaced by wave mechanics, the problem of the relativistic
treatment of the electron still remained. This problem was solved by
Dirac, who derived an equation, the Dirac electron equation, which
is one of the most important developments in modern physics. Dirac
required that the equation which represents the motion of an electron be
invariant under the Lorentz transformation (see Sections 6--4 and 6-5).
15-5) ELECTRON-POSITRON PAIR FORMATION 411
Frn. 15-5. Possible values of the energy of an electron according to the

Dirac theory.
The solution of the Dirac equation led in a natural way to the formula for
the fine structure of spectral lines and to the spin of the electron. But to
solve the Dirac equation it was necessary to assume that the electron can
exist in two sets of quantum states, one of positive energy (including the
rest energy), and the other of negative energy. It was found that the possi-
2
ble values of the energy of a free electron are either greater than +m 0 c or
2
smaller than -m0 c , and that no possible energies for the electron exist
between these two limits. This state of affairs is shown in Fig. 15-5, where
the shaded regions are those in which values of the energy exist. Electrons
in states of positive energy behave in the usual manner of electrons that are
ordinarily observed, while electrons in states of negative energy should have
properties which have no classical analogy. These electrons could simply
be neglected if they were considered from a classical viewpoint. For, in
that case, the value of the energy of an electron could change only in a
continuous way, andan electron in a positive energy state could not bridge
the discontinuous gap between a state of energy +m 0 c2 and one of energy
-m 0c2 . In other words, a state of negative energy would have no real
physical meaning. According to quantum theory, however, an electron
can make a discontinuous transition from one energy state to another, so
that there is no way of ruling out a jump from a positive energy state to a
negative energy state.
Dirac avoided the difficulty by assuming that the states of negative
energy are real, but that they are ali usually occupied. The electrons
which are ordinarily observed are those in positive energy states. Suppose
that one electron is missing in the distribution of negative energy states.
The empty state, according to Dirac, would appear as a particle with
positive energy and positive charge, since a particle of negative encrgy
and negative charge is absent. This empty place, or "Dirac hole," would
therefore behave like a positively charged particlc. Dirac first assumcd
10.-------------------,
Unscreened
Photon energy (hv/~)

Fig, 15--6. Pair-formation cross section per atom as a function of the photon
energy; ª" in units of if,. (By permission, from Experimental Nuclear Physics,
Vol. I., E. Segre, ed., New York: Wiley, 1953.)
TABLE 15-4
PAIR-FORMATION CROSS SECTIONS
Photon energy, alC/Z 2,

Mev aK/if, in units of
10-27 cm 2 /atom
1.533 0.086 0.050

2.043 0.327 0.189
3.065 0.905 0.524
5.108 1.98 1.15
Al Pb Al Pb
7.662 2.93 2.86 1.70 1.66
10.22 3.66 3.60 2.12 2.08
17.01 5.07 4.90 2.94 2.84
25.54 6.14 5.96 3.56 3.46
15-5) ELECTRON-POSITRON PAIR FORMATION 413
that it represented a proton, but this assumption had to be discarded

because it could explain neither the difference in mass between a proton
and an electron nor the stable existence of the proton.
The dilemma was resolved in 1933 when Anderson, during the course of
a study of cosmic rays by means of a cloud chamber, observed particles
with the same mass as that of the electron and with an electric charge
equal in magnitude but opposite in sign to that of the electron. This
particle was given the name -positron and identified with the Dirac hole.
In terms of Dirac's theory, the production of a positron is interpreted as
follows. A photon of energy greater than 2m 0c2 can raise an electron from
a state of negative energy to a state of positive energy. The disappearance
of an electron from a negative energy state leaves a hole, which means the
appearance of a positron; the appearance of an electron in a positive energy
state means the appearance of an ordinary electron. Thus, a pair of
particles is created. Since eaeh particle must have a rest mass equal to
m 0c 2 , the photon must have an energy of at least 2m 0 c 2 , or 1.02 Mev. The
creation process usually takes place in the electric field in the neighborhood
of a nucleus, because sorne body must be given recoil energy and momentum
in order that energy and momentum be conserved in the system.
The formulas for the probability of pair formation are more complicated
than those for photoelectric a.bsorption a.nd Compton scattering, and will
not be written down here. Values of the cross section per atom, aK, for
pair production can be listed for certain photon energies and curves for the
variation of aK with energy can be given. The cross section is zero for
photon energies less than 1.02 Mev; for greater energies, it increases at
first slowly, then more rapidly. lt is proportional to Z 2 , so that. for a given
photon energy, pair formation increases quite rapidly with atomic number.
Pair production cross sections are usually expressed in units of the quantity
2 2 2
Z ( moc
q, = 137 e ) = Z 2 X 5.796 X 10 -28 cm 2 . (15-27)
2
Curves of aK/q, as a function of a = hv/m 0c2are shown in Fig. 15-6 for air
and lead; values of aK/ q, and aK/Z 2for Al and Pb are Iisted at severa! photon
energies in Table 15-4. At photon energies above about 10m 0c2 or about
5 Mev, it is necessary to take into account an effect called screening.
At these energies the electron-positron pair is probably formed sorne dis-
tance from the nucleus. When this distance lies outside sorne of the
electron shells, the field in which the pairs are created is smaller than that of
the nucleus alone and the probability oí pair formation is reduced some-
what. Consequently; at higher energies (up to 25 Mev) the values for
lead are somewhat smaller than those for air or aluminum; at still greater
energies, the screening effect becomes more marked.
It follows from the way in which the cross section for pair formation
depends on Z and on the energy that this process is important for high
energies and for heavy elements. The contribution of pair formation to
the total absorption coefficient µ for Iead is equal to that of the Compton
effect at about 4.75 Mev. Above this energy, pair formation predominates.
The process of pair production is closely related to the reverse process,
that of electron-positron annihilation. A positron, after being formed, is
slowed down by collisions with atoros until it is practically at rest. It then
interacts with an electron which is also practically at rest. The two partí-
eles disappear, and two photons appear moving in opposite directions, each
with an energy of 0.511 Mev, equal to the rest energy of an electron. Two
photons, rather than one, are needed to allow momentum to be conserved
if the annihilation occurs away from nuclei. The photons which appear on
the annihilation of an electron-positron pairare called annihilation radia-
tion, and the absorption of 'Y-rays by the pair-production process is always
complicated by the appearance of this low-energy secondary radiation.
15-6 The absorption of gamma-rays by matter. Comparison of ex-

perimental and theoretical results. Theoretical values of the absorption
coefficient 0 µ or µ can now be obtained from the values of the partial
coefficients given in Tables 15-2, 15-3, and 15-4 and from graphs of those
data. The following procedure is used:
l. Values of 0 r/Z 5 n for different values of n are found from Table 15-2
or from plots of the data in the table. Values of the photoelectric cross
section per atom ,,r are obtained on multiplying by Z 5n.
2. Values of the Compton cross section per atom aU are obtained from
the data of Table 15-3 on multiplying the values of ,u by Z.
3. Values of aK/i[, are obtained from Table 15-4 or Fig. 15--0 for the same
values of n(= 1/a) used in the first two steps, and multiplied by
i[, = 5.796 X 10- 2 sz 2 to give the pair production cross section per atom 0 K.
The total absorption coefficient per atom aµ is then
(15-28)
The mass absorption cóefficient µ/ p is
(15-29)
The absorption coefficient µ is
(15-30)
15-6] THE ABSORPTION OF GAMMA-RAYS BY MATTER 415
Tables of values of aT, a<T, aK, and ªµ have been prepared by Davisson
and Evansc 2 > for 24 elements rangíng from hydrogen (Z = I) to bismuth
(Z = 83). Their values for lead are listed in Table 15-5 together with
the values of µ/p and µ. The values of the cross section per atom are in
units of 10-24 cm 2 /atom. Mass absorptíon coefficíents µ/p (cm 2 /gm) are
listed in Table 15-6 for four commonly used absorbers (water, aluminum,
iron, and lead).
The total and partial absorptíon cross sections per centímeter for lead
are plotted in Fíg. 15-7, and the general features of the absorptíon of
'Y-rays can be seen by considering these curves. At very low energies, ab-
sorptíon by photoelectrons predomínates, but decreaf¡es rapidly wíth in-
creasing energy. As ít decreases, attenuation by the Compton effect
becomes relatively more ímportant, until the two effects are equal at about
0.5 Mev. At energies slightly less than 1 Mev most of the attenuation is
TABLE 15-5
VALUES OF THE ABSORPTION COEFFICIENTS FOR LEAD*
Mass
Photon Photo- Pair Coefficient
Compton Total coefficient
energy, electric formation per cm,
aO' .µ µ, cm- 1 µ/p,
Mev aT aK
cm 2/gm
0.1022 1782 40.18 1822 59.9 5.30

0.1277 985 38.01 1023 33.6 2.97
0.1703 465 35.04 500 16.4 1.45
0.2554 161 30.70 192 6.31 0.558
0.3405 75.7 27.63 103.3 3.39 0.300
0.4086 47.8 25.74 73.5 2.42 0.214
0.5108 27.7 23.50 51.2 1.68 0.149
0.6811 14.5 20.73 35.2 1.16 0.102
1.022 6.31 17.14 23.45 0.771 0.0682
1.362 3.86 14.81 0.1948 18.87 0.620 0.0549
1.533 13.91 0.3313
2.043 2.08 11.86 1.247 15.19 0.499 0.0442
2.633
3.065 9.313 3.507
4.086 0.869 7.761 5.651 14.28 0.469 0.0415
5.108 0.675 6.698 7.560 14.93 0.491 0.0434
6.130 5.917 9.119
10.22 0.316 4.115 14.04 18.47 0.607 0.0537
15.32 0.206 3.042 18.00 21.25 0.698 0.0618
25.54 0.122 2.044 23.24 25.41 0.835 0.0739
*In units of 10- 24 cm 2 /atom.

Photon energy (hv/rnoé')
Fw. 15-7. Total absorption coefficient of lead, showing the contributions of

photoelectric absorption, Compton scattering, and pair formation.
TABLE 15-6
MAss ABsoRPTlON CoEFFICIENTS*
Photon
energy, Water Aluminum Iron Lead
Mev
0.1 0.167 0.160 0.342 5.29

0.15 0.149 0.133 0.182 1.84
0.2 0.136 0.120 0.138 0.895
0.3 0.118 0.103 1U06 0.335
0.4 0.106 0.0922 0.0918 0.208
0.5 0.0967 0.0840 0.0828 0.145
0.6 0.0894 0.0777 0.0761 0.114
0.8 0.0786 0.0682 0.0668 0.0837
1.0 0.0706 0.0614 0.0595 0.0683
1.5 0.0576 0.0500 0.0484 0.0514
2.0 0.0493 0.0431 0.0422 0.0451
3.0 0.0396 0.0353 0.0359 0.0410
4.0 0.0339 0.0310 0.0330 0.0416
5.0 0.0302 0.0284 0.0314 0.0430
6.0 0.0277 0.0266 0.0305 0.0455
8.0 0.0242 0.0243 0.0298 0.0471
10.0 0.0221 0.0232 0.0300 0.0503
*µ/p, cm 2/gm.
15-6] THE ABSORPl'ION OF GAMMA-RAYS BY MATTER 417
1.6
1.4
7
e
~
1.2
]
·a
IS
g 0.8
.) 0.6
...c.
~ 0.4
.i:,
< 0.2
1.0 10 100 1000

Photon energy (h,fmr,c2)
F10. 15-8. The total absorption coefficients of aluminum, copper, tin, and
lead.
caused by the Compton effect. Absorption by pair formation starts at

about 1 Mev and increases while the other effects decrease, until at high
energies the absorption is almost completely by pair production. The curve
of total absorption coefficient vs. energy has a minimum at energies where
the Compton effect and pair production become comparable. While these
features are characteristic of the absorption curves for all elements, the
particular energy at which one process or another is important varíes from
element to element. The following table indicates the energy ranges in
which the various effects make their greatest contributions in a light
element (aluminum) andina heavy element (lead).
Photoelectric effect Compton effect Pair formation

Aluminum <0.05Mev 0.05-15 Mev >15 Mev
Lead <0.5 Mev 0.5 -5 Mev >5 Mev
The variation of the absorption coefficient vs. energy curve from element to
element is shown by the curves for aluminum, copper, tin, and lead in
Fig. 15-8.
The simplest method of testing the theory of the absorption of 'Y-rays is
to measure the total absorption coefficient by the method discussed in
Section 15-1. Although measurements of this kind are basically simple,
the experimental resulta depend on the knowledge of the energy of the
photons emitted by the source, on the purity of the absorber, and on many
other experimental details. The production of artificial radionuclides has
provided sources of monoenergetic 'Y-rays with energies in the range from
0.5 to 2.8 Mev, and monoenergetic 'Y-rays beams of higher energy have been
obtained as products of nuclear reactions. It has been found that in the
energy range from 0.1 Mev to 6 Mev the theoretical values of the absorp-
tion coefficient as calculated from the present theories of the photoelectric
effect, Compton effect, and pair production, are in good agreement with
the experimentally measured values. The agreement has been found for a
large number of elements in the periodic table from carbon to lead. The
reader is again referred to the review article by Davisson and Evans for a
detailed discussion, including many experimental curves, of the experi-
mental results. A partial comparison of theory and experiment is given
in Table 15-1.
The partial absorption coefficients have also been measured experi-
mentally, although measurements of this kind are more difficult, and hence
less extensive, than those of the total absorption coefficient. The theory
of the individual effects has been found to give cross sections or partial
absorption coefficients in good agreement with the experimental results.
Thus, the theory of photoelectric absorption, the Klein-Nishina formulas
for the Compton effect, and the theory of pair formation have all been
shown to predict the correct results. The theory of these processes and
the experimental verification are treated in detail in Heitler's book,
The Quantum Theory of Radiatian. ca>
15-7 The measurement of gamma-ray energies. Severa! general

methods can be used to measure the energies of 'Y-rays. Since 'Y-rays are
electromagnetic radiations, the most direct method is to determine the
wavelength and hence the energy by using a crystal as a diffraction grating
(Section 4-3). A high-precision instrument called a curved-crystalfocusing-
type spectrometer has been used by DuMond, c4 , 5> and the principie on
which it is designed is shown schematically in Fig. 15-9. The quartz
diffracting crystal C is bent and clamped so that the diffracting planes
meet, when extended, in a line at {J, normal to the plane of the figure.
The radius of curvature of the crystal is then equal to the diameter of the
focusing circle F shown (2.0 m). If the source of 'Y-rays is at the real
focus R, and if the Bragg condition, Eq. (4-10), is satisfied, there is a
diffracted beam which enters the detector D as if it carne from a virtual
source at V; the Bragg angle is denoted by 8. For each different wave-
length there is a particular position of the source somewhere on the focusing
circle which yields a diffracted beam. The co,mting rate of the detector is
determined as a function of the source position as the source is moved
around the focusing circle; the rate has a sharp maximum at the point for
which a strong diffracted beam occurs. The position of this point deter-
mines the wavelength and hence the frequency and energy of the 'Y-rays.
This method can give results with very high precision, e.g., the energy of
15-7) THE MEASURME:ST OF GAMMA-RAY ENERGIEB 419
/3
1 \
1 \
\ \
1
1 F
1
1
1
FIG. 15--9. Schematic diagram of a crystal 'Y-ray spectrometer (Du-

Mond<4·51).
the 'Y-ray associated with the ,B-decay of Au 198 was determined as

411.770 ± 0.036 kev, and results with similar precision have been obtained
for the energies of many other nuclear 'Y-rays. c5 > Crystal diffraction meth-
ods, however, have two disadvantages. First, the measurements become
more diffi.cult and less precise as tl:.e energy of the 'Y-rays increases and the
wavelength decreases, with the result that they cannot be carried much
above 1 Mev. Second, they require highly active sources, and these are
often not available.
DuMond's 'Y-ray spectrometer was also used to measure the energy of
the annihilation radiation. es> The photons from the annihilation of
positrons from Cu 64 were found to have an energy of 510.941 ± 0.067 kev.
The "best value" for the electron rest energy at the time of the experiment
was 510.969 ± 0.015 kev, so that the measured value of the energy of the
annihilation radiation agreed very well with the value predicted by the
theory of pair formation and annihilation. The two numbers quoted can
be used to calculate the possible difference in mass between the positron
and the electron, and it was found that to a part in 104 there is no evidence
for any difference in masa between positive and negative electrons. This
result has special interest because, together with a recent direct comparison
of the ratio m/e for the electron and positron, it provides independent
experimental proof that the electron and positron are identical except for
the sign of the electric charge. In the comparison experiment, t 5> elec-
trons from a hot filament and positrons from Na 22 were analyzed in a mass
spectrometer with a reversible electric field and a fixed magnetic field.
The particles followed ídentical trajectories in opposite directions in the
spectrometer. The magnitude of the electric field needed for focusing
was taken as a measure of m/e for the particle involved. The result ob-
tained was
(m/e)- - (m/e)+ = (26 ± 71) X 10-B
m/e '
where the ratios with the minus and plus signs refer to the values obtained
by focusing the particles in the experiment, and the denominator is the
accepted value of the ratio for the electron. Thus, both the mass and the
specific charge of the positron and electron have been shown to be equal to
within very small amounts. Although the comparison experiment is not
directly related to 'Y-rays, it has been included here because of the way in
which it supplements the annihilation radiation experiment.
The curved crystal spectrometer has been improved and simplified, <30 • 3 0
and applied to problems such as the study of the 'Y-rays from neutron cap-
ture. Asan example of the precision that can be obtained, a recent measure-
ment <32> of the deuteron binding energy gave the result 2.2255 ±
0.0015 Mev.
Gamma-rays of modera.te energy are most often studied by mes.ns of the
photoelectrons and Compton recoil electrons which they eject from a
suitable material called a radiaWr. The energies of the ejected electrons
are measured with a magnetic spectrometer, as described in Section 14-1.
_Radi&tor
m-Abeorber ~
~ Source of {J-
and 'l'-rays
Counter
FIG. 15-10. Diagram of an arrangement for measuring the energies of 'Y-rays

from a sourcc of /3- and 'Y-raya.
15--7] THE MEASUREMENT OF GAMMA-RAY ENERGIES 421
When the Compton effect is used, the source is enclosed in· an absorber of
material of low atomic number such as aluminum, thick enough to stop ali
the primary electrons or other charged particles but not the 'Y-rays. The
Compton electrons ejected from the radiator, which is usually a thin foil,
are focused in the spectrometer and form a continuous spectrum with a
fairly sharply defined upper limit. The upper limit corresponds to elec-
trons ejected in the forward direction from the surface of the absorber, and
is related to the 'Y-ray energy by Eq. (15-18). If T m is the maximum
electron energy, and hv is the desired 'Y-ray energy [this amounts simply
to dropping the subscript zero in Eq. (15-18)], then
hv = ½ITm + (T!, + 2Tmmoc 2) 112]. (15-31)
A schematic diagram of an arrangement for measuring the energies of
'Y-rays from a source which emits both ,8-particles and 'Y-rays is shown in
Fig. 15-10.
If the radiator is a thin foil of material of intermediate or high atomic
number, the photoelectrons ejected from it forro line spectra. Llnes appear
corresponding to electrons from the K-shell and the L-shell, and for
spectrometers of very high resolution even M electrons may be resolved.
The energy of the electrons in a given line can be determined from the
position of the line (i.e., from the Hr value) and the 'Y-ray energy is ob-
tained by adding the binding energy of the electron in the shell from which
it was ejected to the measured electron energy.
Sorne of the ,8-particle spectrometers discussed in Section 14-1 have been
used to determine 'Y-ray energies by measuring the energies of Compton
Mn52
-r-rays
Pb-radiator
l
N l
o 2 3 4 5 6 X 103 o 2 3 4 5 6 X 103
Hr_ Hr_
(a) (b)
F10. 15--11. (a) Photoelectrons and Compton recoil electrons ejected from a
lead radiator by 'Y-rays from Mn112 ; the peaks are caused by photoelectrons. (b)
Compton recoil electrons ejected from a copper radiator by 'Y-rays from Mn112
(Peacock and Deutsch<8>).
electrons and photoelectrons. Deutsch, Elliott, and Evans< 7 >have treated

the theory, design, and use of a short magnetic lens spectrometer for both
fJ- and 'Y-rays. The spectra of fJ-particles from two radiators, produced by
'Y-rays from Mn 52 , and measured with that instrument, <s> are shown in
Fig. 15-11. Figure 15-ll(a) shows the photoelectrons and Compton
recoil electrons ejected from alead radiator by 'Y-rays from Mn 52 • There
are four pronounced peaks caused by 'Y-rays of energies 0.510 ± 0.01 Mev,
0.734 ± 0.015 Mev, 0.94 ± 0.02 Mev, and 1.46 ± 0.03 Mev. Figure
15-ll(b) shows the spectrum of Compton recoil electrons produced in a
copper radiator indicating the same four 'Y-rays. The lowest energy
'Y-ray is doubtless annihilation radiation, since Mn 52 is a positron emitter.
At higher energies, above 2 or 3 Mev, 'Y-rays are best studied with a pair
spectrometer, because the probability of pair formation increases with
increasing photon energy, while the Compton and photoelectric cross
sections decrease. A collimated beam of 'Y-rays falls on a radiator and
-- -----+--
Pb
5 Ele_ctron
d
Fm. 15-12. Schematic diagram of a pair spectrometer (Walker and Mc-

Daniei<ll)).
15-7] THE MEASUREMENT OF GAMMA-RAY ENERGIES 423
ejects electron-positron pairs from it. The electrons and positrons are
focused separately in a uniform magnetic field into a number of counters.
It is necessary to detect simultaneously the electron and positron produced
by one 'Y-ray, and this can be done by using coincidence techniques. <9 • im
In the coincidence method only the simultaneous detection of two
events in different counters is recorded. It is possible, with suitable
electronic circuits, to achieve a "resolving time" of a fraction of a micro-
second, that is, less than 10- 5 sec, with Geiger counters. With ionization
chambers and electron multipliers, smaller resolving times (~10- 8 sec)
can be achieved. Two events occurring within a time interval less than the
resolving time and detected by separate counters are said to be in ooin-
cidence. The energies of the paired electron and positron are determined
from their Hr values, and the sum of the energies together with the rest
energies (1.02 Mev) gives the energy of the 'Y-ray.
A schematic diagram of a pair spectrometer used by Walker and
McDanie1° 1 > is shown in Fig. 15-12. A uniform magnetic field is applied
perpendicular to the plane of the paper. A parallel beam of 'Y-rays falls on
a thin radiator foil placed perpendicular to the beam. Electron-positron
pairs are produced which, for 'Y-rays of fairly high energy (>3 Mev), go
very nearly in the forward direction. The electron and the positron may
each fall into one of four counters placed in the plane of the radiator foil.
The counters are connected into coincidence circuits in pairs, in such a way
that paif!S produced by 'Y-rays of the same energy are counted and recorded.
Energies of 'Y-rays can also be obtained from measurements of the
energies of interna! conversion electrons. The process of interna! conver-
sion has been discussed very briefl.y in Section 14-4 ¡ in this process the
secondary electrons come, not from a radiator, but from the radioactive
~tom itself. The electron energies are determined by magnetic analysis
and appear as the secondary or line spectrum of ¡3-emitters.
Absorption methods can also be used to determine the energies of 'Y-rays.
In principle, a measurement of the attenuation of a 'Y-ray beam by a known
absorber in a "good geometry" experiment would give a value of the absorp-
tion coefficient. For a given absorber, the absorption coefficient is known
as a function of energy either from theory, or as a result of measure-
ments with 'Y-rays of known energy, and the unknown energy can then be
found from the measured value of the absorption coefficient. A minor
complication is introduced by the fact that a given absorption coefficient
may correspond to two different photon energies because of the mínimum
in the absorption coefficient-energy curve, as shown in Figs. 15-7 and 15-8.
The correct ~nergy can be found by obtaining absorption curves in two
different materials.
Unfortunately, the "good geometry" method requires highly active
sources because only a small fraction of the 'Y-rays emitted by the source
424 GAIOIA-RAYS AND GAIOIA-DECAY [CBAP. 15
lOOOr----------------,
! 100
1.0 10 100 1000 10,000

Half-thickness in Al (mg/cm2)
Fm. 15-13. Half-thickness values in aluminum for low-energy 'Y-raya [L. E.

Glendenin, Nucleonia, 2, No. 1 (1948)).
is used in the experiment. When weak sources only are available,

absorption experiments with "poor geometry" may be used. An arrange-
ment similar to that of Fig. 14-5 is feasible, but with the absorbers placed
against the counter and with the source placed as far away from the counter
as the intensity of the source allows. The tJ-particle counter of the figure
is, of course, replaced by a 'Y-counter. Under these conditions and if the
'Y-radiation is homogeneous, an exponential curve (straight line on semi-
logaríthmic graph paper) is obtained overa range of at least 3 or 4 half-
thicknesses. The energy may be obtained from a half thickness curve of
known energy such as that for aluminum shown in Fig. 15-13, and this
method yields fairly accurate energy values below 1 Mev.
It is also poesible to determine 'Y-ray energies by measuring the maximum
range of Compton electrons ejected from a radiator by the 'Y-rays. The
range measurements may be made in a cloud chamber or with coincidence
methods. In the latter method, the Compton electrons pass through two
thin-walled Geiger counters connected to a coincidence circuit, and ab-
sorbers are placed between the two counters. The coincidence rate is
plotted against the absorber thickness; it decreases rapidly with increasing
absorber thickness and becomes zero when the endpoint of the Compton
electrons is reached. This method can be used for energies up to about
5 Mev. I t has been discussed in detall and extended to higher energies
by Fowler, Lauritsen, and Lauritsen. u:z>
The last method to be mentioned is that of the scintillation spectrom-
eter, <1 3 , 14> the basis of which was discussed in Section 2-8. In this
method, a scintillation counter acts as a proportional counter for measur-
ing the energies of the electrons produced by the 'Y-radiation, and hence
the energies of the 'Y-rays. Thallium-activated sodium iodide is most
often used as the phosphor because a large sodium iodide crystal has a
15-8] GAMMA-DECAY: INTERNAL CONVERSION 425
high efficiency for the absorption of 'Y-rays. The high atomic number of
iodine permits the measurement of "lines" due mainly to photoelectrons
and electron:positron pairs. The secondary electrons produce pulses which
can be displayed on an oscilloscope screen and photographed. The amount
of light resulting from the capture of a 'Y-ray, as well as the pulse height,
is a nearly linear function of the energy of the 'Y-ray, and the counting rate
can be determined as a function of pulse height or energy. The result is a
spectrum whose maxima give the energies of the 'Y-rays.
15-8 Gamma-decay: intemal conversion. lt was supposed in Sec-

tions 13-5 and 14-7 that a nucleus which has been left in an excited state
following a-decay or {3-decay can pass to a less excited state either by
emitting a 'Y-ray or by interna! conversion. It was shown that with this
supposition it is possible to account for the energies of the 'Y-rays asso-
ciated with a-decay and {3-decay, and to write down consistent decay
schemes.
It was not always clear that the emission of the 'Y-rays follows that of
the a- or /3-particles, and considerable work was done in the 1920's to
prove this. Several different, independent proofs were developed, two of
which will be described briefly. One method, used by Rutherford and
Wooster,< 15> depended on the measurement of the secondary x-ray lines
which accompany the decay of RaB. Radium B, with a half-life of 26.8
minutes, emits electrons and 'Y-rays, and the /3-particle spectrum shows a
number of lines corresponding to interna! conversion in various electron
shells. After the internal conversion electrons are emitted, the vacancies
in the electron shells are filled (see Section 7-7) with the emission of char-
acteristic x-rays. Rutherford and Wooster determined the wavelengths
of sorne of the L x-ray lines which accompany the {3-decay of RaB and
found that the wavelengths agreed excellently with the accepted values
for atomic number 83. Since the decay of RaB causes a change of atomic
number from 82 to 83, the result showed that internal conversion takes
place after the change in atomic number.
The emission of an interna! conversion electron after a radioactive decay
implies that a reorganization of the product nucleus takes place after the
decay process has occurred. It was proved unambiguously that the
amount of the energy released in interna! conversion is the same as that
released in 'Y-ray emission, so that both processes represent the same
reorganization of the product nucleus, and the emission of the 'Y-ray must
therefore follow that of the a- or {3-particle. Table 15-7 gives the energies
of a group of {3-ray lines of RaB, as measured by Ellis. <15> The electron
shell in which the conversion took place was identified from the charac-
teristic x-radiation, and the binding energies of the electrons in the different
shells were known. The sum of the kinetic energy of the electron and its
TABLE 15-7
CONVERSION OF A GAMMA-R.AY IN VARIOUS ELECTRON SHELLS
¡3-Ray energy, Binding energy Conversion 'Y-Ray energy,

kev forZ = 83 level kev
36.74 16.34 L1 53.08
37.37 15.67 Lu 53.04
39.63 13.38 L111 53.01
48.85 3.99 M1 52.84
49.10 3.68 Mu 52.78
49.66 3.17 Mm 52.83
51.90 0.93 N1 52.83
52.64 0.20 o 52.84
- -
Average 52.91
binding energy was taken to be the energy of the converted 'Y-ray, and in
the example shown, the same value, 52.91 kev, for the 'Y-ray energy was
inferred from all of the 'Y-ray lines, within the experimental error.
The energies of the 'Y-rays emitted by a given nuclide can also be meas-
ured directly with a crystal spectrometer, as described in Section 15-7.
Independent measurements of this type gave the value 53.3 kev for the
energy of a 'Y-ray associated with the decay of Ra.B, in good agreement with
the value of 52.9 kev inferred from the {J-ray lines. Similar agreement was
obtained for many 'Y-rays between direct measurements of their energies
and inferences from the fJ-ray lines, proving that the emission of a 'Y-ray and
of an interna! conversion electron represent the same nuclear reorganiza-
tion. Since the emission of the interna! conversion electron follows the
radioactive disintegration, the emission of the 'Y-ray must also follow the
particle decay process.
Another ingenious experiment on the time of einission of the conversion
electrons was that of Ellis and Wooster.' 17> They placed a source of
Ra.B inside a thick platinum tube, on the outside of which was placed a
thin coating of Ra.B. The 'Y-rays from the inside source ejected photo-
electrons from the platinum, with energies hv - KPt, hv - Lpt, ... ,
where KPt is the binding energy of the electron in the K-shell of platinum,
and so on. At the same time the 'Y-rays from the source on the outside of
the tube were accompanied by internal conversion electrons with energies
hv - KRaB, hv - LRaB . . . . The electrons were studied by magnetic
deflection, and photographs were taken simultaneously, and on the same
photographic plate, of the photoelectric spectrum and of the interna! con-
version spectrum. The difference in energy of corresponding groups of
electrons is KRaB - KPt, LRaB - Lpt ... , where KRaB is the binding
15-8] GAMMA-DECAY: INTERNAL CONVERSION 427
energy of an electron in the K-shell of RaB, etc. The energy difference for
the K-lines was calculated theoretically as Ks2 - K1s or Ksa - K1s,
depending on whether the internal conversion eleetrons and 'Y-rays were
assumed to come from the parent nucleus RaB ( 82 Pb 214 ) or from the
product nucleus RaC (83Bi 214 ), and compared with experimental results.
The binding energiP,s of the electrons in the K- and L-shells of platinum
(Z = 78) and of the elements with Z = 82 and 83 were well known, and
the experimental values of the energy differences could be obtained from
the spacing between lines (Hr values) as seen on the photographic plate.
This procedure was used for several different lines and in each case the
experimental result agreed with the value calculated for Z = 83 and not
with the value calculated for Z = 82. These results showed again that the
internal conversion electron and, therefore, the 'Y-ray, is emitted after the
~-particle. Other experiments were done with both a-emitters and ~-
emitters, and in each case the 'Y-rays were shown to be emitted after the
particie disintegration.
The proof that the 'Y-rays are emitted by the product nucleus and not
the parent nucleus must now be followed by a brief discussion of a generally
accepted confusion in terminology. The 'Y-rays which have been considered
up to now are all emitted immediately after the a- or /3-particle, and it has
nót been possible to measure the time interval between the two decay
processes, because it is too short. According to theory, the 'Y-ray lifetime
for these nuclei is of the order of magnitude of 10- 13 to 10- 11 sec. Con-
sequently, the 'Y-ray intensity decays with the same half-life as that of the
parent nucleus. In the case of RaB, with a half-life of 26.8 min, the
'Y-ray intensity also deeays with a half-life of 26.8 min, and the practice
arose and has been continued of calling these 'Y-rays the 'Y-rays of RaB
rather than those of RaC. In another case, ThC goes by a-decay to
ThC" and has a spectrum with fine structure, emitting six groups of
a-particles. The product nucleus ThC" emits at least eight different
'Y-rays which can be fitted into a consistent decay scheme as in Fig. 13-11.
Nevertheless, these 'Y-rays are usually spoken of as the 'Y-rays of ThC and
are so listed in tables and charts of nuclides.
The relative frequency with which 'Y-emission and internal conversion
occur in a given nuclide is a matter of great interest from the standpoint of
theory and is expressed by a quantity called the internal conversion coeffi-
cient. This coeflicient is defined as the ratio of the number of conversion
electrons to the number of 'Y-ray photons emitted and there are different
coeflicients for the dif1erent electron shells. Values of the internal con-
version coeflicients can be calculated from quantum mechanics, but
rigorous calculations are very laborious, requiring the use of large elec-
tronic computing machines. The cooflicient should increase rapidly with
the atomic number, approximately as Z 3 over part of the range, and
decreases rapidly as the 'Y-ray energy increases. The conversion coefficient

is a difficult quantity to determine experimentally with any accuracy, but
sorne of the detailed predictions of the theory have been verified by means
of ingenious experiments.<1 S> In general, interna! conversion is found
most often in heavy elements when low-energy ( <0.5 Mev) -Y-rays are
emitted. In the lighter elements and with higher energy -Y-rays, interna!
conversion is relatively infrequent.
15-9 Gamma-decay and nuclear energy levels: theory. The study of

the 'Y-rays and interna! conversion electrons emitted in transitions from
excited levels of nuclei to Iess excited or stable Jevels is one of the main
sources of our knowledge of low energy nuclear levels. These levels are
characterized by their stationary properties, energy, angular momentum
(spin), and parity. The energies of the -Y-rays or conversion electrons
determine the energies of the transitions which give, in tum, the energy
differences between levels. Sorne information about the angular momenta
and parities of the levels can be obtained by combining experimental
results with theoretical relationships between the decay probability and
the energy of the 'Y-rays. More precise information can be obtained with
the aid of certain specialized methods. One such method is the quantita-
tive study of interna! conversion coefficients; another method depends on
the angular correlation between successive 'Y-rays emitted by an excited
nucleus; a third depends on the angular distribution of the -Y-rays emitted
by nuclei aligned in a magnetic field (see Section 14-9).
A complete theory of 'Y-decay would require the use of the quantum
theory of radiation together with a detailed description of the quantum
states of the nucleus. A knowledge of nuclear forces would be needed to
formulate the problem, but these forces are not yet understood. Con-
sequently, the theory must be based on a rough model of the nucleus and
on a method of treating the interaction between the nucleus and radiation.
In the present state of our knowledge of nuclear structure, therefore, the
properties of nuclear levels and the transitions between levels cannot be
predicted accurately and in detail. Nevertheless, the combination oí
experimental results with the available theory makes possible the deter-
mination of angular momenta and parities and the construction of energy
level and decay schemes. We shall, therefore, discuss the theory briefly
with the object of arriving at equations relating the disintegration constant
for 'Y-decay and the energy of the -Y-rays. Equations (15-40) and (15-41)
below, will be seen to be analogous to Eqs. (13-20) and (13-22) for a-decay
and to Eqs. (14-13) and (14-14) for ,8-decay. Even an elementary dis-
cussion of 'Y-decay is complicated, however, by the need to show explicitly
the effects of angular momentum and parity, effects which did not appear
in the relatively simple cases of a-decay and ,8-decay considered earlier.
15-9) GAMMA-DECAY AND NUCLEAR ENERGY LEVELS 429
The derivation of Eqs. (15-40) and (15-41) is beyond the acope of this
book, but the discussion which follows should lead to sorne familiarity
with the nomenclature of 'Y-decay and may serve to introduce the reader
to some of the ideas underlying the theory.
lt has proved to be very useful to express the interaction (force or poten-
tial) between the nucleus and electromagnetic radiation in terma of an
infinite series of powers of the ratio R/'A, where R is a linear dimension
which characterizes the nucleus and is taken as the radius of the nucleus,
while 'A is a quantity which characterizes the radiation and is the wave-
length divided by 21r. The nuclear radius can be written in the forro
(cf. Section 13-6)
R = R 0 A 113 X 10- 13 cm, (15-32)
where Ro is a constant with a value of about 1.5, and A is the mass number
of the nucleus. The wavelength of the 'Y-radiation is given by
,,, ,,, líe
'A--=-=-·
p hv/c E
When E is in Mev and 'A in cm,
197 -13
(15-33)
'A = E (Mev) X 10 cm,
and
R R 0 A 113E (Mev)
(15-34)
197
Now, A 113 has values up to about 6, and E is generally about 1 or 2 Mev,

so that the ratio R/'A is nearly always less than 0.1. The expression for
the decay probability (disintegration constant) is given by theory as an
infinite series in powers of (R/'A) 2 . The successive terms in the series
decrease very rapidly and the decay probability is given almost completely
by the first nonvanishing term in the series, ~s in the case of t,-decay
(Section 14-6). This term gives rise toan allowed transition; the succeed-
ing terms, which are much smaller, give rise to forbidden transitions.
Although the ratio (R/'A) 2 is small for most of the interesting nuclear
transitions, the forbidden transitions are still important in 'Y-decay. One
reason is that the quantity observed in experimenta is the transition rate
and, at nuclear energies, even very low transition rates can be measured.
This situation is in strong contrast with that of atomic transitions where
light is emitted or absorbed. For atoros, the value of R/'A is of the order
10-8¡10-5 = 10-3 , which is much smaller than the values for nuclei, so
that the forbidden atomic transitions are relatively more forbidden than
the nuclear ones. In addition, the experimental material in atomic transi-
tions consists of spectral lines, and the intensities of the forbidden lines
are so small as to make these lines very difficult to observe. Since the
forbidden transitions are important in 'Y-decay, each term in the expan-
sion of the decay probability must be examined to see what it means
physically and under what conditions it becomes important.
The expansion procedure, which is adapted from classical electro-
magnetics, separates the 'radiation from a nucleus into distinct types called
multipol,e radiations. This separation corresponds to sorting the emitted
'Y-rays {photons) according to the amount of angular momentum Lh
carried off by each photon. In other words, for a given 'Y-ray energy,
photons can be emitted which differ in their angular momentum. The
decay probability is the sum of the partial probabilities for the emission
of these different types of 'Y-rays; but in this sum one tenn, the first non-
vanishing term, predominates. The partial probabilities will be seen to
depend very strongly on the angular momentum carried off by the photons,
and this is one reason for the usefulness of the procedure. Photons can
have only integral values of L; the value L = Ois ruled out as a consequence
of the fact that electromagnetic waves are transverse in nature. The
multipole radiation is characterized by its order, given by 2L. For L = 1,
the radiation has been shown by theory to correspond to that emitted by a
vibrating dipole; for L = 2, it corresponds to the radiation from a vibrat-
ing quadrupole, for L = 3 to that from an octupole, and so on. For each
value of L there are two classes of radiation, called electric and magnetic
multipole radiations, which differ in their parity. An electric multipole
has even parity when L is even, and odd parity when L is odd. Magnetic
multipole radiation has odd parity when L is even, and even parity when
Lis odd; parity of electric multipole = (-l)L, parity of magnetic multi-
+
pole = -(- I)L, where 1 means even parity and -1 means odd parity.
A given 'Y-decay process may leave the parity of the wave function of the
nucleus unchanged, or there may be a change in the parity; in either case,
the parity of the wave function of the entire system, nucleus and radiation,
is conserved.
As in atomic transitions, there are selection rules which must be obeyed;
these are shown in Table 15-8 together with the symbols used for the dif-
ferent types of radiation. The selection rules for emission of 'Y-rays (or
for their absorption) define those combinations of L and parity which give
nonvanishing values of the transition probability.
In the case of electric multipole transitions of order 2L it has been
shown ° 9> that the decay probability can be written in the following fonn
when the independent particle or ahell model of the nucleus (see Chapter
17) is used:
">.,EL = e2 (R)2L ;
2,rv ftc S ~ {15-35)
15-9] GAMMA-DECAY AND NUCLEAR ENERGY LEVELS 431
TABLE 15-8
SELECTION RULES AND 8YMBOLS FOR GAMMA R.ADIATION
Change in Parity change

angular momentum Symbol of nuclear wave
Classification
of nucleus, function
in units of ft
Electric 2L-pole L EL No for Leven

Yes for L odd
Magnetic 2L-pole L ML Yes for L even
No for L odd
Electric dipole 1 El Yes
Magnetic dipole 1 Ml No
Electric quadrupole 2 E2 No
Magnetic quadrupole 2 M2 Yes
Electric octupole 3 E3 Yes
Magnetic octupole 3 M3 No
Electric 16-pole 4 E4 No
Magnetic 16-pole 4 M4 Yes
here >-EL is the disintegration constant (or reciproca! mean life) for 'Y-decay
with the emission of EL radiation, and will not be confused with the
wavelength X; 11 is the frequency of the emitted radiation. The quantity
S depends on L and is given by
2(L + 1) ( 3 )
2
(15-36)
S = L[I X 3 X 5 ... (2L + 1)] 2 L +3 ;
it has the following values.
L s
1 2.5 X 10-1
2 4.8 X 10-3
3 6.25 X 10-5
4 5.3 X 10-7
5 3.1 X 10-9
The probability of emission of magnetic multipole radiation is given by
e
2 h )2 (R)2L
(McR (15-37)
XML = 21l"JI líe (10S) ~ ,
where M is the mass of a nucleon. For a given value of L, electric multi-

432 GAIUlA•RA YS AND GAMM:A·DECAY (CHAP. 15
pole radia.tion is more probable than magnetic multipole radia.tion, with

2
A1n _ _.!_ ( ~ ) _ 44 A21a. (15-38)
>.M L - 10 A/Me - · '
we have used A/Me= 0.211 X 10- 13 cm and R = 1.4 A 113 X 10- 13 cm,
which is considered to be suitable for 'Y-decay calculations. The dis-
integration constants can be expressed in terms of the 'Y-ray energy and
the atomic :mass of the nucleus if we use Eqs. (15-32) and (15-33) and
note that e2 /Ac = 1/137; for example,
2
>.
BL
= 21rV
137
S(R A 11a)2L (E (Mev)) L.
O 197 (15-39)
The frequency P is proportional to the energy so that, on multiplying

the numerator and denominator of the right side of the last equation by
h, and inserting numerical values, we get
).BL 24S(R A
• o
11a)2L (E (Mev))
197
2
L+i 1021 sec-1.
' (15-40)
where R0 A 113 is just the radius of the nucleus in units of 10- 13 cm.
· Equation (15-37) becomes
AML = 0.55SA-21a(RoA 11a)2L (E i~7ev)yL+t 1021 sec-1. {15-41)
Equations (15-40) and (15-41) are convenient forms for calculation and
for comparison with experimental results.
It is evident from Eqs. (15-35) and (15-37) or from Eqs. (15-40) and
(15-41) together with the values of S, that the probability of 'Y-decay
should decrea.se very rapidly with increasing values of L. Since the energy
of the 'Y-rays is at most a few Mev, Eqs. (15-40) and {15-41) show that,
for a given value of L, the probability of 'Y-decay should increase with
increasing values of the energy of the 'Y-rays. The theoretical resulta
are illustrated by the numerical examples given in Table 15-9; values of
the mean life (the reciproca! of the disintegration constant) for 'Y-decay
with emission of different types of multipole radiation are given for a
nucleus with a radius of 6 X 10- 13 cm (A ""' 80). The resulta show that
when the angular momentum carried off by the 'Y-ray is small (L = 1 or
2), that is, when the change in the angular momentum of the nucleus is
small, the decay probability is large and the mean life is very short. But
when the change in the angular momentum of the nucleus is large (3 or
more units), the decay probability may become very small and the mean
life may become long.
TABLE 15-9
APPROXIMATE THEORETICAL VALUES OF THE
MEAN LIFE FOR GAMMA-DECAY
Change in Mean life, sec

angular
Type of
momentum
radiation Gamma-ray energy
of nucleus,
L: in units of fi 1.00 Mev 0.20 Mev 0.05 Mev
El l 3 X 10- 16 3X 10- 14 2 X 10- 12

MI 1 3 X 10- 14 3X 10- 12 2 X 10- 10
E2 2 7 X 10- 12 3X 10- 8 3 X 10- 5
M2 2 8 X 10- 10 3X 10-6 4 X 10-3
E3 3 6 X 10- 7 5X 10- 2 8 X 10 2
M3 3 7 X 10- 5 6 9 X 104
E4 4 6 X 10- 2 1X 10 5 3 X 10 1 º
M4 4 7 1X 10 7 3 X 10 12
E5 5 8 X 104 4X 10 12 2 X 10 19
M5 5 9 X 106 4X 10 14 2 X 1021
The values given by Eqs. (15-40) and (15-41) are rough estimates
because of the approximate nature of the theory; for example, a factor
which depends on the actual value of the angular momentum of the
initial nucleus is not included. The use of a particular nuclear model,
the independent particle model, may be a more serious limitation. But
the theory should give the general trends of the decay probability with
angular momentum, energy, and mass number, and its usefulness will
depend on how well the theoretical predictions agree with experiment.
The theory of the more specialized methods mentioned at the beginning
of this section (interna! conversion, angular correlations, and angular dis-
tributions) is still more complicated than that relating decay constant
and energy and will not be discussed.
15-10 Gamma-decay and nuclear energy levels: experimental results

and nuclear isomerism. l. M easured lifetimes and nuclear isomerism.
Most known -Y-decay rates have been detennined by the direct measure-
ment of the lifetimes T of the excited states involved. <20 > The total decay
rate X of an excited state is given by
1
(15-42)
T
where >.'Y is the rate of emission of photons and >., is the rate of emission
of conversion electrons. The interna! conversion coefficient a = >.,/>...,
can be measured, <21 > or it can be calculated theoreticauy< 22 > by means of
a method which is independent of the calculation of ;\y, The transition
rate by 'Y-decay alone is then
>. -
'Y - T(l
l
+ a)
= .!._ ,
T-y
(15-43)
and it is this experimental value which is to be compared with theory. But,

before we make this comparison, we must look more closely at the ex-
perimental lifetimes.
The theoretical values of lifetimes shown in Table 15-9 indicate that
sorne mean lives for 'Y-decay may be long enough to be measurable.
Although most 'Y-ray lifetimes have been found to be very short, more than
250 cases are known in which the hfetime for 'Y-decay is measurable,
with observed half-lives from 10- 10 sec to many years. These "delayed"
transitions are called isomeric transitions and the states from which they
originate are called isomeric states or isomeric levels. Because of these
delayed transitions, pairs of nuclear species exist whicb have the same
atomic and mass numbers, i.e., they are isotopic and isobaric, but have
different radioactive properties; these nuclides are called nuclear isomers
and their existence is referred to as nuclear isomerism. <23 > The existence of
excited states with long lifetimes makes possible the direct study of such
stationary properties as tbe angular momentum and magnetic moment.
It must be noted, however, that the isomeric property of a state is not
absolute but is only relative to properties of states of lower energy. An
isomeric state is usually denoted by tbe letter m attached to the mass
number of the nuclide; thus, In 113m represents an isomeric level of In 113 .
The first case of nuclear isomerism to be discovered was that of UX2
and UZ, which ha.ve the same atomic number and the same mass number
but have dífferent half-lives and emit different radiations. Both UX 2
and UZ grow out of UX 1 ( 90 Th 234 ) by p--decay, which gives them the
same mass number 234, and both have the atomic number 90 + 1 = 91.
Uranium X 2 has a half-life of 1.18 minutes; it emits three groups of
electrons with endpoint energies of 2.31 Mev (90%), 1.50 Mev (9%),
and 0.58 Mev (1%). On the other hand, UZ emits four groups of P-
particles with endpoint energies of 0.16 Mev (28%), 0.32 Mev (32%),
0.53 Mev (27%), and 1.13 Mev (13%), and has a half-life of 6.7 hours.
Uranium X 2 is more highly excited than UZ, and about 0.15% of the UX 2
nuclei decay to the ground state UZ by emitting a 'Y-ray with an energy
of 0.394 Mev. Uranium Z, the ground state of 91 Pa 234 , then decays to
UII ( 92 U 234 ) by electron emission. Nearly 100% of the UX 2 atoms
decay directly to UII by electron emission.
15-10) GAMMA-DECAY AND NUCLEAR ENERGY LEVELS 435
This example of isomerism was unique for many years, until the dis-
covery of artificial radioactivity, and especially the production uf many
new radio-nuclides by neutron bombardment, led to the discovery of
many isomeric pairs. The first well-established case of artificially produced
isomers was found in bromine. When a target containing bromine was
bombarded with slow neutrons, the product was found to show three
different half-lives for ,8-decay: 18 min, 4.9 hr, and 34 hr. Chemical
analysis identified the carrier of each of these activities as bromine. This
result was surprising because there are only two stable isotopes of bro-
mine, Br 79 and Br 81 , and in this region of the periodic table the usual
reaction with slow neutrons is the (n, 'Y) process with the formation of
the next higher isotope of the element in question. With bromine, there-
fore, only two radioactive products would be expected
79
asBr + on 1 - asBr
80
+ 'Y,
81 1 82
asBr + on - asBr + 'Y,
and one of the new bromine isotopes seemed to have two half-lives for
,B-decay. The identity of this isotope was determined by bombarding
bromine with 17-Mev 'Y-rays, with the following reactions:
asBr + 'Y -
79
asBr
78
+ on1,
asBr + 'Y - + 0n 1.
81 80
asBr
Two products were obtained, again with three half-lives, 6.4 min, 18 min,
and 4.4 hr. Since Br 80 was formed in both cases, and the 18 min and 4.4 hr
half-lives were observed in both cases, the two half-lives were attributed
to two isomeric states of Br 80 •
lt has been found possible to separate the bromine isomers chemically,
and it has been shown that the 4.4 hr isomer decays into the 18 min isomer
by emitting two 'Y-rays in cascade, with energies of 0.049 Mev and 0.037
. Mev, respectively. The 18 min isomer, which is 35Br 80 in its ground state,
then decays by electron emission (92%) to 36Kr 80, which is stable, and
by positron emission (3%) and orbital electron capture (5%) to the ground
state of 34Se 80 .
An explanation of isomerism was proposed by von Weizsacker in 1936 <24 >.
He suggested that an isomer is an ato"m whose nucleus is in an excited state
and has an angular momentum which differs by severa! units from that
of any lower energy level including the ground state. Von Weizsacker
showed that while 'Y-ray transitions requiring only one or two units of
spin change could be expected to occur with half-lives of the order of
10- 13 sec, those with larger spin change might result in longer and meas-
urable half-lives; each additional unit of spin change could increase the
436 GAMMA·RAYS AND GAMMA-DECAY [CHAP. 15
half-life of a transition of given energy by a factor of the order of 10 6 •

Le.ter, more detailed ca.lculations, such as those discussed in Section 15-9,
as well as experimental results ha.ve verified von Weizsacker's hypothesis.
2. Experimental relations between lifetime and energy. The mea.sured
quantities in studies of 'Y-decay are X-r (or T-r) and E; A is a parameter
which varies from nuclide to nuclide. Equation (15-40) for electric multi-
pole radiation may be written
T-r = constant X A- 2 LJ 3 E-< 2L+ 1>, {15-44)
or
log (T-rA 2 LJ 3 ) = constant - (2L + 1) log E. (15-45)
Hence, for a given type of electric multipole radiation, a graph of log
(T-rA 2L 13 ) against log E should be a straight line with slope - (2L 1). +
Similarly, for magnetic multipole transitions, we have
log (T 1A <2L- 2>' 3) = constant - (2L + 1) log E. (15-46)
The compa.rison of experimental and theoretical results can be discussed
more conveniently if we write out Eqs. {15-40) and (15-41) explicitly
for the different types of 'Y-radiation. These equations are listed in Table
15--10; the values of the constants were actually obtained from an im-
proved version< 25 > of Eqs. {15-40) and (15-41). It has been found ex-
perimentally<25> that the mea.sured lifetimes for L = 2, 3, and 4 fit the
following equations.
L = 2:logT-r(sec) = 4 - 5logE(kev) (15-47a)
L = 3:logT-r(sec) = 17.5 - 7logE(kev) (15-47b)
L = 4:logT-r(sec) = 27.7 - 9logE(kev) {15-47c)

The general form of the dependence of the experimental lifetime on the
'Y-ray energies is, therefore, that predicted by theory, although the results,
as expressed by Eqs. (15-47), are not detailed enough to test the depend-
ence of lifetime on mass number. The best quantitative agreement between
theory and experiment is obtained for M 4 transitions for which the experi-
mental data conform to the relation
(15-48)
in this case the dependence of lifetime on both energy and mass number is
justas predicted by theory. The factor (2/¡ +
1), where /¡ is the angular
momentum (in units of '1.) of the isomeric state is a statistical weight
factor; with this factor the mean deviation of the experimental lifetime
from the value given by Eq. {15-48) is estimated to be less than 30%,
15-10) GAMMA-DECA Y ANO NUCLEAR ENERGY LEVELS 437
TABLE 15-10
THEORETICAL EXPRESSIONS FOR THE DECAY CONSTANT
Disintegration constant
Type of
for gamma-decay,
gamma-radiation
;>..,.: cm- 1
El 1.5 X 10 14 A 213 E 3
MI 2.8 X 10 13E 3
E2 1.6 X 10s A 413 Es
M2 1.2 X 108 A 213 E5
E3 1.1 X 102A2E1
M3 1.8 X 102 A 413 E7
E4 5.0 X 10-sA8l3E9
M4 1.5 X 10-4 A 2E 9
E5 1.6 X 10-11 A l0l3Ell
M5 7.5 X 10-llA8l3Ell
which represents good agreement for this kind of nuclear data. This
agreement is especially imp0rtant because most long-lived isomers emit
'Y-radiation of the M4 type.
The experimental lifetimes for MI, M2, and M3 transitions also depend
on energy and mass number in the general way predicted on the basis
of the independent particle model; but the actual numerical values of
the lifetimes may be two or three orders of magnitude greater than the
theoretical values. Similar results have been found for the electric multi-
pole transitions; the general form of the dependence of lifetime on energy
and mass number agrees with theory. For E3, E4, and E5 transitions,
the decay rates are generally smaller than those predicted by theory,
but there are many E2 transitions, especially in certain even-even nuclei,
which have considerably higher transition probabilities than predicted
by theory.
lt will be remembered that the theory mentioned refers to a particular
model of the nucleus, the independent particle, or shell, model. The experi-
mental data have been interpreted as showing that this model provides
a useful basis for the prediction of electromagnetic transitions in nuclei,
although it is not satisfactory in all its details. The discrepancies between
theory and experiment have, therefore, been used to improve the model;
modifications have been made in it and their validity tested by comparison
with experiment. The study of isomers has been especially useful in this
way and, during the last few years, there has been a gradual shift of
emphasis. lsomers, which were at first of interest as a semi-empirical
branch of radioactivity, are now playing an increasing role in the investi-
- - - - - - - , . . . . . 3 . 0 0 Mev
--------2.52
\ ...-......,,.....-+--+-+-+- 2.41
.........,.""'"+---+-+-+- 2.08
....._ __._...,._ 1.015 Mev
- - - - - - 0.834
j+ _...__....___ o _______........... º
o+
Af27 Cel40
(a) (b)
Pu236 (2.8y)
/
/
/
j
/
/
(6+)------:~
(4+ ) - - - - - - - 56
2+------
0+_......_____
7
lJ232
(e)
F10. 15-14. Energy leve! and decay schemes with emission of 'Y-rays from
short-lived excited states. (a) Mg 27 . (b) La 140 . (e) Pu 236 • [Strominger, Hol-
lander, and Seaborg, Revs. Mod. Phys. 30, 585 (1958).]
gation of nuclear structure. The interest has shifted from the general
relations between lifetimes and energies to details of the nuclear structure
which determine the multipole character and probability of 'Y-ray transi-
tions. One result< 27 •28> has been the accumulation of evidence in favor of a
more highly developed model of the nucleus, the collective model which
will be discussed in Chapter 17.
3. Energy level8 and decay schemes. Sorne decay schemes involving the
emission of 'Y-rays by short-lived excited states are shown in Fig. 15-14.
Values of the angular momentum and parity assigned to a level are in-
dicated at the left end of the horizontal line representing the level; the
energy above that of the ground level is given at the right end of the lines.
When the angular momentum of a state has been measured directly, it
is denoted by underlining that quantum number in the decay scheme,
for example, f + , for the ground state of Al 27 in (a). If the angular
momentum is determined uniquely by other methods, it is designated
by the quantum number without any other modification, e.g., ½+, for
the first excited state of Al 27 . Ground states of even-even nuclei are
known to have zero angular momentum and often fall into this category.
Probable values of the angular momentum are indicated with parentheses,
as in the case of the ground state of Mg 27 in (a). Parity assignments,
indicated by a plus sign (even parity) or minus sign (odd parity), depend
heavily on theory and the underlining and parentheses do not apply to
them. Figures 15-14 (a) and (b) contain information not given for Al 27
and Ce 140 in Fig. 14-13. For sorne states, it has not yet been possible to
assign any values of angular momentum and parity, e.g., the 3.00 Mev
leve} of Ce 140 . Sorne decay schemes are very complicated, and extensive
and careful analysis is needed to work them out. Thus, Bi 206 decays by
orbital electron capture to Pb 206 , and 12 excited levels and 28 'Y-rays have
been found so far. Values of the energy, angular momentum, and parity
have been established for 10 of the levels, and probable values assigned
to the other two; the 28 observed 'Y-rays have been fitted into a decay
scheme which meets all the experimental and theoretical requirements. <29 >
Examples of energy level and decay schemes for nuclear isomers are
shown in Fig. 15-15; these schemes show that members of isomeric pairs
can decay in different ways. In the simplest mode of decay the isomeric
leve} decays by 'Y-emission to the ground level, which is stable: In 113m,
with a half-life of 104 min, emits a 'Y-ray with an energy of 0.392 Mev and
becomes stable In 113, as shown in Fig. 15-15(a). The angular momenta
of the stable and isomeric levels have both been measured in this case;
they differ by four units and the transition has been shown to be of the
M 4 type. In sorne isomeric pairs the ground state, instead of being stable,
may be radioactive and decay by ,s--emission. Thus, 14 hr Zn 69 m emits
a 0.438 Mev 'Y-ray and goes to the ground state of Zn 69 , which then
decays by electron emission to the ground state of Ga 69 ; the decay scheme
is shown in Fig. 15-15(b). The decay of both members of an isomeric pair
may be more complex, as in the case of Br 80, Fig. 15-15(c); Br 80m decays to
Br 80 by emitting two 'Y-rays in cascade; Br 80 then decays by electron emis-
sion to Kr 80, and by positron emission and orbital electron capture to
Se 80 • When the excited leve} decays only by 'Y-emission to the ground
440 GAU:'MA-RAYS AND GAU:'MA-DECAY [CHAP. 15
1 foll3m (104m)
Zn69ftl (14h)
~--r-0.392 e;+>--- --- o.438
~+_L o
2 In113
<½->--......--
Zn69(57m)
(a)
(b)
fü&Om ( 4.4h)
(5-)------ 0.086
<½->--....;....;._
( 2 - ) - -.............-0.037 e¡+>---------
(l+) o
0.475
8% , _ _ __.__0.265
o+--_.I___
Se8º
(e) (d)
-----o
~-..¡..--0.2 15
As77
Sbl24••2 (21m)
- - -Sbl24m,
- - - -(1.3m
- - - 0.019
\....,,-.---1. 42
,\_._ _ _ _..,._ 2.296
1-----1.ao
.,.__ _ _........i.-1.972
(e) (í)
Fm. 15-15. Decay schemes of nuclear isomers. (a) In 113 •. (b)Zn 6 D"'.
(e) BrBO•. (d) Ge 77•. (e) Cd 11 s... (f) Sb 124"'1 and Sb 124 "'2. [Strominger,
Hollander, and Seaborg, Revs. Mod. Phys. 30, 585 (1958).)
REFERENCES 441
level, as in the above cases, the members of the isomeric pair are said to
be genetically related. The more excited member of an isomeric pair may
decay in two ways: Ge 77m decays to Ge 77 by emission oi a 'Y-ray; it can
also decay by electron emission to the ground level of As 77 orto an excited
level; the ground level of Ge 77 is unstable and decays to excited levels of
As 77 by electron emission; the decay scheme for Ge 77m and Ge 77 is shown in
Fig. 15-15(d). The members of an isomeric pair may decay independently
by p--emission: Cd 11 11m and Cd 116 are not genetically related, and both
have complex ,B-spectra. The decay scheme is shown in Fig. 15-15(e); it
is especially interesting because two of the energy levels of the product
nucleus, In 116, form an isomeric pair. Triple isomerism has also been
observed and Fig. 15-15(f) shows an example; Sb 124m,, with a half-life of
21 min, decays by 'Y-emission to the ground state, while Sb 124m 1, with a
half-life of 1.3 min, decays both by 'Y-emission and p--emission. The ground
state of Sb 124 is unstable and also decays by p--emission.
For a detailed summary of isomeric decay schemes the reader is referred
to a recent review by Alburger (see gen. ref.).
REFERENCES
GENERAL
RuTHERFORD, CHADWICK, and ELLis, Radiatiom from Radioactive Sub-
stanceB. New York: The Macmillan Co., 1930, Chapters 12, 15.
5, 23, 24, 25.
K. S1EGBAHN, ed., Beta- and Gamma-Ray Spectroacopy. New York: Inter-
science; Amsterdam: North-Holland Publishing Co., 1955.
M. DEUTSCH and O. KoFOED-HANSEN, "Gamma-Rays," Experimental
Nuclear PhyBics, E. Segrll, ed., New York: Wiley, 1959. Vol. 111, Part X.
D. E. ALBURGER, "Nuclear lsomerism," Encyclopedia of PhyBics. Vol. 42
(Nuclear Reactions 111), pp. 1-108. Berlin: Springer, 1957.
F. A.TzENBERG-8ELOVE, ed., Nuclear Spectroacopy. New York: Academic Press,
1960. Parts A, B.
G. W. GRODSTEIN, X-Ray Attenuation Coeffi,cients from 10 kev to 100 Mev,
National Bureau of Standards Circular 583. Washington, D.C.: U.S. Govern-
ment Printing Office, 1957.
H. A. BETHE and J. AsHKIN, "Passage of Radiations Through .Matter,"
Experimental Nuclear PhyBics, E. Segre, ed., New York: Wiley, 1953 (Vol.
I. Part 11).
STROMINGER, HoLLANDER, and SEABORG, "Table of lsotopes," Revs. Mod.
Phys. 30, 585 (1958).
WAY, K1NG, McGINNIS and VAN LIESHOUT, Nuclear Level Schemes, A = 40
to A = 92, U.S. Atomic Energy Commission Document TID-5300, Sept. 1955.
H. GOLDSTEIN, Fundamental AapectB of Reactor Shielding. Reading, Mass.,
U.S.A.: Addison-Wesley, 1959.
PARTICULAR
l. C. M. DAvissoN and R. D. EvANs, "Mea.surements of Gamma-Ray
Absorption Coefficients," Phys. Rev. 81, 404 (1951).
2. 9. M. DAv1ssoN and R. D. EvANs, "Gamma-Ray Absorption Coefficients,"
Reva. Mod. Phys. 24, 79-107, 1952.
3. W. HEITLER, The Quantum Theory of Radiation, 3rd ed. Oxford University
Press, 1954.
4. J. W. M. DuMoND, "High Resolving Power, Curved-Crystal Focussing
Spectrometer for Short Wave-Length X-Rays and Gamma-Rays," Rev. Sci.
lnstr. 18, 626 (1947); aiso, "Gamma-Ray Spectroscopy by Direct Crystal
Diffraction," Ann. Revs. Nucl. Sci. 8, 163 (1958).
5. MuLLER, HoYT, KLEIN, and DuMOND, "Precision Mea.surements of Nu-
clear Gamma-Ray Wavelengths of Ir 192 , Ta 182 , RaTh, Rn, w1s1, C8 137, Aul98,
and Annihilation Radiation," Phys. Rev. 88, 775 (1952).
6. PAGE, STEHLE, and GuNsT, "Direct Comparison of m/e for the Positron
and the Electron," Phys. Rev. 89, 1273 (1953).
7. DEUTSCH, ELLIOTT, and EvANs, "Theory, Design and Applications of a
Short Magnetic Lens Electron Spectrometer," Rev. Sci. Instr. 15, 178 (1944).
8. W. C. PEACOCK and M. DEUTSCH, "Disintegration Schemes of Radioactive
Substances, IX. Mn 52 and V 48 ," Phys. Rev. 69, 306 (1946).
9. J. W. DuNWORTH, "Coincidence Methods in Nuclear Physics," Rev. Sci.
lnstr. 11, 167 (1940).
10. A. C. G. MITCHELL, "The Use of Coincidence Counting Methods in De--
termining Nuclear Disintegration Schemes," Revs. Mod. Phys. 20, 296 (1948).
11. R. L. WALKER and B. D. McDANIEL, "Gamma-Ray Spectrometer
Mea.surements of Fluorine and Lithium Under Proton Bombardment," Phys.
Rev. 74, 315 (1948).
12. FowLER, LAURITSEN, and LAURITSEN, "Gamma-Radiation from Excited
States of Light Nuclei," Revs. Mod. Phys. 20, 236-277 (1948).
13. P. R. BELL, "The Scintillation Method," Beta- and Gamma-Ray Spec-
troscopy, K. Siegbahn, ed. (see gen. ref.), p. 133. ..
14. ERIKSEN, JENSSEN, and SUNDE, "Scintillaíion Counter as Gamma-Ray
Spectrometer," Physica 18, 9 (1952).
15. E. RuTHERFORD and W. A. WoosTER, "The Natural X-Ray Spectrum of
RaB," Proc. Cambridge Phil. Soc. 22, 834 (1925).
16. C. D. ELLIS, "The 'Y-Rays of Radium (B+ +
C) and of Thorium (C C'),"
17. C. D. ELLIS and W. A. WoosTER, "The Atoinic Number of a Radio-
active Element at the Moment of the Emission of the 'Y-Rays," Proc. Cambridge
Phil. Soc. 22, 844 (1925).
18. G. D. LATYSHEV, "lnteraction of 'Y-Rays with Matter and the Spectros-
copy of 'Y-Radiation," Revs. Mod. Phys. 19, 132 (1947).
19. J. M. BLATT and V. F. WEISSKOPF, Theoretical Nuclear Physics. New
York: Wiley, 1952, Chapter 12.
20. R. E. BELL, "Measurement of Short Lifetimes of Excited States: Measure-
ment by Delayed Coincidence; Comparison Methods," Beta- and Gamma-Ray
Spectroscopy, K. Siegbahn, ed. (see gen. ref.), p. 494.
REFERENCES 443
21. M. DEUTSCH and O. KoFOED-HANSEN, "Gamma-Rays," Experimental

Nuclear Physics, E. Segre, ed. (see gen. ref.), p. 360.
22. M. E. RosE, "Theory of lnternal Conversion," Beta- and Gamma-Ray
Spectroscopy, K. Siegbahn, ed. (see gen. ref.), p. 396.
23. E. SEGRE and A. C. HELMHOLZ, "Nuclear Isomerism," Revs. l\fod.
Phys. 21, 271 (1949).
24. C. F. VON WErzs.iCKER, "Metastabile Zustande der Atomkerne," Natur-
wiss. 24, 813 (1936).
25. S. A. Moszxowsxr, "Theory of Multiple Radiation," Beta- and Gamma-
Ray Spectro8copy, K. Siegbahn, ed. (see gen. ref.), p. 373.
26. M. GoLDHABER and A. W. SuNYAR, "Classification of Nuclear Isomers,"
Phys. Rev. 83, 906 (1951).
Beta- and Gamma-Ray Spectroscopy (see gen. ref.) p. 453.
28. L. I. RusrNov and D. A. VARSHALOVICH, "Electromagnetic Transitions
in Isomeric Nuclei," J. Nucl. Energy, Part A: Reactor Science, 10, 170 (1959).
Translated from Atomnaya Energiya S, 432 (1958).
29. D. E. ALBURGER and M. H. L. PRYCE, "Energy Levels in Pb 2 º6 from the
Decay of Bi2 º6 ," Phys. Rev. 95, 1482 (1954).
30. CHUPP, DuMoND, GoRDON, JoPsoN, and MARK, "Precise Determination
of Nuclear Energy Levels in Heavy Elements," Phys. Rev. 112, 518 (1958}.
31. KAz1, RASMUSSEN, and MARK, "Six-Meter Radius Bent-Crystal Spectro-
graph for Nuclear Gamma Rays," Rev. Sci. Imtr., 31, 983 (1960).
32. KAzr, RASMUSSEN, and MARK, "Measurement of the Deuteron Binding
Energy using a Bent-Crystal Spectrograph," Phys. Rev. 123, 1310 (1961).
444 GAMMA-RAYS AND OAHMA-DECAY [CHAP. 15
PROBLEMS
l. When a beam of 'Y-rays pa.sses through aluminum under "good geometry"

conditions, the intensity varíes with the energy of the raye and the thickness
of the aluminum, us ahown in the following te.ble.
Relative intensity
Thicknesa of
aluminum, cm Energy Energy Energy
= 0.835 Mev = 1.14 Mev = 2.75 Mev
o 1.00 1.00 1.00

2.5 0.634 0.675 0.788
5 0.401 0.456 0.622
7.5 0.254 0.308 0.490
10 0.161 0.208 0.387
12.5 0.102 0.141 0.305
15 0.065 0.095 0.240
20 0.026 0.043 0.150
25 0.010 0.020 0.093
30 0.0042 0.0090 0.058
35 0.0017 0.0041 0.036
40 ..... 0.0019 0.022
Find, graphically, the value of the absorption coefficient µ(cm- 1 ) in each case.
Compare the resulta with the values given in Table lf'rl.
2. When 'Y-raya of energy 2.76 Mev pass through varying thicknesses of
aluminum, tin, and lead, the intensity varíes with the thickness of each aub-
at&nce as ahown in the te.ble on the next page. Find, graphically, the value
of ¡.&(cm- 1) in each case and compare the resulta with the values given in
Table 15-1.
3. Calcule.te the values of the cross section 11,T for photoelectric absorption
in aluminum and lead at the photon energies listed in Table 15-2. Plot the
values obt&ined ag&inst the photon energy in Mev.
4. Calcule.te the values of the cross sections <T, <T., and <T"' for Compton scat-
tering in aluminum at 10 values of a which cover the energy range of Table
lf'r3. Plot the resulting values ag&inst the photon energy in Mev.
5. From the curve of Fig. 15-6, calculate the values of the cross section
•" for pair formation in aluminum and lead at photon energies of 1, 5, 10, 15, 20,
and 25 Mev.
6. Calcule.te, from theory, the values of µ{cm- 1) for copper for 'Y-raya of
energy 0.835 Mev, 1.14 Mev, and 2.76 Mev. Compare the resulta with the
theoretical and experimental values given in Table lf'rl.
7. A heterogeneous beam of 'Y-raya from a radioactive source is allowed to
fall on an aluminum radiator and the electrons ejected are analyzed in a ma.gnetic
spectrograph. The electrons fall into three groups with maxima at 11960,
5845, and 2750 gauss-cm, respectively. What are the energies of the three 'Y-raya?
PROBLEMS 445
Aluminum Tin Lead

Thickness, Thickness, Thickness, Intensity
Intensity Intensity
cm cm cm
o 1.00 o 1.00 o 1.00

2.5 0.788 2 0.586 1 0.620
5.0 0.622 3 0.449 2 0.384
7.5 0.490 5 0.263 3 0.238
10.0 0.387 7 0.154 4 0.147
12.5 0.305 8 0.118 6 0.057
15 0.240 10 0.069 8 0.022
20 0.150 12 0.041 10 0.008
25 0.093 16 0.014 12 0.003
30 0.058 18 0.008
35 0.036
40 0.022
8. In an experiment on the 'Y-rays from Ba 131 , the rays are allowed to fall
on a lead radiator. Four groups of photoelectrons are observed, corresponding
to Hr values of 1250, 1445, 2050, and 2520 gaUBB-Cm, respectively. The binding
energy of the K-shell electrons in lead is 0.0891 Mev. What are the energies of
the 'Y-rays?
9. In an early experiment on the 'Y-rays from RaB, Ellis measured the energies
of the photoelectrons ejected from radiators of woHram, platinum, lead, and
uranium. In each case, three groups of K-electrons were observed with the
following energies, in Mev.
w Pt Pb u
a. 0.166 0.158 0.149 0.122
b. 0.220 0.212 0.203 0.174
c. 0.276 0.269 0.260 0.231
The binding energies of the K-electrons in the four radiators are 0.0693, 0.0782,
0.0891, and 0.1178 Mev, respectively. Find the mean energy and the wave-
length of each of the three 'Y-rays.
10. Terbium-160 decays by tf-emiBBion to excited states of dysprosium-160.
Two of the 'Y-rays emitted by the latter nuclide give rise to an interna! conversion
line spectrum with the following energies: 32.9, 78.0, 84.7, 86.1, 143.0, 187.0, and
194.7 kev. The binding energies for K-, L-, M-, and N-shell electrons in dys-
prosium are 53.4, 8.6, 1.9, and 0.4 kev, respectively. Find the energies of the
two 'Y-rays, and the shell and 'Y-ray with which each conversion line is associated.
11. Derive an expreBBion for the kinetic energy (in ev) with which the nucleus
recoils when a 'Y-ray is emitted¡ express the energy of the 'Y-ray in kev, and the
mass of the recoil nucleus in amu.
12. The reactions Cr(p,n)Mn, and Fe(d,a)Mn both lead to activities of

21.3 min and 5.8 days, while the (a,n) reaction on vanadium givea a product with
a half-life of 310 days. To what nuclide, or nuclides, do the two ahorter activities
belong?
13. The nuclide Cr49 decays by positron emission, emitting three groupa of
particlea with endpoint energiea of 1.54 Mev, 1.39 Mev, and 0.73 Mev, respec-
tively. Gamma-rays with energies of 0.15 Mev, 0.061 Mev, and 0.089 Mev, are
observed. Construct a decay scheme to fit these data.
14. The nuclide Cr 51 decays by orbital electron capture to V 51 and the Q-
value for the ground state transition is 0.75 Mev. Three 'Y-raya are observed,
with energies of 0.325 Mev, 0.320 Mev, and 0.65 Mev respectively, the first
two of which are in cascade. Conatruct an energy leve) and decay scheme to
fit these data.
15. The irradiation of Co 59 with neutrons produces a pair of isomers of Co 60 ,
one with a half-life of 10.4 min, the other with a half-Jife of 5.2 years. The
10.7-min leve! has an excitation energy of 0.059 Mev and transforma mainly
to the ground state of Co 60 by interna) conversion; it alao decaya by emitting
¡'3--particles with an endpoint energy of 1.54 Mev. The ground atate leve! of
Co 60 decays by 13--emiasion with an endpoint energy of 0.31 Mev. Two 'Y-raya
are observed, in cascarle, with energies of 1.17 Mev and 1.33 Mev, respectively.
The Q-value for the transition from the isomeric state of Co60 to the ground
state of Ni 60 is 2.876 Mev. Construct an energy leve! and decay scheme based
on these data.
CHAPTER 16
NUCLEAR REACTIONS
In the discussion of nuclear reactions in Chapter 11, the emphasis was

on the production of new nuclear species and the interest was mainly in
the change of one nucleus into another. The quantitative aspects stressed
were those of the mass and energy balance. The study of nuclear reac-
tions is also important for other reasons. lnformation about the relative
probabilities of different reactions provides clues to the problem of nuclear
structure and offers a testing ground for ideas about nuclear forces. Any
nuclear transmutation process may lead toan excited state of the product
nucleus, and the d€cay of the excited state then gives information about
energy levels and decay schemes. The process may be caused by any
kind of incident particle and may result in the emission of any kind of
particle (proton, neutron, deuteron, a-particle, or 'Y-ray). The incident
and emitted particle may even be of the same kind, as in scattering proc-
esses. The great number oi nuclear reactions which are possible provide,
therefore, a wealth of experimental data for the field of nuclear spectro-
scopy and for the theory of nuclear structure. It is from this standpoint
that nuclear reactions will be considered in the present chapter.
16-1 Nuclear reactions and excited states of nuclei. In Chapter 11,

the nuclear reactions considered were of the type
x+X-+ Y+y, (16-1)
or, in more compact notation, X(x, y) Y. The equation and the notation
both mean that particle x strikes nucleus X to produce nucleus Y and
particle y. The particles x and y may be elementary particles or 'Y-rays,
or they may themselves be nuclei, e.g., a-particles or deuterons. The
transmutation represented by Eq. (16-1) <loes not cover all the reactions
of interest. In the general case, more than one particle may emerge,
or the outgoing particle may be the same as the incident particle. For
the purpose of this chapter, the nuclear reactions to be considered may
be represented by the set of equations
X+x,
X*+ x, (16-2)
X +X -
{ Y+ Y,
Z + z, etc.
447
448 NUCLEAR REAcrIONS {CHAP. 16
In the first two reactions of the set, the outgoing particle is of the same
kind as the incident particle, and the procesa is called scattering. The
first reaction representa elastic scattering, in which the total kinetic energy
of the system, projectile plus target, is the same before the collision as
after. Sorne kinetic energy is usually transferred from the projectile to
the target nucleus, but the latter is left in the same interna}, or nuclear,
state as before the collision. A collision between two billiard halls, in
which neither ball is damaged or otherwise changed, is an example of an
elastic collision. The second reaction representa inelastic scattering, in
which the target nucleus X is raised into an excited state X*, and the
total kinetic energy of the system is decreased by the amount of the excita-
tion energy given to the target nucleus. The other reactions of the set
represent different possible nuclear transmutations in which the product
nuclei may be formed in their ground states or, more often, in excited
states. The excited product nucleus· usually decays very quickly to the
ground state with the emission of 'Y-rays.
The fact that the product nucleus in a transmutation procesa can be
left in an excited state was discovered by measuring the energies of the
protons emitted in (a, p) reactions on light elementa, the first nuclear
reactions studied in detail. When a given light element was bombarded
with monoenergetic natural a-particles, one or more groups of protons,
each containing particles of the same energy, was observed for a particular
direction of emission.<1· 2 > The existence of the distinct proton groups
was demonstrated by their different ranges. When boron was bombarded
with a-particles from polonium with a range of 3.8 cm, or an energy of
5.30 Mev, two groups of emergent protons were found at right angles to
the incident a-particle beam; these protons had ranges of 20 cm and 50 cm,
respectively. Similar resulta were obtained with other light elementa such
as fluorine and aluminum; when the latter was bombarded with a-particles
from RaC', four groups of protons were observed< 3 > for each -of severa}
different energies of the incident a-particles.
The energy balance of the reactions, or Q-value, can be calculated for
each energy group by the methods of Section 11-2. The proton group
with the greatest energy gives the greatest Q-value, and this value is
supposed to correspond to the ground state of the product nucleus. A
proton group of lower energy gives a lower Q-value, and the difference
between the greatest Q and a smaller one gives the excitation energy of
the product nucleus after emission of the lower energy proton. These
ideas and their application involve some of the basic procedures of nuclear
physics and will be illustrated by an analysis of two experimenta, on the
reaction
16-1) NUCLEAR REACTIONS AND EXCITED STATES OF NUCLEI 449
which are representative of work relating nuclear reactions to energy

levels. In one experiment, <4 > an aluminum foil was bombarded with
a:-particles accelerated in a cyclotron toan energy of 7.3 Mev. The ranges
of the protons emerging at an angle of Oº with the direction of the incident
beam were measured by counting coincidences between protons and "Y-raya.
Four groups of protons were observed with ranges in air of 101.6 cm,
60.8 cm, 40.8 cm, and 25-30 cm, respectively. The corresponding energies,
obtained from range-energy curves for protons, are 9.34, 6.98, 5.55, and
4.2-4.65 Mev. The energies of the incoming a:-particle and the outgoing
proton are known for each group, as are the masa numbers of the initial
and final nuclei and, since the angle at which the protons are observed is
Oº, Q-values can be calculated from Eq. (11-10). The greatest proton
energy gives Q = 2.22 Mev, which corresponda to the formation of Si 30
in its ground state; the three lower-energy proton groups give Q-values of
-0.06, -1.44, and about -2.4 Mev, respectively. There are, therefore,
three excited states of Si 30 with energies 2.28, 3.66, and about 4.6 Mev
above that of the ground state. The ground state energy is taken as the
zero point of the energy scale for the various states. The resulta, which
are collected in the first part of Table 16-1, are in good agreement with
those of earlier experimenta. <3 , 5 > It was also found in this experiment
TABLE 16-1
ExcITED 8TATES OF 8i
3
ºFROM THE BOMBARDMENT
OF Al.UMINUM'. BY ALPHA-PARTICLES
Exr,eriment 1<4 > Experiment t< 6>

Energy of a:-particles = 7.3 Mev Energy of a:-particles = 21.54 Mev
Energy of Energy of
Energy of Energy of
protons protons Q-value,
Q-value, excited excited
emitted emitted
Mev state, Mev state,
at Oº, at 90°,
Mev Mev
Mev Mev
9.34 2.22 o ..... 2.22 o

6.98 -0.06 2.28 .. .. . .... . ...
5.55 -1.44 3.66 16.85 -1.27 3.49
4.2--4.65 -2.4 4.6
14.96 -3.22 5.44
13.28 -4.96 7.18
12.29 -5.98 8.20
11.27 -7.04 9.26
10.68 -7.65 9.87
9.72 -8.64 10.86
450 NUCLEAR REACTIONS [CHAP. 16
5
4
3
2
I 150
Total absorption in proton bea.m, mg/cm2 of Al
F10. 16--1. Proton groups from the reaction Al 27 (a, p)Si 30 (Brolley et al.<6>).
that no 'Y-rays accompanied the highest-energy group of protons, while

the lower-energy protons did show coincidences with 'Y-rays; the lower-
energy protons are indeed associated with excited states of the Si 30 nucleus,
while the highest-energy group is associated with the ground state.
In another experiment, <&> a thin aluminum foil was bombarded in
vacuum with a-particles accelerated in a cyclotron to an energy of 21.54
Mev. The protons emerging atan angle of 90° with the direction of the
incident beam were count¡ed with two proportional counters placed in
series and arranged for coincidence counting. The ranges of the proton
groups in aluminum were measured by allowing the protons to pass through
aluminum foils before entering the counting system. Under the conditions
of the counting procedure, the number of protons counted in a particular
group showed a peak at the end of the range so that when the number of
counts is plotted against the thickness of aluminum, each peak in the curve
corresponds to the range or energy of a different proton group. The
change of the counting rate with the thickness of the aluminum absorber
is shown in Fig. 16-1. The energies of the proton groups were obtained
from the ranges in aluminum by means of the known theoretical relation-
ship between range and energy in that material, and are listed in the
16-1] NUCLEAR REACTIONS AND EXCITED STATES OF NUCLEI 451
second part of Table 16-1. The Q-values are obtained from Eq. (11-9).
The range of the most energetic proton group could not be determined
well in this experiment because the counting rates became too low, so
the value Q = 2.22 Mev was used for the reaction leading to the ground
state of Si 30 , and gave the energy levels listed in the table. The energy
level values at 3.66 Mev and 3.49 Mev correspond to the same excited
state, the disr.repancy being experimental. The level at about 4.6 Mev
observed in the first experiment was not observed in the second experi-
ment, but other work has established the existence of a level at 4.75 Mev.
There are, therefore, at least 9 excited energy levels of Si 30, as shown in
the table, and it is probable that there are more which have not been
excited by the (a, p) reaction.
The emission of groups of particles is not limited to (a, p) reactions,
but is also found in many other reactions. In each case, the different
energy groups correspond to different states of the product nucleus, the
lower-energy groups corresponding to excited states which decay to the
ground state by emitting 'Y-rays.
Another important feature of many nuclear reactions was discovered
during the investigation of {a, p) reactions when the energy of the incident
a-particles was varied. lt was found in the early experiments that when
the energy of the incident a-particles was increased the yield of protons
did not increase monotonically, as might be expected on intuitive grounds,
but showed sharp maxima at certain discrete values of the energy. <3 > In
the bombardment of Al by a-particles, the yield of protons had peaks at
a-particle energies of 4.0, 4.49, 4.86, 5.25, 5.75, and 6.61 Mev, and was
markedly lower in the energy range between any two of these peaks. The
occurrence of maxima in the reaction rate at different energies is called
resonance, and the particular energies at which the maxima occur are called
resonance energies. The reason for the use of these terms will be discussed
later. At each resonance energy of the incident a-particles the different
proton groups are observed, and Q-values can be calculated from the
known resonance energy and the corresponding energies of the proton
groups. These Q-values are the same as those obtained for nonresonance
a-particle energies, and lead to the same excited states of the product
nucleus. The experimental results obtained at the resonance energies
are often more useful than those obtained at other energies because more
events occur at the resonance energies and more precise information can
be obtained about the Q-values and excited levels in the compound nucleus.
The resonance phenomenon does not, however, add anything new about
the energy levels of the product nucleus.
Resonance is important in nuclear reactions because knowledge of the
resonance energies in a reaction yields information about certain energy
levels of the nucleus referred to in Chapter 11 as the compound nucleus.
The transmutations discussed in Chapter 11 were written in a way which

indicates that the incident particle combines with the target nucleus to
forro an intermediate compound nucleus. When Al is bombarded with
a-particles, the compound nucleus is [P 31 ], and it turns out that the exist-
ence of resonance energies is related to the energy levels of P 31 . Reso-
nance is not limited to (a, p) reactions but is also found in other nuclear
reactions.
16-2 The compound nucleus. In 1936, Bohr proposed his theory< 7 .s>
of the compound nucleus, which has been extremely useful in the correla-
tion and interpretation of nuclear reactions. The basic ideas of this theory
will therefore be considered before a detailed discussion of nuclear reactions
is undertaken. Bohr assumed that a nuclear reaction takes place in two
steps:
(1) The incident particle is absorbed by the initial, or target, nucleus
to form a compound nu.clem.
(2) The compound nucleus disintegrates by ejecting a particle (proton,
neutron, a-particle, etc.) ora "Y-ray, leaving the final, or product, nucleus.
Bohr assumed also that the mode of disintegration of the compound
nucleus is independent of the way in which the latter is formed, and de-
pends only on the properties of the compound nucleus itself, such as its
energy and angular momentum. The two steps of the reaction can then
be considered separate processes:
1. Incident particle +
initial nucleus -+ compound nucleus.
2. Compound nucleus -+ product nucleus +
outgoing particle.
Bohr's assumptions are in accord with many of the facts of nuclear
transmutation. It was shown in Section 11-9 that when a given nuclide
is bombarded with particles of a single type, severa} different new nuclides
can be fonned. When Al 27 is bombarded with protons, the new nuclide
may be Mg 2 ', Si 27, Si 28, or Na 24 • According to Bohr, the interaction
between Al 27 and a proton is assumed to give the compound nucleus
[i 4Si 28], which can then disintegrate in any one of several ways: into Mg 24
andan a-particle; or into Si 27 and a neutron; or into Si 28 and a "Y-ray;
or even into Na 2 ', 3 protons, anda neutron. Bohr's assumption is also
in accord with the picture of the nucleus as a system of particles held to-
gether by very strong short-range forces. When the incident particle
entera the nucleus, its energy is quickly shared among the nuclear particles
before l'EHlmission can occur, and the state of the compound nucleus is
then independent of the way it was formed. That this conclusion is
reasonable can be shown by the following arguments. On being captured,
the incident particle makes available a certain amount of excitation energy,
which is nearly equal to the kinetic energy of the captured particle plus
1&-2) THE COMPOUND NUCLEUS 453
its binding energy in the compound nucleus. The magnitude of the excita-
tion energy can be calculated from the masses of the incident particle,
target nucleus, and compound nucleus, and the kinetic energy of the
incident particle. Consider, for example, the capture of a neutron by
AI 27 to form the compound nucleus [Al 28],
1sAI 27 + on 1 -+ (13Al ].
28
The masses of the interacting neutron and nucleus are 1.00898 and 26.99008
amu, ora total of 27.99906 amu; that of the compound nucleus is 27.99077
amu. The mass excess is 0.00829 amu, corresponding to 7.72 Mev, to
which must be added the kinetic energy of the incident neutron. In the
case of a slow neutron, the kinetic energy may be neglected. If the incident
neutron has a kinetic energy of 1 Mev, the excitation energy is nearly
7.72 + 1 = 8.72 Mev and the energy of the compound nucleus [13AI 28]
is greater than the energy of the ground state of Al 28 by this amount.
Immediately after the formation of the compound nucleus, the excitation
energy may be considered to be concentrated on the captured particle,
but this additional energy is rapidly distributed among the other particles
in the compound nucleus as a result of interactions among the nuclear
particles. The distribution presumably takes place in a random way.
At a given instant, the excitation energy may be shared among severa}
nucleons; at a later time it may be shared by other nucleons, or it may
eventually again become concentrated on one nucleon or combination of
nucleons.. In the Iatter case, if the excitation energy is large enough,
one nucleon, or a combination of nucleons, may esca.pe, and the compound
nucleus disintegrates into the product nucleus and outgoing particle. The
energy that must be concentrated on a single nuclear particle or group of
particles in order to separate it from the compound nucleus is called the
separation or dis8ociation energy, and is usually about 8 Mev.
As a result of the random way in which the excitation energy is dis-
tributed in the compound nucleus, the latter has a lifetime which is rela-
tively long compared with the time that would be required for a particle
to travel across the nucleus. The latter time interval, sometimes called
the natural nuclear time, is of the order of magnitude of the diameter of
the nucleus divided by the speed of the incident particle. If the incident
particle is a 1-Mev neutron, its speed is about 10 9 cm/sec. Since the
diameter of the nucleus is of the order of 10- 12 cm, the time required for a
1-Mev neutron to cross the nucleus is of the order of 10-21 sec. Even a
slow neutron with a velocity of 105 cm/sec would need only about 10- 17
sec to cross the nucleus. It will be seen below that the lifetime of a com-
pound nucleus may be as long as 10- 111 or 10- 14 sec, which is long com-
pared with the natural nuclear time. During its relatively long lifetime,
the compound nucleus "forgets" how it was formed, and the disintegration
is independent of the mode of formation. The compound nucleus may

be said to exist in a. "qua.si-sta.tiona.ry" sta.te, which mea.ns tha.t a.lthough it
exista for a. time interva.l which is very long compa.red with the natural
nuclear time, it can still disintegra.te by ejecting one or more nucleons.
These qua.si-sta.tiona.ry ata.tes a.re usua.lly ca.lled virtual atates or virtual
levela in contra.st to bound atatea or bound levela, which can deca.y only by
emitting 'Y-ra.dia.tion.
There a.re many wa.ys in which the excita.tion energy of the compound
nucleus can be divided a.mong the nuclear pa.rticles a.nd, since ea.ch dis-
tribution corresponda to a. virtual level, there a.re many possible virtual
levels of the compound nucleus. These levels a.re closely rela.ted to the
phenomenon of resona.nce. It is rea.sonable to a.ssume tha.t if the energy of
the incident pa.rticle is such tha.t the total energy of the system, incident
pa.rticle plus ta.rget nucleus, is equa.l to the excita.tion energy of one of the
levels of the compound nucleus, the proba.bility tha.t the compound nucleus
will be formed is much grea.ter tha.n if the energy fa.lis in the region between
two levels. The system is a.na.logous to tha.t of a. radio wa.ve a.nd a. tuned
receiver circuit. When the frequency (energy) of the incoming wa.ve is
equa.l to tha.t of the cir,:iuit, the wa.ve a.nd circuit a.re in resona.nce a.nd the
reception is good; when the two frequencies (energies) a.re not equa.l, the
reception is poor. The occurrence of a. resona.nce peak in the ra.te of a.
nuclear rea.ction when the energy of the incoming pa.rticles is va.ried shows
tha.t the compound nucleus has a.n energy level whose excita.tion energy is
very nea.rly the sum of the binding energy of tha.t pa.rticle a.nd its kinetic
energy. In the Al 27 (a, p)Si 30 rea.ction discussed a.hove, 18 resona.nces ha.ve
been observed a.t a-particle energies from 3.95 Mev to 8.62 Mev; the bind-
ing energy of the a-pa.rticle in the compound nucleus [P 31 ) is 9.68 Mev.
The excita.tion energy of a leve! ca.nnot be obta.ined just by a.dding the
resona.nce energy to the binding energy, beca.use tha.t would neglect the
motion of the compound nucleus. Sorne of the energy of the incident
particle is used in supplying kinetic energy to the compound nucleus and
is not a.va.ita.ble for excita.tion.
The fra.ction of the pa.rticle energy tha.t must be added to the binding
energy in order to get the excita.tion energy can be derived as follows. If
the ta.rget nucleus is a.ssumed to be a.t rest before the collision, the momen-
tum of the incident pa.rticle must be equa.l to tha.t of the compound nuecleus,
m.,v = McNV,
where v, V, a.nd M CN a.re the speed of the incident particle, a.nd the speed
and ma.ss of the compound nucleus, respectively. Then
V= v mz •
McN
16-2] THE COMPOUND NUCLEUS 455
The portion of the energy of motion of the incident particle that goes
into exciting the initial nucleus will be called E;, and is given by
, 1 2 1 2 1 2 1 2 m!
Ex = 2 m:,:V - 2 McN V = 2 ma:V - 2 McNV M2
CN
(16-3)
= !2 m.,v 2
(1 - ~) = E., (1 - ~) ·
McN McN
In calculations of this kind, the masses may be replaced by mass numbers

without introducing any serious errors. Equation (16-3) may also be
written
(16-4)
since M cN = Mx m.,. +
The energies of the excited levels of [P 31 ] which correspond to a-particle
resonances are given in Table 16-2. The excitation energies of these
levels are considerably greater than the dissociation energy ( ~8 Mev)
for a particle and the compound nucleus can disintegrate by particle
emission so that the levels are virtual. The reactions Al 27 (a, p)Si 30 and
Al 27 (a, n)P 3 º have been observed, a result which is consistent with the
idea that the compound nucleus [P 31 ] formed by the bombardment of
aluminum with a-particles can disintegrate in more than one way.
Each excited state of the compound nucleus, whether bound or virtual,
has a certain mean lifetime T; there is a certain period of time, on the
average, during which the nucleus remains in a given excited state before
decaying by emission of either a particle or a 'Y-ray. The reciprocal of
the mean life is the disintegration constant, which gives the probability
per unit time of the emission of a particle or 'Y-ray. In the discussion of
energy states excited by nuclear reactions it is customary to use, instead
of the disintegration constant, a quantity proportional to it, called the
level width, and defined by the relation
h
r=-·
211T
(16-5)
The level width has the units of energy, and its use is based on an applica-
tion of the Heisenberg uncertainty principie (Section 7-8). According
to this principie, the accuracy with which the energy and time can be
determined for a quantized system such as an atomic nucleus is limited by
the relationship
h
(,IB) (.:it) ~ 2,r ' (16-6)
456 NUCLEAR REACTIONS [CHAP, 16
TABLE 16-2
LEVELS OF THE COMPOUND NUCLEUS [P 31 ]
ExcITED IN THE BoMBARDMENT OF ALUMINUM BY
a-PARTICLES
a-particle Energy to be
Excited leve!
added to the
resonance energy, of [P3 1 ),
Mev binding energy,
Mev
Mev
3.95 3.44 13.12

4.53 3.95 13.63
4.70 4.09 13.77
4.84 4.22 13.90
5.12 4.46 14.14
5.29 4.61 14.29
5.64 4.91 14.59
6.01 5.23 14.91
6.38 5.56 15.24
6.57 5.72 15.40
7.00 6.10 15.78
7.20 6.27 15.95
7.34 6.39 16.07
7.60 6.62 16.30
8.04 7.00 16.68
8.24 7.18 16.86
8.42 7.33 17.01
8.62 7.51 17.19
where llE is the uncertainty in the energy and At is the uncertainty in the
time. The mean lifetime of an excited state may be identified with the
uncertainty At corresponding to an uncertainty llE in the energy. The
latter is defined as the width r, in energy, of the excited level, giving Eq.
(16-5). A state with a very short mean lifetime is poorly defined in energy
and the width r is relatively large, while a long-lived state is sharply
defined in energy and the width is relatively small. For each possible
mode of decay, there is a different probability of decay and, therefore,
a different partial width r, for each decay product. The total width r of an
energy level is then the sum of the individual partial widths;
r = r., + r. + r,, +. ••, (16-7)
where r .,, the radiation width, is a measure of the probability per unit
time for emission of a "f-ray, r ª is the a-particle width, r,, is the proton
width, and so on.
16-3) CROSS SECTIONS FOR NUCLEAR REACTIONS 457
The concept of level widths is useful because values of these widths can
often be obtained from measurements of resonances, as will be shown
later. When the total width is known, the value of the mean lifetime is
given by
h. (erg-sec)
T (sec ) = -'---=,-..:..
2rr (erg)
1.06 X 10- 27
- r (ev) x 1.60 x 10-12
6.6 X 10- 16
(16-8)
r (ev)
Level widths are usually given in ev; for a wide level, r may be of the
order of 10'' ev, as in the case of light nuclei, and the lifetime is 6.6 X 10-20
sec, while a sha.rp level with ar of 0.1 ev has the rela.tively long mean life
of 6.6 X 10- 15 sec. When the narrow level is that of a compound nucleus,
the lifetime of that level can be as long as 10- 15 sec or 10- 14 sec, as men-
tioned earlier in this section. Along with the level width, the level spac-
ing D, or mean distance between levels, can be obtained from resonance
measurements and is an important quantity in nuclear spectroscopy.
The level width and level spacing are useful not only for characterizing
compound nuclei, but for any excited nuclear states, and can be applied
to both bound and virtual levels.
The stndy of resonance phenomena in nuclear reactions can now be
seen to ha.ve great importance beca.use it provides information about the
width and spacing of the energy levels of the compound nucleus. The
dependence of the width and spacing on the mass number and the excita-
tion energy of the nucleus provides a test of theories of nuclear reactions
and nuclear models. The partial widths of a l~vel of the compound nucleus
give the relative probabilities for different modes of disintegration, and
these probabilities also yield information about nuclear structure.
16-3 Cross sections for nuclear reactions. To study nuclear reactions

in detail, it is necessary to ha.ve a quantitative measure of the probability
of a given nuclear reaction. This quantity must be one which can be
measured experimentally and calculated in such a way that the theoretical
and experimental values can be compared readily. The quantity that is
most often used for this purpose is the cross section of a nucleus for a par-
ticular reaction, usually denoted by <r with an appropriate subscript. The
term "cross section" has already been met in connection with the attenua-
tion of a 'Y-ray beam, when it was used as a measure of the probability
that a photon is removed from the beam; the cross section was regarded
there as a form of absorption coefficient. In the discussion of nuclear re-
Area = A
-
Projected area
of nucle11S = rW
Number of
nuclei = Nr
Surface density Nuclear
of nuclei = Nr/A density = N
= Nt
1--,-1
Fm. 16-2. Cross sections for nuclear reactions as geometrical areas
(Hughes02>).
actions, a more detailed consideration of the idea of cross sections will

prove useful. The concept of a nuclear cross section can be most easily
visualized as the cross-sectional area, or target area, presented by a nucleus
to an incident particle. lf the nuclei are considered as spheres of radius
R cm and the incident particles as point projectiles, then the target area,
or cross section u, of each nucleus is given by
(16-9)
A particle which passes normally through a thin sheet of material of area A

containing N T nuclei has a probabílity N TU/ A of colliding with a nucleus
provided that there is no "overlapping" of the nuclei, i.e., that N T'1' / A is
small (Fig. 16-2). The quantity NT/ A, which is the number of nuclei
per square centimeter or the surface density of nuclei, is equal to Nt, where
N is the number of nuclei per cubic centimeter and t is the thickness of the
sheet. For an incident beam containing n particles per cubic centimeter,
moving with velocity v, the number of particles passing through the sheet
is nv per square centimeter per second, and the collision rate can be ex-
pressed as
collisions per square centimeter per second = nv · N¡r = nvuNt;
the collision cross section is then
collisions per square centimeter per second
(/'=---~-~~-~---~---· nvNt
(16-10)
The quantity nv, which is the number of particles in the incident beam
crossing one square centimeter of area each second, is called the particle flux.
Equation (16-10) then shows that the cross section for collision is the
number of collisions per unit volume per second for unit incident flux and
unit nuclear density.
16-3] CROSS SECTIONS FOR NUCLEAR REACTIONS 459
The above discussion can be applied to any nuclear reaction, and the
term "collision" can be replaced by the appropriate term for any nuclear
process. Thus, u. may denote the cross section for scattering of a given
kind of particle, and may consist of two parts, <r., the cross section for
elastic scattering, and <r¡, the cross section for inelastic scattering. It is
possible to speak of a cross section <r a for the absorption of a particular
kind of particle, defined as the number of particles that are absorbed, or
disappear, per cubic centimeter per second foc unit incident flux and unit
nuclear density. There are also cross sections for individual reactions,
suih as <r(a, p), <r(a, n), and o-(p, a). The number of particles removed
from the beam is obtained by adding the numbers removed by all the
processes which can take place; particles can be removed by being ab-
sorbed in nuclear reactions or by being scattered out of the beam. In an
experiment, the number of events of a given kind (collisions, &eatterings,
absorptions, or other processes) is counted, and the quantities nv and N
are measured; the cross section for the particular process is then given by a
relationship of the type of Eq. (16-10). The collision cross section, which
corresponds to the effect of ali possible processes, is usually called the total
cross section, o-1 and is the sum of the cross sections for the individual reac-
tions. There is a direct analogy between 0-1 and the total absorption co-
efficient for 'i'-rays, expressed in units of square centimeter per atom, which
is the sum of a cross section for photoelectric absorption, a cross section
for Compton scattering, and a cross section for pair formation. As in the
case of 'Y-radiation, the total cross section can be found by measuring the
transmission of a particle beam in a "good geometry" experiment.
The expression for the number of processes of a given kind, say the ith
kind, contains the product No-¡. This product represents the cross section,
for the ith process, of all the atoms in a cubic c:ientimeter of material. It
is sometimes called the macroscO'f}'ic cross section and denoted by
it has the unit cm - 1 • ·

A rough idea of the magnitude of cross sections for nuclear reactions
can be obtained from Eq. (16-9) for the geometrical cross section of a
nucleus. In the discussion of the theory of a-decay, it was found that
the radii of a-emitting nuclei may be represented by the formula
R = 1.5A 113 X 10-13 cm,
where A is the atomic weight. If it is assumed that this formula may be

applied to all nuclei, which will be shown to be a reasonably good approxi-
mation, then
o- = 1rR2 = 1r(l.5)2 A 2/3 X 10-20 cm 2.
For a nucleus of intennediate mass, say A = 125,
<T = 1r(2.25)(25) X 10-26 cm 2 = 1.8 X 10- 24 cm 2,
and the geometrical cross section of a nucleus is of the order of 10- 24 cm 2.

Cross sections for nuclear reactions are often expressed in tenns of the
unit 10- 24 cm 2 , and this unit is called the barn, abbreviated as b.
Although it is convenient and simple to introduce the cross section as
a target area, and to get a rough idea of its magnitude by calculating
the geometrical cross section, this procedure must not be taken seriously.
The experimental meaning of the cross section comes from its use as a
measure of the number of nuclear events which occur under a given set
of experimental conditions, as expressed by Eq. (16-10) and the subse-
quent discussion. Nuclear cross sections are found to have values ranging
from small fractions of a barn to hundreds of thousands of barns, and
these values often differ greatly from the geometrical cross section. A
given nucleus can have widely different cross sections for different nuclear
reactions and the values represent the relative probabilities of those re-
actions. Under certain special conditions scattering cross sections and
geometrical cross sections can be related directly, and the measured values
of the scattering cross sections are then used to determine the values of
the nuclear radii. The elastic scattering of neutrons with energy greater
than 10 Mev is a case in point, and will be discussed in Section 16-5E.
The results of an experimental study of a nuclear reaction can be ex-
pressed in terms of the number of processes which take place under the
conditions of the experiment or in terms of the cross section for the re-
action. The advantage of using the cross section lies in the fact that its
value is independent of the flux of incident particles and th~ density of
the material used for the target, and these quantities are not essential to
the nuclear reaction itself. When the cross section for a particular reaction
between a given particle and a given nuclide is known, it is possible to
predict the number of reactions that will take place when a sample of
the nuclide is exposed to a known flux of those particles for a given length
of time. This kind of information is needed in many practical problems,
as in the manufacture of artificial radionuclides, and cross sections are a
valuable form of nuclear data. In many cases, the results of the study
of a nuclear reaction are expressed directly in terms of the number of
processes that take place, or of the number of outgoing particles. If the
main interest is in the determination of the energies of the different groups
of particles emitted, it is enough to measure the energies by measuring
the ranges or by magnetic deflection, without knowing exactly how many
processes have taken place. The relative intensities of different groups
of product particles can also be roeasured by roethods which are inde-
16-3) CROSS SECTIONS FOR NUCLEAR REACTIONS 461
pendent of the number of processes, as can the energies of resonances in

nuclear reactions.
The concepts of cross section and level width can be applied to resonances
in a quantitative way. The probability of the reaction X(x, y) Y may be
denoted by the cross section a-(x, y). According to the two-step view of
nuclear reactions,
a-(x, y) = a-c(x) • (relative probability of the emission of y), (16-11)
where a-c(x) is the cross section for the formation of the compound nucleus
in a collision between the particle x and the target nucleus X. The rela-
tive probability of the emission of the particle (or "Y-ray) y is just rv/r,
where r" is the partial level width for y and r is the total level width.
Then
a-(x, y) = O"c(x) ~· (16-12)
In general, the values of the cross sections and level widths depend on the
energy of the incident particle, and on the charge and mass of the target
nucleus. One of the problema of nuclear theory is the calculation of
pc(x), r, rv, and therefore of a-(x, y). In the particularly important case
of resonance processes, a theoretical formula for the cross section was
derived by Breit and Wigner. 19 • 10> The rigorous derivation of the Breit-
Wigner formula is a difficult problem, but the physical meaning of the
formula can be discussed qualitatively. In its simplest form, the Breit-
Wigner formula gives the value of the cross section in the neighborhood
of a single resonance level formed by an incident particle with zero angular
momentum. Under these conditions, the formula is
( ) X2 r~rv
a- x, Y = 41r (E - E 0 ) 2 +
(r/2)2 (16-13)
where >. is the de Broglie wavelength of the incident particle defined in

Section 7-8, E is the energy, E 0 is the energy at the peak of the resonanc~,
and the r's are the widths already defined. Equation (16-13) may be
regarded as containing three factors. The first is a measure of the prob-
ability of forming a compound nucleus and, according to wave mechanics,
is proportional to >. 2• The second factor,
1
(E - E 0) 2 + (r/2)2'
is the mathematical expression for the resonance property and may be
called the resonance factor. The denominator of this factor has its smallest
value when E = E 0 , and the cross section then has its greatest value.
As the value of E departs from E 0 , the denominator increases in value,
and the resonance factor and cross section decrease. The third factor is
the probability for definite types of disintegration of the compound
nucleus and is expressed by the partial widths r"' and r 11 •
The Breit-Wigner formula is an example of the application of wave
mechanics to nuclear physics, and it is typical of such applications that
the wavelength of the incident particle appears in the formula. The
de Broglie wavelength of a particle is defined as
h
X=-, (16-14)
mv
where h is Planck's constant and mv is the momentum of the particle.
It is convenient to express X in terms of the kinetic energy rather than
the momentum and, since E = ½nz.v 2 ,
h
X=--• (16-15)
V2mE
An idea of the magnitudes of X and u may be obtained by considering the
incident particle to be a neutron; when the numerical values of h and m
are inserted, and E is expressed in electron volts, the result is
9
X (cm) = 2.87 X 10- •
(16-16)
v'E (ev)
A neutron with a kinetic energy of 104 ev has a wavelength of 2.87 X 10- 11
cm, while a 1-Mev neutron has a wavelength of 2.87 X 10- 12 cm. Since
u(x, y) is proportional to X2 , its value should be of an order of magnitude
similar to that obtained from the geometrical estima.te in this section.
I ts actual value in any particular case will, of course, depend also on the
other qua.ntities in Eq. (16-13). It is found that the values of resonance
cross sections vary over a wide range, from fractions of a barn to the order
of 10 6 barns.
Values of u 0 (x, y) (the cross section at the peak of the resonance), E 0 ,
and r can be obtained from experiments in which the pa.rticle transmission
through the target is measured as a function of the energy of the incident
particles in the neighborhood of the resonance energy. The values of u 0
and E 0 can be deduced from the t~nsmission curve. lt can be shown
from Eq. (16-13) that the "full width at half-maximum," that is, the
energy interval between the two points at which the cross section is u 0 /2,
is just the total width r. The total width can thus be determined directly
from the actual observed width of the cross secti.on resonance. When the
total width is known, the values of the partial widths can be obtained by
further experimentEI, or may sometimes be inferred with the aid of certain
theoretical considera tions.
HH] EXPERIMENTAL RESULTS. GENERAL CONSIDERATIONS 463
The rigorous form of the Breit-Wigner formula is considerably more

complicated than Eq. (16-13). It contains an additional factor, of order
unity, which depends on the angular momenta of the particles involved
in the reaction. It also takes into account the fact that when two or more
resonances are close together, the simple "one-level formula" (16-13) is
no longer correct because the resonances may "overlap" and affect each
other. Fortunately, for many practica! purposes, the one-level formula is
adequate.
16-4 Experimental results. General considerations. To discuss the

experimental information about nuclear reactions in an orderly way, it
is necessary to classify the reactions in sorne consistent scheme. Blatt
and Weisskopr< 11 > have shown that nuclear reactions can be classified
into different groups characterized by the nature of the incident particle,
its energy, and the atomic weight of the target nucleus. Their classifica-
tion and general treatment have proved successful, and will be followed
during the remainder of this chapter.
The incident particles to be considered include neutrons, protons,
a-particles, and deuterons; reactions initiated by 1'-rays will also be
treated.
The range of energy E of the incident particle may be divided roughly
into five regions:
I. Low energies: O < E < 1000 ev.
II. Intermediate energies: 1 kev < E < 500 kev.
III. High energies: 0.5 Mev < E < 10 Mev.
IV. Very high energies: 10 Mev < E < 50 Mev.
V. Ultrahigh energies: E > 50 Mev.
The target nuclei may be divided into three groups according to the values
of the mass number A :
A. Light nuclei: 1 < A < 25.
B. Intermediate nuclei: 25 < A < 80.
C. Heavy nuclei: 80 < A < 250.
This classifica.tion of nuclear reactions serves to separate them according
to their general character, the particles emitted, and their energy distribu-
tion, as well as other properties. Although the distinctions between the
different groups are not clear-cut, they are sufficiently marked to make
the experimental results fall into definite patterns.
It is not surprising that reactions initiated by neutrons have, in general,
different characteristics from those initiated by charged particles. The
neutron, because of its lack of charge, can penetrate relatively easily into
a nucleus, and the probability of forming a compound nucleus is relatively
TABLE 16-3
REACTIONS WITH lNTERMEDIATE AND liEAVY NUCLEI
(Reprinted by permission írom Theoretirol Nuclear PhyBica by J. M. Blatt ~
and V. F. Weisskonf. 1952. John Wiley and Sons. Inc.)
Tbia table lista the nuclear reactions occurring in each group. The symbola 1iated refer to the emerging particle in a reaction characterised by the type
of target, the type of incident particle (columna), and the energy range (roWII). The order of ayrnbola in each group corresponda roughly to the order of
the yielda of the com,eponding reactions.. Reactions whooe yield is uaually 1- than about 10-• of the leading one an, omitted.
Abbreviations: el - elaatic, inel - inelaatic, ree = reeonances. The abbreviation (ree) refera to ali reactioo. liated in th" box. The elutic 1e&ttering
ia
of cba.rged particlee omitted, IÍllce it cannot euily be aeparated from the non-nuclear Coulomb 1e&ttering. Fiaaion ÍII a1ao omitted, oince it occura only
with a few of the heavieet elcmenta.
Intermediate nuclei Heavy nuclei

Incident particle n p d n p a d
" :;:
Energy of incident particle e:
l. Low: 0-1 kev n (el)
,,
,,
(rea)
n (el)
(ree)
G
::o
II. Intennediate: 1-600 kev n (el) n n p n (el) ::o
,, ,, ,, n 'Y ~
(rea) a p (ree) ~
(ree) (ree) o
111. Hi¡h: 0.!>-10 Mev n (el) n n p n (el) n n p ~
n (inel) p (inel) p n n (inel) p (inel) p n
a a a (inel) pn p 'Y 'Y pn
p (ree for (rea for 2n 'Y 2n
(rea for lower lower
lower energiee) energiee)
energiee)
IV. Very high: lo--60 Mev 2n 2n 2n p 2n 2n 2n p
n (inel) n n 2n n (inel) n n 2n
n (el) p (inel) p pn n (el) p (inel) p np
p
np
2p
a
np
2p
a
three or
np
2p
a (inel)
three or
3n
d (inel)
tritona
three or
p
pn
2p
a
np
2p
a
three or
np
2p
a (inel)
three or
3n
d (inel)
tritona
three or
!....
Q)
three or more more more three or more more more
more particlee particlee particlee more particlee particlea particlea
particlea .. particlee
'
16-5] NEUTRON-INDUCED REACTIONS 465
high. Positively charged particles must overcome the repulsive Coulomb

forces exerted by the positively charged nucleus; they must either pene-
trate the Coulomb barrier, as discussed in Chapter 13, or they must have
enough energy to go over the top of the barrier. Consequently, a neutron
can cause a reaction when it has a very low kinetic energy, while charged
particles need much higher energies, and many more neutron reactions
have been observed than charged particle reactions.
The light nuclei (group A) are characterized by the fact that it is ex-
tremely difficult to describe nuclear reactions in that group in terms of
general rules. The character and yield of most of the reactions seem to
depend strongly on the structure of the individual nucleus; and separation
energies and Q-values differ by much larger amounts among the light
elements than within any other group. The reactions with intermediate
or heavy nuclei are of different character in the different energy regions.
For example, charged particles with energies less than 500 kev (corre-
sponding to the energy groups I and II) cannot penetrate to the nucleus
in groups B and C, and neutron reactions are practically the only ones
observed. In region I (low energy) resonance capture of neutrons by heavy
nuclei predominates, while in energy region II the most important reaction
is the resonance elastic scattering of neutrons. In energy region III charged
particle reactions become important; the residual nucleus can be left in
severa! excited states and the emitted particles can form groups with
different energies. In region IV the energy is high enough for reactions
in which more than one particle is emitted for each incident particle
[(p, 2n), (p, np), etc.]. Finally, in region V, the energy of the incident
particle is so high that many nucleons can be ejected by a single incident
particle, and the theory of the compound nucleus breaks down.
In Table 16-3, taken from Blatt and Weisskopf, (JO> the different re-
action types are arranged according to the groups in which they appear.
Reactions with light nuclei are not listed because, as indicated above, they
cannot be fitted into any general scheme. The reactions in each group are
listed in the order of their importance. In general, the reaction which has
the larger cross section precedes the one with lower yield and, although
there are exceptions to this scheme, it is a useful one.
In the following sections of this chapter, the different groups of reactions
will be discussed in a general way and illustrated by specific examples.
The emphasis will be on the use of nuclear reactions as sources of informa-
tion about nuclei rather than on the experimental techniques or the theo-
retical details. The latter can be found in the literature references.
16-5 Neutron-induced reactions. Reactions induced by neutrons

will be discussed first beca use of their great importan ce in both "pure"
and "applied" nuclear physics. Experimental methods have been devel-
466 NUCLEAR REACTIONS [CHA.P. 16
12
Aluminwn
11
10
9
J
8
7
I ,~
1 Total cross section
6
5
4
,
I '
\ 'V
~
\
\
/\
f 1 f\
A
1\ f "'r '-
I ., \ ..... I \ - J
3
2 .
"--j
I \, \.
..... f ., ... V r\.?
j
V
1
o
Wt---+--~15--+--+----+---1---+---4--+---- +--1
Neutron energy (kev)

F10. 16-3. Total and capture cross sections of aluminum for neutrons in the
intermediate energy region (Henkel and Barscha.110 3>).
oped for obtaining neutrons of different energies, for detecting them, and
for measuring the cross sections of the different neutron-induced reactions.
These methods form a highly specialized branch of nuclear physics and
have been treated in detail by Hughes/ 12 > they will be discussed in Chap-
ter 18, which is devoted to the subject of neutron physics. For the purposes
of the present -::hapter, it will simply be stated that beams of monoener-
getic neutrons can be obtained at energies ranging from small fractions
(~10-4 ) oí an electron volt to many millions of electron volts, and that
the cross sections or yields of the different reactions induced by these
neutrons can be measured by means of appropriate devices.
A~ LO'W and intermediate e:nergy, intermediate nuclei. The most impor-
tañt reactions between neutrons of low or iiitermediáte energy and nuclei
of intermedia.te atomic weight are elastic scattering (n, n), and radiative
capture (n, 'Y). The bombardment'of AI 27 by neutrons provides an inter-
esting example of these reactions. The total cross section u 1 and the
16-5) NEUTRON-INDUCED REACTIONS 467
radiative capture cross section u(n, 'Y) have been measureJ as functions of
the incident neutron energy. The total cross section was obtained from
transmission measurements. The capture cross section was obtained by
making use of the fact that the product Al 28 is an electron emitter and its
yield can be determined by measuring the induced radioactivity. Sorne
of the results 03 > are shown in Fig. 16-3, in which the total and radiative
capture cross sections are plotted as functions of energy in the range from
10 to about 500 kev. The values of the total cross section lie between
1 and 12 X 10-24 cm 2, while the capture cross sections are of the order of
10- 27 cm 2 , or three orders of magnitude smaller. Since the only reactions
that have been observed when Al 27 is bombarded with intermediate-
energy neutrons are elastic scattering and radiative capture, the total
cross section is practically equal to the elastic scattering cross section. The
capture cross section shows resonances at the same energies as the total
cross section. This result would be expected according to the compound
nucleus picture and serves as a confirmation of that picture. It follows
from the values of the cross section, and from the Breit-Wigner formula
that the scattering or neutron width r nis about 1000 times as great as the
radiation width r 'Y' The experimental results show that the neutron
widths are in the range 5-20 kev, so that the radiation widths are of the
order of 10 ev.
Over twenty resonances have been found for neutron energies in the
range from 0.009 Mev to 0.8 Mev and, since Al has only one stable isotope,
they must correspond to excited levels of Al 28 . They are listed in Table
16-4, together with the corresponding excited levels of the compound
nucleus [Al 28) computed as shown in Section 16-2. The binding energy
of the neutron in this nucleus is 7.724 Mev, as calculated from the difference
mass of Al 27 + mass of neutron - mass of Al 28 •
The observed level spacing for Al 28 , the distance between neighboring
resonances, varies from 10 kev to 90 kev, with an average value of about
40 kev.
The scattering cross section has an appreciable value between reso-
nances, being still of the order of 10-24 cm 2 • At neutron energies where
there is no resonance, this cross section represents the probability that the
neutrons are scattered without the formation of a compound nucleus. In
this case, the nucleus acts like a hard sphere of radius R, and it has beén
shown that for neutrons with energies up to about 1 Mev the scattering
cross section is
u,= 41rR 2 •
This type of scattering is called potential scattering, as distinct from reso-
nance scattering, because the force between tlie nucleus and the neutrons
TABLE 16-4
LEVELS OF THE CoMPOUND NucLEUS [AI 28 ] ExcITED
IN THE BOMBARDMENT OF ALUMINUM BY N EUTRONS.
Neutron Energy to be Excited level of

resonance added to the [AI 28 ] = 7.724 +
energy
energy, binding energy, in second column,
Mev Mev Mev
0.0091 0.0088 7.733

0.040 0.039 7.763
0.095 0.092 7.816
0.145 0.140 7.864
0.155 0.149 7.873
0.210 0.203 7.927
0.225 0.217 7.941
0.265 0.256 7.980
0.290 0.280 8.004
0.315 0.304 8.028
0.370 0.357 8.081
0.425 0.410 8.134
0.445 0.429 8.153
0.480 0.463 8.187
0.500 0.482 8.206
0.530 0.511 8.235
0.570 0.550 8.274
0.620 0.598 8.322
0.650 0.627 8.351
0.705 0.680 8.404
0.795 0.767 8.491
can be described mathematically by a potential function analogous to the

Coulomb potential in the scattering oí a-particles by nuclei. The potential
which describes the neutron scattering is not electrical in nature, but is a
characteristically nuclear potential.
One part oí the low energy region I, the so-called thermal energy region,
has a special property in the case oí the (n, 'Y) reaction. A beam
oí thermal neutrons can be obtained from a neutron-producing apparatus
in which the neutrons are slowed down in sorne medium and emerge with
energies in equilibrium with those oí the thermal motion oí the atoms oí
the medium. At room temperature, the thermal neutron energy region
is in the neighborhood oí 0.025 ev. Although the Breit-Wigner formula
is strictly valid only near a resonance, it may be applied in the thermal
energy region in the absence oí a resonance ií it is assumed that the energy
16-5] NEUTRON-INDUCED REACTIONS 469
E O is the resonance energy nearest to the thermal region. It turns out

that ali of the factors in the Breit-Wigner formula are constant compared
with r 0 , which is proportional to the neutron velocity v, and >., which is
proportional to 1/v. The formula for the (n, 'Y) reaction then reduces to
constant
tT (n, 'Y ) = V ' (16-17)
and the cross section for radiative capture is inversely proportional to

the neutron velocity. In the thermal energy region, the radiative capture
of neutrons by aluminum is described by the "1/v law"; at a neutron
energy of 0.025 ev, the value of the (n, 'Y) cross section is 0.24 X 10- 24 cm 2,
and is much greater than the values in the resonance region.
B. Low-energy, heavy nuclei. The only reactions possible for low-energy
neutrons on heavy nuclei are, -with a few exceptions, elastic scattering
and radiative capture. When the positive charge of the nucleus is large,
the effect of the Coulomb barrier prohibiting the emission of charged par-
ticles of low energy is even greater than with intermediate nuclei. In
reactions between low-energy neutrons and heavy nuclei, the cross sections
often show resonances very close together in energy. The level distances
are often of the order of 10 to 100 ev and the excited states of the com-
pound nuclei in this region are usually close together. There are sorne
important exceptions to this rule. In certain nuclides, no resonances have
been found, and in others, especially lead and bismuth, the spacing between
resonances is very large (many kev). These nuclides are now thought
to have a special structure interpreted in terms of the shell model of
nuclear structure, which will be discussed in the next chapter.
An example of a heavy nucleus with resonances very close together is
given in Fig. 16-4, which shows the variation of the total cross section
of Ag 109 in the energy range from one ev to about 5000 ev.< 14 > At the
higher energies, the resonances cannot be resolved well because of ex-
perimental difficulties; the energy resolution which can be achieved is
shown by the triangles. In heavy elements such as silver, the resonances
are mainly capture rather than scattering resonances, and the neutron
width r O is small compared with the radiation width r -r· The total cross
section is practically equal to the capture cross section u(n, 'Y) near reson-
ance; between resonances, the total cross section is approximately equal
to the (potential) scattering cross section 41rR 2 . The total cross section< 14 >
of an exceptional nucleus Bi 209 , between one ev and 10,000 ev, is shown
in Fig. 16-5; only two widely spaced resonances are seen, and the curve
is very different in character from the curve for Ag 109 .
Neutron capture resonances have been analyzed extensively, and have
been used to test the Breit-Wigner formula. Figure 16-6 shows a low-
energy capture resonance in cadmium¡< 15 > the solid line was calculated
~
,1Ag ~
(ev)
z
~
o
t<
.-,(b) f;!
l;,:l
l;,:l
f;!
~
~
r,,
~
....
O>
Fm. 16-4. The total neutron cross section of silver in the low-energy region.0 4>
._,.
100
....
a3Bi
(ev) :f t kr l·i 1¡UIIH ;· J , J :l·l.l Il 11 · { 1 111 l!H !, lfflll!M :. 9'
~
-1
_,
..'º·º zt,J
..
-1
C'.1
>,3
l;é
. o
1 , l 1: 1:LI 11-l lc•tiJ.~:lii~ i:~ I O't(b} ....~z

u,(b) ~I ·· K. ~; le JJ 1 li 1 11U .
1 1 1 · 1 · 1 1 1 1 111
1
t:,
. C'.1
n
t,J
t:,
l;é
._,.'·º t,J
>
~
.. o
z
Ul
◄
.
.
4
Fm. 16-5. The total neutron cross section of bismuth in the low-energy region.0 >
~
....
10000
,41Cd
~
6" (slowl
}{
f I;
~ - .....,_ ..,lo
~f 'l
""-.
\
IOOO
4'-• rtOOlt
'6•0.IN•w
' -
000
1
f•0.1115
\
l
u,{b)
utlb)
\" ¡
o Co
', \ 100
~z
lJl
100
-¡:;-
~
o
10 '
10
00001 4001 0,01 º' ....
Elev) Q:,
5
Fxo. 16--6. The low-energy reeonance in cadmium (Rainwater et al.0 >).
1&-5] NEUTRON-INDUCED REACTIONS 473
from the one-level formula (16-13) with the constants o- 0 = 7200 b, Eo =

0.176 ev, and r = 0.115 ev, and the circles represent the experimental
points. The agreement between the formula and the experimental results
is excellent and provides a quantitative proof of the validity of the formula.
At thermal energies, in the absence of resonances, o-(n, 'Y) follows the
1/v law, as was the case with intermediate nuclei.
C. Intermediate-energy, heavy nuclei. In the intermediate-energy region,
the reactions are similar in nature to those with intermediate nuclei.
Resonance scattering is more important than resonance capture and
r n > r -r· The spacing between resonances, however, is smaller than for
intermediate nuclei. It is hard to explore the intermediate-energy region
because of experimental difficulties with neutrons of these energies. In-
dividual resonances cannot be resolved and the experiments give informa-
tion about cross sections averaged over many resonances. These cross
sections in turn yield information about the average level spacing. When
the Breit-Wigner formula (16-13) is averaged over many resonance levels
and applied to (n, 'Y) reactions at neutron energies of about 1 Mev, it is
found that the averaged (n, 'Y) cross section is given by
o-(n, 'Y) = (>..2 /2)(r-r/ D), (16-18)
where D is the level spacing. If o-(n, 'Y) is measured, and if r -r is calculated

theoretically, values of the level spacing can be obtained. The radiative
capture cross section has been measured for a large number of nuclides
by Hughes< 15i for 1-Mev neutrons. In the scheme of Section 16-4, this
energy lies in region III (high energy) rather than in region II (inter-
mediate). These reactions will be discussed here because they are a
borderline or overlapping case which could be discussed in either place,
but it is more convenient to discuss them here. The measured values
of the capture cross section are plotted against the mass number of the
target nucleus in Fig. 16-7; the cross sections are given in units of milli-
barns (thousandths of a barn). As the mass number increases, the cap-
ture cross section increases until A is about 100. The radiation width
r -r decreases, according to theory, but less rapidly than the cross section
increases. The result is that the level spacing D, from Eq. (16-18), de-
creases quite rapidly as the number of particles in the nucleus increases.
At higher values of the mass number, the cross section does not seem to
behave regularly; but if the cross section is plotted against the number
of neutrons in the nucleus, as in Fig. 16-8, it is evident that the irregu-
larities are caused by sharp drops in the value of the cross section that
occur at neutron numbers of 50, 82, and 126. These exceptionally low
values of the cross section correspond to unusually large leve! spacings
such as those mentioned previously for lead and bismuth, both of which
100.0
10.0
a{mbl
1.0
0'---2~0,--""40,,_....,60,,,.......,80~-,oo=-=--""12~0,......,l40~-l60-"""1ao!:'.:--=200:==--2~20"='
A
Fm. 16-7. Neutron capture cross sections atan eft'eetive energy of 1 Mev
(Hughes0 2>).
.Jn
.Ag
100.0
.Nd
.Nd
IQO
a{mb)
Ceolo
'Ce Bí •
• Bo Pb,
.x.
Na
10 20 30 40 50 60 70 80 90 100 IIO 120 130

Number of neumons
FIG. 16-8. Neutron capture cross sectiolll! at l Mev as a function of neutron
number (HughesCl 2 ').
16-5) NEUTRON-INDUCED REACTIONS 475
have 126 neutrons. The values of the level spacing obtained in this way
vary from about 5 X 104 ev for Na 23 to about 1 ev for the ordinary heavy
nuclides. The nuclides with 50, 82, or 126 neutrons have much greater
level spacings than neighboring nuclides, and it will be seen in the next
chapter that these neutron numbers seem to be associated with particular
nuclear structures.
D. High-energy, heavy and intermediate nuclei. When the energy of
the incident neutron is about 1 Mev or higher, new types of processes can
occur, such as inelastic scattering and the emission of charged particles.
The emission of neutrons is more probable than that of charged particles
because of the Coulomb barrier. The emitted neutrons can leave the
residual nucleus in excited states so that sorne of the scattering is inelastic
rather than elastic. The excited states, which are bound states, decay
by 'Y-emission, and information about them can be obtained from the
energies of the 'Y-rays and from the energy distribution of the scattered
neutrons. The determination of inelastic neutron scattering cross sec-
tions is a difficult experimental problem, and only a relatively small amount
of information has been obtained in this way. The cross sections for the
reactions in which charged particles are emitted are much smaller than
those for inelastic scattering and radiative capture, especially when the
target nucleus has a high value of Z, because of the effect of the Coulomb
barrier.
E. Very high energy, heavy and intermediate nuclei. The scattering of
neutrons with energies greater than 10 Mev has been used to determine
values of the nuclear radius. <17 • 18 > According to theory, the total cross
section at high energies approaches the value
(16-19)
so that if u, is determined by transmission measurements, R can be ob-

tained. The rigorous relationship between u, and R is more complicated
than the last equation.<1 9> The formula for u 1 actually includes a term
which depends on the energy of the neutrons, and the value of the cross
section given by Eq. (16-19) is an asymptotic one which is reached only
for extremely high energies. The results of two sets of experiments, one
with 14-Mev neutrons, the other with 25-Mev neutrons, are listed in
Table 16-5; the values of R were obtained from the rigorous formula.
The quantity Ro is defined by the relationship
R = R 0 A 113 X 10- 13 cm, (16-20)
and should be a constant if the nuclear radius is proportional to the cube
root of the nuclear mass. The results show that Ro is indeed practically
constant, and the values of nuclear radii conform with the formula (16-20)
TABLE 16-5
NUCLEAR RADII FROM NEUTRON CROSS SECTIONS
AT HIGH ENERGIES
Energy, Observed <11, Nuclear radius, R,

Element Ro
Mev 10- 24 cm 2 10- 13 cm
Be 14 0.65 2.4 1.17

B 14 1.16 3.4 1.54
e 25 1.29 3.8 1.65
o 25 1.60 4.3 1.71
Al 14 1.92 4.6 1.53
25 1.85 4.6 1.52
s 14 1.58 4.1 1.30
Cl 25 1.88 4.7 1.44
Fe 14 2.75 5.6 1.46
Cu 25 2.50 5.5 1.38
Zn 14 3.03 5.9 1.48
Se 14 3.35 6.3 1.46
Ag 14 3.82 6.8 1.44
25 3.70 6.9 1.46
Cd 14 4.25 7.2 1.48
Sn 14 4.52 7.4 1.52
Sb 14 4.25 7.3 1.46
Au 14 4.68 7.5 1.33
Hg 14 5.64 8.3 1.42
25 5.25 8.4 1.44
Pb 14 5.05 7.8 1.32
Bi 14 5.17 7.9 1.34
over the entire range of the elements, with the possible exception of the
very lightest ones. The average value of Ro is close to 1.4 and agrees
reasonably well with the value 1.53 obtained from a-decay data in Section
13-6. The two results should not be expected to be the same because
they are not the results of direct measurements of the nuclear radius.
The quantity R, as it appears in the theory, is the distance at which nuclear
forces begin to act. These forces should act differently on a neutron
approaching the nucleus and an a-particle leaving it, and it is not surpris-
ing that different values are obtained for nuclear radii from the analysis
of different kinds of experiments. The fact that the values obtained by
different indirect methods for the dimensions of atomic nuclei agree as
well as they do represents a triumph for nuclear theory.
Values of nuclear radii have also been obtained recently from studies of
the scattering of neutrons with energies in the neighborhood of 1 kev. <2 oi
16-6] REACTIONS INDUCED BY PROTONS AND ALPHA-PARTICLES 477
The interpretation of the experimental results is more complicated than

at the higher energies, but the results are similar, with Ro = 1.35.
In the very high energy region, the incident particle has enough energy
to cause the emission of more than one particle from the compound nucleus,
and (n, 2n), (n, np), (n, 3n), etc., reactions, may be expected. Again,
because of the Coulomb barrier, the emission of neutrons is more likely
than that of charged particles. The (n, 2n) reaction has been studied in
a number of cases,< 21 > and its threshold is usually in the neighborhood
of 10 Mev. Above the threshold, the cross section increases with energy, <22 >
reaching a maximum somewhere near 20 Mev, and then decreases with
increasing energy of the incident neutrons.
16-6 Reactions induced by protons and alpha-particles. A. High energy.

The cross sections of nuclear reactions induced by protons or a-particles
are extremely small for particle energies below 0.1 Mev except for a few
cases among the very light nuclei. At low or intermediate energies, the
Coulomb barrier prevents any appreciable interaction between positively
charged particles and nuclei. Wheri. the energy of the incident particles is
high enough, protons and a-particles may be scattered elastically or in-
elastically; they may undergo radiative capture, or they may cause the
emission of a charged particle. The (a, p} reaction, the first one studied
in detail, has already been discussed in Section 16-1. The properties of
reactions induced by protons and a-particles can be illustrated by the
reactions of protons with aluminum. The reaction Al 27 (p, 'Y)Si 28 yields
information about the excited levels of the compound nucleus [Si 28 ]. A
large number of very sharp resonances has been found, <23 > sorne of which
are shown in Fig. 16-9. In these experiments, an aluminum target was
bombarded with protons accelerated in a Van de Graaff electrostatic
generator, and the 'Y-ray yield was measured as a function of the proton
energy. The sharpness of the resonances (the low value of the total width)
shows that the lifetimes of the excited states are relatively long. About
100 resonances have been found at proton energies between 0.2 Mev and
4 Mev. <24 > The binding energy of the proton as obtained from the mass
difference, Al 27 + p - Si 28 , is 11.59 Mev, and the excited Ievels Iie
between 11.7 and 15.6 Mev.
The reaction Al 27 {p, a)Mg 24 has also been studied and more than 30
resonances have been observed< 24 , 25 > in the energy range from 1.4 Mev to
4 Mev. These resonances also give information about the excited levels
of [Si 28 ] and should occur at the same energies as those of the (p, 'Y)· re-
action. The elastic scattering of protons by aluminum also shows reso-
nances and, with few exceptions, at every resonance obtained with the
(p, 'Y) or (p, a) reactions, there is an elastic scattering resonance. This
result is expected because, according to the compound nucleus theory,
6
l5l 3
tü
5
4
::
\ ! . L 1 1 . lr..i
.. ~
500 600 700 800 900 1000 1100 1200 1300 1400 kev
Proton energy
Fm. 16-9. Yield of the (p, -Y) reaction in aluminum, as a function of the
proton energy (Brostrom, et al. <23 >).
To r28 1 hhc
'
22
1
20
1
1
-
.o 18 -~ ~-~ ---
f~
16
14
i
·a.., 12 -~
8.
:, lO --- ----- ---
o::
u5 8
(\ 111 1f
6
4
f V1
M
1
V
\ ,,,.A.
2
n 1A /1 1 . •
'.l\J ' ,..._,; ~¡v"- ..,J
" ' ""'

\J
1820 1840 1860 1880 1900 1920
Photon energy (kev)
Fm. 16-10. Yield of the (p, n) reaction in Cl 37, as a function of the proton
energy (Schoenfeld et al.< 28>).
l&-6] REACTIONS INDUCED BY PROTONS AND ALPHA-PARTICLES 479
any state of the compound nucleus formed by proton capture must be

capable of decay by emission of a proton with the full energy available.
The inelastic scattering of protons by aluminum shows resonances and
gives information about excited states of the compound nucleus [Si 28 ] ¡
it also provides an important method for determining the level spectrum
of the target nucleus Al 2 7 • This reaction may be described by the equations
13Al
27
+ 1H1 --+ [uSi 28] --+ 13Al27* + 1H 1 ,
27
13Al * --+ 13Al
27
+ "Y,
where the asterisk denotes an excited state of Al 27 . The method generally
used for studying the inelastic scattering of protons is to bombard the
scatterer with a collimated beam of monoenergetic protons, and then, by
means of range or magnetic deflection measurements, to determine the
energy distribution of the protons scattered through a known angle. The
scattered protons usually appear in several groups with different energies,
and the corresponding nuclear energy levels can be obtained from the
equations of conservation of energy and momentum. In this way, Al 27
has been found to have more than 20 excited levels at energies between
0.84 Mev and 5.95 Mev above the ground state. <25 > They are bound
levels and decay by "Y-emission.
When the energy of the incident protons or a-particles is high enough,
a neutron may be emitted. The (p, n) and (a, n) reactions are threshold
reactions; if the target nucleus is stable, the threshold proton energy for
neutron emission is always greater than 782 kev, the neutron-proton mass
difference. One effect of the Coulomb barrier is to increase the threshold
to still higher values. Below the threshold for neutron emission, radia-
tive capture, scattering, and the emission of charged particles are impor-
tant. Above the threshold, the relative importance of neutron emission
increases with increasing energy of the bombarding particle, and when
the compound nucleus has enough excitation energy to emit a neutron
of 1 Mev or more, neutron emission becomes the dominant reaction.
Threshold energies have been determined< 27> for a number of (p, n)
reactions among the light and intermediate nuclides. In these experi-
ments, protons were usually accelerated in an electrostatic generator and
the neutron yield was measured. In the case of Cl 37 , the threshold for
the reaction Cl 37 (p, n)A 37 is at 1.64 Mev. Between this energy and 2.51
Mev over 100 resonances have been founct< 28 > corresponding to excited
states of the compound nucleus [A 38]. A sample of the experimental
results in the range from 1.80 Mev to 1.94 Mev is shown in Fig. 16-10.
The resonances were resolved well enough so that an average separation
between levels of about 5 kev was observed. Similarly, in the reaction
Cr 53 (p, n)Mn 53 , 261 resonances have been observed, <29 J corresponding
to excited levels of the compound nucleus [Mn 54 ], for proton energies

between 1.42 Mev and 2.47 Mev, with an average spacing of 3.9 kev.
These results are an indication of the complexity of nuclear spectra
and of the great difficulty of accounting in detail for these spectra on the
basis of a theory of nuclear structure. One method of avoiding this problem
is to develop an approximate theory which does not attempt to reproduce
individual resonances but simply tries to describe the general trend of the
magnitude and energy dependence of nuclear cross sections on the basis
of a very rough picture of nuclear structure. A schematic theory of this
kind has been proposed by Feshbach and Weisskopf 09 > and has been
applied to the (p, n) and (a, n) reactions, among others. When the energy
of the incident particle is high enough, the emission of a neutron is the
dominant reaction, and the other reactions are negligible in comparison.
The approximate theory can then be applied to the calculation of the cross
sections averaged over many resonances, giving the values for the cross
sections for the formation of the compound nuclei. If the latter are de-
noted by <Tc(p) and <Tc(a), then
<T(p, n) ~ <Tc(p); <T(a, n) ~ <Tc(a).

When the energy levels are very close together, as in the heavier-inter-
mediate and heavy nuclei, it is very difficult to resolve the resonances,
and experimental measurements give values of <T(p, n) and <T(a, n) aver-
aged over many resonances. If the resolution is sufficiently poor, the
detailed resonance structure is not seen, and curves such as those in Fig.
16-11 are obtained. The circles represent experimental points for the
reactions Ag 1 º7 (p, n)Cd 1 º7 and Ag 1 º9 (p, n)Cd 109 , and the dashed curves
are theoretical results; the latter depend on the val ue chosen for the
nuclear radius. Good agreement was obtained when the formula (16-20)
was used for the radius with Ro = 1.5. This agreement is typical of results
obtained for (p, n) and (a, n) reactions on a series of nuclides; in nearly
every case, the experimental results could be accounted for by the theory
provided that a value of Ro between 1.3 and 1.5 was used. cao> These values
of Ro are in good agreement with the values obtained from a-decay and
neutron scattering so that the application of the approximate theory of
Feshbach and Weisskopf to (p, n) and (a, n) reactions seems to be suc-
cessful.
B. V ery high energy. When the energy of the incident particles is greater
than 10 Mev, more than one particle can be emitted. The compound
nucleus decays most often by emitting a neutron, but the residual nucleus
may be left in a sufficiently excited state to emit a second neutron; the
most common reactions of this type are the (p, 2n) or (a, 2n) reactions.
At still higher energies, three neutrons may be emitted. When bismuth
16-6] REACTIONS INDUCED BY PROTONS AND ALPHA-PARTICLES 481
u
(cm 2)
10--u~~--~-~--~~~
2.0 3.0 4.0 5.0 6.0 6.5

Ep (Mev)
Fm. 16-11. Cross sections for the reactions Ag 1º7{p, n)Cd 1º7 and Ag 1º9-
(p, n)Cd 1º9 • The circles are experimental points and the dashed curves are the
theoretical resulta of Shapiro. <30 >
is bombarded by a-particles,< 31 > the (a, 2n) reaction is first observed

when the a-particle energy is about 20 Mev. The cross section increases
with increasing energy, and reaches its maximum value of about one barn
at about 30 Mev. At this energy, the (a, 3n) reaction is first observed.
As the energy is increased above 30 Mev, u(a, 2n) decreases, and reaches
a value of on~uarter of a barn at about 40 Mev, while u(a, 3n) increases
to 1¼ barns.
In this energy region, Ghosha1< 32> has made a set of measurements
which serves as an interesting test of the theory of the compound nucleus.
The idea underlying the experiment can be developed from Eq. (16-12).
For a given excited state of a compound nucleus r is a constant and the
ratio r 1,/r may be replaced, for convenience, by 'Y11 • The cross section
for a reaction of the type A + a -+ [C*] -+ B + b may be written
{16-21)
where u.,(E) is the cross section for the absorption of the particle a with
kinetic energy E by the target nucleus A to forro a particular state [C*]
of the compound nucleus; 'Yh(E) is the probability of disintegration of
[C*] into the final nucleus B and the particle b. The excitation energy E
+
of the compound state [C*] is E = E /Ja, where /Ja is the binding energy
482 NUCLEAR REACl'IONS [CHAP. 16
of the particle a to the target nucleus A. lf the compound nucleus [C*]

is formed in the same state of excitation, but by the reaction between a
different initial particle and target nucleus, then the cross section for the
reaction A'+ a' - [C*] - B +bis
(16-22)
where E' is the kinetic energy oí the incident particle a'. Since the excita-
iion energy is assumed to be the same in the two cases, but the binding
energies of the incident particles are different, E' is different from E; but
-Yb(E) is the same in the two cases because of Bohr's basic assumption
that the mode of decay of the compound nucleus [C*] is independent of
the way in which it is formed.
If [C*] disintegrates in a different way to give the particle d and the
residual nucleus D, Eqs. (16-21) and (16-22) are replaced by
(16-23)
and
<r(a', d) = <1'a•(E')-Yd(E). (16-24)
It follows from the last four _equations that
<r(a, b) -Yb(E) <r(a', b)
(16-25)
u(a, d) = -Yd(E) = <r(a', d) ·
An experimental test of Eq. {16-25) provides a direct test of the validity

of Bohr's theory of the compound nucleus.
Ghoshal bombarded Cu 63 with protons and Ni 60 with a-particles, pro-
ducing the compound nucleus (Zn 64 ] in both cases. The energy (mass)
difference between the two target nuclei is such that a proton energy E and
an a-particle energy E' = E + 7 Mev produce the compound nucleus
[Zn 64 ] in the same state of excitation. Ghoshal measured the cross sections
of the reactions
Ni 60 (a, n)Zn 63 , N~ 60 (a, 2n)Zn 62 , Ni 60{a, pn)Cu 62 ;
Cu 63 (p, n)Zn 63 , Cu 63 (p, 2n)Zn 62 , Cu 63 {p, pn)Cu 62 .

According to the analysis leading to Eq. (16-25), the ratios <r(a, n):
u(a, 2n):u(a, pn) for Ni 60 should agree, within the limits of experimental
errors, with the ratios <1(a, n):11(a, 2n):<1(a, pn) for Cu 63 . The experi-
mental resulta are shown in Fig. 16-12, where the six cross sections are
plotted as functions of the kinetic energy of the a-particles and protons.
The proton energy scale has been shüted by 7 Mev with respect to the
a-particle energy scale in order to make the peaks of the proton curves
correspond with those of the a-particle curves. It can be seen from the
16-6] REACTIONS INDUCED BY PROTO~S AND ALPHA-PARTICLES 483
Energy of protons (Mev)
5 9 13 17 21 25 29 33
120~~~~~~~~~~r-,~~ --r-r--,
100
80
60
40
20
o '-----'=-----'-___._-'--_..............,__..,__.,__,..___.c.......,c.......,__.__.___,
8 12 16 20 24 28 32 36 40
Energy oí a (Mev)
FIG. 16-12. Experimental cross sections for the (p, n), (p, 2n), and (p, pn)
reactions on Cu 63 and for the (a, n), (a, 2n), and (a, pn) reactions on Ni 6 º as
functions of the proton and a-particle energies, respectively (Ghoshal< 32 >).
figure that the ratios agree quite well, providing evidence for the validity
of the compound nucleus hypothesis.
C. Coulomb Excitation. A reaction different in nature from those
discussed so far can occur when a charged particle interacts with a nucleus
without specific nuclear forces coming into play; there can be inelastic scat-
tering and excitation of the nucleus because of the Coulomb forces. <33 , 34 >
When a charged particle (a-particle or proton) with a few Mev of energy
moves past a target nucleus, the latter is subjected to a rapidly varying
electric field. This field may give rise to transitions from the ground state
to excited states of the target nucleus. An excited state produced in this
way then decays by emission of 'Y-rays or conversion electrons, and the
decay can be described in terms of appropriate multipole radiation. A
cross section for the production of the 'Y-ray can be measured; it is directly
related to the transition probability, and is determined from the particle
beam current, the thickness of the target, and the efficiency of the detector.
(4+)----------
2+ --...-----+---
2+ ---+---.___.._..........
o+ -----------o
4&Cd110
F10. 16-13. Low-lying energy states of In 11 °"'

showing data obtained by
Coulomb excitation. [Strominger, Hollander, and Seaborg, Revs. Mod: Phys. 30,
585 (1958).)
The cross section for electric excitation is small (about 10-29 or 10-28 cm 2 )
compared with those for direct nuclear transformations, and can be ob-
served only when the latter are extremely improbable, as in collisions of
protons or a-particles of intermediate energy with intermediate or heavy
nuclei. Because of the large Coulomb forces, the classical distance of
closest approach (cf. Chapter 3) is then large compared to the nuclear
radius and the probability of direct nuclear reaction is very small.
Coulomb excitation has been studied in many nuclei, and the theory of
the process has been developed extensively. <35> The nuclear states most
strongly produced in these reactions are the low-lying states (1 Mev or
lower) induced by the electric quadrupole field of the incident particles,
and the 'Y-rays emitted in this case are of the E2 type. The Coulomb ex-
citat.ion reaction is, therefore, a useful tool for the study of low-lying
states; new information has been obtained for previously known levels,
and many new levels have been identified. A decay scheme including data
obtained by means of Coulomb excitation is shown in Fig. 16-13; the arrows
pointing upward represent the transitions excited in the reaction.
16-7 Limitations of the compound nucleus theory. The information

concerning nuclear reactions discussed in Chapter 11 and so far in this
chapter is qualitatively in harmony with the theory of the compound
nucleus. The existence of sharply defined resonances (in agreement with
16-7] LIMITATIONS OF THE COMPOUND NUCLEUS THEORY 485
theory based on Bohr's ideas) shows that the assumption that the mode
of disintegration of the compound nucleus is independent of the way in
which the latter is formed is valid within a certain limited region of excita-
tion energy. When the energy of this incident particle is within the region
of sharp and well-defined resonances and if the energy coincides with or is
near to a resonance, the independence assumption is justified. The nuclear
reaction then produces only one quantum state of the compound nucleus,
and the properties of a given quantum state are independent of the way
it is produced. But, when the energy of the incident particle is high,
several states of the compound nucleus may be excited. The uncertainty
t:.E [cf. Eq. (HH3)] may be large enough to overlap several states, and their
relative excitation may depend on how the nucleus is excited. The decay
of these states may then depend on the mode of excitation, and the inde-
pendence assumption requires further testing.
At these higher energies the independence hypothesis is hard to test
quantitatively, in a direct and general way, because it is not easy to reach
the same range of excitation energy in the compound nucleus with different
reactions. Ghoshal's experiments, discussed in the last section, form one
of a few such tests. The results show that the independence assumption
seems to be valid for a compound nucleus with an excitation energy of
15 to 40 Mev when produced by protons on Cu 63 or by a-particles on Ni 6°.
There are, however, reactions for which the independence hypothesis seems
to break down in that the decay probabilities of the compound nucleus
depend upon the way in which the compound nucleus is formed. For
example, a comparison has been made< 35> of the relative probabilities of
emission of protons and of neutrons from compound nuclei formed either
with protons or with neutrons. When nuclei with mass numbers in the
range 48-71 were bombarded with 21-Mev protons and 14-Mev neutrons,
proton emission was found to be more probable than neutron emission
when the reaction was initiated by protons.
There is also indirect evidence against the independence assumption. <S>
According to Bohr's picture of nuclear reactions, independence is related
to the sharing of the excitation energy by all the particles in the compound
nucleus; the direction of incidence or the position in the nucleus of the
particle initiating the reaction should have no effect on the decay of the
compound nucleus. But reaction products are sometimes emitted with
much greater energy than would be expected ü the energy of the incoming
particle were shared among all the constituents of the nucleus; and the
reaction products often show an angular distribution with more of them
in the direction of the incident particle than would be expected. In these
cases the compound nucleus seems to have more memory of the initial
procesa than is compatible with the independence hypothesis. Certain
experimental resulta have also led to a further examination of the assump-
Proton energy (Mev)
3.0 3.5 4'.o 4.5 5.0 5.5 6.0 6.5 7.0
()16
n1 ~
00
O)
Proton groupe from AJ27 target

c12
Ed = 2.1 Mev
J[300 ~
~¡ z
'o ~
.s
~
:;) ~
:e
z :e
~
200
r'J §
o
~~ B
z
lll
D
o
100 o
¿,
:13
s¡zs
ñ
370 380 390
~
Fm. 16-14. Proton groups from the reaction Al 27 (d, p)Al 28 , in which aluminum was
barded, with 2.1--Mev deuterons (Enge tt•al.< 371 ).•
bom-
-
O>
16-8) DEUTERON-INDUCED REACTIONS 487
tion that the compound nucleus is formed immediately after the incident
particle enters the nucleus, and it has been found that this assumption
must be changed. At excitation energies above the region of sharply
defined resonances, it has therefore become necessary to introduce a modi-
fied picture of nuclear reactions. This treatment will be discussed in Chap-
ter 17 after the subject of nuclear models has been introduced.
There are additional limitations of the compound nucleus hypothesis.
Bohr, in his original statement of the hypothesis, pointed out that his
two-step analysis of nuclear reactions would not apply to light nuclei;
there are too few particles in such nuclei to make possible the large number
of energy exchanges needed for the independence hypothesis to be valid.
It will also be seen in later sections of this chapter that certain reactions
initiated by deuterons are not adequately described in terms of Bohr's
hypothesis, and that sorne photon-induced reactions may occur through a
mechanism other than the formation of a compound nucleus.
16-8 Deuteron-induced reactions: intermediate nuclei, high energy.

Reactions induced by deuterons have been a fruitful source of information
about the energy levels of nuclei. The product particles, protons, neu-
trons, or a-particles, often appear in groups with different energies and
are related to the low-lying excited levels of the product nucleus. Up to
now, the most important results from the standpoint of nuclear spectro-
scopy have been obtained with the light and lighter-intermediate nuclides
in the high-energy region. Because of the special nature of the light nuclei,
the discussion of deuteron-induced reactions will be limited to intermediate
nuclei and the high energy region III.
The reaction Al 27 ( d, p )Al 28 is an example of the way in which deuterons
have been used in the study of nuclear levels. In one experiment, <37 > an
aluminum target was bombarded with deuterons accelerated in an electro-
static generator. The target was placed between the pole pieces of a 180º
magnetic spectrograph, and charged particles leaving the target at 90º
with respect to the incident beam were analyzed in the spectrograph.
Nuclear emulsion plates were used for recording the charged particles, and
proton, deuteron, and a-particle tracks were recognized by the track
lengths. The proton tracks were counted under a microscope in small
strips of plate material and the number of tracks in each strip was plotted
against the product Hr, where H is the magnetic field strength and r is the
radius of the path of the protons in the magnetic field. A sample of the
results is shown in Fig. 16-14 for incident deuterons with an energy of
2.1 Mev; the proton peaks that have been assigned to the Al 27 (d, p)Al 28
reaction are designated by capital Ietters. Other peaks caused by the
presence of impurities were also observed. These impurities were known
and their peaks were identified by comparison with the results obtained
with other target materials. Fifty
~,
2.1
different proton groups were found
corresponding to the ground state
6.J07
and 49 excited states of AI 28 . An •·••o
energy level diagram of these states
J
--6.011
is shown in Fig. 16-15. These states -s.11s

S.755
,.,.,
J.715
decay by 'Y-emission and are bound
states of Al 28 ; together with the vir- AP7+rp .1.415
5.172
,..-1.111
...
tual levels listed in Table 16-3, they s., ••
........... ....., ,
s.u,
5,111
indicate how the spectrum of a • r,,.._ 5.007
lighter-intermediate nucleus looks. '·!.

4.759
'-•····
4.714
In a later experiment,<asi 7.0-Mev 4.112
4.457
deuterons were used as well as a 4.J07
,.,..
magnetic spectrograph with greater •. 115
resolution; 100 proton groups were • J.tJI
4.011
,... ,
._,,111
resolved with nearly all of the new a.,,s
groups having energies lower than J,517
1.111
J.451
those shown in Fig. 16-14. The new 1.1,1
J.142
1.120
proton groups correspond to energy SflO+ra 1.00,
levels of Al 28 about 6 Mev or more -1.tlO
above the ground state. 1.,11

1,571
The 'Y-rays resulting from the de- t.414
cay of excited states of Al 28 could 1.1,1

2.1,1
not be studied under the conditions 2.117
2-
of the Al 27 (d, p)Al 28 reaction dis-
cussed above. They have been ob- 1.,1.s
served in studies of the 'Y-rays from
excited states of Al 28 formed as a 1.167
result of neutron capture in Al 27 •

When the captured neutron is a
1 I.OIJ ...,.
slow one, its kinetic energy is neg-
ligible compared with its binding
energy in the compound nucleus,
and the binding and excitation ener-
gies are practically identical. The
o
,.,.. 0.011
~2.N
compound nucleus, in this case
~
[Al 28 ], decays by 'Y-emission through
excited states of the product nucleus
Al 28, and the energies of the 'Y-rays
give information about these excited
states. About 30 'Y-rays have been
...U ....
observed, and their energies meas- FIG. 16-15. Energy-level diagram
ured with a pair spectrometer. <39l for AI 28 (Enge et al.< 37l).
16-8] DEUTERON-INDUCED REACTIONS 489
Energy spectrum of AI28

Al27 (n,'Y)Al28 (Kinsey)
(x4)
Capture
state
o 1.0 2.0 3,0 4.0 5.0 6.0 7,0 H.O
Al27 (d,p)A128
Ed = 2.1 Mev, 8 = 90°
2~ a~ ~o .;.O 6,0
Excitation energy
Fm. 16-16. Energy-level diagram for Al 28 with relative intensities (Enge

et al. <37>).
The most energetic 'Y-ray hadan energy of 7.72 Mev, which is just the
binding energy of the captured neutron, and there can be little doubt that
this 'Y-ray represents the direct transition from the capture state to the
ground state. The differences in energy between this 'Y-ray and those
with less energy can be related to bound states of Al 28 . The states inferred
in this way are compared in Fig. 16-16 with those given by the (d, p)
reaction. The upper part of the diagram was constructed from (n, 'Y)
studies, and the height of the lines represents the relative intensities of
the 'Y-rays. The lower diagram shows the states as deduced from the (d, p)
studies and indicates the relative proton yields. The levels given by the
two methods are in good agreement, and the maxima and mínima in the
relative intensities have the same locations in the two diagrams.
The (d, a) reaction has been used in a similar way as, for example, in
the reaction Al 27 (d, a)Mg 25 . Excited states of Mg 25 can also be obtained
from the reaction Mg 24 (d, p)Mg 25 and the results of a studyc 4 o> of the
two reactions are listed in Table 16-6. The levels of Mg 25 obtained from
the two reactions agree very well.
The (d, n) reaction often yields groups of neutrons whose energies
lea.d to levels of the product nucleus. An examplecrn is the reaction
Mg 26 (d, n)Al 27 , which gives excited states of Al 27 at 0.88 ± 0.07,
1.92 ± 0.07, 2.75 ± 0.07, 3.65 ± 0.07, 4.33 ± 0.07, 5.32 ± 0.07, and
5.81 ± 0.07 Mev. These levels are in good agreement with levels found
from the inelastic scattering of protons. The inelastic scattering of deuter-
ons, like that of protons, gives information about the levels of the target
490 NUCLEAR REACTIONS [CHAP.16
TABLE 16---6
ENERGY LEVELS IN Mg 26 *
Mg24(d, p)Mg2s A127(d, a)Mg25
0.582 ± 0.006 0.584 ± 0.006

0.976 ± 0.006 0.977 :±: 0.010
1.612 :±: 0.006 1.610 :±: 0.010
1.957 :±: 0.006 1.958 :±: 0.010
2.565 :±: 0.006 2.558 ± 0.010
2.742 ± 0.008 2.729 :±: 0.010
2.806 :±: 0.007 2.791 ± 0.015
3.405 + 0.007 3.404 ± 0.012
3.899 + 0.008 3.896 ± 0.015
3.972 ± 0.010 3.965 ± 0.015
*Excita.tion energy in Mev.
nucleus. Levels of Al27 bave been found in this way<Ol with excitation
energies of 0.97, 2.39, 3.17, 4.74, and 5.76 Mev. The experimental error
in this case was about 0.2 Mev, and within the limits of this error, the levels
agree with some of those obtained from the inelastic scattering of protons
by Al 27, and from the (d, p) reaction on Mg 26 . Finally, the reaction
Si 21 (d, a)Al 27 gives levels at 0.84 and 1.01 Mev, <Ol in agreement with the
two lowest levels obtained by the three other methods.
The theoretical treatment of deuteron-induced reactions is more difficult
tban the treatment of the reactions discussed earlier in this cbapter. One
reason is tbat new types of processes occur with deuterons which cannot
be accounted for on the basis of the compound nucleus picture. These
processes result from the following properti~ of the deuteron: (a) the
deuteron is a very loosely bound structure with a binding energy of only
2.23 Mev as compared with 28.3 Mev for the a-particle and 7 or 8 Mev
for the average particle in a nucleus; (b) its charge distribution is very
unsymmetric, i.e., the center of mass and the center of charge (the proton)
do not coincide as they do in the a-particle.
According to the compound nucleus theory, a nucleus X(Z, A) bom-
barded by deuterons should form the compound nucleus [C(Z +
1, A +
2)),
which can then decay in different ways:
Y(Z +
1, A 1) + + n,
X(Z, A) + d -+ [C(Z + 1, A + 2)) -+ ¡Y(Z, A 1) + +
p,
Y(Z - I, A - 2) + a.
(16-26)
The special properties of the deuteron make it possible for other processes
to take place in addition to those of the acherne (16-26). Because of the
16-9] REACTIONS INDUCED BY GAMMA-RAYS 491
finite distance between the proton and neutron in the deuterQn one of
these particles may reach the nuclear surface before the other. The
nuclear interaction energies (average binding energy per nucleon) are
much higher than the binding energy of the deuteron, and the constituent
of the deuteron that arrives first at the nuclear surface is quickly sepatated
from its partner. If the kinetic energy of the deuteron is not very high, the
Coulomb field of the nucleus keeps the proton away and the neutron is
more likely to enter the nucleus than the proton. If the second constit-
uent of the deuteron hits the nucleus an instant later, the compound
nucleus [C(Z + 1, A + 2)] is formed just as in the scheme (16-26). But
if the second nucleon (proton) misses the nucleus, the process that takes
place is
X(Z, A) + d __. [C'(Z, A + 1)] + p. (16-27)
The compound nucleus [C'(Z, A+ 1)] may decay by emission of sorne
other particles, or by 'Y-decay. If the proton penetrates the Coulomb
barrier and hits the target nucleus first, the process may be
X(Z, A) + d __. [C"(Z + 1, A + 1)] + n. (16-28)
The processes (16-27) and (16-28) are called stripping reactions because a
nucleon is stripped out of the deuteron by the target nucleus. At relatively
low energies, the reaction (16-27) is called the "Oppenheimer-Phillips
process" after the two workers who first described it in 1935.
The additional processes which can take place with deuterons make the
cross sections for these reactions greater than those for reactions with
other charged particles of similar energy. As a result, deuteron reactions
are of great practica} importance in the production of radioactive nuclides
and in the study of nuclear reactions. The characteristics of stripping
reactions are closely related to the properties of the state of the product
nucleus, and the investigation of these reactions has been of special use in
nuclear spectroscopy. <44 , 59 >
16-9 Reactions induced by gamma-rays. The disintegration of nuclei

by 'Y-rays occurs when a photon is absorbed whose energy hv is greater
than the separation energy of a particle (proton, neutron, or a-particle).
When this condition is satisfied, a ('Y, p), ('Y, n), or ('Y, a) reaction is ob-
served. A process of this type is often called a plwtodisintegration; the
term nuclear plwtoejfect is sometimes used in analogy with the atomic
photoelectric effect.
The simplest example of this type of reaction is the photodisintegration
of the deuteron. This reaction has a threshold at a 'Y-ray energy just
equal to the binding energy of the deuteron, and very careful determina-
tions of the threshold energy have been made because of the great impor-
tance of the binding energy of the deuteron in the theory of the nucleus.
In a highly precise measurement< 46> of the threshold, deuterium was
bombarded with x-rays from the action of electrons on a gold target;
the electrons from a cathode-ray "gun" were accelerated in an electrostatic
generator. In the neighborhood of 2 Mev, the energy of the electrons is
practically identical with that of the most energetic x-rays, and if the elec-
tron energy is mea.sured very carefully, that of the x-rays is known with
high precision. The neutrons resulting from the reaction d + 'Y -+ p + n
were counted as a function of the electron energy, and the pkoúmeutron
threshold was obtained by extrapolating the neutron yield versus energy
curve to zero yield. The results of the experimentare shown in Fig. 16-17.
The open circles give the neutron count; the dots represent the square
root of the difference between the total neutron count and the background
count. According to theory, this quantity should be proportional to the
energy above threshold. The straight lines show that this is the case and
allow an accurate extrapolation to zero yield. The binding energy of the
deuteron was determined as 2.226 ± 0.003 Mev. The shape of the yield
curve is typical for photodisintegration reactions at energies just a.hove
the threshold.
In the experiment just described, the threshold for the photodisintegra-
tion of beryllium was also determined. The reaction in this case is
9 4 4
4Be + 'Y -+ 2He + 2He + n,
and the threshold energy gives the binding energy of the neutron in the
Be 9 nucleus as 1.666 ± 0.002 Mev.
Photoneutron thresholds ha.ve been mea.sured for many ('Y, n) reac-
tions<46> over the entire range of elements, chiefly to get neutron binding
energies. The measurement of neutron binding energies is one method
of establishing mass differences with great accuracy and also provides
information about nuclear structure. The neutron whose binding energy
is determined is the "last" neutron, that is, the one removed from a stable
nucleus or added to a stable nucleus. The energy needed to remove a
neutron from a stable nucleus is given by the threshold of the ('Y, n) or
(n, 2n) reaction, or by the Q-value of the ground-state group from a (p, d)
reaction or a (d, H 3) reaction. These reactions furnish neutron binding
data for the target nucleus. The energy relea.sed on the addition of a
neutron to a stable nucleus is given by the 'Y-rays from the capture of slow
neutrons (n, 'Y) or can be obtained from the Q-value of the ground-state
proton group from a (d, p) reaction. <47 >
With the exception of deuterium and beryllium, the binding energy of
the last neutron líes between about 5 and 13 Mev. The experimental
values of the binding energy have been tabulated and analyzed as a func-
tion of the number of neutrons in the nucleus. <47 > It was found that
60 1 1 1 1 1600
50
D(-r,N)
l11 1400
.(Normalized neutron counts
minus backgroll1ld)i
40 o Normalized neutron counts
11 1200
Electron current 15¡,a y) 1000
¡ "' e
J
800
r
1/
j 600 §
~
5
o
{
I
V
I l)
I/
8
~
~
5
../ I /
)
v OO"ál"
12
l
,~ ✓ / ~
1oooE
o
V 1/
z
'- ~lectron e ~rren t 10¡,a 800
5
V V 600
)
V V
o 400
lJ'
) ,(
/ 200
/· l,d
~
- - •/ ~
2.240
o
2.225 2.230 2.235
Electron energy (Mev absolute)
F10. 16-17. The photoneutron yield from deuterium near threshold (Mobley
and Laubenstein< 45 >).
nuclei with 28, 50, 82, and 126 neutrons have special properties; in each
case there is a sharp break in the binding energy of the neutron added to
the nucleus. Thus the binding energy of the 127th neutron is about
2.2 Mev less than that of the 126th neutron. The binding energy of the
50th neutron is about 2 Mev greater than that of the 51st neutron; simi-
larly, there is a break at 82 neutrons and probably one at 28 neutrons.
These results indicate that the nuclei with 28, 50, 82, and 126 neutrons
are markedly more stable than their neighbors. In Section 16-5C, the
special nature of nuclei with 50, 82, and 126 neutrons was shown by the
low values of-their capture cross sections for 1-Mev neutrons. A theory
of nuclear structure which accounts for the particularly stable nature of
these nuclei will be discussed in the next chapter. These results provide
another example of the way in which the study of nuclear reactions is
related to the problem of nuclear structure.
The ('Y, n) cross section has been measured as a íunction oí the energy
oí the 'Y-rays< 4 s> for a series oí elements covering the periodic table. In
the intermediate and heavy nuclei, the cross section increases quite
rapidly above the threshold energy, reaching a maximum somewhere near
20 Mev. The maximum value oí the cross section increases with increasing
mass number; ~t is oí the order oí 10- 26 cm 2 for the lighter-intermediate
nuclei such as aluminum, and reaches a value oí nearly 10- 24 cm 2 in
bismuth.
Studies of ('Y, p) reactions< 49 > show that they are .similar in nature to
the ('Y, n) reactions in that the cross section versus energy curves have the
same general shape. An interesting theoretical problem arises because
oí the relative magnitudes oí ('Y, n) and ('Y, p) cross sections. In reactions
involving the intermediate and heavy nuclei, it seems natural to assume
that the Bohr theory oí the compound nucleus should apply, and that
the process is divided into two parts: (I) the absorption of the 'Y-ray
photon forming a compound nucleus with the excitation energy E = h11,
and (2) the decay of the compound nucleus by the emission of any one
of several different particles. Beca.use of the blocking effect of the Cou-
lomb barrier, the most probable process is the emission of a neutron lead-
ing to a ('Y, n) reaction; protons are expected to be emitted much more
rarely. In many cases, the relative yield of ('Y, n) and ('Y, p) reactions does
not agree with these ideas, and proton emission has been found to be much
stronger than expected. These unexpectedly large ('Y, p) cross sections
have been interpreted in terms of a direct photodisintegration process. <so>
lt is supposed that the energy of the 'Y-ray photon is absorbed by a single
proton in the nucleus which is then emitted before the energy can be
shared with the other nucleons. Theoretical calculations show that only
a small number of protons would have to be ejected by this direct process
to give rise to a ('Y, p) cross section appreciably larger than the one ex-
pected from thé compound nucleus alone, and there seems to be evidence
supporting the idea of a direct photoproton effect.
At 'Y-ray energies in the neighborhood of 20 Mev, the ('Y, 2n) and ('Y, np)
reactions are observed and their cross sections increase with increasing
energy, while the ('Y, n) and ('Y, p) cross sections decrease.
In contrast with ('Y, p) reactions, ('Y, a) reactions can be interpreted ade-
quately at present in terms of the compound nucleus theory, as is shown,
for example, by recent experiments' 51 > on the reaction V 51 ('Y, a)Sc 47 .
A sample of vanadium was bombarded with 'Y-rays with energies which
could be varied from 10.5 to 25 Mev. The yield was determined as a func-
tion of 'Y-ray energy by measuring the activity of the Sc 47 produced; the
ntensity oí the 160-kev 'Y-ray characteristic of Sc 47 was determined with
ys1 ('Y, a) Sc47
•
0.10
I
f
•
0.001~-~-~~~-~-~-~-~
10 14 18 22
Energy (Mev)
Frn. 16-18. Excitation function for the reaction V 51 (-Y, a)Sc 47 .< 51 >
a Nal(Tl) scintillator and a pulse-height analyzer. Vanadium consists

almost entirely (99.75%) of V 51 , and the products of the ('Y, n) and ('Y, p)
reactions on that nuclide are stable. Hence those reactions, which are more
probable than the ('Y, a) reaction, do not interfere with the analysis of the
latter. The (-Y, a) reaction was not observed below 15.5 Mev because of
the effect of the Coulomb barrieron the emission of a-particles. The excita-
tion function (curve of yield or cross section against energy) has a maxi-
mum of 0.8 millibarns at 23 Mev and agrees with the theoretical curve
based on the compound nucleus theory, as shown in Fig. 16-18. This
result seems to be a general property of ('Y, a) reactions in medium-weight
nuclei.
It is not surprising that the idea of the compound nucleus applies to
these reactions; a-particles probably do not exist as such in the nucleus
so that direct ejection of an a-particle is very unlikely. It is reasonable
to suppose that the excitation energy brought by the incident 'Y-ray is
shared among many particles and that, under favorable conditions, four
nucleons fuse to form an a-particle in a kind of interna! conversion process.
The emission of the a-particle from the excited compound nucleus is then
analogous to the emission of a-particles from radioactive nuclei.
16-10 Reactions at ultra-high energies. At energies of about 50 Mev

and up, nuclear reactions are quite different in character from those at
lower energies. When a particle with ultra-high energy strikes a nucleus,
it may be absorbed to forro a very highly excited compound nucleus. It
is more likely, however, that the particle will collide with one or more
nucleons and then escape with part of its original kinetic energy. By
means of this scattering process, one or a few nucleons may get a large
fraction of the energy of the incident particle. The latter nucleon or group
of nucleons may be ejected from the nucleus without further collision, or
they may share their energy with a few other nucleons. In most collisions
with ultra-high energy particles, the target nuclei are excited to energies
far greater than the binding energies of individual nucleons. N ucleons
or groups of nucleons such as a-particles or lithium nuclei escape until
the residual nucleus is left in a relatively stable state. At energies of
100 Mev or more, a dozen or more nucleons may escape and many com-
binations of protons and neutrons have been observed. Reactions of this
kind are described by the term spallation. The theory of the compound
nucleus does not apply to these reactions because there is not enough time
for the excitation energy to be shared among ali the nucleons before
particles start being emitted. It is no longer possible to separate the reac-
tion into two independent phases and it is necessary to treat separately
the interaction of the incoming particle with the individual nucleons of the
target nucleus. The region of ultra-high energy also includes reactions
induced by cosmic rays, including the production of new kinds of particles
called mesons. The complexity of both the experimental information and
the theoretical ideas is such that further discussion of ultra-high energy
nuclear physics is beyond the scope of this book. The interested reader
is referred to the literature< 52 , 53 > for relatively easy treatments of these
subjects.
16-11 Reactions with light nuclei: energy levels of light nuclei. A great
deal of information about the excited states of light nuclei has now been
accumulated from nuclear reactions. Sorne of the most recent compila-
tio11s<54, 57> include more than 400 energy levels of the 40-odd nucleí of
atomic number less than 11; another< 25 > includes many levels of nuclei
with Z between 11 and 20. The progress made in recent years in the
methods of studying excited states, together with the establishment of
an accurate and consistent set of nuclear masses (Table 11-1), has made
it possible to locate these states with high precision. In the majority
of the nuclei in this region, the first 10 Mev of excitation have been studied
thoroughly enough so that most of the stationary states are probably
known. Although the reactions do not fall into the kind of patterns that
were found for intermediate and heavy nucleí, the energy levels provide a
16--11) REACTIONS WITH LIGHT NUCLEI 497
body of experimental data in which important clues to the problems of

nuclear forces and nuclear structure can be found. In the discussion of
light nuclei, therefore, the emphasis will be on the levels rather than on
the reactions. With a few exceptions, the reactions are of the types which
have already been discussed, and the treatment of the light nuclei in this
section can be regarded as a further application of the methods and ideas
which have been developed in the earlier sections of this chapter.
The energy levels of the light nuclei are sufficiently well defined and
separated so that it is possible to present the level structure for each nucleus
in the forro of a diagram showing the known Ievels and indicating the
nuclear reactions in which they are involved. In addition to the position
of the level, other information is also desired in nuclear spectroscopy, such
as the width or lifetime of an excited state. Two additional parameters
that are important in the description of a nuclear state are the angular
momentum quantum number and the parity. The latter quantities can
really be understood only in terms of quantum mechanics, and their effect
on nuclear reactions has therefore not been mentioned in this chapter.
They influence the angular distribution of the reaction products, and
information about them can be obtained from the way in which the yield
of the product particle varíes with the angle between the direction of
emission of the product particle and the direction of the incident particle.
Information about the angular momentum and parity can also be obtained
from the probability of ¡9-decay and that of "Y-decay. When the values
of these parameters for a state are known, they are also shown in the
level diagram.
The way in which the energy-level diagram is constructed can be illus-
trated by treating a typical nucleus in sorne detail. The nucleus NI 5
has been chosen because of the large number of reactions in which it is
either the compound nucleus or the product nucleus. AH of the data
which will be considered have been taken from the compilation by Ajzen-
berg and Lauritsen referred to above, <57> where the references to the
original sources are given. The energy level diagram for NI 5 is shown in
Fig. 16-19. Although it appears formidable at first sight, it will be seen
to provide a concise and easily interpreted summary of the available data.
Altogether, 25 reactions among various light nuclei provide information
about the excited states of NI 5. These are: B 11 (a, n)N 14, B 11 (a, p)C 14,
C 12 (a, p)N 15, C 12 (t, p)C 14 , C 13 (d, n)N 14 , C 13 (d, p)C 14, C 13 (d, a)B 11 ,
CIª(d, t)c12, Cl4(d, n)N1s, Cl4(p, n)Nl4, c1s(¡9-)N1s, Nl4(n, n)Nl4,
N 14 (n, a)B 11 , N 14 (n, p)C 14, N 14 (n, "Y)N 15, N 14 (n, 2n)N 13 , N 14 (n, t)C 12 ,
Nl4(d, p)Nl5, Nl4(t, pn)Nl5, O1s(¡9+)N1s, ois(p, a)Nl5, c14(p, "Y)Nl5,
O 16 (-Y, p)N 15 , O 16 (n, d)N 15, and O 17 (d, a)N 15. In 12 reactions, N 15 is
the product nucleus, and in 13 it is the compound nucleus. Low-lying
levels (bound states) are given by the proton groups from the reactions
ti"
9.777 ,.,
C' 87
d-n 7.700
o'"+t-a
t
T 5.2a
W1º--
o +p-a
Fio. 1&-19. Energy levels of N 15 (Ajzenberg-Selove and Lauritsen< 57>).
N 14 (d, p)N 15 and C 12 (a, p)N 15 , by the neutron groups from the reaction
C 14 (d, n)N 15, and by the 'Y-rays from the reaction N 14 (n, 'Y)N 15. These
levels are listed in Table 16-7. Although no single reaction gives all the
levels, there is good agreement among the values obtained from the
different reactions. The middle column of Fig. 16-19 shows the energy
states of N 15 • The numbers indicate the energy in Mev above the ground
state of N 15, which serves as the zero of the energy scale for the whole
diagram. The N 15 spectrum is remarkable for the rather great excitation
(5.28 Mev) of the lowest excited level and for the fact that this level is a
doublet with a spacing of 30 kev. The angular momenta are indicated by
the J-values as obtained from the analysis of the angular distribution of
16-11) REACTIONS WITH LIGHT NUCLEI 499
TABLE 16-7
Low-LYING LEVELS OF N
15 FROM VARIOUS REACTIONs*
N14(d, p)N1s N14(n, 'Y)N1s c12(a, p)Nl5 c14(d, n)Nl5

proton groups 'Y-ra.ys proton groups neutron groups
5.276 5.275
5.305 5.3 5.34
6.328 6.325 6.5 6.32
7.164 7.164
7.309 7.356 7.46
8.315 8.278
9.156
* Excita.tion energy in Mev.

TABLE 16-8
RESONANCE LEVELS IN N 15 IN VARIOUS REACTIONS*
C 14(p,n)N 14 N 14(n, p)C 14 N 14(n, n)N 14 N 14(n, a)B 11 C 13(d, n)N 14 B 11 (a, n)N 14
11.235
11.294 11.30 11.293
11.430 11.431 11.430
11.760 11.761 11.765
11.877 11.879
11.943
11.964 11.964
12.096 12.094
12.147 12.155 12.142 12.16
12.328 12.323 12.38
12.495 12.47 12.494
12.94 12.94 12.95
13.22
13.45 13.40 13.34
13.63
14.05
14.26 14.29
14.50
14.63 14.71
14.90
15.62
15.89
16.13
16.47 16.65
17.04 16.89
17.30 17.49
17.78 17.69
18.43
18.97
* Excit&tion energy in Mev.

500 NUCLEAR REACTJONS [CHAP. 16
the protons from the reaction N 14 (d, p)N 15 • When two numbers are
given for the angular momentum, it means that the assignment is not yet
certain but the choice has been narrowed down to the values given. Even
parity is indicated by a plus sign, odd parity by a minus sign. The reactions
which give the low-lying levels are indicated at either side oí the central
column, together with the binding energy of the incident particle in the
compound nucleus. For example, the reaction N 14 (d, p)N 15 is shown to
the left of the central column together with the Q-value for the reaction,
which is 8.609 Mev. The formation oí N 15 by the t3+-decay oí 0 15 and
the ¡S--decay oí C 15 are also shown.
Virtual levels oí N 15 , as given by six different reactions, are listed in
Table 16-8. About 30 levels, obtained from resonances in these reactions,
are shown in the upper part oí the central column oí Fig. 16-19. The
reactions which give rise to these resonances are shown on the sides oí the
diagram, together with the binding energy oí the incident particle in the
compound nucleus and a sketch oí the cross section curve for each reaction.
Thus, the reactions N 14 (n, n)N 1 4, N 14 (n, p)C 1 4, and N 14 (n, a)B 11 are
indicated to the leít oí the upper part oí the central column. The binding
energy of the last neutron in [N 15] is 10.834 Mev and this value serves as
_L
,-
1 Mev
[0.14)
Frn. 16-20. Energy levels of N 15 and 0 15 .< 58>

REFERENCES 501
a zero point for the energies of the particles in the various neutron-induced
reactions. The energy of the excited level is obtained from the binding
energy and the resonance energy as shown in Section 16-2. The values
of the resonance levels as given by the different reactions are in excellent
agreement. The levels which are cross-hatched are known to be particu-
larly broad, that is, to have a particularly large r.
The comparison of the energy level diagrams of the light nuclei shows
that there are similarities in sorne of the level spectra which have a bearing
on the problem of nuclear forces. <55 , 56 • 58> It has been found, for example,
that the distribution of low-lying levels in mirror nuclei is similar. In
Fig. 16-20, these levels<su are shown for N 15 and 0 15 • The former nucleus
has 7 protons and 8 neutrons, while the latter has 8 protons and 7 neutrons.
Results like these provide evidence on the relative magnitudes of the force
between two protons and that between two neutrons and, in fact, support
the hypothesis that these two forces are equal.
REFERENCES
GENERAL
R. D. EvANs, The Atomic Nuclem. New York: McGraw-Hill, Chapters 12,
13, 14.
P. MoRRISON, "A Survey of Nuclear Reactions," Experimental Nuclear
Physics, E. Segre, ed. New York: Wiley, 1953, Vol. 11, Part VI.
L. E1sENBUD and E. P. WIGNER, Nuclear Structure. Princeton, New Jersey:
Princeton University Press, 1958, Chapters 4, 9, 10.
J. M. BLATT and V. F. WEISSKOPF, Thearetical Nuclear Physics. New York:
Wiley, 1952, Chapters 8, 9, 10.
P. M. ENDT and M. DEMEUR, eds. Nuclear ReactiQns. Amsterdam: North-
Holland Publishing Co.; New York: Interscience Publishers, 1959. ·
F. ÁJZENBER0-8ELOVE, ed., Nuclear Spectroscopy. New York: Academic
Press, 1960.
ARTICLES IN "The Encyclopedia of Physics" (Handbuch der Physik) Berlín:
Springer, 1957-1959.
l. W. E. BuRCHAM, "Nuclear Reactions, Levels, and Spectra of Light Nuclei,"
Vol. 40, Nuclear Reactims I, pp. 1-201.
2. B. B. KINSEY, "Nuclear Reactions, Levels, and Spectra of Heavy Nuclei,"
Vol. 40, Nuclear Reactims I, pp. 202-372.
3. J. RAINWATER, "Resonance Processes by Neutrons," Vol. 40, Nuclear
Reactions I, pp. 373-409.
4. A. WATTENBERG, "Nuclear Reactions at High Energies," Vol. 40, Nuclear
Reactions I, pp. 450-537.
5. G. BREIT, "Theory of Resonance Reactions and Allied Topics," Vol. 41,
Nuclear Reactims II, pp. 1-407.
6. G. R. BrsHoP and R. WILSON, "The Nuclear Photoeffect," Vol. 42, Nuclear
Reactims III, pp. 309-361.
C0LLECTI0NS OF ENERGY LEVEL DATA

l. F. AJzENBERG and T. LAURITSEN, "Energy Levels of Light Nuclei, IV,"
Revs. Mod. Phys. 24, 321 (1952).
2. F. AJZENBERG and T. LAURITSEN, "Energy Levels of Light Nuclei, V,"
Revs. Mod. Phys. 27, 77 (1955).
3. F. AJzENBERo-8ELOVE and T. LAURITSEN, "Energy Levels of Light Nu-
clei," Nuclear Physics 11, 1-340 (1959).
4. P. M. ENDT and J. C. KLYVER, "Energy Levels of Light Nuclei (Z 11
to Z = 20)," Revs. Mod. Phys. 26, 95 (1954).
5. P. M. ENDT and C. M. BRAAMS, "Energy Levels of Light Nuclei (Z 11
to Z = 20), 11," Revs. Mod. Phys. 29, 683 (1957).
6. WAY, K1No, McGINNIS, and VAN LIESHOUT, Nuclear Energy Levels (A 40
to A = 92; Calcium to Zirconium). U.S. Atomic Energy Commission Document
TID-5300. Washington, D.C.: U.S. Government Printing Office, 1955.
7. STROMINGER, HoLLANDER and SEABORG, "Table of Isotopes/' Revs. M od.
Phys. 30, 585 (1958).
PARTICULAR
l. CHADWICK, CoNsTABLE, and PoLLARD, "Artificial Disintegration by a-
Particles," Proc. Roy. Soc. (London) Al30, 463 (1931).
2. J. CHADWICK and J. E. R. CoNsTABLE, "Artificial Disintegration by a-
Particles. Part 11, Fluorine and Aluminum," Proc. Roy. Soc. (London) Al35,
48 (1932).
3. W. E. DuNCANSON and H. MrLLER, "Artificial Disintegration by Ra.C' a-
Particles: Aluminum and Magnesium," Proc. Roy. Soc. (London) Al46, 396
(1934).
4. B. B. BENSON, "Proton-Gamma Re.y Coincidence Counting with Cyclotron
Bombardment," Phys. Rev. 73, 7 (1948).
5. M. S. LrvrNGSTON and H. A. BETHE, "Nuclear Physics. C. Nuclear Dy-
namics, Experimental," Revs. Mod. Phys. 9, 245 (1937).
6. BROLLEY, SAMPSON, and MrTCHELL, "Energy Levels of Si3°," Phys. Rev. 76,
624 (1949).
7. N. BoHR, "Neutron Capture and Nuclear Constitution," Nature 137, 344
(1936).
8. F. L. FRIEDMAN and V. F. WEISSKOPF, "The Compound Nucleus," Niels
Bohr and The Development of Physics, W. Pauli, ed. London: Pergamon Press,
1955, pp. 134-161.
9. G. BREIT and E. WmNER, "Capture of Slow Neutrons," Phys. Rev. 49,
519 (1936).
10. J. M. BLATT and V. F. WErssKOPF, Theoretical Nuclear Physics. New
York: Wiley, 1952, Chapter 8.
11. J. M. BLATT and V. F. WEISSKOPF, Ref. 10 a.hove, Chapter 9.
12. D. J. HuGHES, Pile Neutron Research. Rea.ding, Ma.ss.: Addison-Wesley,
1953.
13. R. L. HENKEL and H. H. BARSCHALL, "Capture Cross Sections for Fast
Neutrons," Phys. Rev. 80, 145 (1950).
REFERENCES 503
14. D. J. HUGHES and R. B. ScHWARTZ, "Neutron Cross Sections." Brook-

haven National Labqratory Repqrt BNWB5, 2nd ed. Washington, D.C.: U.S.
Government Printing Office, July 1958.
15. RAINWATER, HAVENs, Wu, and DuNNING, "Slow Neutron Velocity Spec-
trometer Studies, l. Cd, Ag, Sb, Ir, Mn," Phys. Rev. 71, 65 (1947).
16. D. J. HuGHES, Ref. 12 above, pp. 109-114.
17. R. SHERR, "Collision Cross Sections for 25-Mev Neutrons," Phys. Rev.
68, 240 (1945).
18. J. M. BLATT and V. F. WEISSKOPF, Ref. 10 above, pp. 481-483.
19. H. FESHBACH and V. F. WEISSKOPF, "A Schematic Theory of Nuclear
Cross Sections," Phys. Rev. 76, 1550 (1949).
20. SETH, HuoHEs, ZIMMERMAN, and GARTH, "Nuclear Radii by Scattering
of Low-Energy Neutrons," Phys. Rev. 110, 692 (1958).
21. B. L. COHEN, "(n, 2n) and (n, p) Cross Sections," Phys. Rev. 81, 184 (1951).
22. BROLLEY, FowLER, and ScHLACKS, "(n, 2n) Reactions in C 12, Cu 63 and
Mo 92 ," Phys. Rev. 88, 618 (1952).
23. BROSTR0M, Huus, and TANGEN, "Gamma-Ray Yield Curve of Aluminum
Bombarded with Protons," Phys. Rev. 71 1 661 (1947).
24. P.M. ENDT and J. C. KLYVER, "Energy Levels of Light Nuclei, (Z = 11
to Z = 20)," Revs. Mod. Phys. 26, 95 (1954).
25. P. M. ENDT and C. M. BRAAMS, "Energy Levels of Light Nuclei. II."
Revs. Mod. Phys. 29, 683 (1957).
26. BROWNE, ZIMMERMAN, and BuECHNER, "Energy Levels in Al 27 ," Phys.
Rev. 96, 725 (1954).
27. RxcHARDs, SMITH, and BROWNE, "Proton-Neutron Reactions and Thresh-
olds," Phys. Rev. 80, 524 (1950).
28. ScHOENFELD, DuBORG, PRESTON, and GooDMAN, "The Reaction
Cl37 (p, n)A37 ; Excited States in A38," Phys. Rev. 85, 873 (1952).
29. LovrNGTON, McCuE, and PRESTON, "The (p, n) Reaction on Separated
Isotopes of Chromium," Phys. Rev. 85, 585 (1952).
30. M. M. SHAPIRO, "Cross Sections for the Formation of the Compound
Nucleus by Charged Particles," Phys. Rev. 90, 171 (1953).
31. E. L. KELLY and E. SEGRE, "Sorne Excitation Functions of Bismuth,"
Phys. Rev. 75, 999 (1949).
32. S. N. GHOSHAL, "An Experimental Verification of the Theory of the Com-
pound Nucleus," Phys. Rev. 801 939 (1950).
33. C. L. McCLELLAND and C. GooDMAN, "Excitation of Heavy Nuclei by the
Electric Field of Low-Energy Protons," Phys. Rev. 91, 760 (1953).
34. T. Huus and C. ZUPANcrc, "Excitation of Nuclear Rotational States by
the Electric Field of lmpinging Particles," Kgl. Danske Videnskab, Selskab.
llfat. fys. Medd. 28, No. 1 (1953).
35. ALDER, BoHR, Huus, MoTTELSON, and W1NTHER, "Study of Electro-
magnetic Excitation with Accelerated lons," Revs. llfod. Phys. 28, 432 (1956).
36. B. L. CoHEN and E. NEWMAN, "(p, pn) and (p, 2n) Cross Sections in
Medium Weight Elements," Phys. Rev. 99, 718 (1955).
37. ENGE, BuECHNER, and SPERDUTO, "Magnetic Analysis oi the Al 27 (d, p)Al 28
Reaction," Phys. Rev. 88, 961 (1952).
38. BuECHNER, MAZARI, and SPERDUTO, "Magnetic Spectograph Measure-

ments on the Al 27 (d, p)Al 28 Reaction," Phys. Rev. 101, 188 (1956).
39. KINSEY, BARTHOLOMEW, and WALKER, "Neutron Capture of 'Y-Rays
from F, Na, Mg, Al, and Si," Phys. Rev. 83, 519 (1951).
40. ENDT, ENGE, HAFFNER, and BUECHNER, "Excited States of Mg 25 from the
Al 27 (d, a)Mg 25 and Mg 24 (d, p)Mg 25 Reactions," Phys. Rev. 87, 27 (1952).
41. SwANN, MANDEVILLE, and WHITEHEAD, "The Neutrons from the Disinte-
gration of the Separated Isotopes of Magnesium by Deuterons," Phys. Rev. 79,
598 (1950).
42. K. K. KELLER, "The Al 27 (d, p)Al 28 and Al 27 (d, d')Al 27 Reactions," Phys.
Rev. 84, 884 (1951).
43. D.M. VAN PATTER and W. W. BUECHNER, "lnvestigation of the (d, p)
and (d, a) Reactions of the Silicon Isotopes," Phys. Rev. 87, 51 (1952).
44. S. T. BuTLER, "Angular Distributions from (d, p) and (d, n) Nuclear Re-
actions," Proc. Roy. Soc. (London) A208, 559 (1951).
45. R. C. MoBLEY and R. A. LAUBENSTEIN, "Photon-Neutron Thresholds of
Beryllium and Deuterium," Phys. Rev. 80, 309 (1950).
46. SHER, HALPERN, and MANN, "Photo-Neutron Thresholds," Phys. Rev. 84,
387 (1951).
47. J. A. HARVEY, "Neutron Binding Energies from (d, p) Reactions and
Nuclear Shell Structure," Phys. Rev. 81, 353 (1951).
48. MoNTALBETTI, KATZ, and GoLDEMBERG, "Photoneutron Cross Sections,"
Phys. Rev. 91, 659 (1953).
49. J. HALPERN and A. K. MANN, "Cross Sections of Gamma-Proton Reac-
tions," Phys. Rev. 83, 370 (1951).
50. E. D. CouRANT, "Direct Photodisintegration Processes in Nuclei," Phys.
Rev. 82, 703 (1951).
51. P. DYAL and J. P. HuMMEL, "Excitation Function for the V51 ('Y, a)Sc 47
Reaction," Phys. Rev. 115, 1264 (1959).
52. R. B. LEIGHTON, Principles of Modern Physícs. New York: McGraw-
Hill, 1959, Chapters 20, 21.
53. RICHTMYER, KENNARD, and LAURITSEN, lntroductwn to Modern PhyBics,
5th ed. New York: McGraw-Hill, 1955, Chapter 11.
54. F. AJZENBERG and T. LAURITSEN, "Energy Levels of Light Nuclei. V,"
Revs. Mod. Phys. 27, 77 (1955).
55. W. E. BuRCHAM, "The Low-Lyipg Excited States of Light Nuclei,"
Progress in Nuclear Physics, O. R. Frisch, ed. New York: Academic Press,
Vol. 2, pp. 174-234 (1952).
56. T. L.~URITSEN, "Energy Levels of Light Nuclei," Annual Review of Nu-
clear Science, Vol. l. Stanford: Annual Reviews Inc., 1952.
57. F. AJZENBERG-8ELOVE and T. LAURITSEN, "Energy Levels of Light
Nuclei," Nuclear Physics 11, 1 (1959).
58. T. LAURITSEN and F. AJZENBERG-8ELOVE, "Energy Levels of the Light
Nuclei," Annual Reviews of Nuclear Science 10, 409 (1961).
59. M. K. BANERJEE, "The Theory of Stripping and Pickup Reactions,"
in Nuclear Spectroscopy, F. Ajzenberg-Selove, ed. New York: Academic Press,
1960. Part B, p. 695.
505
PROBLEMS
l. To which states of the nucleus Mg 25 do the different groups of a-particles
considered in Problem 4 of Chapter 13 belong? Draw a diagram to scale showing
the location of these states.
2. A beryllium target was bombarded with a-particles accelerated to a kinetic
energy of 21.7 Mev in a cyclotron. lt was reported that four groups of protons
were observed corresponding to Q-values of -6.92, -7.87, -8.57, and -10.74
Mev, respectively. (a) What were the energies of the proton groups observed
at an angle of 90º with the incident beam? (b) Which of the Q-values corre-
sponda to the ground state of the product nucleus, and how <loes it compare with
the value calculated from the atomic masses? (e) To which levels of the product
nucleus do the different proton groups correspond? (d) What are the threshold
energies for the excitation of the different states of the product nucleus?
3. The reaction C 13 (d, p)C 14 has a resonance at a deuteron energy of 2.45
Mev. At what a-particle energy <loes this datum lead you to predict a resonance
leve! for the reaction B 11 (a, n)N 14 ?
4. The nucleus N 14 has an excited state at an energy 12.80 Mev above the
ground leve!. At what energy of the incident particle would you expect a reso-
nance to occur in each of the following reactions: (a) B 1º(a, n)N 13 ; (b)
Cl2(d, p)Cl3?
5. A target of boron enriched in the isotope B 1º was bombarded with
deuterons with an energy of 1.510 Mev, and the protons from the reaction
B 1º(d, p)B 11 were analyzed with a magnetic spectrograph. Eleven groups of
protons were observed with the following Hr values, in kilogauss-centimeters:
448,399,338,322,266,264, 245, 190, 171, 156, and 151. Calculate (a) the energy
in Mev of the protons in each group, (b) the Q-value corresponding to each group
of protons, (c) the energy leve! of the product nucleus corresponding to each
proton group.
6. The nuclide F 19 was bombarded with protons of varying energy and the
yields of neutrons from the (p, n) reaction were measured. Resonances were
observed at the following proton energies, in Mev: 4.29, 4.46, 4.49, 4.57, 4.62,
4.71, 4.78, 4.99, 5.07, and 5.20. The threshold energy for the reaction was found
experimentally to be 4.253 Mev. Calculate from the data (a) the mass differ-
ence Ne 19 - F 19 , (b) the endpoint energy of the ,8-decay of Ne 19 , (e) the
excitation energies of the states of Ne 20 which correspond to the resonances.
7. The nuclide Na 23 was bombarded with deuterons in a cyclotron and the
yield of the (d, p) reaction was determined in terms of the deuteron current in
microampere hours (µah) and the radioactivity of the product nucleus. For
each µah, 1500 microcuries of activity were produced owing to the formation of
Na 2 4 , which has a half-life of 15.06 hr. (a) How many particles are there in 1 µah
of deuterons? (b) What was the yield of the reaction in protons per million
deuterons?
8. A gold foil 0.02 cm thick is irradiated for 5 min with a beam of
thermal neutrons with flux of 10 12 neutrons/cm 2-sec. The nuclide Au 198 , with
a half-life of 2.7 days, is produced by the reaction Au 197 (n, 'Y)Au 198 . The
density of gold is 19.3 gm/cm3 and the cross section for the reaction is 98.7
X 10- 24 cm 2. (a) How many atoms of Au 198 are produced per squa.re centimeter
of foil? (b) What is the activity of the foil in millicuries per square centimeter?
9. A tantalum foil 0.02 cm thick and with a density of 16.6 gm/cm3 is ir-
radiated for 2 hours with a beam of thermal neutrons of flux 1012 neutrons/
cm 2-sec. The radionuclide Ta 182 , with a half-life oí 114 days, is formed as a
result of the reaction Ta 181 (n, 'Y)Ta 182, and the foil has an activity of 12.3
rd/cm 2 immediately after the irradiation. Find (a) the number of atoros oí
Ta 182 produced, (b) the cross section for the reaction Ta 181 (n, 'Y)Ta 182 .
10. When the nuclide C 12 is bombarded with slow neutrons, 'Y-raya with an
energy of 4.947 Mev are produced by the reaction C 12(n, 'Y)C 13 . Compare this
energy with the binding energy of the last neutron in C 13 (a) as calculated from
the atomic masses oí the nuclei involved, (b) as deduced írom the Q-value,
2.723 Mev, of the reaction C 12(d, p)C 13 , (e) as deduced from the Q-value,
-0.281 Mev, of the reaction C 12(d, n)N 13 and the endpoint energy 1.200 Mev
of the positron decay of N 13 .
11. The effect of a (d, p) reaction is to add a neutron to the target nucleus.
Show that the binding energy of the last neutron in the product micleus is given
by the sum of the Q-value for the (d, p) reaction and the binding energy of the
deuteron. Q-values of 4.48 Mev, 5.14 Mev, and 1.64 Mev have been obtained
º º º º
for the reactions Pb 2 6(d, p)Pb 2 7 , Pb 2 7(d, p)Pb 208, and Pb 2 8(d, p)Pb2 9,º
respectively. What are the binding energies of the last neutron in Pb 2 07, Pb 208,
and Ph 2 º9 ?
12. A sample of lead was bombarded with thermal neutrons and the capture
'Y-raya were analyzed with a pair spectrometer. Two 'Y-rays were observed, one
with an energy of 6.734 ± 0.008 Mev, the second with an energy of 7.380
± 0.008 Mev. What reactions were responsible for these rays?
13. The difference in mass between Fe 54 and Fe 56 has been shown mass-
spectroscopically to be 1.99632 amu. What is the binding energy of the last two
neutrons in Fe 56 ? How does this value compare with the value obtained from
the following set of nuclear reactions?
Fe 54(n, 'Y)Fe 55 ; Q = 9.28 Mev,

Fe 55(K capture)Mn55 ; Q = 0.21 Mev,
Mn55(n, 'Y)Mn56 ; Q = 7.25 Mev,
Mn 56~-)Fe 56 ; Q = 3.63.
14. The reactions N 14 (n, p)C 14 and C 14(p, n)N 14 show resonances at the
energies listed below.
Nl4(n, p)Cl4 Cl4(p, n)Nl4
495 kev 1163 kev

639 kev 1312 kev
998 kev 1668 kev
1120 kev 1788 kev
1211 kev 1884 kev
PROBLEMS 507
(a) Show that each pair of resonances corresponds to the same excited level
of N 15 • (b} Why is each resonance in the reaction C 14 (p, n)N 14 higher by the
same amount (close to 670 kev) than the corresponding resonance in the
N 14(n, p)C 14 reaction?
15. Is it ever possible to use neutrons from the reaction C 14(p, n)N 14 to
produce the inverse reaction N 14(n, p)C 14 in another target? Explain.
CHAPTER 17
NUCLEAR FORCES AND NUCLEAR STRUCTURE
A large amount of experimental information about atomic nuclei is now

available, examples of which have been given in preceding chapters.
Although the data and their interpretation are consistent with the idea
that nuclei are built up of protons and neutrons, there is as yet no satis-
factory explanation of how nuclei are held together. It has not yet been
possible to analyze the forces which hold neutrons and protons together
and the structure of nuclei is not understood. There is no picture or theory
of the nucleus which can unite the available information about nuclei into
a consistent body of knowledge like that of the quantum-mechanical
theory of atomic structure. It is possible only to show how the problem of
nuclear forces and nuclear structure has been attacked, to point to sorne
successes, and to indicate sorne limitations and dilemmas.
Information about the nature of nuclear forces has been sought in two
ways. The first is through the study of the interaction between two
nucleons in order to discover the nature of the forces between nuclear
particles. The properties of the deuteron and the scattering of protons and
neutrons by protons have been analyzed with this objective. It is hoped
that when the properties of the interactions between nucleons are known,
it will be possible to derive from them the properties of complex nuclei;
this program has so far been only partially successful. The second method
is the study of the properties of complex nuclei and the attempt to deduce
from these properties the nature of the interactions between the particles
which make up the nuclei; this method also has been only partially success-
ful. Between the two methods, however, something has been learned about
nuclear forces, and this information can be used to suggest models of the
nucleus. Although no single model can account for the complete range of
nuclear properties, there are severa! which are useful.
In this chapter, the following procedure will be adopted. It will first
be shown how sorne qualitative and semiquantitative ideas about nuclear
forces and structure can be deduced from the properties of complex nuclei.
The analysis of the interactions between nucleons will then be discussed in
order to get sorne quantitative information about nuclear forces. Finally,
sorne nuclear models and their applications will be discussed.
17-1 Nuclear binding energies and the saturation of nuclear forces.

It has been shown (Table 9-3 and Fig. 9-11) that the average binding
energy per particle in a nucleus has approximately a constant value for all
508
17-1] NUCLEAR BINDING ENERGIES 509
but the lightest nuclei. In other words, the total binding energy of a
nucleus is nearly proportional to the number of particles in the nucleus.
Now, if every particle in the nucleus is supposed to interact with every
other particle, the interaction energy, and therefore the binding energy,
should be approximately proportional to the number of interacting pairs.
Since each of the A particles in the nucleus would interact with (A - 1)
other particles, the number of interacting pairs would be A(A - 1)/2,
and the binding energy would be proportional to this quantity. In heavier
nuclei, A may be neglected in comparison with A 2 and the binding energy
would be proportional to the square of the number of particles in the
nucleus. This calculated result is in sharp contrast with the experimental
result that the total binding energy is nearly proportional to A. To get
out of this difficulty, it is necessary to assume that a nuclear particle <loes
not interact with all the other particles in the nucleus, but only with a
limited number of them. This situation is analogous to that in a liquid
or solid, in which each atom is linked by chemical bonds to a number of
nearest neighbors rather than to all the other atoms, and the chemical
binding energy is practically proportional to the number of atoms present.
The consideration of different types of chemical binding leads to the
conclusion that the best analogy to nuclear forces is represented by horno-
polar binding like that of the hydrogen molecule. There is a strong attrac-
tion between two hydrogen atoms to form the molecule H 2 , but a third
hydrogen atom cannot be bound strongly to a hydrogen molecule. The
hydrogen molecule is said to be saturated. An assembly of many hydrogen
atoms, as in a drop of liquid hydrogen, has a chemical binding energy
approximately equal to that of the corresponding number of hydrogen
molecules and is, therefore, approximately proportional to the number of
hydrogen atoms present.
The correct dependence of nuclear binding energies on the number of
particles in the nucleus can be obtained, then, if it is assumed that the
forces between nucleons show saturation in much the same way as the
forces of homopolar chemical binding. Furthermore, if a particular nuclear
particle <loes not interact with all the other particles in the nucleus but
only with sorne of its neighbors, then nuclear forces must have a short
range. The force, or interaction energy, between two nucleons must fall
off rapidly as the particles are separated; molecular forces also behave in
this way. In a nucleus with many particles, a given nucleon can interact
only with sorne of its neighbors and not with nucleons which are relatively
far from it. The range of the specifically nuclear forces must consequently
be smaller than the radius of any but the lightest nuclei.
The above discussion shows how certain properties of nuclear forces,
namely, their saturation property and their short range, can be deduced
from the variation of the binding energy per particle with the number of
510 NUCLEAR FORCES AND NUCLEAR STRUCTURE [CHAP. 17
particles in the nucleus. In this discussion, the repulsive electrostatic

forces between the protons have been neglected. The magnitude of the
Coulomb energy between two protons is simply e2 /r, where e is the charge
on a proton, 4.8 X 10- 10 esu, and r is the distance between the protons.
A rough idea of the magnitude of this energy can be found by taking
r = 3 X 10- 13 cm. Then
(4.8) 2 X 10- 20 ~
r (3 X IQ-13)(1.60 X 10-6) = 0.5 Mev.
The Coulomb energy between two protons is therefore small compared
with the average binding energy per particle, which is about 8 Mev.
Despite its smallness, the Coulomb repulsion becomes important for
heavier nuclei because of the saturation of the attractive nuclear forces.
For the Coulomb force shows no saturation, and the total energy of the
Coulomb interaction is proportional to the number of proton pairs in the
nucleus, ½Z(Z - 1). The total Coulomb energy has been shownw to be
3 e2
Total Coulomb energy = Z(Z - 1) R , (17-1)
5
where R is the nuclear radius. Since R ~ A 113 andZ is rougbly proportional
to A, the Coulomb energy is roughly proportional to A 513 . But the total
binding energy is proportional to A, so that the relative importance of the
repulsive electrostatic energy increases with increasing mass number,
roughly as A 213 .
One result of the increased importance of the Coulomb energy for nuclei
with greater Z (or A) is to decrease the binding energy per particle as the
mass number increases. Figure 9-11 shows that in the neighborhood of
Z = 50 or A = 135, the binding energy per particle starts to decrease.
For very heavy nuclei, with A gres.ter than 200, the effect is strong enough
so that sorne of tbese nuclei are unstable with respect to a-disintegration.
17-2 Nuclear stability and the forces between nucleons. The informa-
tion about stable nuclides contained in Table 9-1 and Fig. 9-9 allows
certain deductions to be made about the forces between pairs of nucleons.
The mass number A is approximately equal to twice the nuclear charge Z.
This relationship is particularly accurate for the light nuclei; for heavier
nuclei, the value of A increases more rapidly than 2Z. Since A is the total
number of particles in the nucleus and Z is the number of protons, the
number of neutrons is very nearly equal to the number of protons in the
light nuclei, but in tbe heavier nuclei the number of neutrons increases
more rapidly than the number of protons. The equality between the num-
bers of protons and neutrons in the light nuclei can be interpreted as
17-2] NUCLEAR STABILITY AND FORCES BETWEEN NUCLEONS 511
showing that there is a strong attractive force between a neutron and a

proton. The stability of the deuteron, which is made up of a neutron and
a proton, supports this conrlusion.
If there is a strong specifically nuclear attractive force between a neutron
and a proton, it is reasonable to suppose that there are similar forces
between two neutrons and between two protons. lf so, the force between
two neutrons must be very nearly equal to that between two protons,
apart from the electrostatic force between two protons. For, if the attrac-
tion between two neutrons were greater than that between a neutron and
a proton or between two protons, the most abundant stable light nuclei
would contain more neutrons than protons rather than equal numbers of
these particles. A similar argument shows that the attractive force be-
tween two protons cannot be greater than that between a neutron and a
proton, or between two neutrons. The existence of a force between two
neutrons may be inferred from the way in which the neutron number in-
creases with the proton number. As the Coulomb energy increases, the
ratio of the number of neutrons to that of protons increases gradually
from 1 for the light nuclei to 1.6 for uranium, with Z = 92. This increase
may be interpreted as a necessary increase in the total attractive forces
in the nucleus needed to overcome the effects of the electrostatic repulsion.
Since each nuclear particle interacts only with a limited number of other
nucleons, it is reasonable to suppose that sorne of the neutrons interact
with other _neutrons.
It will be assumed, in view of the preceding discussion, that there are
specifically nuclear, attractive forces between a neutron and a proton,
between two neutrons, and between two protons. These forces are abbre-
viated as n-p, n-n, and p-p, respectively. The n-n and p-p forces are
assumed to be equal or very nearly equal in magnitude and to be not
greater than the n-p force; they may, however, be much smaller than the
n-p force, or they may have the same, or very nearly the same, magnitude
as the n-p force. Either possibility would be consistent with the observed
mass-to-charge ratio of the stable nuclides. The two possibilities can be
expressed in the form
(a) n-p ..,. n-n ..,. p-p,
or
(b) n-p » n-n; n-p » p-p; n-n ..,. p-p.
Wigner, <2 , 3 , 4 > in a general treatment of nuclear energy levels, made the
assumption that the relationship described by case (a) holds, that is, that
the specifically nuclear n-p, n-n, and p-p forces are equal. This assump-
tion is known as the hypothesis of the charge independence of nucl,ear forces.
The alternative assumption that the n-n and p-p forces are equal, but not
necessarily equal to the p-n force, is known as the hypothesis of the charge
symmetry of nuclear forces. It is less restrictive than the hypothesis of
charge independence.
17-3 Energy levels of light nuclei and the hypothesis of the charge
independence of nuclear forces. The hypothesis of the charge symmetry
of nuclear forces can be tested by studying the energy levels, especially the
ground states, of certain light nuclei. According to this hypothesis, the
difference in the total energy between the ground states of a pair of mirror
nuclei (isobars with a neutron excess of ±1) should be accounted for solely
by Coulomb repulsion and the difference in mass between neutron and
proton. The simplest mirror nuclei are H 3 and He 3 . The former contains
1 proton and 2 neutrons, or two n-p bonds and one n-n bond; He 3 contains
2 protons and 1 neutron, or two n-p bonds and one p-p bond. The binding
energies of H 3 and He 3 , calculated from the data of Table 11-1, are
B.E.(H 3 ) = 8.482 Mev; B.E.(He 3 ) = 7.711 Mev.
The difference in binding energies, 0.771 Mev, is small compared with

either binding energy, so that the n-n force in H 3 is not very different from
the p-p force in He 3 • The small difference between the binding energies is
attributed to the Coulomb repulsion between the two protons in He 3 , since
there is no Coulomb force in H 3 • The effect is in the right direction, since
He 3 is less strongly bound than H 3 . If He 3 is assumed to be a sphere of
radius Re, its Coulomb energy 0 > is given by
e2
Coulomb energy of He 3 = 56 Re = 0.771 Mev. (17-2)
This equation can be sol ved for Re, with the result Re = 2.24 X 10- 13 cm,
which is a reasonable value for the "radius" of a light nucleus such as He 3 •
The hypothesis that the n-n force in H 3 is equal to the p-p force in He 3 is,
therefore, consistent with the known information about these nuclei.
The positron-emitters, zXA, with A = 2Z - 1, for which the parent
and daughter nuclei are mirror nuclei, provide more evidence for the
equality of the n-n and p-p forces. In each pair of mirror nuclei, the parent
has one more proton than neutron, and the daughter has one more neutron
than proton. The number of n-p bonds is presumably the same in parent
and daughter, but the parent nucleus should have one more p-p bond and
one less n-n bond than the daughter. If the n-n and p-p forces are equal,
it should be possible to calculaté the difference in the ground-state energy
(atomic or nuclear mass) from the simple formula
tlE (cale) = áE (Coulomb) - (n - H 1 ), (17-3)

17-3) ENERGY LEVELS OF LIGHT NUCLEI 513
where !!,,.E (Coulomb) is the difference between the Coulomb energies uf the
pair, and (n - H 1) represents the energy equivalent of the difference in
mass between a neutron and a proton. If Z is the nuclear charge of the
parent nucleus, then the difference in energy is given by
3 e2
1) - (Z - l)(Z - 2)] - (n - H 1).
5 R [Z(Z -
!!,,.E (cale) = (17-4)
The radius R is taken to be the same for the parent and daughter nuclei
because the mass number is the same and the difference in the nuclear mass
should have a negligible effect on the radius. The value of R used in
calculating !!,,.E comes from the formula
R = 1.42A 113 X 10- 13 cm. (17-5)
The constant 1.42, obtained from measurements of the scattering of fast

neutrons by nuclei, is used because those measurements iricluded work on
light nuclei; it is, therefore, more suitable than the value obtained from
the decay of heavy radionuclides. The value of !!,,.E (cale) should be the
same as the experimental value of the positron decay energy if the n-n
and p-p forces are equal as assumed.
Equation (17-4) can be written
2
!!,,.E (cale) = 56 eR (Z - 1) - (n - H 1 )
2
6 e (Z - 1) 1
=5 (l.42)AI13lQ-13 - (n - H ). (17-6)
When numerical values for e2 and (n - H 1) are inserted, and AE is ex-

pressed in Mev, the result is
6 (4.80) 2 X 10- 20 (Z - 1)
!!,,.E (cale) =5 (1.42)10-13(1.60)10-6 Atta - º· 78 Mev,
or
(Z - 1)
!!,,.E (cale) = 1.22 AVa - 0.78 Mev. (17-7)
The results of the calculations are given in Table 17-1, together with the
experimental values of AE; the agreement is good, and the results are con-
sistent with the assumption that the n-n and p-p forces are equal.
Although the mirror nuclei provide evidence in support of the hypothesis
of the charge symmetry of nuclear forces, they tell nothing about the
magnitude of the n-p force as compared with the n-n and p-p forces. There
are light nuclei whose energy levels do yield such information. Consider,
for example, the three isobars Be 10 , B 1º, and C 10 , which contain 4 protons
514 NUCLEAR FORCES AND NUCLEAR STRUCTURE [CHAP, 17
TABLE 17-1
ENERGY D1FFERENCE BETWEEN MrnROR NuCLEI
Charge of Cube root Energy difference

Parent Daughter
nucleus nucleus parent of mass AE (cale), AE (exp),
nucleus, Z number Mev Mev<73,74J
en Bll 6 2.22 1.95 1.98
Nl3 Cl3 7 2.35 2.33 2.22
Ql5 Nl5 8 2.47 2.68 2.76
Fl7 011 9 2.57 3.02 2.77
Nel9 Fl9 10 2.67 3.33 3.26
Na21 Ne21 11 2.76 3.64 3.52
Mg23 Na23 12 2.84 3.94 4.06
Af25 Mg2s 13 2.92 4.24 4.26
s¡21 AI27 14 3.00 4.51 4.78
p29 Si29 15 3.07 4.77 4.98
sa1 p31 16 3.14 5.05 5.41
CI33 Sªª 17 3.21 5.30 5.53
Aª5 CI35 18 3.27 5.57 5.95
Caª9 Kª9 20 3.39 6.07 6.45
Sc41 Ca41 21 3.45 6.29 5.96
and 6 neutrons, 5 protons and 5 neutrons, and 6 protons and 4 neutrons,

respectively. Because of the equality of p-p and n-n forces, Be 1º and C 1º
should correspond in much the same way that mirror nuclei do; their
ground-state energy difference should be obtained from the difference in
Coulomb energy and the energy equivalent of the dift'erence in mass
between two neutrons and two protons. The calculated and observed
energy differences do indeed agree, as has been shown by a calculation
similar to that made for the mirror nuclei. <5> This kind of argument can
be extended to B 1 º, but only if the n-p force is equal to the n-n and
p-p forces, since the number of n-p bonds in B 1 º should differ from the
number of n-p bonds in Be 16 and C 10 • The theoretical treatment is more
complicated than that for the mirror nuclei and will not be given here. It
turns out that a certain excited state of B 10 should bear a close relation to
the ground state of Be 10 and C 10 ; there is reasonably good evidence that
this excited state does exist. An analogous result is obtained for the triad
C 14-N 14-0 14 , and these results provide evidence for the charge independ-
ence of nuclear forces, that is, that p-p = n-n = n-p. Although the evi-
dence from energy levels is not as clear as that for the equality of p-p and
n-n forces, a. more detailed survey<&> of the known energy levels of light
isoba.rs indica.tes that the assumption of the charge independence of the
17-3] ENERGY LEVELS OF LIGHT NUCLEI 515
specifically nuclear forces is a good approximation to the true state of

affairs within the nucleus.
The hypothesis of the charge independence of nuclear forces has proved
increasingly fruitful: It has led, for example, to the introduction of a new
quantum number T, the isotopic or isobaric spin quantum number, which
has helped in the understanding of nuclear energy levels and nuclear reac-
tions. <2 , 4 ,s> If nuclear forces are, in fact, charge independent, that is, if
the forces between nucleons are the same for neutrons and protons, we
may regard a nucleon as a single entity having two states, the proton and
the neutron. The charge of a nucleon may then be treated as a variable
anda new variable, the isotopic or isobaric spin T, may be introduced for
nucleons. The adjective isotopic or isobaric comes from the fact that the
variable refers to protons and neutrons, and "isobaric" is preferable
because the proton and neutron are isobars, not isotopes; the term "i-spin"
is often used for convenience. The new variable has a role similar to that
of the electron spin variable u, which is the reason for the "spin" part of
the name. The isobaric spin variable T distinguishes between protons and
neutrons justas the electron spin variable u is used to distinguish electrons
with spins parallel to a given axis from electrons with spins in the opposite
direction. Thus, T may be assigned the value +½ for a neutron and -½
for a proton, and the charge of a nucleon may be expressed by (½ - T).
The quantum number T bears the same relation to the isobaric spin vari-
able T that the spin quantum number S does to the electron spin variable u.
In the atomic case there are 2S + 1 independent states belonging to the
same quantum number S, with each state having a different value of the
z-component, s., of the spin angular momentum. Similarly, there are
2T + 1 independent nuclear states associated with the isobaric spin
quantum number T, each state belonging to a different value of the quan-
tity Tr, where Tr = T, T - 1, ... , - T. The quantity Tr may be
called the "component of T in the direction of positive charge." In analogy
with s., which is given for an atom by the sum of the S.'s of the individual
electrons, Tris given for a nucleus by
Tr = ¿T,,
i
where the sum is over all the nucleons. Hence, if the nucleus has N neu-
trons and Z protons,
Tr = ½(N - Z), (17-8)
or just half the neutron excess of the nucleus. The values of Tr associated
with a particular T describe states of severa! isobaric nuclei, and all of
these states depend in the same way on the space and ordinary spin
variables of the nucleons. If the forces are the same for neutrons and pro-
/ / /
_L e'º
-,
1 M1v
Fm. 17-1. Energy levels of the isobara Be 10, B 1 º and c10_< 76>
tons, these states should have the same energy (apart from their Coulomb
energy and neutron-proton mass difference) and are said to form a T
multiplet. Two states of a T multiplet differ only in that neutrons are
substituted in one of these states for sorne of the protons in the other state,
or vice-versa; and the total angular momentum is the same for ali states
of a T multiplet, as is the parity.
An example of the application of the concepts oí charge independence
and isobaric spin to the energy levels of light nuclei m is shown in Fig.
17-1. The level diagrams for the nuclides Be 1 º, B 1 º, and C 1 º are grouped
into an isobaric set to show the correspondence of levels comprising iso-
baric spin multiplets. The relative positions of the ground states in each
set have been adjusted to the extent that the first order Coulomb energy
differences and the n-H mass differences have been removed. Levels for
which the correspondence seems well established are connected by dashed
lines.
17-4] THE NUCLEAR RADIUS 517
The concept of the T multiplet has been applied to /3-decay, 'Y-decay,

and to nuclear reactions. There are selection rules involving the permis-
sible change in T in /3-decay, in 'Y-decay, and in nuclear reactions, and
there is increasing evidence that total isobaric spin is conserved in nuclear
reactions<s> in a manner analogous to the conservation of total nuclear
angular momentum. The success of these applications supplies additional
support for the hypothesis of the charge independence of nuclear forces.
17-4 The nuclear radios. The treatment of the Coulomb energy in the
last section is an approximate one, based on elementary ideas about the
size and shape of the atomic nucleus. These ideas have had to be modi-
fied, because of both theoretical and experimental work, <9 • 10> although
the conclusions concerning the charge independence of nuclear forces are
not affected. It was assumed in the last section that the nuclear charge is
spread uniformly over a sphere of radius R, with R obtained from experi-
ments on nuclear reactions; certain quantum-mechanical effects were also
neglected. More accurate calculations have been made in which the ex-
perimental data on mirror nuclei were used to determine the nuclear
radius.<11l The radius may still be expressed by the relation
(17-9)
but with Ro = (1.28 ± 0.05) X 10- 13 cm. These results are in good
agreement with those of precise calculations of nuclear Coulomb ener-
gies. <12 > Nuclear radii have also been determined experimentally with
methods which yield information about the electric charge distribution.
For example, electrons interact very little with nuclei through specifically
nuclear forces and are affected only by the electric charge distribution of
the nucleus. Experiments on the scattering of fast electrons by nuclei
can consequently be interpreted 03 > to give the radius of the equivalent
sphere of electric charge with the result
R = (I.2 ± 0.1) X 10- 13A 113 cm.
Similar results have been obtainedu 4 > from experiments on the interaction
between nuclei and µ-mesons (see Section 21-4).
The values for nuclear radii obtained by means of these methods are
somewhat smaller than those obtained from experiments involving nuclear
forces, e.g., from cross sections for the scattering of neutrons or protons
or from a-radioactivity. The former methods give information about the
size and shape of the charge distribution while the latter give information
about the nuclear potential energy distribution. The two distributions
are not necessarily the same, and should not be confused in precise work.
In particular, the assumption of a uniform spherical charge distribution

seems to be an oversimplification.
The extent of recent work in this field and the scope of the results is
indicated by the material in References 15 and 75.
17-5 The interaction between two nucleons. It has been shown in the
earlier sections of this chapter how qualitative information about nuclear
forces can be obtained from the properties of complex nuclei. Quantitative
information can best be obtained from the experimental and theoretical
investigation of phenomena involving only two nucleons, and two-nucleon
problems are, therefore, the basic problems of nuclear physics. These
problems include the scattering of protons by protons, which gives informa-
tion about the strength and the range of the p-p forces, the scattering of
neutrons by protons, the properties of the deuteron, and the capture of
neutrons by protons, ali of which give information about the strength and
range of the n-p force. The force between two neutrons cannot be studied
directly because there is no stable nucleus which consists only of neutrons,
and because experiments on the scattering of neutrons by neutrons are not
practical. The properties of complex nuclei are the only source of informa-
tion about the n-n force.
The analysis of the experimental results and the theory are both com-
plicated, involving the quantum-rnechanical theory of collision processes
and of the deuteron. These problems are treated in sorne of the books
listed at the end of the chapter and in review articles 06 • 17 • 18> and will not
be discussed in detail here. Instead, the ideas underlying the experiments
and theory will be sketched and sorne of the results will be indicated. We
start with the problem of the deuteron.
The successful application of wave mechanics to a-radioactivity showed
that nuclear problems as well as atornic problems should be understandable
in terms of wave rnechanics. There is, however, a serious difference between
atornic and nuclear problems; in the former the force between the posi-
tively charged nucleus and an electron is known to be derived from the
Coulomb electrostatic potential; in the latter the details of the force law
are unknown. In the case of the deuteron, the simplest nucleus with more
than one nucleon, certain facts are known, and it is necessary to seek a.
force law, or potential between the proton and neutron, which will a.ccount
for the facts. The most iroportant experimental ha.sis for the theory of
the deuteron is its binding energy, 2.23 Mev; in addition, the spin, roag-
netic moment, and qua.drupole rnoment are known. The Schroedinger
equation for the deuteron is
2m
2
V i/¡ + fi2 [E - V]i/1 = O, (17-10)
17-5) THE INTERACTION BETWEEN TWO NUCLEONS 519
where V2 is the Laplacian second order differential operator which is, in

rectangular cartesian coordinates,
a2 a2 a2
v2 = ax2 + ay2 + ¡}z2 ;
m is the reduced mass,
m = n
+
M "'" -1 M
M MnM
11
JI 2
(of proton or neutron) . (17-11)
The quantity E is known; for the ground state of the deuteron it is negative
and numerically equal to the binding energy. The potential V is not
known and the procedure used is to start with sorne rea.sonable assumed
function and modify it as necessity requires, keeping it as simple as the
experimental fa.eta permit. The simplest assumption that can be ma.de
about the force between the neutron and proton is tha.t it is a central force,
acting a.long a line joining the two particles, and tha.t it can be derived
from a potential V(r), where r is the distance between the particles. Since
the force is attractive, V(r) is negative and decrea.ses with decrea.sing r.
The wave function is then a function of r only, and ma.y be written as
1/t(r). It is convenient to use spherical coordina.tes in the Schroedinger
equation. The lowest, most sta.ble, quantum state is tha.t for which the
angular momentum l = O (the S-state). There is then spherical symmetry
and terms which involve the polar and azimutha.l angles drop out so that
the Schroedinger equation for the deuteron becomes
dir) + ~ d~~) + !': [E - V(r)]i/t(r) = O. (17-12)
b r-
o r-----------------------
1
1
1
1
L....J
-Vo
F10. 17-2. Attractive potential between a neutron anda proton as a function

of the distance between the two particles. The dotted lines represent the square-
well potenti&l.
Equation (17-12) can be simplified by making the substitution
..J,(r) = u(r), (17-13)

r
with the result
d 2u(r) M
dr2 + h 2 [E - V(r)]u(r) = O. (17-14)
Various potentials can be used subject, of course, to the requirement that

the force has short range. The potentials which have been used most
frequently are
(a) square well: V(r) = - V0 (r< b)
=O (r > b)¡
(b) exponential: V(r) = - V oe-•;
(c) GauBBian: V(r) = - V oe_,.1 ' 1 ;
(d) Yukawa: V(r) = -V 0 (e-,.'lµr).
The simplest of these is the square well shown in Fig. 17-2; the magnitude
of the potential is assumed to be significantly large only within the short
range O < r < b, and to be zero for r > b. This kind of potential was
used in the theory of a-decay (Section 13-6) to describe the interaction
between the a-particle and the residual nucleus except that a repulsive
(positive) potential was used there. The binding energy alone does not
permit the determination of both the magnitude and the range, but gives
only a relationship between these two quantities. The result of the analysis
may be written
(17-15)
lf the range bis assumed to be 2 X 10- 13 cm, a reasonable value for the
deuteron, then V O would be 37 Mev. The other short-range potential
functions give about the same results as the square well.
The exa.ct solution of the problem of the deuteron is complicated by
several factors. The n-p potential does not depend on r alone, since the
deuteron is known to have a quadrupole moment (see Section 8-5) and is,
therefore, not spherically symmetrical. The potential depends also on the
relative orientation of the spins of the neutron and the proton. In fact, a
detailed analysis of the ground state of the deuteron shows that a neutron
and a proton can form a stable deuteron only if their spins are parallel.
This stable state is known as the triplet state, in contrast to the unstable
singlet state in which the spins are antiparallel. To account for the shape
(quadrupole moment) of the deuteron, it is necessary to postulate that the
17-5] THE INTERACTION BETWEEN TWO NUCLEONS 521
n-p potential depends on the angle between the spins and the line of con-
nection of the two particles. The n-p force is consequently noncentral or,
more precisely, not completely central. Finally, the analysis of the deu-
teron alone does not give all of the desired information about the n-p
force, although it does provide useful clues to the solution of the problem.
More detailed information about the n-p force is obtained from the
analysis of the scattering of neutrons by protons. In sorne typical scatter-
ing experiments, n 9 , 2 º• 2 1> cyclohexane, C 6 H 12, was bombarded with
neutrons of known energy, and the transniission was measured in a "good
geometry" experiment. The neutron transmission is given by
1/10 = e-"• 1"',
where <1 1 is the total cross section in cm 2, n is the number of scattering
nuclei per cubic centimeter, and x is the thickness of the scatterer in cm.
Cyclohexane contains carbon in addition to hydrogen, and the effect of
the carbon was determined by doing separate transmission measurements
on a carbon target. The total cross section per atom of hydrogen is then
(17-16)
The absorption cross sections of hydrogen and carbon are negligible com-
pared with the respective scattering cross sections in the neutron energy
ranges studied (0.35 to 6.0 Mev), so that the total cross section can be
taken equal to the scattering cross section.
The measured cross sections can be analyzed by methods which lead in
a not very direct way to information about the properties of the n-p po-
tential. The difficulties mentioned in connection with the problem of the
deuteron are met also in the scattering problem. At neutron energies less
than 10 Mev, which are relatively low for scattering experiments of this
kind, the range of the n-p potential turns out to be close to 2 X 10- 13 cm
for the triplet potential. For a square-well potential, the corresponding
depth is 35 Mev. The results for the singlet potential are 2.8 X 10- 13 cm
for the range, and 11.8 Mev for the depth. It has been found, however;
that for neutron energies less than 10 Mev, the details of the shape of the
potential cannot be obtained from neutron-proton scattering experiments,
in spite of the difficult and complicated analytical methods used. The
results of the experiments are now analyzed by means of a theory called
the ejfective range theory which yields two quantities called the effective
range and the scaUering length. These quantities represent average proper-
ties of the potential and determine the scattering cross section; they are
related to both the range and magnitude of the potential, and are used
becauoo they make the analysis of experimental data easier. In particular,
they allow calculations to be made on the basis of very general assumptions
about the nature of nuclear potentials. Both quantities have different
522 NUCLEAR FORCES AND NUCLEAR STRUCl'URE [CHAP. 17
values in the triplet and singlet cases. Recent results for the effective
range are, for the n-p forces,
triplet effective range = (1.70 ± 0.03) X 10- 13 cm,
(17-17)
singlet effective range = (2.4 ± 0.4) X 10- 13 cm.
Values for these quantities can also be obtained from other measure-
ments. An estimate of the triplet effective range based on measurements
of the cross section for the photodisintegration of the deuteron< 22 , 23 > gives
a result in good agreement with the value 1.70 X 10- 13 cm. The cross
section for the capture of thermal neutrons by protons< 24> leads to a value
of 2.5 X 10- 13 cm for the singlet effective range.
Recent values for the scattering lengths are
triplet scattering length = 5.38 X 10- 13 cm,
(17-18)
singlet scattering length = -2.37 X 10- 12 cm.
There is only one proton-proton force, or potential, corresponding to
antiparallel spins. This singl.et interaction is a consequence of the Pauli
exclusion principie (Section 7-6) according to which two similar particles
cannot have the same quantum numbers. Two protons can interact
strongly only when their quantum states are the same except for the spins,
which must be antiparallel. The hypothesis of the charge independence
of nuclear forces would then require that the values of the effective range
and scattering length_for the p-p potential be the same as those for the
singlet n-p potential.
Many proton-proton scattering experiments bave been done with protons
of relatively low energy (10 Mev or less).< 16> Well-collimated, mono-
energetic beams of protons can be obtained from a Van de Graaff electro-
static generator. The protons are scattered by the protons in a sample of
hydrogen gas, and the scattered particles are detected at different angles by
means of an ionization chamber. The cross section for the proton-proton
scattering process can be determined as a function of the energy of the
incident particles. The differences between p-p scattering and the Ruther-
ford sca.ttering of a-particles by nuclei a.re interesting. In the latter case,
the measured scattering is caused by the Coulomb repulsion between the
a-particle and the nucleus. The repulsive forces are very large and devia-
tions from the Coulomb force law are observed only for highly energetic
particles and light nuclei. The occurrence of deviations points to the
existence of specifically nuclear forces. In p-p scattering, the Coulomb
potential is relatively small, and the bombarding particles are scattered
in a way which dependa on the combined effect of the specifically nuclear
forces and the Coulomb forces. The effect of the Coulomb forces must be
17-6) THE STATUS OF THE PROBLEIII OF NUCLEAR FORCES 523
subtracted from the total effect to obtain the desired information ahout
the nuclear p-p force.
The resulta of the scattering experiments can be analyzed by methods
similar to those used for n-p scattering. Additional complications are
introduced into the theory by the Coulomb repulsive force between the
protons and by the fact that the bombarding and target particles are of
the same kind. When the experimental data are analyzed in terms of the
range and depth of the potential which fit the measured values of the cross
sections, the results agree quite well with those obtained for n-p scattering.
Thus, a square-well potential with a range of 2.8 X 10- 13 cm would have
a depth of 11.35 Mev as compared with a singlet n-p depth of 11.8 Mev for
the same range. If other shapes of the potential well are used, equally
good agreement is obtained between the strengths of the potential needed
for n-p and p-p forces. These results, together with those of Section 17-3
on the equality of the n-n and p-p forces (apart from the Coulomb effect),
provide experimental evidence for the validity of the assumption of the
charge independence of nuclear forces.
Justas in n-p scattering, p-p scattering results below 10 Mev do not give
details of the shape of the potential. The effective range theory has also
been applied to p-p scattering, and gives the results
effective p-p range = (2.65 ± 0.07) X 10- 13 cm,
{17-19)
p-p scattering length = (-7.67 ± 0.5) X 10- 13 cm.
The value for the effective range agrees, within the stated limits of error,
with the value for the singlet n-p effective range [Eq. {17-17)]. The
scattering lengths are quite different in the two cases, a result which seems
to imply that the depths of the potentials are significantly different in the
two cases. lt has been shown, however, that there are magnetic interac-
tions between the particles, and that when these are taken into account, the
p-p and singlet n-p scattering lengths agree.
The results of the scattering work can be summarized, then, by the
statement that at relatively low energies (less than 10 Mev) the nuclear
potentials for the singlet n-p interaction and the p-p interaction are
identical, in agreement with the hypothesis of the charge independence of
nuclear forces.
17-ó The status of the problem of nuclear forces. The information

collected in the last section about the interaction between two nuclear
particles at low energy can be summarized as follows: there exists a force
between nuclear particles which is attractive (except for the electro-
static repulsion between two protons), and which has a range of about
2 X 10- 13 cm. The force is spin dependent, being different in the parallel
spin (triplet) and antiparallel spin (singlet) cases. It is not cornpletely

central, but depends to sorne extent on the angle between the axis of the
spins and tl,i.e line between the two particles. The nuclear force does not
seern to depend on the type of nucleon; the sarne force can be assurned
between any pair of nucleons, two protons, two neutrons, or a neutron and
a proton, without corning into conflict with any low-energy result. In
other words, the nuclear forces seern to be charge independent at low
energies.
Unfortunately, the information just listed is incornplete and leads to
serious problerns, sorne of which have not yet been resolved. One of these
problerns has to do with the saturation of nuclear forces. The study of
cornplex nuclei has shown that the binding energy per nucleon is approxi-
rnately constant and it has been deduced frorn this fact (Section 17-1) that
nuclear forces rnust have the property of saturation. It has also been
found that the nuclear radius is proportional to the cube root of the nuclear
rnass, that is, the nuclear volurne is proportional to the number of particles
in the nucleus, a result which also irnplies the saturation of nuclear forces.
lt turns out, however, that if the nuclear forces are attractive, short-range
forces, they cannot have the saturation property required by the binding
energies and nuclear radii. A study of the potential and kinetic energy of
the nucleus as a function of the radius has shown< 25> that under the action
of purely attractive forces, the nucleus would be stable only if its radius
were about half as great as the range of nuclear forces. Each nucleon
would then interact with every other nucleon and the binding energy
would be proportional to A 2 , in contradiction with the observed fact that
the binding energy is proportional to A. Nuclear radii would be smaller
than the values obtained from experiment.
The difliculty has been partially resolved by assuming that the force, or
part of the force, between two nucleons is sometimes attractive, sometimes
repulsive, depending on the state of the two nucleons with respect to each
other. A force of this kind is of a more general type than those used in
ordinary atomic physics, and it has been necessary to broaden the concepts
of potential energy and force in order to apply them to the nucleus. A clue
to the kind of generalization needed was supplied by the treatment of the
molecular ion Ht, which consists of two protons and one electron. To
account for the properties of H t, a force was needed which is a mixture of
ordinary and exchange terms. The exchange terms arose from the possi-
bility that the electron could exchange positions between the protons. It
was assumed that the electron could jump from one proton to the other,
and that the H! ion could be regarded as consisting of a hydrogen atom
and a proton, which can change their positions. The possibility of this
exchange introduced new terms into the mathematical expressions for the
potential and the force, and the new contribution to the force is known
17-6] THE STATUS OF THE PROBLEM OF NUCLEAR FORCES 525
asan exchange force. Exchange forces have the property of being attractive
or repulsive depending on the states of the particles involved, and it was
postulated that forces of this kind exist between nucleons. Although there
is not yet a satisfactory explanation for the existence of nuclear exchange
forces, these forces can be treated mathematically and lead to the correct
saturation properties. The introduction of exchange forces does not affect
the low-energy interaction between two nucleons, and removes the con-
tradiction with the saturation requirement which follows from the assump-
tion of purely attractive nuclear forces.
Although exchange forces cannot be described accurately in terms of
elementary ideas, a simple, if not entirely correct, picture can be given of
how they lead to saturation. Immediately after the discovery of the neu-
tron (1932), Heisenberg suggested that when a neutron interacts with a
proton the single electric charge jumps from one nucleon to the other so
that, in the first such jump, the original proton changes into a neutron and
the neutron into a proton. In this way, a neutron may interact with only
one proton at a time, and the n-p force has the property of saturation.
After Fermi proposed his theory of ¡'.3-decay, these ideas were extended.
According to this theory, a neutron can change into a proton by emitting
an electron and a neutrino, and a proton can change into a neutron by
emitting a positron and a neutrino. Heisenberg suggested that nuclear
forces are exchange forces in which electrons (or positrons) and neutrinos
are exchanged between two nucleons. The strength of these nuclear forces
could then be calculated from the observed probability of ¡'.3-decay. Un-
fortunately, when these calculations were made, the nuclear forces that
should result from the electron-neutrino interchange were weaker by a
factor of about 10- 14 than those required by a square-well potential with
a range of the order of 10- 13 cm. Although the theory was pretty, it did
not work.
After the failure of this theory, Yukawa (1935) predicted the existence
of a particle, now called a meson, having a rest mass between the electron
and nucleon masses. He proposed a theory of nuclear forces which involves
these mesons and is referred to as the meson theory of nuclear forces. The
discovery of such a particle in the cosmic radiation stimulated a great
deal oí experimental and theoretical research in the fields of nuclear theory,
cosmic rays, and high-energy particle physics. Several different mesons
with different charges and rest masses have been discovered, and many
attempts have been made to fit them into a consistent scheme of nuclear
forces. These attempts, however, have not yet succeeded in reproducing
quantitatively the known properties of nuclear forces.
Another serious problem has to do with the properties of nuclear forces
deduced from nucleon scattering experiments at high energy. It has been
emphasized that the hypothesis of the charge independence of nuclear
forces seems to be valid at uno energies. It has also been mentioned that
the low-energy data do not give definite information about the detailed
properties of' the n-p and p-p potentials. It was expected on theoretical
grounds that the results of scattering experiments at high energies would
give results which are very sensitive to the details of the interaction, its
dependence on the distance between the particles, and to its exchange
properties. The development of high-energy particle accelerators has
made it possible to extend the energy range of neutron-proton and proton-
proton scattering to energies up to 400 Mev.
Severa! difficulties have appeared as a result of the work on high-energy
scattering, and the data available at present have led to confusion rather
than to clarity. The proton-proton scattering at high energies shows sorne
features which are quite different from the neutron-proton scattering. For
example, the p-p scattering cross section at energies greater than 100 Mev
is almost independent of the angle of scattering, in strong contrast to the
n-p scattering. Although certain differences between the p-p and n-p
scattering are to be expected even if the nuclear force law is the same, the
observed differences are such that they cannot be explained in terms of the
theory and forces that account for the low-energy results. The charge-
independent forces which account for the low-energy results seem to fail
when applied to the high-energy scattering data. This dilemma has not
yet been resolved; it is receiving a great deal of attention and it may still
be possible to find a force law which can account for both types of scattering.
Thus, in spite of the existence of a large amount of experimental material,
it has not yet been possible to analyze the forces which hold neutrons and
protons together. Although considerable progress has been made in recent
years, knowledge of the nature of nuclear forces is still in an early, if not
rudimentary, stage. It is, therefore, not possible to develop a theory of
nuclear structure based on the present knowledge of nuclear forces.
17-7 Nuclear models. The shell, or independent particle, model. In

the absence of a detailed theory of nuclear structure, atter,1pts have been
made to correlate nuclear data in terms of various models. Severa! models
have been proposed, each based on a set of simplifying assumptions and
useful in a limited way. Each model serves to correlate a portion of our
experimental knowledge about nuclei, usually within a more or less narrow
range of phenomena, but fails when applied to data outside of this range.
The nuclear shell model is one of the most important and useful models
of nuclear structure. Its conception and name come from the results of
empírica! correlations of certain nuclear data. It has become apparent in
recent years that many nuclear properties vary periodically in a sense
similar to that of the periodic system of the elements. Most of these
properties show marked discontinuities near certain even values of the
17-7) NUCLEAR MODELS. THE SHELL MODEL 527
proton or neutron number. The experimental facts indicate that especially

stable nuclei result when either the number of protons Z or the number of
neutrons N = A - Z is equal to one of the numbers' 26 • 27 • 28 ¡ 2, 8, 20,
50, 82, 126. These numbers are commonly referred to as magic numbers.
Although this name is not a strictly scientific one, it serves the purpose of
identifying the numbers simply and conveniently. The magic numbers of
neutrons and protons have been interpreted as forming closed shells of
neutrons or protons in analogy with the filling of electron shells in atoros,
and the neutron and proton shells appear to be independent of each other.
It has been shown that the existence of the closed neutron and proton shells
corresponding to the magic numbers can be described theoretically in terms
of a particular nuclear model. The remainder of this section will be de-
voted to a summary of sorne of the evidence which led to the nuclear shell
model, and to a brief discussion of the ideas in volved in the theory.
One of the most important pieces of evidence for the magic numbers
comes from the study of the stable nuclides. The nuclei for which Z and
N are 2 or 8, He 4 and 0 16 respectively, are more stable than their neigh-
bors, as can be seen from the binding energies (Table 9-3). Evidence for
sorne of the other magic numbers has been obtained from very careful
studies< 29 , 3 o> of the details of the binding energy curve in the region of
intermediate and heavy nudei. When the average binding energy per
nucleon (obtained from mass spectrographic measurements and nuclear
disintegration data) is plotted against the mass number, the curve is not
really smooth as in Fig. 9-11, but has severa! kinks in it. These kinks or
breaks correspond to sudden increases in the value of the binding energy
per nucleon. They have been found to occur at Pb 2º8 (Z = 82, N = 126),
Ce 14º(Z = 58, N = 82), Sn 12 º(Z = 50, N = 70), and Sr 88 (Z = 38,
N = 50). In other studies, <31 , 32 > masses have been determined in the
neighborhoods of Z = 20 and 28, and N = 20 and 28, and maxima have
been found at these values for the binding energy per nucleon. As in other
properties, the evidence for the special stability of nuclei with N or Z equal
to 28 is less marked than that for the magic numbers. Along with the
numbers 14 and 40, 28 is sometimes called a semimagic number.
Stability is usually related not only to high binding energy, but also to
high natural abundance; in all cases where the binding energies are known,
a definite correlation is found between the binding energies and the abun-
dances. A table of the relative abundances of nuclei compiled from data
on the composition of the earth, sun, stars, and meteorites has been pub-
lished by Brown;< 33 > pronounced peaks are found at O 16 (N, Z = 8),
Ca 40 (N Z = 20) Sn 118 (Z = 50) · Sr 88 Y 89 Zr 89 (N = 50) · Ba 138
' 14. º(N =' 82); and at Pb 2 'º 8 (Z =' 82, N' = 126).
La 1 39 , Ce
' '
The relative stability of different elements is also indicated by the
number of stable isotopes per element. A graph of this number is shown
~
@ ® @
¡ ¡ ! z
10 e::
¡;a
9 l".I
>
:tJ
8 "'l
o
:tJ
[7 ('")
l".I
rfJ
.i
Q)
6 >
zt::,
:o
~
., 5 z
e::
'o ¡;a
1l 4 ~
z~
:tJ
3 1
2 ~
e::
~
e::
o l l lll l lll l ll f l ll l lll r l l
:tJ
l".I
10 20 30 40 50 60 70 80 90
Atomic number n:i:
;,.
FIG. 17-3. The number of stable isotopes per element as a function of the atomic number (Flowers< 27 >). :o
....
--.:,
..
1
~
10
9 @ @ @ @
¡ ¡ ¡ ¡
8
z
Q
[ 7
3
~
-~ 6
., ~
:o
is
'o
1
$4 ~
l!'.I
§ a,
za =
~
¡i:
2 ot:,
~
10 20 30 40 50 60 70 80 90 100
Number of neutrons
Fm. 17-4. The number of stable nuclides as a function of the neutron number (FlowersC 27>). 01
~
in Fig. 17-3 as a function of z. The lightest element with more than two
stable isotopes is oxygen (Z = 8) ; the fi.rst element with four stable iso-
topes is sulfur, whose heaviest isotope is S 36 with 20 neutrons. Five iso-
topes occur first in calcium (Z = 20); seven appear in molybdenum, of
which the lightest is Mo 92 with 50 neutrons. Ten isotopes occur only
once-in tin with 50 protons. Figure 17-4 represents the number of
stable nuclideR having a given number of neutrons; such nuclei are some-
times called isotoms. The peaks at 20, 28, 50, and 82 neutrons are clear.
The energies of the a-particles emitted by heavy radioactive nuclei
provide strong evidence' 34 > for the ma.gic numbers Z = 82 and N = 126.
The fact that lead (Z = 82) is the stable end product of the three natural
radioactive series points to the special nature of nuclei with Z = 82. The
nuclides 8 ~t 213 and 84 Po 212 , which have 128 neutrons, emit exceptionally
energetic a-particles, 9.2 Mev and 8.78 Mev, respectively; Po 212 has the
extremely short half-life of 3 X 10- 7 sec, while that of At 213 has not yet
been measured. In both of these cases, the daughter nucleus has 126
neutrons; there seems to be a strong tendency for nuclei with 128 neutrons
to emit an a-particle and change into much more stable nuclei with 126
neutrons. When the parent nucleus has 126 neutrons, the a-energies
should be particularly low, since the decay starts from a nucleus with low
energy (high binding energy). These expectations are borne out by the
experimental data; Po 210 and At 211 , with 126 neutrons, have relatively
long half-lives and low a-energies; Bi 2 º9 (N = 126) is stable, whereas the
neutron-richer isotopes Bi 21 º and Bi 211 are a-radioactive.
Similar relations have been found in /3-decay. c35> The energies of the
emitted /3-particles are especially large when the number of neutrons or
protons in the product nucleus corresponds to a magic number. The
results of the investigations on a-decay and /3-decay show that there are
energy discontinuities of about 2 Mev in the neighborhood of the magic
numbers 28, 50, 82, and 126. Since the average binding energy per particle
in a nucleus is about 8 Mev, the effect in energy is about 25%.
It has already been noted in Section 16-5C that the absorption cross
sections for neutrons of about 1 Mev in energy are unusually low for nuclei
containing 50, 82, or 126 neutrons. c35> The cross sections for these nuclei
are lower by a factor of about fi.fty than those of neighboring nuclei; the
low cross section is an indication of large leve! spacing in the compound
nucleus formed by the neutron capture. Similar evidence has been ob-
tained from thermal neutron capture cross sections. <37> The observed
values are unusually small for nuclei with magic neutron numbers, and
nuclei with 50 protons (tin) and 82 protons (lead) have smaller capture
cross sections than their neighbors. Figure 17-5 is a graph of the thermal
neutron cross sections of nuclides with even mass number as a function of
the number of neutrons. The solid lines join the largest values of the cross
17-7] NUCLEAR MODELS. THE SHELL MODEL 531
Z 10,000
barns 1 1 1
Nuclei of even A only ~
1000
20 50 126
~2
+
.¡ 100
5l
~ Z---=-----.50
U 10 A
., ..
W\ . ~. ,) ~ \
r .. N-'
!
20 30
.
40
·V\ l.
50 60 70
V
80
Number of neutrons
90
/·
".
100 110
\
120 130
Fm. 17-5. Thermal neutron cross section as a function of the neutron

number (Flowers <27 >).
sections when two or more nuclei exist, for a given value of N, whose
cross sections have been measured. This procedure emphasizes the larger
values and tends to mask the appearance of the low values associated with
the magic number nuclei. Nevertheless, real dips in the curve appear at
or near the neutron numbers 20, 50, 82, and 126. A similar trend is shown
by a graph for nuclei with odd values of A.
The binding energy of the last neutron added to a nucleus has been
determined from measurements of the Q-values of (d, p) reactions, (n, "Y)
reactions, ("Y, n) reactions, and (d, t) reactions. These reactions and sorne
of the results have been mentioned in Section 16-9. Values of this
binding energy can be calculated from the semiempirical binding energy
formula (17-30) below. The difference between the observed and calcu-
lated values was studied as a function of the number of neutrons, <33> and
the results are shown in Fig. 17-6; AE is the difference between the ob-
served neutron binding energy for a nucleus with N + 1 neutro:ris and the
neutron binding energy calculated from the semiempirical mass formula.
The figure shows sharp discontinuities in the value of AE at 50 and 126
neutrons, and another drop in the region of 82 neutrons. There is a break
near, but not exactly at, 28 neutrons; the evidence for the semimagic
neutron number 28 is not as clear in this study as in sorne of the others
discussed above.
532 NUCLEAR FORCF.S AND NUCLEAR BTRUCTURE [CHAP. 17
-1.0
50 82
30 40 50 60 70 80
N +l
1.0 •
o
•
¡ t
e-1.0
~
-2.0
126
80 90 100 110 120 130 140

N + l
17--6. Binding energy of the last neutron added to a nuclellB as a function

of the neutron number. t:.E is the difference between the theoretical and experi-
mental values (Harvey< 38>).
The correlations discussed are examples of the simpler kinds of evidence

for the special properties of nuclei containing magic numbers of neutrons
or protons. The results have been interpreted as an indication that neu-
trons and protons within the nucleus are arranged into shells within the
nucleus like electrons in atoms. Ea.ch shell is limited to a certain maximum
number of protons or neutrons. When a shell is filled, the resulting con-
figuration is particularly atable and has an unusually low energy.
Severa! theories have been proposed to account for the neutron and
proton shells. <39 .•o.u.u> lt is assumed that each nucleon moves in its
orbit within the nucleus, independently of all other nucleons. The orbít
is determined by a potentíal energy function V(r) whích representa the
17-7) NUCLEAR MODELS, THE SHELL MODEL 533
average effect of all interactions with other nucleons, and is the same for
each particle. Each nucleon is regarded as an independent particle and the
interaction between nucleons is considered to be a small perturbation on
the interaction between a nucleon and the potential field. This model is
often called the independent-particl,e model as well as the nuclear-shell
model. There is a direct analogy between the theoretical treatment of a
nucleon in a nucleus and an electron in an atoro. The potential energy
V(r) is analogous to the Coulomb energy, and the orbit (or quantum state)
of a nucleon is analogous to an orbit (or quantum) state of an electron.
The theoretical treatment of the nucleon parallels that of the electron as
described in Chapter 7, and closed nucleon shells are found analogous to
the closed electron shells in the atoro. The analogy cannot be carried too
far, however, for several reasons. The potential which describes the nuclear
attractions is quite different from the Coulomb potential; it has a forro
between the square-well potential V = - V 0 , and the so-called oscillator
potential V = - V O + ar 2 , where r is the distance between the nucleon and
the center of force anda is a constant; the Coulomb potential is propor-
tional to 1/r. Furthermore, two different kinds of particles must be con-
sidered, protons and neutrons, rather than a single kind, electrons. The
Pauli exclusion principie (Section 7--6) must be applied to both protons
and neutrons; it excludes two protons from occupying the same quantum
state, and two neutrons from having the same quantum numbers.
The theoretical results are shown in Fig. 17-7. The energy levels for a
three-dimensional rectangular well potential are shown on the right (the
well has actually been assumed to be infinitely deep). The levels for the
simple harmonic oscillator potential are shown on the left. The numbers
in parentheses are the total number of electrons in the indicated shell and
the shells below. For the oscillator potential there would be closed shells
at 2, 8, 20, 40, 70, 112 and 168 nucleons. Although closed shells are pre-
dicted at 2, 18, and 20 nucleons, the numbers 50, 82 and 126 are missing
and closed shell numbers appear which have no apparent connection with
experimental relations. The infinitely deep rectangular well potential also
predicts closed shells at 2, 8 and 20 nucleons but, in addition, at many
other nonmagic numbers, with 50, 82 and 126 again missing. lt has been
shown that to get the correct shell numbers, i.e., those deduced from the
experimental data, another assumption must be made. For reasons which
are not yet fully understood, it is necessary to assume that there is a
marked difference in the energy of a nucleon according to whether its
spin angular momentum is parallel or antiparallel to its orbital angular
momentum. When this assumption of slrong spin-orbit coupling is made
and its effects are taken into account, <41 , 43 > the theory gives closed nucleon
shells at 2, 8, 20, 50, 82, and 126 particles, in agreement with the
experimental data. The occurrence of magic numbers of protons and
·-;
-;;..;,_
~
:8
38-2-(92
h-22-(90
.
o
d-10-<&a
~
g
g-18-(58
t
Í;i;l
3p-6-(40)
lf-H-(34
28-2~) ~
Sd-10-(18
~
&-<a>-{p
2p-6-(8) "'
FIG. 17-7. Nuclear shells. The energy levels of an infinitely deep rectangular
well are shown on the right, and those of a simple harmonic oscillator potential
on the left. The numbers in parentheses give the total of electrons in the in-
dicated and lower shells. The levels in the center are at energies intermediate
between those of a square-well potential and the simple harmonic oscillator; in
addition a strong spin-orbit interaction is assumed for these levels.<·U>
17-7] NUCLEAR MODELS. THE SHELL MODEL 535
neutrons can then be said to be understood in the sense that a model has
been devised which gives the correct numbers when treated mathematically.
The nuclear-shell model has been applied successfully to a variety of
nuclear problems. A detailed discussion of these applications cannot be
given here, but sorne of the successes of the shell theory will be mentioned.
The model makes it possible to predict the total angular momenta of nuclei
and the results are in good agreement with experiment. It is then possible
to assign values of the total angular momentum to nuclei for which this
quantity has not been measured, in particular to ,S-radioactive nuclei.
This assignment is very useful in the study of ,S-decay. lt has been men-
tioned (Section 14-6) that the probability of ,S-decay depends on the dif-
ference between the total angular momenta of the initial and product
nuclei. The theoretical predictions with regard to the allowed or forbidden
nature of many ,S-decay processes, based on shell mo<lel assignments of the
angular momentum, have been compared with the experimental results,
and reasonably good agreement has been obtained. <44 • 45 •45 >
The subject of nuclear isomerism has been studied from the viewpoint of
the nuclear-shell model, <47 •4 s. 49 > anda correlation has been found between
the distribution of isomers and the magic numbers. When the known
long-lived isomers (T 112 ~ 1 sec) with odd mass number A are plotted
against their odd proton or odd neutron number, as in Fig. 17-8, it is clear
that there are groupings or islands of isomers just below the magic numbers
50, 82, and 126. There is a sharp break at each of these numbers; isom-
erism disappears when a shell is filled and does not appear again until the
next shell is about half full. An example of the discontinuity in the dis-
tribution of isomers is given by the odd-A isotopes of xenon: 5 ~e 127,
s4 Xe 129 , s4Xe 131 , s4 Xe 133 , and 54Xe 135 with 73, 75, 77, 79, and 81 neu-
trons, respectively, ali show isomerism, but no isomerism has been found
in 54Xe 137 and 54Xe 139 with 83 and 85 neutrons, respectively. Isomerism
(the existence of long-lived excited states) occurs when neighboring energy
levels of a nucleus have a large difference in total angular momentum, so
that the transition is highly forbidden. The shell model can be used to
predict the total angular momenta of low-lying excited levels. It predicts
that the conditions for isomerism should exist below the magic numbers
50, 82, and 126, but not immediately above them; in the latter regions,
neighboring levels are expected to have small angular momentum differ-
ences. The model also predicts correctly when isomerism should appear in
the unfilled shells. The classification of isomers according to the type of
transitions (Section 15-10) is also in excellent agreement with level assign-
ments based on the shell model.
The experimental data on magnetic moments and electric quadrupole
moments have also been interpreted with sorne success in terms of the
nuclear shell model. Asan example, at the proton numbers 2, 8, 20, 50,
240_.....,.....,_____________
220
200
180
..
160
140
¡ 120
,-.: 100
T
.I
80
60
Distribution of long-lived
40 nuclear isomers of odd A
20
010 40 60
20 50
Odd nucleons • N or • Z -
Fm. 17-8. Distribution of long-lived isomers of odd A as a function of either

atomic number or neutron number (adapted from Goldhaber and Hill<47 >).
and 82, the quadrupPle moment is zero or small. When a new shell begins
to form, the quadrupole moment is negative; as the number of protons in
the unfilled shell is increased, Q becomes positive and increases until it
reaches a maximum when the shell is about f filled; Q then decreases to
zero at the magic proton numbers, <60 > after which it becomes negative.
This behavior is in accord with the predictions of shell theory. In some
cases, however, the quadrupole moment is much larger than expected from
the independent-particle model. <60 The magnitude of the quadrupole
moment is a measure of the deviation of a nucleus from spherical shape,
and the shell model seems to underestimate this deviation. The large
values of the moment indicate that in sorne cases the nucleus is far from
spherical. These large values can be explained if it is assumed that the
part of the nucleus consisting of the filled shells forros a core which can be
deformed by the nucleons in the unfilled shell. When a spheroidal, rather
than a spherical, core is treated mathematically, the quadrupole moments
17-8) THE LIQUID DROP MODEL 537
come out closer to the experimental values. <52 > This modification leads
to an extension of the shell model which will be discussed very briefly in
Section 17-9.
17-8 The liquid drop model and the semiemphical binding energy
formula. A great deal of importance has been attached to the mass and
binding energy of the nucleus, and a formula which would allow the cal-
culation of nuclear masses or binding energies would be very useful. Such
a formula has been developed and is called the semiempirical mass formula,
or the semiempirical binding energy formula;< 53 , 54 > it is related to the liquid-
drop model of the nucleus. Sorne of the properties of nuclear forces (satu-
ration, short-range), which have been deduced from the approximately
linear dependence of the binding energy and the volume on the number of
particles in the nucleus, are analogous to the properties of the forces
which hold a liquid drop together. Hence, a nucleus may be considered
to be analogous to a drop of incompressible fluid of very high density
(~10 14 gm/cm 3 ). This idea has been used, together with other classical
ideas such as electrostatic repulsion and surface tension, to set up a semi-
empirical formula for the mass or binding energy of a nucleus in its ground
state. The formula has been developed by considering the different factors
which affect the nuclear binding, and weighting these factors with con-
stants derived from theory where possible and from experimental data
where theory cannot help.
The main contribution to the binding energy of the nucleus comes from
a term proportional to the mass number A; since the volume of the nucleus
is also proportional to A, this term may be regarded as a volume energy:
(17-20)
The Coulomb energy between the protons tends to lower the binding
energy and its effect appears as a term with a minus sign. The total Cou-
lomb energy of a nucleus of charge Z is given by Eq. (17-1), and its effect
on the binding energy is represented by the term
, 3 Z(Z - l)e 2
Ec = -ac 5 R
Z(Z - 1)
-4ac A113 , (17-21)
where
2
3 e
ac = 20 Ro X 10-1a (17-22)
In Eq. (17-21), the radius R has becn expressed in terms of A 113 , and the
factor 4 has been introduced in order to conform to the notation often

found in the literature.
The binding energy is also reduced because the nucleus has a surface.
Particles at the surface interact, on the average, only with half as many
other particles as do particles in the interior of the nucleus. In the attrac-
tive term (17-20), it was assumed that every nucleon has the same access
to other nucleons and it is necessary, therefore, to subtract a term pro-
portional to the surface area of the nucleus. This surface energy is repre-
sented by
(17-23)
and is analogous to the surface tension of a liquid.
The binding energy formula needs another term to represent the so-
called symmetry effect. For a given value of A, there is a particular value
of Z which corresponds to the most stable nuclide. For light nuclides,
where the Coulomb effect is small, this value is A/2, as is seen from the
fact that the numbers of protons and neutrons are equal in the most
abundant light nuclides. In the absence of the Coulomb effect, a departure
from the condition Z = A/2 would tend to lead to instability anda smaller
value of the binding energy. A term proportional to sorne power of the
neutron excess (A - 2Z) would represent the magnitúde of this effect.
The second power is chosen because the term then vanishes for Z = A/2,
as does its derivative with respect to Z. The latter condition corresponds
to the maximum value of the binding energy in the absence of the Coulomb
energy. A detailed study of the symmetry effect shows that it is also
inversely proportional to A, with the result that the symmetry energy can
be written
E,. = -a.,. (A - 2Z)2. (17-24)
A
lt was noted in Section 9-4, during the discussion of the stable nuclides,
that those with even numbers of protons and neutrons are the most abun-
dant and presumably the most stable. N uclei with odd numbers of both
protons and neutrons are the least stable, while nuclei for which either the
proton or neutron number is even are intermediate in stability. lt has
been shown that this "odd-even" effect can be represented by a term
E, whose value depends in the following way on the numbers of protons Z
and neutrons Nin the nucleus:
A z N Ea
even even even +a/2A
odd even odd o (17-25)
odd odd even o
even odd odd -6/2A
The formula for the binding energy is obtained by combining all of the
terms just discussed into the equation
2 2
B.E. = avA - 4ac Z(!;; l) - aaA213 - ar (A ~ Z) + E¡. (17-26)
Equation (17-26) can also be rewritten to give the nuclear (or atomic)
mass, since the mass and binding energy are related by the equation
where M p, M n, and M are the masses of the proton, neutron, and nucleus,
respectively. The semiempirical mass formula is then
M = ZM
P
+ (A _ Z)M _ av A
n c2
+ 4ac Z(ZA -
c2 1/3
1) + ~ A 213
c2
+ ar (A - 2Z) 6
2
(17-27)
c2 A - 2c 2 A ·
The equations (17-26) and (17-27) express the binding energy and the
atomic mass of a nuclide as a function of A and Z. The values of the con-
stants can be determined by a combination of theoretical calculations and
adjustments to fit experimental values of the masses (or binding energies).
The values of the constants a.,, ac, ar, a8 , and 6 have been determined, with
the results
a., = 14.0 Mev, ac = 0.146 Mev,
ar = 19.4 Mev, ª• = 13.1 Mev;
the quantity o is 270 Mev for nuclides with even values of A and Z, zero
for odd A nuclides, and -270 Mev for even A, odd Z nuclides. In prac-
tice, masses of atoms are usually measured, and Eq. (17-27) is used to
calculate atomic masses by using the mass of the hydrogen atom instead of
the mass of the proton. When the values M n = 1.008982 amu, and M H
(instead of Mp) = 1.008142 amu, are inserted, Eq. (17-27) can be written
M = 0.99395A - 0.00084Z + 0.0141A 213

- 2Z) + 0.00063Z(Z -
2
(A 1) _ !__
+ O.021 Á A 113 Á'
(17-28)
where
0.145 amu, even A, even Z,
o'= {O, odd A, even Z; odd A, odd Z, (17-29)
-0.145 amu, even A, odd Z.
Equation (17-26) becomes
Z(Z - 1) 21a
B .E. (Mev) = 14.0 A - 0.584 Alla - 13.lA
19.4 (A ~ 2z)2 + E,, (17-30)

where
135
A, even A, even Z,
Ea= O, odd A, even Z; odd A, odd Z, (17-31)

-135
---X-, even A, odd Z.
Several formulas similar to Eq. (17-28) have been developed with some-
what different values for the various constants depending on the amount of
effort put into the determination of those values. One of these formulas
has been used to calculate the mass values corresponding to about 5000
pairs of Z, A values. <55> The semiempirical mass or binding energy formula
reproduces quite accurately the values for many nuclei but does not
account for all of the features of the nuclear binding energy, as will be
shown later. It does give the main features of the dependence of binding
energies on mass number and charge number.
The binding energy formula belps account for sorne of the important
features of the stability properties of nuclei, in particular the P-activity
anél stability properties of isobars. This application of the binding energy
formula will now be illustrated. An examination of the nuclide cbart
shows that isobars (nuclides with the same mass number) are subject to
interesting stability rules. <55> One of these rules is that pairs of stable
isobaric nuclides differ in charge number much more frequently by two
units than by one. More than 60 examples of isobaric pairs are known for
which liZ = 2, but only 2 pairs are known for which liZ = 1; the
latter are 4sCd 113 and 49 lnu 3 , and 51Sb 123 and 52Te 123. The values
A = 113, 123 are the only odd values of A for which there are more
than one stable nuclide. For even values of A, there may be two stable
isobars, and in a few cases (A = 9!_3, 124, 130, 136) even three. In
accordance with the rule stated above, the isobars differ in atomic number
by two units, as in 5oSn 12 4, s2Te 124, and 54 Xe 124 ; the isobars with odd
Z, 51Sb 124 and s 3 I 12 4, are unstable.
Consider now the binding energy formula (17-30) for a given odd value
of A. The only variable terms are those which contain Z, and these depend
on Z 2 ; the odd-even term vanishes. If the binding energies for a series of
.,~
~ .,s::
<l) bl)
s::
<l) .8
"O
bl)
:gs:: ¡a
i:i:l
l
l Ge73
29 30 31 32 33 34 35 26 27 28 29 30 31 32
Atomic number Atomir number
Frn. 17-9. Binding energies and Frn. 17-10. Binding energies and
decay properties of nuclides of odd decay properties of nuclides of even
mass number (A = 73). mass number (A = 64).
nuclides with constant odd A and varying Z are plotted against Z, the
result is a parabola-like curve such as that shown in Fig. 17-9, in which A
has been taken as 73. The binding energies decrease with increasing values
of the ordinate, so that the bottom of the curve represents the most stable
nucleus of the series, with the greatest binding energy. In general, there
will be one isobar for which the binding energy is at or near the bottom
of the curve, and this will be the only stable member of the series. All
isobars whose binding energies are less than that of the stable one will lie
on the two arms of the curve. Their masses will be greater than that of
the stable isobar and they will decay by the emission of an electron or a
positron or by K-capture, subject to the conditions expressed in Eqs.
(12-4), (12-7), and (12-10). Isobars to the left of the stable one have
fewer protons than the stable one and decay by electron emission,
73
aoZn 73 L 31 Ga
73
L 32Ge (stable).
Isobars which lie to the right of the lowest point in the binding energy
curve contain an excess of protons and decay by positron emission, or by
K-capture, or by both, e.g.,
34Se
73 11+ ,K
- a3A s73 ____.
K a2
Ge73 (
s
t able ) •
The results are different for isobars with even A because of the odd-even
effect. The binding energy curve, considered as a function of Z, will still
be parabolic in shape. But nuclides with even numbers of protons and
neutrons have an additional positive energy contribution (135 Mev / A),
while nuclei with odd numbers of protons and neutrons will have binding
energies decreased by this amount. Hence, there are two curves, one for
even Z-even A nuclides and another for odd Z-even A nuclides, as in Fig.
17-10. The odd-even nuclides lie on the upper curve and are therefore
unstable with respect to those on the lower curve, with the result that no
stable odd Z-even A nuclides should exist. The only exceptions to this rule
are the light nuclides H 2 , Li6, B 10 , and N 14 . As in other respects, light
nuclei must be treated separately; a model like the liquid-drop model
would not be expected to apply to these nuclei, which have too few particles
to be treated as a fluid drop with a continuous distribution of material. It
can be seen from the curve for even Z-even A nuclides, that two isobars
with Z-values differing by two units can lie close to the bottom of the lower
curve; these constitute stable isobaric pairs. N uclides to the left of the
stable nuclides decay by electron emission, and those to the right by
positron emission or K-capture or both. The isobar at the bottom of the
odd-even curve may decay by either positron or electron emission.
The binding energy or mass formula <loes not include closed shell effects;
but it can be used to provide a base line from which shell effects can be
calculated. Studies of the energetics of ,B-decayc 35 J and mass-spectroscopic
measurementsc 31 · 32 J show that the mass of any isobar which has Z or
N = 20, 28, 50, 82 or 126 lies about 1 or 2 Mev below the value predicted
by the smoothly varying mass formula. There are, therefore, disconti-
nuities in the atomic masses at the magic numbers and a shell-structure
term can be added to the semiempirical mass or binding energy formula
to take the effect into account. <57 l
The liquid-drop model has other important applications; the theory of
the compound nucleus is based on this model, and the phenomenon of
nuclear fission, which will be discussed in Chapter 19, can be described in
terms of the distortion and division of a liquid drop.
17-9 The collective nuclear model. The successes of the liquid-drop

and nuclear-shell models seem to lead to a serious dilemma. The liquid-
drop model can account for the behavior of the nucleus as a whole, as in
nuclear reactions and nuclear fission. In the latter, certain nuclei actually
divide into two smaller nuclei and the division can be described in terms of
the deformation of a drop; an explanation in terms of the motion of a single
nucleon, or any number of nucleons, seems impossible. The liquid-drop
model is said to describe the collective behavior of the nucleus, and the
excitation of the nucleus is treated as surface oscillations, elastic vibrations,
and other such "collective" modes of motion. On the other hand, many
nuclear phenomena seem to show that nucleons behave as individual, and
nearly independent, particles. Hence there are two entirely different ways
of regarding nuclei, with a basic contradiction between them. The reason-
able conclusion is that the two different models or pictures are incomplete
17-9] THE COLLECTIVE NUCLEAR MODEL 543
parts of a larger or more general model. The problem of the electric

quadrupole moments has led to such a model, which includes both the
liquid-drop and independent-particle aspects; this new model is called
the collective model. ' 58 • 59 •60 > lt is assumed, in this model, that the particles
within the nucleus exert a centrifuga! pressure on the surface of the nucleus
as a result of which the nucleus may be deformed into a permanently
nonspherical shape; the surface may undergo oscillations (liquid-drop
aspect). The particles within the nucleus then move in a nonspherical
potential Iike that assumed in order to account for the quadrupole mo-
ments. Thus, the nuclear distortion reacts on the particles and modifies
somewhat the independent-particle aspect. The nucleus is regarded as a
shell structure capable of performing oscillations in shape and size. The
result is that the collective model can be made to describe such drop-like
properties as nuclear fission' 61 • 62 > while at the same time preserving the
shell model characteristics and, in fact, improving on the earlier shell
mode!.
The simplest type of collective motion which has been identified experi-
mentally is connected with rotations of deformed nuclei. ' 63 •64 > If the
rotational collective motion is sufficiently slow, it will not affect the internal
structure of the nucleus. According to classical physics, the rotational
energy is proportional to the square of the angular velocity w, so that
E,ot = ½Iw2 , (17-32)
The parameter I denotes an effective moment of inertia of the nucleus

which can be calculated if a specific model is assumed for the interna!
structure. Rotational energy levels are obtained when the angular mo-
mentum is quantized and the result is, for even-even nuclei,
E,ot = "'22¡ J(J + 1), (17-33)
where J is the total angular momentum quantum number of the nucleus.

In the case of a spheroidal nucleus, the deformation is symmetric with
respect to reflection in the nuclear center. As a result, J is restricted to
even values, O, 2, 4, 6, • • • , and the parity should be even. According to
the theory, the deformation (deviation from spherical shape) should be
greatest and the rotational levels most easily observed for nuclei with
numbers of nucleons far from closed shells. The first excited state should
be a 2+ state; the second excited state should be a 4+ state with energy
5•4/3·2 = 3½ times that of the first excited state. The third and fourth
states should have I = 6+ and s+, respectively, and energies 7 and 12
times that of the 2+ state. The experimental data agree quite well with
these predictions, as is evident from Fig. 17-11, which shows the experi-
• Er,/E2
15 1
1
1 .a. EsfE2
1
12 1
• 1
••
10
7 . - .- o o -
5
3! ,- - -
" -
3
---
o i
140 160 180 200 220 240 260 A
208
FIG. 17-11. Experimentally observed energy ratios for rotational levels in

even-even nuclei. <7 4>
mentally observed energy ratios for excited rotational states in even-even

nuclei with mass numbers iv. the regions 150 < A < 190 and A > 220.
These are just the regions farthest from closed shells. The theory also
predicts correctly the properties of the low-lying levels of nearly spherical
nuclei close to closed shells. <54 > Another successful application is to the
problem of the cross sections for Coulomb excitation and for the probability
of E2 transitions. <55 > Finally, the theory has been applied successfully to
problems of magnetic moments, quadrupole moments, and isomeric
transitions. <5 o. 66 >
17-10 The optical model for nuclear reactions. In the study of nuclear
reactions in Chapter 16, it became evident that the theory of nuclear
reactions based on the concept of the compound nucleus (the statistical
model for nuclear reactions) is not always reliable. The statistical nature
of the compound nucleus theory implies that its predictions are at best
averages, and do not take into account the differences between specific
nuclei. It is not surprising, therefore, that a more detailed model is needed
for the description of nuclear reactions.
One of the most serious failures of the compound nucleus theory is its
inability to account for the large-scale energy dependence of total neutron
cross sections. The statistical model assumes that the compound nucleus
is formed immediately when the incident neutron reaches the nuclear
surface. The cross section for reaching the surface turns out to be a mono-
tonically decreasing function of the energy, varying as E- 112 for small
energies and reaching the asymptotic value 21rR 2 for large energies.< 57 >
--
"¡l
.2.
-3
=
t;J
~
~t:"
12 t(
230 240
o
10 210 220 o
t;J
200 t:"
8 "'J
180 o
¡;, 170 ~
160 z
rR2 6 150
140 q
4 130 ~
t;J
110
120
"G ..,e~·9"''' >
~
100 f>. \o'<:J(('' R = 1.45 X 10-13 Al/3 cm ~

t;J
>
~
o
z
Ul
tion of energy and mass number.( 6 71

9, >
Frn. 17-12. Observed tota l neut ron cross sections as a func wavclength of the neutrons.
2
(R/'A) , whcre 'A is the de
2 Broglie c.,,
The encrgy is expressed as x = ,¡..
c.,,
.
í
o
X t
11'1~ M
""NO
o.. ...:"" ó
11 II 11
oc-:: ,_
·1&
lb lo
....
1....
p .s
s
"3
::e:
t<J
s o
p
~
s ~
¿~s~P
10
(\ . ... 230 24(
~
210 220 i:,
u, 8
190 200 ~
,rR2 "!l
170 180
6 150 160
~
140 z
~
4 120 130
2
o
100 110 "G ~e\~'(\\~
/). \.Or¡clo('~ R = 1.45 X 10-13 A1/ 3 cm
!~
Vo = 42 Mev ~
r = o.os ~
...o
......
... 6 z
f¾fh r/1
o
(b)
FIG. 17-13. (a) Calculated neutron total cross sections as a function of energy and mass number, for
awelldepth Vo = 42Mev,radiusR = 1.45 X 10- 13 A 1l 3 cm,t = 0.03. (b) Thesamefort = o.os.crn e,,
~
..._:¡
548 :SUCLEAR FORCES AND NUCLEAR STRUCTURE [CHAP. 17
The experimental results do not agree; at excitation energies above the

region of sharply defined resonances, the observed total cross sections show
a more complicated behavior which seems to point to a combination of
single particle and compound nucleus properties. When the cross sections
are plotted against energy E and atomic number A in a three-dimensional
graph, as in Fig. 17-12, certain regularities are observed, with maxima
and mínima at values of E which shift in a systematic way with increas-
ing A. <os. 59 > At the neutron energies involved, between 0.1 Mev and
several Mev, and for intermediate or heavy nuclei, individual resonances
cannot be resolved, and the measured cross sections are averages over
many levels. The problem is that of accounting for this gross structure of
the total neutron cross section.
The large-scale maxima are similar to those predicted by the theory of
a single particle moving in an attractive potential V(r), e.g., a square
well, <70 > but such a potential does not permit absorption of the particle
as required by the compound nucleus concept. The single particle and
compound nucleus pictures have been combined successfully in the optical
model. <71 • 12 > This model describes the effect of the nucleus on the incident
particle by a potential well -V 0(r), but allows for the possibility of com-
pound nucleus formation by adding to the potential a negative imaginary
part, -iV 1 (r). This part produces an absorption oL the incident particle
within the nucleus, and this absorption is supposed to represent the forma-
tion of the compound nucleus. On the optical model, compound nucleus
formation does not occur "immediately" nor with complete certainty.
Even if the incident particle has entered the nucleus, it is removed from
its free particle state only with sorne delay and with a certain probability.
If V 0 (r) and V 1 (r) are reasonably constant over the nucleus, it is possible
to define a coalescence coefficient which is the probability per unit length
for the incident particle in nuclear matter to form the compound nucleus.
In this picture, if an incident particle gets into the nucleus, it is reflected
back and forth before escaping or being absorbed. The theory accounts
very well for the behavior of the neutron cross sections, as is evident in
Fig. 17-13, with the rather simple potential,
Vo = -U, far r < R,
VO = O, for r > R,
and the values
Vo = 42 Mev, t= 0.03 to 0.05,
R = 1.45 X 10- 13 A. 113 CTYJ..

REFERENCES 549
In the optical model, the nucleus is not "black" to the wave representing
the incident particle; instead it acts like a gray sphere, partly absorbing
and partly refracting the incoming wave. There is an analogy with physical
optics in that the nucleus acts like a spherical region with a given refractive
index (attractive potential well) and opacity. lt is this analogy which is
responsible for the adjective optical, and for the name cloudy crystal ball
model, which is sometimes used. The model has also been. extended and
applied with more complicated potential functions to other nuclear reac-
tions and cross sections. <72 >
REFERENCES
GENERAL
2, 10, 11, Appendix C2.
H. A. BETHE and P. MoRRISON, Elementary Nuclear Theory, 2nd ed. New
York: Wiley, 1956.
L. R. B. ELTON, Introductory Nuclear Theory. New York: Interscience Pub-
lishers, Inc., 1959, Chapters 3, 4, 5.
L. EisENBUD and E. P. WIGNER, Nuclear Structure. Princeton, New Jersey:
Princeton University Press, 1958, Chapters 5, 6, 7, 8.
L. LANDAU and YA. SMORODINSKY, Lectures on Nuclear Theory (translated
from the Russian). New York: Plenum Press, 1959.
E. FERMI, Nuclear Physics. Notes Compiled by J. Orear, A. H. Rosenfeld,
and R. A. Schluter. Revised Edition. Chicago: University of Chicago Press,
1950.
J. M. BLATT and V. F. WEISSKOPF, Theoretical Nuclear Physics. New York:
Wiley, 1952.
R. G. SACHS, Nuclear Theory. Reading, Mass.: Addison-Wesley, 1953.
N. F. RAMSEY, "Nuclear Two-Body Problems and Elements of Nuclear
Forces," Experimental Nuclear Physics, E. Segre, ed., Vol. I, Part IV. New
York: Wiley, 1952.
K. B. MATHER and P. SwAN, Nuclear Scattering. Cambridge: Cambridge Uni-
versity Press, 1958.
M. G. MAYER and J. H. D. JENSEN, Elementary Theory of Nuclear Shell
Structure. New York: Wiley, 1955.
E. FEENBERG, Shell Theory of the Nucleus. Princeton, New Jersey: Princeton
University Press, 1955.
ARTICLES IN THE Encyclopedia of Physics (Handbuch derPhysik) Vol. 39. Berlin:
Springer Verlag, 1957.
l. L. HuLTHÉN and M. SuoAWARA, "The Two-Nucleon Problem," pp. 1-143.
2. D. L. H1LL, "Matter and Charge Distribution within Atomic Nuclei,"
pp. 178-240.
3. J. P. ELLIOTT and A. M. LANE, "The Nuclear Shell Model," pp. 241-410.
4. S. A. MoszKOWSKI 1 "Models of Nuclear Structure," pp. 411-559.
PARTICULAR
l. H. A. BETHE, "Nuclear Physics A. Stationary States of Nuclei," Revs .
.Mod. Phys. 8, 82 (1936); esp. p. 96.
2. E. WmNER, "On the Consequences of the Symmetry of the Nuclear Hamil-
tonian on the Spectroscopy of Nuclei," Phys. Rev. SI, 107 (1937).
3. E. FEENBERG and E. WIGNER, "On the Structure of the Nuclei Between
Helium and Oxygen," Phys. Rev. SI, 95 (1937).
4. E. WIGNER and E. FEENBERG, "Symmetry Properties of Nuclear Levels,"
Rep. Prog. Phys. London: Physical Society. Vol. 8 (1941), pp. 274-317.
5. D. R. INGLIS, "The Energy Levels and the Structure of Light Nuclei,"
Revs . .Mod. Phys. 25, 390 (1953).
6. T. LAURITSEN, "Energy Levels of Light Nuclei," Ann. Revs. Nuc. Sci. 1,
67 (1952).
7. F. AJZENBERG-SELOVE and T. LAURITSEN, "Energy Levels of Light Nuclei,"
(a) Nuclear Physics 11, 1-340 (1959); (b) Ann. Revs. Nuc. Sci. 10, 409 (1961).
8. W. E. BuRCHAM, "lsotopic Spin and Nuclear Reactions," Progress in
Nuclear Physics, O. R. Frisch, ed. London: Pergamon Press, 4, 171 (1955).
9. K. W. FoRo and D. L. H1LL, "The Distribution of Charge in the Nucleus,"
Ann. Revs. Nuc. Sci. S, 25 (1955).
10. J. M.C. ScoTT, "The Radius of a Nucleus," Progress in Nuclear Physics,
O. R. Frisch, ed. London: Pergamon Press, 5, 157 (1956).
11. O. KoFOED-HANSEN, "Mirror Nuclei Determinations of Nuclear Size,"
Revs . .Mod. Phys. 30, 449 (1958).
12. A. E. S. GREEN, "Nuclear Sizes and the Weiszacker Mass Formula,"
Revs. Mod. Phys. 30, 569 (1958).
13. R. HoFSTADTER: (a) "Electron Scattering and Nuclear Structure," Revs.
Mod. Phys. 28, 214 (1956); (b) "Nuclear and Nucleon Scattering of High-Energy
Electrons," Ann. Revs. Nuc. Sci. 7,231 (1957).
14. E. M. HENLEY, "Nuclear Radii from Mesonic Atoms," Revs. Mod. Phys.
30, 438 (1958).
15. "Papers from the International Congress on Nuclear Sizes and Density
Distributions held at Stanford University, Dec. 17-19, 1957," Revs. Mod. Phys.
30, 412-584 (1958).
16. J. D. JACKSON and J. M. BLATT, "The Interpretation of Low-Energy
Proton-Proton Scattering," Revs. Mod. Phys. 22, 77 (1950).
17. G. L. SQUIRES, "The Neutron-Proton Interaction," Progress in Nuclear
Physics, O. R. Frisch, ed. London: Pergamon Press, 2, 89 (1952).
18. G. BREIT and M. H. HuLL, JR., "Advances in Knowledge of Nuclear
Forces," Am. J. Phys. 21, 184-220 (1953).
19. BAILEY, BENNET, BERGSTRAHL, NUCKOLLS, RICHARDS, and ,VILLIAMS,
"The Neutron-Proton and Neutron-Carbon Scattering Cross Sections for Fast
20. D. H. FRISCH, "The Total Cross Sections of Carbon and Hydrogen for
Neutrons of Energies from 35 to 490 kev," Phys. Rev. 70, 589 (1946).
21. LAMPI, FREIER, and WILLIAMS, "The Total Scattering Cross Section of
Neutrons by Hydrogen and Carbon," Phys. Rev. 80, 853 (1950).
REFERENCES 551
22. BARNES, CARVER, STAFFORD, and WILKINSON, "The Photodisintegration

of the Deuterorr at Intermediate Energies, I," Phys. Rev. 86, 359 (1952).
23. D. H. WILKINSON, "The Pl\otodisintegration of the Deuteron at Inter-
mediate Energies, II," Phys. Rev. 86, 373 (1952).
24. HARRis, MuEHLHAUSE, RosE, ScHROEDER, THOMAS, and WEXLER,
"Thermal Neutron-Proton Capture," Phys. Rev. 91, 125 (1953).
25. J. M. BLATT and V. F. WEISSKOPF, Theoretical Nuclear Physics. New
York: Wiley, 1952, Chapter III.
26. M. G. MAYER, "On Closed Shells in Nuclei," Phys. Rev. 74, 235 (1948).
27. B. H. FLOWERS, "The Nuclear Shell Model," Progress in Nuclear Physics,
O. R. Frisch, ed. London: Pergamon Press, 2, 235 (1952).
28. M. H. L. PRYCE, "Nuclear Shell Structure," Reports on Progress in
Physics. London: Physical Society, 17, 1 (1954).
29. H. E. DucKWORTH and R. S. PRESTON, "Sorne New Atomic Mass Measure-
ments and Remarks on the Mass Evidenee for Magic Numbers," Phys. Rev.
82, 468 (1951).
30. H. E. DucKWORTH, "Evidence for Nuclear Shells from Atomic Mass
Measurements," Nature 170, 158 (1952).
31. CoLLINS, NIER, and JoHNSON, "Atomic Masses from Titanium through
Zinc," Phys. Rev. 86, 408 (1952).
32. B. G. Hooo and H. E. DucKWORTH, "An Atomie Mass Study of Nuclear
Shell Structure in the Region 28 :5 N :5 50, 28 :5 Z :5 40," Can. J. Phys. 31,
942 (1953).
33. H. BROWN, "Table of Relative Abundances of Nuclear Species," Revs.
Mod. Phys. 21, 625 (1949).
34. PERLMAN, GHIORSO, and SEABORG, "Systematics of Alpha-Radioaetivity,"
Phys. Rev. 77, 26 (1950).
35. C. D. CoRYELL, "Beta-Decay Energetics," Ann. Revs. Nuc. Sci., Vol 2.
Stanford: Annual Reviews, Inc., 1953.
36. HuGHES, GARTH, and LEVIN, "Fast Neutron Cross Sections and Leve!
Density," Phys. Rev. 91, 1423 (1953).
37. B. H. FLOWERS, op. cit. Ref. 27, p. 241.
38. J. A. HARVEY, "Neutron Binding Energies from (d, p) Reactions and
Nuclear Shell Strueture," Phys. Rev. 81, 353 (1951).
39. E. FEENBERG and K. C. HAMMACK, "Nuclear Shell Strueture," Phys. Rev.
75, 1877 (1949).
40. L. W. NoRDHEIM, "On Spins, Moments and Shells in Nuclei," Phys. Rev.
75, 1894 (1949).
41. HAXEL, JENSEN, and 8Euss, "Modellmii.ssige Deutung der ausgezeich-
neten Nukleonzahlen in Kernbau," Z. Physik. 128, 295 (1950).
42. M. G. MAYER, "On Closed Shells in Nuclei," Phys. Rev. 75, 1969 (1949).
43. M. G. MAYER, "Nuclear Configuration in the Spin-Orbit Coupling Model.
I. Empirical Evidence," Phys. Rev. 78, 16 (1950).
44. MAYER, MoszKOWSKI, and NoRDHEIM, "Nuclear Shell Structure and Beta-
Decay. I. Odd-A Nuclei," Revs. Mod. Phys. 23, 315 (1951).
45. L. W. NoRDHEIM, "Nuclear Shell Structure and Beta-Decay. 11. Even-A
Nuclei," Revs. Mod. Phys. 23,322 (1951).
552 NUCLEAR FORCES AND NUCLEAR STRUCTURE {CHAP. 17
46. M. G. MAYER, "Nuclear Shell Systematics: Classification of ~-Transi-

tions," Beta- and Gamma-Ray Spectroscopy, K. Siegbahn, ed. Amsterdam: North-
Holland Publishing Co.; New York: Interscience. 1955, pp. 433-452.
47. M. GoLDHABER aod R. D. H1LL, "Nuclear Isomerism and Shell Structure,"
Revs. Jlod. Phys. 24, 179-239 (1952).
Phys. Rev. 83, 906 (1951).
Beta- and Gamma-Ray Spectroscopy, op. cit. Ref. 46, pp. 453-467.
50. W. GoRDY, "Relation of Nuclear Quadrupole Moment to Nuclear Struc-
ture Theory," Phys. Rev. 76, 139 (1949).
51. ToWNES, FoLEY, and Low, "Nuclear Qu&drupole Moments and Shell
Structure," Phys. Rev. 76, 1415 (1949).
52. J. RAINWATER, "Nuclear Energy Leve! Argument for a Spheroidal Nuclear
Model," Phys. Rev. 79, 432 (1950).
53. E. FEENBERG, "Semi-empirical Theory of the Nuclear Energy Surface,"
Revs. Mod. Phys. 19, 233 (1947).
54. A. E. S. GREEN and N. A. ENGLER, "Mass Surfaces," Phys. Rev. 91, 40
(1953).
55. N. METROPOLIS, Table of Atomic Jl,laases. Chicago: lnstitute for Nuclear
Studies, 1948.
56. N. FEATHER, Nuclear Stability Rules. Cambridge: University Press, 1952,
Chapter l.
57. A. E. S. GREEN and D. F. EowARDS, "Discontinuities in the Nuclear Masa
Surface," Phys. Rev. 91, 46 (1953).
58. A. BoHR, "The Coupling of Nuclear Surface Oscillations to the Motion of
Individual Particles," Kgl. Danske Videnskab. Selskab, llfat-fys. llfedd. 26, No.14
(1952).
59. A. BoHB and B. R. MoTTELSON, "Collective and Individual-Particle
Aspects of Nuclear Structure," Kgl. Danske Videnskab. Selskab, llfat-fys. llfedd.
27, No. 16 (1953).
60. A. BoHR and B. R. MoTTELSON, "Collective Nuclear Motion and the
Unified Model," Beta- and Gamma-Ray Spectroscopy, K. Siegbahn, ed., op. cit.
Ref. 46, pp. 469-493.
61. D. L. H1LL and J. A. WHEELER, "Nuclear Constitution and the Interpreta-
tion of Fission Phenomena," Phys. Rev. 89, 1102 (1953).
62. J. A. WHEELER, "Nuclear Fission and Nuclear Stability," Niels Bohr and
the Development of Physics, W. Pauli, ed. London: Pergamon Press, 1955,
pp. 163-184.
63. A. BoHR, "Rotational States of Atomic Nuclei." Thesis, Copenhagen:
E. Munksgaards Forlag, 1955, 53 pp.
64. B. R. MoTTELSON, "Collective Motion in the Nucleus." Revs. llfod. Phys.
29, 186 (1957).
65. ALDER, IloHR, Huus, MoTTELSON, and W1NTHER, "Study of Electro-
magnetic Excitation with Accelerated lons," Revs. Mod. Phys. 28, 432 (1956).
66. F. VILLARS, "Thc Collective Model of Nuclei," Ann. Revs. Nuc. Sci. 7,
185 (1957).
REFERENCES 553
67. H. FESHBACH and V. F. WEISSKOPF, "A Schematic Theory of Nuclear

Cross Sections," Phys. Rev. 76, 1550 (1949).
68. H. H. BARSCHALL, "Regularities in the Total Cross Sections for Fast
69. ÜKAZAKI, DARDEN and W ALT0N, "Total Cross Sections of Rare Earths
for Fast Neutrons," Phys. Rev. 93, 461 (1954).
70. F. L. FRIEDMAN and V. F. WEISSKOPF, "The Compound Nucleus," Niels
Bohr and the Development o/ Physics, W. Pauli, ed. London: Pergamon Press,
1955, pp. 134-162.
71. FESHBACH, PoRTER, and WErssKOPF, "Model for Nuclear Reactions with
72. H. FESHBACH, "The Optical Model and Its Justification," Ann. Revs.
Nuc. Sci. 8, 49 (1958).
73. STROMINGER, HoLLANDER, and SEABORG, "Table of Isotopes," Revs. !Jfod.
Phys. 30, 585 (1958).
74. R. WALLACE and J. A. WELCH, Ja., "Beta Spectra of the Mirror Nuclei,"
Phys. Rev. 117, 1297 (1960).
75. L. R. B. ELTON, Nuclear Sizes. Oxford University Press, 1961.
:
PROBLEMS
l. The total binding energies (obtained from the measured atomic masses)
of the nuclei Jisted below are given in Table 9-3. Compute the Coulomb energy
of each of these nuclei and the ratio of Coulomb energy to total binding energy.
Plot all three quantities against the atomic number and the mass number. The
nuclei are Be 9 , Al 27 , Cu 63 , Mo 98 , Xe 130, w1s4, u2as.
2. Calculate the binding energies of the nuclei of Problem 1 from the semi-
empirical binding energy formula (17-29), and compare the results with the
experimental values.
3. Calculate the atomic masses of the nuclides of Problem 1 from the semi-
empirical mass formula (17-28) and compare the resulta with the experimental
value f Table 9-3.
4. ~epeat the calculations of Problems 1, 2, and 3 for the nuclides Ca40,
Sn 12 º, and Pb 208 • What differences are shown by these nuclides from the be-
havior of the nuclides of Problem 1? How would you account for these differ-
ences?
5. From the semiempirical mass formula, Eq. (17-28), derive a condition for
the most stable nucleus ata given mass number. [Hint: The most stable nucleus
with a given mass number should occur for that value of Z, denoted by Zo, for
which the nuclear mass, as given by Eq. (17-28), is a minimum.] What is the
value of Zo for A = 27, 64, 125, and 216? Compare the results with the table of
stable nuclides, Table 9-1.
6. Plot A - Zo against Zo, where Zo is determined from the equation of
Problem 5. Compare the resulting curve with the plot of the known nuclides,
Fig. 12-1.
7. Calculate, from the semiempirical mass formula, the energy available for
the following decay processes, and compare the results with the experimental
values. (a) the a-decay of radium, (b) the a-decay of Po 214 (RaC'), (c) the
/j-decay of Hf 170 (experimental endpoint energy = 2.4 Mev), (d) the /j (plus 'Y)-
decay of Au 198 (experimental decay energy = 1.374 Mev).
8. Calculate the Q-value of the reaction Fe 54 (n, 'Y)Fe55 from the semiempirical
mass formula, and compare the rcsult with the experimental value given in
Problem 13 of Chapter 16.
9. Calculate, from the semiempirical binding energy formula, the binding
energies of the last neutron in Pb 207 , Pb 2 º8 , and Pb 209 • Compare the results
with the experimental values given in Problem 11 of Chapter 16.
10. Calculate the approximate kinetic energy of a nucleon inside a nucleus
under the following assumptions: (1) Because of the short-range interaction with
neighboring nucleons, a nucleon occupies a volume, the radius of which is roughly
half the range of nuclear forces, with the range taken as b ~ 2 to 3 X 10- 13 cm.
(2) This radius is equal to A = >./2r, whcre >. is the de Broglie wavelength of a
nucleon inside the nucleus.
11. When a thermal neutron (a neutron with practically zero kinetic energy)
is captured in a nucleus with Z protons and N neutrons, the excitation energy
in the compound nucleus is given by
S,. = M(Z, N) + M,. - M(Z, N + 1).
PROBLEMS 555
(a) What is the difference in the excitation energy of the compound nucle11s
when a thermal neutron is captured by an odd-N target and by an even-N target
of approximately the same mass number? (b) Calculate the difference in excita-
tion energy for the cases
and
12. Show that the average binding energy per nucleon can be represented by
B.E.
A = ª• -
ª•
A1l3 - ª•
Z
2
A4l3 - a,.
(
l -
2Z) E&
2
A + A .
Calculate the value of the average binding energy per particle in Xe 124 and
compare the result with that listed in Table 9-3.
13. Show that the Q-value of a nuclear reaction may be written
Q = l: (B.E.) 11 -1: (B.E.);,

JI i
where the subscripts i, p represent initial and product nuclei, respectively.

Then show that the Q-value for a (d, a) reaction on the target nucleus with atomic
number Z and mass number A is given, to a good approximation, by
Q = (B.E.)a - (B.E.)d - 2a. + a,[A 213

- (A - 2)2 13 ]
2 2
z (Z - 1) ] [2<A - 2z)2]
+ 4a. [ All3 - (A - 2)113 - aT A(A - 2)
6/ A: A even, Z odd
+ -6/ A: A even, Z even
{
O: A odd, Z anything.
Part III
Special T opics and Applications
CHAPTER 18
NEUTRON PHYSICS
Since the discovery of the neutron in 1932, the scope and importance
of neutron physics have grown remarkably, and there is now a wide
interest in the ideas, methods, and applications which have been developed
in this field. Neutrons, because they are uncharged heavy particles, have
properties which make them especially interesting and important in
contemporary science and technology. The many nuclear reactions
induced by neutrons are a valuable source of information about the nucleus,
and have produced many new nuclear species. These artificially made
nuclides yield further information about nuclei, and have applications in
other branches of science, such as chemistry, biology, and medicine.
Neutrons have direct uses as research tools; for example, their optical prop-
erties make them more useful than x-rays for certain analytical purposes.
The most striking use of neutrons is in the chain reactions involving fissile
materials. These chain reactions have had epoch-making military ap-
plications ("atom bombs"), and may be developed into an important
industrial source of heat and electric power. These uses of neutrons and
applications of neutron physics depend on knowledge of the properties of
neutrons and on an understanding of their interaction with matter. Sorne
aspects of neutron physics and its place in nuclear physics have already
been discussed; the discovery of the neutron was treated in Chapter 11,
as was the production of new nuclear species by means of reactions in-
itiated by neutrons. The P-radioactivity of the free neutron was dis-
cussed in Chapter 14. The properties of neutron-induced reactions were
treated in Chapter 16, and examples were given of the ways in which these
reactions give new information about nuclei. Finally, the role of neutrons
as constituents of the nucleus was discussed in Chapters 8 and 17. In the
present chapter the emphasis will be on the neutrons themselves, their
production and detection, their interaction with matter in bulk, and on
methods for measuring neutron energies and cross sections for neutron-
induced reactions.
18-1 The production of neutrons. Nuclear reactions are the only source
of neutrons, and the (a, n) reactions on light elements which led to the
discovery of the neutron are still used to produce these particles. 0 > When
one gram (one curie) of radium is mixed with severa! grams of powdered
559
560 NEUTRON PHYSICS [CHAP, 18
..._ String to su pport so urce
'Y-ray source
Fm. 18-1. A photoneutron source in which neutrons are produced by the

disintegration of beryllium or deuterium by 'Y-ra.ys from radioactive nuclides
made in a. cha.in-rea.cting pile (Wa.ttenberg< 2 , 3>).
beryllium, about 10 7 fast neutrons are emitted per second as a result of

the reaction
Be 9 + He 4 -+ C 12 + n 1 + 5.71 Mev.
The mixture can be contained in a volume of 6 or 7 cubic centimeters,
and provides a convenient source of neutrons. Radium and its decay
products emit a-particles with energies from 4.79 Mev to 7.68 Mev (RaC'),
and the neutrons have energies from about 1 Mev to 12 or 13 Mev. The
half-life of radium is long, about 1600 years, and the Ra-Be mixture pro-
vides neutrons at a suffi.ciently steady rate to be used as a standard of
neutron emission. Radon is sometimes used to supply the a-particles
instead of radium, since it is a gas and can be made into a more compact
source then Ra-Be, but the Rn-Be source has the disadvantage of decaying
rapidly because of the short half-life (3.8 days) of radon. Polonium-
beryllium sources are used when a neutron source with relatively few
'Y-rays is desired; one curie of polonium mixed with powdered beryllium
18-1] THE PRODUCTION OF NEUTRONS 561
yields about 3 X 10 6 neutrons/sec. The use of polonium has severa! dis-

advantages; the neutron output is much lower than that of the Ra-Be
source, special chemical facilities are needed to prepare pure polonium,
and the half-life (140 days) is short compared with that of a Ra-Be source.
Photoneutron sources< 2 > are used frequently because the neutrons
produced are practically monoenergetic, and it is not necessary to rely on
natural radioactive substances. Most of these sources are based on the
Be 9 ('Y, n)Be 8 and the H 2 ("Y, n)H 1 reactions. The binding energy of the last
neutron is particularly low in H 2 and Be 9 , and the ('Y, n) reactions have low
thresholds (Section 16-9), 2.23 Mev for the H 2 ("Y, n)H 1 reaction, and 1.67
Mev for the beryllium reaction. Very intense "Y-emitters can be made
cheaply in nuclear reactors, and can be made into much more intense
neutron sources than are possible with natural "Y-emitters. A diagram of a
photoneutron source is shown in Fig. 18-1 ; a source of this kind can be used
with Be or D 2 0 and a "Y-emitter, and under appropriate conditions can sup-
ply up to 10 7 neutrons/sec. The properties<a, 4 , 5> of sorne photoneutron
sources are listed in Table 18-1; the values for the neutron yield are taken
from Feld's article. <5> The yield is the number of neutrons per second
from 1 gm of target at 1 cm from 1 curie of the source.
Neutrons can also be produced in particle accelerators, <6 , 7 > and sources
of this kind have been used a great deal in neutron research. The bombard-
ment of heavy ice (D 20), or deuterium-containing paraffin, with deuterons
accelerated in a Van de Graaff electrostatic generator provides a simple
and efficient source. The reaction H 2 (d, n)He 3 is exoergic, with a Q-
value of 3.28 Mev, and good neutron yields can be obtained with deuteron
energies as low as 100 to 200 kev; approximately monoenergetic neutrons
can be obtained if the conditions of the experimentare carefully controlled.
Other reactions which can be used are
H3 + H 2 -+ He 4 + n 1 + 17.6Mev,
C + H -+ N + n
12 2 13 1
0.26 Mev,
-
Li 7 + H 1 -+ Be 7 + n 1 - 1.65 Mev,
H 3 + H 1 - He 3 + n 1 - 0.764 Mev.
The ti.ve reactions mentioned have proved to be the most useful sources
based on bombardment with artificially accelerated charged particles,
and they can provide beams of approximately monoenergetic neutrons in
the range 5 kev to 20 Mev.
The neutrons produced by the methods discussed have intermediate
or high energies. Slow _neutrons, with energies from about 0.01 ev to
a few ev, are needed for many applications; they are obtained from higher
562 NEUTRON PHYSICS [CHAP. 18
TABLE 18-1
PROPERTIES OF PHOTONEUTRON 80URCES
Neutrons per sec.

Source Half-life Average neutron
per curie, 1 gm of target
energy, kev at 1 cm, X 10-4
Na24 + D20 14.8 hr 220 (±20) 27

Na 24 +Be 14.8 hr 830 (±40) 13
Mn 56 + D20 2.6 hr 220 0.31
Mn 56 + Be 2.6 hr 100 (90%); 2.9
300 (10%)
Ga 72 + D20 14 hr 130 6
Ga 72 + Be 14 hr 5
Jn116 + Be 54 min 100 (40%); 0.82
300 (60%)
Sbl24 + Be 60 days 30 19
Lal40 + D20 40 hr 140 0.8
Lal40 + Be 40 hr 620 o.a
energy neutrons by allowing the latter to move through a material in
which they can lose most of their energy in scattering collisions. Neutrons
lose energy through inelastic scattering collisions with medium or heavy
nuclei and through elastic scattering collisions with light nuclei; the first
process is more effective for neutrons with energies greater than about
1 Mev and the second for neutrons with lower energies. If a source of
fast neutrons is embedded in a large mass of material containing light
nuclei, the neutrons move through the material and lose energy as they
undergo scattering collisions. If the absorption ·cross section of the
material for neutrons is small compared with the scattering cross section,
the neutrons continue to lose energy until their speeds are comparable
to those of the thermal motion of the nuclei of the material. They are
then called thermal neutrons. A light material with a suitably low absorp-
tion cross section for neutrons is called a moderatnr and the neutrons are
said to be "slowed. down" or "moderated" to thermal energies. The
most frequently used moderators are water, paraflin, and graphite; heavy
water and beryllium, which are expensive materials, are sometimes used.
A sotirce of fast neutrons together with a moderator, suitably arranged,
provides a source of slow neutrons. Sorne of the details of the slowing-
down process will be given in Section 18-3, and the properties of the
thermal neutrons will be discussed in Section 18-4.
The most powerful sources of neutrons are those associated with nuclear
reactors (chain-reacting piles). It will be shown in Chapters 19 and 20 that
many neutrons are produced in a chain-reacting assembly of a fissile
18-2] THE DETECTION OF NEUTRONS 563
material and a moderator. The neutrons resulting from the fission of

uranium atoms are fast, and are slowed down in the moderator. If the
reactor is based on fission by slow neutrons, it contains a mixture of fast,
intermediate, and slow neutrons. For sorne purposes, as for the produc-
tion of artificial radionuclides, the mixture of neutrons with different
energies can be used, but for many experiments thermal neutrons are
needed. These neutrons can be obtained by the use of a thermal column,
which is simply a large block of graphite moderator placed directly ad-
jacent to the nuclear reactor. Neutrons escaping from the reactor enter
the column and are slowed down in it; after diffusing an appropriate
distance in the column, they are reduced to thermal energies. The thermal
neutrons can be used inside the column to bombard targets exposed there,
or they can be allowed to escape through a hole, forming a beam of particles
which can be used outside the column.
18-2 The detection of neutrons. N eutrons are uncharged particles and

produce a negligible amount of ionization in passing through matter,
with the result that they cannot be detected directly in any instrument
(Geiger counter, cloud chamber) whose action depends on the ionization
caused by the particle which enters it. The detection of neutrons depends
on secondary effects which result from their interactions with nuclei. es>
Sorne of the reactions are:
l. the absorption of a neutron by a nucleus with the prompt emission
of a fast charged particle;
2. the absorption of a neutron with fission of the resulting compound
nucleus;
3. the absorption of a neutron with the formation of a radioactive
nuclide whose activity can be measured;
4. the scattering of a neutron by a light nucleus, such as a proton, as
a result of which the recoiling light nucleus produces ionization.
A neutron detector based on the first type of interaction can be an
ionization chamber ora proportional counter. One of the most frequently
used detectors is based on the reaction B 1 º(n, a) Li 7 , which is exoergic, with
Q = 2.78 Mev. The target nucleus B 10 , which has an abundance of
18.8% in natural boron, is responsible for the high cross section (755
barns) of the latter for thermal neutrons. A chamber or counter can be
filled with BF 3 gas or lined with a compound of boron; the gas detector
is used more extensively. When high sensitivity is needed, natural BF 3
is replaced by B 1ºF 3 made from the separated isotope B 10 . The cross
section for the (n, a) reaction follows the 1/v law (Section 16-5A) and
falls off with increasing neutron energy. The sensitivity of BF 3 counters
consequently decreases with increasing neutron energy, and these counters
are most useful for slow neutrons.
The fission process can be used to detect neutrons, and this method
may be regarded as a special case of the first method discussed. The
following reactions are typical:
U 235 + n (slow or fast) --+ A + B,
U 238 + n (fast) --+ C + D;
A, B, C, and D represent strongly ionizing fission fragments, which are
heavily charged and highly excited nuclei. Fission fragments have kinetic
energies of the order of 100 Mev and their intense ionization is easy to
distinguish from that caused by protons, a-particles, or other ionizing
radiations. The fissile nuclei may be introduced into an ionization chamber
as a gas (UF 6) or as a wall coating. Chambers containing materials
fissionable by thermal neutrons (U 235 , U 233 , Pu 239 ) are efficient thermal
neutron detectors. Chambers containing natural uranium, or uranium
from which sorne of the U 235 has been removed, can be used as fast neutron
detectors at energies greater than about 1.0 Mev. Thorium, protoactinium,
and neptunium can also be used in this way to detect fast neutrons.
The third method is based on the fact that many nuclear reactions
induced by neutrons result in radioactive product nuclei, and the neutrons
are detected by means of the activity of an exposed substance. The
feasibility of detecting the induced radioactivity depends on its lifetime,
which cannot be appreciably shorter than the time which must elapse, for
experimental reasons, between exposure to neutrons and measurement of
the induced reactivity; on the other hand, the lifetime must not be so
long that the decay rate is negligible. The cross section for the reaction
involved must be large enough so that enough radioactive nuclei are
formed during the exposure. Many substances have been found in which
these requirements are met, among them being indium, gold, manganese,
and dysprosium, and these can be used as detectors. Foils of the detector
material are exposed to the neutron source, usually for a given length of
time; they are then removed and the extent of the induced activity is
determined by counting the emitted radiations with an appropriate
Geiger counter, ionization chamber, scintillation counter, or other radia-
tion detector. This activation method can be used for neutrons in the
different energy ranges, provided the detector materials are chosen care-
fully.
The most common method for detecting fast neutrons is based on the
observation of the ionization produced by the recoiling protons in the
elastic scattering of neutrons by hydrogenous materials. An ionization
chamber or counter filled with a hydrogen-containing gas, or having a
window made of a solid hydrogenous material, may be used. An incident
neutron can impart enough energy to a hydrogen nucleus so that the
18-3] INTERACTION OF NEUTRONS WITH MATTER IN BULK 565
ionization caused by the moving proton can operate the detector. The
hydrogenous material can also be contained in a cloud chamber or in a
nuclear emulsion.
18-3 The interaction of neutrons with matter in bul]{ : slowing down.

The interaction between neutrons and matter in bulk is very different
from that of charged particles of 'Y-rays and is a subject that requires
special treatment. lt was found by Fermi {1934) that the radioactivity
induced in targets bombarded with neutrons is increased when the neutrons
are made to pass through a hydrogenous material placed in front of the
target. Fermi and his colleagues showed that the neutrons are slowed
down in the hydrogenous material, apparently without being absorbed,
and the slower neutrons have a greater probability of inducing radioac-
tivity than do more energetic neutrons. The conversion of fast neutrons
into slow neutrons has been investigated in great detail both experimentally
and theoretically, and the importance of slow neutrons and the slowing
down process has been demonstrated beyond question. A complete and
rigorous treatment of neutron slowing down involves complicated mathe-
matics, but the underlying principles and sorne of the most useful results
are quite simple and can be illustrated with little difficulty. For most
practica! purposes, the important slowing-down process is elastic scattering
by light nuclei. Inelastic scattering by intermediate or heavy nuclei is
important for neutrons with energies above 1 Mev, but becomes practically
negligible below this energy. For simplicity, therefore, only elastic scat-
tering by light nuclei will be considered.
The amount of energy that a neutron loses in a single collision can be
calculated by solving the equations of conservation of energy and momen-
tum for the energy of the neutron after the collision. But there is another
method which is simpler and neater, and also illustrates sorne ideas which
are very useful in the treatment of nuclear collisions. This method in-
volves the use of two reference systems: the first is the laboratory or L-
system, in which the target nucleus is assumed to be at rest before the
collision, and is approached by the incident neutron; the second is the
center of mass or C-system, in which the center of mass of the neutron and
nucleus is considered to be at rest and both the neutron and the nucleus
approach it. The collision process can be described from the viewpoint
of an observer moving with the center of mass, and it turns out that the
equations needed for the description are relatively simple. They can be
solved quite easily and the results can then be transformed back to the
L-system. In view of the importance of the slowing-down process and
of the usefulness of the center of mass reference system in nuclear physics,
sorne of the less complex aspects of the process will be treated in detail.
The treatment does not require the use of quantum mechanics, as the
m, vo, Eo
L (laboratory) system
before collision
-+ M
L-system after
collision
--o
M
___X m
m V,
M /v: e
C (center of mass) system C-system

before collision after collision
Fm. 18-2. An elastic scattering collision between a neutron and a nucleus as
described in the laboratory and center of mass reference systems.
classical picture of ''billiard ball" collisions is accurate enough, and the

mathematics needed do not go beyond trigonometry and elementary
calculus.
The relationship between the L- and C-systems is shown in Fig. 18-2.
In the L-system, before the collision, the neutron of mass m moves to-
wards the nucleus (to the right in the figure) with speed v0 , momentum
mv 0 , and energy E 0 ; the nucleus of mass Mis assumed to be at rest. The
speed of the center of mass Ve is given by
m
V.= Vo M +m• {18-1)
After the collision, the neutron moves with speed v and energy E, at an
angle (J with its original direction, and the nucleus moves off at sorne
angle with the original direction of the neutron.
In the C-system, before the collision, the neutron moves to the right
with speed
M
Vo - Ve = Vo M +m 1 (18-2)
and the nucleus moves to the left with speed V •. The total momentum, as
measured in the C-system, is
Mvo ) ( mvo )
m(M +m - M M + m = O,
since momentum is a vector quantity, and the velocity of the nucleus is op-
posite in direction to that of the neutron. After the collision, the neutron
moves at an angle et, with its initial direction. Since the total momentum
must be conserved, its value must be zero after the collision, and the
nucleus must move off at an angle (180º + et,) with the direction of the
incident neutron. The fact that the momentum is zero in the C-system
before and after the collision makes the arithmetic in this system simpler
than in the L-system. The observer in the C-system sees only a change in
the directions of the neutron and the nucleus as a result of the collision,
and the two particles depart in opposite directions. In an elastic collision,
in which the kinetic energy is also conserved, the speeds of the particles
in the C-system must be the same as they were before the collision; other-
wise there would be a change in the total kinetic energy of the two particles.
The total effect in the C-system is, therefore, to change the directions of
the velocities but not their magnitudes. In the L-system, in which the
nucleus was originally at rest, the magnitudes of the velocities are changed,
and the directions are not opposite. The neutron, which is scattered
through an angle 8, has a velocity v which is the vector sum of the velocity
of the neutron in the C-system and the velocity of the center of mass.
The relationship between the different velocities is shown in the vector
diagram of Fig. 18-3. There are two cases of special interest for which
the speed of the neutron after the collision is readily obtained from the
figure. In a glancing collision et, ""' O and
M m
v = v0 M + m + v0 M + m = vo.
The amount of energy lost by the neutron is negligible, and E = E 0 •

In a head-on collision, et, = 180º, and the speed of the neutron is
V = Vo M: m - Vo M~ m = Vo (! ~ :) '
or
Emin
Eo
= ½mv =
½mv~
2
(MM +- mm) 2
• (18-3)
The neutron loses the most energy in a head-on collision; when the moder-
ator is graphite, M = 12, m = 1, and
2
= (12 - 1) = 072
12 + 1
Emin
E0 ..
A neutron can therefore lose up to 28% of its energy in a collision with a

carbon nucleus; a 1-Mev neutron can lose as muchas 0.28 Mev per col-
lision, anda 1-ev neutron up to 0.28 ev.
r ,=r'() (Jf+m
m )
'
''
l'
/ /
,'
,, /
,
/
,
/·¡
''
'
1
,
'
1
B
1 / '
1 , '
e ____ l_¿,,,, __________ J
/)
Fm. 18-3. Vector diagram for the Fm. 18-4. The relationship be-
velocity of the neutron after the col- tween the scattering angle in the center
lision in the laboratory and center of of mass system and that in the labora-
mass systems. tory system.
For intermediate values of t/>, the neutron speed after the collision can
be found as a function of t/> by applying the trigonometric law of cosines
to Fig. 18-3;
2 2
2 2 2 2
v = v0 ( M M
+ m) + Vo ( M m+ m ) + 2vo ( M M+ m) ( m m ) cos t/>.
M+
The ratio of the neutron energy E after collision to the initial energy
E 0 is then
M 2 + m 2 + 2Mmcos t/>
2
E v
(18-4)
Eo - v3 (M + m) 2
If the ratio of moderator mass to neutron mass M/m is called A, the

last equation becornes
E A2 + 1 + 2A cos t/> (18.-5)
Eo (A+ 1) 2
The mass ratio A can be taken equal to the mass number of the moderator
without introducing any significant error, since m is close to unity and
M is very close to an integer. lt is convenient to express the energy
ratio in terms of the quantity
(A_ 1) 2
r= (18--6)
A+ 1 '
which is a measure of the maximum energy that can be lost by the neutron
in a single collision. Equation (18-5) then becomes
E l+r 1-
Eo = -2- + - 2 -r cos t/>. (18-7)
18-3] l:",TERACTION OF NEUTRONS WITH MATTER IN BULK 569
The greatest energy loss occurs for </> = 180º, when cos </> = -1, and
E = rE 0 ; for </> = O, cos </> = 1, and E= Eo,
The scattering angle </> in the center of mass system can now be related
to the scattering angle in the laboratory system. It is evident from Fig.
18-4 that cot 0 = D/B, where
M m
D = v0 M + m cos </> + Vo M + m '
B M .
= Vo M + m sm<f>,
so that
cot 0 = cos + 1/ A .
<1>_
(18-8)
sm</>
The cosine of an angle can be obtained from the cotangent by means of

the relation
cot 0
cos 0 =- -- ---
(1+cot 2 0) 1 12
When the expression (18-8) for cot 0 is inserted into the last equation, the
result, after sorne simple arithmetic, is
1 + A cos </>
cos 0 = (1 + A2 + 2A cos <f>)l/2. (18-9)
The quantity needed is the average value of cos 0, which can be obtained
by integrating Eq. (18-9) over the possible values of </>, the scattering
angle in the C-system. This integration depends on the probability
that a neutron will be scattered through an angle between </> and </> + d</>,
but this probability is not known a priori. Both experimental results
and a rigorous theoretical treatment of the collision process show that the
scattering is spherically symmetric (isotropic) in the C-system, provided
that the initial energy of the neutrons is less than 10 Mev. This condition
is satisfied in most cases of interest, in particular for the neutrons resulting
from nuclear fission. Equation (18-9) can then be integrated over the
element of solid angle 271' sin</> d</>, and
-
cos 0 =
1
71' )
f" cos 0 2 71' sin </> d</>
4 0
1 {" 1 + A cos </> •

= 2 )0 (1 + A2 + 2A cos <1>)112 sm <J,d<f>.
If cos q, is set equal to x,

+1
-- 1 1 + Ax 2
cos(J = 2 / _1 (1 + A2 + 2Ax)11 2 dx = 3A · (18-10)
When A is large, that is, for heavy scattering nuclei, cos 8 is small and
the scattering in the L-system is practically isotropic. Neutrons which
collide with heavy nuclei are scattered forward as often as they are scat-
tered backward. When A is small, as for light nuclei, more neutrons are
scattered forward than backward.
The average energy loss per collision can now be calculated; the calcula-
tion involves a more useful quantity, the average decrease per collision in
the logarithm of the neutron energy, denoted by ~- This quantity is
a convenient one to use in neutron slowing-down calculations because
it is independent of the neutron energy. Since E/E 0 , as given by Eq.
(18-7), is a linear function of cos q,, and all values of cos q, are equally
probable, it follows that all values of E/E 0 are equally probable. The
probability P dE that a neutron of initial energy E 0 will have an energy,
after one collision, between E and E + dE is given by
dE
p dE = E 0 (1 - r) ' (18-11)
where E 0 (1 - r) represents the entire range of energy values which a

neutron can have after one collision. By definition,
~
~ = ln E 0 - ln E = ln E·
Then,
~ = fEo
rEo
Eo
ln E p dE =
f Eo (
rEo In
Eo)
E
dE
Eo(l - r) ·
If x is set equal to E/E 0 ,
~= -1
l
- r 11
f' ln x dx,
or
r
~=1+ 1 _rlnr
2
_ l + [(A - 1)/(A + 1)] 2 ln [(A - 1)/(A + 1)]
(18-12a)
- 1 - [(A - 1)/(A + 1)]2 '
and ~ is independent of energy, as stated above.

Equation (18-12a) can be rewritten in the form
~= 1- (A -
2A
1)2 ln (AA +- 1). 1
(18-12b)
For A > 10, a convenient approximation for ~. good to about one percent,
may be used,
(18-13)
The formulas for t, Eqs. (18-12a) and (18-12b), break down for two
special cases, A = 1 (hydrogen) and A = oo, because the functions on
the right sides of the equations are no longer determínate. By taking
the appropriate limits as A ---+ 1, and as A ---+ oo, values of ~ can be
determined. For the important case A = 1 (hydrogen), ~ = 1, and the
average value of In (E 0 /E) is unity. For A---+ oo, ~---+ O, so that a neutron
loses practically no energy in an elastic collision with a heavy nucleus.
This result can also be obtained from Eq. (18-6) which shows that r ---+ 1,
and Eq. (18-7) which shows that E/E 0 ---+ l. In fact, for large values of
A, r can be expanded in the series
and for values of A greater than 50, r ~ 1 - 4/ A. Remembering that

the maximum energy loss occurs for cos q, = - 1, we get E ~ rE 0 ~
Eo(l - 4/A).
It is sometimes useful to consider the average energy after a collision
Eo
E=
!rEo
EP(E) dE = Eo(l + r) ·
2
(18-14)
For hydrogen (A = 1), r = O, and the average energy of a neutron after

a collision with a proton is just half of the initial energy. In a head-on
collision between a neutron and a proton, which have very nearly equal
masses, the former can lose all its energy; this result follows from Eq.
(18-7), since cos q, = - 1 and r = O. This possibility distinguishes
hydrogen from other moderators. For a collision with a carbon atom,
r = 0.72, and E= 0.86E0 • For A = 200, r = 0.98, E= 0.99E 0 , and
Em,n = 0.98Eo.
When ~ is known, the average number of collisions needed to bring about
a given decrease in neutron energy can easily be calculated. If the neutrons
start out with an average energy of 2 Mev, and if they are to be slowed
down to 0.025 ev (thermal energy at ordinary room temperature), the
572 NEUTRON PHYSICS [CHAP, 18
TABLE 18-2
ScATTERING PROPERTIES oF SoME NucLEI
Element A Number of collisions

E from 2 Mev to 0.025 ev
Hydrogen 1 1.00 18
Deuterium 2 0.725 25
Helium 4 0.425 43
Lithium 7 0.268 67
Beryllium 9 0.208 87
Carbon 12 0.158 114
Oxygen 16 0.120 150
Uranium 238 0.0084 2150
total logarithmic energy loss is ln(2 X 10 6/0.025). Since the average

loss per collision is E, the number of collisions is given by
. ln [{2 X 106 )/0.025] 18.2
Average number of collis1ons = E = -E- • (18-15)
(2 Mev to 0.025 ev)
The values of E and of the average number of elastic collisions needed to
reduce the neutron energy from 2 Mev to 0.025 ev are listed in Table
18-2 for severa! nuclear species.
Although the quantity E is a measure of the moderating ability of a
given substance, it does not tell the whole story. According to Table
18-2, hydrogen would be the best moderator, but the probability of
a collision between a neutron and a hydrogen nucleus in hydrogen gas is
small because of the small density of the gas. The number of atoms
per unit volume must therefore be taken into account. The probability
that a scattering collision will occur, the cross section for scattering, must
also be considered. Finally, for a good moderator, the absorption cross
section must be small, otherwise too many neutrons would be lost by
absorption. As a result of these considerations, two other quantities are
used to express the properties of moderators. The first of these is the
slowing-down power, defined by
Slowing-down power = ENu, = E~, = N ~Ea-,, (18-16)
where N is the number of atoros per cubic centimeter, p is the density,

No is the Avogadro number (6.02 X 10 23 ), and u, is the scattering cross
section. The slowing-down power has the dimension cm- 1. The quantity
Nu • = 2:, is called the macroscopic scattering cross section, and is the
18-3) INTERACTION OF NEUTRONS WITH MATTER IN BULK 573
TABLE 18-3
VALUES oF THE 8Low1Na-DowN PowER AND
MooERATING RATIO
Slowing-down Moderating
Moderator
power, cm- 1 ratio
H20 1.53 72
D20 0.370 12,000
Be 0.176 159
Graphite 0.064 170
probability per centimeter that a neutron will be scattered. Since ~ is the

average loss in log E per collision, the slowing-down power may be inter-
preted as the average loss in log E per centimeter of neutron travel; it
should have a relatively large value for a good moderator.
The second quantity is the moderating ratio, defined by
. ratº10
Moderat mg = ~:2:.
~ = -~- , (18-17)
.,a (1 a
"ª
where is the absorption cross section. The moderating ratio is a measure
of the relative slowing down power and absorbing ability of a substance.
Of the light elements listed in Table 18-2, lithium has by far the highest
absorption cross section (about 65 b at thermal energy) and is ruled .out
as a moderator. Hydrogen and deuterium can be used in the form of
water (ordinary and heavy, respectively), since oxygen is a good moderator,
and the number of hydrogen or deuterium atoms per cubic centimeter of
water is large compared with that in the respective vapors.
The properties of sorne good moderators are given in Table 18-3. In
calculating the values of the moderating ratio, the thermal neutron
absorption cross sections have been used. These values are greater than
those at higher energies, so that the values of the moderating ratio are
lower limits rather than accurate values, but are useful for purposes of
comparison. According to these results, D 2 0 is the best of the moderators
listed and H 2 0 the least effective. The relatively poor moderating ratio
of H 2 0 is caused by the relatively high absorption cross section of hydrogen.
In practice, D 20 is an extremely expensive substance and can be used only
for special applications in which cost is not the primary consideration;
beryllium is also expensive and is rarely used. Graphite and water are
used often, as is paraffin, which is made up of hydrogen and carbon; these
materials offer a satisfactory compromise between moderating ability
and cost.
The above discussion of neutron moderation has touched only on sorne

of the elementary aspects of the process. There are other problems which
are important, such as the determination of the energy distribution of
the neutrons during moderation, and of the distance traveled by the
neutrons during the process. These problems have been treated in detail,
both theoretically and experimentally, by various authors, <9 , 10 , 11 > and
are also discussed in the general references cited at the end of this chapter.
18-4 Thermal neutrons. As a result of the slowing-down process, neu-

trons reach the state in which their energies are in equilibrium with
those of the atoms or molecules of the moderator in which they are moving.
In a particular collision with an atom, a neutron may then gain or lose a
small amount of energy; in a large number of collisions between neutrons
and atoms energy gains are as probable as energy losses. The neutrons
are then said to be in thermal equilibrium with the atoms or molecules
of the moderator. Their behavior is similar to that of gas atoms and
can be described quite accurately by the kinetic theory of gases. It is
necessary, of course, that the neutrons be able to collide with the moderator
atoms a sufficient number of times to reach equilibrium before they are
absorbed, and this requires that the absorption cross section at low energies
must be small compared with the scattering cross section. When the
conditions for thermal equilibrium are satisfied, the neutrons have the
well-known Maxwell distribution of velocities, <12 >
n(v) dv = 41m ( -m-

2rkT
)ª' 2
v2e-mv2 /2kT dv
'
(18-18)
where n is the total number of neutrons per unit volume, n(v) dv is the
number per unit volume with velocities between v and v +
dv, mis the mass
of the neutron, k is Boltzmann's constant, equal to 1.380 X 10- 16 erg
per degree absolute, and T is the absolute temperature in degrees Kelvin.
The Maxwell distribution has certain interesting properties, one of
which is the value of the most probable velocity. lt is evident from
Eq. (18-18) that n(v) = O for v = O and for v = oo, so that n(v) must
have a maximum value for sorne finite value of v. This value of vis the
most probable velocity, denoted by v0 ; ít can be determined by the usual
procedure of differentiating the right side of Eq. (18-18) with respect to
v and setting the derivative equal to zero. The right side can be written
as Cf(v), where C represents the coefficient 4rn(m/2rkT) 312 , which is
independent of v; it is also convenient to set a = (m/2kT) 112 . Then
n(v) = Cf(v) = Cv 2e_,,
2
v2,
and
18-4] THERMAL NEUTRONS 575
1.0
-j
.s
j
o 2000 4000 6000
v (m/s)
FIG. 18-5. The Maxwell distribution of thermal neutrons, showing the

density n and the flux nv as functions of neutron velooity.
For the last equation to be satisfied for some value v = v0 , it is necessary

tbat
{18-19a)
for each value of T. Since n(v) has its maximum value when v = v0 , this
velocity is the most probable one. The energy correaponding to the most
probahl,e vel,ocity is denoted by Eº and is
Eº = ½mvi = kT. (18-19b)
The numerical value of v0 is given in centimeters per second by
The most probable velocity at room temperature, 20ºC or 293ºK, is

2200 m/sec (more precisely, 2198 m/sec), and this velocity corresponds
toan energy of 0.0252 ev. The Maxwell distribution of thermal neutrons
at 20ºC is shown by the left-hand curve in Fig. 18-5, in which the relative
neutron density n(v) is plotted against the velocity in meters per second.
The neutron energy corresponding to the most probable velocity is given
by
Eº(ev) = 8.61 X 10-5 T. {18-21)
Values of v0 and Eº at different temperatures are listed in Table 18-4.

In addition to the most probable velocity v0 and the corresponding
energy Eº there are severa! other properties of the Maxwell distribution
TABLE 1&-4
MosT PROBABLE VELOCITIES oF THERMAL NEUTRONS
Temperature Most probable velocity, Energy corresponding to

ºC ºK vo: meters/sec the most probable velocity,
Eº: ev
o 273 2120 0.0235
20 293 2200 0.0252
27 300 2220 0.0258
127 400 2570 0.0344
227 500 2870 0.0431
327 600 3140 0.0517
427 700 3400 0.0603
527 800 3630 0.0689
727 1000 4060 0.0861
which are of interest. The average velocity is

2
l1 = 7rl/ 2 Vo = l.1284v 0, (18-22)
a'nd is about 13% greater than the most probable velocity. The energy
distribution of the neutrons is given by
n(E) dE = 2m e-B/l:TE112 dE
(1rkT)3!2 , (18-23)
where n(E) dE representa the number of neutrons with kinetic energy

between E and E + dE. The average energy is
(18-24)
and is half again as great as the energy corresponding to the most probable
velocity. The average energy does not correspond to the average velocity,
as can be verified by comparing Eqs. (18-22) and (18-24), but to the root
mean square velocity v., given by
v: = 2Em = ~(ª
m2
kT) = ª(2kT)
2m
=ª
2'
v~
or
v. = l.2248v 0 = 1.0854'1. (18-25)
Although any of the quantities discussed can be used to characterize the

Maxwell distribution of thermal neutrons, it is customary to describe
thermal neutrons in terms of the most probable velocity and the corre-
OIC..:---'----2000~-~-4-000~-~~6000
v (m/s)
FIG. 18-6. Directly measured velocity distribution of thermal neutrons

(circles) compared with the Maxwell distribution at room temperature (Ringo,
cited t>y Hughes).
sponding energy. Thus they are usually referred to as "kT neutrons,"

and the cross sections quoted for thermal neutrons are usually those for
a speed of 2200 m/sec, or for an energy of 0.025 ev.
The velocity distributions of thermal neutrons from different sources
have been measured experimentally. It is possible to determine the rela-
tive numbers of neutrons with different speeds by means of special instru-
ments which will be described in Section l!H>. The maximum of the
measured distribution gives the most probable speed directly, and the
"temperature" of the neutrons can be obtained from Eq. (18-19). For
various experimental reasons, highly accurate measurements are hard to
make, but the measured velocities do follow the Maxwell distribution
within the experimental accuracy of about 10%. An example of a directly
measured velocity distribution° 3 > is given in Fig. 18-6; the neutrons
carne from the thermal column of the nuclear reactor at the Argonne
National Laboratory. The experimental points indicate a "neutron
temperature" slightly higher than 293°K (20ºC), for which the solid,
theoretical curve was calculated.
Another quantity which has great importance in neutron physics is the
flux of neutrons; it is usually defined as the product n(v)v and is obtained
by multiplying the neutron density n(v) by the velocity. The flux dis-
tribution for thermal neutrons is given by
4n _,,2,,,2O
n(v)v dv = 31/2 v3e dv, (18-26)
Vo1r
and is plotted in Fig. 18-5 for room temperature neutrons. The flux has
578 NEUTRON PHYSICS (CHAP. 18
the dimensions cm- 2 sec- 1 and is usually deecribed geometrically as the

number of neutrons crossing one square centimeter of area in one second.
This description is correct for a unidirectional beam of neutrons of density
n and velocity v crossing aplane 1 cm 2 in area normal to the beam. Under
some conditions, however, this description is wrong. Suppose, for example,
that the plane 1 cm 2 in area is inside a nuclear reactor or thennal column
in which there is a density n of neutrons moving with velocity v, and that
the neutrons can cross the area in any direction with the same probability,
that is, the flux nv is isotropic. It can be shown that nnder these condi-
tions the number of neutrons crossing the plane unit area per second from
each side is nv/4. The simple geometrical description of the flux can be
retained ü the area considered is that of a properly chosen volume ele-
ment as, for example, the surface of a sphere whose cross-sectional area
is 1 cm 2. This interpretation of the flux is illustrated in Fig. 18-7, taken
from Hughes' book. The concept of neutron flux is not limited to thennal
neutrons but is used for neutrons with any density and velocity distribution.
The neutron density and flux, the Maxwellian distribution, and the
1/v-law for certain neutron cross sections are linked in an instructive way
in the absorption of thermal neutrons. Suppose that a sample of a certain
stable material is exposed to the flux n(v)v of thennal neutrons, and that
a nuclear reaction occurs in which the neutron disappears, and which has
.a cross section <Ta, In general, <Ta depends on the energy of the neutrons,
or on their velocity, and can be written as <Ta(v). The number of reactions
induced per second by neutrons of velocity v is n(v)vtTa(v)N, where N is
the number of nuclei in the target sample which can undergo the given
reaction. If a new nuclide is produced in the reaction, as in the case of
an (n, 'Y) reaction, the rate of production is
d:a = n(v)vtT0 (v)N, {18-27)
where N (v) is the number of nuclei of the new kind produced by neutrons
0
of velocity v. If now, the cross section u 0 obeys the 1/v-law, it can be

written in the form
<TaoVo
( ) = -V-,
<T,.V {18-28)
where v0 represents sorne standard velocity and u ,. 0 is the value of the

cross section at that velocity. Equation (18-27) becomes
d:i,. = n(v)CTaovoN; (18-29)
the v's cancel and the absorption rate depends on the neutron density n
rather than on the flux nv, regardless of the neutron velocity.
Sphere of
1 cm2 cross-
sectional area
nv/2 neutrons nv neutrons cross

cross plane per sec sphere per sec
Fm. 18-7. The geometrical description of neutron flux as the number of

neutrons crossing a sphere of unit cross-sectional area per second (Hughes).
If the neutron distribution is Maxwellian, n(v) is given by Eq. (18-18),

and the formula for the total number Na of new nuclei formed per second
is obtained by integrating Eq. (18-29) over all possible values of v,
dNa
dt = NuaoVo ¡00
O
m
41m ( 2 1rkT )ª'2 V e-v·
2 2
/2kT
dv. (18-30)
The integral on the right side of the last equation is just equal to n, the
total number of neutrons per cubic centimeter, because the Maxwellian
distribution just shows how the n neutrons are divided among the dif-
ferent possible velocities. This•result can be verified by noting t.hat
['° 2 -IJ2v2 7f'l/2
JO ve dv = 4 fJ 3 ,
where fJ is a quantity independent of v. Then the rate at which new nuclei

are formed by thermal neutrons with a Maxwellian velocity distribution,
in a reaction with a 1/v-cross section, is simply
dNa _llT
dt = nvo<r.,on. (18-31)
The same result would be obtained for any distribution n(v), provided the
cross section <Ta follows the 1/v-law, but the Maxwell distribution is the
important practica! one. It is customary to take "º
as 2200 m/sec, that
is, as the most probable velocity in a Maxwellian distribution at 20ºC,
and the thermal cross sections listed in tables and compilat:ions are
usually those measured at neutron speeds of 2200 m/sec. When the
material being activated has a 1/v-cross section, the thermal neutrons may
be treated as monoenergetic with energy Eº corresponding to the most
probable speed. The thermal flux is expressed simply by multiplying the
neutron density by the velocity v0 = 2200 m/sec.
When the new nuclear species is radioactive, with a decay constant >.,
the rate of change of the nurober of new radioactive nuclei is given by
dNa _71.T ,u
dt = nvo(Ta()-" - l\lYa, (18-32)
This equation has exactly the saroe forro as Eq. (lo-6), which was met
in the discussion of natural radioactivity, and can be integrated to give
Na(t) = nv~aoN {l - e-At), (18-33)
where t is the length of time for the irradiation of the sarople. The rate
at which the radioactive nuclei decay is then
(18-34)
Equation {18-34) is the basis of the roethod usually used to determine
the activation cross section of a nuclide if the neutron density n is known,
or for deterroining the neutron density when the activation cross sec-
tion is known. The target, or detector, is a thin foil of a material which
undergoes a reaction obeying the 1/v-law, so that Eq. (18-34) is satisfied.
The rate of decay >..Na of the induced activity is measured by standard
counting procedures, and if t, >., and uªº are known, n can be obtained.
In an actual experiment, a finite time (J must elapse between the end of
the irradiation period t and the time of the actual counting. The measured
activity is then given by
A,,.= nvoUaoN(l - e-->.1)e--AI. {18-35)
The additional exponential factor corrects for the decay which occurs
between the ·end of the irradiation period and the time of the counting
procedure. This method will be discussed further in Section 18-7.
18-5 The diffusion of thermal neutrons. In work involving thermal

neutrons, it is often necessary to know the spatial distribution of the
neutrons, that is, the dependence of the neutron density on position.
This problem can be treated with the aid of the theory of diffusion, and
sorne of the ideas involved will be developed and then illustrated by roeans
of an easy example. It will be assumed, for the sake of simplicity, that
the neutrons are monoenergetic ¡ in view of the results of the Iast section,
this assumption is a reasonable one.
Suppose that there is a distribution of neutrons whose density varies
only in the x-direction, and consider a plane element of area perpendicular
to this direction. If the neutron density n(x) were uniform (independent
of x) the nurober of neutrons crossing this element from one side would
18-5] THE DIFFUBION OF THERMAL NEUTRONB 581
be equal to the number crossing from the other side. But if there are
more neutrons on one side of the surface element, more neutrons cross
from that side, and there is a net flow of neutrons across the plane area
from the side with the greater neutron density to the side with the smaller
density. Neutrons are then said to diffuse from the region of higher neutron
density to the region of lower neutron density. The motion of the neutrons
may be represented by Fick's law of diffusion, according to which the net
number J of neutrons flowing per unit time through a unit area normal
to the direction of flow is given for the one-dimensional case by
dn
J - -D-· (18-36)
- dx'
J is called the current densüy and D is the diffusi.<m coefficient. When n
has the units cm -a and x is in centimeters, D has the units square centi-
meters per second. The fact that thermal neutrons in a moderator have
the same velocity distribution as the molecules in a gas suggests that the
motion of the neutrons may be treated in the same way as that of gas
molecules. The assumption is made that neutrons, in a medium which
does not absorb many of them, act like the molecules of a very dilute gas,
and the penetration of neutrons through matter is assumed to be similar to
the diffusion process in gases. With these assumptions, neutron diffusion
theory, at least in its elementary form, consists of the application of the
ideas and equations of the kinetic theory of gases to neutrons. In particular,
the diffusion coefficient D can be interpreted in terms of concepts taken
from kinetic theory. The application of these concepts is not limited to
thermal neutrons but can be used for neutrons of any velocity (as, for
example, in the problem of the spatial distribution of neutrons during the
slowing-down process), provided that a suitable diffusion coefficient can
be defined.
One of the concepts which is useful in kinetic theory is that of the mean
free path, usually denoted by >.. Although this symbol is also used for
the disintegration constant, there will be no confusion because the mean
free path >. will always have a subscript denoting a particular process.
The mean free path for a process is related to the cross section for that
process; thus the average distance that a neutron moves between scatter-
ing collisions is called the scaUeri,ng mean free path >.., defined by the
relation
>..(cm) = 1/N<T., (18-37)
where Nis the number of nuclei per cubic centimeter and "• is the scatter-
ing cross section per nucleus. The absorption mean free path is given by
(18-38)
and is the path length traveled, on the average, by thermal neutrons

before being absorbed. There is also a transport mean free path >-tr defined
as
1 1
Atr = --------..=- = - - 1 (18-39}
Ntr.(1 - coa 8) NtTtr
where coa 8 is the average value of the scattering angle in the laboratory
reference system, and is given by Eq. {18-10). The transport mean free
path is introduced in order to take into account the fact that the scatter-
ing is preferentially in the forward direction; it is greater than the scat-
tering mean free path, which means that, on the average, a neutron will
travel farther in a given number of collisions than if there were no pre-
ferred direction. The transport crosstection tTtr = tTa{l - coa(}) is then
a measure of the rate at which the neutron loses ita forward momentum
or the "memory" of ita original direction. According to kinetic theory,
the diffusion coefficient is given by
(18-40)
Hence, if the neutron velocity is known, and if >-tr is determined [by means
of Eq. (18-39)) in terma of the properties N, coa 8, and tTa of the medium
in which the neutrons are diffusing, then the rate of diffusion can be cal-
culated from Eq. {18-36), and the current density is
J = _ >-trV dn. (18-41)

3 dx
It is now possible to show how the spatial dependence of the neutron

density can be found. The example to be given is that of monoenergetic
neutrons diffusing in the x-direction in a large slab of moderator with an
Neutron source
lnfinite
plane
,,,,,..,,,.. source
~ at z = O
I
I
I
I
,,///
,,
--------
-a;, 4 - - % z----.+IIO
Fio. 18-8. Diffusion of thermal neutrons from an infinite plane source.
18-5] THE DIFFUSION OF THERMAL NEUTRONS 583
absorption cross section <Ta• This problem arises frequently in research

with thermal neutrons, especially when a thermal column is used in con-
nection with a nuclear reactor. It is assumed that there is a source which
produces Q neutrons per square centimeter per second, spread uniformly
over the plane boundary of the slab at x = O, as shown in Fig. 18-8. If
the slab is large enough (infinite), variations in the y- and z-directions can
be neglected. Consider the element of the slab between x and x +
dx.
The net number of neutrons which leak into the slab per unit time is
denoted by L,, dx and is given by
L,, dx = J(x) - J(x+ dx)
= _ >-trV (dn) + XtrV (dn)
3 dx ,, 3 dx z+dz
2
= d n dx
>-trV
3 dx 2 •
At equilibrium, the number of neutrons leaking into the element dx of the

slab must be equal to the number of neutrons absorbed in it, which is
nvu aN dx, so that
or
(18-42)
A new quantity, the thermal dijfusion length L, is now defined by means

of the relationship
1/2
L = + ( >-tta) ' (18-43)
so that Eq. (18-42) becomes
(18-4.4)
It is easily verified by direct substitution that the general solution of

Eq. (18--44) is
(18-45)
where a and b are constants to be determined by the conditions on the

problem. One condition is that n must be finite for ali values of x, includ-
ing x = oo; it follows that since Lis positive, b must be zero. The second
TABLE 18-5
DIFFUSION CONSTANTS OF MODERATORS
tTa, milli-
Density, N, tTu, b&r1111, per

Material molecules/ L,cm b&rllll/ molecule
Xt,, Xa,
gm/cm 3 cm cm
cm 8 molecule at 2200
m/-.
H20 1.00 0.0334 X 10" 2.73 69.6 660 0.43 51.2

D,O 1.10 0.0331 171 12.0 0.93 2.52 36,700
Be 1.85 0.1235 20.8 5.65 10.1 1.43 906
e 1.60 0.0803 52.0 4.54 4.5 2.74 2,900
condition is that the current density at x = O is J(O) = Q/2, since only

half of the neutrons produced by the uniform source enter the rnoderator.
This condition determines the value of the constant a
J(O) _ Q _ _ XtrV
2 3
(dn)
dx
_ 0
XtrV
3 L
a,
so that
3 QL QL
a = 2 XtrV = 2D ·
The solution of the problern, therefore, is
n(x) = (QL/2D)e-s/L = noe-s1L, (18-46)
and the neutron density decreases exponentially with the distance frorn
the source. The variation in n(x) is shown in Fig. 18-8. The diffusion
length L can be shown to be the average (air-line) distance a neutron
rnoves frorn the plane x = O before absorption, in contrast to X0 , which is
the total path a neutron traces before absorption. The diffusion length
is important from a practica! standpoint because it can be rneasured experi-
rnentally. The transport and absorption cross sections can also be meas-
ured in sorne cases, and other constants can then be determined with the
aid of the diffusion theory equations. The various diffusion constants
for several rnoderators are listed in Table 18-5.
There are rnany other irnportant applications of the theory of neutron
diffusion which are beyond the scope of this book and are treated in sorne
of the general references listed at the end of the chapter.
18-6 Cross sections for neutron-induced reactions: measurement of

the total cross section. The use of the cross section for a nuclear reaction
as a quantitative measure of the probability that the reaction will occur
18-6] CROSS SECTIONS FOR NEUTRON-INDUCED REACTIONS 585
was treated in Chapter 16. The importance of neutron-induced reactions

in pure and applied physics and technology makes it necessary to have a
detailed knowledge of the cross sections for these reactions. The efficient
production of new nuclides by neutron bombardment requires accurate
information about cross sections for (n, 'Y) reactions. The possibility of
achieving a chain reaction and the design of a nuclear reactor depend
strongly on the values of the absorption and fission cross sections of fissile
materials, and on the absorption and scattering cross sections of moder-
ators and structural materials. The intelligent use of experimental data
on neutron cross sections requires understanding of the methods used for
measuring them, of the results of the measurements, and of the limitations
of the methods.
The fact that a given nucleus generally has a different cross section
for each type of neutron reaction in which it can take part, together with
the fact that each cross section may vary with the neutron energy, sets
obstacles in the way of building up an adequate store of cross-section
values. The total cross section u 1 is the sum of the partial reaction cross
sections and the partial scattering cross sections. Although the total
cross section is the easiest one to measure, it does not, in general, yield
enough information, and measurements of the various partial cross sections
are often needed. As a result of the variation of cross section with energy
it is desirable to measure values of cross sections for monoenergetic neu-
trons. One of the main problems which then arise is that of obtaining
beams of monoenergetic neutrons of sufficient intensity to allow precise
measurements. The experimental techniques involved depend on the
neutron energies; they are different for slow and for fast neutrons, and it
is necessary to consider measurements in different energy ranges separately.
The total cross section can be determined by measuring the neutron
transmission in a "good geometry" experiment. The idea on which the
determination is based is similar in principie to that discussed previously
(Section 15--1) in connection with the measurement of the total attenuation
coefficient or cross section for 'Y-rays. The neutron flux observed after a
beam of neutrons has passed through a thickness x of material is given by
(18--47)
where (nv) 0 is the flux incident on the material, Nis the number of nuclei
per cubic centimeter, and u 1 is the desired total cross section. Then
U1 =_!_In (nv)o' (18--48)

Nx nv
and the value of the cross section can be determined from the response of
a detector in the presence and absence of the material. The geometrical
1
1
Photoelectric cell
'
--
rotatable 1
about shutter 1
, ,- 1 .
, Tr1gger pulse
1
1
1
I ' 4
I ', '
Rotating~
laminated
Cd-AI
shutter ,,,, Double-sided mirror
~ 1 attached to shutter
1
1
Lamp
'I
1
Neutron beam
, from nuclear reactor
1
1
1
'
1
Fm. 18-9. Schematic diagram of a neutron velocity selector.0 6>
requirements can usually be met without difficulty, and the emphasis is on

getting sufficiently intense beams of monoenergetic neutrons, or of analyz-
ing the effect of a nonhomogeneous beam in terms of the effects of its com-
ponents. This problem has been solved for slow neutrons, with the aid of
devices called velocity selectors.
In one type of instrument, <1 4 , 15 • 16> the mechanical velocity selector or
slow neutron chopper, the design is based on the fact that cadmium is a
strong absorber of neutrons with energies less than 0.3 ev, while other
metals such as aluminum are weak absorbers in this energy region. A
schematic diagram of a slow chopper< 15> at the Harwell Laboratory
(England) is shown in Fig. 18-9. A collimated beam from the thermal
column of a reactor is incident on a rotating cylindrical shutter which
contains a series of laminae, alternately cadmium and aluminum, mounted
in a steel case. Because of the strong absorption by the cadmium, neutrons
can pass through and be detected only when the laminae are parallel to
the direction of the beam. Two bursts of neutrons are released at each
revolution; the neutrons travel along a measured path to a detector
(BF 3 counter) and are recorded, according to their time of arrival, in the
following way. A double-sided mirror fixed to the end of a shutter reflects
18-6) CROSS SECTIONS FOR NEUTRON-INDUCED REACTIONS 587
a beam of light from a fixed lamp into a photocell at an instant between

successive bursts given by the angular setting of the cell. The pulse from
the photocell is then fed into an electric gate unit which passes those
pulses from the detector which occur during a short preset time interval
immediately following the photocell impulse. As this gate closes it opens
a second gate and so on through six gates, thus recording neutrons which
arrive at six different times. The circuit therefore constitutes a six-channel
analyzer which can be moved over a greater range of velocities by adjust-
ing the position of the photocell. The relative transmission or neutron
counting rate can then be obtained for different neutron velocities (energies)
and the cross section u 1 can be determined as a function of energy. The
limitations on the mechanical chopper are set by the limited speed with
which the shutter can be rotated and by the use of cadmium, with its
low "cut-off" energy. As a result, a slow chopper, as an instrument of
this type is often called, can be used for neutrons with energies from about
10- 4 ev to about 0.2 ev.
An example of results obtained with slow choppers is given in Fig. 18-10,
in which the total cross section of Au 197 is shown at energies between
2 X 10- 4 ev and about 0.1 ev. The values denoted by black circles were
obtained with the instrument described; the values denoted by open
circles were obtained with a slow chopper at the Brookhaven National
Laboratory. The curve is taken from a recent compilation of neutron
cross sections.° 7 >
Monoenergetic beams of neutrons can be obtained through the use of a
crystal spectrometer velocity selector, the design of which is based on the
wave prpperties of neutrons. It can pe seen from Eq. (16-16) that neutrons
with energies in the range 0.01 ev to about 10 ev have wavelengths in the
neighborhood of 10-s or 10- 9 cm. These wavelengths are of the same
order of magnitude as the wavelengths of x-rays and of the distance
between crystal planes, with the result that optical interference effects
are observed when slow neutrons are reflected from crystal surfaces.
When a beam of slow neutrons of wavelength X is reflected from the surface
of a crystal, reflection maxima are observed at angles given by the Bragg
relation
nX = 2d sin fJ, (18-49)
where n is an integer which represents the order of the reflection (n = 1,

2, 3, ... ), d is the distance between crystal planes, and fJ is the glancing
angle for the nth order reflection. If the de Broglie relation, X = h/mv,
for the wavelength is inserted into Eq. (18-49), the velocity of the neutrons
at the nth maximum is
v- nh , (18-50)
- 2md sin fJ
and the corresponding energy is

1 n 2h 2
E= 2 mv2 = 8md2 sin2 (J (18-51)
A schematic diagram of a crystal spectrometer for neutrons would look

much like the diagram of an x-ray spectrometer (Fig. 4--4), but the neutron
instrument is usually larger because the source area and the detector are
larger than in the case of x-rays. In the Brookhaven crystal spectrom-
eter, os> the detector is a BF 3 cylindrical proportional counter, con-
taining B 1 ºF 3 , oriented so that neutrons travel down the axis of the counter
in order to get a high detecting efficiency. In a measurement of the total
cross section, a thin sheet of material is placed between the crystal and
the detector, and the transmission is measured as a function of the glanc-
ing angle 8 and, therefore, of the neutron energy. The values of u, for
Au denoted by open triangles and squares in Fig. 18-10 were obtained with
the Brookhaven crystal spectrometer. With this instrument, measurements
can be extended into the energy range in which the slow chopper is no
longer useful. Precise measurements can be made at neutron energies up
to 10 ev with a beryllium crystal and useful, but less precise, resulta can
be obtained between 10 and 50 ev. At higher energies, the glancing angles
become very small and the uncertainty in the measurement of these angles,
and consequently in the energies, is large. At low energies (large glanc-
ing angles) higher order reflections introduce difficulties; precise measure-
ments can be made at energies down to about 0.05 ev with a crystal of
rock salt. The energy range which can be covered by the crystal spectrom-
eter makes this instrument especially useful for the study of the details
of low-lying neutron resonances. Accurate values of the Breit-Wigner
parameters (Section 16-3) can be obtained 09> and details of the shape
of the resonance can be determined. The importance of information of
this kind in connection with the understanding of nuclear reactions and
of the structure of the nucleus has been discussed in Chapter 16.
It is necessary, for various reasons, to measure cross sections at neutron
energies greater than those covered by the slow chopper and crystal spec-
trometer, and instrumenta have been devised for this purpose. The
time-of-flight method for measuring the neutron velocity, mentioned in
the discussion of the slow chopper, can be used with a cyclotron or linear
accelerator, which produces a neutron beam. Fast neutrons are produced
in periodic bursts lasting only a few microseconds and separated by longer
intervals. They are slowed down in a paraffin block and then allowed to
pass through a thin slab of material. The time required for the neutrons
to reach a detector about 10 m away from the source is measured elec-
tronically. The timing is usually done by distributing the counter pulses
to a series of recorders (channels) according to time and, in this way,
0.6-----~--....---~--"T'""--..----..----,
5
e
o
10
·1
~
u
15
30 25
0.1 20
25
30
ºo 2 4 6 8 10 12 14 16
Neutron time of fli¡¡:ht (microseconds per meter)
Fm. 18-11. Neutron transmission and total cross section of bromine as

functions of the neutron energy, as measured with a pulsed-cyclotron, time-of-
flight, velocity spectrometer (Havens and Rainwater< 23 >).
neutrons of certain velocities are selected from a nonhomogeneous beam.

An instrument of this kind has been used< 20 • 21 • 22 • 23 > at Columbia
University and a transmission curve obtained with it is shown in Fig.
18-11. The transmission and the total cross section of bromine are shown
as functions of the neutron time of flight and energy; the dips in the
transmission curve correspond to the resonances. lt is possible with
this instrument, to work with neutron energies from 0.003 ev to over
5000 ev.
The time-of-flight techniques have also been applied to fast mechanical
choppers used with neutrons from a nuclear reactor. Bursts of neutrons
are produced by mechanical interruptions of a beam of neutrons with
energies up to severa! thousand electron volts. The f ast chopper first
developed at the Brookhaven National Laboratory< 24 , 25 > consists of a
stationary slit system (the stator) and a moving slit system (the rotor).
The stator produces two sharply defined neutron beams from a roughly
collimated beam from the BNL graphite-moderated research reactor. The
rotor intercepts these beams except at the instant when the slits of the
rotor are aligned with those of the stator. The rotor is 30 in. in diameter,
is made of aluminum and a plastic material, weighs 250 lb and can
attain a speed of 12,000 revolutions per minute. Channels through the
rotor permit eight bursts of neutrons to pass for each revolution. At
10,000 rev /min, the burst length is slightly less than 1 µsec ¡ the flight path
is 20 manda scintillation neutron detector is used. Under these conditions,
high precision can be obtained in transmission measurements and in the
determination of total cross sections. The curve for the total cross section
of silver between 1 ev and 10 4 ev (Fig. 16--4) is an example of data ob-
tained with severa} instruments including a crystal spectrometer (open
circles) and two fast choppers (black circles and triangles). There are
many resonances which lle close together, and it was not possible to separate
(resolve) such close levels before the development of these new, high-resolu-
tion, choppers. .
The different methods used for measuring total cross sections for neu-
trons with energies up to 10 4 ev are compared in Reference 26.
The methods discussed so far can be used for neutrons with energies
from a fraction of an electron volt to several thousand electron volts, and
these energies are said to lie in the "low to intermedia te" energy range.
The energy range from a few kev to about 20 kev is usually referred to as
the "intermediate to fast" range. Although the distinction between slow,
intermediate, and fast neutrons is an arbitrary one, and depends on the
particular purposes for which the neutrons are used or studied, a change
in the methods of measuring cross sections occurs at energies of a few kev,
and is caused by a change in the method of producing monoenergetic
neutrons. At energies of a few kev and above, monoenergetic neutrons
are produced directly by reactions in the target of a charged particle
accelerator or by monoenergetic 'Y-rays producing photoneutrons in
deuterium or beryllium ¡ these methods were discussed in Section 18-1.
When a beam of monoenergetic intermediate or fast neutrons has been
produced in a cyclotron or Van de Graaff generator, its transmission can
be determined by measuring the neutron fiux< 27 > with detectors designed
for the purpose, and the cross section is obtained from the transmission
in the usual way. For further details about the measurement of total
neutron cross sections at higher energies, the reader is referred to a review
article by Barschall. <2 s> The cross section curves of Fig. 16-3 for alu-
minum are examples of results obtained in the energy range 10 4 ev to 1
Mev, with monoenergetic neutron beams produced with a Van de Graaff
generator. Figure 18-12 shows the total cross section of bismuth< 17 > in the
range 10 4 ev to 15 Mev. Figure 18-12 together with Fig. 16--5 shows
how the total cross section of bismuth varies between 1 ev and 15 Mev.
For completeness, Fig. 18-13 is included to show the total cross section
of bismuthn 7 > between 1 ev and 10-4 ev. In the case of bismuth, the total
cross section has been measured as a function of energy between 10-4 ev
and 15 Mev. This element is especially interesting to physicists because it
contains a magic number of neutrons and is monoisotopic, and most
nuclides have not been studied so carefully. In particular, the energy
{§
uBI
(Mev)
z
l!!I
o
~
z
.,,
1T1(b) [¾™~4JjJ~"~=~F~=;f",.¡'i·gf~FHttt1~c.;_ . • E3,'.-i;ffJclJ{fffl1 ♦ ~~~ 5 __ . 1HUI rtll 6
' <bJ ;...
~
J....;f .J ';:t:=¡f:µ:-.:r:-41,.
1 1 11110.1 1 1 1 111.0 1 1 1 l lto 1 1 1 1 I IIOO
E(Mev)
Fm. 18-12. The total cross section of bismuth as a function of neutron energy from
~...
10• ev to 15 Mev.U7>
00
uBi
(slow) ....
cr
~
C"l
i;II
orll
rll
gJ
~
~
D'l
..,<b)
~
z
1
f
z
1
;
i
rll
f (ev)
Fm. 18-13. The total cross section of bismuth as a function of neutron energy trom e;,
10-4 ev to 1 ev.< 17> ~

range between 1 and 10 kev has not yet been explored well for most
nuclei because it has been a region in which precise measurements could
not be made; the energies were too high for time-of-flight methods and
too low for direct neutron beams from accelerators. The development of
the fast chopper allows measurements to be made well into this region from
below, while improvements in the techniques based on the use of Van de
Graaff accelerators< 29> allow transmission measurements to be made with
reasonably good precision at energies as low as 1 kev. Hence the gap in
a.vailable cross-section values in the 1-10 kev region should be closed in
the near future.
18-7 Scattering, absorption, and activation cross sections. The total

cross section can be expressed as the sum of the scattering and absorption
cross sections u, and uª'
<Tt = <T,+ <Ta, (18-52)
The partial cross sections are harder to determine than the total cross
section for several reasons. The total cross section can be obtained from
the ratio of two counting rates, with and without the sample, and it is
not necessary to know the absolute efficiency of the detector. The partial
cross sections, however, cannot be expressed in terms of flux ratios, and
absolute measurements are usually needed. The scattering cross section
can be measured, in principie, by placing the detector outside the direct
beam, say at an angle of about 90º with the direction of the beam, as
shown in Fig. 18-14. If the scattering is isotropic or nearly so, the total
scattering can be obtained from a single mea.surement in which the detector
subtends a known solid angle and the number of neutrons scattered through
that angle is counted. Since the scattered neutrons are spread over a
4,r solid angle, the scattered intensity at any particular angle is very low.
Furthermore, to avoid multiple scattering within the sample, which would
increase the difficulty of interpreting the experimental results, a thin
scatterer must be used and nearly all the neutrons in the beam are trans-
mitted. Hence, only a small fraction of the neutrons is scattered and only
a small fraction of these is detected. Because of these complications,
scattering cross sections cannot in general be determined with the accuracy
possible for total cross sections.
Scattering experiments such as that just outlined have, nevertheless,
been made. <30 > The result obtained for the cross section usually repre-
sents an average value over sorne neutron velocity distribution. This
average value has no simple relationship to the cross section at any par-
ticular neutron velocity beca.use the scattering cross section, unlike the
absorption cross section, often varies in an indefinite and irregular way
at thermal energies. The average scattering cross section for the Maxwell
18-7) SCA'ITERING, ABSORPTION, ACTIVATION CROSS SECTIONS 595
(nv)o
o
:,;
nv
Detector
(transmission)
/ \ Detector
O (scattering)
Fm. 18-14. Geometrical arrangement for measurement of total or scatter-

ing cross section.
distribution has been measured for many substances, and values are listed
°
in the AEC Compilation 7 > of cross sections for neutron reactions, because
of their usefulness in certain practical applications such as design calcula-
tions for nuclear reactors. With a few exceptions these values lie between
1 and 10 b.
To overcome the difficulty of the low counting rate in thin-target scat-
tering measurements, a thick-target method has been developed.< 31 , 32 , 33 >
Although this method has the advantage of greater counting rates for the
scattered neutrons, it has the disadvantage that the measured counting
rate of scattered neutrons cannot be interpreted directly in terms of the
scattering cross section. The ratio <T,/<T1 can, however, be determined as
a function of energy if the corresponding ratio for a standard material
is known; the lower curve of Fig. 16-4 was determined in this way. <33 >
If <T1 is known, <T, is determined. There is also a thick-thin target< 34 • 35>
method in which an attempt is made to combine an increased counting
rate with a mínimum of difficulty in the interpretation of the observed
rates. A thin sample is placed in the neutron beam in aplane at a small
angle to the incident beam; the sample then appears thick for the trans-
mitted neutrons but thin for neutrons scattered at right angles to the direc-
tion of the incident beam. The scattered and transmitted neutrons are
observed as functions of neutron energy for a standard target as well as
for the sample whose scattering cross section is desired. The ratio of
<T,/<T 1 for the sample is related to <T,/<T1 for the standard more simply than
in the thick-target method, and is determined if the latter ratio is known.
The scattering cross section is again determined if <T1 is known from trans-
mission measurements.
The thin-target method has also been improved recently. <36> Resonance
scattering of neutrons from a linear accelerator was measured by comparing
the number of neutrons counted in an annular detector with the number

scattered by a sample of pure lead, which has a known and constant scat-
tering cross section. Eighteen BF3 proportional counters were arranged
to forro the annular detector. The comparison with the known scatterer
me.de it unnecessary to know the absolute values of the incident flux and
the counter efficiency. The detected scattered neutrons were recorded in
a multi-channel time-of-flight analyzer so as to give the yield of scattered
neutrons as a function of the incident neutron energy.
Measurements of the scattering cross sections for slow neutrons can
also be me.de by studying optical interference effects. <3 n Since the wave-
length of slow neutrons is comparable to the interatomic distances in
crystals, the scattered neutrons produce a wave capa.ble of interfering
with the incident neutron wave. <39 > This coherent scattering has a prob-
ability which can be expressed in terms of the ooherent scattering cross
sections of the individual atoms. These cross sections can be determined
from the analysis of the interference effects, although the theory is too
involved to be discussed here. Values of the coherent scattering cross
section ha.ve been determined for a large number of nuclides; a.ge.in, with
a few exceptions, the measured values lie between 1 and 10 b. <17 > In
addition to the coherent scattering there is sorne incoherent, or diffuse,
scattering which does not show interference effects. The absence of inter-
ference between incident and scattered waves can be attributed to such
factora as the presence of two or more isotopes of an element and the
effect of the interaction between the spins of the neutrons and those of
the scattering nuclei. When the neutron energy rea.ches severa} electron
volts, the wavelength of the neutrons is too small to give coherent crystal
effects and the scattering is entirely incoherent. The scattering atom
acts as if it were free and not bound to other atoros; the scattering cross
section measured under these conditions is called the free atom cross section
and is the se.me as the total scattering croes section at theae energies.
It is apparent that the study of neutron scattering is complicated at low
energies, but the existence of effects peculiar to neutrons helps make
them valuable as research tools. <39>
The absorption cross section u,. refers to those reactions in which a
neutron is not l'EHlmitted. The me.in absorption procesaes are the (n, 'Y),
(n, p), and (n, a) reactions, which ha.ve already been discussed in Chapters
11 and 16. The absorption cross section of thermal neutrons can be
measured most accurately with a method that is sensitive to the disap-
pearance of a neutron, but is not affected by scattering. One such method
involves the use of a chain-reacting pile. <•o> If a pile is operating at
a constant neutron flux level, the inaertion of a substance which can abiorb
neutrons wiU result in a gradual decrease in the neutron flux. The decrease
can be measured accurately and is not sensitive to the scattering properties
18-7] SCATTERING1 ABSORPTION 1 ACTIVATION CROSS SECTIONS 597
of the absorber even when the absorption cross section is small. This
method is called the danger coejficient metlwd because it was first used to
measure the harmful effects (with respect to the reactivity of a pile) of
various materials. A modification in which a pile oscillator is used in-
creases the sensitivity of this method{ 41 ·42> the absorber is moved period-
ically in and out of the pile, or from one point to another inside the pile,
with the result that the neutron flux in the pile oscillates; the magnitude
of the flux oscil~tion is a measure of the absorption cross section of the
- sample.
The individual reactions whose cross sections make up u,. may be meas-
ured by detecting the particles produced in the reaction, such as protons
_ or a-particles, or by determining the activity of the product nucleus if
it is radioactive. In the last case, the cross section measured is called
the activation cross section, u act, although the particular reaction produc-
ing the activation may be an (n, 'Y), (n, p), (n, a), or (n, 2n) reaction. The
name uact refers to the method of measuring the cross section rather than
the reaction involved. The ideas on which this method are based were
discussed in Section 18-4. The activation method is not restricted to
thermal neutrons, but can be applied to neutrons of any energy, orto a
distribution of energies; an example of the use of this method with fast
neutrons was given in Section 1&-5C. The activation cross section usually
refers to a specific isotope of an element, since the activity of a particular
isotope of the target element or of another element is usually measured. <43>
In many cases, the absorption and activation cross sections are equal;
for this to happen, it is necessary that one particular absorption process be
much more important than the others and that the product of the reaction
be radioactive. Thus, Cu 63 , a naturally occurring isotope of copper,
captures a neutron to give Cu 6 4 , which has a half-life of 12.9 hr; the meas-
ured activation cross section is 4.3 ± 0.2 b; the absorption cross section
of separated Cu 63 measured by the pile oscillator method is 4.5 ± 0.1 b,
so that the two cross sections agree well.
In sorne cases, the absorption and activation cross sections are not the
same. For example, monoisotopic Bi 269 has an absorption cross section
of 0.034 ± 0.002 b as measured by the pile oscillator method. The activa-
. tion cross section for the production of Bi 210, which is an electron emitter,
is only 0.019 ± 0.002 b, or about half as great as the absorption cross
section. It is n,ow known that Bi269, on capturing a neutron, can also
form an isomer of Bi 216, which is an a~mitter with a half-life of about
10 6 years. Hence, the determination of the {HLctivity of the product of
neutron absorption by Bi 269 aoeounts for only about half of the neutrons
absorbed. The cross section compilation, <17> which has now been re,ferred
to several times, consequently lists absorption and activation cross sec-
tions for thermal neutrons.
The interpretation of the total cross section is often made easy because
one partial cross section is much greater than the other. In moderators
(graphite, D 2O, H2O) the absorption cross section is very small compared
with the total cross section. Graphite, for example, has a total cross
section of about 4.8 b in the energy range 0.02 ev to 400 ev, while its
absorption cross section at 0.025 ev is only 0.0045 b and decreases accord-
ing to the 1/v-law as the energy increases. The scattering cross section
is practically identical with the total cross section, and is constant in this
range. Similarly, bismuth (Figs. 16--5, 18-12, and 18-13) has a total
cross section of about 9 b over most of the energy range. Its thermal
absorption cross section is 0.034 b and decreases according to the 1/v-law
with increasing energy, so that scattering is responsible for practically all
of the total cross section. On the other hand, boron has a total cross
section of 755 b at 0.025 ev, caused almost entirely by the (n, a) reaction,
since the scattering cross section is only 4 b. Boron follows the 1/v-law
for absorption, and scattering does not become significant until the neutron
energy reaches about 10 ev. If the absorption cross section of a nuclide
is known at 0.025 ev (2200 m/sec) and there are no resonances in u, at
low energies, it is usually safe to assume that Ua follows the 1/v-law (Sec-
tion 16--5A) so that the absorption cross section can be calculated for
other energy values. If u, is also known as a function of energy, u, can
be obtained at energies other than 0.025 ev by subtracting the calculated
values of Ua from u,. This procedure saves a good deal of experimental
work on the measurement of u, as a function of energy.
In the region of neutron energies in which the total cross section shows
resonances, it is sometimes necessary to distinguish between scattering
and absorption resonances, and this can be done with the aid of
activation measurements. Thus, indium and dysprosium have absorption
(activation) resonances at 1.45 ev and 1.74 ev, respectively, which make
these substances especially useful as neutron detectora. Manganese
(Mn 66 ), Co 69 , and W 186 have strong scattering resonances at 337 ev, 132
ev, and 19 ev, respectively. Resonance cross sections can reach many
thousands of barns, and one nuclide is known, Xe 136, a fission product,
with an absorption cross section of about 3 X 10 6 b, the largest known
cross section.
At high energies in the range 0.01 to 10 or 20 Mev, additional difficulties
are met.< 28 • 46 > Scattering is no longer isotropic and the scattering cross
section dependa on the angle of scattering. Inelastic scattering becomes
important, and is complicated by the fact that the scattered neutrons
have energies which can be distributed over a wide range, and the sensi-
tivity of the detector must be known as a function of energy. Activation
measurements are harder to make at high energies than at low energies
because the cross sections involved, mainly (n, 'Y), are relatively small and
REFERENCES 599
Iow ,B--activities are produced. For all of these reasons, the cross section
values available in the range 0.01 Mev to about 20 Mev are mainly those
foru 1•
There are, of course, many other interesting and important properties
and applications of neutrons which cannot be discussed in this book;
the reader is referred for them to the books listed in the bibliography
which follows.
REFERENCES
GENERAL
B. T. FELD, "The Neutron," Experimental N'UClear Physics, E. Segre, ed.
New York: Wiley, 1953, Vol. 11, Part VII.
D. J. HuoHEB, Pile Neutron Research. Reading, Mass.: Addison-Wesley, 1953.
E. AMALDI, "The Production and Slowing Down of Neutrons," The Encyclo-
pedia of PhyBics. (Handbuch der PhyBik). Berlin: Springer Verlag, 1959, Vol. 38,
Part 2, pp. 1-659.
L. F. CuRTiss, lntrod'UCtion to Neutron Phy8ÍC8. New York: Van Nostrand,
1959.
D. J. HuoHES, Neutron Cross Sections. London: Pergamon Press, 1957.
D. J. HuoHEs, Neutron Optics. New York: Interscience Publishers, 1954.
F. AJZENBERo-SELOVE, ed. N'UClear Spectroscopy. New York: Academic Press,
1960. Part 111. Neutron Spectroscopy. Articles by various authors, pp. 335-
490.
Proceedings of the International Conference on the Peaceful Uses of Atomic
Energy, Geneva, 1955, Vol. 4. "Cross Sections Important to Reactor Design."
New York: United Nations, 1956 (Many articles on equipment, techniques, and
results).
Proceedings of the Second International Conference on the Peaceful Uses o/ Atomic
Energy, Geneva, 1958, Vol. 14. "Nuclear Physics and Instrumentation" (Many
articles on nuclear theory, nuclear reactions, and neutron spectrometry).
PARTICULAR
l. H. L. ANDERSON, Neutronsfrom Alpha-Emüters, Preliminary Report No. 3
in Nuclear Science Series. Washington, D.C.: National Research Council, 1948.
2. A. WATTENBERG, Photo-Neutron Sources, Preliminary Report No. 6 in
Nuclear Science Series. Washington, D.C.: National Research Council, 1949.
3. A. WATTENBERG, "Photo-Neutron Sources and the Energy of Photo-
4. RussELL, SAcHs, WATTENBERG, and FIELDS1 "Yields of Neutrons from
Photoneutron Sources," Phys. Rev. 73, 545 (1948).
5. B. T. FELD, "The Neutron," op. cit. gen. ref., p. 369
6. A. O. HANSON and R. F. TASCHEK, Monoenergetic Neutrons from Charged
Particle Reactions, Preliminary Report No. 4 in Nuclear Science Series. Wash-
ington, D.C.: National Research Council, 1948.
7. HANSON, TASCHEK, and WILLIAMS, "Monoenergetic Neutrons from Charged
Particle Reactions," Revs. Mod. Phys. 21, 635 (1949).
8. B. T. FELD, "The Neutron," op. cit., gen. ref., pp. 404--426.

9. A. M. WEINBERG and E. P. WIGNER, The Phyaical Theory of Neutron
Chain Reactors. University of Chicago Press, 1958, Chapters 10, 11.
10. R. E. MARSHAK, "Theory of the Slowing Down of Neutrons by Ela.stic
Collisions with Atomic Nuclei," Revs. Mod. Phys. 19, 185 (1947).
11. S. GLAssTONE and M.C. EoLUND, The Elements o/ Nuclear Reactor Theory.
New York: Van Nostrand, 1952, Chapter 6.
12. R. D. PRESENT, Kinetic Theory of Gases. New York: McGraw-Hill,
1958, Chapter 5.
13. R. RINGO, unpublished work, cited by D. J. Hughes (op. cit., p. 89).
14. FERMI, MARSHALL, and MARSHALL, "A Thermal Velocity Selector and
its Application to the Measurement of the Cross Section of Boron," Phys. Rev.
72, 193 (1947).
15. T. BRILL and H. V. LICHTENBERGER, "Neutron Cross Sections with the
Rotating Shutter Mechanism," Phys. Rev. 72, 585 (1947).
16. P. A. EGELSTAFF, "The Operation of a Thermal Neutron Time-of-Flight
Spectrometer," J. Nuclear Energy, 1, 57 (1954).
17. D. J. HuoHEB and R. B. ScHWARTZ, "Neutron Cross Sections," Brook-
hauen National Laboratory Report BNL-SB6, 2nd ed. July 1, 1958. Washington,
D.C.: U.S. Government Printing Office.
18. (a) L. B. BoRsT and V. L. SAILOR, "Neutron Measurements with the
Brookhaven Crystal Spectrometer," Rev. Sci. Instr. 24, 141 (1953); (b) Sailor,
Foote, Landon, and Wood, Rev. Sci. Instr. 27, 26 (1956).
19. V. L. SAILOR, "The Parameters for the Slow Neutron Resonance in
Rhodium," Phys. Rev. 91, 53 (1953).
20. L. J. RAINWATER and W. W. HAVENs, JR., "Neutron Beam Spectrometer
Studies of Boron, Cadmium, and the Energy Distribution from Paraffin," Phya.
Rev. 70, 136 (1946).
21. RAINWATER, HAVENs, Wu, and DuNNING, "Slow Neutron Velocity
Spectrometer Studies, l. Cd, Ag, Sb, Ir, Mn," Phys. Rev. 71, 65 (1947).
22. E. MELKONIAN, "Slow Neutron Velocity Spectrometer Studies of 02,
N 2, A, H2, H2Ü and Seven Hydrocarbons," Phya. Rev. 76, 1750 (1949).
23. W. W. HAVENS, JR., and L. J. RAINWATER, "Slow Neutron Velocity Spec-
trometer Studies IV. Au, Ag, Br, Fe, Co, Ni, Zn," Phys. Rev. 83, 1123 {1951).
24. SEmL, HuoHEs, PALEVSKY, LEVIN, KA.ro, and 8.J0sTRAND, "Fast Chopper
Time-of-Flight Measurements of Neutron Resonances," Phys. Rev., 95, 476
(1954).
25. FLUHARTY, S1MPSON, and SIMPSON, "Neutron Resonance Measurements
of Ag, Ta, and U 238," Phys. Rev. 103, 1778 (1956).
26. L. M. B0LLINGER, "Techniques of Slow Neutron Spectroscopy," Nuclear
Spectroscopy, F. Ajzenberg-Selove, ed., op. cit. gen. ref., pp. 342-357.
27. BARSCHALL, RosEN, TASCHEK, and WILLIAMS, "Measurements of Fast
Neutron Flux," Reva. M od. Phys. 24, 1 (1952).
28. H. H. BARSCHALL, "Method for Measuring Fast Neutron Cross Sections,"
Reva. Mod. Phys. 24, 120 (1952).
29. HIBDON, LANGSDORF, and HoLLAND, "Neutron Transmission Cross Sec-
tions in the Kilovolt Region," Phys. Rev. 85, 595 (1952).
REFERENCES 601
30. M. GoLDHABER and G. H. BRIGGS, "Scattering of Slow Neutrons," Proc.

Roy. Soc. (London), Al62, 127 (1937).
31. J. TrrrMAN and C. SHEER, "The Energy Dependence of the Resonant
Scattering of Slow Neutrons from Gold," Phys. Rev. 83, 746 (1951).
32. B. N. BROCKHOUSE, "Resonant Scattering of S!ow Neutrons," Can. J.
Phys. 31, 432 (1953).
33. C. SHEER and J. MooRE, "Measurement of Scattering Cross Sections for
Low-Energy Neutron Resonances," Phys. Rev. 98, 565 (1955).
34. J. A. MooRE, "Resonance Scattering of Slow Neutrons on Indium,"
Phys. Rev. 109, 417 (1958).
35. H. L. FooTE, JR., "Neutron Scattering Cross Section of U 235," Phys.
Rev. 109, 1641 (1958).
36. RAE, CoLLINs, KINSEY, LYNN, and WIBLIN, "An Analysis of Slow Neutron
Resonances in Silver," Nuclear Physics 5, 89 (1958).
37. C. G. SHULL and E. O. WoLLAN, "Coherent Scattering Amplitudes as
Determined by Neutron Diffraction," Phys. Rev. 81, 527 (1951).
38. E. O. WoLLAN and C. G. SHULL, "Neutron Diffraction and Associated
Studies," Nucleonics, 3, No. 1, 8-21, July 1948; 3, No. 2, 17, August 1948.
39. G. E. BACON, Neutron Dif!raction. Oxford: Clarendon Press, 1955.
40. ANDERSON, FERMI, WATTENBERG, and ZINN, "Method for Measuring
Neutron Absorption Cross Sections by the Effect on the Reactivity of a Chain-
Reacting Pile," Phys. Rev. 72, 16 (1947).
41. HoovER, JoRDAN, MoAK, PARDUE, PoMERANCE, STRONG, and WoLLAN,
"Measurement of Neutron Absorption Cross Sections with a Pile Oscillator,"
Phys. Rev. 74, 864 (1948).
42. H. PoMERANCE, "Thermal Neutron Capture Cross Sections," Phys. Rev.
83, 64 (1951).
43. SEREN, FRIEDLANDER, and TuRKEL, "Thermal Neutron Activation Cross
Sections," Phys. Rev. 72, 888 (1947).
44. B. T. FELD, op. cit., gen. ref., pp. 291, 323-325.
45. J. L. FowLER and J. E. BROLLEY, JR., "Monoenergetic Neutron Tech-
niques in the 10-30 Mev Range," Revs. Mod. Phys. 28, 103 (1956).
PROBLEMS
l. The Q-value for the photodisintegration of the deuteron is 2.227 ± 0.003

Mev. The mass doublet 2H 1-H 2 is, from nuclear data, 1.5494 ± 0.0024 milli-
mass units (mmu), and the mass of the hydrogen atom is 1.008142 ± 0.000003
amu. Find the mass of the neutron from these data.
2. It is customary to describe neutrons of different energy by the adjectives
shown below:
Energy, ev Neutron type
0.001 cold
0.025 thermal
1.0 slow (resonance)
100 slow
104 intermediate
106 fast
108 ultrafast
1010 ultrafast (relativistic)
Compute the speed, de Broglie wavelength, and absolute temperature for each
neutron type.
3. The reaction Be9 (a, n)C 12 with polonium a-particles with an energy of
5.30 Mev is a useful source of neutrons. Calculate (a) the Q-value of the reac-
tion, (b) the energy of the neutrons which emerge at angles of 0°, 90°, and
180° with the direction of the incident beam.
4. The following(d, n) reactions are good sources of fast neutrons: H 2(d, n)He3,
º.
H 3(d, n)He4 , C 12 (d, n)N 13 , N 14 (d, n)0 15, Li7 (d, n)Be8, Be9 (d, n)B 1 Assume
that the product nucleus is left in the ground state, and calculate the energy of the
neutrons which would correspond to zero kinetic energy of the incident deuterons.
5. The following (p, n) reactions are useful sources of relatively low-energy
neutrons: H 3(p, n)He 3, Li7 (p, n)Be 7, Be9 (p, n)B 9, C 12 (p, n)N 12, Na 23 (p, n)Mg 23,
V51 (p, n)Cr 51 . lt has been shown that the minimum neutron energy is given by
En. min = E,/(A + 1) 2, where E, is the threshold energy and A is the mass
number of the target nucleus. Calculate the value of E,., min for each reaction.
The Q-value for the (p, n) reaction on vanadium is -1.532 Mev.
6. A cylindrical BFa counter, 10 in. long and ¼-in. in diameter, contains
BFa at a pressure of 20 cm Hg (corrected to N.T.P.). The counter has an
over-all efficiency of 2%, that is, it detects 2 out of every 100 neutrons incident
on it. When the counter is exposed to neutrons with a Maxwellian velocity
distribution at 20ºC, the counting rate is 20,400 counts/min. (a) What is the
neutron flux? (b) What would be the neutron flux if the neutrons had a Max-
wellian distribution at 127ºC, and the counting rate were the same? A counter
of this type is usually used for relative flux measurements. (c) What would
be the relative counting rates if the flux were kept constant and the temperature
of the neutrons in different experiments had the values OºC, 25ºC, 50ºC, lOOºC,
200°C? Take the rate at 25ºC equal to 100 units.
PROBLEMS 603
7. Compute the average logarithmic energy loss per collision, the number of
collisions needed to reduce the neutron energy from 2 Mev to 0.025 ev, the
slowing-down power, and the moderating ratio of fluorine, magnesium, and
bismuth at room temperature. Compare the results with those for the materials
listed in Tables 18-2 and 18-3. Use the following values of cross sections.
O-a o-.
F 9mb 5b
Mg 59mb 6b
Bi 30mb 9b
8. A gold foil 0.02 cm thick and 1 cm 2 in cross section is irradiated for one
hour in a nuclear reactor. The thermal neutrons have a Maxwellian energy
distribution and their flux is 10 12 neutrons/cm 2 -sec; it is assumed that ali of the
neutrons at the irradiation position are thermal. The gold has a density of
19.3 gm/cm 3 ; its thermal absorption cross section is 98.7 b and the slight devia-
tion from the 1/v-law may be neglected. (a) What is the activity of the foil on
being removed from the reactor? (b) If the activity of the foil is measured
60 days after the irradiation, with a counter whose over-all efficiency is one per-
cent, what is the counting rate?
9. Consider the problem of the diffusion, in a weakly absorbing medium, of
monoenergetic neutrons from a point source which emits Q neutrons per second.
(a) Show that the differential equation for the neutron density is
d2n + ~ dn _ ~ = 0
dr 2 r dr 1,2 '
where r is the radial distance from the source at r = O, and L is the diffusion
length of thermal neutrons in the medium. (b) Show that the neutron density
is given by
Qe-r/L
n = ---,
4rDr
where D is the diffusion coefficient.
10. A crystal spectrometer has a beryllium crystal with a lattice spacing of
0.7323 A for the reflecting planes used for slow neutrons. What are the Bragg
angles for the first arder reftection of neutrons of energy 1, 3, 5, 10, 30, and 50 ev,
respectively?
11. In measurements of neutron wavelength with a, crystal spectrometer,
there is an uncertainty /:,.). in the wavelength because of the uncertainty t:.8 in
the Bragg angle. Consequently, there is an uncertainty t:.E in the nteasured
energy of the neutrons. The resolution of the instrument is defined as the
fractional uncertainty t:.E/E in the energy. Show that (a) t:.E/E = 2 cot 8 t:.8
... 2 t:.8/8, (b) t:.E/ E ... k t:.8E 112 , where k is a certain constant. In the spectrom-
eter of Problem 10, the uncertainty in the Bragg angle, t:.8, is 7.8 min. Calcu-
late the uncertainty in the energy for each of the energies of Problem 10.
12. In the time-of-flight velocity selector, the following relations hold for
the energy E (ev), the velocity v (m/sec) and the time-of-flight for a 1-m path
t (µsec/m):
V = 10 6 /t, E = 51.5 X 10 2/t 2 , t = 71.5/E 112 •
Derive the following relations.
6
10 At
Av= - -t2-
Av At
--,
V t
AE 2At
E= - -t , AE = -o.02sE312 At
Calculate the energy spread AE at energies of 1 ev, 10 ev, 100 ev, 1 kev, and
10 kev for each of the time uncertainties At = 0.5, 0.1, and 0.05 p.see/m.
13. A beam of thermal neutrons with an initial intensity of 106 neutrons/
cm 2 -sec is passed through a tantalum sheet 1 mm thick. The intensity of the
beam emerging from the tantalum is 8.65 X 105 neutrons/cm 2--sec. The activity
of the foil after irradiation for 2.0 days and a delay of 2 hr to allow for the
decay of short-lived isomeric activity is 1224 counts/cm 2/sec because of the
activity of Ta 182 • The density of tantalum is 16.6 gm/cm3 and the half-life of
Ta 182 is 114 days. What are (a) the total thermal cross section of tantalum,
(b) the activation cross section for the production of Ta 182 ?
14. Arsenic has an absorption cross section for thermal neutrons of 4.1 b and
a scattering cross section of 6 b; its density is 5.73 gm/cm 3 • A beam of thermal
neutrons is passed through a slab of arsenic 2 cm thick. By what fraction is
the intensity of the beam reduced? How much of this reaction is caused by
scattering? How much by absorption?
15. The total cross section of a nuclide with an. isolated resonance for which
the resonance scattering is negligible can be expressed in the form
2 2
Eo)u uor
u,(E) = u.+ ( E 4(E - Eo) 2 + 1'2 1
where 11', is the scattering cross section, and uo is a composite quantity involving
parameters discussed in Section 16-3. The total cross section of rhodium has
been measured with the crystal spectrometer of Problema 10 and 11, and the
following values he.ve been obtained for the resonance parameters: Eo = 1.260
± 0.004 ev, a-0 = 5000 ± 200 b, r = 0.156 ± 0.005 ev, and <1, = 5.5 ± 1.0 b.
Plot the value of the total cross section for values of the energy between 0.2 ev
and 40 ev. Calcule.te, from the formula, the thermal absorption cross section.
How does the result compare with the measured value of 156 ± 7 b.
16. How many collisions are needed for neutrons to lose, on the average, 99%
of an initial energy of 2 Mev in graphite? What does this result show, as coro-
pared with the number required to modere.te the neutrons to thermal energies?
PROBLEMS 605
17. Show that the neutron velocity at which the maximum of the Maxwellian
flux distribution occurs is given by v3kT/m.
18. The expression (18-43) for the thermal diffusion length was obtained for
monoenergetic neutrons. Suppose that the thermal neutrons have a Maxwellian
velocity distribution at 20ºC in a very large block of graphite. Graphite has a
transport croBB section which may be taken to be constant and its absorption
croBS section varies as l/11. Show that the thermal diffusion length of the neu-
trons is given by
L = 0.613 ,
N~
where the subscript zero denotes the cross section for 2200 m/sec neutrons.
19. Suppose that the only two reactions that a certain nuclide undergoes when
bombarded with neutrons are elastic scattering and radiative capture. Show
that the value of the total cross section at the peak of the resonance is given by
>.~ r,.
u,o =-¡ r'
where the subscript zero refers to values at E = Eo, the energy at the peak.
[Hint: Apply Eq. (16-13) to each of the reactions.)
20. lt is known from experiment that when a given nuclide is bombarded
with neutrons, r,. has a constant value, while r,. is proportional to E 112 • Show
that the croBB section for radiative capture may be written in the form
112
u,(E) = Eo)
u,o ( E
[
1 + [(E - 1
Eo)/(f/2)) 2
]
'
where the subscript zero refers to values at the peak of the resonance.
CHAPTER 19
NUCLEAR FISSION
19-1 The discovery of nuclear fission. The discovery of nuclear fission

was one of the results of attempts to make transuranium elements of
atomic number greater than 92 by means of (n, 'Y) reactions followed by
,S-decay of the product nucleus. These attempts eventually suceeded,
and some of the methods used and results obtained were discussed in
Section 12-3. The interpretation of the early experiments was difficult,
however, because of unexpected results which could be explained only in
tenns of nuclear fission.
To understand some of the difficulties that were met, it is necessary to
consider briefly how radioactive elements are separated from inactive ones.
A radioelement fonned by a nuclear reaction is usually available only in
a very small amount, possibly as small as 10- 12 gm, and cannot be sep-
arated by means of ordinary chemical methods. But the separation
can often be made with the aid of a carrier which is a stable substance
with chemical properties similar to those of the radioelement. The element
and the carrier usually belong to the same subgroup of the periodic system
and can undergo similar chemical teactions. If an appreciable amount,
perhaps 10 to 100 mg, of the carrier is added to a solution containing the
radioelement, and if the carrier is then precipitated from the solution by
the fonnation of an insoluble salt, the radioelement is precipitated along
with the carrier. The carrier and the radioelement can then be separated
by means of other chemical reactions.
In the separation of radium from other members of the uranium series,
barium serves as a carrier. Radium and barium both belong to group IIA
of the periodic system and form insoluble sulfates. When a solution of
a barium salt is added to a solution containing a very small amount of
radium and a sulfate is added, barium and radium sulfates are precipi-
tated together. A neater method depends on the fact that barium and
radium chlorides are precipitated together from concentrated solutions of
hydrochloric acid. The precipitate can be dissolved in water, and the
barium and radium can then be separated by repeated fractional crystal-
lization from hydrochloric acid. In analogous ways, lanthanum acts as a
carrier for actinium (Z = 89). Sometimes the carrier is a stable fonn of
the radioelement, e.g., stable iodine can be used as a carrier for radioactive
isotopes of iodine.
In the early experiments on the fonnation of transuranium elements,
uranium was bombarded with neutrons and several different tJ,-activities,
606
l!H] THE DISCOVERY OF NUCLEAR FISSION 607
distinguished by their half-lives, were detected. Carrier techniques were

used to separate the elements responsible for the activities, but the number
and properties of the new radioelements were such that they could not be
fitted into a scheme consistent with the known properties of the heavy
elements and the predicted properties of the transuranium elements. One
difficulty in the analysis of the products of the bombardment of uranium
with neutrons led to a remarkable conclusion. In addition to the elements
which seemed to be real transuranium elements, there were four which
were supposed to be ~-radioactive iostopes of radium because they were
precipitated with barium when the latter was used as the carrier. The
decay products of these nuclides seemed to be isotopes of actinium because
they were precipitated with lanthanum, the carrier for actinium. Although
these results appeared to be consistent, since actinium follows radium in
the periodic system, they raised two serious questions. First, the produc-
tion of an isotope of radium by the neutron bombardment of uranium
would require an (n, 2a) reaction, but this reaction is a very unlikely one
(cf. Section 16-5), especially at low neutron energies. Second, further
chemical experiments showed that the "radium" activities could not be
separated from the barium carrier, and the daughter activities could not
be separated from the lanthanum carrier. In 1939, Hahn and Strass-
mann0•2> performed a beautiful and thorough set of experiments which
proved beyond a doubt that the "radium" isotopes are really isotopes of
barium and the "actinium" isotopes are isotopes of lanthanum. Further-
more, they showed that one of the barium isotopes resulting from the
neutron bombardment of uranium could be identified, beca.use of its half-
-life of 86 min, with the previously known nuclide Ba 139 , which has the
same half-life. Similarly, one of the lanthanum isotopes from the neutron
bombardment of uranium was identified with the known nuclide La 140 ,
which has a half-life of 40 hr.
The production of the nuclides 57 La 140 and 56 Ba 139 from uranium, which
has the atomic number 92 andan atomic weight of nearly 240, required
a hitherto unknown kind of nuclear reaction in which the uranium nucleus
is split into fragments which are themselves nuclei of intermediate atomic
weight. If such a process really occurs, it should also be possible to find
nuclei with masses between 90 and 100 and atomic numbers of about 35.
Hahn and Strassmann were able to find an active isotope of strontium
(Z = 38) and one of yttrium (Z = 39) which met these requirements, as
well as isotopes of krypton (Z = 36) and xenon (Z = 54). It was clear
from the chemical evidence that uranium nuclei, when bombarded with
neutrons, can indeed split into two nuclei of intermedia.te atomic weight.
It was then predicted, from the systematics of stable nuclides and from
the semiempirical binding-energy formula, that the product nuclei would
have very great energies and would produce large numbers of ion pairs in
608 NUCLEAR FISSION [CHAP. 19
pa.ssing through a gas. When a thin layer of ura.nium was put in a suit-
able ionization chamber connected to an a.mplifier and irradiated with
neutrons, grea.t bursts of ionization were obeerved13 ·'>corresponding to
energies up to 100 Mev. These pulses of ionization are extremely large
compa.red with those of single a-particles and are easy to recognize. Further
chemical work then showed that besides the rea.ction products mentioned
above, other elements of medium mass number were formed, including
bromine, molybdenum, rubidium, antimony, tellurium, iodine, and cesium.
There was, therefore, ample chemical a.nd physical evidence for the split-
ting of the uranium nucleus, and this procesa was called fi88Í<m. <5 >
lt is now known that fission can be produced in various nuclides under
different conditions, and sorne of the resulta will be mentioned. When
small samples of the separated isotopes of uranium, prepared in a mass
spectrograph, were bombarded, it was found 16> that slow neutrons cause
fission of U 235 but not of U 238 ; fast neutrons, with energies gres.ter than
one Mev, cause fission of both U 235 and U 238 • Thorium and Pa 231 undergo
fission, but only when bombarded with fast neutrons. Fission can also
be produced in uranium and thorium by high-energy a-particles, protons,
deuterons, and 'Y-rays. The nuclei Pu 239 and U 233 , formed by (n, 'Y) reac-
tions on U 238 and Th 232 , respectively, followed by ¡9-decay of the products,
undergo fission when bombarded with slow or fast neutrons, as do other
artificial heavy nuclides. Finally, sorne heavy nuclei have been found to
undergo spontaneous fission; in this procesa, the nucleus divides in the
ground state without bombardment by particles from outside.
In addition to the two large fission fragments, neutrons and 'Y-rays are
emitted. Division into three fragments of comparable size (temary fis-
sion) has been obeerved but is a very rare event, occurring about five times
per million binary·fissions. Long-range a-particles are sometimes emitted,
about once in every 400 fissions. The emission of light nuclei, with masses
gres.ter than 4 and probably lesa than 12, is a relatively common event,
occurring once in about every 80 fissions.
19-2 Fission cross sections and thresbolds. The probability of fission,

as compa.red with that of other reactions, is a matter of theoretical interest
and practical importance. The U 235 nucleus may capture a thermal neu-
tron to form the compound nucleus [U 236], or the neutron may be scattered.
The compound nucleus may either undergo fission, or it may emit 'Y-rays
and decay to the ground state of U 236 , which emits a 4.5-Mev a-particle
and has a half-life of 2.4 X 10 7 years. The thermal cross sections< 7 > for
the different reactions for U 235 , U 238 , natural uranium, and Pu 239 are given
in Table 1~1. The total cross section of U 235 for 2200 m/sec neutrons
is 698 b, the total absorption cross section is 683 b, and the fission cross sec-
tion is 577 b. The ratio of the radiative capture cross section to the
19-2) FISSION CROSS SECTIONS AND THRESHOLDS 609
TABLE 19--1
PROPERTIES OF FISSIONABLE MATERIALS< 7 ,B>•
u2aa u2as Natural Pu239

Uranium
Uabe 578 ± 4 683 ± 3 7.68 ± 0.07 1028 ± 8

u¡ 525 ± 4 577 ± 5 4.18 ± 0.06 742 ± 4
Ur 53 ± 2 101 ± 5 3.50 286 ± 4
u. 15 ± 2 8.3 ± 0.2 9.6 ± 0.5
2.51 ± 0.02 2.44 ± 0.02 2.89 ± 0.03
"
(average
number of
neutrons per
fission)
a = u,/u¡ · 0.101 ± 0.004 0.18 ± 0.01 0.39 ± 0.03
,,,=--
l+a
" 2.28 ± 0.02 2.07 ± 0.01 1.34 ± 0.02 2.08 ± 0.02
*For 2200 m/sec neutrons.

fission cross section is 0.18, so that the probability of radiative capture
is about 18% that of fission. The natural abundance of U 235 is only
O. 72%, with the result that the fission and radiative capture cross sections
of natural uranium are much smaller than those of the separated isotope
U 235 . In Pu 239 the ratio of the capture cross section to that for fission is
0.39; although the fission cross section is greater than in U 235 , the ratio
of capture to fission is also greater.
Sorne of the artificially produced radioactive nuclides undergo fission with
thermal neutrons, as do sorne of the shorter-lived naturally occurring
radionuclides. Values of the fission cross section for 2200 m/sec neutrons
are listed in Table 19--2, together with values of activation cross sections
and total absorption cross sections where these are available. <7 > lt is
evident that many heavy nuclides are fissionable, and that the probability
of fission varies overa wide range of values. Only U 235 , U 233 and Pu 239
have high cross sections as well as long half-lives, and either occur natu-
rally (U 235) or can be produced in significant amounts in practical lengths
of time (Pu 239 and U 233 ). Hence, only these three fissionable materials
are important in the large-scale applications of nuclear fission.
The fission cross section varies with energy in a complicated way, as
is shown in Fig. 19--1 for U 235 . In the thermal region <T¡ varies approxi-
mately as 1/v; starting at 0.28 ev, there are many closely spaced reso-
nances, with at least 20 resolved resonances below 20 ev. At high energies,
the fission cross section is relatively small, only about one bam in the
neighborhood of 1 Mev; the fission cross sections of U 233 and U 235 in the
19-2] FISSION CROSS SECTIONS AND THRESHOLDS 611
E
rñ
o
º5ii
...o
s::o
i,!:
.Q
~
ol
"'..."';::,
-
.s
o
s::
~
o
... E c.,
00
00
00
...oo
o
s::
ºéil
00
<i::
"'
..s::
E-<
...;
-
d,
~
d
612 NUCLEAR FISSION [CHAP, 19
TABLE 19-2
THERMAL CRoss SECTIONS oF SoME HEAVY NuCLIDEsm•
Total
Fission Activation
absorption
Nuclide Half-life cross section, cross section,
cross section,
u¡, b u act, b
ua, b
g0Th221 18.2d 1500 ± 1000

g0Th22s 7.34 X 103 y 45 ± 11
s1Pa2ao 17.3d 1500 ± 250
91Pa2a1 3.4 X 104y (10 ± 5) X 10-a 200 ± 15
91Pa2a2 1.3d 700 ± 100 760 ± 100
92u2ao 20.8d 25 ± 10
92u2a1 4.2d 400 ± 300
g2u2a2 74y 80 ± 20 300 ± 200
92u2aa 1.62 X 105y 525 ± 4 578 ± 4 53 ± 2
92u::34 2.52 X 10 5y <0.65 105 ± 4 90 ± 30
92u2as 7.1 X lOSy 582 ± 4 683 ± 3 101 ± 5
gaNp234 4.4d 900 ± 300
gaNp236 22h 2800 ± 800
gaNp2a1 2.2 X 106y (19 ± 3) X 10-3 170 ± 5 169 ± 6
93Np2as 2.ld 1600 ± 100
gaNp2a9 2.3d <l 35 ± 10
94Pu2as 86.4y 16.8 ± 0.3 400 ± 10
94Pu2a9 2.44 X 104y 742 ± 4 1028 ± 8 286 ± 4
114Pu2•0 6.6 X 103y 0.030 ± 0.045 286 ± 7 250 ± 40
94Pu241 13y 1010 ± 13 1400 ± 80 400 ± 50
94Pu242 3.75 X 105y <0.2 30 ± 2 19 ± 1
95Am 241 458y 3.2 ± 0.2 630 ± 35 750 ± 80
95Am2•2 lOOy 6400 ± 500
•For 2200 m/sec neutrons.
neighborhood of 1 Mev are shown in Fig. 19-2a. Sorne heavy nuclides, for
example, U 234, U 236, and U 238 , do not undergo fission with slow neutrons,
but only with fast neutrons. Fission is a threshold reaction in these
nuclides, and the fission cross section varies with energy in much the same
way that the cross sections for other threshold reactions do. The fast
fission thresholds of U 23 4, U 236 , and U 238 , and the variation of the fission
cross section with energy, above threshold, have been determined by
bombarding the nuclides with neutrons from the reaction H 3(p, n)He 3.
The results are shown in Fig. 19-2b.
Threshold energies and cross sections have been measured for fission
processes induced by charged particles.<1°> The threshold for fission by
deuterons is close to 8 Mev for Th 232 , U 235 , and U 238 ; that for fission by
19-2) FISSION CROSS SECTIONS AND THRESHOLDS 613
5
eb' 3
2
1
4 kev 10 kev 100 kev 1 Mev 3 Mev
E,.
(a)
1.75
1.50
tT/ ((J234)
1.25
~ 1.00
e tT/ (U236)
.;-,o.75
tT/ (l.1238)
0.50
0.25
o
O 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
E,. (Mev)
(b)
Fm. 19-2. The fission cross sections of several uranium isotopes in the
Mev region.< 7 ,9> {a) U233, U 236 {b) u2 a., U236, U23 ª.
a-particles is in the neighborhood of 21 Mev for these nuclides. In the
range of energies explored so far, the value of the cross section increases
with energy above the threshold and then either levels off or passes through
a maximum. In the case of U 2311 bombarded with deuterons, the cross
section rises to a value of one barn at about 20 Mev and then increases
slowly to a value between 1.5 and 2 b at deuteron energies between 120
and 200 Mev. The cross section for the fission of U 2311 by a-particles rises
to a value of 1.5 b at about 50 Mev and remains at about this value up to
energies of 400 Mev.
Gamma-raya from nuclear reactions and high-energy x-rays from a
betatron can cause fission, and sorne of the measured thresholds 0 u are
TABLE 19-3
THRESHOLD ENERGIES FOR PHOTOFISSION
Photofission
Nuclide threshold,
Mev
o2u2as 5.08 ± 0.15
o2u2as 5.31 ± 0.25
o2u2aa 5.18 ± 0.27
94Pu2ao 5.31 ± 0.27
o0Th2a2 5.40 ± 0.22
shown in Table 19-3. The cross sections for these reactions are generally
smaller than those for neutron-induced fission. When U 238 is bombarded
by monoenergetic 6.3-Mev 'Y-rays from the F 19 (p, 'Y)Ne 20 reaction, the
cross section is about 3 mb; for the 17.5-Mev 'Y-rays from the Li 7(p, 'Y)Be 8
reaction, the cross section is about 30 mb.
Many heavy nuclides undergo spontaneous fission, 02 > and this process
is an alternative, less probable, method of nuclear disintegration than
a-particle emission. Nuclei which undergo fission with slow neutrons have
smaller spontaneous fission rates than their isotopes which undergo fission
only with fast neutrons. For example, U 235 has a half-life for spontaneous
fission of about 1.8 X 10 17 years, corresponding to a rate of about one
fission per gram per hour while U238 has a fission half-life of 8.0 x 10 15
years, ora rate of 25 fissions/gm/hr. Similarly, Pu 239 has a spontaneous
fission rate of 36 fissions/gm/hr, or a fission half-life of 5.5 X 10 15 years,
while Pu 240 has a rate of 1.6 X 10 6 fissions/gm/hr and a half-life of
1.2 X 10 11 years.
Finally, fission can be produced in bismuth, lead, thallium, mercury,
gold, and platinum by neutrons with very high energies, e.g., 40 Mev.
19-3 The fission products. It was indicated in Section 19-1 that a

number of nuclides of intermediate charge and mass are formed when a
uranium nucleus undergoes fission. The study of the nuclei formed in
fission was evidently a promising source of information about the mecha-
nism of the fission process and offered the possibility that new, hitherto
unknown, nuclides might be discovered. The latter possibility became
apparent when the neutron-to-proton ratios of the uranium isotopes were
compared with those of some of the fission products. The compound
nucleus [92U 236 ] has 144 neutrons and 92 protons, and the value of the
ratio is 144/92 = 1.57. The values of the ratio for the atable isotopes of
some typical fission products (krypton, iodine, xenon, and cesium) vary
19-3) THE FISSION PRODUCTS 615
from 1.17 to 1.52, and are appreciably lower than that for [U 236 ]. When
that excited nucleus splits into two smaller nuclei, the neutron-to-proton
ratio for at least one of them must be greater than the value compatible
with stability. Such an unstable nucleus might be expected to approach
stability by electron emission or, if the excitation energy is high enough,
by ejection of one or more neutrons; it has been found experimentally that
both of these processes occur.
The investigation of the products of the fission of U 235 has shown that
the range of mass numbers is from 72, probably an isotope of zinc with
atomic number 30, to 158, thought to be an isotope of europium with
atomic number 63. About 97% of the U 235 nuclei undergoing fission
yield products which fall into two groups, a "light" group with mass num-
bers from 85 to 104, anda "heavy" group with mass numbers from 130
to 149. The most probable type of fission, which occurs in about 7%
of the total, gives products with mass numbers 95 and 139. There are
87 possible mass numbers between 72 and 158, which may represent
the total number of different nuclides formed as direct fission fragments.
If this were the case, the uranium nucleus should be capable of splitting
in over 40 different ways. More than 60 primary products have actually
been detected, so that there are at least 30 different modes of fission, a
different pair of nuclei being formed in each mode.
The fission fragments have too many neutrons for stability and most of
them decay by electron emission. Each fragment starts a short radio-
active series, involving the successive emission of electrons. These series
are called fisaion decay chains, and each chain has three members, on the
average, although longer and shorter chains occur frequently. The prob-
lem of determining the masses and atomic numbers of the fission products
and of ídentifying the members of the many decay chains is an extremely
difficult one. Nevertheless, as the result of careful and persistent work,
more than 60 chains have been established, and about 200 different radionu-
clides have been assigned to them. <13, 14> An example of a long chain is
.veuo~ 5508 1 • 0 ~ 56Bal4º ~

5 4-"- 161 661 12.Sd
- - + 57 La140 ~h 58Ce
140
(stable).
40
This chain is especially interesting because it contains two of the nuclides,

Ba 140 and La 140, whose appearance led to the discovery of fission. The
short chain
60
Nd147 ..!:._
lld 61
Pm1•1..!:._ ªm1u (~ 10 11y)
62"'
47
is important beeause one of its members is an isotope of the element with

atomic number 61. This element had not been clearly isolated oefore the
discovery of fission, and has now been named promethium (symbol, Pm).
Radioactive isotopes of the element with atomic number 43, which has
not been found to occur in nature, also have been identified as fission
products. This element is now called technetium (symbol. Te), and its
longest-lived isotope occurs in the chain.
Mo 99 ~ 43Tc
99
_!:__._ oRu 99 (atable).
O 88h 2.2XlO"y
19-4 The mass and energy distributions of the fission products. The
mass distribution of the fission products is shown most conveniently in the
form of a fission yiel.d curoe, in which the percentage yields of the different
products are plotted against mass number. The yield of any given mass
can be found by measuring the abundance of a long-lived nuclide near the
end of a chain, or that of the atable end product. Yield curves< 16 • 181 for
the fission of U 236 by thermal neutrons and by 14-Mev neutrons are shown
in Fig. 19-3a¡ curves for thermal fission of U 233 and Pu 239 are shown in
Fig. 19-3b. The fission yield for a particular nuclide is the probability
(expressed as a percentage) of forming that nuclide or the chain of which
it is a member; it may also be regarded as the percentage of fissions yield-
ing the nuclide or chain. Since two nuclei result from each fission, the total
yield adds up to 200%. The yields_ vary from about 10-6% to about 7%,
10 10
4 Mev: T, T ~
U233 \ Pu239
1.0
' al
1.0
j
' Pu:139
1
~
- 0.1
1
Thennal t- i .-
,I
.¡0.01 - Pu239 U23:3..
ti;
0.001 0.001
• Thennal neutrons ~ •lJ233 •Pu:139
•14-Mev neutrons
0.0001 11 11 0.0001
1, 1 rr
80 100 120 140 80 100 120 140 160
M888 number M888 number
(a) (b)
FIG. 19-3. Fission yields from U 236, U233 , and Pu 239 : (a) thermal and fast
fiasion of U 236 ; (b) thermal fiasion of U 233 and Pu239 .U 6>
19-4] MASS AND ENERGY DISTRIBUTIONS OF FISSION PRODUCTS 617
and the range is so large that a logarithmic scale is used for the ordinate
of the yield curve. Each curve shows two peaks, corresponding to the light
and heavy groups of products. In the case of U 235 the maxima lie near
mass numbers 95 and 135; fission induced by slow neutrons is a highly
asymmetric process, and division into two equal fragments occurs in only
about 0.01% of the fissions. Small "fine-structure" peaks are evident in
sorne of the yield curves; for U 235, these peaks are at masses 100 and
134. This fine structure is, for the most part, a shell effect, and can be
accounted for by a detailed analysis of the neutron binding energies in
nuclei with neut.ron numbers close to those corresponding to closed shells,
i.e., 50 and 82 neutrons. The U 235 curves illustrate the effect on the mass
distribution of increasing the neutron energy above thermal. The greatest
change is the increase in the probability of symmetric fission; for fission by
14-Mev neutrons, the increase is about 100-fold. The other changes are
a small drop in the peak yields and a moderate increase in the most asym-
metric modes of fission, i.e., a rise in the wings of the yield versus mass
curve.
The mass distribution of the fission fragments can also be obtained from
the distribution of their kinetic energies, which can be determined by
measuring the ionization produced in an appropriate ionization chamber.
In one type of chamber, the fissile material is placed on one of the electrodes
and the ions which result when a fission fragment enters the region be-
tween the electrodes are collected. Since the fission fragments occur in
pairs, an experiment of this kind measures the energy of one fragment only.
In another type of chamber, a very thin foil is made the common cathode
of two back-to-back ionization chambers. The two fragments resulting
from neutron bombardment travel in opposite directions into the two
chambers and the ionizations they cause are measured simultaneously.
The nucleus undergoing fission can be considered to be initially at rest;
and, if the neutrons emitted are neglected, the law of conservation of
momentum gives
M1V1 = M2V2, (19-1)
where the subscripts refer to the two fission fragmenta. The energies of
the fragmenta are then in the ratio
E1 ½M1 Vf
E2 = ½M 2v: = MM2 1'
(19-2)
and the masses are inversely proportional to the kinetic energies; when
the energy distribution has been measured, the mass distribution is ob-
tained from Eq. (19-2).
The energy distribution of the fission products has been measured for
the fission of U 233 , U 2311, and Pu 239 by thermal neutrons, 0 7 • 1S> and for
900
800
700
~::, 600
o
.:: 500
o
,3 400
§ 300
z
200
100 \
o
Energy (Mev)
Frn. 19-4. Energy distribution of the fragments from fission of U 235 by

thermal neutrons (Brunton and Hanna.0 7>).
the fission of U 235 , U 238 , Th 232 , and Pu 239 by fast neutrons.° 9 • 20 > The
energy distribution in the case of the thermal fission of U 235 is shown in
Fig. 19-4; this energy spectrum is typical for fission induced by slow
neutrons, and tne curves for U 233 and Pu 239 are similar to the curve shown.
The peaks near 60 Mev and 95 Mev show the asymmetry of the fission
process. The best ionization chamber measurements<17l give 154.7 Mev
for the average total kinetic energy of the fragments from U 235 . The
energy distribution for fission induced by fast neutrons (2.5 and 14 Mev)
show the same two peaks, and the maxima differ only slightly from
the values for fission induced by slow neutrons. The most important differ-
ence in the shapes of the spectra is an increase in the height of the curve
in the region between the two peaks when the energy of the bombarding
neutrons is in the Mev range. This rise is interpreted as indicating in-
creased probability of symmetrical fission when the neutron energy is
high. In the case of U 235 , symmetrical fission is about 100 times more
probable for 14-Mev neutrons than for thermal neutrons. When the
energy of the bombarding neutrons is raised to 45 Mev, there are still
two peaks, but the dip between them is small, and the probability of sym-
metrical fission is still greater than at the lower energies. With 90-Mev
neutrons, only one peak is observed, corresponding to division into two
equal fragments; at these very high energies symmetrical fission is the
most probable mode.< 20
The velocity distribution of the fission fragments has been studied
directly by means of a time-of-flight method. <22 > A thin foil with a film
of U 233 , U 235 , or Pu 239 is irradiated by thermal neutrons, and pairs of
fragments are detected by scintillation detectors. One fragment travels
only about 1 cm before striking a detector, while the other travels about
19-5] NEUTRON EMISSION IN FISSION 619
350 cm along an evacuated tube. The pulses from the two detectors are
shown on the screen of a cathode-ray tube and photographed; the velocity
of the fragment can be obtained from the distance between the two peaks,
and the kinetic energy is then calculated from the velocity. Results
obtained with this method show that the energies are greater than those
given by the ionization measurements, and the average total kinetic energy
of the fission fragments from U 235 is increased by 12.4 Mev, from 154.7
Mev to 167.1 Mev. This value agrees with the result, 167.1 ± 1.6 Mev,
of a calorimetric measurement. c23 i
The energy dístribution of the fragments from spontaneous fission has
also been determined< 24 > and is similar to that found in neutron-induced
fission.
The determination of the charge distribution in fission is more difficult
than that of the energy distribution because in an ionization chamber or
cloud chamber there is no obvious way of determining the nuclear charge
at the instant of fission. Radiochemical methods based on attempts to
measure the primary fission yield have been used;<1 6 • 25 l these methods are
indirect, and the problem has not yet been solved. ' 26 >
19-5 Neutron emission in fission. The comparison of the value of the

neutron-to-proton ratio of the compound nucleus [U 236 ] with the values
for sorne of the fission products, discussed in Section 19-3, indicated also
that neutrons might be emitted in fission. The fact that neutrons are
emitted was shown in a simple experiment shortly after the discovery of
fission. ' 26 > A neutron source was placed at the center of a large vessel,
with detectors at various distances from it to determine the neutron density
in the vessel. The vessel was filled first with a uranyl sulfate solution,
and then with an ammonium nitrate solution for comparison. The aver-
age neutron density was found to be greater when uranium was present,
showing in a rough way that more neutrons were formed, when fission
occurred, than were used up. The average value of the number of neu-
trons released in fission, usually denoted by v, has been measured for
various fissionable materials, and sorne of the results' 27 • 28 > are listed in
Table 19-4. The average number of neutrons released is always greater
than two and increases with the energy of the neutrons that induce fis-
sion. The number of neutrons released in any one fission process must,
of course, be an integer but, since the fissionable nucleus can divide in at
least 30 different ways, the average value v of the number of neutrons
does not have to be an integer.
In addition to v, there is another property of fissionable materials which
has practical importance, the average number of neutrons emitted per
neutron absorbed by a fissionable nuclide. It is evident from the cross
section values of Table 19-1 that all of the neutrons absorbed by a fis-
620 NUCLEAR FISSION [CBAP, 19
TABLE 19-4
THE AVERAGE NUMBER OF FISSION NEUTRONS AS A
FUNCTION OF THE ENERGY OF THE NEUTRONS lNDUCING F1ss10N< 27 >
Neutron
energy, Mev
u233 +n u23s +n u2as +n Pu239 +n
Thermal 2.51 ± 0.02 2.44 ± 0.02 2.89 ± 0.03
0.08 2.58 ± 0.06 2.47 ± 0.03 3.05 ± 0.08
1.3 2.69 ± 0.05 2.61 ± 0.09 3.08 ± o.os
1.5 2.57 ± 0.12 2.65 ± 0.09
1.8 2.75 ± 0.06 2.72 ± 0.06 3.28 ± 0.06
2.0 2.80 ± 0.15
4.0 3.06 ± 0.12 3.01 ± 0.12 3.11 ± 0.10 3.43 ± o.u
14.1 3.86 ± 0.28 4.52 ± 0.32 4.13 ± 0.25 4.85 ± o.so
sionable material do not induce fission; sorne absorptions result in the

emission of 'Y-rays, i.e., radiative capture competes with fission. With
U 235, for example, the reaction U 235 (n, 'Y)U 236 also occurs, and U 236 is
an a-emitter with a half-life of 2.4 X 10 7 years. The ratio of the radiative
capture cross section to the fission cross section is usually denoted by a,
(19-3)
and the number of fission neutrons released per neutron absorbed in a
fissionable nuclide is given by
'I = v/(1 + a)· (19-4)
Values of a and 'I for thermal neutrons are listed in Table 19-1 for the
important fissionable materials.
The average number of neutrons emitted per spontaneous fission has
been determined for various heavy nuclides.<1 2> The rate (number of
fissions per gram per second) can be measured 129> as well as the total
number of neutrons. <ao> For U 238 , the rate is {6.90 ± 0.24) X 10-3 fis-
sions/gm/sec, or about 25 fissions/gm/hr, and v = 2.4 ± 0.2, close to
the value for fission induced by thermal neutrons; for Th 232, v = 2.6 ±
0.3, and for Cf 252 , v = 3.53 ± 0.15.
The neutrons emitted as a result of the fission process can be divided
into two classes, prompt neutrona and delayed neutrom. The prompt
neutrons, which make up about 99% of the total fissíon neutrons, are
emitted within an extremely short interval of time, possibly as low as
10-14 sec, of the fission process. It is thought that the compound nucleus
[U 238 ] first splits into two fragments, each of which has too many neu-
trons for stability, and has also the excess energy (6 Mev or more) needed
to expel a neutron. The excited, unstable nucleus consequently ejects
19-5) NEUTRON EMISSION IN FISSION 621
TABLE 19-5
33
PROPERTIES OF DELAYED NEUTRONS< >
Yield, % of total neutrons emitted in fission,

Half-life, Energy, Thermal and fast fission Fast fission
sec Mev
u2as u2aa Pu2ao u23s Th2a2
55 0.25 0.021 0.022 0.007 0.020 0.075

22 0.46 0.140 0.078 0.063 0.215 0.330
5.6 0.41 0.125 0.066 0.046 0.254 0.341
2.1 0.45 0.253 0.072 0.068 0.609 0.981
0.6 0.42 0.074 0.013 0.018 0.353 0.378
0.2 0.027 0.009 0.009 0.118 0.095
Total yield 0.64 0.26 0.21 1.57 2.20
one or more neutrons within a very short time after its formation; prompt
'Y-rays are apparently emitted at the same time. These ideas are con-
sistent with the resulta of experimenta on the angular correlation between
the direction of the neutrons and that of the fragmenta, <rn the neutrons
being emitted preferentially in the same direction as the fragments.
The delayed neutrons, which constitute about 0.64% of the total neu-
trons from the fission of U 235 , are emitted with gradually decreasing
intensity for several minutes after the actual fission process. Six well-
defined groups of delayed neutrons have been observed by studying the
rate of decay of the neutron intensity.< 32 , 33 > The rate of decay of each
group is exponential, just as for other forms of radioactive change, and
a specific half-life can be assigned to each group, as well as a mean life
and decay constant. From the intensity, the fraction fh which the group
constitutes of the total (prompt and delayed) fission neutrons can be deter-
mined. The properties of the prominent groups of delayed neutrons are
listed in Table 19-5. Three additional low-intensity delayed neutron
groups from uranium have been reported,< 34> with half-lives of 3, 12, and
125 min and yields, per fission, of 5.8 X 10-s, 5.6 X 10- 10 , and 2.9 X
10- 10 , respectively.
The same groups of delayed neutrons are found in the fission of different
heavy nuclides by slow or by fast neutrons, but the yields vary from
nuclide to nuclide. Of the three important fissionable materials, U 235
has the greatest yield, 0.64%, about three times that of Pu 2 39 • Although
the yield of delayed neutrons is less than one percent of the total number
of neutrons emitted, the delayed neutrons have a strong influence on the
time-dependent behavior of a chain-reacting system based on fission and
play an important part in the control of the system.
622 NUCLEAR FISSION (CHAP, 19
The mechanism for the emission of delayed neutrons is understood.

Since sorne of the fission products are rich in neutrons and are very un-
stable with respect to ¡3-emission, a product with Z protons and N neutrons
may have a ¡3-decay energy greater than the binding energy of the last
neutron in the daughter product with Z + 1 protons and N - 1 neutrons.
In the ¡3-decay, the product nucleus can be left either in the ground state
or in one of the many excited states. If the excitation energy of one of the
excited states of the Z + 1, N - 1 nucleus is greater than the binding
energy of the last neutron, de-excitation may occur by the emission of a
neutron, leaving a nucleus with Z + 1 protons and N - 2 neutrons. The
neutron emission will be delayed and will appear to have the half-life of
the ¡3-decaying nuclide (Z, N), i.e., that of the delayed neutron precursor.
Severa! of the delayed neutron groups have been related to the decay of
bromine and iodine fission products.<3 5 - 39> Thus, Br87 is the precursor
of the 55-sec delayed neutron emitter. Iodine-137, with a half-life of
24 sec, is associated with the 22-sec group. This group may consist of two
or more precursors with similar half-lives, and may also contain Br 88
which has a half-life of 16.3 sec. The 5.6-sec group includes 1 138 (T 112 =
6.3 sec) and possibly Br 89 (T 112 = 4.4 sec). The decay scheme of Br 87
is shown in Fig. 19-5. Sorne 70% of the ¡3-decays of Br 87 go to the 5.4
Mev excited level of Kr 87 ; this level lies about 0.3 Mev above the level
corresponding to the binding energy of the last (51st) neutron in Kr 87 •
o
{J-, <2.6 Mev
Kr86 +n (0.3 Mev)
-Y,-2 Mev
-Y,-3.4 Mev
FIG. 19-5. Decay scheme of Br87 and its products, showing delayed neutron
emission.
19-6] ENERGY DISTRIBUTION OF NEUTRONS EMITTED IN FISSION 623
About 3% of the Kr 87 nuclei in the 5.4 Mev leve} emita neutron in the
transition Kr 87 - Kr 86 + n + 0.3 Mev, while the rest decay by 'Y-emis-
sion to the ground state of Kr 87 • The delayed neutrons are delayed entirely
by the 55-sec half-life of their ,8-ray parent Br 87 • It should be noticed
that Kr 87 has 51 neutrons and the emission of the delayed neutron leaves
a closed shell of 50 neutrons. Similarly 53 1 137 decays by ,8-emission to
54Xe
137
which can emit a neutron, leaving 54Xe 136 , which has a closed
shell of 82 neutrons; Br 89 and 1 139 yield, on ,8-decay followed by n-emis-
sion, nuclei with 50 + 2 and 82 + 2 neutrons, respectively, which are
again especially stable neutron configurations.
19-6 The energy distribution of the neutrons emitted in fission. The

energies of the prompt neutrons emitted in fission vary from values less
than 0.05 Mev to more than 17 Mev. The determination of neutron
energies over so wide a range is not an easy problem, and severa! methods
have been used in order to cover the range. The neutrons from the ther-
mal fission of U 235 have been studied in great detail, and this case provides
a good example of the methods and results. <4 o> In the energy range 0.05
to 0.7 Mev, a cloud chamber was used containing hydrogen gas and water
vapor. <41 > A thin foil of U 235 was bombarded with a beam of thermal
neutrons from the thermal column of a nuclear reactor. The fission
neutrons were allowed to enter the cloud chamber, where they collided
with protons, and the range of the recoil protons was determined by meas-
uring the track lengths. The energy distribution of the recoil protons
was obtained from the range measurements, and the energy distribution
of the neutrons was obtained from that of the recoil protons. The results
70
e60
00
1:, 50
~
'e 40
...o,
] 30
::,
Z 20
10
OO J 00 200 300 400 500 600 700

Neutron energy (kev)
F10. 19-6. Energies of fission neutrons: the low-energy region (Bonner
et al.<rn).
624 NUCLEAR FISSION (CHAP. 19
10¡
106 '\N(E)
N(E) = 4.75Xl06sinh(2E)½e-E
t
x-xN(E) Calculated
&pe,im~•• - W•U
•N(E) Experimental - Hill
105
).
§w ~
~\¡
>-:
101
102
10
0 4 8 12
E (Mev}
\ 16 20
Fm. 19-7. Energies of fission neutrons: the spectrum from 0.4 Mev to 17 Mev
(Watt< 43>).
are shown in Fig. 19-6, where the relative number of neutrons in a given
energy interval is plotted aga.inst the energy a.nd compared with an empiri-
cal formula which will be given below.
The neutron energy spectrum between 0.4 Mev and 17 Mev was studied
by measuring, with coincidence counters, the range distribution of recoil
protons from hydrogenous materia1s.< 42 , 43 > The neutron intensity shows
a broad maximum near 0.75 Mev, and decreases nearly exponentially at
energies greater than 2 Mev. The energy distribution between 0.075
Mev and 17 Mev can be described by the empirical formula
N(E) = e-B sinh(2E) 112 , (19-5)
where N(E) representa the relative number of neutrons per unit energy
range as a function of the neutron energy. The experimental resulta
between 0.4 Mev and 17 Mev are shown in Fig. 19-7, a.nd agree with the
empirical formula. The solid curve of Fig. 19-6 representa the same em-
pirical formula, and the agreement is good a.lso for the low-energy neutrons.
The average value of the neutron energy is 2.0 ± 0.1 Mev.
The fission neutron spectra of U 236 and Pu 2 39, in the range from 0.5 to
8 Mev, ha.ve been studied with nuclear emulsion platea, <44 , 46> a.nd the
19-7] THE ENERGY RELEASE IN FISSION 625
results agree with those just discussed. The spectrum for the neutrons
from Pu 239 shows a maximum in the 0.6- to 0.8-Mev region and the em-
pirical formula of Eq. (19---5) fits the experimental results; the average
energy is again 2.0 Mev.
19-7 The energy release in fission. One of the most striking properties
of the fission process is the magnitude of the energy released per fission,
which is about 200 Mev as compared with several Mev for other nuclear
reactions. An estimate of the amount of energy released per fission can
be made by considering the binding energy curve, Fig. 9---11. The value
of the average binding energy per particle has a broad maximum of about
8.4 Mev in the range of mass numbers from 80 to 150, and it has been shown
that nearly all of the fission products have mass numbers in this range.
The average binding energy per particle is about 7.5 Mev in the neighbor-
hood of uranium. Hence, the average binding energy per particle is
about 0.9 Mev greater in the fission products than in the compound
nucleus [U 236 ], and the excess, 0.9 Mev, is liberated in the fission process.
The total amount of energy released per fission should be roughly equal
to the product of the number of particles (236) multiplied by the excess
binding energy per particle (0.9 Mev), or approximately 200 Mev.
The total energy release per fission can also be calculated from the
nuclear masses of [U 236] anda typical pair of fission products. It has been
shown that the fission products with the greatest yields have mass numbers
near A = 95 and A = 139. If 42 Mo 95 and 57La 139 are taken as a pair of
stable products at the ends of the chains for their respective masses, their
combined mass is 138.955 amu + 94.946 amu = 233.900 amu, as given by
the semiempirical mass formula. The corresponding mass number is 234,
and 2 neutrons are apparently released in this particular fission process.
When 2.018 amu are added for these two neutrons, the products of the
reaction have a total mass of 235.918 amu. The mass of U 235 from the
semiempirical mass formula is 235.124 amu, so that the mass of the com-
pound nucleus [U 236 ] is close to 235.124 + 1.009 = 236.133 amu. The
mass excess that is converted to energy is equal to 236.133 amu - 235.918
amu = 0.215 amu. Since one amu is equivalent to 931 Mev, the energy
liberated in the process is 931 X 0.215 = 198 Mev. Although there are
at least 30 different ways in which the nucleus can divide, the mass excess
is approximately the same for ali of these processes, and 200 Mev is a good
value for the average amount of energy released per fission, as calculated
in this way.
The predicted value of about 200 Mev can be compared with experi-
mental values. The total amount of energy released per fission is the
sum of the kinetic energy of the fission fragments, the kinetic energy of
the emitted neutrons, the kinetic energy of the prompt 'Y-rays, and the
TABLE 19-6
THE ENERGY RELEASE IN THE FISSION OF
U 235 BY THERMAL NEUTRONS
Kinetic energy of the fission fra.gments 167 Mev

Kinetic energy of fission neutrons 5
Prompt 'Y-rays 7
/3-decay energy 5
'Y-decay energy 5
Neutrino energy 11
Total fission energy 200 Mev
total energy of the decay processes in the fission decay chains. It was
shown in Section 19-4 that the average value of the total kinetic energy
of the fission fragments from the thermal fission of U 235 is 167 Mev; the
uncertainty in this number is about 5 Mev. The average value of the
kinetic energy carried off by the neutrons is egua! to the product of the
average number of neutrons emitted per fission and the average kinetic
energy of the neutrons; for the thermal fission of U 235 , this product is
2.5 X 2.0 Mev = 5 Mev. The kinetic energy of the prompt 'Y-rays is in
the neighborhood of 5 to 8 Mev. <45 • 47 > Finally, the average energy< 4 S>
of ali radiations (/3-rays, 'Y-rays, and neutrinos) of the fission products
is 21 ± 3 Mev; about half of this energy escapes in the form of
neutrinos, and the other half is divided, approximately equally, between
the /3-rays and the 'Y-rays. These results are summarized in Table 19-6;
the total fission energy determined in this way is 200 Mev, with an
uncertainty between 5 and 10 Mev, in good agreement with the calculated
values.
The energy release in fission has also been measured calorimetrically, <49 >
in a way which did not include the energy carried off by 'Y-rays, neutrons,
and neutrinos. The result obtained was 177 ± 5 Mev; if 26 Mev are
added (cf. Table 19-6) for the energies missed, the total is 203 ± 8 Mev,
in good agreement with the previous results. Another calorimetric meas-
urement yielded a value of the energy released in the fission of U 235
except for the energy carried off by neutrinos;< 5 o> the result was 190 ± 5
Mev, which also agrees very well with the results of Table 19-6 and with
the calculated values.
Results obtained with other fissile materials, such as Pu 239 and U 233 ,
are similar to those for U 235 .
19-8 The theory of the fission process. It is not surprising, in view

of the remarkable properties of the fission process, that a great deal of
19-8] THE THEORY OF THE FISSION PROCESS 627
o
0008
A 8 8 e e
o D
Frn. 19-8. Possible steps in the process of nuclear fission according to the
liquid-drop model.
effort has been put into theoretical studies of the process. The first
thorough treatment was that of Bohr and Wheeler<so who accounted
for many of the properties of fission on the basis of the liquid-drop model
of the nucleus. This model was used in the last section in calculating the
energy release in fission from the atomic masses of the fissile and product
nuclei as given by the semiempirical mass formula; that formula was
derived by considering that a charged nucleus is analogous to a liquid drop
(Section 17-6). Bohr and Wheeler showed that the liquid-drop model
could be used to describe the process of fission in sorne detail, and made
successful predictions about the existence of spontaneous fission, and about
the ability of various heavy nuclei to undergo fission with slow or fast
neutrons.
In the spherical, liquid-drop nucleus, the shape of the drop depends on
a balance involving the surface tension forces and the Coulomb repulsive
forces. If energy is added to the drop, as in the form of the excitation
energy resulting from the capture of a slow neutron, oscillations are set up
within the drop; these tend to distort the spherical shape so that the drop
may become ellipsoidal in shape. The surface tension forces tend to make
the drop return to its original shape, while the excitation energy tends to
distort the shape still further. If the excitation energy is sufficiently
large, the drop may attain the shape of a dumbbell. The Coulomb repul-
sive forces may then push the two "bells" apart until the dumbbell splits
into two similar drops, each of which then becomes spherical in shape.
The sequence of steps leading to fission is shown in Fig. 19-8. If the
excitation energy is not large enough, the ellipsoid may return to the
spherical shape, with the excitation energy being liberated in the form of
'Y-rays, and the process is one of radiative capture rather than fission.
The potential energy of the drop in the different stages of Fig. 19-8 can
be calculated under the Bohr-Wheeler theory, as a function of the degree
of deformation of the drop< 51 , 52> The form of the results is shown in
Fig. 19-9, where the potential energy E is plotted against a parameter r
which is a measure of the degree of deformation. At r = O, which repre-
La148
1
m ...::.~:1
---
Ec .,,' ",-"'¡
t II
.,,
/ :1
1 E '
1
7.9 X 6.7 X 1
1
10~13 10~13 1
cm cm 1
1- ..¡ 1
1
1.46 X 10-12 cm
57e X 35e
r = R1 + R2
Epot = l. 46 X 10 _12 = 197 Mev Separa tion (r)
Fro. 19-9. Potential energy of two fission fragments as a function of distance

between centers and the configuration at contact, where the potential energy is
197 Mev.
sents the initial spherical nuclear drop, there is available an amount of

energy E 0 , given by
(19-6)
where (A, Z) represents the nucleus that may undergo fission, and the
subscripts 1 and 2 stand for the possible final products. The value of E 0
corresponds to the ground state of the compound nucleus formed when
the target nucleus captures a neutron, but does not include the excitation
energy resulting from the neutron capture. lt was shown in the last
section that the value of E 0 for U 236 is about 200 Mev.
The meaning of the graph can now be seen more easily if the process
which would be the reverse of fission is considered. When the drop has
split into two fragments, r is the distance between their centers; if R 1 and
R 2 are the radii of the two product drops, r = R 1 + R 2 is the value of the
deformation parameter for which the two drops just touch. For values
of r smaller than R 1 +
R 2 , r represents the degree of deviation of the
original drop from its spherical shape. For values of r greater than
R 1 + R 2 , the energy is just the electrostatic energy resulting from the
mutual repulsion of the two positively charged nuclear fragments. The
value of E in this region is (Z 1Z 2 e2 /r), and E increases as the distance
between the two fragments is decreased; the energy at r = oo is taken
equal to zero. When the two fragments just touch, at r = R 1 + R 2 , the
Coulomb energy is denoted by Ec and is
2
E _ Z1Z2e
(19-7)
e - R1 + R2
When r is smaller than R 1 +
R 2 , the energy depends not only on the
electrostatic forces, but also on the surface tension forces; the calculation
19-8) THE THEORY OF THE FISSION PROCESS 629
of E is then complicated, and three different curves (I, II, and III) are
shown for the possible variation of E in this region. The shapes of these
curves and the values of E 0 are related to the mass of the nucleus, that is,
to the value of A in Eq. (19--6). Stable nuclei with values of A some-
what greater than 100 are of type I, with E 0 about 50 Mev smaller than
Ec. Nuclei like those of uranium, plutonium, or thorium are of type II,
for which Ec - E 0 is about 6 Mev. For still heavier nuclei, E 0 may be
greater than Ec, as in case III. Nuclei of type III should undergo fission
spontaneously and would not be expected to last long.
On a classical basis, nuclei of type II should be stable with respect to
fission, but according to quantum mechanics, there is a certain probability
that they will undergo spontaneous fission. The fragments may "leak
through" the barrier represented by Ec in the same way that a-particles
do in a-decay. In U 238 , about 25 spontaneous fissions occur per gram
per hour, and the half-life for this process is about 10 17 years.
The activation energy Ec - E 0 needed to induce fission in nuclei of
type II has been calculated from the Bohr-Wheeler theory, and can be
compared with the excitation energy resulting from the capture of a
neutron, or other particle, ora 'Y-ray. The excitation energy of a compound
nucleus can be computed with the aid of the semiempirical binding energy
(or mass) formula, or from the masses of the nuclei involved if they are
known. When the compound nucleus is formed by the capture of a slow
neutron, the excitation energy differs by a negligible amount from the
binding energy of the last neutron. In the case of [U 236 ], the binding
energy is calculated from the known masses as follows:
Mass of U 235, 235.11704

Mass of neutron, 1.00898
236.12602
Mass of U 236 236.11912
t:.M = 0.00690 amu = 6.4 Mev.
The values of the excitation energy calculated in this way for a number
of heavy nuclei are listed in Table 19-7 and compared with the correspond-
ing values of the activation energy obtained from the Bohr-Wheeler
theory. In U 233 , U 235, and Pu 239 the excitation energy is considerably
greater than the activation energy, and these nuclei would be eKpected
to undergo fission with thermal neutrons; they do, and the thermal fission
cross sections of U 235 and Pu 239 are large, 577 b and 742 b, respectively.
In U 238 and Th 232, the excitation energy is smaller than the activation
energy and these nuclei do not undergo fission with thermal neutrons, as
predicted. To induce fission, the neutrons must have considerable kinetic
TABLE 19-7
FISSIONABILITY OF HEAVY NucLEI WITH THERMAL NEUTRONS
Target Compound Excitation Activation

nucleus nucleus energy, Mev energy, Mev
u2aa ¡u2a4¡ 6.6 4.6
u2as ¡u2a6¡ 6.4 5.3
u2as ¡u2a9¡ 4.9 5.5
Th2a2 [Th2aa¡ 5.1 6.5
pa,2a1 [Pa.232] 5.4 5.0
Np2a1 [Np2as¡ 5.0 4.2
Pu239 [Pu240] 6.4 4.0
energy, and fission should be a threshold reaction in these nuclei, as it is.

In Pa 231 and Np 237 , the excitation energy is greater than the activation
energy, and these nuclides should undergo fission with thermal neutrons.
Early experiments indicated that they do not; recent measurements show
that there is sorne thermal fission, but the cross sections are very small
(about 10-2 b) compared with those of U 235 and Pu 239 . The fact that the
calculated values of the excitation energy are greater than those of the
activation energy for Pa 231 and Np 237 <loes not agree with the very low
probabilities of fission by thermal neutrons.
The values of the excitation energy listed in Table 19-7 indica.te that
the binding energy of the neutron added to a nucleus with an odd number
of neutrons (U 233 , U 235 , Pu 239 ) is greater than that for a nucleus with an
even number of neutrons (Th 232 , U 238 , Pa 231 , Np 23 7). It is found generally
that the excitation energy is greater for target nuclei with odd numbers
of neutrons than for nuclei with even numbers of neutrons, and this effect
is related to the presence of the "odd-even" term in the binding energy
formula. Fission with thermal neutrons occurs much more often in
nuclei with an odd number of neutrons than in nuclei with an even number
of neutrons, as is shown by the values of the fission cross section listed in
Tables 19-1 and 19-2. Nearly all of the target nuclei with large values
of the cross section for fission by thermal neutrons have odd numbers of
neutrons, while nearly all of the nuclei with very small cross sections have
even numbers of neutrons. There is, therefore, sorne correlation between
the values of the excitation energy and the occurrence of thermal fission,
in qualitative agreement with the predictions of the Bohr-Wheeler theory.
There are sorne obvious exceptions; 96Cm 240 has a very large cross section
(~20,000 b), and 92 U 230 and 92 U 232 have appreciable cross sections (25 b
and 80 b, respectively); each of these nuclei has an even number of neu-
trons. The nuclides Pa 231 and Np 237 have low excitation energies, corre-
REFERENCES 631
sponding to their even numbers of neutrons, but their activation energies

are still lower, so that they should undergo fission with slow neutrons in
spite of their odd neutron numbers. The theory based on the Jiquid-
drop model gives incorrect values of the activation energy in these cases.
The above discussion indicates that the liquid-drop modcl of the nucleus
has been applied with sorne success to the problem of nuclear fission.
The theory gives a qualitative, and sometimes quantitative, picture of
the process and has made possible the description of sorne of the properties
of fission, but it also has sorne serious defects in addition to those already
mentioned. According to the theory, the most probable mode of division
of the liquid-drop nucleus is into two equal fragments, so that it fails to
account for the observed asymmetry of the process; it also yields values
for photofission thresholds and for spontaneous fission rates which are
wrong quantitatively, and whose variations from nuclide to nuclide do
not agree with experiment.
Attempts have been made to modify the liquid-drop theory of fission
in order to overcome these difficulties. The application of the shell model
has been suggested, <53 > and has had sorne success in accounting for the
asymmetry. <54 > Another promising approach seems to be that of Hill
and Wheeler, <55 > which is based on the new collective model of the nucleus,
and the reader is referred to their article for a complete and detailed dis-
cussion of the mechanism of fission. Many additional papers on the theory
of fission can be found in the general references listed below.
REFERENCES
GENERAL
L. A. TuRNER, "Nuclear Fission," Revs. Mod. Phys. 12, 1 (1940).
W. J. WHITEHOUSE, "Fission," Progress in Nuclear Physics, Vol. 2, O. R.
Frisch, ed. New York: Academic Press, 1952.
Radiochemical Studies: The Fission Products, National Nuclear Energy Ser.,
Div. IV, 9, C. D. Coryell and N. Sugarman, eds. New York: McGraw-Hill,
1951.
Physics of Nuclear Fission, Supplement No. 1 of the Soviet Journal, Atomnaya
Energiya. London: Pergamon Press, 1958.
l. HALPERN, "Nuclear Fission," Ann. Revs. Nuc. Sci. 9, 245 (1959).
Proceedings of the United Nations International Conference on the Peaceful
Uses of Atomic Energy, Geneva, 1955, Vol. 2: Physics: Research Reactors. New
York: United Nations, 1956.
Proceedings of the Second United Nations International Conference' on the
Peaceful Uses of Atomíc Energy, Geneva, 1958, Vol. 15: Physics in Nuclear
Energy. Geneva: United Nations, 1958.
Proceedings of the Symposium on the Phyaics of Fi8Bion, held at Chalk River,
Ontario, May, 1956, Atomic Energy of Ganada, Ltd, Report AECL-329, Chalk
River, Ontario, July, 1956.
PARTICULAR
l. O. HAHN and F. STRASSMAN, "Uber den Nachweis und das Verhalten der
bei der Bestrahlung des Urans mittels Neutronen entstehenden Erdalkali-
metalle," Naturwús 27, 11, 89 (1939).
2. O. HAHN, New Atoms. New York: Elsevier (1950).
3. O. R. FRISCH, "Physical Evidence for the Division of Heavy Nuclei under
Neutron Bombardment," Nature 143, 276 (1939).
4. ANDERSON, BooTH, DuNNING, FERMI, GLASOE, and SLAcK, "The Fissíon
of Uranium," Phys. Rev. SS, 511 (1939).
5. L. MEITNER and O. R. FR1scn, "Disintegration of Uranium by Neutrons:
a New Type of Nuclear Reaction," Nature 143, 239 (1939).
6. N1ER, BooTH, DuNNING, and GRossE, "Nuclear Fission of Separated
Uranium Isotopes," Phys. Rev. 57, 546, 748 (1940).
1: (a) D. J. HuoHES and R. B. ScHWARTZ, Neutron Cross Section8, BNL-St5,
2nd ed. Washington, D.C.: U.S. Government Printing Office, July, 1958.
(b) D. J. HuoHES, B. A. MAGURNO, and M. K. BRUSSEL, BNL-3!5, 2nd ed.
Supplement I, Jan., 1960.
8. D. J. HuonEs, "New World-Average Thermal Cross Sections," Nucleonics,
17, No. 11, 132 (Nov., 1959).
9. R. W. LAMPHERE, "Fission Cross Sections of the Uranium Isotopes,
233, 234, 236, and 238, for Fast Neutrons," Phys. Rev. 104, 1654 (1956).
10. J. JuNGERMAN, "Fission Excitation Functions for Charged Particles,"
Phys. Rev. 79, 632 (1950).
11. KocH, McELHIN_NEY, and GASTEIGER, "Experimental Photo-Fission
Thresholds in U235 , U238 , U233 , Pu239 , and Th 232," Phys. Rev. 77, 329 (1950).
12. K. A. PETRZHAK, "Spontaneous Fission oí Heavy Nuclei," Physic8 of
Nuclear Fission, op. cit. gen. ref., pp. 129-152.
13. "Plutonium Project, Nuclei Formed in Fissíon: Decay Characteristics,
Fission Yields, and Chain Relationships," Revs. M od. Phys. 18, 513 (1946);
J. Am. Chem. Soc. 68, 2411 (1946).
14. Radiochemical Stúdies: The Fission Products, op. cit. gen. ref., Book 3,
Appendixes A, B, C.
15. S. KATCOFF, "Fission-Product Yields from U, Th, and Pu," Nucleonics
16, No. 4, 78 (April 1958).
16. A. N. MuRIN, "Charge and Mass Distributions oí Fission Products,"
Physics of Nuclear Fission, op. cit. gen. ref., pp. 26-44.
17. D. C. BRuNTON and G. C. HANNA, "Energy Distribution of Fission
Fragments from U 235 and U 233 ," Can. J. Research A28, 190 (1950).
18. D. C. BRUNTON and W. B. TnoMPsoN, "Energy Distribution of Fission
Fragments from Pu 239," Can. J. Research A28, 498 (1950).
19. S. S. FRIEDLAND, "Energy Distribution of Fission Fragmenta of U 235
Produced by 2.5 Mev and 14 Mev Neutrons," Phys. Rev. 84, 75 (1951).
20. J. S. WAHL, "Energy Distribution of Fragments from the Fission of U 235 ,
U 238 , and Pu 239 by Fast Neutrons," Phys. Rev. 95, 126 (1954).
21. J. JuNGERMAN and S. C. WRIGHT, "Kinetic Energy Release in Fission of
u2 as, u2as, Th232, and Bi2 º9 by High Energy Neutrons," Phys. Rev. 76, 1112
(1949).
REFERENCES 633
22. R. B. LEACHMAN, "Velocities of Fragmente from Fission of U 233 , U 235 ,

and Pu 239," Phys. Rev. 87, 444 (1952).
23. R. B. LEACHMAN and W. D. ScHAFER, "A Calorimetric Determination
of the Average Kinetic Energy of the Fragmenta from U 235 Fission," Can. J.
Phys. 33, 357 (1955).
24. W. J. WHITEHOUSE and W. GALBRAITH, "Energy Spectrum of Fragmente
from the Spontaneous Fission of Natural Uranium," Phil. Mag. 41 1 429 (1950).
25. L. E. GLENDENIN, "The Distribution of Nuclear Charge in Fission,"
Technical Report No. 35, Laboratory for Nuclear Science and Engineering,
M.I.T., Cambridge, 1949.
26. V0N HALBAN, JotIOT, and KovARSKI, "Liberation of Neutrons in the
Nuclear Fission of Uranium," Nature 143, 470, 680 (1939).
27. R. B. LEACHMAN, "The Fission Process-Mechanisms and Data," Paper
No. 2467, Proceedings oJ the Second United Nations lnternational Conference on
the Peaceful Uses of Atomic Energy, op. cit. gen. ref., Vol. 15, 229 (1958).
28. BoNDARENKO, KusMINOV, KUTSAYEVA, PROKHOR0VA, and SMIRENKIN,
"Average Number and Spectrum of Prompt Neutrons in Fast-Neutron-Induced
Fission," Paper No. 2187, Proceedings of the Second United Nations lnternational
Conference on the Peaceful Uses of Atomic Energy, op. cit. gen. ref., Vol. 15, 353
(1958).
29. E. SEGRE, "Spontaneous Fission," Phys. Rev. 86, 21 (1952).
30. D. J. L1TTLER, "A Determination of the Rate of Emission of Spontaneous
Fission Neutrons by Natural Uranium," Proc. Phys. Soc. (London) 65A, 203
(1952).
31. R. B. LEACHMAN, "Neutrons and Radiations from Fission," Paper No.
665, Proceedíngs of the Second United Nations lnternational Conference on the
Peaceful Uses of Atomic Energy, op. cit. gen. ref., Vol. 15, 331 (1958).
32. HuGHES, DABBS, CAHN, and HALL, "Delayed Neutrons from Fission of
U 235," Phys. Rev. 73, 111 (1948).
33. KEEPIN, W1111ETT, and ZIEGLER, "Delayed Neutrons from Fissionable
Isotopes of Uranium, Plutonium, and Thorium," Phys. Rev. 107, 1044 (1957);
J. Nuclear Energy 61 1 (1957).
34. KuNSTADTER, FLOYD, and BoRST, "Long-lived Delayed Neutrons from
Fission," Phys. Rev. 91, 594 (1953).
35. SNELL, NEDZEL, lBSER, LEVINGER, WILKINSON, and SAMPSON, "Studies
of the Delayed Neutrons. l. The Decay Curve and Intensity of the Delayed
36. SNELL, LEVINGER, MEINERB, SAMPSON, and WILKINSON, "Studies of the
Delayed Neutrons. II. Chemical Isolation of the 56-Second and 23-Second
Activities," Phys. Rev. 72, 545 (1947).
37. N. SuoARMAN, "Short-Lived Halogen Fission Products," J. Chem. Phys.
17, 11 (1949).
38. (a) G. J. PERLOW and A. F. STEHNEY, "Halogen Delayed-Neutron Activi-
ties," Phys. Rev. 113 1 1269 (1959).
(b) A. F. STEHNEY and G. J. PERLow, "Halogen Delayed-Neutron Activi-
ties," Paper No. 691, Second International Conference on the Peaceful Uses of
Atomic Energy, op. cit. gen. ref., Vol. 15, 384 (1958).
39. A. C. PAPPAS, "The Delayed Neutron Precursors in Fll!Sion," Paper No.

583, Second lnternational Conference on the Peaceful Usea of Atomic Energy,
op. cit. gen. ref., Vol. 15, 373 (1958).
40. B. G. EROZOLIMSKI, "Fission Neutrons" in Physics of Nuclear FisBion,
op. cit. gen. ref., p. 64.
41. BoNNER, FERRELL, and RINEHART, "A Study of the Spectrum of the
Neutrons of Low Energy from the Fission of U 23 5," Phys. Rev. 87, 1032 (1952).
42. D. L. HILL, "The Neutron Energy Spectrum from U 235 Thermal Fission,"
Phys. Rev. 87, 1034 (1952).
43. B. E. WATT, "Energy Spectrum of Neutrons from Thermal Fission of
U2 35 ," Phys. Rev. 87, 1037 (1952).
44. N. NERESON, "Fission Neutron Spectrum of U 235," Phys. Rev. 85, 600
(1952).
45. N. NERESON, "Fission Neutron Spectrum of Pu 239 ," Phys. Rev. 88, 823
(1952).
46. KINSEY, HANNA, and VAN PATTER, "Gamma Rays Produced in the
Fission of U 235," Can. J. Research 26A, 79 (1948).
47. MAIENSCHEIN, PEELLE, ZOBEL, and LovE, "Gamma Rays Associated with
Fission," Paper No. 670, Second lnternational Conference on the Peaceful Applica-
tions of Atomic Energy, op. cit. gen. ref., Vol. 15, 366 (1958).
48. K. WAY and E. P. WIGNER, "The Rate of Deca.y of Fission Products,"
Phys. Rev. 73, 1318 (1948).
49. M. C. HENDERSON, "The Hea.t of Fission of Uranium," Phys. Rev. 58,
744 (1940).
50. J. L. MEEM, "Energy Relea.se per Fission in the Bulk Shielding Reactor,"
Nucleonics 12, No. 5, 62 (Me.y 1954).
51. N. BoHR and J. A. WHEELER, "The Mechanism of Nuclear Fission,"
Phys. Rev. 56, 426 (1939).
52. S. FRANKEL and N. METROPOLIS, "Calculations in the Liquid-Drop Model
of the Nucleus," Phys. Rev. 72, 914 (1947).
53. L. MEITNER, "Fission a.nd Nuclear Shell Model," Nature 165, 561 (1950).
54. P. FoNG, "Sta.tistical Theory of Nuclear Fission: Asymmetric Fission,"
Phys. Rev. 102, 434 (1956); also, Phys. Rev. 89, 332 (1953).
55. D. L. HILL and J. A. WHEELER, "Nuclear Constitution and the Interpreta.-
tion of Fission Phenomena," Phys. Rev. 89, 1102 (1953).
635
PROBLEMS
1. The nuclides listed below are formed in fission. Indicate the fission product
chain, if any, to which each nuclide gives rise: Zn72, Kr9º, Kr 92, Kr97, Mo 105,
Snl26, Xel44, Pml56.
2. Two of the stable end products of fission chains resulting from the thermal
fission of U 235 are Sr91 and La 139 . What weight of each of these nuclides is
formed by the complete fission of 1 gm of U 235 ?
3. Two possible sets of products of the fission of Pu 239 by thermal neutrons
are (a) Ce 14 1, Mo 96 and 3 neutrons, (b) Pd 1º8, Xe 129 and 3 neutrons, (e) Gd 155,
Br8 1 and 4 neutrons. Compute the energy release in each case.
4. In a calorimetric measurement of the energy released in the fission of
uranium, it was found that a 13.36 gm sample of uranium produced heat at
the rate of 40 microwatts when bombarded with slow neutrons. A thin !ayer
of uranium weighing 54 µgm, also in the calorimeter and subjected to the same
neutron flux, was found to undergo fission at the rate of 340 processes/min.
The fissions were counted with an ionization chamber in the calorimeter. Calcu-
late the energy release in Mev/fission from these data.
5. Plot, against time, the total intensity of the delayed neutrons from the
fission of U 235 in a sample of uranium which has been irradiated for a time
interval long compa.red with the period of the longest-lived delayed neutron
emitter. Continue the plot until the intensity rea.ches 0.1 % of the initia.l activity.
Use a semilogarithmic pa.per.
6. It has been shown, on the basis of the liquid-drop model, that the limit of
stability of nuclei against spontaneous fission is given by the quantity
2
Z 10 Ro
A= aª•e2'
where a, is the coefficient of the surface energy term in the semiempirica.l binding
energy formula., Ro is the constant coefficient in the A 113 formula. for the nuclear
radius, and e is the electronic charge.
(a) Compute the limiting va.lue of Z 2/ A.
(b) Calculate the value of Z2/ A for Ra 226 Th 232 u 2a2 U 23 8 Pu238 Am241
and Cu 242 . ' ' ' ' ' '
(e) Which of these nuclides might be expected to have significant rates of
spontaneous fission? The observed rates, in fissions per gram per second, for
the nuclides in (b) are «0.6, 4.2 X 10-5, 16, 6.9 X 10-3, 2.1 X 103, 40, and
7.6 X 106, respectively.
7. Compute the binding energy of a thermal neutron to each of the following
nuclei· Th221 u229 Np2as Np2a1 Np2as Pu241 Pu242 Am241 Am242 a.nd
Cm 240 . Whi;h of these nucÍides w~uld be :xpected to hav~ rela.tiv;ly large'cross
sections for fission by thermal neutrons? Compare the predictions with the
va.lues listed in Table 19-2.
[Hint: For Problems 3 a.nd 7, the necessary masses can be obtained from
K. T. Ilainbridge, "Cha.rged Particle Dynamics and Optics, Relative Isotopic
Abundances of the Elements, Atomic Masses," Experimental Nuclear Physics,
Vol. I, E. Segre, ed. New York: Wiley, 1953.]
8. How would you account for the difference in the kinetic energy of the
neutrons that can cause fission of U235 and U238, respectively? [Hint: Use the
resulta of Problem 11, Chapter 17.J
9. Suppose that an even-even compound nucleus (Z, A) such as U 236 under-
goes symmetric fission. (a) Show that the prompt energy relea.se is given, to
a good approximation by
z2
E1 (Mev) = 0.22 A 113 - 3.42A
2,a
(b) Show that the condition for spontaneous symmetric fission to be possible
energetically is Z2/ A ~ 15.
10. Calculate the contributions of the Coulomb and surface energies, re-
spectively, to the energy release when the compound nucleus [U 236] formed by
the capture of a slow neutron by U 235 undergoes symmetric fission.
11. Show that the height of the Coulomb barrier for symmetric fission is
given by
z2
E. (Mev) = 0.15 Alta·
12. Show that the condition for stability with respect to spontaneous sym-
metric fission is Z 2/ A ::; 49. [Hínt: Note that the condition is given by
E& - E1 ~ O.
13. Bohr and Wheeler<51 > showed that, for a small deformation of a liquid
drop, the net change in the potential energy A.E is given by
A.E= A (1 - 0.022~)-
For A.E > O, the nucleus would return to its original shape; for A.E < O, the
nucleus would be unstable. Show that the condition that a nucleus be just
unstable with respect to a small deformation is Z 2/ A ~ 45. Note that
!2a.
ª• >
-
45
'
where a, and ª• are the coefficients which appear in the surface and Coulomb
energy terms of the semiempirical binding energy formula.
14. Bohr and Wheeler< 51 > also showed that the activation energy E,. needed
for nuclear drop to undergo fission is
213
E,.= A X 0.89(1 - 0.022~)·
Use this formula, and also the difference E& - E¡ to calculate the activation
energies for the nuclei listed in Table 19-7.
CHAPTER 20
NUCLEAR ENERGY SOURCES
2~ 1 Nuclear fission as a source of energy. The large amount of energy

released in fission, together with the emission of more than one neutron,
has made it possible to use the fission process as a source oí energy. The
emission, on the average, of 2.5 neutrons in the fission of a U 235 nucleus
permits a chain reaction in which these neutrons produce more fissions
and more neutrons, and so on. Under sorne conditions, the numbers of
fissions and neutrons increase exponentially with time because each fission
produces more neutrons than the one absorbed, and the amount of energy
released can become enormous. The time interval between successive
generations of fissions can be a very small íraction of a second and the
energy released in the chain reaction can take the forro of an explosion;
the result is an "atomic bomb." Under other conditions the chain reaction
can be controlled, anda steady state can be attained in which justas many
neutrons are produced per unit time as are used up. The rate at which
fissions occur and energy is released is kept constant, and the result is a
chain-reacting pile, or nuclear reactor, which can be used as a source of
neutrons or of power.
These ideas can be illustrated by sorne simple calculations. The fission
of one U 235 nucleus liberates 200 X 1.6 X 10-6 erg = 3.2 X 10- 4 erg.
When this quantity is multiplied by the Avogadro number, the product
gives the energy released in the fission of all the nuclei in 1 gram atom
(235 gm oí U 235 ), and is equal to 1.93 X 10 2 º ergs. The energy liberated
by the complete fission of 1 kg (2.2 lb) of U 235 would be 8.21 X 10 20 ergs,
or about 2 X 10 10 kilocalories, which is roughly equivalent to the energy
released in the explosion oí 20,000 tons of TNT. Even if only a small
fraction of this energy could be released explosively, the result would be
a formidable bomb.
The energy release in fission can also be expressed in terms of power
units with interesting results. Since 1 Mev = 1.60 X 10-6 erg = 1.6 X
10- 13 watt-sec, the fission oí one U 235 nucleus frees 3.2 X 10- 11 watt-
sec of energy, and 3.1 X 10 10 fissions/sec give one watt of power. The
complete fission of 1 gm of U 235 would release 8.2 X 10 10 watt-sec of
energy, or 2.3 X 104 kilowatt-hours, or nearly 1 megawatt-day. If the
energy release were spread out over the period of a day, the complete
fission of one kilogram of U 235 would produce energy in the forro of heat
637
638 NUCLEAR ENERGY SOURCES [CHAP. 20
at a rate of 1000 megawatts. If this heat could be changed to electricity

at a conversion efficiency of 30%, electrical energy would be supplied at
the rate of 300,000 kilowatts. This output is equivalent to that from a
large power plant which consumes about 2500 tons of coal per day. The
fact that 1 kg of a fissionable material such as U 235 is equivalent to 2500
tons of coal as a source of electric power is responsible for research and
development work on the use of nuclear chain reactions and for the con-
struction of nuclear powerplants.
Although useful nuclear power is a worth-while objective, another reason
why nuclear reactors have been built has to do with the military applica-
tions of the fission process. It was realized in 1940 that an atomic bomb
would have to depend on fission by fast neutrons, because only in this case
would the energy release be rapid enough to make a bomb effective.
But experiments indicated that the cross section for the fission of U 238
by fast neutrons is sma.11 compared with the total cross section, with the
result that an explosive fast neutron chain reaction with normal uranium
seemed improbable. The relatively rare isotope U 235 appeared to have
satisfactory nuclear properties for an atomic bomb, but the separation of
U 235 in amounts large enough to be useful for military purposes presented
a very difficult problem. The transuranium nuclide Pu 239, however,
offered an alternative solution. There was evidence, in 1940, which indi-
cated that the radiative capture o( neutrons by U 238 would probably lead
by two successive ,B-decays to the formation of a nucleus for which Z = 94
and A = 239. It was predicted from the Bohr-Wheeler theory of fission
that this nuclide would be a fairly long-lived a-emitter which would prob-
ably undergo fission when bombarded by slow or fast neutrons, as does
U 235 • If a method could be devised for converting sorne of the U 238 to
Pu 239 , a chemical separation of the plutonium from uranium would avoid
the difficulties of the isotopic separation of U 235 from U 238 • It was then
suggested that Pu 239 could be made in sufficiently large amounts in a
controlled chain reaction based on the fission by slow neutrons of the
U 235 in natural uranium.
It is now well known that both the isotopic separation of U 235 and the
manufacture of Pu 239 in a chain-reacting pile have been accomplished on
a large scale. Many atomic bombs have been exploded but, for various
reasons, this application of nuclear physics will not be discussed further
in this book. A considerable number of nuclear reactors has been built
for both military and peaceful purposes. Their designa depend on the
nuclear properties of materials and involve the application of the methods
and ideas of nuclear physics. These applications often serve to sharpen
the distinctions among different nuclear properties and among different
reactions and make use of the available knowledge about heavy atomic
nuclei in surprising and fascinating ways.
2~2] CHAIN-REACTING SYSTEM OR NUCLEAR REACTOR 639
20-2 The chain-reacting system or nuclear reactor. The achievement

of a chain reaction with uranium depends on a favorable balance among
four competing processes:
l. fission of uranium nuclei, with the emission of more neutrons than
are captured,
2. nonfission capture of neutrons by uranium,
3. nonfission capture of neutrons by other materials,
4. escape of neutrons without being captured.
If the loss of neutrons by the last three processes is less than or equal to the
surplus produced by the first, the chain reaction occurs; otherwise it does
not.
The need for a favorable neutron balance sets certain conditions on
any system in which a chain reaction is sought; one condition "is on the
size. If the uranium is distributed in a regular way throughout the as-
sembly, the neutron production depends on the volume of the system,
whilc the probability of escape depends on the surface area. lf the system
is very small and the surface-to-volume ratio is large, most of the neutrons
can escape, and this process alone may make it impossible to achieve a
chain reaction. The loss of neutrons by nonfission capture is a volume
effect like the production, and its relative importance does not change
with the size of the system. The result is that the greater the size of the
assembly, the less likely it is that the escape of neutrons and their loss by
absorption will outweigh their production and prevent a chain reaction.
There is a certain size, called the critical size, for which the production of
neutrons by fission is just equal to their loss by nonfission capture and
escape, anda chain reaction is possible. If the size of the system is smaller
than the critical size, a chain reaction cannot be sustained. The existence
of a critica! size below which a chain reaction will not "go" contrasts
sharply with systems based on chemical reactions in which the possibility
of a reaction is independent of the size of the system.
The relative probabilities of fission, radiative capture, and scattering
are expressed and discussed in terms of cross sections. Values of the cross
sections of U 235 for thermal neutrons are listed in Table 19-1; the varia-
tion with energy of the fission cross section of U 235 in the epithermal or
"intermediate" energy range is shown in Fig. 19-l(b). Cross seetiolis of
natural uranium for thermal neutrons are listed in Table 19-1. The varia-
tion of the total cross section of U 238 in the epithermal energy range is
shown in Fig. 20-1; there are many resonances with strong absorption,
that is, radiative capture. Average values of the cross sections of U 238
for neutrons with energies above the fission threshold (taken for practica!
purposes to be 1.4 Mev) are listed in Table 20-1. Sorne of the basic re-
quirements for the achievement of chain reactions can be deduced from
this information about cross sections.
~
I➔ nlJ2"
(ev}
Total
(n, 'Y)
z
d
(')
t'
l:!':I
>
l:l;1
o-,(b) ",, 1
,; 1*':t!,"".:<
".,
~t'.!.!i
! "
1.~:,r;,i'
.--:...
"
' ~:"-;, i!,. ""'· _,¡,¡,
¿
. "'"t'~
.
o(;¡..
,~ • >..,.
•t. ~~ '\ .¡t ~ ,,:1~ 1'~ ~f ""1'
fil .t¡ ,. ' ,
'· ,>t..
,l¡ ~
o-, (b)
l:!':I
z
l:!':I
tll
Q
wj
1))
o
d
::,,
(')
l:!':I
rJJ
'"i'i'
::i:
~
1:-.?
o
Fm. 20-1. The total cross section of U238 as a function of neutron energy, showing the
absorption resonances (BNL-325, 2nd ed,).
20-2) CHAIN-REACTING SYSTEM OR NUCLEAR REACTOR 641
TABLE 20-1
AVERAGE VALUES OF Caoss SECTIONS OF U 238
FOR NEUTRONS WITH ENERGIES ÜREATER
THAN THE FISSION THRESHOLD
Fission Ü¡ = 0.50b
Radiative capture ¡¡, = 0.05b
Elastic scattering a-. = 4.55b
Inelastic scattering qi = 2.10b
Total ¡¡ = 7.2 b
Average number of
neutrons per fission ji = 2.7 b
We consider first the question of the possibility of a chain reaction in a.

mass of natural uranium large enough so that the loss of neutrons by
leakage may be neglected. Some spontaneous fissions will occur, releasing
about 2.5 fast neutrons per fission, with a.n average energy of about 2 Mev.
Now, for practica! purposes, natural ura.nium consists of U 238 with an
abundance of 99.28 atom percent and U 236 with an abundance of 0.72
atom percent; the fission cross sections of the two isotopes in the energy
range of the fission neutrons are not very different. Hence, if the fission
neutrons cause further fissions, these will occur chiefly in U 238 , with a.
negligibly small amount of fast fission in U 235 • lt is only necessary, there-
fore, to consider the interaction of the initial fission neutrons with U 238•
For these neutrons, fission is more probable than radiative capture but less
probable than either elastic or inelastic scattering. Neutrons which undergo
radiative capture are lost. Elastic scatteríng by a uranium nucleus has
very little effect on the energy of a fission neutron and, after such a col-
lision, the neutron is free to make another collision almost as though noth-
ing had happened. Inelastic scattering, however, has a serious effect in
that a single process of this kind can reduce the neutron energy to a value
below the fission threshold, to about 0.3 Mev on the average. Conse-
quently, the original fission neutrons are only slightly increased in number
by fast fission in U 238 , and nearly all of the neutrons are left with energies
below the fission threshold. Inelastic scattering is also a threshold reaction,
and doesn't occur for neutrons with energies below about 1 Mev. The
neutrons which have been degraded in energy below the thresholds for
fission and inelastic scattering can undergo either radiative capture or
elastic scattering. Those which undergo radiative capture are, of course,
lost. Neutrons which are scattered elastically lose energy very slowly,
and many collisions are needed to reduce the energy from about 10 5 ev to
1 ev. In this energy range, the probability of resonance capture in U 238
is much greater than the probability of fission in U 235 , as can be seen by

comparing Figs. 19-1 and 20-1 and remembering the small abundance of
the U 235 . Since the neutrons lose energy very slowly, the probability is
very small that they will reach therroal energies without being captured.
Although the fission cross section of natural uranium at therroal energies
is greater than the radiative capture cross section, very few neutrons can
reach these low energies and cause fission, with the result that a chain
reaction is not possible in uranium alone.
A chain reaction can be achieved if the U 238 is removed, that is, in pure
U 235 , or in Pu 2 39• Neutrons of any energy can cause the fission of U 235 ,
so that inelastic scattering of fission neutrons does not reduce the fission
probability. In fact, even neutrons which might reach resonance energies
would contribute to the chain reaction because resonance fission of U 235
is much more probable than radiative capture. Analogous arguments hold
for Pu 239 . Thus, bombs made of U 235 or Pu 239 have been exploded, and
nuclear reactors have been operated which have U 235 or Pu 239 as the fuel
and in which the fissions are induced by fast neutrons. In such a "fast"
reactor, there is very little slowing-down of neutrons, and the presence of
materials which can cause moderation is avoided as far as possible.
A chain reaction can be achieved with natural uranium and a suitable
moderator in an appropriate arrangement. The fission cross section of
U 235 is so Iarge that, in spite of the small abundance of U 235, fission by
therroal neutrons competes favorably with radiative capture. The require-
ment is then that enough neutrons reach thermal energies. This result can
be achieved by using a moderator; a neutron can lose enough energy in a
single collision with a moderator nucleus so that it skips over many reso-
nances, with the result that the resonance absorption of neutrons can be
made small. Heavy water and graphite have been used successfully as
moderators with natural uranium. When D 2O is the moderator, the ura-
nium can be in the forro of a solution of a salt such as uranyl sulfate, or
very small particles of uranium oxide can be suspended uniforroly in the
D 2 O; this type of reactor is said to be homogeneous. When graphite is tbe
moderator, the uranium must be in the forro of large lumps for reasons
which will be seen later, and rods of uranium can be distributed in a regular
way throughout the graphite, forming a lattice. This kind of assembly is
referred to as heterogeneous because of the separation of the fue! and the
moderator. Natural uranium and ordinary water cannot be made to sus-
tain a chain reaction in either a homogeneous or a heterogeneous system
because the absorption cross section of hydrogen for thermal neutrons is
too large.
When fission was discovered (1939), the only material available for chain
reacting systems was natural uranium. The enrichment of uranium in the
20-2] CHAIN-REACTING SYSTEM OR NUCLEAR REACTOR 643
isotope U 235 in gaseous diffusion plants and the production of Pu 239 in

nuclear reactors have provided highly purified nuclear fuels and have
greatly extended the range of possible chain-reacting systems. For ex-
ample, a solution in water of a salt containing enriched uranium can form
a small, homogeneous reactor. Enriched uranium or plutonium can also
be used in partially moderated assemblies in which the fissions are caused
mainly by neutrons of intermediate energy. In these "intermediate" re-
actors there is only enough moderator to slow neutrons down part of the
way, but not all of the way, to thermal energies. The nuclide U 233 , which
can be made in a reactor by irradiating Th 232 with neutrons, can also be
used in thermal, intermediate, or fast reactors.
The rate at which fissions occur in a reactor determines the number
of neutrons produced per unit time and the rate at which heat is produced
-the power level. For a reactor to operate at a steady power level, the
energy released in fission must be removed from the assembly. The fission
energy, originally in the form of the kinetic energy of fission fragments,
neutrons, {3-rays, and 'Y-rays, is converted to heat when these particles are
stopped in the reactor materials. The heat is removed by circulating a
coolant through the reactor, and water, gas, or a liquid metal can be
used as the coolant. The choice of the coolant depends on the purpose of
the reactor and is limited by nuclear and engineering considerations. The
coolant is an impurity from the standpoint of the nuclear properties of the
system, and adds to the nonfission capture of neutrons; the resulting
damage to the neutron economy must be balanced against the efficiency
of heat removal. In a reactor designed to serve as a research tool, to supply
neutrons for experiments, there is no premium on the total power output;
a coolant which is relatively inefficient but absorbs few neutrons can be
used, and air is good enough. In reactors for the production of plutonium,
the rate of production depends on the power level and more efficient cooling
is needed; water is often used because it is a better coolant than air, even
though it absorbs many more neutrons. When the emphasis is on making
useful power, liquid metals may be used to increase the efficiency of heat
removal still further.
Nuclear reactors can be classified according to the characteristics of
the chain-reacting system, and the following features are used to describe
them.m
l. The neutron energies at which most of the fissions occur.

(a) High energy, that is, most of the fissions are induced by the fast
neutrons from fission.
(b) Intermediate energy.
(c) Low energy (thermal).
2. The fuel.
(a) Natural uranium, containing 0.72% of U235 .
(b) Enriched uranium, containing more than 0.72% of U235 .
(e) Pu2a9.
(d) u2aa.
3. The fuel-moderator assembly.
(a) Heterogeneous.
(b) Homogeneous.
4. The moderator.
(a) Graphite.
(b) Water.
(e) Heavy water (D 2 0).
(d) Beryllium or beryllium oxide.
5. The coolant.
(a) Air, C0 2 , or He.
(b) Water or other liquid.
(e) Líquíd metal.
6. The purpose.
(a) Research tool.
(b) Productíon of fissile material.
(e) Power.
The description of nuclear reactors based on the factors listed above
can be illustrated by the following examples. The reactor at the Brook-
haven National Laboratory was originally a thermal, natural uranium,
graphíte-moderated, heterogeneous, aír-cooled research reactor; to obtain
a higher thermal neutron flux, the fuel has recently been changed to highly
enriched uranium. The NRX reactor at Chalk River, Canada, is a thermal,
normal uranium, heterogeneous, heavy water-moderated, ordinary water-
cooled, research reactor. The EBR (Experimental Breeder Reactor at
Arco, Idaho) is a fast, enríched uranium, heterogeneous, liquíd metal-
cooled, power-breeder reactor on a pilot-plant scale. (The term breeder will
be discussed in Section 20-7.)
It is evident that a detailed treatment of nuclear reactors would have
to cover a wide variety of problems, and such a treatment is beyond the
scope of this book. It is possible, however, to indicate by means of sorne
relatively simple examples how·the conditions set by nuclear requirements
affect the design of nuclear reactors. The examples will be limited to
thermal reactors because most of the reactors built so far are of this type,
because they are the best understood reactors, and because most of the
available information is about them.
2<r-3 Thermal nuclear reactors. The neutron cycle. The neutron

balance in a thermal reactor can be described in terms of a cycle that
20-3) THERMAL NUCLEAR REACTORS. THE NEUTRON CYCLE 645
Fission Fission
fragment fragment
//2issio~
/ ):utrons\
Fission
Fission frjagmen t
FiSl!ion
fragment byfast~
neutro~ v•
Fast neutrons
VE
¡ 1-
(1 - l¡) fast neutrons
v..Z¡
me.y escape
-
captured in
U-238 (Resonance D
U
vt(l - 1¡)(1 - p ) - ~
absorption) ecay~ to
neptumum-
239 and Pu-239
v,(1 - l¡)p
vt(l - l¡)p(l - l¡)f
u
~ _ _ _ _ - - ; l o w neutrons
absorbed in U
vt(l - l¡)p(l - li)f(l - g)

Slow neutrons absorbed in U238
ll
ve(l - l¡)p(l - l1)fg
é
Slow neutrons cause fission
of U2M, etc.
9
Decays to neptunium-239
and then to plutonium-239
FIG. 20-2. Schematic representation of a chain reaction based on the fission

of uranium nuclei by thermal neutrons.
shows what happens to the neutrons. Consider this cycle, as illustrated

in Fig. 20-2, starting with the fi.ssion of a U 235 nucleus by a thennal
neutron. In this fi.ssion process, 11 fast neutrons are emitted. These neu-
trons have an average energy above the fission threshold for U 238 , and
some of them can cause the fi.ssion of U 238 nuclei. The chance for such
additional fi.ssions depends on whether the fi.ssion neutrons can collide with
U 238 nuclei before colliding with moderator atoros. A small fraction of the
neutrons that do collide with U 238 nuclei can cause fi.ssions, and for a given
number of neutrons from the thermal fi.ssion of U 235 , some additional
neutrons result from the fast fi.ssion of U 238 nuclei. The total number of
fast neutrons from fi.ssion is increased from II to 11E, where E is a quantity
which may be equal to, or greater than, unity. The quantity E is called
the fast ji8sion factor; it is very close to unity in a homogeneous reactor,
but can be as high as 1.1 in certain heterogeneous reactors. In several
natural-uranium, graphite-moderated reactors, E is 1.03; the fi.ssion of
U 238 nuclei by fast neutrons from the thermal neutron fi.ssion of U 235
nuclei increases the total number of fission neutrons by 3%, and this effect
can be very important.
The PE neutrons diffuse through the pile; most of them are slowed down,
but a fraction l¡ escapes before being slowed down to thermal energies;
11El¡ neutrons are lost, so far as a chain reaction is concerned. The remain-
ing IIE(l - l1 ) neutrons are slowed down by collisions with moderator
atoros, but during the slowing-down process, sorne of them may be cap-
tured by U 238 to form U 239 , which decays to Np 239 and then to Pu 239 •
The capture process is a resonance reaction, as can be seen from Fig. 20-1,
which shows the total cross section of uranium as a function of energy for
neutron energies from one to several thousand electron volts. The reso-
nances have been shown to include radiative capture reactions. Of the
IIE(l - l¡) neutrons which start the slowing-down process, a fraction
11E(l - Z1 )pescapesresonancecapturewhilethefractionvE(l - l1 )(1 - p)
is captured and goes to fonn Pu 2 39 • The quantity pis called the resonance
escape probability. Although the absorbed neutrons cannot cause further
fissions of U 235, and in this sense are lost, they are important because they
contribute to the making of Pu 239 •
The neutrons which escape resonance absorption are slowed down to
thennal energies, where either of two processes occurs. Sorne of the neu-
trons diffuse without being captured and eventually escape from the
system. If the fraction escaping is denoted by z,, the number of thennal
neutrons, per fi.ssion of U 235 , that escape is equal to vE(l - l1)pl1• Of the
remaining IIE(l - l¡ )p(l - l1) thennal neutrons, a fraction f is absorbed
in uranium; the fraction (1 - f) is absorbed in other materials, such as
the moderator or structural materials, and these neutrons are lost. The
number of neutrons still available for carrying on the chain reaction is then
20-3) THERMAL NUCLEAR REACTORS. THE NEUTRON CYCLE 647
vc;(l - l¡)p(l - l,)f. The quantity f is called the thermal utilization.

Not all of the neutrons absorbed in the uranium cause fission of U 236
nuclei. Sorne of the neutrons are absorbed by U 238 to form U 239 and
eventually Pu 239 , while others are absorbed in U 235 to form U 236 . The
fraction of the thermal neutrons absorbed in uranium that cause fission
of U 235 is just <T¡(U)/ua(U).
The number of second generation fissions in U 235 per fission of U 235 by
a first generation neutron is called the reproduction factor, or muUiplication
factor, and is denoted by k. It follows that
u¡(U)
k = vE(l - l¡)p(l - l1)f <Ta(U) · (20-1)
The product v[u¡(U)/u0 (U)) represents the number of (fast) fission neu-
trons produced per thermal neutron absorbed in uranium; it is called
eta (,¡). The multiplication factor can be written as
k = ,¡Epf(l - l¡)(I - l1), (2Q-2)
The quantity k, as given by Eq. (20-2), is usually called the ejfecti11e mul-
tiplication factor and corresponds to an assembly of finite size. To have a
chain-reacting system operating in a steady state, it is necessary that k
be equal to unity; if k is less than unity there can be no chain reaction;
if k is greater than unity the numbers of neutrons and fissions increase
from cycle to cycle and the chain reaction is said to be divergent. The
critical size of a chain-reacting system is the size for which k is equal to
unity; for k < I, the system is suhcritical, and for k > I it is supercritical.
The value of k for a system which is infinitely large is especially useful
in the theory and design of nuclear reactors. There can be no leakage of
neutrons from such a system, and the quantities l¡ and l 1 are both zero.
The multiplication factor is then denoted by k and is given by
00
k.,, = ,¡Epf. (20-3)

This formula is called the f our-factor formula, and the calculation of kao is
one of the main problems in reactor design theory.
In the special case of a reactor in which the fuel contains only U 235 and
no U 238 , both the fast fission factor E a~d the resonance escape probability
pare practically unity. The formula for kao reduces to
(2Q-4)
Of the four factors which make up k ,¡ depends only on the• nuclear

00 ,
properties of the fuel; E depends on the nuclear properties of the fuel as

well as on its size and shape. The remaining quantities, p and f, depend
on the nuclear properties of the fuel, moderator, and any other materials
present in the reactor, and also on the way in which all of these materials
are arranged. One of the basic design problems is that of finding the
relative amounts of ali the materials and the way in which they can be
assembled to yield the largest value of the product pf. For a given fuel,
this arrangement gives the largest value of k,,,. The above discussion of the
neutron cycle, leading to the four-factor formula, is an over-simplified
description of the interactions between neutrons and matter in a chain-
reacting system, because many details have been omitted. But this treat-
ment provides a convenient way of looking at the physics of a reactor
and, for homogeneous thermal reactors, the four factora TJ, E, p, f can be
calculated without too much trouble. Sorne simple examples will be con-
sidered in the next section.
2<>-4 The calculation of the multiplication factor for a homogeneous

thermal reactor. The multiplication factor will be calculated for three
homogeneous systems consisting of natural uranium moderated with
graphite, ordinary water, and heavy water, respectively. In the first of
these, very small particles of uranium are assumed to be dispersed uni-
formly throughout a large block of graphite. When H20 or D 20 is the
moderator, the uranium is assumed to be in the form of a suspension or
slurry of U0 2 in the fluid. lt will be seen that in each case the independent
variable, or design pa.rameter, may be taken to be the ratio of the numbers
of moderator atoros and uranium atoms.
The value of TJ for natural uranium and thermal neutrons is obtained
from the formula
TJ = v Iu,(U)/u,.(U)), (20-5)
and is given in Table 19-1 as 1.34 ± 0.02. For the purposes of later cal-
culations, it is desirable to use consistent values of various constants and
to calculate TJ to three decimal places. For example, the value of TJ changes
when the uranium is enriched with respect to the isotope U 235, and it is
useful to express TJ in a form which shows its dependence on the concentra-
tion of U 235 . Let N 28 and N 25 represent the numbers of atoms of U 238
and U 235 , respectively, per cubic centimeter of uranium. Equation (20-5)
can be written, in greater detail, as
= V
N
25 /
u (U2a11)
(20-6)
TJ N 25CT¡(U235) + N 25a'r(U235) + N 2sCTa(U238) 1
where u,(U 235 ) stands for the cross section for radiative capture by U 235 .
When the numerator and denominator on the right side of Eq. (20-6) are
divided by N 25CT¡(U 235 ), the result is
V
TJ = 1 + [u,(U235)/u,(U235)) + [N 2sCT,.(U238)/N 25u¡(U235)). (2(}-7)
20-4] THE CALCULATION OF THE MULTIPLICATION FACTOR 649
The value of u,(U 235 )/u¡(U 235) is 0.180, and that of u,.(U 238)/u¡(U 235)
is 0.00470, so that Eq. (20-7) becomes
2.44
(20-8)
The ratio N 28 /N 25 is obtained from experimental values of the concentra-

tion of U 235 in uranium, usually expressed in terms of weight percent.
For natural uranium, this concentration is 0.00715 weight percent, or
234.lN 25 ) _ O 00715
( 238.lN 28 + 235.lN 25 Nat. U - • '
and
Hence, for natural uranium, 71 = 1.335. When the ratio N u!N 28 is in-
creased, as in enriched uranium, the ratio N 28 /N 25 is decreased; 71 is in-
creased and more neutrons are produced per thermal neutron absorbed in
uranium.
In the homogeneous systems to be considered the fuel is in the form of
very small particles, of the order of microns in diameter. I t is almost cer-
tain that the fission neutrons will escape from the uranium before colliding
with a U 238 nucleus. The neutrons will enter the moderator and will be
slowed down before they can cause fission of U 238, and the value of E will
therefore be unity. This loss of the additional neutrons contributed by the
fast effect is one of the disadvantages of a homogeneous system.
The condition that a chain reaction be possible in a homogeneous system
of finite size can be written as k,z, > 1, or 71fp > 1, or
pf > -1 = 0.75.
"
This condition expresses a difficult design problem. The resonance escape
probability and the thermal utilization must both be in the neighborhood
of 0.85. In other words, not more than about 15% of the neutrons may be
lost to resonance absorption during the slowing-down process, and not
more than about 15% of the thermal neutrons may be absorbed in materials
other than the nuclear fuel. It will be seen that when the fuel is natural
uranium, the condition (20-9) may be very hard to meet.
The thermal utilization is defined by the equation
f = Thermal neutron absorption in uranium .

(20-10)
Total thermal neutron absorption
If we consideran assembly consisting only of fuel and moderator, and if

both of these materials are 1/v-absorbers, we may write
f = N o<Tao = 1 , (20-11)
No<Tao + N1<Ta1 1 + (N1<Tai/No<Tao)
where the subscripts O and 1 stand for fuel and moderator, respectively,
and N 0 and N 1 represent the number of atoms of fuel and moderator per
cubic centimeter of reactor material, respectively. When absorption in
other materials must be taken into account, additional terms must be
included in the denominator of Eq. (20-11). This equation is actually a
simplified form for the thermal absorption ratio, in which the values of
the cross sections correspond to neutron velocities of 2200 m/sec. The
thermal absorption should really be expressed in terms of the thermal
neutron flux and the absorption cross section, and depends on the Max-
wellian velocity distribution of the neutrons; each term in f should be an
integral of the flux and cross section weighted with the velocity distrihu-
tion, with the integral taken over all possible neutron velocities. It was
shown, however, in Section 18-4, that if the cross section follows the
1/v-law, each absorption term reduces to N,'nll<Tg;, where a superscript
zero has been written to denote the cross section for 2200 m/sec neutrons,
and i stands for either O or l. The superscript has been omitted in Eq.
(20-11) to simplify the expression. In a homogeneous reactor, the flux is
the same for the fuel and the moderator, and the ratio of the absorption
in two 1/v-absorbers reduces simply to the ratio of the macroscopic ab-
sorption cross sections. The absorption cross section of graphite follows
the 1/v-law exactly, while that of uranium deviates from 1/v-behavior by
about one percent. If we neglect this deviation, the error introduced is
again small in illustrative calculations, although a correction would be
made in real design calculations.
The values of the parameters needed for the calculation of the thermal
utilization are listed in Table 20-2. The determination of the absorption
cross section of heavy water presenta a problem because, in practice, heavy
water always contains sorne H 2 O; it may, for example, consist of 99.75
atom percent of D and 0.25 atom percent of H. But the absorption cross
section of H is so much greater than that of D that the effect of the H
impurity must be taken into account. We, therefore, write for the macro-
scopic absorption cross section of the heavy water
pract
¿ a
(D2O) = mN(D2O)<1a(D2O) + (1 - m)N(H2O)<1a(H2O), (20-12)
where m is the fraction, in atom percent, of D in the heavy water, and

(1 - m) is the fraction, in atom percent, of H; N(D 2O) and N(H 2O) are
the numbers of molecules in one cubic centimeter of D 2O and H 2O, re-
20-4] THE CALCULATION 0F THE MULTIPLICATION FACTOR 651
TABLE 20-2
QUANTITIES NEEDED F0R THE CALCULATION 0F THE
THERMAL UTILIZATION IN CERTAIN HoMOGENEOUS
REACTORS
Absorption
cross section Number of atoms/cm 3
Material per atom for of material,
2200 m/sec X 10- 24
neutrons, b
u 7.68 0.0478 (density: 18.9 gm/cm 3 )

H 0.332 0.0668 (in füO)
D 0.00046 0.0662 (in D20)
e 0.0045 0.0803 (density: 1.60 gm/cm 3 )
o <0.0002 (""O)
spectively; u a(D2O) and u a(H2O) are the absorption cross sections per
molecule for 2200 m/sec neutrons. On inserting values from Table 20-2
into the expression (2Q-12), we get
pract
I: a
(D2O) = (0.9975)(0.0331)(9.2)10- 4 cm- 1
+ (0.0025) (0.0334) (0.664) cm - 1
= (0.304 + 0.554)10- cm- =4 1
0.858 X 10- 4 cm - 1
•
Thus, the small impurity of H 2O in the D 2O contributes more to the total

macroscopic cross section of the heavy water than <loes ali of the D 2O.
lt is convenient to introduce an effective value for the absorption cross
section per molecule of heavy water, defined by the relation
N(D2O)u!ff (D 2O) = 0.858 X 10-4 cm- 1 ,
so that
Since the absorption cross section of oxygen is practically equal to zero,

we may write
u!ff (D) = 0.00130b.
This cross section has been introduced because it is convenient to use an
effective absorption cross section per atom of deuterium together with the
number of atoms of deuterium per cubic centimeter of reactor material
when we compare the results for the three reactors. Similarly, for the
H 2 0-moderated reactor we shall use the absorption cross section per hy-
drogen atom and the number of hydrogen atoms per cubic centimeter.
We may finally write for the thermal utilization for the three reactors,
uranium-graphite,
1
f = 1 + (Ni/ N o)(0.000579) ; (20-13)
uranium-heavy water,
1
f = l + (N i/N )(0.000169) ;
0
(2CH4)
uranium-water,
1
f = 1 + (Ni/N )(0.0432)
0
(20-15)
The resonance escape probability can be calculated by means of a for-

mula <2 > the derivation of which is beyond the scope of this book,
1
p=exp[-- No(/uadE) ]· (20-16)
~.1 N1 E eff
The quantity <fua[dE/EDer1 is called the effective resonance integral and is
a measure of the total neutron absorption in the resonance energy region.
It has been determined experimentally for homogeneous mixtures of
uranium and scattering materials and may be expressed by the empirical
formula
( /<Ta
dE)
E el! = 3.9 No (L•)º· 41
5
b. (20-17)
In this formula, L,INO represents the total macroscopic scattering cross

section of the reactor material per atom of uranium, in the resonance
energy range. For the uranium-graphite reactor we have
L, = N 1u,(C)
No N0
+ N o<T,(U) = N1
N 0 u,
(C) + u, (U).' (20-18)
for the uranium-D 20 reactor,
L, N(D20)u,(D20) + N(U02)u,(U02)
No No
= N1
No u,(D) + u,(U) + (12 N1
No+ 2) u,(O); {20-19)
20-4) THE CALCULATION OF THE MULTIPLICATION FACTOR 653
TABLE 20-3
ScATTERING CRoss SECTIONS IN THE
RESONANCE ENERGY REGION
Scattering Average logarithmic

cross section energy loss per
Material
per atoro or collision
molecule, b
u 8.3
e 4.7 0.158
H 20.5 1.00
D 3.3 0.725 1
1
o 3.8 0.120
H20 44.8 0.926
D20 10.4 0.508
for the uranium-H 2 0 reactor,
:E,
No = N
No1 CT,(H) + CT,(U) + (12 No+ 2
N1 )
CT, ( O) . (20-20)
In the last three equations, the scattering cross sections are those in the
resonance region, about 1 ev to 10 5 ev; the appropriate values are listed
in Table 20-3.
The results of the calculations are collected in Table 2o-4, where the
values off, p, pf and k,., are listed for different values of N i/N 0 , the ratio
of the number of moderator atoms to that of uranium atoms. For small
values of N i/N 0 , f approaches unity, while pis small. This result is ex-
pected because both thermal and resonance absorption should increase as
the amount of uranium in the reactor increases. We expect, therefore,
that an arrangement which increases f should decrease p, and that the
product pf should pass through a maximum for a given moderator. Hence,
k,., should also pass through a maximum as Ni/N O is varied, and the
calculations show that this is indeed the case. The maxima are quite flat,
occurring in the range Ni/No= 4 to 10 for H 2 0, 150 to 500 for D 2 0 and
300 to 600 for graphite.
The most important result is that values of k,., greater than unity can
be attained with homogeneous mixtures containing natural uranium and
heavy water, but not with homogeneous mixtures of natural uranium and
graphite, or natural uranium and ordinary water. It is this result, espe-
cially in the case of graphite, which led to the consideration of heterogeneous
systems. When graphite is the moderator, the failure results from the low
value of the resonance escape probability. Adequate values off can be
~
TABLE 2Q-4
VALUES OF THE THERMAL UTILIZATION, RESONANCE ESCAPE PROBABILITY AND MULTIPLICATION

FACTOR FOR THREE Ho111:ooENEous FuEL-MonERATOR SYSTEMS
N1 U ranium-graphite U02-D20 U02-H20

N2 f p pf k.,, f p pf kao f p pf kao z
q
(")
1 0.999 o 1.000 0.018 0.018 0.02 0.958 0.317 0.304 0.41 t<
2 0.999 o 1.000 0.102 0.102 0.14 0.920 0.510 0.469 0.63 f:!
::o
4 0.998 0.006 0.999 0.267 0.267 0.36 0.853 0.680 0.580 0.78 t,J
5 0.997 0.012 0.999 0.333 0.333 0.45 0.822 0.724 0.595 0.80 z
t,J
10 0.994 0.062 0.062 0.08 0.998 0.509 0.508 0.63 0.698 0.832 0.581 0.78 ::o
Cl
20 0.989 0.167 0.165 0.22 0.997 0.651 0.649 0.87 0.536 0.903 0.484 0.65 -<
50 0.972 0.364 0.354 0.47 0.991 0.787 0.780 1.04 0.316 0.974 0.308 0.41 fg
100 0.945 0.523 0.494 0.66 0.984 0.853 0.839 1.12 q
::o
150 0.920 0.586 0.539 0.72 0.976 0.882 0.861 1.15 (")
200 0.896 0.641 0.574 0.77 0.968

gJ
0.900 0.871 1.16
300 0.852 0.704 0.600 0.80 0.952 0.920 0.876 1.17
400 0.812 0.744 0.604 0.81 0.938 0.932 0.874 1.17
500 0.775 0.772 0.598 0.80 0.924 0.940 0.869 1.16
600 0.742 0.793 0.588 0.79 0.909 0.946 0.850 1.13
800 0.683 0.820 0.560 0.75 0.882 0.954 0.841 1.12
1000 0.633 0.841 0.532 0.71 0.857 0.960 0.823 1.10 ñ
=
~
t,;)
o
20-4] THE CALCULATION OF THE MULTIPLICATION FACTOR 655
attained for values of Ni/N O up to about 300 or 400; but a value of p of

0.82 is not reached until N ¡/NO = 800, with the result that the maximum
value of pf is about 0.6, and the maximum value of k~ is about 0.8. For-
tunately, the properties of slowing-down and resonance absorption are such
that the resonance escape probability can be increased significantly by
lumping the uranium. In the case of water, the difficulty is in the small
value of the thermal utilization obtained for all but the smallest values of
Ni/N O ; to get useful values off there must be so much uranium that there
is too much resonance absorption. The preceding arguments can be put
in another way. With D 20, the thermal utilization is favorable, say 0.85
or more, for values of Ni/N O up to about 1000, because of the very small
absorption cross section of D 2 0. In this wide range it is possible to find a
fairly wide range of values over which p is also favorable, and a chain
reaction can be attained. With graphite as moderator, the slowing-down
is in such small steps that the probability is high that neutrons will have
energies corresponding to the resonance energies of U 238 ; if U 238 atoms
are readily available to the neutrons, the resonance capture is high, and
the escape probability is low. The resonance escape is small until the
uranium is very dilute, but by then the thermal utilization has decreased
toomuch.
When uranium enriched in U 235 is available, a chain reaction can be
achieved with either graphite or water as the moderator. With enriched
uranium, both f and 71 are increased, the increase in 71 being more impor-
tant. We consider the graphite-moderated reactor. The maximum value
of k~ with natural uranium and graphite was about 0.8, so that an increase
of about 20% in 71 would be needed to bring k~ up to unity, ora value of 71
between 1.6 and 1.7. Since 71 = 2.07 for pure U 235 , it is clear that the
necessary value of 71 can be attained for sorne concentration of U 235
between that in natural uranium and that of pure U 235 . The value of 71
can be computed from Eq. (20-7). The absorption cross section of uranium
also increases with increasing U 235 concentration so that f will increase;
<Ta can be calculated from the formula
N(U235)<Ta(U235) + N(U238)<Ta(U238)
<Ta(U) = N(U235) + N(u2as) (20-21)
ua(U
235
) + (N 2s/N 25)<Ta(U 238) (20-22)
= 1 + (N2s/N2s) .
The values of the cross sections to be used are
The calculation of k~ has been made for N ifN O = 400, with p equal to
0.744, the value for natural uranium. It is assumed that the decrease in
TABLE 20-5
THE MULTIPLICATION FACTOR FOR A HOMOGENEOUS MIXTURE OF
ENRICHED URANIUM AND GRAPHITE, N¡/N 0 = 400
N2s
- ua, b f pf 77 k.,,
N2s
137.8 7.68 0.810 0.603 1.335 0.80

130 8.01 0.816 0.607 1.362 0.83
120 8.45 0.824 0.613 1.399 0.86
110 9.04 0.834 0.620 1.439 0.89
100 9.57 0.841 0.626 1.479 0.93
90 10.3 0.852 0.634 1.523 0.97
80 11.3 0.862 0.641 1.569 1.01
70 12.5 0.874 0.650 1.618 1.05
60 14.1 0.887 0.660 1.670 1.10
50 16.3 0.901 0.670 1.726 1.16
the concentration of U 238 over the range of U 235 concentration is too small
to affect the resonance escape probability; a more careful calculation would
take this effect into account. The results are shown in Table 20-5. lt is
seen that a chain reaction can be achieved when the U 235 atom concen-
tration is increased from 1: 138 to about 1: 80. Similarly, a chain reaction
can be achieved in a homogeneous system with partially enriched uranium
and ordinary water. <3 >
20-5 The heterogeneous thermal reactor. lt wa.s seen in the la.st section
that a homogeneous natural uranium-graphite assembly cannot support a
chain reaction no matter how large it is, but that when the uranium is
sufficiently enriched with respect to the isotope U 235, a chain reaction is
possible. In the absence of enriched uranium, the question a.rose as to
whether natural uranium and graphite could be arranged in sorne way
which could lead to a chain reaction. The results listed in Table 20-3
indicate that the main difficulty is the small value of the resonance escape
probability. The value of p does not become appreciable until the uranium
concentration becomes very low, but it is hard to conceive of a chain
reaction in a system with only a very small amount of natural uranium.
lt was apparent, therefore, that something had to be done to increase the
value of the resonance escape probability in an assembly which has enough
uranium to yield an adequate value (0.8 to 0.9) for the thermal utilization
factor. The solution to this problem was suggested by Fermi and Szilard,
who proposed the use of lumps of uranium of considerable size embedded
in the moderator. The use of lumps also affects the thermal utilization
20-5] THE HETEROGEXEOUS THERMAL REACTOR 657
a11d the fast fission factor, and a qualitative discussion of the various
effects of lumping will be useful.
A consideration of the way in which resonance absorption takes place
shows that the lumping of the uranium increases the resonance escape
probability. Resonance absorption occurs when neutrons, during the
course of the slowing-down procesa, reach energies which, together with
239
the binding energy, correspond to levels of the compound nucleus [U ].
238
When a neutron with such an energy collides with a U nucleus, there
is an appreciable chance that it will be absorbed. In a homogeneous
uranium-graphite assembly, each uranium particle has the same chance of
capturing a certain fraction of the resonance neutrons, but when the
uranium is present in relatively large lumps this is no longer the case.
For, suppose that neutrons collide with a lump of uranium; those neutrons
which happen to have energies within a resonance energy band may be
absorbed at the surface of the lump, while neutrons with energies which
do not correspond to a resonance band pass into the lump. Sorne of these
neutrons may undergo elastic collisions with the uranium nuclei and,
although the energy loss per collision is very small, a fraction of them
may be scattered into resonance energy bands and absorbed. Many of
the neutrons, however, can pass through the lump without being absorbed;
they may either escape elastic collisions with uranium nuclei altogether,
or they may be scattered, but not into a resonance band. The probability
of resonance absorption inside the lump is therefore smaller than that at
the surface of the lump. In other words, U 238 nuclei inside the lump have
a smaller chance of capturing neutrons during the slowing-down process
than U 238 nuclei near the surface of the lump, and there is sorne "self-
shielding" of the uranium against resonance absorption. The neutrons
which re-enter the moderator undergo further slowing down in a region
devoid of uranium, and they may lose enough energy to miss several
resonance bands before entering another lump or reaching thermal energies.
The magnitude of this effect depends on the spacing of the uranium lumps
in the moderator matrix, and its probability is appreciably greater than in
a homogeneous assembly. Thus, one advantage of the heterogeneous
assembly is the lower resonance absorption owing to filtering at the lump
surface of neutrons whose energies lie within the resonance bands of
uranium.
It follows from the above discussion that the resonance absorption may
be divided into two parts, a surface absorption and a volume absorption,
and the size and shape of the uranium iumps can be chosen to make the
resonance escape probability relatively high. Sorne qualitative notions
about the size of the lumps can be obtained quite easily. To cut down
the volume absorption, the size of the lump should not be too large com-
pared with the mean free path of resonance neutrons in the lump. Al-
though this mean path is not known accurately, a rough estimate can
be made by considering thermal and fission neutrons as limiting cases.
The total thermal cross section of normal uranium is 16.0 b and uranium
has 0.048 X 10 24 atoms/ cm 3 ; the total mean free path for thermal neu-
trons is then
1
= (.048) (16.0) = 1·3 cm.
The average total cross section for fast neutrons is 7.2 b (from Table 20-1),
and the mean free path is about 3 cm. A good lump size corresponds,
therefore, to a diameter between 1.3 cm and 3 cm, or about one inch.
The flux of resonance neutrons is lower inside the lump than near the
surface because a considerable part of the resonance absorption occurs
near the surface of the lump. This effect favors large lumps but the choice
of lump size depends also on the effect of lumps on the fast fission factor
and the thermal utilization. The value of the fast fission factor increases
with the size of the lump because the probability of a collision between a
fast neutron and a U 238 atom increases with the lump size. This effect
also tends to favor larger sizes.
The main disadvantage of a heterogeneous assembly of uranium and
a moderator is the decrease in the thermal utilization, as compared with
that in a homogeneous mixture with the same fuel concentration. Suppose
that monoenergetic thermal neutrons are incident on the surface of a
uranium lump. Sorne of these neutrons are absorbed near the surface
and the thermal flux is decreased; as the remaining neutrons penetra te
the lump, more are absorbed. The thermal neutron flux is then reduced
inside the lump and the probability of further absorption is decreased.
The effect is directly analogous to that in resonance absorption, except
that it is not desirable in the case of thermal absorption. In other words,
the self-shie/,ding of the uranium lump against the absorption of thermal
neutrons results in a decrease in the thermal utilization.
The choice of a suitable lump size depends on the balance among the
effects on p and E which tend to increase the value of k,. and the decrease
in f which tends to decrease the value of k,.. The actual calculation of
E, p, and f is much more complicated for a heterogeneous assembly than
for a homogeneous one. Appropriate methods can be found in the general
references listed at the end of this chapter. Values up to k,. = 1.075 have
been obtained for natural uranium, graphite-moderators, <4 > and values up
to 1.21 have been otitained< 5 > for natural uranium, heavy water-moderated
reactors.
Cylindrical rods of natural uranium have been used in many reactors.
At Brookhaven, the fuel elements were originally 1.10 in. in diameter
(1.40 cm in radius) and 11 ft long; there were two such rods, covered with
20--5) THE HETEROGENEOUS THERMAL REACTOR 659
TABLE 20-6
SLOWING-DOWN LENGTH OF FISSION
NEUTRONS IN 8EVERAL MODERATORS
Slowing-down Slowing-down
Moderator length, L. area, L~,
cm in cm 2
H20 5.3 2.1 28

D20 11.2 4.4 125
Be (density: 1.85 gm/cm3 ) 9.8 3.9 96.
C (density: 1.60 gm/cm 3 ) 19.1 7.5 364
thin aluminum jackets, in a channel through which air could be pumped

to cool the uranium. The French heavy water-moderated reactor, desig-
nated as P2, has natural uranium rods 1.3 cm in radius, also ciad in alumi-
num. The British Experimental Pile (BEPO), which is a heterogeneous,
natural-uranium, air-cooled, graphite pile has rods 0.9 in. in diameter.
In a heterogeneous fuel-moderator assembly (lattice), the choice of the
lattice spacing, or the distance between fuel elements, is also important.
From the standpoint of neutron economy, the choice involves a balance
between the slowing-down and diffusion properties of the assembly. If
the moderator is a very weak absorber (D 2 0 or graphite), the distance
between lumps should be large enough so that, on the average, the fast
neutrons from one lump are slowed down to thermal energies before reach-
ing another lump. There is a quantity L,, called the slowing-down length,
which can be either determined experimentally or calculated theoretically,
and which is a measure of how far a fission neutron travels on the average
while being slowed down to thermal energies. The values of the slowing-
down length for sorne moderators are listed in Table 20-6, together with
L:, the slowing-down area. For reactors moderated with D 2 0 or graphite,
we would expect the lattice spacing to be at least one slowing-down length.
In the BNL reactor, the spacing is 8 in, in BEPO it is 7½ in, in P2 it
is 5.9 in, so that our estimate of the lattice spacing is a reasonable
one. Ordinary water is a much stronger absorber than heavy water or
graphite and a lattice spacing of one slowing-down length would permit
too much absorption of slow neutrons. Hence, in a reactor with uranium
enriched in U 235 and moderated with H 2 0, the rods would be quite close
together and Ni/No would have to be much smaller than for D 20 or
graphite. <5 >
In detailed design calculations, the choices of the lump size and the lattice
spacing are not independent and must be considered together. The final
choice of design parameters may also be strongly influenced by factors

other than the nuclear properties of the fuel and moderator. Those factors
may include the ease or efficiency of heat removal, structural properties,
and effects of emphasis on a special purpose for the reactor.
2o-6 The critica! size of a thermal reactor. The value of the multipli-
cation factor alone does not determine the critica! size of a chain-reacting
system. The multiplication factor k.,, tells how the neutron concentration
would increase from generation to generation if no neutrons leaked out of
the system, and it must be combined in sorne way with a quantity which
tells something about the neutron leakage. The mathematical theory of
the critica! size is an important part of reactor theory and it will be illus-
trated by means of a very simple problem. We shall considera homogene-
ous reactor in the form of a slab, infinite in the y- and z-directions but of
finite thickness in the x-direction. It will be assumed that the nuclear
properties, including k.,,, are known, and the problem is to determine the
critica! thickness a at which the slab will sustain a chain reaction. The
treatment is an extension of that of Section 18-5, where we considered
the diffusion of thermal neutrons from an infinite plane source, and sorne of
the basic ideas of the earlier treatment can be used again. In-the new prob-
lem, the source is distributed uniformly throughout the slab; neutrons are
both produced and absorbed throughout the slab and can leak out of the
boundary of the slab. We shall assume at the start that all neutrons are
thermal and monoenergetic, that the neutrons released in fission have a
Maxwellian velocity distribution which can be represented by the most
probable velocity. This assumption may be made if the reactor materials
are 1/v-absorbers, because of the arguments of Section 18-4. The treat-
ment based on this assumption is called the one-group model because all
the neutrons are taken to have the same energy. The use of the one-group
model simplifies the mathematics, and the model can easily be improved
to give satisfactory results for certain types of reactors.
We seek the thickness needed for the reactor to operate in a steady
state. The condition for this equilibrium may be written as a neutron
balance equation,
Rate of neutron leakage + rate of neutron absorption
= rate of neutron production. (20-23)
If we apply this balance condition to a slab of thickness dx between x and
x +dx, we have
2
Net rate of leakage out of element dx = - >-tri!
3
ddx2
n dx1 (20-24)
20-61 THE CRITICAL SIZE OF A THERMAL REACTOR 661
where nis the neutron density, v is the constant neutron velocity; Xtr is
the transport cross section for the reactor material, and is given by
1
= (20-25)
>-tr N tUtrl + N oO"trO
The leakage rate gives the net number of neutrons leaving the slab of
thickness dx per second across a unit area (1 cm 2 ) normal to dx.
The absorption rate is given by
Rate of absorption of neutrons in element dx = nv :Ea dx, (20-26)
where
(20-27)
is the total absorption cross section of the reactor material. The expression
on the right of Eq. (2ü-26) gives the number of neutrons absorbed per
second in a volume element one square centimeter in area and with thick-
ness dx. The production rate is obtained if we note that each thermal
neutron absorbed results in the emission of k.,, new neutrons,
Rate of production of neutrons in the element dx = nv :Ea k.,, dx. (20-28)
The neutron balance equation is then
>-trV d 2n
- 3 dx 2 + nv :Ea = nv :Ea k.,,, (20-29)
or
d2n + -,-
-d2
3:Ea (k.,, - l)n = O. (20-30)
X "tr
We define a diffusion length L for the reactor by analogy with the defini-
tion used in Section 18-5 for a single substance,
{20-31)
where Xtr and :Ea are defined by Eqs. (20--25) and (20--27), respectively.
The neutron balance equation then becomes
(20-32)
lt is now possible to take into account the fact that fission neutrons are
released as fast neutrons and must be slowed down to thermal energies.
lf the reactor is well moderated, that is, if nearly all the fissions (say, 95%)
are produced by thermal neutrons, L 2 can be replaced by
(20-33)
where L, is the slowing clown length introduced in Section 20-5; M 2 is

called the migration area. The balance equation then becomes
d 2n kao - 1
dx2+~n=0. (2~34)
Equation (2~32) is the modified one-group equation. A more detailed theo-

retical treatment< 7 > shows that this model gives reasonably good results
for well moderated reactors for which kao is close to unity, or k 001 « 1;
-
natural uranium, graphite moderated reactors are an example.

The solution of Eq. (2~34) requires boundary conditions, and the
accurate formulation of these conditions is a serious problem. We adopt
the simple, although not quite accurate, condition that the neutron density
must vanish at the outer boundary of the reactor. If the center, x = O,
of our coordinate system is taken at the central plane of the reactor, the
condition is that
n(x) = O at x ± ½a. (20-35)
lf the reactor is uniform, the neutron density must be symmetric about

x = O; physical intuition also tells us that the neutron density should be
positive throughout the interior of the reactor. The general solution of
Eq. (2~35) is
n(x) = A cos Bx + C sin Bx, (20-36)
where
2 _ k,. - 1
B - M2 ' (2~37)
and A and C are arbitrary constants. Now, si,n Bx is not symmetric about
x = O so that C must vanish, and the solution is just
n(x) = A cos Bx. (20-38)
The condition (2~35) requires that
or
7r 1rM
a= B = (kao - 1)112
(2~39)
20-6] THE CRITICAL SIZE OF A THERMAL REACTOR 663
Equation (20-39) gives the critical thickness of the reactor in terms of

+
the nuclear properties as expressed in k.. and M = (L 2 L:) 112. The
neutron density distribution as a function of position is given by
1rX
n(x) = A cos - ,
a
{2Q-40)
where A is an arbitrary constant whose value is determined by the rate

of heat removal or the power level of the reactor. Equation (2o-39) shows
that if k .. is less than unity, there is no real value of the critical thickness;
if k.. = 1, the critical thickness is infinite (a result consistent with the
definition of k .. ); if k .. > 1, a chain reaction can be achieved in an as-
sembly of finite size. The critical radius Re of a spherical reactor is
1rM
Re= (k .. - 1)112' (2Q-41)
and the side of a critical cube is given by
v'3 1rM (2Q-42)

Ae = (k .. - 1)112
The sphere has a smaller surface-to-volume ratio than the cube, and a
smaller critical volume. For the sphere, the critical volume is
Ve=; 1rR~ = 130 (k ..M~ 1r12' (2Q-43)
while for the cube,
Ve = Ae
3= 161 (k..M2_ l )ª'2 , {2Q-44)
or about 24% greater than the critical volume of the sphere. The critical
mass of uranium can be determined from the critical volume and the
mass of uranium per unit volume; the latter is, of course, one of the basic
design parameters.
It is usual to build a reactor so that its size is greater than the critical
size, that is, so that k = 1 + 6, where 6 is called the excess reactivity.
The excess reactivity is needed to allow neutrons to be used, either in
the form of beams coming out of the reactor or by being absorbed inside
the reactor to make new radionuclides, and for other reasons too in-
volved to discuss here. The existence of the excess reactivity makes it
necessary to have a control system, usually in the form of strong neutron
absorbers such as boron-steel rods or cadmium strips. The value of the
actual multiplication factor in the operating reactor can be kept at unity
: by proper adjustment of the control rods.
Frn. 20-3. Photograph of the model of the uranium-graphite reactor at the

Brookhaven National Laboratory.
The leakage of neutrons from a reactor can be reduced by surrounding

the reactor with a reflector made of a weak neutron absorber; graphite is
often used for this purpose. The reflector knocks neutrons which leave
the reactor back into it, and the decreased leakage reduces the critical
size of the reactor, with a resulting saving in fuel. The reactor system,
including the reflector, must be enclosed in a shield, usually of concrete,
to reduce the intensity of the radiations (neutrons and 'Y-rays) leaving
the reactor to values below the biological tolerance limits.
Figure 20-3 is a schematic diagram of the uranium-graphite reactor at
the Brookhaven National Laboratory. The diagram shows how sorne of
the features which have been discussed are incorporated into the design of
an actual reactor.
20-7 Power and breeding. The possibility that a power industry based
on nuclear fission will be developed raises the question of the availability
and cost of fissionable material. <sJ In the discussion so far, the fissionable
materials considered have been U 235 , Pu 239 , and U 233 , none of which is
available naturally in large amounts. lt would be inefficient to get power
on a large scale from the fission of U 235 because only 0.7% or less of tl:ie
20-7] POWER AND BREEDING 665
available uranium could be used directly. Plutonium has been made by

the fission of U 235 in a reactor containing U 238 , but this process is involved
and expensive. Similarly, U 233 , which <loes not occur in nature, can be
made in a reactor containing Th 232 • Although the problem of availability
and cost of nuclear fuels is a serious one, a unique and remarkable solution
is possible because of one of the nuclear properties of the fission process.
This property is the emission, on the average, of more than two neutrons
per fission.
Suppose, for the sake of example, that 3 neutrons are emitted per fission.
One of these is needed to keep the chain reaction going, to induce fission
in another fuel atom, say an atom of U 235 • Of the two neutrons that are
left, one may be used to convert an atom of uranium to one of plutonium,
leaving one neutron. If this last neutron could be used to produce another
atom of plutonium from another atom of U 238, one atom of U 235 would
have been used and two atoms of Pu 239 would be made, leaving a profit
of one atom of fissionable material. In other words, more fissionable
material would be produced than consumed. This example is, of course,
an oversimplified one. <9> It is inevitable that sorne neutrons should be
lost by leakage or by parasitic absorption, but so long as more than two
neutrons are produced in fission there is a possibility that more nuclear
fuel may be produced than is consumed. This process is called breeding;
if it could be made to work, a reactor could make new fuel for itself and,
in addition, a stockpile of fissionable material could be built up for use in
new reactors.
The process of breeding fissionable material depends on two materials,
one of which is fissionable and one of which is jertile. Uranium-238 and
Pu 239 form such a pair, and Th 232 and U 233 form another pair. Consider
the first pair and suppose that enough Pu 239 is available to achieve a chain
reaction in a system which can be used to produce power. The thermal
fission of Pu 239 yields 3 neutrons, so that if sorne of these neutrons could be
absorbed in U 238 to form more Pu 2 39 , a breeding cycle might be achieved.
In this case, U 238 is the fertile material because although it is not itself
usefully fissionable, it can be converted into a good fissionable material,
and it is conceivable that a reactor system could be built in which more
plutonium is produced than is consumed. The nuclear property that
determines whether the possibility of breeding exists is not the value of
the number of neutrons emitted per fission, v, but the number produced
per neutron absorbed in the fue!, 7/, because the nonfission capture of
neutrons by the fue] can be large enough to interfere seriously with breed-
ing. The value of 7/ for plutonium and thermal neutrons (Table 19-1) is
2.08 ± 0.02 and turns out to be only slightly greater than two. This
result seems to make impractical the breeding of Pu 239 in a thermal
reactor. It is expected, on theoretical grounds, that the ratio of the fission
cross section to the total absorption cross section should be closer to unity
at high neutron energies than for slow neutrons, and that ,¡ should be
close to II in a fast reactor. Consequently, breeding of Pu 239 should be
possible in a fast neutron, Pu 239-U 238 system. A reactor, the Experimental
Breeder Reactor, or EBR, has been built to study the feasibility of power
production and breeding in such a system. A large power-breeder reactor,
the Enrico Fermi Atomic Power Plant has been built near Detroit, Michi-
gan to combine production of electricity with the breeding of Pu 239 . 00 >
The value of ,¡ for U 233 and thermal neutrons (Table 19-1) is 2.28 ± 0.02
which makes the breeding of U 233 from Th 232 in thermal reactor systems
appear feasible. Research and development work are in progress on such
systems. <ll> One design involves the use of a solution of U 233 0 2 S0-1 in
D 2 0; another involves the use of a solution of U 233 Bi in liquid bismuth,
with graphite as moderator. The U 233-Th system could also be used in a
fast reactor, but the greater value of II for Pu 239 favors the use of the latter
material in fast breeder reactors.
If breeding could be achieved, U 238 and Th 232 could be converted into
fissionable materials. The resulting increase in available nuclear fuels
would be a step in the direction of economic nuclear power. Much informa-
tion is needed concerning the cross sections for reactions of fast and inter-
mediate energy neutrons with fissionable and fertile materials, and much
work is being done in this field.
2~ Energy production in stars. Thermonuclear reactions. One of

the most intriguing problema in physical science is that of the source of
the energy emitted by stars. The sun, which is not a particularly impres-
sive star, emits electromagnetic energy ata rate of about 4 X 10 23 ergs/sec
or 2 ergs/gm/sec, and astronomical and geological evidence shows that the
sun has been radiating energy at about its present rate for severa! billion
years. Chemical reactions cannot possibly be the source of this energy,
because even if the sun consisted of pure carbon its complete combustion
would supply enough energy to maintain the radiation for only a few
thousand years. Another possible source of energy is the conversion of
gravitational energy into heat· energy, a process that is analogous to the
production of electricity from the falling of water, as in a hydroelectric
plant. But it has been shown that if contraction were taking place it could
supply not more than one percent of the total energy output needed, and
if it were the only energy source, the sun could be not more than 20 million
years old.
The inability to account for the energy emitted by the sun on the basis
of ordinary energy sources led to the suggestion that subatomic or nuclear
processes provide the necessary energy. Many exothermic nuclear reac-
tions liberate energies of severa! Mev per nuclear particle and if all the
20-8] ENERGY PRODUCTION IN STARS 667
nucleons in a gram of matter could be used, about 10 19 ergs would be

produced per gram of matter. At the rate of 2 ergs/gm/sec, reactions of
this kind could supply energy for about 10 11 years and are the only kind
of reactions which could do so. The problem is to find a nuclear reaction
or set of reactions which result in the observed rate of energy emission
under the conditions of temperature and density inside the sun, and which
are consistent with the available information about the chemical composi-
tion of the sun. When the cross sections for the reactions are known,
either from experiment or theory, the rates of the reactions and the rate
of energy production can be calculated.
Stars can be classified according to their luminosity, or brightness, and
their radiation spectra. The spectral properties of a star are related to
the effective temperature at which the star radiates energy; the effective
temperature is a convenient way of describing the total radiation per unit
area. The sun is an example of one class of stars, the so-called main se-
quence stars. The effective surface temperature of the sun is about 6000ºK,
and the interna} temperature may be as high as 2 X 10 7 ºK. The stars of
the main sequence have effective surface temperatures from about 2000ºK
to about 50,000ºK, and densities between about rr; and 10 times that of
the sun. In addition to the main sequence stars there are white dwarfs,
red giants, variables, novae, and supernovae. The white dwarfs have
extremely great densities, possibly 100,000 times as great as the sun, and
are very faint, that is, have very small luminosities. The red giants have
very low densities and high luminosities. The variable stars show recurring
variations in luminosity and surface temperature; the variations may be
periodic or irregular. Finally, there are novae and supernovae, which
show sudden large increases in luminosity. There are wide variations in
interna! conditions among the different kinds of stars, in their tempera-
ture, density, pressure, and chemical composition. There may, therefore,
be different mechanisms of energy production; different nuclear processes
may occur in different types of stars, and more than one set of reactions
may be responsible for the energy production in stars. The sun is the best
known and most carefully studied star, and its energy production as
calculated from nuclear reactions can be compared with well-known
astrophysical data.
It can be stated first that the fission of heavy elements cannot supply
the sun's energy because the abundance of these elements in the sun is
much too sma11< 12> to account for the rate of emission and for the lifetime
of the sun. The light elements predominate, and hydrogen and helium
together form about 90% by weight of the sun's matter, with approxi-
mately equá.l fractions of those two elements. It seems likely, from the
chemical composition, that the nuclear processes should involve hydrogen
and helium, and the nuclear properties of these elements support this
idea. Assume, for example, that 4 protons can be combined to form a

helium nucleus. The process would be exothermic, as can be seen from the
mass balance,
atomic mass of 4 hydrogen atoms = 4.13258 amu

atomic mass of 1 helium atom = 4.00387
difference in mass = 0.02871 amu
= 26.7 Mev
= 42. 7 X 10- 5 erg.
Thus, about 10- 5 erg could be produced for each proton destroyed and,
since one gram of the sun's matter contains about 2 X 10 23 protons, the
available energy supply would be about 2 X 10 18 ergs/gm-of the order of
magnitude needed.
The possibility that four protons could collide to form a helium nucleus
is ruled out because the probability for such a reaction under solar condi-
tions is too low. It seems more likely that the four protons are formed
into a helium nucleus by a series of reactions, that is, in a cyclic nuclear
reaction. The rates of the reactions depend on the number of nuclei present
per unit volume and also upon the temperature; the higher the tempera-
ture, the faster the thermal motion of the particles, and the more frequent
and energetic are the collisions. At stellar temperatures of 10 to 20 million
degrees, the kinetic energies resulting from thermal motion are in the
neighborhood of 1 kev, as compared with ir; ev for particles at room
temperature on earth, and the nuclei in the sun's interior are assumed to
have a Maxwellian velocity distribution with a kT energy of about 1 kev.
The reactions that occur under these conditions are called thermonuclear
reactions.
Two sets of thermonuclear reactions have been proposed as sources of
energy in the sun and other stars of the main sequence. One set, sorne-
times called the proton-proton chain, consists of the reactions
(H 1 + H 1 -+ H 2 + 13+ + 11 + 0.42 Mev) X 2,

(H 1 + H 2 -+ He 3 + -Y + 5.5 Mev) X 2, (20--45)
He 3 + He 3 -+ He 4 + 2H 1 + 12.8 Mev.
For the third reaction to occur, each of the first two reactions must occur
twice. The effect of the reactions is
4H 1 -+ He 4 + 213+ + 2-Y + 211, (20--46)
with a total enefgy release of 26.7 Mev, as calculated previously; when

the kinetic energy of the neutrinos is subtracted, the energy is 26.2 Mev.
20-8] ENERGY PRODUCI'ION IN STARS 669
The positrons emitted are annihilated by free electrons with the production
of i'-rays.
Another proton-proton chain consists of the reactions
Hl + Hl --+ H2 + p+ + v,
H 1 + H 2 --+ He 3 + 'Y, (2o--47a)
7
He 3 + He" --+ Be + 'Y.
Be 7 + p- --+ Li 7 + v + 'Y,
(2o--47b)
Li 7 + H 1 --+ He"+ He 4 •
Be 7 + H 1 --+ B8 + 'Y,
B 8
--+ Be 8
+ p+ + v, (2o--47c)
8 4 4
Be --+ He + He •
It is thought 05> that the chain (2o--45) is important at lower tempcratures,

corresponding to those in the sun when it was first formed, and that the
chain (2o--47) is more important in the present state of the sun, with its
: higher central temperature and larger He 4 concentration.
The production of one helium nucleus from four protons is an example
of a process called jusion in which a heavier element is built up from one
or more lighter elements. When this process occurs among the light ele-
ments, energy is usually released because the mass of the product nucleus
is less than the sum of the masses of the nuclei which are fused. This
property of the light elements is shown by the binding-energy curve,
Fig. 9-11, in which the binding energy rises rapidly with mass number in
the region of small mass numbers. Under appropriate conditions, reac-
tions such as those discussed can liberate vast amounts of energy, amounts
much greater than that released in an atom-bomb explosion, and thermo-
nuclear reactions form the basis of the so-called thermonuclear or hydrogen
bomb.
Another set of reactions, the well-known carbon-nitrogen cycle was
proposed by Bethe 06> to account for energy production in the sun and
other stars of the main sequence. Bethe showed that reactions starting
with Li, Be, B, C, N, and F have mean reaction times less than 109 years
and had to be ruled out, while the calculated rates of energy production
involving O, Ne, Mg, and other light elements are too low. But the re-
actions involving carbon and nitrogen were found to have the remarkable
property that they could be formed into a cycle in which the carbon and
nitrogen nuclei are not used up, but are regenerated. These nuclei act
as catalysts in a series of reactions in which 4 protons are converted into

a helium nucleus and about 26 Mev of energy are liberated. The sequence
of reactions is
c12 + H1 _ N13 + "Y,
N13 _ c13 + fJ+ + v,
c13 + H1 - Nu + "Y,
(20-48)
Nl4 + Hl ----+ 015 + "Y,
o 15 ----+ N 15 + fJ+ + v,
Nl5 + Hl----+ c12 + He4.
The mean lifetime of hydrogen in the cycle as a whole varíes rapidly with
the temperature but is in the correct range for the sun and other main
sequence stars.
For sorne years it was thought that the C-N cycle was responsible for
nearly ali of the solar energy production but, because of recently obtained
nuclear data, the proton chain is now considered to be more important in
the sun than the carbon cycle. It is thought that the carbon cycle produces
more energy in main sequence stars which are much more luminous than
the sun and whose central temperatures are higher, while the proton chain
is more important for main sequence stars less luminous than the sun.
The application of nuclear physics to the problem of energy production
in stars depends on nuclear data and on astrophysical data and calcula-
tions. The latter have to do with such matters as structure, temperature
distribution, chemical composition, and density. Knowledge of the nuclear
reactions has reached the stage where the nuclear calculations are probably
more reliable than the astrophysical calculations. lf the energy produc-
tion is calculated for different stellar models and the results compared
with observations, deductions can be made about the astrophysical
aspects of stars, and sorne of the present uncertainties about stellar models
should be removed in the near future.
Nuclear reactions have been suggested for stars which differ greatly
from the sun. It is probable that there are stars which were like the sun
but had such a high luminosity and rate of conversion of hydrogen into
helium that they have already exhausted their hydrogen supply. A star
of this kind would be expected to undergo gravitational contraction until
its central density and temperature become very great (T ~ 2 X 10 8 ºK).
The following reactions could then occur:
He 4 + He 4 + 95 kev ----+ Be 8 ,
(20-49)
He 4 + Be 8 ----+ C 12 + 7.4 Mev.
20-9] CONTROLLED THERMONUCLEAR REACTIONS 671
When a C 12 nucleus has been produced, it can undergo another (a, 'Y)
reaction, giving an 0 16 nucleus.and about 7 Mev, and so on. In this way,
fusion reactions among nuclei heavier than helium may produce energy
in sorne stars.
20-9 Controlled thermonuclear reactions. The fact that nuclear fusion

reactions can release great amounts of energy, as in stars, has led to the
search for practical means for the controlled release of such energy. The
reactions of the proton-proton chain (20-45) occur extremely slowly,
which is fortunate because the sun can then have a reasonably long life.
There are thermonuclear reactions which occur much more rapidly and
depend on an abundant material, deuterium. The commercial separation
of deuterium (H 2 or D) from hydrogen permits the use of the following
reactions:
H + H2 -
2
H3 + H 1 + 4 Mev, (20--50)
H +H2 2
- He + n + 3.3Mev,
3
(20-51)
H + H3 -
2
He4 + n + 17.6Mev, (20-52)
H 2 + He 3 - He 4 + H 1 + 18.3 Mev. (20-53)
The D-D reaction cango in two equally probable ways, the first of which
produces tritium (H 3 or T) while the second produces He 3 • If the D-D
reactions take place in a chamber, and if the products remain in the
chamber, deuterium can react with tritium, Eq. (20-52), to give an
a-particle and a neutron, or with He 3 to give an a-particle and a proton.
The net result of the reactions is the burning of six deuterium nuclei and
the formation of two helium nuclei, two neutrons, and two protons. The
detailed energy balance shows that of the 17.6 Mev of kinetic energy re-
leased in the reaction (20--52), 14.1 Mev are carried off by the neutron
and 3.5 Mev by the a-particle; of the 18.3 Mev released in the reaction
(20--53), 14.7 Mev are carried off by the proton and 3.6 by the a-particle.
The total energy output per deuteron burned is approximately 7.2 Mev of
which the neutrons carry off 2.76 Mev.
The energy released in the above fusion reactions is much less than that
released in a fission reaction but the energy yield per unit mass of material,
e.g., per gram, is slightly greater from fusion than from fission. Deuterium
occurs in nature with an abundance of about one part in six thousand of
hydrogen, and can be separated from the lighter isotope quite cheaply.
One gallon of water contains about one-eighth of a gram of deuterium, and
the cost of its separation is less than 4 cents; but its energy content if it
could be burned as a fuel in a thermonuclear reactor would be the equiva-
lent of about 300 gallons of gasoline. The total amount of deuterium in the
672 ~UCLEAR ENERGY SOURCES [CHAP, 20
oceans is estimated to be about 10 17 lb, and the energy content of this

fuel would be 10 20 kilowatt-years, so that the controlled use of thermo-
nuclear energy might provide a practically infinite source of energy with
a fuel of negligible cost. There are, however, sorne difficult problems to be
solved, and sorne of these will be discussed briefly.
The nuclei which interact in the fusion processes are positively charged
and, therefore, repel one another. They must be made to collide with
enough relative velocity to overcome the Coulomb barrier which tends to
keep them apart. It can be shown that this can occur when the interacting
particles have energies of about 0.1 Mev or more. Since the energy released
in the reactions is between 3 Mev and 18 Mev, the net energy gain in
any one fusion process is appreciable and most of the gain appears in the
kinetic energies of the neutron and-proton. The problem that must be
solved is that of making the reactions go in such a way that useful energy
can actually be obtained. Thus, the nuclear reactions can be produced
in the laboratory by accelerating deuterons or tritons on to a target;
but this process will not yield a net gain of energy because the energy
is lost in ionizing the target material. To produce a net gain, the target
must be ionized before the reactions occur. But, if a beam of deuterons
is incident on an ionized target, the probability that the deuterons will
undergo elastic scattering is much greater than the probability of fusion.
To achieve fusion, the deuterium nuclei (or deuterium and tritium nuclei)
must be confined somehow in a region where they can undergo many
collisions, enough so that fusion eventually occurs. It is necessary,
therefore, for the nuclei to be in a container and moving at high kinetic
energies with respect to one another; in other words, the particles must
be in a container at very high temperature. Sorne idea of the temperatures
required can be obtained from Fig. 20-4 which shows the cross sections
for the D-D and D-T reactions as functions of the deuteron energy. The
probability of fusion is very small at low energies but rises rapidly as the
energy of the interacting particles increases toward 100 kev. The relation
E = kT implies that 1 ev of energy corresponds to a temperature of
1.16 X 104 ºK, and 10 kev corresponds roughly to 10 8 ºK. Consequently,
the interacting particles must be contained under conditions at which
their thermal energies correspond to temperatures of the order of 10 8 ºK.
At the temperatures required for fusion, the atoms are entirely stripped
of their electrons. The result is a completely ionized gas, or plasma
consisting of atomic nuclei (deuterons, tritons, protons, etc.,) and electrons
in rapid random motion. No wall of ordinary material can, of course,
contain such a plasma. In the sun, the proton fusion reactions are con-
tained by a tremendous gravitational pressure. Such a high pressure is
not available in experiments on controlled thermonuclear reactions, but
the plasma can be contained in a magnetic field. The problem then is to
20-9] CONTROLLED THERMONUCLEAR REACTIONS 673
10-24
!. 10-25
d
·B
"~
g
d
10-26
o
·~
...
10-21
Deuteron energy (kev)
FIG. 20-4. Fusion cross section as a function of deuteron energy. (From

A. S. Bishop1 Project Sherwood-The U.S. Program in Controlled Fusion. Reading,
Mass.: Addison-Wesley, 1958.)
devise an apparatus in which the plasma can be contained by means of a

magnetic field at the temperature required for the fusion reactions to
proceed. The plasma pressure must be kept from becoming too high,
which requires that the density of the plasma be low, in the range from
10 14 to 10 18 particles/cm 3 •
Another requirement for the controlled release of useful thermonuclear
energy is that the reaction be self-sustaining; once the plasma temperature
has reached the point at which fusion occurs at an appreciable rate, the
energy generated in the plasma must be at least high enough to maintain
that temperature. This energy includes the energy of the protons and
a-particles produced by the reactions (since these particles are contained
by the magnetic field), but does not include the kinetic energy of the
neutrons which escape from the magnetic container. The plasma loses
energy by radiation, for example, by bremsstrahlung resulting chiefiy from
collisions between electrons and ions. The rates of both power generation
and power loss by radiation increase with temperature; but the generation
rate increases faster with the result that a certain critica! temperature will
exist above which the reactions should be self-sustaining. According to
theory, in the case of the D-D reaction, this so-called ignition temperature
is about 4 X 10 8 ºK; for the D-T reaction it is about 4.5 X 10 7 ºK.
The cross sections shown in Fig. 20--4 and the ignition temperatures
just cited indicate that the D-T reaction is more favorable than the D-D
reaction. Although deuterium is plentiful in nature, tritium does not occur
naturally. But tritium can be made, for example, by bombarding lithium
with slow neutrons in a reactor,
Li 6 + n -+ H 3 + He 4 + 4.8 Mev. (20-54)
This reaction can serve two useful purposes. In a thermonuclear reactor

burning a mixture of deuterium and tritium, the neutrons which escape
from the container carry off much of the energy released. This energy
may be converted to heat by slowing-down the neutrons in a bl,anket sur-
rounding the reactor in which the plasma is contained. The slow neutrons
could then be captured in lithium, producing tritium in the process. Thus
the tritium would be recovered, and deuterium and lithium consumed.
The blanket might consist of moderator, coolant, and lithium. The heat
generated by the moderation and absorption of the neutrons could then
be transferred by the coolant to external heat exchangers and equipment
for generating electricity.
The problem of producing and maintaining deuterium plasmas at the
temperatures required is an extremely difficult one, and the major effort
in the field of thermonuclear research is the investigation of the behavior
of totally ionized gases {plasmas) at ultra-high temperatures. Although
the problem of confining a plasma and heating it to thermonuclear tem-
peratures has not been solved, extensive experimental and theoretical
research programs are under way. Details can be found in the general
references listed at the end of the chapter.
The necessarily brief discussion of thermonuclear reactions has indicated
sorne of the ways in which nuclear physics can be applied to astrophysics
and fusion. Another interesting application, which cannot be treated here,
is in the problem of the origin and abundance of the elements. o 7>
675
REFERENCES
GENERAL
FISSION
H. D. SMYTH, Atomic Energy for Military Purposes. Princeton: Princeton
University Press, 1945; also, Revs. Mod. Phys. 17, 351 (1945).
A. M. WEINBERG and E. P. WIGNER, The Physical Theory of Neutron Chain
Reactors. Chicago: University of Chicago Press, 1958.
S. E. LIVERHANT, Elementary Introduction to Nuclear Reactor Physics. New
York: Wiley, 1960.
S. GLASSTONE and M. C. EoLUND, The Elements o/ Nuclear Reactor Theory.
New York: Van Nostrand, 1952.
R. L. MuRRAY, Nuclear Reactor Physics. Englewood Cliffs, New Jersey:
Prentice-Hall, 1957.
D. J. L1TTLER and J. F. RAFFLE, An Introduction to Nuclear Reactor Physics,
2nd ed. London: Pergamon Press, 1957.
R. L. MuRRAY, Introduction to Nuclear Engineering. Englewood Cliffs, New
Jersey: Prentice-Hall, 1954.
R. STEPHENSON, Introduction to Nuclear Engineering, 2nd ed. New York:
McGraw-Hill, 1958.
R. V. MEGHREBLIAN and D. K. HoLMES, Reactor Analysi8. New York:
McGraw-Hill, 1960.
C. F. BONILLA, ed., Nuclear Engineering. New York: McGraw-Hill, 1957.
H. ETHERINGTON, ed., Nuclear Engineering Handbook. New York: McGraw-
Hill, 1958.
S. GLASSTONE, Principles of Nuclear Reactor Engineering. New York: Van
Nostrand, 1955.
J. R. DIETRICH and W. H. ZINN, eds., Solid Fuel Reactors. Reading, Mass.:
Addison-Wesley, 1958.
J. A. LANE, H. G. MAcPHERBON, and F. MASLAN, eds., Fluid Fuel Reactors.
Reading, Mass.: Addison-Wesley, 1958.
Proceedings of the International Conference on the Peaceful Uses of Atomic
Energy, Geneva, 1955. New York: United Nations, 1956.
Vol. 2. Physics; Research Reactors.
Vol. 3. Power Reactors.
Vol. 5. Physics of Reactor Design.
Proceedings of the Second International Conference on the Peaceful Uses of
Atomic Energy, Geneva, 1958. Geneva: United Nations, 1958.
Vol. 8. Nuclear Power Plants, Part l.
Vol. 9. Nuclear Power Plants, Part 2.
Vol. 10. Research Reactors.
Vol. 12. Reactor Physics.
Vol. 16. Nuclear Data and Reactor Theory.
FUSION
R. F. PosT, "Controlled Fusion Research-An Application of the Physics of
High Temperature Plasmas," Revs. Mod. Phys. 28, 338 (1956).
R. F. PosT, "High-temperature Plasma Research and Controlled Fusion,"

Ann. Rev. Nuc. Sci. 9, 367 (1959).
A. S. BisHOP, Project Sherwood-The U.S. Program in Controlled Fusion.
Reading, Mass.: Addison-Wesley, 1958.
A. S1MoN, An Introduction to Thermonuclear Research. London: Pergamon
Press, 1959.
D. J. RosE and M. CLARK, JR., Plasmas and Controlled Fusion. Cambridge,
Mass.: Technology Press of M.I.T.; New York: Wiley, 1961.
Proceedings of the Second International Conference on the Peaceful Uses of
Atomic Energy, Geneva, 1958. Geneva: United Nations, 1958.
Vol. 31. Theoretical and Experimental Aspects of Controlled Nuclear Fusion.
Vol. 32. Controlled Fusion Devices.
W. P. ALLis, ed., Nuclear Fusion. New York: Van Nostrand, 1960.
C. LoNGMIRE, J. L. TucK, and W. B. THOMPSON, eds., Progress in Nuclear
Energy, Series XI. Plasma Physics and Thermonuclear Research, Vol. l. London:
Pergamon Press, 1959.
PARTICULAR
l. H. S. lsBIN, "Catalogue of Nuclear Reactors," Proceedings of the Second

International Conference on the Peaceful Uses of Atomic Energy, Geneva, 1958,
op. cit. gen. ref., 8, 561 (1958).
2. A.M. WEINBERG and E. P. WIGNER, op. cit. gen. ref., p. 312.
3. "Research Reactors," U.S.A.E.C. Report T.I.D. 5B75. Washington, D.C.:
U.S. Government Printing Office, 1955; New York: McGraw-Hill, 1955, Chap. l.
4. l. KAPLAN and J. CHERNICK, "Uranium-Graphite Lattices-The Brookhaven
Reactor," Paper No. 606, Proceedings of the lntemational Conference on the
Peaceful Uses of Atomic Energy, Geneva 1955, op. cit. gen. ref., 5, 295 (1956).
5. J. HoROWITZ, "Studies of the Neutronics of Two French Heavy Water
Reactors," Paper No. 361, Proceedings of the lntemational Conference on the
Peaceful Uses of Atomic Energy, Geneva 1955, op. cit. gen. ref., 5, 256 (1956).
6. KouTs, PRICE, DowNES, SHER, and W ALSH, "Exponential Experimenta
with Slightly Enriched Uranium Rods in Ordinary Water," Paper No. 600,
Proceedings of the lnternational Conference on the Peaceful Uses of Atomic Energy,
Geneva 1955, op. cit. gen. ref., 5, 183 (1956).
7. R. L. MuRRAY, Nuclear Reactor Physics, op. cit. gen. ref., Chapter 3.
8. J. A. LANE, "Econoinics of Nuclear Power," Ann. Rev. Nuc. Sci. 9, 473
(1959).
9. F. T. MILES and l. KAPLAN, "Optimizing and Comparing Reactor De-
signs," Chemical Engineering Progress Symposium Series, Vol. 50, No. 11, p. 159
(1954).
10. J. R. D1ETRICH and W. H. ZINN, eds., op. cit. gen. ref., Chapters 2, 3, 4.
11. LANE, MAcPHERSON, and MAsLAN, eds., op. cit. gen. ref., Parts I, 111.
12. H. E. SuEss and H. C. UREY, "Abundances of the Elements," Revs. Mod.
Phys. 28, 53 (1956).
13. H. A. BETHE and C. L. CRITCHFIELD, "The Formation of Deuterons by
Proton Combination," Phys. Rev. 54, 248 (1938).
PROBLEMS 677
14. E. E. SALPETER, "Nuclear Reactions in the Stars. l. Proton-Proton
Chain," Phys. Rev. 88, 547 (1952).
15. A. G. W. CAMERON, "Nuclear Astrophysics," Ann. Rev. Nuc. Sci. 8, 299
(1958).
16. H. A. BETHE, "Energy Production in Stars," Phys. Rev. SS, 434 (1939).
17. BuRBIDGE, BuRBIDGE, FoWLER, and HoYLE, "Synthesis of the Elements
in Stars," Revs. Mod. Phys. 29, 547 (1957).
PROBLEMS
l. A thermal nuclear reactor containing a mixture of U 235 and U 238 operates

ata power leve! of 1000 megawatts. At what rate is U 235 consumed by fission?
What is the total rate of consumption of U235 ? What is the maximum rate at
which Pu 239 might be produced if ali neutrons released in fission over those
required to maintain fission were absorbed by U238 ?
2. What is the smallest value of the ratio of the number of U235 atoms to
moderator atoms, in a homogeneous mixture of uranium and graphite, for which
a thermal neutron chain reaction will be possible in an infinitely large assembly?
3. What is the smallest value of the ratio of the number of U 235 atoms to
moderator atoms, in a homogeneous mixture of U 235 and beryllium, for which a
thermal neutron chain rcaction will be possible in an infinitely large assembly?
water molecules in a homogeneous mixture of U235 and water for which a ther-
mal neutron chain reaction will be possible in an infinitely large system?
D20 molecules in a homogeneous mixture of U 235 and D20 for which a thermal
neutron chain reaction will be possible in an infinitely large system?
6. The migration area in a homogeneous mixture of uranium and moderator
may be written as
where r, the "age" of the thermal neutrons, is the square of the slowing-down
length in the moderator, and LM is the thermal neutron diffusion length in the
moderator. From the known values of the cross section, the diffusion lengths,
and the slowing-down lengths, calculate the migration area corresponding to the
solution to each of Problems 2, 3, 4, and 5.
7. Calculate the ratio of U 235 atoms to moderator atoms needed to make
k.,, = 1.10 in each of the cases of Problems 2, 3, 4, and 5. Then finá the corre-
sponding values of the migration area, critica! radius, and critica! mass of a
spherical reactor in each case.
8. If no is the value of the neutron density at the central plane of an infinite
slab reactor and iJ is the average value, show that
2
iJ = - no.
11'
9. Derive the differential equation for the neutron density in a spherical reactor
on the modified one-group model. Then show that the neutron density is given by
sin TT/R
()
nr = A ----,
r
where A is an e.rbitrary constant. Eve.luate A in terms of no, the neutron density
at the center of the sphere, and calcula.te the ve.Iue of the ratio ff/no. [Hint: See
Problem 9, Chapter 18.]
10. It can be shown the.t the neutron density in a. reactor which is a rectangu-
lar parallelopiped is given by
rx ,ry rz
n(x1y1z) = no cos - cos -b cos - ,
a e
where a, b, e a.re the sides, and the center of the reactor is a.t (O, O, O). Find the
ratio of the average neutron density to the maximum neutron density. Compare
the result with tha.t for a. spherica.l reactor.
11. Ca.lculate the Q-value of each rea.ction in the carbon-nitrogen cycle and
find, from the results, the amount of energy, in Mev, liberated in the formation
of one helium a.toro.
CHAPTER 21
THE ACCELERATION OF CHARGED PARTICLES
The development of machines for the acceleration of charged particles

is closely related to the advance of nuclear physics and is an examp_le of
the interplay between the invention of new instruments and the progress
of physical science. Inventions like the cloud chamber, the Geiger-Mueller
counter, the cyclotron, and the proton synchrotron extend the range of phe-
nomena which can be studied quantitatively, and the new information
obtained leads to new developments in theory. The theoretical advances
lead, in turn, to predictions about phenomena which líe outside the
available range of experiment and inspire improvements in existing instru-
ments as well as the invention of new ones.
The design and construction of accelerators depend strongly on elec-
trical and mechanical engineering, and are not, in themselves, a part of
nuclear physics. But accelerators are intimately connected with nuclear
physics and are far enough fr.om ordinary industrial engineering so that
they are usually regarded as a part of nuclear physics. The subject of
particle accelerators will therefore be discussed briefly, with the object of
presenting sorne of the basic ideas underlying the design of accelerators
rather than a detailed treatment of accelerator design.
21-1 The Cockcroft-Walton macbine. Charged particles. can be accel-

erated by applying a steady potential between a sourre of particles and a
suitable electrode. A hot filament can be used as a source of electrons,
while a gaseous discharge tube containing hydrogen, deuterium, or helium
can supply protons, deuterons, or a-particles, respectively. When a dis-
charge tube is used, the particles can emerge from a small hole in the tube.
The ion source is placed at one end of an evacuated accelerating tube and
the target or exit hole for the beam is at the other end. The problem is to
make the potential across the accelerating tube as large as feasible. One
of the simplest and cheapest methods of getting a potential which can
accelerate charged particles to energies great enough to cause nuclear
transmutations was devised by Cockcroft and Walton, 
Cockcroft and Walton used a condenser-rectifier voltage multiplier
based on a principle illustrated in Fig. 21-1. Three condensers, Ki, K 2 ,
679
680 THE ACCELERATION OF CHARGED PARTICLES [CHAP. 21
and Ka, each of capacity C, are connected in series, and condenser K 3 is

connected to a source of steady potential E. If two other condensers
X 1 and X 2 are connected to condensers K 1 , K 2 , and Ka, first as shown by
the dotted lines S 1 , S 2 , and Sa, and then as shown by the full lines S~,
S~, Si, then in the first cycle, when X 1 and X 2 are connected to K 2 and
Ka, condenser X 2 is charged to voltage E. When the switches are moved
to the upper position, condenser X 2 shares its charge with condenser K 2
and both are charged to E/2 if they have equal capacity. In the next
reversa} of the switches, condensers K 2 and X 1 are connected and take up
potentials E/4, while condenser X 2 is recharged to potential E. The
process can be continued and charge can be transferred gradually to all
the condensers until, in the absence of loss, a potential 3E is developed
across the condensers K 1 , K 2 , Ka in series. By adding more condensers,
large multiples of a given steady voltage can be obtained.
3E
K1
X1
3R K2
E X2
1
1
Supply Ka
1
1
FIG. 21-1. The principie of the Cockcroft-Walton condenser-rectifier voltage

multiplier. m
Although the system just described shows the idea underlying the
Cockcroft-Walton machine, the actual apparatus they used was more
complicated. An alternating voltage from a transformer was applied and
the switching action was accomplished by the use of vacuum-tube recti-
fiers. By using their voltage multiplier in conjunction with an accelerating
tube, Cockcroft and Walton succeeded in getting protons with energies
up to 800 kev in their early apparatus. Particle energies greater than
1 Mev have been obtained with later forms of the apparatus.
Tbe Cockcroft-Walton machine is a relatively simple one, with no
moving parts; the maximum energies which can be obtained are low coro-
pared with those from other accelerators, but it provides fairly large ion
currents at constant voltage, and is very useful for experimental work at
moderate particle energies, especially below 1 Mev. Asan example, the
H 2 (d, n)H 3 reaction, in which ''heavy ice" is bombarded with deuterons
accelerated in a machine of this kind, is a good source of fast neutrons
21-2) THE ELECTROSTATIC GENERATOR 681
because good yields can be obtained with bombarding energies as low as

200 kev.
Newer types of high voltage rectifier equipment have also been devel-
opedc3> because of the continued usefulness of relatively low-energy, but
simple, machines.
21-2 The electrostatic generator or Van de Graaff machine. To in-

crease the probability of nuclear transmutations and the yield of arti-
ficial nuclides, it is necessary that the charged particles used as the pro-
jectiles be available in adequate amounts and with enough energy to
penetrate the nucleus. The quantitative study of nuclear reactions re-
quires, in addition, that the particles be homogeneous in energy, and that
they emerge from the apparatus in which they are produced in a parallel
beam and with little accompanying stray radiation. Furthermore, it
should be possible to measure the energy of the particles accurately and
to vary it over a wide range. These requirements can be met, in the energy
range up to about 10 Mev, by the combination of an electrostatic generator,
or Van de Graaff machine, and an accelerator tube.
The design of the Van de Graaff generator< 4 • 5 ,o> is based on the prin-
cipie that if a charged conductor is brought into interna! contact with a
hollow second conductor, ali of its charge transfers to the hollow conductor
no matter how high the potential of the latter may be. m Except for insu-
lation difficulties, the charge, and hence the potential, of a hollow con-
ductor could be raised to any desired value by successively adding charges
to it by internal contact. The conductor must be supported in sorne way,
and its maximum potential is limited to that at which the rate of leakage
of charge from it through its supports or through the surrounding air
equals the rate at which charge is delivered to it.
A schematic diagram of a Van de Graaff generator operated to accelerate
positive ions is shown in Fig. 21-2. One terminal, which consists of a
number of sharp points projecting from a horizontal rod, is maintained
at a positive potential of about 5000 to 20,000 volts relative to ground.
This terminal is indicated in the diagram by the spray points. Because
of the large elcctrostatic field in the air near the points, ions (positive and
negative) are formed, and a spray of positive ions is repelled from the
points. These positive ions attach themselves to the surface of•a flexible
belt made of paper, silk, rayon, or sorne other nonconducting material.
The belt passes over pulleys, the lower pulley being driven by a motor,
while the upper pulley is an idler. The positive ions are carried upward
by the belt toward the metal shell that forms the upper terminal. Another
set of points draws off the charge and transfers it to the shell. Negative
ions are carried downward by the belt and migrate to the spray points.
The charge-transfer process is repeated, and the potential of the shell rises
High-voltage terminal
Charge remover points

Positive ion source
Charge conveyor belt Pressure tank
Acceleration tube
Driving motor
Controllable
spray voltage
Ground plane
Spray points
High-vacuum pump Target
Fio. 21-2. Schematic diagram of a Van de Graaff generator operated to

accelerate positive ions.
until the rate of loss of charge by leakage equals the rate at which charge
is introduced. The leakage can be minimized and the maximum attainable
voltage increased by enclosing the apparatus in a gas-tight steel chamber
and operating at pressures up to about 15 atmospheres. The gas in the
chamber may be air, nitrogen, methane, or freon.
A gas discharge tube, which is the source of the positive ions used as
projectiles, is inside the shell and the ions are accelerated down the tube
21-3] THE CYCLOTRON 683
onto the target located at the low-voltage end of the tube. The accelerat-
ing tubes are usually made up of sections of glass, porcelain, or other
insulating material joined end to end by vacuum-tight seals, and must
be long enough to avoid the possibility of a spark or other discharge pass-
ing from one end to the other when the potential is applied.
In the first Van de Graaff machine the maximum voltage attained was
1.5 Mev. In later machines< 3 ,s, 9 > this voltage was increased to 5 Mev,
and the newest machines should reach still higher energies, possibly
20 Mev. <45 > These machines give good supplies of charged particles, at
voltages which can be held constant within about 0.1 %- Proton beams
from these machines have been used to study proton-proton scattering,
to study nuclear reactions, to make new nuclides, and to produce beams
of monoenergetic neutrons. The Van de Graaff generator can also be used
to accelerate electrons.<10>
21-3 The cyclotron. The methods discussed so far for obtaining ener-
getic charged particles depend on the direct use of high voltages and are
subject to certain practica} limitations. The experimental problems
increase rapidly with increasing voltage because there are difficulties of
insulation and corona discharge. Lawrence and Livingston° 0 developed
a method that avoids these difÍiculties by means of the multiple accelera-
tion of ions to high speeds without the use of high voltage. Their first
instrument was the earliest version of the magnetic resonance accelerator
or cyclotron, which is the most famous of ali accelerators.<1 2 • 13 • 14 >
The cyclotron consists of two flat, semicircular metal boxes, called
"dees" or "D's" because of their shape. These hollow chambers have their
diametric edges parallel and slightly separated from each other, as shown
in Fig. 21-3. A source of ions is located near the midpoint of the gap
between the <lees. The <lees are connected to the terminals of a radio-
frequency oscillator so that a high-frequency alternating potential of
several million cycles per second is applied between the dees, which act
as electrodes. In this way, the potential between the dees is made to
alternate rapidly, sorne millions of times per second, and the electric field
in the gap is directed first toward one dee and then toward the other. The
space within each dee is, however, a region of zero electric field because
of the electrical shielding effect of the dees. The dees are enclosed within,
but insulated from, a larger metal box containing gas at low pressure, and
the whole apparatus is placed between the poles of a strong electromagnet
which provides a magnetic field perpendicular to the plane of the dees.
When an ion· with a positive charge q is emitted from the source, it is
accelerated by the electric field between the <lees toward the dee which is
negative at that instant. Since there is a uniform magnetic field H acting
at right angles to the plane of the dees, the ion travels in a circular path
To vacuum pump
Frn. 21-3. The accelerating chamber in a cyclotron.
of radius ri, given by

mv1
r1 = Hq, (21-1)
where m is the mass of the ion, H is the strength of the magnetic field,
and v 1 is the speed of the ion. While the ion is inside the dee, its speed
stays constant, but after describing a semicircle through the dee, the ion
reaches the gap, where it is again subject to the action of the applied
potential difference. If now, in the time needed for the ion to complete
the semicircle, the electric field has reversed so that its direction is toward
the second dee, the ion is again accelerated as it crosses the gap between
the dees and enters the second dee with an increased speed v2 • It then
moves in a semicircle of larger radius r 2 within the second dee to emerge
again into the gap.
The angular velocity of the ion is

w=':.=H.!!.., {21-2)
r m
where v is the velocity of the ion in a semicircle of radius r. It is seen

that the angular veloeity is independent of the speed of the ion and of
the radius of the circle in which it travels, depending only on the strength
of the magnetic field and the charge-to-mass ratio (q/m) of the ion. If,
therefore, the electric field reverses at regular intervals, each equal to
the time needed for the ion to make a half revolution, the field in the gap
is always in the right direction to accelerate an ion each time the gap is
crossed. In other words, if the oscillation frequency is adjusted to the
properties of the given ion and to the strength of the magnetic field, the
charged particles always keep in phase with the changes of electric potential
between the <lees. Each time the ions cross the gap they receive an addi-
tional kick, with the result that the energy of the ions is steadily increased,
and the ions describe a flat spiral of increasing radius. Eventually, the
ions reach the periphery of the dee, where they can be brought out of the
chamber by means of a deflecting plate charged to a high negative poten-
tial. The attractive force acting on the positive ions draws them out of
their spiral paths, and they can be used to bombard a properly placed
target.
The energies of the ions produced by a cyclotron can be calculated from
the equation for the motion of an ion in a magnetic field,
mv 2
Hqv=R, {21-3)
where R is the radius at which the particles leave the machine, and H and
q are expressed in electromagnetic units. After an ion has been accelerated,
its energy is given by
{21-4)
where V is the equivalent potential difference (also in emu) through which
the ion has been accelerated When v is eliminated from Eq. (21-4) by
inserting its value from Eq. (21-3), the result is
V = !.2 H 2 R 2 .!!..
m
· (21-5)
If q is in esu and V in volts, then
{21-6)
since e is the velocity of light. The value of q/m for a proton is 2.87 X 10 14
esu/gm, and, since H and r are known from the design, the equivalent
voltage of the particles can be calculated.
The value of the magnetic field strength cannot be chosen arbitrarily
but depends on the frequency of the voltage applied to the dees, and must
be chosen so as to give resonance. For resonance, the time taken for
an ion, traveling with an angular velocity w, to describe a semicircle within
one of the dees must be equal to half the time period T of oscillation of
the applied high frequency voltage; that is,
1r T
W = 2'
and the frequency n is
1 w H q
n=-=-=--· (21-7)
T 21r 21r m
For ions with a given value of q/m, the magnetic field needed to give
resonance for any given frequency of the applied voltage is given by
Eq. (21-7). The higher the frequency, the stronger must be the magnetic
field, and therefore from Eq. (21-6), the greater will be the equivalent
voltage through which the ions will have been accelerated at any given
radius. With a radiofrequency voltage of 10 7 cycles/sec, the magnetic
field must be about 6500 gauss for protons and 13,000 gauss for deuterons.
When these values are inserted into Eq. (21-6), and if the radius is taken
to be 30 cm, the resulting values of V are 1.8 million volts for protons and
3.6 million volts for deuterons. If the radius is doubled, the equivalent
voltage through which the ions are accelerated is raised by a factor of four.
The voltage applied to the dees does not appear in Eq. (21-6), so that
the energy which a charged particle can acquire in a particular cyclotron
is independent of this voltage. When the voltage is small, the ion makes
a large number of turns before reaching the periphery; when the voltage
is high, the number of turns is small. It is this property of the cyclotron
that made it possible to accelerate charged particles to relatively high
energies by means of small applied voltages.
Cyclotrons are usually described in terms of the diameter of the pole
faces of the magnet. The first machine built by Lawrence and Livingston
had a magnet with pole faces 11 in. in diameter and produced 1.2-Mev
protons. A 60-in cyclotron can produce protons of about 10 Mev, deu-
terons of about 20 Mev, and a-particles of about 40 Mev. o 5> The particle
beams are strong, but the voltage is neither as constant nor as uniform
as it is for the Van de Graaff generator. The cyclotron is a powerful too!
for studies requiring particles of high energy but where it is not essential
to know the exact value of energy. The use of particles from the cyclotron
to produce beams of neutrons for experiments with a time-of-flight neutron
(a)
(b)
Frn. 21-4. The cyclotron at the Massachusetts Institute of Technology.

688 THE ACCELERATION OF CHARGED PART[CLES [CHAP. 21
spectrometer is an excellent example of the way in which a cyclotron can

be employed. The high-energy particles can, of course, be used directly
in the study of nuclear reactions and in the production of radionuclides.
Figure 21-4(a) is a photograph of the cyclotron at the Massachusetts
Institute of Technology, and Fig. 21-4(b) shows the dees removed from
the gap between the poles of the electromagnet; the cover of the outer
vacuum chamber has been removed in this photograph. The bar in the
gap between the dees supports the ion source. Accelerated particles can
be brought out of the chamber through a thin foil window in the short
tube at the lower right. In this machine, R is 48 cm, H is 1.8 X 10 4 gauss,
and 18-Mev deuterons can be obtained.
The energies to which particles can be accelerated in a cyclotron are
limited by the relativistic increase of mass with velocity. The mass is
given by the formula
m = ---::==m=º====
v'I - (v2/c2)
where m 0 is the rest mass. According to Eq. (21-2), the angular velocity
of the ions becomes
qR
w=H- 2
1- -2, (21-8)
mo c
and decreases &.s the velocity increases. The frequency of ion rotation
then decreases; the ions take longer to describe their semicircular paths
than the fixed period of the oscillating electric field and they arrive too
late at the gap between the dees. Consequently, they lag in phase behind
the voltage applied to the dees until finally they are no longer accelerated.
The ion velocities at which the relativistic mass increase leads to this
effect represents a limiting size and energy for the cyclotron. If the limiting
energy is taken to be about one percent of the rest mass, the limits for
different particles are as shown in the table below. It is clear from these
values that a cyclotron operates successfully only with relatively heavy
particles and cannot be used to accelerate electrons. This difficulty does
not arise in the Van de Graaff generator, which can be used with electrons.
1% of Radius, cm, for

rest mass, Mev H = 10,000 gauss
electrons 0.005 0.02

protons 10 40
deuterons 20 80
a-particles 40 80
21-4] THE FREQUENCY-MODULATED CYCLOTRON 689
21-4 The frequency-modulated cyclotron or synchrocyclotron. The

loss of resonance caused by the relativistic increase in mass can, in prin-
cipie, be balanced in two ways, as can be seen by considering the resonance
condition, Eq. (21-7). That relationship can be written
(21-9)
It should be possible to compensate for the decrease in the ion frequency

either by increasing the field strength H at such a rate that the product
Hyl - (v 2 /c 2 ) remains constant, or by gradually decreasing the fre-
quency n of the applied electric field. It was thought for sorne years that
the experimental difficulties involved in applying these methods would
make them impractical. But it was shown theoretically by Veksler 07> and
McMillan° 8> that these methods could be applied successfully and that
this possibility depends on a property called phase stability possessed by
certain orbits in a cyclotron.
The phase stability inherent in a variable-frequency cyclotron can be
discussed by describing qualitatively the motion of the charged particles.
Suppose that an ion moves in a circular path in a uniform magnetic field,
and on each revolution crosses the gap between accelerating electrodes
to which is applied an oscillating electric field with a frequency identical
with the frequency of rotation of the ion. Under these conditions there
are certain "stationary" or stable orbits in which the ion can move. Con-
sider first ions which cross the gap at instants of time when the oscillating
electric field is just passing through the value zero. The ions are then
said to ha.ve zero phase, and this condition is indicated by the points O,
21r, 41r, and so on in Fig. 21-5; these ions will neither gain nor lose energy
and will continue to revolve at constant frequency and in the same orbit.
An ion which crosses the gap atan earlier instant such as t 1 has a posi-
tive phase; it will gain speed and energy, since the voltage is positive,
and its rotation frequency will decrease in accordance with Eq. (21-8).
o 1r
Pha.se
Fm. 21-5. Phase stability in a cyclic magnetic accelerator.
The ion will take a slightly longer time to return to the gap, as shown by
points t 2 and t 3 in subsequent accelera.tions, when the volts.ge a.cross the
gap is smaller. Eventua.lly the pa.rticle will cross the gap at zero pha.se
but with the excess energy it has received, and will continue the pha.se
shift into the decelerating part of the cycle when the volts.ge is negative.
The particle then loses energy and speed, its frequency increa.ses, and it
is returned to the zero phase. There has been a phase oscillation about
the equilibrium pha.se 8 = O, and the ion has oscillated about a "sta.tion-
ary" or equilibrium orbit.
Veksler and McMillan pointed out that by increasing the magnetic
field or reducing the frequency of the dee volts.ge, or both, the orbits of
the ions could be roa.de to expand and the energy to increa.se. If the
variation is roa.de sufficiently slowly, the phase sta.bility is preserved dur-
ing the acceleration, and the limita on the ordinary cyclotron can be
passed. Particles can be accelerated by holding the electric field frequency
fixed and varying the ma.gnetic field; in this case the roa.chine is called a
synchrotron. Protons, deuterons, and a-particles can be accelerated to
high energies by keeping the magnetic field constant and varying the
frequency of the electric field; in this case the roa.chine is called a f requency-
modulated (FM) cyclotron or synchrocyclotron. <14> The na.me &1.Jmhrotron
was suggested by McMillan beca.use the beha.vior of the roa.chine is similar
in sorne respects to that of a synchronous motor.
The first FM cyclotron was the 184-in roa.chine at the University of
California at Berkeley,us,iu> which produced 200-Mev deuterons and
400-Mev a-pa.rticles. For these particles, the frequency was modulated
from 11.5 million cycles per second at the instant of injection to 9.8 million
cycles per second when the ions reached the periphery of the dee. Protons
with energies of about 350 Mev were obtained with an oscillator fre-
quency modulated from 23 to 15.6 million cycles per second. The weight
of the ma.gnet is about 4000 tons and the energies cited were obtained
with a dee voltage of 15. kv and a ma.gnetic field of 15,000 gauss at the
gap center. A diagra.m of the Berkeley roa.chine is shown in Fig. 21--6.
There is only one dee and the oscillating potential is applied between it
and a ground connection. An ion pa.th is shown, distorted to indica.te
only a few turns; in pra.ctice the ions make about 50,000 revolutions. The
ion beam can be removed with the a.id of electrosta.tic and magnetic
deflectors.
The process of spallation <20> was discovered and investigated by bom-
barding targets of various elements with particles from this roa.chine, and
fission of bismuth and neighboring elements has been observed with 190-
Mev deuterons. Mesons ha.ve also been produced as a result of the bom-
bardment of carbon by 390-Mev a-particles or 345-Mev protons. Mesons
are fundamental particles which are related to nuclear forces, and the
investigation of nuclear reactions which produce mesons or are induced
21-41 THE FREQUENCY-MODULATED CYCLOTRON 691
Ion
source
Electro-
static
deflector
F10. 21-6. Schematic diagram of the 184-in frequency modulated cyclotron

at the University of California.
by mesons is one of the main problems of nuclear physics. Before the

development of the large accelerators, of which the 184-in FM cyclotron
was the first example, these reactions could be studied only in cosmic rays,
where they are relatively rare, but they can now be made to occur at will
and in great numbers with the aid of large accelerators.
Many synchrocyclotrons have now been built in the United States and
in Europe in order to study the nature and constitution of the atomic
nucleus, and many interesting and important results have already been
obtained. It has been found, for example, that µ-mesons which have a
rest mass of close to 200 electron masses can be used to study the structure
of the nucleus.< 21 •22 > Beams of these particles from the Columbia Uni-
versity 164-in synchrocyclotron have been shown to interact with nuclei
in a way which is sensitive to the nuclear charge distribution. Experi-
ments indicate that the nuclear charge is distributed in a uniform sphere
of radius 1.2 X 10- 13 A 113 cm, a value which seems to differ significantly
from earlier values. <23 > Results such as these are leading to a closer exam-
ination of the distribution of particles in the nucleus, as well as of other
nuclear properties (cf. Section 17-4).
692 THE ACCELERATION OF CHARGED PARTICLES [CHAP, 21
21-5 The acceleration of electrons. The betatron and the electron

synchrotron. The emphasis so far in this chapter has been on the accelera-
tion of positively charged particles because of their importance in nuclear
reactions. Beams of energetic electrons are also needed for various pur-
poses, the most important of which is the production of x-rays of very
high energy. These x-rays can be used to produce nuclear reactions, such
as the ('Y, n), ('Y, p), ('Y, 2n), and ('Y, np) reactions, oras highly penetrating
radiations for the study of the properties of solids. The energetic elec-
trons can also be used directly.
The voltage multiplier (Cockcroft-Walton) and the Van de Graaff
electrostatic generator can both be used to accelerate electrons, but the
energies are limited to a few million electron volts. The ordinary cyclo-
tron cannot be used with electrons because of the large relativistic increase
of mass at low energies; an electron with an energy of 1 Mev has a speed
more than nine-tenths that of light and its mass is about 2.5 times the
rest mass. It has been mentioned that this relativistic effect can be over-
come by varying the magnetic field, and the design of the electron synchro--
tron is based on this idea. But before that instrument is discussed, it is
convenient to consider another electron accelerator, the betatron, in which
magnetic induction is used to accelerate the electrons. <24 , 14 >
The action of the betatron depends on the same principie as that of the
transformer, in which an alternating current applied to a primary coil
induces a similar current in the secondary windings. The primary current
produces an oscillating magnetic field which, in turn, induces an oscillating
potential in the secondary coi!. The betatron is a transformer in which a
cloud of electrons, located inside an annular, doughnut-shaped vacuum
chamber, takes the place of the secondary winding. The chamber is placed
within the poles of an electromagnet energized by an alternating pulsed
current, and the magnet produces a strong field in the central space, or
hole, of the doughnut. The electrons move in a circular orbit óf constant
radius within the vacuum chamber, as shown in Fig. 21-7, and they gain
energy by induction, because of the change with time of the magnetic
flux el> linking the orbit.
For this method to work, the induced accelerating field must have just
the right strength at the stable orbit to ensure that the electrons remain
in that orbit when the magnetic field in creases; that is, the flux linking
that orbit must have the right magnitude. The required relationship
between the linking flux and the magnetic field can be derived without
much trouble. The induced voltage per turn is d<l>/dt, as for a transformer,
and the electric field (voltage per unit length) is given by
1 del>
E= 21rR dt' (21-10)
21-5) THE ACCELERATION OF ELECTRONS 693
To vacuum pump Electron gun
Doughnut-shaped
vacuum chamber
Fm. 21-7. The accelerating chamber in a betatron.
where R is the radius of the atable orbit. The force on the electron is eE,
and the law of motion for the electron is
d e d<P
-(mv)=eE=--· (21-11)
dt 21rR dt
To maintain motion in a circular orbit of constant radius R, the magnetic

field H at the orbit must increase as the electron energy increases. The
momentum of the electron in the field is given by the familiar relation
mv2
Hev=R,
or
mv = eRH,
694 THE ACCELERATION' OF CHARGED PARTICLES [CHAP. 21
FIG. 21-8. The acceleration cycle in a betatron.
so that
d dH
dt (mv) = eR dt • (21-12)
Acceleration by magnetic induction at constant radius is obtained when

the time rate of change of momentum as given by the change in the link-
age flux is the same as that given by the change in the magnetic field.
In other words, the expressions on the right sides of Eqs. (21-11) and
(21-12) must be equal, with the result that
d4> 2dH d 2
dt = 21rR dt = 2 dt (1rR H). (21-13)
Equation (21-13) is the betatron condition, which states that in any

time interval the linking flux 4.> must change at a rate twice that which
would occur if the central magnetic field were uniform and equal to the
field at the orbit. This condition holds for relativistic energies as well as
in the nonrelativistic range because the law oí force as written in Eqs.
(21-11) and (21-12) satisfies the requirements of special relativity (Sec-
tion 6-6). The betatron condition makes it necessary to have a central
iron core with high flux density inside the orbit. Since the induced poten-
tial is determined by the rate oí ehange of flux, the iron core is laminated
as in a transformer, and alternating power at 60 or 180 cycles is used to
produce the varying magnetic field.
Electrons are injected into the chamber from a "gun" in which thermo-
electrons from a hot cathode are accelerated and focused by a potential
of sorne thousands of volts. The electrons are injected at ai;i instant when
the magnetic field is just rising from its zero value in the first quarter
cycle, as shown in Fig. 21-8. The increasing magnetic field induces a
21-5] THE ACCELERATION OF ELECTRONS 695
potential within the doughnut which increases the energy of the clectrons.
Since they are traveling in a magnetic field, the electrons move in a cir-
cular path, and the increasing magnetic field keeps them moving in a
stable orbit. When the field strength passes its maximum value and
starts to decrease, the direction of the induced electromagnetic force is
changed and the electrons start to slow down. This effect is avoided by
removing the electrons from their orbit when the magnetic field strength
has its maximum value. A pulse of current is sent through an auxiliary
coil and changes the magnetic field; the electrons are pulled out of their
stable orbit and either strike a target, producing x-rays, or emerge from
the apparatus.
The betatron at the General Electric Research Laboratories< 25 > pro-
duces 100-Mev electrons. The diameter of the pole face is 76 in and that
of the stable orbit is 66 in; the magnet weighs 130 tons and the maximum
magnetic field at the orbit is 4000 gauss. Electrons are injected at energies
of 30-70 kev; they travel around the doughnut about 250,000 times be-
tween injection and removal, gaining 400 ev of energy at each turn. The
100-Mev electrons have a speed which is more than 0.9999 that of light
and their mass is nearly 200 times the rest mass. Electrons with energies
of more than 300 Mev have been obtained with the betatron at the Uni-
versity of Illinois. <26 • 27 J
The betatron has the disadvantage that a large magnet is nceded to
supply the variable flux which accelerates the electrons. If a radiofre-
quency electric field could be used, the size of the magnet would be reduced.
The radiofrequency would have to be locked to the electron frequency, and
if the electrons started at low energies, the change in the radiofrequency
would have to be very large, beyond practica! limits. Elcctrons with
energies of about 2 Mev, however, have speeds within a few percent of
that of light. Beyond this energy the electron speed and its frequency of
revolution change very little, and the radiofrequency can remain prac-
tically constant. As the electron energy increases, the electrons can be
held in a stable orbit by increasing the strength of the magnetic field
proportionally to the increase in the mass of the electron. Veksler and
McMillan showed, independently, that the electron motion has the prop-
erty of phase stability, and the theory is similar to that for positively
charged particles in the FM cyclotron. Since it is not necessary to satisfy
the betatron condition, a ring-shaped magnet can be used. The instrument
based on these ideas is the electron synchrotron, a diagram of which is
shown in Fig. 21-9.
Electrons can be injected at high enough speeds so that the orbit radius,
although not constant, does not increase greatly. In practice, however,
it has turned out to be more satisfactory to accelerate electrons up to
1 or 2 Mev by betatron action with the aid of flux bars located inside the
Laminated steel
ring magnet
To vacuum pump Electron gun
Doughnut-shaped
vacuum chamber
F10. 21-9. The accelerating chamber in a synchrotron.
orbit. These bars are made of high permeability metal and do not have to
be large; they short the magnetic field at low inductions but become
saturated at high inductions, and the transition from betatron action to
synchrotron action can be made smoothly.
About 20 electron synchrotrons have either been built or are under
construction, <2 s> and electrons have been obtained with energies of more
than 300 Mev. The machine at the Massachusetts Institute of Tech-
nology has a 50-ton magnet and an rf of 46.5 megacycles. The orbit radius
is 40 in. Electrons are injected at 80 kev, accelerated to 7 Mev by beta-
tron action, and finally reach energies of 330 Mev.
The limitation on the energies to which electrons can be accelerated in
a betatron or a synchrotron is set by radiation losses. An electron loses
energy by radiation when it is accelerated, and the rate increases with
the fourth power of the energy. The maximum energy is reached when
21-6] THE PROTON SYNCHROTRON 697
the energy lost per turn by radiation is equal to the maximum practica}
acceleration energy per turn. lt is thought that the maximum electron
energy that can be reached under practical conditions is about 1000 Mev
or 1 Bev.
21-6 The proton synchrotron. Up to 1959, the highest energies to which

charged particles were accelerated were attained in proton synchrotrons.
In 1952, protons were accelerated to 3 Bev in the Cosmotron at the Brook-
haven National Laboratory; in 1954, 6-Bev protons were obtained in the
Bevatron at the Radiation Laboratory of the University of California;
in 1957, proton energies of 10 Bev were reached in a Russian proton syn-
chrotron at Dubna. Protons can be accelerated to these high energies
without losing appreciable amounts of energy by radiation because the
loss of energy is proportional to the fourth power of the ratio of total energy
to rest energy. The rest mass of a proton is nearly 2000 times that of an
electron, so that a proton would have to have an energy of 10 Bev in
order to lose as much energy by radiation as a 5-Mev electron, and the
loss by an electron of this energy is not important. Hence the acceleration
of protons will not be limited by radiation loss until energies much higher
than 10 Bev are reached.
The principies on which the proton synchrotron is designed and operated
are basically the same as those of the electron synchrotron. <14 , 29 , 3 o>
Protons revolve in an orbit of constant radius in a doughnut-shaped
vacuum chamber. A ring-shaped magnet produces a magnetic field normal
to the chamber. Protons do not reach a speed of 0.98c (the speed of a
2-Mev electron) until their energy is 4 Bev, so that the rotation frequency
for protons in an orbit of constant radius increases by a large factor during
acceleration. The frequency of the applied electric field must also increase
by a large factor and this is the only real difference, apart from size,
between the proton and electron synchrotrons. Protons, at low energy,
are injected ·in periodic pulses into the synchrotron orbit. They are ac-
celerated by an oscillating magnetic field in resonance with the motion of
the ions while the magnetic field increases to its maximum. The principie
of phase stability can be utilized, and if the applied frequency is correct,
the ions, on crossing the accelerating gap, oscillate in phase about an
equilibrium phase which provides the proper average acceleration to keep
the radius of the orbit constant. The magnet is excited periodically, and
the protons are accelerated during the time that the magnetic field is in-
creasing. When they reach maximum energy, the frequency is distorted
so that the orbit expands or contracts and the protons strike a target or
an ejection device at the periphery.
The Brookhaven Cosmotron has an orbit radius of 30 ft with a magnetic
field, at injection, of about 300 gauss and a maximum magnetic field of
21-6] THE PROTON SYNCHROTRON 699
Van de Graff
accelerator
288 magnet blocks
Targets
~-11r1,...-- ---+--+--+-- ----fifY-- and
~--~a==.,-.1a11- ejection
Pick-up electrodes
F10.
1/ 20 11 dia. diffusion and
backing pumps
21-11. Schema.tic dia.gra.m of the Cosmotron.
14,000 gauss. <31 , 32> The magnet, with a maximum diameter of 75 ft, is a
steel ring about 8 ft by 8 ft in cross section, and weighs about 2000 tons.
It is built in four quadrants separated by 10-ft gaps to allow straight sec-
tions of the vacuum chamber free from magnetic field to be used for injec-
tion, acceleration, and ejection of the ions. Pulses of protons with an
energy of a.bout 3.5 Mev accelerated in a horizontal electrostatic generator
are injected into the vacuum chamber. The acceleration interval in the
cosmotron is one second and about 800 ev are added per revolution; the
protons make about 3 million revolutions and travel more than 100,000
miles before reaching their maximum energy. The radiofrequency changes
from about 0.37 megacycle/sec to about 4 megacycles/sec, during the
acceleration period. Figure 21-10 is a picture of the Cosmotron, and
Fig. 21-11 is a diagram which indicates sorne of the main features of the
machine.
The Brookhaven machine is called Cosmotron because it makes possible
the study of sorne of the nuclear reactions that occur with particles with
energies comparable to those of the primary cosmic ray particles. These
studies are essential to an understanding of nuclear forces and nuclear
structure. The Bevatron< 33 •34 > has been used to create proton-antiproton
700 THE ACCELERATION OF CHARGED PARTICLES (CHAP. 21
pairs as well as neutron-antineutron pairs, and the creation and study of

these particles< 35 i has been a major contributíon to the understanding of
fundamental particles.
21-7 Linear accelerators. In the linear accelerator, charged particles

are accelerated by an oscillating electric field applied to a series of elec-
trodes, with an applied frequency which is in resonance with the motion
of the particles. In its earliest forro, which illustrates the principies of
this type of machine, the linear accelerator consisted of a set of cylindrical
electrodes, of increasing length, arranged as shown in Fig. 21-12. Alter-
nate cylinders, enclosed in a glass vacuum chamber, are connected to-
gether, the odd-numbered cylinders being joined to one terminal and the
even-numbered to the second terminal of a high-frequency power supply.
Ions from a discharge tube at one end move along the axis of the tubes
and are accelerated on crossing the gaps between the tu bes; they are not
accelerated inside the tubes because the potential is constant there. Sup-
pose that the ions are positively charged and moving from left to right, as
in the diagram. If the first cylinder is positive and the second negative,
the ions are accelerated and travel through the second cylinder ata speed
which is constant but greater than the speed in the first cylinder. The
second cylinder is just long enough so that when the ions reach the gap
between the second and third cylinders, the potentials are reversed. The
second cylinder is now positive and the third negative, so that the ions
are again accelerated, this time in the gap between the second and third
cylinders. The ions can be kept in phase with the reversals of potential
by making successive cylinders longer to allow for the increasing speed
of the ions, with the result that the ions gain additional energy each time
they pass from one cylinder to the next.
The early linear accelerators< 36J could be used only for heavy, slowly
moving ions such as mercury ions, because the oscillators available did
not have high enough frequencies. The intensive development oí high-
frequency methods in the field of radar during recent years has made
possible the design of linear accelerators with which protons and even
electrons can be accelerated to energies up to severa! hundred Mev. <37 • 3 s. 39>
These machines have both advantages and disadvantages as compared
with circular machines. A smaller ~ e t can be used than is needed for
~~_gpdeg;reniÍffl<J:P-< Me tp _]
Particle path
FIG. 21-12. Schematic diagram of a linear accelerator.

21-8] THE ALTERNATING-GRADIENT SYNCHROTRON 701
a circular machine producing particles of the same energy. The size and
cost of the linear machine are roughly proportional to the final energy of
the particles rather than to a higher power of the energy, as in circular
machines, so that for very high energies the linear machines may be more
economical. The particles automatically emerge from a linear accelerator
in a well-collimated beam, while ejection of the particles is one of the main
difficulties in the circular machines. The main disadvantage of the linear
accelerator is the fact that an individual particle, instead of passing
through the same alternating field again and again, and using the same
power source and accelerating gap many times, as in the circular machines,
must pass through a number of alternating fields and a number of power
sources. This multiplicity of power sources and fields requires many
pieces of high-frequency apparatus, with the resulting expense and com-
plication. A very long and elaborate tube must be built with its many
power sources nearby, and there is a difficult design problem in the adjust-
ment of the phase of the oscillating field overa great length of accelerator,
so as to ensure that the field will keep in step with the particles.
The linear accelerator can be used to obtain energies in the range above
1 Bev. In this energy range, the betatron is impractical because of the
huge magnet size and the energy loss by radiation, while the synchrotron
will probably be limited by radiation loss. This loss, which is caused
mainly by circular motion, is not serious in the linear accelerator. The
proton linear accelerator is useful at energies in the cyclotron energy range,
but seems to offer few advantages in the ranges covered by the FM cyclo-
tron and proton synchrotron.
Beams of fast electrons from a linear accelerator can be used to probe
the structure of the nucleus. The elastic scattering of electrons with
energies up to 125 and 150 Mev depends on the distribution of charge
in the nucleus, and experiments indica.te that there may be a variable
density of charge in the nucleus. <4 o, 41 >
21-8 The altemating-gradient synchrotron. The increases in the energy

to which charged particles have been accelerated have been accompanied
by increased complexity and cost of the machines needed. Although the
energies which could be reached in the proton synchrotron are much
greater, in theory, than severa! Bev, practical limits are set by the size
of the machine and by its cost. The Bevatron, which has reached 6 Bev,
has a magnet weighing over 10,000 tons, an orbit radius of 80 ft, and has
been estimated to cost 15 to 20 million dollars. A 30-Bev synchrotron
would need a magnet of about 100,000 tons, and its cost would be exorbi-
tant. Engineering studies indicate, for example, that direct scaling from
the Cosmotron would be practica! for energies up to about 20 Bev but,
for higher energies, the size and cost of the ring magnet become impractical.
It is clear that new ideas are needed in order to increase the energies to
which particles can be accelerated while staying within the bounds of
engineering feasibility and possibly available funds. Such ideas would
have to improve the efficiency of focusing the ions enough so that the size
of the magnet could be greatly reduced.
The problem of improving the focusing can be illustrated in terms of
sorne of the properties of the Brookhaven Cosmotron, in which the protons
move through a pipe 36 in by 7 in. in cross section with a magnet 8 ft
by 8 ft in cross section. The magnet is needed to keep the particles in
the desired orbit by mea.ns of corrective forces that push the particles
back when they begin to stray because of collisions with gap molecules
in the tube or because of fluctuations in the accelerating voltage orfre-
quency. Now, if the straying of the particles could be controlled so closely
that the particles stay almost exactly in a circular orbit, the pipe in which
they move could be very narrow and only a thin magnet would be needed
around it. It has recently been found that this effect can be achieved by
the use of an apparatus called the alternating gradient 81J11,Chroton or strong-
f ocusing synchrotron. <42> This apparatus depends on the principie that
alternating focusing and defocusing ion lenses can be arranged so as to
have a net focusing effect. The effect is analogous to the focusing of a
beam of light by a series of alternately converging and diverging lenses.
Thus, if a number of C-shaped magnets are arranged in a circle so that
alternate magnets face in opposite directions, the back of one C toward
the center of the circle and the back of the next toward the outside, the
arrangement should keep the particles in a stable orbit. Moreover, it
should be possible with proper design to make the focusing forces very
strong. The first machine of this type to go into operation (1959) was
that of CERN, the European Organization for Nuclear Research, near
Geneva; 50-Mev protons werc ,tecelerated to 28 Bev.< 43>
The alternating gradient synchrotron (AGS) of the Brookhaven National
Laboratory recently (1960) accelerated protons to 30 Bev.< 44 > A plan
drawing of this machine is shown in Fig. 21-13. The acceleration takes
place in three stages. Protons are obtained from hydrogen gas by strippíng
the negatively charged electrons from the gas molecules, leaving bare
positively charged protons in the ion source (A at the upper left of the
figure). A high constant voltage of 750,000 volts is maintained between
the ion source and ground by means of a Cockcroft-Walton generator.
This voltage imparts an initial impulse to the protons by accelerating them
through an evacuated ceramic tube. As they emerge, at a velocity of
0.04c (e is the velocity of light) they are directed into the linear accelerator
B. This device, called the linac, is a long cylindrical tank about 3 ft in
diameter and llO ft long containing 124 drift tubes along its axis. The
second stage of acceleration takes place while the protons travel through
21-8) THE ALTERNATING-GRADIENT SYNCHROTRON 703
Bubble chamber house
Access
tunnel
Target - - 1 - - - - H - .
building (D)
X denotes positions of
. acceleration stations ·
Orbit cen
throug
240 magn
Service building
Undergound magnet tunnel (administration, power, controls,
laboratories, machine shop, etc.)
o 100 200 400 600 800 1000
Feet
Fm. 21-13. Plan of the Brookhaven alternating gradient synchrotron (by

permission of the Brookhaven National Laboratory).
these tubes. In each successive gap between adjacent drift tubes, the par-
ticles are accelerated by an electric field which is set up by feeding high
frequency (2 X 10 8 cycles/sec) power into the tank. The strength of the
electric field is such that the protons emerge from the linear accelerator
at a velocity of one third that of light and an energy of 50 Mev. The
proton beam is guided from the linac into the synchrotron proper through
an elaborate system of debunching, deflecting, focusing, and monitoring
equipment.
The circular orbit of the synchrotron, where the third stage of the
acceleration is performed, is 842.9 ft in diameter (about one-half mile in
circumference). The main magnet, which is divided into 240 units, each
weighing sorne 16 tons, must bend the protons into a path of just this
diameter, and apply strong focusing forces which always tend to bring the
protons back toward their orbit within the vacuum chamber whenever
as
they tend to stray away from it. Thus, the particles increase in energy,
the synchrotron's magnetic field must become stronger to hold the beam
in the center of the vacuum chamber. The vacuum pipe is only 7 in wide
and 2¾ in high, a small fraction of the ring's diameter. The vacuum
chamber is maintained ata pressure of less than 10- 5 mm Hg.
Besides providing a guide field, the synchrotron magnets perform an-
other extremely important function, which embodies the alternating
gradient feature of the machine. The pole pieces of each successive pair
of magnets are shaped so that the magnetic field alternately increases and
then decreases in the radial direction. This alternation of the magnetic
field's gradient causes the circulating proton beam alternately to focus
and defocus vertically and horizontally and, after many traversals of the
magnets, yields a tightly focused beam. Thus, the magnets, in forming
a series of alternately converging and diverging lenses, present a beam
path with much greater focusing power than the corresponding effect in
a conventional constant gradient machine.
As a consequence, the beam can be contained in a much &naller volume
of space around the circumference of the accelerator. Hence, the magnets
can be much smaller with an accompanying saving in the amount of steel
and copper required. For instance, the total amount of steel in BNL's
30-Bev machine is roughly 4000 tons, compared to Russia's 10-Bev machine
of conventional design whose magnet weighs sorne 36,000 tons.
The large magnetic gradients require that the individual magnets be
very precisely built and aligned to avoid errors in the magnetic field that
would cause the beam to stray from the desired path and strike the walls
of the narrow vacuum chamber.
As the protons circulate around the ring, they are accelerated by electric
fields produced in 12 radio-frequency acceleration stations C. At these
stations, a high frequency voltage is impressed across two gaps. Protons
which cross the gaps when the electric field is in the forward direction are
accelerated. If the applied frequency is correct, the same protons will
always be accelerated at each gap. In this way, after many traversals,
very large energies can be acquired by relatively small increments. On
the average each of the 12 stations accelerates the beam by about 7,500
volts each time it passes. Therefore, the protons gain about 90,000 ev
of energy per revolution. To reach 30 Bev, sorne 325,000 revolutions
around the ring are required, a distance in the neighborhood of 160,000
miles. Toward the end of this one-second acceleration period, the protons
are traveling at a velocity within less than one-tenth of one per cent of
the speed of light, or over 186,000 miles per second. At this velocity, the
21-8] THE ALTERNATING-GRADIENT SYNCHROTRON 705
mass of the protons has been increased more than 30 times, as predicted
by the theory of relativity.
After a two second recovery period, the synchrotron is ready to pulse
again; 20 pulses of 30-Bev protons are produced each minute of full opera-
tion, with about 10 1 º protons per pulse.
The proton beam can be directed at appropriate target substances, and
the resulting reactions can be studied by means of the emitted radiations.
The primary beam can also be deflected out of the vacuum chamber and
guided into the experimental area, where a number of separate experiments
can be installed along the path of the beam and different portions used
simultaneously. Most of the experiments will be conducted in the 25,000
square foot target building D.
Most of the elementary particles produced by the interaction of the high
energy protons with the nuclei of the target atoros exist outside the atom
for only a fraction of a millionth of a second before they decay into other
particles or change their mass completely into energy. These brief life-
times, however, are long enough to detect the particles by sensitive instru-
ments (bubble chambers, counters, or photographic emulsions) and to
determine their mass, electrical charge and other properties, or to observe
the results when they impinge on secondary targets.
REFERENCES
GENERAL
E. M. McMILLAN, "Particle Accelerators," Experimental Nuclear Physics,
E. Segre, ed., Vol. III, Part XII, pp. 639-785. New York: Wiley, 1959.
Encyclopedia o/ Physics (Handbuch der Physik), Vol. 44, Nuclear lnstrumenla-
tion l. Berlin: Springer Verlag, 1959.
l. E. BALDINGER, "Kaskaden generatoren," pp. 1-63.
2. R. G. HERB, "Van de Graaff Generators," pp. 64-104.
3. B. L. CoHEN, "Cyclotrons and Synchrocyclotrons," pp. 105-169.
4. R. R. W1LS0N, "Electron Synchrotrons," pp. 170-192.
5. D. W. KERST, "The Betatron," pp. 193-217.
6. G. K. GREEN and E. D. CouRANT, "The Proton Synchrotron," pp. 218-
340.
7. L. SMITH, "Linear Accelerators," pp. 341-389.
M. S. L1v1NGSTON, High Energy Accelerators. New York: Interscience Pub-
lishers, 1954.
D. L. JuDD, "Conceptual Advances in Accelerators," Ann. Rev. Nuc. Sci. S,
181-216, 1958.
J. J. LIVINGOOD, Principles o/ Cyclic Particle Accelerators. New York: Van
Nostrand, 1961.
S. L1v1NGSTON and J. P. BLEWETT, Particle Accelerators. New York: McGraw-
Hill, 1961.
PARTICULAR
l. J. D. CocKCROFT and E. T. S. WALTON, "Experiment.s with High Velocity
Positive Ions. l. Further Development.s in the Method of Obtaining High
Velocity Positive Ions," Proc. Roy. Soc. (London) Al36, 619 (1932).
2. J. D. CocKcROFT and E. T. S. WALTON, "Experiment.s with High Velocity
Positive Ions. 11. The Disintegration of Element.s by High Velocit.y Protons,"
3. R. L. FoRTESCUE, "High Voltage Direct Current Generators," Progresa in
Nuclear Physics, O. R. Frisch, ed., Vol. 1, New York: Academfo Press, 1950.
4. R. J. VAN DE GRAAFF, "A 1.5 Mev Electrostatic Generator," Phya. Rev.
38, 1919 (1931).
5. VAN DE GRAAFF, CoMPTON, and VAN ATTA, "The Electrostatic Production
of High Voltage for Nuclear Investigations," Phys. Rev. 43, 149 (1933).
6. VAN DE GRAAFF, TRUMP, and BUECHNER, "Electrostatic Generators for the
Acceleration of Charged Particles," Rep. Prog. Phya. Vol. XI. London: Physical
Soc., 1946, p. l.
7. F. W. SEARS, Electricity and Magnetism. Reading, Mass.: Addison-Wesley,
1951.
8. PARKINSON, HERB, BERNET, and McKIBBEN, "Electrostatic Generator
Operating under High Air Pressure. Operation Experience and Accessory
Apparatus," Phys. Rev. 53, 642 (1938).
REFERENCES 707
9. WELLS, HAXBY, STEPHENs, and SHOUPP, "Design and Preliminary Per-

formance Tests of the Westinghouse Electrostatic Generator," Phys. Rev. 58,
162 (1940).
10. VAN DE GRAAFF, SPERDUTO, MclNTOSH, and BuRRILL, "Electrostatic
Generator for Electrons," Rev. Sci. Jnstr. 18, 754 (1947).
11. E. O. LAWRENCE and M.S. L1v1NGSTON, "The Production of High Speed
Light Ions without the Use of High Voltages," Phys. Rev. 40, 19 (1932).
12. W. B. MANN, TheCyclotron. 4thed. New York: Wiley; London: Methuen,
1953.
13. M.S. LIVINGSTON, "The Cyclotron." J. Appl. Phys. 15, 2, 128 (1944).
14. J. H. FREMLIN and J. S. GooDEN, "Cyclic Accelerators," Rep. Prog. Phys.
Vol. XIII. London: Physical Soc., 1950.
15. LAWRENcE, ALVAREZ, BRoBECK, CooKSEY, CoRSON, McMILLAN, SALI~
BURY, and THORNTON, "lnitial Performance of the 6~inch Cyclotron of the
William H. Crocker Radiation Laboratory, University of California," Phys. Rev.
56, 124 (1939).
16. V. VEKSLER, "A New Method of Acceleration of Relativistic Particles,"
J. Phys. (U.S.S.R.), 9, 153 (1945).
17. E.M. McMILLAN, "The Synchrotron-A Proposed High Energy Particle
Accelerator," Phys. Rev. 68, 143 (1945).
18. BRoBECK, LAWRENCE, MAcKENZIE, McMILLAN, SERBER, SEWELL, S1MP-
soN, and THORNTON, "lnitial Performance of the 184-inch Cyclotron of the Uni-
versity of California," Phys. Rev. 71, 449 (1947).
19. HENRICH, SEWELL, and VALE, "Operation of the 184-inch Cyclotron,"
Rev. Sci. Jnstr. 20, 887 (1949).
20. D. H. TEMPLETON, "Nuclear Reactions Induced by High Energy Par-
ticles." Ann. Rev. Nuc. Sci. Stanford: Annual Reviews, Inc., 1953, Vol. 2.
21. V. L. F1TCH and J. RAINWATER, "Studies of X-Rays from Mu-Mesonic
Atoros," Phys. Rev. 92, 789 (1953).
22. J. A. WHEELER, "Mu Meso;3 as a Nuclear Probe Particle," Phys. Rev. 92,
812 (1953).
23. L. N. CooPER and E.M. HENLEY, "Mu-Mesonic Atoms and the Electro-
magnetic Radius of the Nucleus," Phys. Rev. 92,801 (1953).
24. D. W. KERST, "Acceleration of Electrons by Magnetic Induction," Phys.
Rev. 60, 47 (1941).
25. W. F. WESTENDORP and E. E. CHARLTON, "A l~Million Volt Induction
Electron Accelerator," J. Appl. Phys. 16, 581 (1945).
26. KERsT, AoAMs, KocH, and RoBINSON, "An 8~Mev Model of a 3~Mev
Betatron," Rev. Sci. Jnstr. 21, 462 (1950).
27. KERST, ADAMs, KocH, and RoBINSON, "Operation of a 3~Mev Beta-
tron," Phys. Rev. 78, 297 (1950).
28. THOMAS, KRAUSHAAR, and HALPERN, "Synchrotrons," Ann. Rev. Nuc. Sci.
Stanford: Annual Reviews Inc., 1952. Vol. l.
29. ÜLIPHANT, GoooEN, and HrnE, "The Acceleration of Charged Particles to
Very High Energies," Proc. Phys. Soc. (London) 59, 666 (1947).
30. GoooEN, JENSEN, and SYMONDs, "Theory of the Proton Synchrotron,"
Proc. Phys. Soc. (London) 59, 677 (1947).
31. LrvrNGSTON, BLEWITT, GREEN, and HA WORTH, "Design Study for a 3-Bev
Proton Accelerator," Rev. Sci. Imtr. 21, 7 (1950).
32. "The Cosmotron," Rev. Sci. Imtr. 24, No. 9, 723-870 (1953).
33. W. M. BRoBECK, "Design Study for a 10..Bev Magnetic Accelerator," Rev.
Sci. Imtr. 19, 545 {1948).
34. E. J. LoFGREN, "Bevatron Operational Experiences," Proceedinga of the
CERN SympoBium on High Energy Acceleratora and Pwn Phyaics, Vol. l. Geneva:
CERN, 1956, p. 496.
35. E. SEGRE, "Antinucleons," Ann. Rev. Nuc. Sci. 8, 127 (1958).
36. D. H. SLOAN and E. O. LAWRENCE, "The Production of Heavy High
Speed Ions Without the Use of High Voltage," Phys. Rev. 38, 2021 (1931).
37. J. C. SLATER, "The Design of Linear Accelerators," Reva. Mod. Phys. 20,
473 (1948).
38. D. W. FRY and W. WALKINSHAW, "Linear Accelerators," Rep. Prog. Phys.
Vol. XII. London: Physical Soc., 1949.
39. J. C. SLATER, "Linear Accelerators," Ann. Rev. Nuc. Sci. Stanford:
Annual Reviews lnc., 1952, Vol. l.
40. HoFSTADTER, FECHTER, and MclNTYRE, "High-Energy Electron Scatter-
ing and Nuclear Structure Determination," Phys. Rev. 92, 978 (1953).
41. L. l. ScHIFF, "lnterpretation of Electron Scattering Experiments," Phys.
Rev. 92, 988 (1953).
42. R. CoURANT, LmNGSTON, and SNYDER, "The Strong Focusing Synchro-
tron-A New High-Energy Accelerator," Phya. Rev. 88, 1190 (1950).
43. Phyaics Today, 13, Ño. 4, 70 (April 1960).
44. "Brookhaven National Laboratory, lnformation Release," Aug. 2, 1960.
45. R. J. VAN »EGRAAFF, "Tandem Electrostatic Accelerators," Nuclear lnatr.
and Methods, 8, 195 (1960); also, J. J. Livingood, op. cit. gen. ref., Chapter l.
709
PROBLEMS
l. Show that in the fixed-frequency cyclotron, the following relations hold:

H = 6.55 X 10-4 n, for protons; H = 1.31 X 10-3 n, for deuterons, where H
is the field strength in gauss and n is the frequency of the applied voltage in
cycles per second. Derive a similar relationship for a-particles.
2. If the frequency of the applied voltage is 1.2 X 107 cycles/sec, find the
values of the field strength for resonance when protons, deuterons, and a-particles,
respectively, are to be accelerated. If the values of the radius at ejection of the
particles is 50 cm, what is the energy of the particles in each case?
3. Devise reasonable sets of parameters, that is, values of the frequency, field
strength, and radius, for the production of beams of the following particles.
(a) 30-Mev a-particles, (b) 20-Mev deuterons, (e) 10-Mev protons.
4. Show that when a particle of charge q and rest mass M o is accelerated to a
potential of V volts, high enough so that relativistic effects are no longer negli-
gible, the mass ratio may be written
M = 1+--
-Mo 10 ( -
c2 Mo
8
q) V= I+aV,
where a is a constant and e is the velocity of light. Find an expression for the
velocity v of the particle, relative to that of light. Then show that the radius
of the path of the particle in a magnetic field of strength H is given by the
expression
In a frequency-modulated cyclotron with a field of 15,000 gauss and a path

radius of 80 in, what is the energy of the deuterons produced?
5. Assume that the magnetic flux linking the orbit in a betatron varies with
time according to the relation
'1> = 'l>osinwt,
and that the acceleration of the electrons occurs during one-fourth of the cycle,
that is, during a time interval 1r/2w. Find formulas for (a) the energy per turn
available when the flux changes, (b) the average value of this energy during the
acceleration period, (e) the distance traveled by the electrons under the assump-
tion that these speeds differ by only a negligible amount from that of light,
(d) the number of turns, if the radius of the orbit is R, (e) the final energy of the
electrons.
6. The following conditions obtain in a particular betatron: maximum mag-
netic field at orbit = 4000 gauss, operating frequency = 60 cycles/sec, atable
orbit diameter = 66 in. Show that the average energy gained per turn is about
400 ev and the final energy is about 108 ev. [Hint: It follows from Eq. (21-13)
that if the initial flux through the orbit and the initial field strength at the
orbit are both zero, then '1> = 21r R 2 H, where H is the field strength at the orbit.]
Appendixes
APPENDIX I
ALPHABETIC LIST OF THE ELEMENTS
Atomic Atomic
number number
Element Symbol z Element Symbol z
Actinium Ac 89 Holmium Ho 67
Aluminum Al 13 Hydrogen H 1
Americium Am 95 Indium In 49
Antimony Sb 51 Iodine I 53
Argon A 18 Iridium Ir 77
Arsenic As 33 Iron Fe 26
Astatine At 85 Krypton Kr 36
Barium Ba 56 Lanthanum La 57
Berkelium Bk 97 Lead Pb 82
Beryllium Be 4 Lithium Li 3
Bismuth Bi 83 Lutetium Lu 71
Boron B 5 Magnesium Mg 12
Bromine Br 35 Manganese Mn 25
Cadmium Cd 48 Mendelevium Md 101
Calcium Ca 20 Mercury Hg 80
Californium Cf 98 Molybdenum Mo 42
Carbon c 6 Neodymium N<l 60
Cerium Ce 58 Neon Ne 10
Cesium Cs 55 Neptunium Np 93
Chlorine Cl 17 Nickel Ni 28
Chromium Cr 24 Niobium Nb 41
Cobalt Co 27 Nitrogen N 7
Copper Cu 29 Nobelium No 102
Curium Cm 96 Osmium Os 76
Dysprosium Dy 66 Oxygen o 8
Einsteinium E 99 Palladium Pd 46
Erbium Er 68 Phosphorus p 15
Europium Eu 63 Platinum Pt 78
Fermium Fm 100 Plutonium Pu 94
Fluorine F 9 Polonium Po 84
Francium Fr 87 Potassium K 19
Gadolinium Gd 64 Praseodymium Pr 59
Gallium Ga 31 Promethium Pm 61
Germanium Ge 32 Protactinium Pa 91
Gold Au 79 Radium Ra 88
Hafnium Hf 72 Radon Rn 86
Helium He 2 Rhenium Re 75
713 (Continued)
714 APPENDIX I
ALPHABETIC LIST OF THE ELEMENTS------l'ontinued
Atomic Atomic
number number
Element Symbol z Element Symbol z
Rhodium Rh 45 Terbium Tb 65
Rubidium Rb 37 Thallium TI 81
Ruthenium Ru 44 Thorium Th 90
Samarium Sm 62 Thulium Tm 69
Scandium Se 21 Tin Sn 50
Selenium Se 34 Titanium Ti 22
Silicon Si 1.4 Tungsten w 74
Silver Ag 47 Uranium u 92
Sodium Na 11 Vanadium V 23
Strontium Sr 38 Xenon Xe 54
Sulfur s 16 Ytterbium Yb 70
Tantalum Ta 73 Yttrium y 39
Technetium Te 43 Zinc Zn 30
Tellurium Te 52 Zirconium Zr 40
APPEND IX II
VALUES OF PHYSICAL CONSTANTS

(The values listed below are taken from the book, The Fundamental Con-
stants o/ Physics, by E. R. Cohen, K. M. Crowe, and J. W. M. DuMond. New
York: Interscíence Publishers, Inc., 1957.)
Constant Value
Avogadro's number No = (6.02486 ± 0.00016) X 1023 per gram

mole
Velocity of light e = (2.997930 ± 0.000003) X 10 10 cm/sec
Standard volume of a Vo = 22420.7 ± 0.6 cm 3/mole
perfect gas
Gas constant per mole Ro = (8.31696 ± 0.00034) X 10 7 erg/moleº C
Boltzmann 's constant k = Ro/No
= (1.38044 :l:: 0.00007) X 10- 16 erg/ºK
= (8.6167 ± 0.0004) X 10-5 ev/°K10
Charge on the electron e = (4.80286 ± 0.00009) X 10- esu
20
= (1.60206 ± 0.00003) X 10- emu
Faraday constant F = Noe = (2.89366 ± 0.00003) X 10 14
(physical scale) esu/ (gm mole)
= (9652.19 ± 0.11) emu/(gmm ole)
Planck's constant h = (6.62517 ± 0.00023) X 10-27 27
erg-sec
h/21r = (1.05443 ± 0.00004) X 10- erg-sec
Rydberg constants R,,, = 109737.309 ± 0.012 cm- 1
RH = 109677.576 ± 0.012 cm- 1
Ro = 109707.419 ± 0.012 cm- 1
RHo = 109722.267 ± 0.012 cm - 1
Rest masses neutron: 1.008982 ± 0.000003 amu
= (1.67470 ± 0.Q0004) X 10-24 gro
proton: 1.007593 ± 0.000003 amu
= (1.67239 ± 0.00004) X 10-24 gm
hydrogen atoro: 1.008142 ± 0.000003 arou
ratio: roass of hydrogen atom
mass of proton
= l.000544613 ± 0.000000006
electron: (5.48763 ± 0.00006) X 10-4 arou
= (9.1083 ± 0.0003) X 10- 2 8 gro
ratio: proton mass
electron mass
= 1836.12 ± 0.02
deuteriuro atoro: 2.014735 ± 0.000006 arou
715 (Continued)
716 APPENDIX II
VALUES OF PHYSICAL CONSTANTS-C ontinued
Constant Value
Mass-energy conversion 1 amu = 931.141 ± 0.010 Mev
factors 1 proton mass = 938.211 ± 0.010 Mev
1 neutron mass = 939.505 ± 0.010 Mev
1 electron mass = 0.510976 ± 0.000007 Mev
1 gm = (5.61000 ± 0.00011) X 1026 Mev
Energy conversion factor 1 ev = (1.60206 ± 0.00003) X 10- 12 erg
Energy of a 2200 m/sec E2200 = 0.0252973 ± 0.0000003 ev
neutron T2200 = 293.585 ± 0.012 ºK
= 20.435 ± 0.012 ºC.
First Bohr radius ao = h 2/4r 2me 2
= (5.29172 ± 0.00002) X 10-9 cm
Classical electron radius ro = e2;mc2
= (2.81785 ± 0.00004) X 10-13 cm
Thomson cross section !rr3 = (6.65205 ± 0.00018) X 10-25 cm2
Compton wavelength electron: >-•• = (24.2626 ± 0.0002) X 10- 11 cm
proton: Xcp = (13.2141 ± 0.0002) X 10-14 cm
neutron: X.,. = (13.1959 ± 0.0002) X 10- 14 cm
APPENDIX III
JOURNAL ABBREVIATIONS FOR REFERENCES
Journal Abbreviation
American Journal of Physics Am. J. Phys.

Canadian Journal of Physics Can. J. Phys.
Canadian Journal of Research Can. J. Research
Comptes rendus hebdomadaires des séances de
l'académie des sciences Compt. rend.
Journal of the American Chemical Society J. Am. Chem. Soc.
Journal of Applied Physics J. Appl. Phys.
Journal of Chemical Physics J. Chem. Phys.
Journal of lnorganic and Nuclear Chemistry J. Inorg. Nucl. Chem.
Journal of the Chemical Society (London) J. Chem. Soc.
Journal de physique et le radium J. phys. et radium
Journal of Physics (U.S.S.R.) J. Phys. (U.S.S.R.)
Kongelige Danske Videnskabernes Selskab, Kgl. Danske Videnskab,
Matematisk-fysiske Meddelelser Selskab, Mat-fys. Medd.
Nature Nature
Naturwissenschaften Naturwiss.
Philosophical Magazine and Journal of Science Phil. Mag.
Physical Review Phys. Rev.
Physikalische Zeitschrift Physik. Z.
Proceedings of the Cambridge Philosophical
Society Proc. Cambridge Phil. Soc.
Proceedings of the Physical Society (London) Proc. Phys. Soc. (London)
Proceedings of the Royal Society (London) Proc. Roy. Soc. (London)
Reviews of Modern Physics Revs. Mod. Phys.
Review of Scientific Instrumenta Rev. Sci. lnstr.
Zeitschrift für Naturforschung Z. N aturforsch.
Zeitschrift für Physik Z. Physik
Zeitschrift für Physikalische Chemie Z. Physik. Chem.
717
APPENDIX IV
THE RUTHERFORD SCATTERING FORMULA
The treatment to be given here of the scattering of a-particles by a

positively charged atornic nucleus is more rnathernatical than that given
in Chapter 3, but it contains sorne ideas which are important in nuclear
physics; for exarnple, that of the differential scattering cross section. For
the purposes of this derivation we consider the infinitely heavy nucleus
of charge Ze to be at the origin of a system of polar coordinates r, 8 as
shown in Fig. A-IV-1. The initial path of the a-particle is along the
line DC, and p is the shortest distance between this line and the nucleus;
the position of the a-particle is given generally by r and 8. Although the
a.ngle 8 is different from what it was in Section 3-2, there should be no
confusion. Other quantities are defined as in Section 3-2.
The angular momentum L about the nucleus is given by
2 do
L = Mr dt; (A4-1)
it is a constant of the rnotion, and we rnay write

2 do
L = Mr dt = MVp. (A4-2)
Fm. A-IV-1. Hyperbolic orbit under an inverse square law of repulsion.

718
APPEXDIX IV 719
The coulomb repulsive force, F = zZe 2 /r 2, gives rise to a radial acceleration
a, -
2
dr
dt2
(d8)
r dt
2
-
F zZe
M - Mr 2
2
(A4-3)
Equations (A4-2) and (A4-3) are two simultaneous equations in r, 8, and t;

if t is eliminated between them, we get a single differential equation in r and
8 which gives the orbit of the a-particle. For convenience, we put
1
r = -, (A4--4)
u
and get, from Eq. (A4-1),
(A4-5)
It follows that
(A4--6)
and that
(A4-7)
If we now substitute for d 2r/dt 2 and d8/dt in Eq. (A4-3), the result is
d 2u zZe 2M b
d82 +u = - -¡y:- = - 2p2 ' (A4-8)
where
(A4-9)
Equation (A4-8) is the differential equation of the orbit; it has the

general solution
b
u = -1r = A cos (8 - 80) - -
~2
, (A4-10)
where A and 80 are arbitrary constants. The constant 80 just determines

the orientation of the orbit in the (r, 8) plane. We can choose the x-axis
,.
720 APPENDIX IV
so that 80 = O, and the equation of the orbit becomes
u = -rI = - -2pb 2 + A cos 8. (A4-11)
The equation
1
u=-=B+Acos8 (A4-12)
r
is known, from analytic geometry, to representa conic section (parabola,
ellipse, or hyperbola) with eccentricity given by
A
E= IBI. (A4-13)
The orbit is a hyperbola if E > l. Since Bis known and is equal to -b/2p 2 ,
it remains only to determine A or E. On setting
b
A= E!BI =2p2
-, (A4-14)
we get
U = -rI = - -2pb 2 (1 - ECOS 8). (A4-15)
We can determine E (or A) by making use of the equation for the con-
servation of energy,
.!. MV2 =
2
_!. Mv2
2
+ zZe2'
r
(A4-16)
where v is the speed of the a-particle when it is at a distance r from the

nucleus. The last equation may be written
2
-Vv 2 = 1 - -·
b
T
(A4-17)
The quantity v2 is given by
(A4-18)
From Eq. (A4-15), we get

dr Eb z.
d8 = 2p2 r sm 8; (A4-19)
and, from Eq. (A4-2),
(A4-20)
APPENDIX IV 721
so that
v2 E2b2 . 2 p2 b
y2 = 4p2 sm 8 + r2 = 1- r.
If we now insert for r the expression (A4-15), and then simplify, we get
2
4p2
E = 1 + fi2'
or
4p2)1/2
E = (1 + ----¡j2 • (A4-21)
Since b and p are positiv~, E > 1, and the orbit is a hyperbola. The
constant A is then given by
1 ( b2 )112
A=-p 1+-
4p2
,
and
1
(A4-22)
r
The orbit is thus expressed in terms of the two quantities p, the impact
parameter, and b, the closest distance which an a-particle of initial speed
V can approach to a nucleus of charge Z. When r - oo, 8 - ,j,, the angle
between the x-axis and the asymptote of the hyperbola, and
b
cos ,¡, = 2p[l + (b2/4p2)]1i2
On solving for b, we get
b = 2pcot,j,,
or
p = 2b tan,j,.
But, the scattering angle q, = 7r - 2,j,, and
p = 2b cot 24' , (A4-23)
which is Eq. (3-11).

If collisions occur for values of the impact parameter within a range dp
at p, they will result in deflections within the angular range dq, at q,, with p
722 APPEXDIX IV
FJG. A-IV-2. Effective cross section for scattering into the solid angle dO (by
permission from S. W. McCuskey, Introduction to Advanced Mechanics, Addison-
Wesley Publishing Co., lnc., 1959).
and cf> related through Eq. (A4-23). The probability of such collisions is
proportional to the area of an annulus of radius p and width dp, because
this is the effective target area presented by the nucleus to the incident
beam for producing deflections of this particular type. We can therefore
define a collision area, or cross section du(q,) for scattering into dq,:
du(<t,) = 21rp dp
_ 1rb 2 cos <t,/2 d

= 4 sin3 q,/2 q,. (A4-24)
In an experiment, the scattering into a detector which subtends a fixed

solid angle is usually observed. The element of solid angle is
díl = 21r sin q, de/>

= . e/> cos e/> d <f,,
4 1r sm
2 2
and
du(q,) b2 Z 2e 4
(A4-25)
dn = 16 sin4 q,/2 = (MV 2 ) 2 sin 4 q,/2
The Rutherford scattering formula is often written in this form; du(q,)/díl

.is called the dijferential scattering cross section and is the effective collision
area for producing scattering into a unit solid angle at cf,.
The Rutherford formula seems to show that the probability that an
a-particle ·is scattered through an angle cf, ~eco mes infinite as cf> --+ O.
The reason for this result is that the presence of negatively charged particles
APPENDIX IV 723
in the atom has been neglected. These particles have the effect of "screen-
ing" the nucleus from the a-particles so that the latter move in a modified,
or "screened" Coulomb potential. The more rigorous calculation of the
scattering cross section for this potential removes the infinity at t/, = O
without changing any of the results of Chapter 3.
It is important to note that the Rutherford scattering formula does not
depend on the geometrical details of the orbit of an a-particle even though
the derivation involves sorne of those details. The formula contains only
quantities which can be determined experimentally, such as the charge
and incident speed of the a-particle, the nuclear charge of the target
atoms (see Section 4-4), macroscopic properties of the scattering material
and the scattering angle.
The effective cross section 21rp dp for scattering into dtf, is illustrated in
Fig. A-IV-2.
Answers to Problems
ANSWERS TO PROBLEMS
CHAPTER 1
l. Eight parta by weight of oxygen combine with 14.00, 7.00, 4.67, 3.50 and
2.80 parta of nitrogen, respectively. Then, 14.00 parta by weight of nitrogen
combine with 8( = 8 X 1), 16( = 8 X 2), 24( =- 8 X 3), 32( = 8 X 4), and
40( = 8 X 5) parta of oxygen, respectively.
2. The weight of nitrogen in one molecular weight of each compound, as cal-
culated from the data, is listed below.
nitrous oxide 28.010 =2X 14.005, nitric oxide 14.006 = 1 X 14.006,
nitrous anhydride 28.006 = 2 X 14.003, nitrogen tetroxide 14.005 = 1 X 14.005,
nitric anhydride 28.010 =2X 14.005.
The smallest weight of nitrogen present in one molecular weight of these
compounds is (averaged) 14.005, which is then the atomic weight of nitrogen
as obtained from the data of Problema 1 and 2.
3. (a) 10.32, (b) 31.0, (c) 31.0, (d) 3, (e) 30.96, (f) 137.33. The element is
phosphorus; the compound is PCla.
4. 195.2
5. (a) 57.1, (b) 18.617, (c) 3, (d) 55.85. The metalisiron, thecompound FeCl3.
6. The atomic volume curve has sharp maxima at sodium, potassium, rubidium,
and cesium-the alkali metals. The atomic volume decreases in passing
from lithium to boron, after which it increases through carbon, oxygen, and
fl.uorine to sodium, when it again decreases through magnesium down to
aluminum, and then increases to potassium. The elements boron, alumi-
num, cobalt and nickel, rhodium, etc., lie at the troughs of the curve. The
melting point curve shows minima at sodium, potassium, cesium, and
rubidium, and maxima at elements of groups between III and V, approxi-
mately midway between the alkali metals.
7. The new element is technicium, with an approximate atomic weight of 99.
It is in group VII of the periodic table and resembles manganese in its chem-
ical properties.
CHAPTER 2
l. (e) charge,
Element Valence (a) gm/amp-hr (b) coul/mg
esu/gm
Cr 6 0.3234 10.132 3.34 X 10 13

Cu 2 1.186 3.036 0.910 X 1013
In 3 1.427 5.223 0.756 X 10 13
Fe 3 0.6945 5.184 1.55 X 1013
Hg 2 3.742 0.9621 0.288 X 1013
Ni 2 1.095 3.288 0.986 X 1013
Ag 1 4.0245 0.8945 0.268 X 1013
7'1:T
728 ANSWERS TO PROBLEMS
2. v = 3.20 X 109 cm/sec; e/m = 1.73 X 107 emu/gm.

3. H = 15.34 ga1188¡ displacement = 19.4 cm. 4. R = 14.8 cm.
5. v = 2.60 X 109 cm/sec; e/m = 1.74 X 107 emu/gm.
6. H = 55 gall88. 7. 56.8 volts/cm.
8. Equation (2-8) may be written t:.q = Constant X (V B' - V B), where the
V's are obtained by dividing 0.5222 cm by the time of rise or fall. The suc-
cessive ratee are listed in the second column of the following table.
Rate of rise or fall,

Time
cm/sec Rate + 0.0030 cm/sec
12.45 0.0419 14.0

21.5 0.0243 8.1
34.7 0.0150 5.0
85.0 0.00614 2.0
34.7 0.0150 5.0
16.0 0.0326 10.9
34.7 0.0150 5.0
21.85 0.0239 8.0
The third column shows that the ratee are all integral multiples of the great-
est common divisor 0.0030 cm/sec. Hence the differences V B' - V B are inte-
gral multiples of a certain constant quantity, and t:.q is alwaye an integral
multiple of a certain quantity of charge.
9. 0.08323 cm/sec; 0.08571 cm/sec.
10. (a) 4.98 X 10-10 esu; (b) 4.76 X 10- 10 esu.
CHAPTER 3
2. tp
tp Counts/hr (rounded off) Counts/hr (rounded off)
5º 67,630 5º 28,700
10º 4,240 10º 1,960
15° 840 15° 370
30º 54 30º 24
45° 11 45° 5
60º 4 60º 2
3. tp Counts/hr (rounded off) 4. Z(Cu) = 29.
5º 23,800
10º 1,490
15° 295
30º 19
45° 4
60º 1
ANSWERS TO PROBLEMS 729
5. Distance of closest approach, cm

Element z
5.30 Mev 7.0 Mev
Gold 79 4.3 X 10-12 3.2 X 10- 12

Silver 47 2.6 1.9
Copper 29 1.6 1.2
Lead 82 4.5 3.4
Uranium 92 5.0 3.8
6. 8.56 X 10- 6 7. 5.7 X 10-12 cm 8. 1.3 X 10-4 cm

CHAPTER 4
l. (a) n = 4.3 =4, (b) n = 4.5, or between 4 and 5, µ. = 0.34 cm - 1 •
2. >. = 0.842 A, 82 = 17º 22'; 83 = 26° 36'.
3. Rock salt: 10º 13', calcite: 9° 29', quartz: 6° 45', mica: 2° 53'.
4. Element 81 Element 61
Cr 7° 52' Co 6° 8'
Mn 7º 13' Ni 5°40'
Fe 6° 39' Cu 5° 17'
Zn 4° 56'
5. Element Wavelength, A Wave number Frequency
Al 8.364 1.196 X 107 3.586X 10 17/sec

Si 7.142 1.400 4.197
CI 4.750 2.105 6.311
K 3.759 2.660 7.974
Ca 3.368 2.969 8.901
Ti 2.758 3.626 1.087 X 10 18
V 2.519 3.970 1.190
Cr 2.301 4.346 1.303
Mn 2.111 4.737 1.420
Fe 1.946 5.184 1.554
Co 1.798 5.562 1.667
Ni 1.662 6.017 1.804
Cu 1.549 6.456 1.935
Zn 1.445 6.920 2.074
y 0.838 1.193 X 108 3.577
Zr 0.794 1.259 3.774
Nb 0.750 1.333 3.996
Mo 0.721 1.387 4.158
Ru 0.638 1.567 4.698
Pd 0.584 1.712 5.132
Ag 0.560 1.786 5.354
6. X = 0.682 A, Z = 43 7. d = 2.816 cm; No = 6.048 X 1023

8. Z = 90
CHAPTER 5
l.
T, ºC W, erg/cm 2 sec W, gm-cal/cm 2 sec
o 3.15 X 105 7.53 X 10-3

100 1.10 X 106 2.63 X 10-2
300 6.11 X 106 0.146
500 2.02 X 107 0.483
1000 1.49 X 108 3.56
1500 5.60 X 108 13.4
2000 1.51 X 109 36.1
3000 6.51 X 109 156
2. T, ºC Energy loss, gm-cal/cm 2 sec
o o
100 0.0188
300 0.138
500 0.475
1000 3.552
1500 13.39
2000 36.1
3000 156
3. h = 6.625 X 10-27 erg sec.

5 4
8 r k 4 27
4. u = 15 cªh3 T = aT"', h = 6.625 X 10- erg sec.
5. Energy
X 11, sec- 1
ergs ev
1 km 2.99793 X 105 1.9862 X 10-21 1.25 X 10-9

lm 2.99793 X 108 1.9862 X 10- 18 1.25 X 10-6
1cm 2.99793 X 10 1º 1.9862 X 10- 16 1.25 X 10-4
1 mm 2.99793 X 10 11 1.9862 X 10- 15 1.25 X 10-3
10-4 cm 2.99793 X 10 14 1.9862 X 10- 12 1.25
5000A 5.99586 X 10 14 3.9724 X 10- 12 2.50
1000 A 2.99793 X 10 15 1.9862 X 10- 11 12.5
lA 2.99793 X 10 18 1.9862 X 10-8 1.25 X 104
10-3 A 2.99793 X 1021 1.9862 X 10-5 1.25 X 107
10- 13 cm 2.99793 X 1023 1.9862 X 10-3 1.25 X 109
10-15 cm 2.99793 X 1025 1.9862 X 10- 1 1.25 X 10 11
6. (a) (b) (e)

>.thr, A Vmax, cmlsec V, ev
Ag 2760 3.88 X 107 1.7

Ba 4970 5.71 X 107 3.7
Li 5400 5.86 X 107 3.9
Pt 3030 4.3 X 10 7 2.1
Ni 2485 3.26 X 107 1.2
7. h. = 6.602 X 10-27 erg sec.

8. llmax = 5.93 X 107 -VV (volts).
9. (a) 3.0 ev, (b) 4167 A, (e) 9.5 ev
10. >..,,A , .., sec- 1 h,,,., ev

Voltage
50,000 0.248 1.21 X 10 19 5 X 104

105 0.124 2.42 X 10 19 105
5 X 10 5 0.0248 1.21 X 102 º 5 X 105
106 0.0124 2.42 X 102 º 106
CHAPTER 6
2. u
,2
= 1 [
[l _ (t1uslc 2 )] 2 (us - 11)
2
+( 1 - 11 )
c2 2
(u 11 + u,)2 ] .
If u = e, then u: + u~ = c2 - u! and the expression in brackets becomE".s
+ 11 c4u!) (i _
2 2
c2 ( 1 _ 211Us = c2 11Us) ,
c2 c2
so that u' 2 = c2 , and u' = c.

3. 1.8c; 0.9945c.
4. vlc mimo 11lc mimo 11lc mimo
0.1 1.005 0.7 1.400 0.99 7.089

0.2 1.021 0.9 2.294 0.995 10.013
0.5 1.155 0.95 3.203 0.999 22.366
5.
Electron energy
v/c Proton energy
ergs Mev
0.1 4.10 X 10- 9 2.56 X 10-3 4.70 Mev

0.2 16.75 1.05 X 10-2 19.3 Mev
0.5 125.7 7.85 X 10-2 1.44 X 102 Mev
0.7 325.2 0.203 3.73 X 102 Mev
0.9 1,052 0.657 1.21 Bev
0.95 1,790 1.12 2.06 Bev
0.99 4,950 3.09 5.67 Bev
0.995 7,320 4.57 8.39 Bev
0.999 17,360 10.84 19.90 Bev
6. 0.115; 0.365; 0.816.
7. 20Mev 400 Mev 2.5 Bev

1
M/Mo 1.0002 1.43 3.66
v/c 0.020 0.713 0.962
E 958 Mev 1338 Mev 3438 Mev
8. lt follows from Eq. (6-26) that

mo mo
m1 = ----;:==::::;::::::;:: = ----;:==~~,
Vl - (u~/c2) Vl - (U2/c2)
mo m1
Vl - (v2/c 2)Vl - (U2/c2) Vl - (v2/c 2)
9. (a) Square the relation

m = mo .
1 - (112 /c 2 ) '
the result is
2 2
2 e - 11
m _c_2_ = mo,2 and
or
(pc)2 = (mc2)2 - (moc2)2.
(b) Combining the Jast result with the relation T = mc2 - moc 2 gives the
required relation. Solution of (b) for T gives (d).
(e) Use of the result of (a) for (pc) 2 gives the required formula.
10. Energy a.nd momentum conserva.tion give

2 2
PI+ P2 = 0, or Pl p2.
Hencé Ef - Ei = Mfc 4 - M~4, a.nd solution of the two equa.tions for

E1 a.nd E2 gives the required result.
11. The ma.ss is M = (1/c2}v'E2 - p2c2, where Eand pare theenergy a.nd mo-
mentum of the composite particle, equa.l respectively to the sum of the energy
a.nd momentum of the two colliding particles. The velocity is V = pc 2/ E,
and the re8Ult follows.
12. (a) The ma.ximum energy transfer to the electron occurs when q, = r, a.nd
the result (a) follows directly.
(b) For q, ~ O, e.g., for q, = r, and a »
1,
Tm..; = hvo ( 1 - ~ + · · -) ,
and
1 1 2
(hv')min = hvo - T max ,.. hvo · 2a = 2 moc .
(e) Solve Eq. (&-30) for
• hv~.
sm8 = -----'--smq,·
e mofk '
divide Eq. (&-29) by sin 8 and collect.
13. >.o = 2h/moe; hvo = ½moe 2 = 0.255 Mev; v = Je = 1.8 X 10 10 cm/sec.
14. (a) >. = 0.1000 A.
(b) hvo = 137.9 kev; hv = 124.1 kev.
(e) 7.37 X 10- 18 gm cm/sec¡ 6.63 X 10- 18 gm cm/sec.
(d) T = 13.8 kev¡ p = 6.39 X 10-18 gm cm/sec¡ v = 6.84 X 10 9 cm/sec.
(e) 8 = 57°.
CHAPTER 7
l. n = 20, 25, 30.
2. RD = 109707.42cm- 1. '1>. = 1.79AforH.. ;1.32AforHp;l.18AforHr
3. In Eqs. (7-9) and (7-10) repla.ce v by rw, where w is the angular velocity;
this gives
2 h
mrw=n-, (7-9')
2r
and
mrw2 ze2
= r2. (7-10')
The frequency of orbital revolution is then found by solving these equa.tíons

for w¡ it is
734 ANSWERS TO PROBLEM:S
For comparison, Eq. (7-16) may be written
JI =
2
h3
2
4r me'Z (n1 +n2) (
n1 -
)
n2 •
2 2
2n1n2
Since n1 > n2, it follows that
Hence, if n1 - n2 = 1, the frequency JI of the emitted radiation is inter-

mediate between the frequencies of orbital revolution in the initial and final
states. For very large n, the orbital frequencies in successive orbits are
practically equal and the emitted and orbital frequencies practically coincide.
4. n = 1: energy = -109,678 cm- 1 ; -217.8 X 10- 13 erg; -13.00ev;
-313.4 Kcal/gm atom.
n = 5: energy = -4.387 cm- 1 ; -8.71 >< 10- 13 erg; -0.544 ev; -12.52
Kcal/gm atom.
5. MH!m = 1836.0; e/m = (e/MH)(MH/m) = 5.27 X 10 17 esu/gm.
6.
Excitation Quantum numbers, n
energy, ev Wavelength, A
Initial state Final state
10.19 1216.6 2 1
12.08 1026.3 3 1
12.75 972.4 4 1
13.05 950.0 5 1
13.12 944.9 6 1
13.31 931.4 7 1
The lines belong to the Lyman series of hydrogen.

7. For the derivations, which are somewhat in volved, see Born, Atomic Phyaica,
Appendix XIV.
8. The tables asked for can be worked out directly from Table 7-4. Tbe oc-
currence of the rare earths can be explained in terms of the filling of the
4f subshell. At xenon (Z = 54), the 4s, 4p, 4d, 5s, and 5p subshells are
filled, which accounts for the chemical inertness of that element. The fill-
ing in of the 5d subshell is started with lanthanum, which has one 5d elec-
tron. At this point, the 4f subshell with room for 14 electrons is still empty.
The successive additíon of electrons up to the number of 14 would fill this
shell; the resulting elements should resemble each other in chemical prop-
erties because these properties are deterinined mainly by the outermost
electrons, i.e., those corresponding to n = 5. The elements from lanthanum
(Z = 57) to lutecium (Z = 71) correspond to the filling of the 4f subshell,
as shown in Ta.ble 7-4, a.nd are the 15 rare earths.
9. Wavelength of K,. line, Wavelength of K,. line,

Element Element Eq. (7-39)
Eq. (7-39)
Al 8.438 A Ni 1.667 A
K 3.750 Cu 1.550
Fe 1.944 Mo 0.723
Co 1.797 Ag 0.574
• h h hv'l - (v/2c2)
10 ). - - =- - - - ----,-...,....,..---
moc(v/c)
--.
· mv mc(v/c)
Compton wavelength of the eleetron - h/moe = 2.4263 X 10-10 cm.
Electron Compton Electron Compton
speed, v/c wavelength speed, v/c wavelength
0.1 2.414 X 10-9 0.95 7.973 X 10-11

0.2 1.188 X 10- 9 0.99 3.457 X 10-11
0.5 4.202 X 10- 1<! 0.995 2.436 X 10- 11
0.7 2.475 X 10-10 0.999 1.086 X 10-11
0.9 1.175 X 10-10
13
The Compton wavelength of the proton is 1.3214 X 10- cm.
11. ). - h/mv = h/mQII when ti«
c. From the result of Problem 8, Chapter 5,
v = 5.93 X 107 vv', where V is the voltage. Substitution of numerical
values of h and mo gives the required formula..
Volta.ge, volts >., cm

1 12.268 X 10-8
10 3.880 X 10- 8
100 1.2268 X 10-s
1000 0.3880 X 10-8
12. The formula ti = 5.93 X 107 v'V is obtained from the relation eV = ½mv •
2
Hence, t1 ~
m- , and). ~ m- v- or m- • The relativistic correetion,
112 1 1 112
therefore, introduces the factor (m/mo)- 112 into the formula for the wave-
length.
Voltage >.,cm
10• 1.215 X 10-9

10s 3.457 X 10-10
106 7.134 X 10-11
107 8.554 X 10-12
13. (a) 0.82 A, {b) 8.3 X 10-24 A.

736 ANSWERS TO PROBLEM:S
CHAPTER 9
l. (a) 1.099 X cm/sec, {b) 2.652 X 10- 23 gm; 15.98 amu, (c) 16.
10 7
2. When the velocity vis eliminated between Eqs. (9-4) and (9--5), the equation
MV = ½qH 2 R2
is obtained. In this equation, M is in grams, R in cm, and V, q, and H in
electromagnetic units. For a singly-charged ion q = 1.60207 X 10-20 emu;
1 volt = 108 emu, and 1 amu = 1.65982 X 10- 24 gm. When these con-
version factora are inserted, the result is
M(amu)V(volts) = 4.826 X 10- 5H 2(gauss 2)R 2(cm2).
For the data of Problem 1, the formula gives M = 15.99 amu.

3. At R = 17.6 cm and R = 18.9 cm, respectively.
4. Maxima are observed at Z = 8, 20, 28, 50, and 82.
5. Maxima are observed at A-Z = 20, 28, 50, and 82.
6. C 12, 16.88 Mev; N 14, 7.54 Mev; F 19, 7.95 Mev; Ne 20, 12.87 Mev; Si 28,
11.59 Mev.
7. C 13, 4.96 Mev; N 13, 1.95 Mev; N 14, 10.27 Mev; 0 16, 7.15 Mev.
8. H 1 = 1.008142 amu; H 2 = 2.014736 amu; C 12 = 12.003821 amu.
9. H 1 = 1.007826 amu; H 2 = 2.014104 amu; 0 16 = 15.994904 amu.
10. C 12 = 12.003816 amu; S32 = 31.982238 amu.
12. 96495.0 coul/gm mole.
10
CHAPTER
1
l. T = 137 min; ;\, = 0.005058 min- . The counting rate at t = O would
have been 11,100 min- 1 •
2. T 1 = 137 min, Ao1 = 11,100 min- 1 ; T2 = 45.7 min, Ao2 = 8000 min-1•
3. Time, days Fraction of sample disintegrated, %
1 16.8
2 30.4
3 42.0
4 51.6
5 59.6
10 83.7
A one-microgram sample of radon contains initially
10-6{~5)1023 = 2.71 X 1015 atoros.
Then, 4.55 X 10 14 atoros disintegrate in the first day; 2.17 X 1014 in the
fifth day, and 0.87 X 10 14 in the tenth day.
4. Time, Fraction changed Time,

Fraction changed
to UX2, % days to UX2, % days
90 80.1 99 160.2
95 104.2 63.2 24.1
5. (a) Ra: 1 curie weighs 1 gm; 1 rd weighs 2.70 X 10-5 gm.

10
(b) Rn: 1 curie weighs 6.45 X 10-6 gm; 1 rd weighs 1.74 X 10- gm.
(e) RaA: 10-6 cu weighs 3.54 X 10-15 gm; 1 rd weighs 9.56 X 10- gm.
14
21
(d) RaC': 10- cu weighs 3.1 X 10- gm; 1 rd weighs 8.4 X
6 10- 20 gm.
6. T1¡2 = 8.0 X 104 years; 0.049 rd.
7. Time, No. of disintegrations Activity,

Activity, rd
days per second millicuries
o 1.664 X 108 4.50 166.4

10 1.58 X 108 4.27 158
30 1.43 X 108 3.86 143
50 1.29 X 108 3.49 129
70 1.17 X 108 3.16 117
100 1.01 X 108 2.73 101
300 3.69 X 107 1.00 37
8. 10.8 hr.
9. Neglect Tl 2º6 because of its short half-life. Then RaE, RaF, and RaG form
a 3-member chain which can be treated by Eqs. (10-16), (10-22), and
(10-23). The greatest amount of RaF will be formed after 24.85 days,
and will be 5.06 X 1011 atoms. At that time the a-activity will be 2.93 X 104
disintegrations/sec, as will be the ¡3-activity.
10. (a) >./J = 1.22 X 10-4 sec- 1, >... = 0.69 X 10-4 sec- 1
(b) T112 (ft) = 94.8 min, T112 (a) = 167.1 min
(c) a, 1.96 X 10 1º sec- 1; {J, 3.46 X 10 10 sec- 1
(d) a, 2.49 X 109 sec- 1; {J, 4.39 X 109 sec- 1
11. 1.3 X 10 11 y
12. Secular equilibrium.
13. During a 10-min period, the activities of Th 234 and Pa. 234 seem to be in
secular equilibrium; but overa 10-day period, the Pa. 234 activity would de-
crease with a half-life of 24.1 days, in transient equilibrium with the Th 234 .
14. The number of atoms N" of a given member of a radioactive series in secular
equilibrium is related to the number of atoms N 1 oí the parent substance by
the formula
If the parent substance is U 238 with a half-life of 4.50 X 10~ years, the
number of atoros of each daughter substance per 10 9 atoms of U 238 is given

in the table which follows:
Substance Atoms/billion of U 23 8 Substance Atoms/billion of U238

Th234 1.47 X 10- 2 Em222 2.32 X 10-3
u234 5.54 X 104 Pb210 4.88
Th23o 1.77 X 104 Bi2I0 3.05 X 10-3
Ra226 3.59 X 102 p 0 210 8.41 X 10-2
The resulta for the series starting with U 235 are
Substance Atoms/billion of U235
Th2a1 3.95 X 10-3

Pa2a1 4.83 X 104
Ac221 30.9
Th221 7.17 X 10-2
Ra223 4.32 X 10-2
15. The disintegration constant of U 238 is 1.54 X 10- 10 year- 1. Hence, one
gram of uranium produces, in one yea.r, an amount of Pb206 equal to
l.54 X 10-10 X 206 = 1.33 X 10-10 gm.

238
Hence, the age in years can be found by dividing the a.mount of Pb20G cor-
responding to one gram of uranium by 1.33 X 10-10, or by multiplying it
by 7.5 X 109. The formula
Age = (Pb 2º6/U) X 7.5 X 109 years, follows directly.
(a) 108 years; corrected 9.9 X 107 years.

(b) 7.5 X 108 years; corrected 7.1 X 108 years.
16. One gram of uranium produces in one year
1.54 X 10-10 X :! = 2.07 X 10-

11
gm of helium = 1.16 X 10- 7 cm3 of
helium at N.T.P., since the density of helium at N.T.P. is 0.1785 gm/liter.

Then the a.ge is equal to
(Be/U) + 1.16 X 10-1 year = (He/U) X 8.6 X 106 years.
This formula may also be written
Age = (He/Ra)(Ra/U ) X 8.6 X 106 years,

where (Ra/U) is the number of grams of radium per gram of uranium and
is equal to 1620/4.50 X 109 • Then
Age = (He/Ra)(1620/4.50 X 109 )8.6 X 10 6 = 3.l(He/Ra).
The age of the mineral is approximately 1.5 X 109 years.
CHAPTER 11
l. and 2.
Q-Value, Q-Value,
Reaction Reaction Type
Mev Type Mev
H 1(n, 'Y)H 2 2.230 exoergic Be9 (d, p)Be 10 4.585 exoergic

H 2(n, 'Y)H 3 6.251 exoergic Bll(d, a)Be 9 8.016 exoergic
Li7(p, n)Be 7 -1.645 endoergic c12(d, n)N1a -0.280 endoergic
Li 7(p, a)He 4 17.339 exoergic N14(a, p)Ol7 -1.198 endoergic
Be 9 (p, d)Be 8 0.559 exoergic N15(p, a)Cl2 4.961 exoergic
Be9 (p, a)Li 6 2.133 exoergic O 16 (d, n)F 17 -1.631 endoergic
3.
Reaction Mass of product nucleus
(a) Al 27 (n, 'Y)Al 28 27.99076

(b) Al 27 (d, p)Al 28 27.99076
(e) Al 27 (p, a)Mg 24 23.99263
(d) Al27 (d, a)Mg25 24.99374
4. Q = 8.04 Mev. 5. E,. = 2.368 Mev.

6. Be 9 (p, n)B 9 : Q = -1.853 Mev; C 13 (p, n)N 13 : Q = -3.003 Mev;
ois(p, n)F 18 : Q = -2.453 Mev.
7.
Reaction Threshold energy, Mev
Bll(p, n)C 11 3.015

ois(p, n)F1s 2.590
Na2ª(p, n)Mg23 5.091
8. Energy change
Reaction Energy change Reaction
c12(n, 'Y)c1a exoergic c12(d, p)c1a exoergic

c12(n, p)Bl2 endoergic c12(a, n)O15 endoergic
C 12 (n, a)Be 9 endoergic Cl2(a, p)Nl5 endoergic
c12(p, n)N12 endoergic c12('Y, n)Cll endoergic
Cl2(d, n)Nla endoergic c12('Y, p)Bll endoergic
740 A'.I.SWERS TO PROBLEMS
9. Reaction Energy change ! Reaction Energy change

1
Bil(p, 'Y)Cl2 exoergic Be 9(d, n)C 12 exoergic

Nl5(p, a)Cl2 exoergic 131o(a, d)Cl2 exoergic
Bil(d, n)Cl2 exoergic Ci3('Y, n)Cl2 endoergic
Nl4(d, a)Cl2 exoergic O16(-Y, a)Cl2 endoergic
ciª(d, t)c12 exoergic Nl4(n, t)CI2 endoergic
10. Equation (11-9) can be written in the form

Eu = Q + Ez[l - (mz/Jfy)).
1 + (mu/My)
The value of Q can be calculated from the masses of the target and product
nuclei; Ez = 1.8 Mev. The mass numbers of mz, mu, and Jfy are known.
Hence, the value of Eu can be calculated. The results are
Reaction Q, Mev Eu, Mev
s¡2s(d, p)Si29 6.246 7.66

Si 29 (d, p )Si 30 8.386 9.75
Si3º(d, p)Si 31 4.367 5.86
Si 29 (d, a)Al 27 5.994 6.66
Si3º(d, a)Al 28 3.120 4.19
11. When the reactions are added and when quantities which appear on both
sides of the resulting equation are canceled, the result is
0 16 + 4d ...... 6a + 33.224 Mev
or
a=
16.000000 + 4(2.014723) - (33.224/931.162) = 4 _003869 amu.
6
12. n p d a
(a) 14.96 Mev 11.95 20.27 17.15

(b) 14.19 11.18 18.84 14.65
13. (a) C 12 , 18.72 Mev; C 13 , 4.96; C 14, 8.17

(b) N 14, 10.55; N 15, 10.84; N 16 , 2.65
(e) 0 16, 15.60; 0 17 , 4.15; 0 18, 8.07
(d) Ne 20 , 16.91; Ne 21 , 6.75; Ne 22 , 10.36
14. (a) C11, 8.69 Mev; C 12 , 16.88; C 13 , 17.54
(b) N 13 , 1.95; N 14 , 7.54; N 15, 10.30
(c) 0 14, 4.61; 0 15, 7.35; 0 16 , 14.82; 0 17 , 13.60
(d) Ne 19, 6.36; Ne 20 , 12.87; Ne 21 , 13.03
16. 1.88 Mev 17. Al 28 - Si 28 = 4.65 ::\fev = 0.00499 amu.
18. From 5.7 Mev (9 = ,r) to 11.2 Mev (9 = O).
CHAPTER 12
l.
Nuclide Type of emission Nuclide Type of emission
(fa73 13-· y4s 13+

Nb96 13- Sc47 13-
c 8 121 13+ Agll0 {:3-
Jrl97 13- xe1a1 {:3-
Aul98 13- xe12a 13+
Br7s 13+ zn63 13+
2.
Nuclide Type of emission Nuclide Type of emission
Na23 stable Na22 13+

p32 /3- Fl8 13+
g¡a1 /3- Bel0 {:3-
Mg27 {3- He 6 13-
3. Decay process Q-Value, Mev Decay process Q-Value, Mev
He 6 ({3-)Li 6 3.215 F2º(/3-) N e20 7.052

Cl4({3-)Nl4 0.155 Na22(i3+)Ne22 2.841
Nlª(/3+)Cl3 2.222 Na24({3-)Mg24 5.531
Fl8(i3+)Ql8 1.671 A12s(/3-)Si2s 4.650
4. To a first approximation, the condition for a-decay is
M(A) > M(B) + M(a),

i.e., the mass of the initial nucleus must be greater than or equal to the sum
of the masses of the final nucleus and an alpha-particle. Actual!y, the final
nucleus always has sorne recoil kinetic energy E(B), so that
M(A) > M(B) + E(B) + M(a).

Since the a-particle must also have sorne kinetic energy E(a),
M(A) ~ Jf(B) + E(B) + M(a) + E(a).

5. At secular equilibrium there would be 6.65 X 10 12 atoros of Mn 56 . In gen-
eral, the number N of atoros of Mn 56 is given by the equation
dN
- = a->..N
dt '
where a is the constant rate of production. This equation has the solution
N(t) = r (1 - e->-').
At (secular) equilibrium N = a/X. At t = 20 hours, 99.5% of secular

equilibrium is reached.
6. The number of atoms of Au 198 at any time t is given by
N(t) = i (1 - e-A'),
where a = 10 10 sec~1, and X = 0.2567 days- 1 = 0.2972 X 10- 5 sec- 1 •

After 100 hr there will be 2.47 X 10 15 atoms of Au 198 ; after 10 days there
will be 3.47 X 10 15 atoms of Au 198 . At secular equilibrium there would be
3.76 X 10 15 atoms. The number of Hg 198 atoms is equal to the total number
of atoms formed minus the number of Au 198 atoms, i.e., to the number of
Au 198 atoms that have decayed. This number is
at-,;;a (l -e-Al) ,
under the assumption that the Hg 198 atoms are not affected by the neutrons.
(Actually, this is not the case, and the problem is more complicated.) At
100 hr there will be 1.13 X 10 15 Hg 198 atoms, and at 10 days there will be
5.17 X 10 15 atoms of Hg 1 98_
7. The irradiation would have to be continued for 11.7 days to rea.ch 95% of
secular equilibrium.
8. Fe 59 • 9. Ge 69 . 10. fj~: F 2º, F 21 ; fj+: Na. 20 , Na. 21 ; EC: none
11. fj~: CI , 8 ; fj+: none; EC: CI 36
36 35 12. K 4 º: {j-, EC; Sc 40 : fj+
13. X(fj~) = 0.585 X 10- 5 sec- 1 , X(fj+) = 0.285 X 10- 5 sec- 1,
X(EC) = 0.630 X 10- 5 sec- 1 , T112(fj~) = 32.8 hr,
T112(fj+) = 67.4 hr, T112(EC) = 30.5 hr
14. 1.25 X 109 y 15. He 5 -+ He 4 + n, Q = 0.95 Mev
Li 5 -+ He 4 + p, Q = 1.68 Mev
Be8 -+ 2He4, Q = 0.093 Mev
CHAPTER 13
l. Speed, Energy, Mev Energy, Mev Hr,
cm/sec
M/Mo (relativistic) (classical) gauss-cm, X 10- 5
108 1.00000 0.0274 0.0274 0.2073

5 X 108 1.00015 0.519 0.519 1.037
7.5 X 108 1.00034 1.117 1.117 1.555
109 1.00058 2.076 2.074 2.074
1.25 X 109 1.00090 3.245 3.241 2.594
1.50 X 109 1.00125 4.676 4.667 3.113
1.75 X 109 1.00170 6.293 6.277 3.634
2.0 X 10 9 1.00220 8.324 8.296 4.155
2.5 X 109 1.00351 13.09 12.96 5.201
3.0 X 109 1.00503 18.81 18.67 6.252
5.0 X 109 1.01420 52.94 51.85 10.51
1010 1.06078 227 207 22.00
2.0 X 10 1º 1.3425 1277 830 55.7
2. Since the reader has had practice in relativistic calculations in Problem 1,

the present problem may be treated classically. The error introduced is
small, as can be seen from the results of Problem l. Equation (13-4) can be
rewritten toread v(cm/sec) = [E(Mev)/2.O74]1 12 X 109 •
Energy, Mev v, cm/sec Energy, Mev v, cm/sec
1 6.94 X 108 6 1.70 X 109

2 9.82 X 108 8 1.96 X 109
4 1.39 X 109 10 2.20 X 109
5 1.55 X 109
3. The classical formulas may again be used, and the following resulta are
obtained; they may be compared with values listed in Table 13-1.
Alpha Energy Speed, Hr,

emitter Mev cm./sec, X 10-9 gauss-cm, X 10-5
p 0 21s 5.9981 1.701 3.526
p 0 216 6.7744 1.807 3.746
p 0 21s 7.365 1.884 3.906
p 0 214 7.680 1.924 3.988
8.277 1.998 4.142
9.066 2.091 4.335
10.505 2.251 4.666
p 0 212 8.7759 2.057 4.264
9.488 2.139 4.434
10.538 2.254 4.673
Po211 7.434 1.893 3.924
p 0 210 5.2984 1.598 3.313
4. When vis eliminated between Eqs. (13-3) and (13-4), the result is
E.,(Mev) = 4.824 X 10- 11 (Hr) 2,

with Hr in gauss-cm. Then Q is given by
Q = Eil + 4/25) - 2.10(1 - 2/25),

to a sufficiently good approximation.
Hr, gauss-cm, Hr, gauss-cm,

X 10- 5 E.,, Mev Q, Mev X 10- 5 E .., Mev Q, Mev
3.93 7.49 6.69 3.29 5.22 4.13

3.79 6.93 6.11 3.25 5.09 3.98
3.69 6.57 5.72 3.23 5.04 3.85
3.54 6.045 5.08 3.05 4.49 3.27
3.45 5.74 4.73 2.90 4.06 2.78
5.
Energy, Mev Range, cm
6.0 4.66
6.28 5.01
8.0 7.35
8.30 7.80
10.0 9.22
6. (a) (c)
Polonium Mean range of
Range in Equivalent
a-particle
isotope aluminum, thickness,
in air, cm
X 103 cm mg/cm2
p 0 21s 4.657 2.60 7.02
p 0 216 5.638 3.15 8.51
p 0 21s 6.457 3.60 9.72
Po2a 6.907 3.85 10.40
7.793 4.35 11.75
9.04 5.04 13.61
11.51 6.42 17.33
p 0 212 8.570 4.78 12.91
9.724 5.43 14.66
11.580 6.46 17.44
Po211 6.555 3.66 9.88
p 0 210 3.842 2.14 5.78
7, 8, 9. The decay schemes are given in Strominger, Hollander, and Seaborg,

"Table of Isotopes," Rev,. Mod. Phy,. 30, 585 (1958): Rn p. 802, u2 aa
p. 816, Am 2 U p. 826.
10. Nucleus Height of Coulomb barrier, Mev
Ne20 3.5
Sn112 10.0
Th2a2 14.1
12. 0.6 sec. 13. 5.401 Mev; 210.0483.
CHAPTEB 14
1. For an electron, the relationship
eHr =mu= p
holds, where p is the momentum. Then
Hr = p/e = (1/ec) (T
2
+ 2mo(:2T) 112•
The desired result is obtained when the values of the constants are inserted.
2. From the result of Problem 11, Chapter 6, it follows, since
p = eHr,
that
The desired result is obtained when the values of the consta.nts are inserted.
3. Range = 7 mg/cm 2 ; T = 65 kev; th112 = 0.6 mg/cm 2 •
4. The range is about 800 mg/cm 2 by inspection. (a) ~1.75 Mev. (b) ~l.77
Mev. The measured value of the maximum energy is 1.71 Mev.
5. In each case, the extrapolated endpoint of the plot should correspond to an
energy close to 1.71 Mev; the Hr-plot should extrapolate to a value close
to 7300 gauss-cm.
6. N 14 : 14.007517 amu, 0 17 : 17.004524, P 29 : 28.990988, S35 : 34.98014
7. 1.483 Mev
8. (a) 0.782 Mev, (b) 0.782, (e) 0.780, (d) 0.771
9. (a) Zn 64 , Ni 64 are stable; Ga 64 : fj+, Cu 65 : fj-, fj+, EC
(b) Cu 64 : endpoint energy and Q-value are both 0.58 Mev for /j--emission;
for fj+-emission, endpoint energy is 0.67 Mev, Q-value is 1.69 Mev;
Ga 64 (fj+): endpoint energy is 6.22 Mev, Q-value is 7.24 Mev.
10. The average energy is given by the ratio
{º TP(T)dT
'1'=-º----
fº P(T) dT
The integrations can be performed analytically, and the result is !, under

the given assumptions.
11. (a) 2.44 X 10 11 sec- 1, (b) 5.69 kev, (c) 0.32.
12. 1 watt.
13. 6000 kilowatts.
14. AI 25 : 3.2 Mev; CI 33 : 4.1 Mev.
15. The acherne is given on page 730 of the "Table of Isotopes," Strominger,
Hollander, and Seaborg, Rev,. Mod. Phys. 30, 585 (1958).
16. See page 644 of the "Table of Isotopes."
17. See page 671 of the "Table of Isotopes."
CHAPTER 15
l. The values of µ. obtained from the graphs should be close to those given in
Table 15-1, namely, 0.182cm- 1 , 0.157cm- 1, and 0.096cm- 1 for 'Y-ray
energies of 0.835 Mev, 1.14 Mev, and 2.76 Mev, respectively.
2. The values of µ. obtained from the graphs should be close to those given in
· Table 15-1, namely, 0.096 cm-1, 0.267 cm- 1 , and 0.478 cm- 1 for aluminum,
tin, and lead, respectively.
3. Photoelectric cross sections.
Energy ª"' 10- 2 " cm 2 /atom

n Mev Aluminum, Z = 13 Lead, Z = 82
0.1 5.108 0.00013 0.675

0.125 4.086 0.00017 0.869
0.25 2.043 0.00042 2.08
0.5 1.022 0.00129 6.31
1.0 0.511 0.00691 Zl.1
1.5 0.341 0.0214 75.7
2.0 0.255 0.0554 161
3.0 0.170 0.196 465
4.0 0.128 0.492 985
5.0 0.102 0.991 1782
4. Compton cross sections,

Energy
10-2 " cm2
a Mev (T u, ITa
0.025 0.0128 8.203 8.203 0.00

0.20 0.102 6.370 5.482 0.888
0.50 0.255 4.868 3.663 1.205
1.0 0.511 3.725 2.443 1.282
2.0 1.022 2.717 1.513 1.204
4.0 2.044 1.880 0.876 1.004
6.0 3.066 1.476 0.620 0.856
10.0 5.108 1.062 0.393 0.669
20.0 10.22 0.653 0.195 0.458
50.0 25.55 0.325 0.086 0.239
5. aK, 10-2 " cm 2

Energy, Mev
Aluminum Lead
1.0 o o
5.0 0.19 7.4
10.0 0.35 14.0
15.0 0.45 17.9
20.0 0.53 21.1
25.0 0.59 23.4
6. The values of " for copper should come out close to the values given in
Table 15--1, namely, 0.578 cm- 1 , 0.486 cm- 1, and 0.316 cm- 1 for energies
of 0.835 Mev, 1.14 Mev, and 2.76 Mev, respectively.
7. 3.35 Mev, 1.53 Mev, and 0.64 Mev, respectively.

8. 0.212 Mev, 0.248 Mev, 0.377 Mev, and 0.490 Mev, respectively.
9. 0.237 Mev, 0.294 Mev, and 0.347 Mev, respectively.
10. The energies are 86.5 kev and 196.5 kev, as can be seen from the following
table.
Electron energy, Shell and binding Energy of 'Y-ray,

kev energy, kev kev
32.9 K (53.4) 86.3

78.0 L (8.6) 86.6
84.7 M (1.9) 86.6
86.1 N (0.4) 86.5
143.0 K (53.4) 196.4
187.0 L (8.6) 196.6
194.7 M (1.9) 196.6
11. E, = (h11) 2/2Mc 2• With h11 in kev, and M in amu

4
E ( ) _ 5.38 X 10- (h11)2
, ev - M
12. 21.3 min: Mn 52 m; 5.8 d: Mn52 •

13, 14. See "Table of Isotopes," op. cit., p. 626.
15. See Evans, The Atomic Nucleus, p. 233.
CHAPTER 16
l.
Q,Mev Energy level, Mev Q, Mev Energy level, Mev
6.69 Ground state 3.98 2.71

6.11 0.58 3.85 2.84
5.72 0.97 3.27 3.42
5.08 1.61 2.78 3.91
4.73 l.96
2. (a)
Q,Mev Q, Mev Eu
-6.92
-7.87 1/:t 7
1
-8.57
-10.74
5.45
3.44
These results were obtained by using Eq. (11-9), with mass numbers rather
than with the precise values of the masses. (b) The Q-value -6.92 Mev
corresponds to the ground state. The value calculated from the masses
(Table 11-1) is -6.88 Mev. (e) The different proton groups correspond
to the ground state and to excited levels in B 12 at 0.95, 1.65, and 3.82 Mev,
respectively. (d) The threshold energies are 9.99 Mev, 11.36 Mev, 12.38
Mev, and 15.51 Mev, for the ground state and the three excited states,
respcctively.
3. 9.94 Mev. 4. (a) 1.64 Mev, (b) 2.95 Mev.
5.
Hr, Q, B 11 Leve!, Hr, Q, B 11 Level,
kg-cm Mev Mev kg-cm Mev Mev
448 9.235 Ground 245 1.937 7.298

399 7.097 2.138 190 0.667 8.568
338 4.776 4.459 171 0.309 8.926
322 4.201 5.034 156 0.045 9.190
266 2.477 6.758 151 -0.041 9.276
264 2.427 6.808
6. (a) 0.003499 amu. (b) 2.24 Mev.

(e)
Resonance Excitation Resonance Excitation
energy, energy, energy, energy,
Mev Mev Mev Mev
4.29 16.95 4.71 17.35

4.46 17.11 4.78 17.41
4.49 17.14 4.99 17.62
4.57 17.21 5.07 17.69
4.62 17.26 5.20 17.81
7. (a) 1.261 X 10 16, (b) 190.

8. (a) 3.5 X 1013 atoms of Au 198 , (b) 2.8 millicuries.
9. (a) 1.75 X 1014 atoms of Ta 182 , (b) 22 b.
10. (a) 4.94 Mev, (b) 4.95 Mev, (e) 4.05 Mev.
11. Product nucleus Binding energy of last neutron, Mev
Pb201 6.71
Pb2os 7.37
Pb200 3.87
12.
Energy of 'Y-ray, Mev Reaction
6.73 ± 0.008 Pb206(n, 'Y)Pb201

7.380 ± 0.008 Pb207(n, 'Y)Pb2os
13. The binding energies of the last two neutrons in Fe 56 are
(a) 20.24 Mev, from the mass difference,
(b) 20.54 Mev from the nuclear reactions.
14. (b) The reaction C 14 (p, n)N 14 has a Q-value of -0.628 Mev and has,
therefore, a threshold energy of 670 kev.
CHAPTER 17
l.
Total binding Coulomb . Coulomb energy
Nucleus energy, Mev
. energy, Mev Ratio: T otalb"mding energy
Be9 58.0 3.3 0.056

AJ27 225 30.0 0.13
Cu6a 552 152 0.28
Mo98 846 221 0.26
xe124 1045 362 0.35
w1s2 1461 549 0.38
u2as 1803 778 0.43
2.
Total binding energy, Mev,
Nucleus
from semiempirical formula
Be9 47
AJ27 224
Cu63 595
Mo98 841
Xel24 1039
w1s2 1457
u2as 1819
Atomic mass, amu Atomic mass from

3. Nucleus
(experimental) semiempirical formula
Be9 9.0149 9.0274

AJ27 26.9901 26.9915
Cu63 62.9486 62.9049
Mo98 97.9361 97.9473
Xel24 123.9458 123.9588
w1s2 182.0033 182.0226
u2as 238.1234 238.1321
4.
Total binding Coulomb Ra . Coulomb energy
Nuclide energy, Mev energy, Mev tio: T otal b"m ding energy
Ca40 342 66 0.19

Snl20 1020 294 0.29
Pb2os 1636 664 0.41
Binding energy from

semiempirical Atomic Atomic mass from
Nuclide
mass semiempirical formula
formula
Ca40 345 39.9753 39.9877

Sn120 1012 119.9401 119.9945
Pb2os 1630 208.0422 208.0688
5. Zo = 34.5A/(A 213 + 66.7)

Zo
A
(closest integral value)
'27 12
64 '27
125 47
216 73
6. This problem involves just a comparison of curves.

7. (a) Mass (Ra 226) = 226.10543, Mass (Em 222) = 222.09803, Mass He4
4.00387. Hence, the energy available for the a-decay of radium is 0.00343
amu or 3.2 Mev, as calculated from the semiempirical formula. The experi-
mental value is 4.88 Mev.
(b) Mass (Po 214 ) = 214.08049, Mass (Pb 21 º) = 210.07368, Mass (He 4) =
4.00387. Hence, the energy available is 0.00304 amu or 2.8 Mev, as
compared with the experimental value of 7.83 Mev.
(c) Mass (Hf 17º) = 170.00774, Mass (W 17º) = 170.01262. The mass
difference is negative so that, according to the semiempirical mass formula,
the tl-decay should not take place.
(d) Mass (Au 198) = 198.05121, Mass (Hg 198) = 198.04944. Hence,
0.01777 amu or 1.65 Mev should be available for the decay process, as com-
pared with an experimental value of 1.37 Mev.
The above results show that the semiempirical mass formula in the forro
used does not predict the decay' energies quantitatively. But it gives semi-
quantitative results which are often useful.
8. Mass (Fe 54 ) = 53.96011, Mass (n) = 1.00898, Mass (Fe 55) = 54.96045.
Energy available is 8.97 Mev, as compared with the experimental value of
9.28 Mev.
9. Mass from semi- Binding energy of

Nucleus empirical formula last neutron, Mev
Pb206 206.06428 --
Pb201 207.06712 5.72
Pb2os 208.06871 6.88
Pb209 209.07182 5.47
º º
The binding energies of the last neutrons in Pb 2 7 , Pb 2 8 , and Pb 2 9 , as ob- º
tained from (d, p} reactions have been found to be 6.70, 7.37, and 3.87 Mev,
respectively.
10. T = 2 M
1 (f¡)2 =
½b 10-20 Mev
11. (a} ASn = 28, (b) ASn = 1 Mev
CHAPTER 18
l. Neutron mass = 1.008984 ± 0.000003 amu.
2.
Energy, ev TºK v, cm/sec X, cm
0.001 11.6 4.37 104

X 9.04 X 10-;--8
0.025 290 2.19 105
X 1.81 X 10-8
1.0 1.16 X 104 1.38 106
X 2.86 X 10-9
100 1.16 X 106 1.38 107
X 2.86 X 10-10
104 1.16 X 108 1.38 108
X 2.86 X 10- 11
106 Ll6 x·1010 1.38 109
X 2.86 X 10- 12
10s 1.16 X 1012 1.28 X 10 1 º 2.79 X 10-13
1010 1.16 X 1014 2.99 X 101 º 1.14 X 10-14
3. (a} Q = 5.65 Mev. (b) 0°: 10.8 Mev; 90°: 8.5 Mev; 180°: 6.7 Mev.
4. H 2(d, n)He3 : E,.o = 2.45 Mev.
H 3(d, n)He4 : E,.o = 14.1 Mev.
C 12 (d, n)N 13 : Q = -0.280 Mev, the reaction has a threshold greater than
0.28 Mev and no neutrons would be ejected by deuterons of "zero" energy.
N 14 (d, n)0 15 : E,.o = 4.8 Mev. Li 7 (d, n)Be8 : E,.o = 13.4 Mev. Be9 (d, n)B 1 º:
Eno = 4.0 Mev.
5. Reaction EJ&, min Reaction E,., min
H 3 (p, n)He3 64 kev c12(p, n)N12 118 key

Li 7 (p, n)Be7 29 kev Na23(p, n)Mg23 8.3 kev
Be 9 (p, n)B 9 21 kev v111(p, n)Cr51 0.58 kev
6. (a) nv = 105 neutrons/cm 2sec. (b) nv = 1.17 X 105 neutrons/cm 2 sec.

(e) 0°:104.4, 25°: 100, 50º: 96.0, 100°: 89.4, 200º: 79.4.
7. Average number Slowing-down Moderating

Element ~ of collisions power ratio
F 0.102 178 2.74 X 10-5 56.7

Mg 0.0807 226 0.021 8.2
Bi 0.0096 1890 0.0024 0.3
8. (a) 3.2 curies, (b) about 14,000 counts/min.

9. (a) The derivation is directly analogous to that given in Section 18--5
except that in spherical coordinates the operator d2/ dx 2 is replaced by
(d 2 /dr2 ) + (2/r)(d/dr).
(b) The general solution of the differential equation in (a) is n = (Ae-•IL/r)
+(Be+rlL/r). One boundary condition is that the total neutron current
through a small sphere of radius a surrounding the source has a limit Q as
the radius of the sphere approaches zero. The second condition is that n
must be finite everywhere, so that B = O.
10. Bragg angle for Bragg angle for

Energy, ev Energy, ev 1
first order reftection first order reftection
1 11° 15' 10 3° 32'

3 6° 28' 30 2° 2'
5 5° O' 50 1° 35'
1
11. Energy, ev t.E, ev Energy, ev t.E, ev
1 0.023 10 0.74
3 0.121 30 3.8
5 0.26 50 8.2
12. E
t.t, µ sec/min
1 ev 10 ev 100 ev 1 kev 10 kev
0.5 0.014 ev 0.44 ev 14 ev 0.44 kev 14 kev

0.1 0.0028 ev 0.09 ev 2.8 ev 88 ev 2.8 kev
0.05 0.0014 ev 0.04 ev 1.4 ev 44 ev 1.4 kev
13. (a) rr, = 26.3 X 10- 24 cm 2 , (b) "•d = 21 X 10- 24 cm 2 •

14. I/Io = e-Nzv, = e-Nz(v,+va> = e-Nzv, X e-Nna = 0.575X 0.685 = 0.394.
15. "" (0.025 ev) = 147 b, in good agreement with the measured value 150 ± 7b.
16. 29 collisions.
CHAPTER 19
72 72 72
l. (a) aoZn -> a1Ga -> a2Ge (stable).
90 90 90 90
(b) a6Kr9°-> a1Rb -> ast:lr -> 39Y -> 40Zr (stable).
2 92 92 92 92
(e) 36Kr9 -> 31Rb -> 3sSr -> 39 Y -> 40Zr (stable).
bo7m
97 97 97 97 41 N
01
(d) 36Kr -> 31Rb ->3sSr ->39Y ->40Zr /' 1 97 .?'42:'.\fo97(stable).
',.41Nb
105 105 105 105 105
(e) 42::\fo -> 43Te -. 44Ru -> 45Rh ---> 45Pd (stable).
126 126 126
(f) 50Sn -> 51Sb -> s2Te (stable).
144 144 144 144 144 44
(g) 54Xe -> 55Cs -> s6Ba -> s1La -> 5sCe ---> 59Pr1 --->
144
50Nd (stable).
156 156 156 156
(h) 61Pm -> 52Sm -> 5aEu -> 54Gd (stable).
91 139
2. Sr : 0.025 gm; La : 0.031 gm.
3. (a) 205 Mev; (b) 211 Mev.
4. 179 Mev.
5. Plot the function
A(t) = (~~) (O.OSSe-0.693t/0.43 + 0 _2416-0.693111.52

+ 0 _2136 -o.H93t/4.51 + 0 _1666 -0.6931122.0 + 0 _025e-0.6931/55.6)
133.Se-I. 611 + 376.6e-0·4561 + 332.9e - - + 259.5e - 0-03151
0 2421
+ 39 _0e-o.01251,
where t is in seconds. The delayed neutron activity is here normalized to
1000 units at t = O.
6. (a) 45. (b) 34, 35, 36.5, 35.5, 37, 37, 38. (e) The nuclides with the greater
values -0f Z 2/ A, namely, U 232 , Pu 238, Am 241 , and Cu 242, would be expected
to have greater spontaneous fission rates than the other nuclides.
7. Binding energy of Thermal fission
Nuclide Fissionability
added neutron, Mev cross section, b
Th221 7.01 yes 1500

u229 7.58 yes ?
Np236 7.05 yes 2800
Np2a1 5.05 no or? = 10-2
Np2as 6.35 yes 1600
Pu24t 6.88 yes 1080
Pu242 5.32 no or? < 0.2
Am241 5.21 no or? 3
Am242 6.45 yes 6400
cm240 6.70 yes 20,000
Since the values of the activation energy are not given in the text, it is
a.ssumed that an excitation energy of 6 Mev or greater indicates fissionability,
while values of a.bout 5 Mev indicate doubtful fissiona.bility. The results
are in good agreement with the known values of the thermal fission cross
section.
10. 170 Mev
CHAPTER 20
l. U235 is consumed by fission at the rate of about 1 kg/day; the total rate of
consumption is about 1.18 kg/day. The ma.ximum conceivable re.te of pro-
duction of Pu 239 would be about 1.1 kg/day.
2. When pure U 235 is the fuel, the fa.st fission effect and the resonance escape
probability are both unity, and
where
Nu O'av
x= --·
NM O'aM
The condition for criticality in an infinitely large assembly is k.,, = 1, or

x = l/(71 - 1). For U 235, 71 = 2.07 and x = 0.93 is the condition for criti-
cality. For graphite and U 235, then,
3. The thermal neutron absorption cross section of beryllium is 0.01 b. Then
4. The thermal neutron absorption cross section of water is 0.66 b/molecule.

Then
5. The thermal neutron absorption cross section of D20 is 0.0093 b. Hence,

4
Nu (0.93):/)10- = 1.2 X 10 -6.
ND20 =
6. Migration a.rea, cm2 Migration length, cm
Moderator
e 1708 41.3
Be 326 18.1
H20 36.8 6.1
D20 15,510 124.5
7. Moderator Nu!NM M 2, cm 2 M,cm Re, cm Critica} ma.ss, kg
e 7.0 X 10- 6 1579 39.7 1247 56.5

Be 1.60 X 10-5 304 17.4 547 16.8
H2O 1.05 X 10-3 36.4 6.0 188 12.3
D2O 1.48 X 10-6 14050 118.5 3723 13.1
Note: The values of the parameters used in this problem are a poor choice
from the etandpoint of efficient reactor design. They were used only
to make possible very simple calculations.
9 (a) d n2
·
2
dr
+ ~ dn + k.. -
r dr M2
1 = 0
'
10. !!_ =
no
(~)ª;
..-
(fi/no)RP
(fi/no), =
8
3..- ·
no
(c) A = ..-/R;
11. Reaction Q-value, Mev Reaction Q-value, Mev
Cl2(p, 'Y)Nl3 1.944 Nl4(p, 'Y)QIS 7.346

Nl3(,6+)Cl3 2.221 O1sc13+)Nl5 2.705
Clª(p, 'Y)Nl4 7.542 NlS(p, 'Y)Cl2 4.961
Total: 4p -. 213+ + a, Q = 26.72 Mev.

CHAPTER 21
l. From Eq. (21-7), it follows that H = 2..-(q/m)-1 n.
The values of (q/m) are proton: 9.581 X 103 emu/gm; deuteron: 4.794 X 103
amu/gm; a-particle: 4.823 X 103 emu/gm.
It follows that, from these values, H = 6.55 X 10-4 n, for protons, H =
1.31 X 10-4 n, for deuterons, H = 1.30 X 10-4 n, for a-particles.
2. Particle H, gauss Energy, Mev
proton 7.86 X 103 7.4 Mev

deuteron 1.572 X 104 14.8 Mev
a-particle 1.56 X 104 14.7 Mev
3. Different sets of parameters can be devised with the aid of Eq. (21-6) and
the relationships of Problem l. When the value of the energy is inserted
into Eq. (21-6), a relationship is obtained between H and R. If R is chosen,
H is determined from this relationship, and nis then given by the appropriate
relation of Problem 1. If the frequency is chosen, H is determined, and
R is then obtained from Eq. (21-6).
4. The kinetic encrgy of a relativistic particle is given by
T = (M - Mo)c 2 = Mo (M ) 2
Mo - 1 e.
But T = 108 q V, if q is in emu and V in volts. When the two expressions

Cor Tare combined,
M - 1) e2 = 10sqV,
llfo ( Mo
or
8
M
Mo
1 + _!_Q___
c Mo
2
...!LV = 1 + aV.
lt is also known that
so that
~ =
e
[1 - (llfº)2]112
M
= (2aV)112[l
1
+ (aV/2))112.
+ aV
Then, since llfV = qH R, it follows that
and
HR = (
~oc (2aV) 112 1 + ª: )
1/2
For deuterons
H 2R2 = 4.172 X lO~V + 1.1127 X 10-5 V2 •
For H = 15,000 gauss and R = 80 in., the deuterons have an energy of
210 Mev.
~
5. (a) eV = e dt = ew<l>o cos wt. (b) ~ ew<l>o. (e) cr · (d) _e_·
r 2w 4wR
ec<l>o
(e) 2 rR = ecHR.
6. Average energy per turn "" 400 ev; final energy "" 108 ev.
INDEX
INDEX
Absorption,. of alpha-particles, 307-316 Annihilation radiation, 379, 414, 419,
of beta-particles, 346-350 422
of gamma-rays, comparison of theory Antineutrino, 378-381
and experiment, 414-418 Artificial nuclear disintegration, 258-
of gamma-rays, experimental, 395- 279
400 Artificial radioactivity, 283-299
Absorption coefficient of beta-particles, Artificial radionuclides, 295-297
347 Aston, 184,198,203,204
Absorption coefficient of gamma-rays, Atomic and molecular radii, 13, 14
measurement of, 395-400 Atomic hypothesis, Dalton's, 5-7
Absorption cross sections for neutrons, Atomic masses, 218, 219, 224, 278
594-599 Atomic number, 12, 78-82
Acceleration of charged particles, 268- Atomic spectra and theory, 135-176
269, 679-705 Atomic structure, 51, 67, 84, 85, 128,
Accelerator (see special types: Cyclo- 139-176
tron, Van de Graaff, etc.) Atomic theory, chemical foundations
Actinium series, 251, 297 of, 3-17
Activation cross section, 580, 594-599 Atomic volume, 14
Activation energy, fission, 629 Atomic weight, 6, 8, 10, 12, 79
Activation method for neutron detec- Atomic weight sea.les, physical a.nd
tion, 564, 580 chemical, 218
Age of earth, 296 Atom smashing, 277
Age of minerals, 256-257 Auger electrons, 287
Allowed orbits, 140-141 Avogadro number, 13, 19, 30
Alpha-decay, 304-335 Avogadro's Hypothesis, 7-8, 13
relation between disintegration con-
stant and energy, 326-335 Barn, 460
theory of, 326-335 Balmer series, 135, 136, 137, 143
Alpha-disintegration energy, 320-321, Barrier permeability, 330
323 Bateman equations, 243
Alpha-particles, absorption of, 307- Becquerel, 32, 231
316 Bergmann series, 138
charge of, 34 Beta-decay, 340-391
charge-to-mass ratio of, 306 allowed and forbidden transitions,
distance of closest approach to nu- 362-373
cleus, 63 average energy, 356
ionization by, 307-316 comparative half-life or ft-value, 370
long-range, 319-322 comparison of theory and experi-
nature of, 34 ment, 363-371
range of, 307-316 conservation of angular momentum,
scattering of, 52--62, 80, 718-723 360, 361
stopping power, 314-316 conservation of energy, 358
velocity and energy of, 304-307 conservation of sta.tistics, 361
Alpha-particle spectra, 304, 319-322 disintegration constant, 364
Alpha-rays, 32-34 endpoint energy, 345, 348, 353, 364
Alternating gradient synchrotron, 701- energy levels (decay schemes), 371-
705 373
Angstrom unit, 86 Kurie (Fermi) plot, 365-368
759
760 INDEX
line spectrum, 344, 345 Breit-Wigner formula, 461-463, 473,

matrix element, 363, 368 589
m9mentum distribution curves, 356, experimental evidence, 469
365 BreDl88trahlung (braking radiation),
parity nonconservation in, 381-388 350
Sargent curves, 362 Bubble counter, 47
selection rules, 369
theory of, 358-371 Canal raye, 31
Beta-decay of free neutron, 359 Carbon-nitrogen cycle, 669--671
Beta-particles, absorption of, 346-350 Carrier methods, 606
energy loes in matter, 346-348 Cathode raye, 20-24, 100
magnetic rigidity, 343 Cathode-ray tube, 21, 28, 67
range of, 346-350 Center-of-mllBII system., 565
range-energy relations, 350-352 Cerenkov detector, 47
straggling, 350 Chadwick, 62, 189--191, 259, 266-268
velocity and energy of, 340-346 Chain-reacting pile (see Nuclear re-
Beta-raye, 32-33 actor), 562, 639--666
Beta-ray spectrometer, 345 Charge independence of nuclear forces,
Beta-spectra, 344, 345, 352-358 511, 512-517
Betatron, 224, 692-697 Charge symmetry of nuclear forces,
Bevatron, 697-700 417, 512, 513
BFa counter, 563 Charge-to-m11811 ratio (e/m), 23
Binding energy formula, 537-542 Chemical combination, laws of, 3-5
Binding energy (nuclear), 192, 218- Chronotron, 205
, 224, 508-510 Cloud chamber, diffusion, 46, 47
Bismuth, total croBB section for neu- Cloud chamber (Wilson), 25, 44-46
trons, 592, 593 Cloudy crystal hall model, 549
Blackbody, 87, 89, 93, 95, 98 Coalescence coefficient, 548
Blackett, 259 Cockcroft-Walton volta.ge multiplier,
Bohr, correspondence principie, 153 269, 679--681
eliptic orbits, 149--154 Coherent scattering croBB section, 596
magneton, 187 Coincidence techniques, 423
postula.tes, 139--140 Collateral radioactive series, 295, 297
quantum conditions, 139, 168-169 Collective nuclear model, 542-544, 631
theory of atomic spectra a.nd atomic Collision croBB section, 458
structure, 139--156 Combining volumes, law of (Gay-
theory of the compound nucleus, LUBBac's), 6-9
451-458 Combining weights, 5
Bohr-Wheeler theory of fission, 507, Comparative half-life (see ft-value)
626-631 Compound nucleus, 260-261, 451-457,
Boltamann's consta.nt, 95, 98 482
Born, derivation of relation between limitations of, 484-487, 544
mase and velocity, 117-121 Compton effect, 127-131, 165, 170
Bose-Einstein statistics, 196 Compton electrons, 424
Bound states (levels) of nuclei, 454 Compton scattering, 400-401, 405-410
Brackett series, 137, 144 Conservation of mase, law of, 3, 5
Bragg's Law, 72-76 Coolidgetube,67,68
Bragg x-ray spectrometer, 74, 128 Cosmic raye, 217, 496
Branched decay, 245, 252, 296 Cosmotron, 269, 697-700
Breeding of fissionable material, 664- Coulomb energy, nuclear, 510, 511-
666 514, 537
INDEX 761
Coulomb excitation, 483-484 Displacement law of radioactivity, 197,

Coulomb force law, 54, 63, 64 252
Critica! size of a nuclear reactor, 639, Dissociation encrgy (see Separation
647, 660-664 energy)
Cross section (see also type: Absorp- Dulong and Petit, rule of, 11, 12
tion, Scattering, etc.), classical,
for scattering by a free electron, Effective radius of the nucleus for
70 alpha-emission, 328
for Compton scattering, 406-407 Effective range, 521, 522
of fissile materials, 608, 609 Effective resonance integral, 652
for neutron-induced reactions, 584- Eigenfunction, 171
594 Eigenvalue (proper value), 172
for a nuclear reaction, 457-463 Einstein, 85, 100-131
for resonance reactions, 461 Einstein, photoelectric equation, 100-
as a target area, 458-460 101
Crystal, lattice spacing of, 7~ Einstein, quantum theory of light, 127
Crystal spectrometer (neutron), 587- Einstein, relativity theory, 113-124
589 Elastic scattering, 448
Curie, Madame, 33 of neutrons, 466, 565-574
Curie (unit), 253-254 Electricity, conduction in gases, 20
Curie and Joliot, 190, 283, 284 Electrolysis, Faraday's laws of, 19
Curved-crystal gamma-ray spectrom- Electromagnetic theory of light, 91
eter, 418 Electron, 19, 24, 25, 39
Cyclotron, 269, 683-688 angular momentum (orbital), 140
angular momentum (spin), 156
Dalton's atomic hypothesis, 5-7 charge of, 24-30
Danger coefficient, 597 "classical radius" of, 71
de Broglie's hypothesis, 165-166, 168 magnetic moment of, 156, 187
de Broglie wavelength, 165, 462 reduced mass, 143
Decay chain (see fission decay chain) rest mass, 30
Decay scheme, 324 variation of mass with velocity, 109
Beta-decay, 371-373 Electron configurations of the atoros,
Gamma-decay, 439-441 160-163
Definite proportions, law of, 3, 5 Electron diffraction, 165
Delayed neutrons emitted in fission, Electron emission, 285-290
620-623 Electron-positron mass and specific
Dempster's mass spectrometer, 207 charge comparison, 420
Deuterium, 214, 272, 273 Electron-positron pair formation and
Deuteron, singlet and triplet states, 520 annihilation, 410-414
Deuteron stripping reactions, 491 Electrons, number per atoro, 70-72
Differential scattering cross section, Electron synchrotron, 692-697
722 Electron volt, 126
Diffusion coefficient of neutrons, 581 Electroscope, 37-38
Diffusion cloud chamber, 46, 47 Electrostatic generator, 681-683
Diffusion constants of moderators, 584 Elements, alphabetical list of, 713-714
Diffusion of thermal neutrons, 580- Elements, origin and abundance of,
584 674
Dirac electron theory, 156, 410-413 Emission spectra, 135-139
Disintegration constant, 231, 234-239 Emissivity, 86
in a.lpha.-deca.y, 326, 330, 332 Energy balance, Q, in a nuclear reac-
in beta-decay, 364 tion, 262
762 INDEX
Energy density of therma.l ra.día.tion, discovery of, 606-608

90, 97 energy distribution oí neutrons in,
Energy leve\s (see Nuclear energy 623-625
levels), concept of, 99 energy relea.se in, 625-626
Energy leve! dia.gra.m, (a.tomic), 147, neutron emission in, 61~23
148,152,153 sponta.neous, 608
for x-ra.ys, 164 terna.ry, 608
(nuclear) for N 15 , 498 theory oí, 626-631
Energy levels, of Al28 , 488-489 Fissionability of heavy nuclei, 630
oí hydrogen a.tom, 142, 147, 152 Fission counter, 563
oí light nuclei, 497-501 Fission <leca.y cha.ins, 615
oí light nuclei a.nd the cha.rge inde- Fission fragments, 618
pendence of nuclear forces, 512- Fission neutrons, energy distribution,
517 623-625
oí Mg 25 , 490 Fission products, 614-619
oí mirror nuclei, 501 ma.ss a.nd energy distribution, 616-
Energy production in sta.rs, 666-671 619
Energy relea.se in fission, 514, 62l>--626 Fission yield curves, 616-617
Equiva.lent electrons, 157 Fitzgera.ld contra.ction, 108, 109
Equiva.lent thickness in mg/cm 2, 316 Force, definition oí, 124
Equiva.lent weights, 5, 10 between two nucleons, 518-523
Eta. (for nuclear reactor), 647-649 Four-fa.ctor formula., 647
Ether a.nd problem oí absolute velocity, Fra.nck-Hertz experiment, 146
105-106 Free a.tom cross section, 596
Excess rea.ctivity of. a. nuclear reactor, Frequency-modula.ted cyclotron, 269,
663 689-691
Excha.nge forces, 525 Funda.mental consta.nts oí physics, 30,
Excita.tion energy 'of compound nu- 715-716
cleus, 452, 454-455 Funda.mental equa.tion oí radioactive
Excited nuclei, 322, 448-449 <leca.y, 232
Excited sta.tes of the a.tom, 146 Fusion, 669
Exclusion principie (Pa.uli), 157, 195,
533 Ga.mma-decay, 425--441
Expectation values, 170 energy levels a.nd deca.y schemes,
439-441
Fa.rada.y (unit), 19 interna! conversion, 425-428
Fa.stchopper, 590,594 mean life (theoretical), 433
Fa.st fission factor, 646 multipole radiations, 430
Fermi, 290 nuclear energy levels, 428-433
theory of beta-decay, 359-371, 374, nuclear isomerism, 433--441
375 relations between lifetime and en-
Fermi-Dira.c sta.tistics, 195, 361 ergy, 4~38
Fertile material, 665 Gamma.-ra.ys, 71, 395-441
Fick's la.w oí diffusion, 581 a.bsorption coefficients (experimen-
Fine structure, 149, 152 ta.!), 395-400
Fine-structure consta.nts, 153 absorption coefficients (theoretical),
Fine structure oí spectral lines, 149 414-418
Fission, 606-631 in a.lpha.-deca.y, 322-326
cross sections a.nd thresholds, 608- in beta.-deca.y, 355
614 Compton scattering, 400, 405-410
delayed neutrons in, 62<Hi23 in decay oí AI 28, 488
INDEX 763
electron-positron pair formation by, Ionization chamber, 39, 42
400, 41()--414 Ionization potential, 146, 160--163
energy measurement, 418-425 Ions, electrolytic and gaseous, 19-21
flux of, 396 !sobar, 198, 285
from (a, p) reactions, 449-450 stability .properties of, 541
interaction with matter, 40Q--402 Isobaric spin, 515, 517
photoelectric absorption, 400, 402- Isobaric transformations, 287
405 Isomerism, nuclear, 434-436
time of emission, 423-425 Isotone, 530
x-rays, 395 Isotopes, 31,184, 197-224
Gamma-re.y spectrometer, 418-422 and hyperfine structure, 186
Gas amplification, 41 in nuclide chart, 299
Gas multiplication, 41 Isotope tables, 297
Gay-Lussac's law (combining vol- Isotopic abundances (see Sta.ble iso-
umes), 7-9 topes)
Geiger and Marsden, 58-62 Isotopic abundances and masses, 203-
Geiger-Mueller counter, 39, 41, 43 209
Geiger-Nuttall rule, 326-329, 331, 362 Isotopic doublet measurements, 204
Geiger point counter, 43 Isotopic masses, 31, 198, 203-209, 218-
"Good geometry" measurements, 397, 224
423, 585 Isotopic spin (see Isobaric spin)
Gram-molecular volume, 12
Ground state of atom, 142 Journal abbreviations (for references),
Ground state of nucleus, 323 717
Groups of particles emitted in (a, p)
reactions, 450-451 K-capture, 373
K,. lines, wavelengths of, 76-81
Half-life, determination of, 237-239 K-radiations (x-rays), 69
Heisenberg, 169, 191, 525 Kinetic theory of gases applied to
Heterogeneous nuclear reactor, 656- neutrons, 580--581
660 Kirchhoff's law, 87, 90
Homogeneous nuclear reactor, 648-656 Klein-Nishina cross sections for Comp-
Hyperfine structure of spectral lines, ton effect, 406
186 kT neutrons, 577
Kurie (Fermi) plot in beta-decay, 365-
Impact para.meter, 55 367
Independent particle model, 533
Inelastic scattering, 448 Laboratory system in neutron scatter-
of deuterons, 489 ing, 565
of neutrons, 475 Lawrence and Livingston (see Cyclo-
of protons, 479 tron)
Infrared radiation, 86 Laws of radioactive transformation,
Interaction between two nucleons, 518- 229-258
523 Lee and Yang, nonconservation of
Interference fringes, 107 parity, 383-388
Interferometer, 107-109 Level width, 455
Internal conversion, 352, 423, 425-428 Linear accelerators, 700--701
Invariant form of physical theories, Linear harmonic oscillator, 94
104, 110--112 Liquid-drop model and fission, 627
Iodine, decay of isotopes of, 289 Liquid-drop model of the nucleus, 537-
• Ionization by alpha-pa.rticles, 307-318 542
764 INDEX
L,. lines, wavelengths of, 78-81 low-lying levels of, 501

Long-range alpha-particles, 319-322 Moderating ratio, 573
Lorentz-Fitzgerald contraction hy- Moderators, 562, 644
pothesis, 109, 112 slowing-down lengths in, 659
Lorentz transformation, 111, 114 Molar volume, 12
derivation of, 114-116 Molecular clocks, 110
Lyman series, 137, 144 Molecular weights, 8, 9, 11, 12-14
Molecules, 7, 8
Macroscopic cross section, 459 size of, 12--14
Magic numbers, 527, 530, 531 Monochromatic emissive power, 88
Magnetic lens spectrometer for beta- Monochromatic energy density, 90
particles, 345-346 Moseley's law, 76-82, 159
Magnetic moment and nuclear-shell Multichannel differential pulse height
model, 535 selector, 313
Magnetic moment of the nucleus, 186 Multiple proportions, law of, 4, 6
Magnetic resonance accelerator (cyclo- Multiple scattering, 53
tron), 683-688 Multiplet structure of spectral linea,
Magnetic spectrograph, for alpha-parti- 139,155
cles, 305 Multiplication factor of a nuclear re-
for beta-particles, 340 actor, 647, 648-656
Mass defect, 221
Mass doublet technique, 204-205 Natura! nuclear time, 453
Mass-energy balance in nuclear reac- Natural radioactive series, 248-252
tions, 261-265 Natural radioactivity, 197-198, 229-
Mass-energy relation, 124-127, 222 254
experimental proof of, 270 Negative energy states of the electron,
Mass number, 185 411
Mass spectrograph, 198, 203-204 Negatron (electron), 283
Mass spectrometer, 207-209 Neptunium series, 296
Mass spectrum, 204, 206, 207 Neutrino, 374-381
Mass standard, 223-224 detection of, 377-381
Mass synchronometer, 206 existence, problem of, 358-359
Mass, variation with velocity, 113-124 hypothesis, 358-359
Maxwellian distribution of thermal mass of, 374-376
neutrons, 574-577 properties of, 359-361, 378
Mean free path of neutrons, 581 Neutron, angular momentum of, 191
Mean life in radioactive decay, 237-239 beta-decay of, 359
Mean lifetime of a state of the com- discovery of, 189-191, 265-268
pound nucleus, 455 mass of, 266-268, 278
Mean range, 310, 317, 318, 320, 327 wavelength of, 462
Mendeléef, periodic law of, 14-15 Neutron balance in a nuclear reactor,
Mesons, 496, 523 639, 644-648
artificial production of, 691 N eutron binding energies, 492
Meson theory of nuclear forces, 525 Neutron capture resonance, 469
Michelson-Morley experiment, 107- N eutron cross sections, absorption,
110, 112, 113 473, 578, 597
Migration area, 662 activation, 580, 597
Millikan, 25-30, 101 capture, 474
Mirror nuclei (also mirror nuclides), 370 scattering, 467, 594-595
Coulomb energy of, 512 total, 545-549
energy difference of, 512-514 transport, 582
INDEX 765
N eutron cycle, 644-648 in liquid-drop model, 537-542
Neutron emission in fission, 623-625 nuclear stability, 510-512
Neutron flux, 577-579 nucleon-nucleon scattering, 518-523
Neutron physics, 559-599 range of, 192, 509, 521
N eutron scattering, coherent and in- saturation of, 508-510, 524
coherent, 596 status of problem, 523-526
potential, 467 Nuclear fuels, 642
by protons, 521 Nuclear gyromagnetic ratio, 194
Neutron transmission, mea.surement Nuclear isomerism, 434-436
of, 585-594 Nuclear isomers, decay of, 435
Neutrons, detection of, 274, 563-565 Nuclear level spectra, complexity of,
diffusion coefficient, 581 480
mean free path of, 581 Nuclear magneton, 188, 194
production of, 274, 559-563 Nuclear ma.sses from nuclear reactions,
reactions induced by, 274-276, 465- 276-279
477 Nuclear models, 526-549
slowing-down of, 565-574 Nuclear photodisintegration (photo-
thermal, 574-580 effect), 275, 491
N eptunium, 292 Nuclear potential, square well, 520
N eptunium series, 295 Nuclear power, 664-666
Nonconservation of parity in beta- Nuclear ra.dii, from alpha-decay, 334
decay, 383-388 from neutron scattering, 475-476
Nuclear atom, 51-64 Nuclear ra.dius, 517-518
Nuclear binding energies, 508-510 Nuclear reactions, 258-279, 447-501
Nuclear charge, 62, 78, 82 analogy to chemical reaetions, 260
Nuclear chemistry, 192, 276-279 balance of ma.ss and energy in, 261-
Nuclear decay schemes, 371-373 265
Nuclear disintegration (transmuta- classification of, 463-465
tion), 189-191, 258-279 cross sections for, 457-463
Nuclear emulsions, 46, 47, 313 and excited states of nuclei, 447-452
Nuclear energy-level diagrams, 324, experimental, 463-465
325,372,488,498,500 induced by a.lpha-particles, 258-261,
Nuclear energy levels, from alpha- 265-268, 277-484
decay, 322-326 induced by deuterons, 272-273, 487-
from beta-decay, 371-373 491
bound, 454 induced by gamma-rays, 276, 491-
from nuclear reactions, 447-452, 495
487-491, 496-500 induced by neutrons, 274-275, 465-
related to gamma-ray emission, 425- 477
441 induced by protons, 269-272, 477-
virtual, 454, 500 484
Nuclear energy sources (see also Nu- of light nuclei, 463, 496-501
clear reactors and Thermonu- threshold energy of, 265
clear reactions), 637-674 at ultra-high energies, 496
Nuclear fission (see Fission), 606-631 Nuclear reactors, 562, 639-666
as a source of energy, 637-638 classification and description of, 642-
Nuclear forces, 64, 192, 223, 508-526 644
charge independence, 511, 512-517 control of, 663
charge symmetry, 512-517 critical size of, 639
electrosta.tic (coulomb), 192, 513 critica! size of thermal reactor, 660-
exchange, 525 664
766 INDEX
excess reactivity, 663 Oscillator potential, 533

heterogeneous, 656-660 Oxygen isotopes, abundance of, 217
homogeneous, 648-656
multiplication factor of, 648-656 Packing fraction, 218-224
neutron balance in, 641, 644 Pair formation, 41G-414
one group model, 660-663 Pair spectrometer, 422
power and breeding, 664-666 Parity, 174-176, 196
power level, 643 Parity, nonconservation in beta-decay,
reflector, 664 381-388
thermal, 644-648 Partial width, 456, 461
Nuclear shell model, 526-537 Particle tracks in cloud chamber, 46
Nuclear spectroscopy, 229 Paschen series, 137, 144
Nuclear spin, 186, 193-194 Pauli (see Neutrino hypothesis)
Nuclear stability, 217, 510 Pauli exclusion principie, 157, 195, 533
Nuclear statistics, 195 Periodic system of the elements, 14-17,
Nuclear structure, problem of, 183, 76-82, 156-159
508-549 Permeability of simple barrier, 330
Nuclear transmutation, 189-191 Pfund series, 137, 144
Nucleon, definition of, 192 Phase stability, 689
Nucleon scattering experiments and Phosphors, 35, 36
nuclear forces, 518-523 Photodisintegration, of beryllium, 492
Nucleus, characteristics of, 62-64 of the deuteron, 491-493
constitution of, 183-196 of nuclei, 275
proton-electron hypothesis, 183- Photoelectric absorption of gamma-
185 rays, 402-405
proton-neutron hypothesis, 190- Photoelectric effect, 99-102, 124
191 Photofission thresholds, 614
properties of angular momentum Photographic emulsions, 46, 313
(total), 193 Photomultiplier tube, 36-37
angular momentum (orbital), 193 Photon, 101,125,126
binding energy, 193, 218-224 Photoproton effect, 494
density, 183 Photoneutron sources, 56o-562
electric quadrupole moment, 194, Photoneutron thresholds, 492, 614
533-536 Physical atomic weight scale, 218
magnetic moment, 194 Physical constants, values of, 715-716
parity, 174-176, 196 Physical theory, meaning of, 110-112
radius, 64, 334, 476 Pile oscillator, 597
spin, 185, 193 Planck, 96-99, 139
statistics, 195 Planck's constant, 96, 98, 140
Nuclide chart, 297-299 Planck's distribution law, 96
Nuclide, definition of, 209 Planck's quantum theory of thermal
radiation, 96-99
Odd-even effect in nuclei, 538 Plutonium, 291, 292, 638
Oil-drop experiment (Millikan), 26-30 Polonium-beryllium neutron source,
Oppenheimer-Phillips process, 491 560
Optical model for nuclear reactions, "Poor geometry" measurements, 424
544-549 Positive rays, 31, 198-202
Orbital electron capture, 285--290, 361, Positron, 283, 284, 413 "
373 Positron emission, 285-290
Orbits (electron) in Bohr theory, 140- Potential barrier in alpha-decay, 328-
142, 144, 149-154, 157 329
INDEX 767
Potential scattering, 467 Quantum theory of radiation, 84-102,

Potential well, depth and range, in 139
alpha-decay, 328 Q-value of a nuclear reaction, 262
for nuclear forces, 520 Q-values and excited states of nuclei,
Probability, in quantum mechanics, 448-451
170 Q-values and nuclear masses, 279
in radioactivity, 233
Prompt neutrons, 620 Radiation, detection and measurement
Proper function, 171 of, 35-47
Proper value (eigenvalue), 172 Radiation, thermal, 86-99
Proportional counter, 39, 41, 42 Radiation width, 456
Proton, 30, 185 Radiative capture of neutrons, 275,
Proton, properties of, angular momen- 466--477
tum (spin), 187, 193 the 1/v law, 469
magnetic moment, 187 Radiative capture of protons, 477
mass, 185, 278 Radioactive transformation (see also
Proton groups in (a, p) reactions, 448- Radioactivity), basis of theory
450 of, 229-234
Proton-electron hypothesis of the nu- fundamental equation oí, 232
cleus, 183-185 Radioactivity, activity or intensity,
Proton-neutron hypothesis of the nu- 231
cleus, 191-192 age of roinerals, 256-257
Proton-proton forces, 511, 522, 526 artificial, 229, 283-299
Proton synchrotron, 697-700 Bateman equations, 243
Prout's hypothesis, 184-185 branching decay, 252
decay and recovery curves, 230
Quadrupole moment, 194, 535-536 detection coefficient, 234
Quantization of electron orbits, 140, detection and measurement of, 35-47
149, 154 discovery of, 32
Quantization of nuclear angular mo- disintegration constant, 231-233,
mentum, 193 234-239
Quantum conditions, Bohr, 139, 168- displacement law of, 197
169 equilibrium, 244-248
Sommerfeld and Wilson, 149 half-life, 237-239
in wave mechanics, 168-169 importance in nuclear physics, 229
Quantum mechanics (wave mechan- mean life, 237-239
ics), 155, 164-171 of mixture, 238-239
Quantum number, 140 natural (see Natural radioactivity)
azimuthal, 150 number of active species, 283
correct, 154 number of atoms, 231-233
magnetic, 154 probability of disintegration, 233
orbital, 154 qualitative description, 32-35
principal, 150, 158 series (artificial), 295-297
radial, 150 series (in fission), 615-616
spin (electron), 156 series (natural), 248-252
spin (nuclear), 185, 193 series (theory), 239-242
Quantum of energy, 85, 99 of "stable" nuclides, 214
Quantum theory and photoelectric ef- successive transformations, 239-244
fect, 99-102 transient equilibrium, 246-247
Quantum theory of atomic structure, of thorium and thorium-X, 230--231
84, 139-145 units, 253-254
768 INDEX
of uranium I and uranium-X, 230- Sargent curves, 362

231 Saturation of nuclear forces, 508-510,
Radiochemistry, 285, 606 524
Radioelements, separation of, 606 Scaler, 44
Radionuclides, artificial, 285-290 Scattering, of alpha-particles, 52--62
Radium-beryllium neutron source, 560 of gamma-rays and x-rays, 69, 128-
Range, of alpha-particles, 307-316 131, 405-410
of beta-particles, 346-350 of neutrons by nuclei, 475, 566-574,
extrapolated, 309 594-599
mean, 310 of neutrons by protons, 521, 525
Range-energy curves, for alpha-parti- of protons by nuclei, 479
cles, 317-319 of protons by protons, 522-523, 525
for beta-particles, 350-352 Scattering cross section, 459, 466, 467,
Rayleigh-Jeans' formula, 94-96 475
Reciproca! proportions, law of, 4, 6-7 macroscopic, 572
Reference system, 110-111 for neutrons, 594-599
Relative abundance of nuclides, 208- Scattering length, 521, 522
218 Scattering mean free path, 581
Relative stopping power, 316 Schroedinger wave equation, 156, 166-
Relativity, general theory of, 113 168
Relativity, special theory of, 104-131 in alpha-decay, 330
kinetic and total energy in, 124-125 Born, 170
mass-energy relation, 124-127 example of solution, 171-176
momentum in, 124-125 for the deuteron, 518
postula.tes, 113 Scintillation counter, 36
variation of mass with velocity, 113- Scintillation method, 36, 60
124 Scintillation spectrometer, 37
Reproduction factor (see Multiplica- Secular equilibrium, 245-246
tion factor) Selection rules, atomic, 153
Resolving time of a counter, 423 in alpha-decay, 270
Resonance cross sections, 461-463, 598 in beta-decay, 368
Resonance energies in nuclear reac- in gamma-decay, 430-431
tions, 451 Self-shielding, 658
Resonance escape probability, 646, 649 Semiempirical binding energy (mass)
Rest mass, 30, 109, 121-122 formula, 537-542
Rest mass and rest energy, 125 Separation energy, atomic, 146
Roentgen, 20, 31 nuclear, 453
Russell-Saunders coupling, 175 Separation of radioelements, 606
Rutherford, 34, 53-58, 189, 258-259, Sharp series, 137
319,425 Shell model, 430, 526-537
Rutherford (unit), 253 Slow chopper, 586
Rutherford alpha-particle scattering, Slowing-down length, 659
differential cross section, 722 Slowing-down power, 572
experimental proof, 58--62 Soddy, 197,229,252
formula, 718-722 Solenoidal magnetic spectrometer, 345-
theory of, 52-58 346
Rydberg constant, 78, 135-138, 143, Sommerfeld fine-structure formula,
144,186 149, 153
Rydberg formula, 136 Space and time, interrelationship be-
Rydberg-Ritz combination principie, tween, 113
155 Spallation, 496, 690
INDEX 769
Special theory of relativity, 104-131 mean free paths in moderators, 584

Specific ionization, 310, 311 velocity distribution, 574-577
Spectra, absorption (atomic), 138 Thermal nuclear reactor, 644-664
of alpha-particles, 322 Thermal radiation, classical theory of,
of beta-partícles, 344-345, 365-368 89-93
emission (atomic), 85, 135-139, 147 emission and absorption of, 86-89
of gamma-rays (see Gamma-decay) Planck's quantum theory of, 96-99
Spectral emíttance, 88 Rayleigh-Jeans' Jaw of, 95
Spectral line, 138, 148 spectral distribution of, 86-89
Spectral notation, 157 Wien's law of, 92-93
Spin, of electron, 156 Thermal utilization, 647, 649-652
of neutron, 191 Thermionic effect, 24
of nucleus, 185-187, 194 Thermoelectrons, 67
of proton, 191 Thermonuclear reactions (see Stars,
Spín-orbit coupling, 533 energy production), 666-671
Spontaneous fission, 608 carbon-nitrogen cycle, 669-671
Square-well nuclear potentíal, 520 controlled, 671-674
Stable isotopes of the elements, 209- proton-proton cycle, 668
218, 528 Thermopile, 88
Stable nuclides, 209-218, 529 Thomson, 22, 23, 24, 51-53, 100, 198-
Stars, energy production in, 666-671 202
Stationary states of an atom, 139, 140, Thomson atom, 51-53
145-149 Thomson parabolas, 200-202
Statistical model for nuclear reactions, Thomson scattcring, 402
544 Thorium-C, 323-326
~ Statistics, in beta-decay, 361 Thorium series, 250-251
of nuclei, 195 Threshold energy, for fission, 608-614
Stefan-Boltzmann constant, 87 for a nuclear reaction, 264
" Stefan-Boltzmann law, 87, 89, 97 Total cross section, 459
Stokes' law of fall, 25, 28, 29 Total neutron cross sections, 469-471
Stopping power, 314-316 measurement of, 584-594
Straggling of alpha-partícles, 310 Total width, 456
Straggling of beta-partícles, 350 Transient equilibrium, 246-247
Strippíng reaction, 491 Transmutatíon by alpha-particles,
Strong-focusing synchrotron, 702-705 258-261, 2ti8-269
Superficial density, 348 by deuterons, 272-273
Symmetrical wave functíon, 106 by neutrons, 273-275
Symmetry and conservation laws, 381- by photons, 275-276
388 by protons, 269-272
Synchrocyclotron, 689-691 Transparency of barrier, 330
Synchrotron, 269, 690 Transport cross section, 582
Transport mean free path, 582
Thermal column, 563, 577 Transuranium elements, 290-295, 606
Thermal cross sections, 612 Tritium, 275
Thermal diffusion Jength, 583 Tunnel effect, 330
Thermal neutrons, 562-584
absorption, 578 Uncertainty principie, 169, 170, 455
cross sections of moderators for, 584 Uranium, and discovery of radioactiv-
diffusion, 580-584 ity, 32
energy distribution, 575-57{¡ effect of neutron bombardment on,
flux distribution, 577-578 290-292
770 INDEX
nuclear constants, 641 Weak interactions in beta-decay, 381,

series, 248, 250, 297 383
separation of isotopes of, 638 Whole number rule (Aston), 184
Wien's displacement law, 92-94, 96
Valence, 10 Wilson cloud chamber, 25, 44
Van de Graaff electrostatic generator,
269, 681-683, 692 X-rays (see also Gamma-rays), absorp-
Velocity selector, crystal spectrometer, tion coefficient, 68, 69
587-589 Characteristic spectra, 76-82, 159,
mechanical (fast), 590, 594 164
mechanical (slow), 586 diffraction and Bragg's law, 72-76
time-of-flight, 589-590 discovery, 20, 31-32
Virtual levels of a nucleus, 454, 500 interaction with matter, 67-69
ionization chamber for detection of,
Wave mechanics, and alpha-decay, 330 42
and atomic spectra, 154-156 scattering (Thomson), 70-72, 128
basic ideas, 164-171 sorne properties of, 67-70
examples and results, 171-176 spectrometer, 74-76
Wave nature of matter, 164
Wave number, 78, 81, 136 Yang, symmetry laws, 381
Wave theory of light and ether, 105-
106 Zeeman effect, 139, 156
fnn1ed 1n H,"mg K•'mJ, h:,,

'11,,ng T,u C'heuntt Prmun1 Co. Lid

Nuclear Physics (Irving Kaplan)

Uploaded by

Copyright:

Nuclear Physics (Irving Kaplan)

Uploaded by

Document Information

Copyright

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Nuclear Physics (Irving Kaplan)

Uploaded by

Copyright:

NUCLEAR

ADDISON-WESLEY PUBLISHING COMPANY

EIGHTH PRINTING 1977

A complete and unabridged reprint of the original American textbook,

Copyright© 1962 by Addison-Wesley Publishing Company, lnc.

I have tried to write an elementary, yet coherent, account of nuclear

of nuclear physics and begins with the chemical foundations of atomic

them individually because of the lack of space. I have, however, exer-

PART l. THE BACKGROUND OF NUCLEAR PHYSICS

CHAPTER l. THE CHEMICAL FOUNDATIONS OF ATOMIC THEORY 3

CHAPTER 2. ATOMS, ELECTRONS, AND RADIATIONS 19

CHAPTER 3. THE NUCLEAR ATOM 51

CHAPTER 4. X-RAYS AND ATOMIC STRUCTURE. 67

CHAPTER 5. THE QUANTUM THEORY OF RADIATION 84

CHAPTER 6. THE SPECIAL THEORY OF RELATIVITY 104

CHAPTER 7. ATOMIC SPECTRA AND ATOMIC STRUCTURE 135

PART 11. THE NUCLEUS

CHAPTER 8. THE CoNSTITUTION OF THE NucLEUS 183

CHAPTER 9. ISOTOPES 197

9-2 Positive-ray analysis and the existence of isotopes 198

CHAPTER 10. NATURAL RADIOACTIVITY AND THE LAWS

10-1 The basis of the theory of radioactive disintegration 229

CHAPTER 11. ARTIFICIAL NUCLEAR DISINTEGRATION . 258

11-1 Transmutation by alpha-particles: alpha-proton reactions 258

CHAPTER 12. ARTIFICIAL RADIOACTIVITY 283

CHAPTER 13. ALPHA-DECAY 304

CHAPTER 14. BETA-DECAY 340

CHAPTER 15. GAMMA-RAYS AND GAMMA-DECAY 395

CHAPTER 16. NUCLEAR REACTIONS • 447

CHAPTER 17. NUCLEAR FORCES AND NUCLEAR 8TRUCTURE . 508

PART III. SPECIAL TOPICS AND APPLICATIONS

ÜHAPTER 18. NEUTRON PHYSICS . 559

18-1 The production of neutrons 559

CHAPTER 19. NUCLEAR F1sSION . 606

CHAPTER 20. NuctEAR ENERGY SouRcEs 637

CHA.PTER 21. THE AccELERA.TION OF CHARGED PARTICLES 679

THE CHEMICAL FOUNDATIONS OF ATOMIC THEORY

1-1 The laws of chemical combination. The experimental information

Compound Percent Percent

1. Calcium oxide Calcium 71.47 Oxygen 28.53

1-2 Dalton's atomic hypothesis. Dalton's atomic hypothesis was

3. Chemical compounds are formed by the union of atoros of different

According to the third postulate, n, m, p, and q are ali small integers;

1-3 Avogadro's hypothesis and the molecule. One of the difficulties

1 volume of hydrogen + 1 volume of chlorine

It follows that if elements in a gaseous state combine in simple proportions

must be simply related. Dalton assumed that equal volumes of different

1-4 Molecular and atomic weights of gaseous elements. The relative

Hydrogen 0.08988 2.000 2.ol3 2.016

* Corrected for deviations from ideal gas laws.

Molecular weight = 2 X density of gas relative to that of hydrogen,

Relative Weight of carbon

1-6 Atomic weights of nongaseous elements. The method of relative

Element Atomic weight Element Atomic weight

Aluminum 26.98 Lead 207.21

is the approximate atomic weight of copper. Chemical analyses of com-

Carbon: 0.77 X 10-8 cm Cesium: 2.6 X 10-8 cm

1-8 The periodic system of the elements. As information about the

Mendeléeff's general conclusions have been verified ancl extended.

Grou- I 11 III IV V VI VII Vlll o

55 Ca 56 !Ja 57-71 72 Hf 73 Ta 74W 75 Re 76 0a 77 Ir 78 Pt

Nitrous-oxide 63.63 36.37

ATOMS, ELECTRONS, AND RADIATIONS

2-1 Faraday's laws of electrolysis and the electron. It was seen in