DEUTERIUM REMOVAL UNIT FOR THE MUCAP EXPERIMENT

I. Alekseev1, T.I. Banks2, O. Fedorchenko1, V. Ganzha1, P. Kravtsov1,

P. Kammel3, С. Petitjean4, V. Trofimov1, A. Vasilyev1 and M. Vznuzdaev1
1. Introduction
A cryogenic unit for highly efficient isotopic separation of hydrogen has been designed and
built at the Petersburg Nuclear Physics Institute (PNPI), Russia. The unit’s main task is to
produce isotopically pure hydrogen of deuterium concentration below 1 ppm, for use in the
MuCap experiment conducted at the Paul Scherrer Institute (PSI), Switzerland. The goal of
MuCap is to measure the rate of the basic electroweak process of nuclear muon capture by
the proton from the hyperfine singlet ground state of muonic hydrogen. MuCap aims to
achieve a 1% measurement of the capture rate, which will in turn provide the most precise
(~7%) estimate to date for the weak nucleonic charged-current pseudoscalar form factor.
MuCap is a high-precision measurement that requires ultra-pure and deuterium-
depleted hydrogen. The issue of chemical purity was solved by introducing a circulating
hydrogen purification system, which is described in detail elsewhere [2]. To meet the iso-
topic purity requirements, the experiment first attempted using deuterium-depleted gas sup-
plied by an outside vendor [3]. However, the deuterium content was still too high to be
completely satisfactory, strongly affecting the interpretation of the first experimental re-
sults. It was therefore decided to specially design and build a device for isotopic purifica-
tion of hydrogen. The principles of the device’s design and the results of its operation are
described in this report.

2. Principles
Selection of the isotopic purification method
The most common method for separating hydrogen isotopes is to distill water and then per-
form electrolytic decomposition, followed by hydrogen isolation with a palladium filter.
This method was inappropriate for our situation, for several reasons. First, it simply could
not achieve the desired level of isotopic purity, as the best commercial hydrogen samples
available contained more than 1 ppm of HD. Second, it was desirable to integrate a new
isotopic separation unit into the existing hydrogen recirculation system, in order to ensure
that the gas was continually cleaned of any possible deuterium contamination. The only
method suitable for an in situ isotopic purification of hydrogen is cryogenic distillation.

1
Petersburg Nuclear Physics Institute (PNPI), 188350, Gatchina/Russia
2
University of California, Berkeley, CA 94720-7300, USA.
3
University of Illinois at Urbana-Champaign, UIUC, IL 61801, USA
4
Paul Scherrer Institut (PSI), CH-5232 Villigen PSI/Switzerland
Principles of cryogenic distillation and main column requirements
The principles of distillation (i.e. rectification) are well-known. Distillation under cryo-
genic conditions conforms to the same rules as high-temperature rectification, which is
prevalent in the chemical industry. The method exploits the difference in the saturation va-
por pressure of mixed species above the surface of the liquid mixture. It can be regarded as
a multi-step fractional distillation that uses a column filled with a special facility for in-
creasing the phase contact surface, where the facility can be either a set of perforated plates
or packing, the latter being more suitable for a column with a relatively small inner diame-
ter. A condenser located in the top of the column converts the vapor mixture to liquid
form; this returning liquid is called “reflux.” The vapor can also be partially siphoned from
the top of the column as a purified product of the process. The reflux drains down along
the column and moistens the packing. The amount of liquid suspended on the packing is
the column holdup. The lower end of the column is equipped with a reboiler. The sepa-
rated mixture boils in the reboiler to form the vapor, which rises upward along the column
and interacts with the counterflow of draining reflux. The liquid is saturated with the high-
boiling component while the gas is saturated with the low-boiling component. In our case,
the mixture is of regular hydrogen (“protium”, or H2) and “deuterohydrogen” (HD). Deute-
rium (D2) molecules are very rare in the source gas and can be ignored.
The column was designed to separate an initial mixture injected into the device
through a feed port, with pure H2 (whose HD content is at least 30 times lower than that of
the feeding gas) being the distillation product. The HD component has to be accumulated
in the reboiler with the possibility to discharge it.
In order to find the optimal operating mode for the column, a number of studies and
verifications were carried out, including investigations of the pressure drop along the col-
umn, assessment of the holdup, and the cleaning performance at different modes.

HETP concept
The separation factor ! defines an elementary separation effect achieved at one contact of
liquid and vaporized hydrogen. The ideal separation factor is a ratio of saturation vapour
pressures above pure components, and depends on temperature: a higher temperature means
a lower separation factor. By passing through the separation column, liquid and vapour are
brought into repeated contact, thereby multiplying the elementary separation effect. The
downflowing liquid hydrogen is thus steadily enriched in its high-boiling component (in
our case, HD) while the rising vapour is steadily depleted.
In accordance with the concept of the Equilibrium Theoretical Plate [3], the Height
Equivalent to a Theoretical Plate (HETP) is the main performance characteristic of a sepa-
ration column. HETP is the height of a column part left by liquid and gas flows which are
in mass-transfer equilibrium (the deuterium concentration in the liquid is α times higher
than in the gas). A lower HETP corresponds to a higher column separation power.
The separation ratio of the column is the ratio of relative concentrations of the com-
ponent of interest at the bottom and top of the column:
X / (1 ! X Bottom )
SR = Bottom . (1)
X Top / (1 ! X Top )
The separation ratio is associated with the separation factor by the Fenske [4] equation:
SR = ! N , (2)
where N = the number of theoretical plates.

The separation factor for hydrogen components

In addition to its isotopic composition, hydrogen has spin-isomeric components [5]. H2
molecules can be grouped according to their nuclear spin orientations: In the first group, the
proton spins are aligned in the same direction (triplet state, orthohydrogen), while in the
other group, the proton spins are aligned in opposite directions (singlet state, parahydro-
gen). Ordinarily, transitions between the ortho and para states is relatively rare, so H2 can
be regarded as a mixture of two distinct components with rather different physical and
chemical properties. The ratio between the ortho and para states is about 3:1 at standard
temperature and pressure, but the para form dominates at low temperatures.
For ortho-para and H2-HD compositions, we can treat factor ! as the ideal one.
Thus, for ortho-para separation the factor ! can be shown to be
P
" o ! p = sat . para . (3)
Psat .ortho
Here Psat . Para and Psat .Ortho are the pressures of saturated vapour above pure parahydrogen
and pure orthohydrogen, respectively.
Further, for the deuterium-protium case,
Psat . H 2
" HD ! H 2 = , (4)
Psat . HD
where Psat .H 2 and Psat . HD are the pressures of saturated vapour above pure H2 and pure HD.
The temperature dependencies of separation factors for HD-H2 and for the ortho-para hy-
drogen system are shown in Figure 1and Figure 2 [6].
Verification of the column’s separating power was complicated by the absence of a
fast, on-line deuterium concentration analysis, and also by the fact that the residual deute-
rium content of the depleted product can lie below the detection limit of any available
analysis methods. This compelled us to find another approach: namely, the use of gas
chromatography (described below) for ortho-para hydrogen analysis. Gas chromatography
proved to be a relatively fast and simple way of evaluating column performance during test-
ing.
The lower ! (compare Figure 1 and Figure 2) makes it possible to measure the
concentrations for ortho and para modifications in probes taken from each control point of
the column (condenser, reboiler, and feeding line). Using the Fenske equation and the data
from the chromatographic analysis, it is easy to determine the HETP for the column.
Knowledge of HETP enables calculation of concentration profiles along the column, using
a mathematical model.
 1200 1.8
Alpha D-H
Saturated vapor pressure, kPa
HD, kPa
1000 H2, kPa 1.7

800 1.6

Alpha
600 1.5

400 1.4

200 1.3

0 1.2
20 21 22 23 24 25 26 27 28 29 30 31 32
Temperature, K
Figure 1. Temperature dependencies of saturated vapor pressures and separation factor for
H2-HD system

1200
1.060
p-H2, kPa
1.055
Saturated Vapor Pressure, kPa

1000 o-H2, kPa

Alpha 1.050
800
1.045

Alpha
1.040
600
1.035
400
1.030

1.025
200
1.020
0
1.015
20 21 22 23 24 25 26 27 28 29 30 31 32
Figure 2. Temperature dependencies of Temperature, K pressures and separation factor for
saturated vapor
ortho-para hydrogen system
3. Design
Layout
A simplified layout of the deuterium separation unit (DRU) is presented in Figure 3. His-
torically, the DRU has been part of a chemical purification system [2] gas scheme, and has
a through sequence of enumeration for the units and devices. The separation column of
2.2 cm inner diameter and 155 cm overall packing height is the main part of the unit. It
consists of two sections connected by flange coupling. The column is encased in a vacuum
jacket containing upper and lover attachments (adaptors) that support various hardware and
vacuum terminals. The jacket and adaptors comprise a common vacuum volume. The vol-
ume is fixed on an adjustable tripod which provides precise vertical positioning of the col-
umn. All working parts of the device are placed inside the vacuum volume.
Cold head

PT12
CV8
V 42 Feed flow

M FC 4

H10
V43 Pure protium
TT10
M FC 5

Conden ser
Counterflow
Radiation shield heat exchanger

Liquid distributor
Packed column
TT12
Reboiler
TT11

H11
PT11 PT10
Needle valve
V41

LNT2
Centering screws

Exhausting
pump

Vacuum
system

Figure 3. Simplified scheme of the DRU

The column has a condenser and a reboiler connected to its upper and lower part, respec-
tively. A COOLPOWER 140T (Oerlikon Leybold Vacuum GmbH5) cryogenerator with a
maximal cooling power of 20 W at 20 K is used for the cold operations.
A control system provides the necessary algorithms for column operation in all
modes. Four modes were tested during our experiments: (1) “feedthrough” with purging of
the deuterium-enriched mixture, (2) “feedthrough” without purging, (3) continuous circula-
tion, and (4) so-called Rayleigh depletion. The differences between these modes are con-
sidered below. In all modes, the system operates two mass-flow controllers (MFC4 and
MFC5 in the scheme), two heaters (H10 and H11), and reads signals from the temperature
and pressure sensors.

Condenser
The condenser is a closed, conical, bimetallic volume (Figure 4) whose upper part is made
of copper plate. The inner part of this plate is processed in the form of lamellas to develop
the heat transfer surface area. The vertical orientation of the lamellas facilitates the down-
ward motion of the condensed hydrogen drops. The conical part has a flange for connect-
ing to the column that is welded to the copper plate by electron-beam. The condenser has
two connections for conducting product hydrogen and its pressure to an upper point of the
differential manometer PT11.

Figure 4. Condenser
The cold head tightly contacts with the outer surface of the copper plate through a thin in-
dium foil to provide better heat flow. The cooled part of the condenser is mounted in the
upper adaptor and therefore in the common vacuum insulation of the column. A counter-
flow heat exchanger is mounted around the condenser. The heat exchanger first cools the
feed flux using the product flux. The feed flux is further cooled and ultimately liquefied by
5
http://www.oerlikon.com/vacuumsystems/
the cooling power of the cryogenerator. The cooled feed flux is delivered to the center of
the column just above the liquid distributor.

Reboiler
The reboiler is used to evaporate the separating mixture. Its rather complicated arrange-
ment (Figure 5) is needed to provide steady and controllable boiling [7]. This unit mainly
consists of a central tube and a massive copper muff equipped with an electric heater. Up-
per and lower collectors are welded to the central tube and connected with the muff by four
lateral tubes (from above) and the manifold (from below). The connection between the steel
parts and the copper body of the muff is provided by electron-beam welding.
Several longitudinal apertures are drilled in the muff around the large central hole.
An electric heater is coiled around its body and a PT100 temperature sensor (TT11) is fixed
on it. The sense of its structure is shown on the scheme of fluxes (Figure 6).
The central tube keeps most of the liquid hydrogen collected in the bottom of the
column. This hydrogen forms the steady lift of the liquid which can be precisely measured
by the differential manometer (upper and lower arms of the manometer are connected to the
lower part of the column and to the lover collector of the reboiler, respectively).

Flange
Lateral tube

Central tube

Upper Heater
collector
Discharge
line

Manometer
line

Manifold Lower
collector

Figure 5. Sectional drawing of the reboiler

 Column
Vapor
Upper Liquid
collector
Manometer
line
Central tube
Lateral tube
Differential
manometer
Temperature
sensor Steady lift

Boiling zone

Heater

Manifold

Figure 6. Scheme of fluxes in the reboiler

The boiling zone is removed to the apertures of the muff. The forming vapour moves from
the muff upward through four lateral tubes to the upper collector and further to the column.
The liquid is delivered to the boiling zone from the central tube through the manifold of the
lower collector under the action of hydrostatic pressure of the steady lift.

Packing
The column is filled with specially designed spiral prismatic random packing made of
0.2 mm stainless steel wire (Figure 7). The packing is intended to provide a maximal sur-
face for phase contact. Given the 2×2 mm size of the prismatic springs and the 560 ml vol-
ume of packing in the column, it provides the surface area of 1.95 m2. The choice of stain-
less steel was dictated by the purity requirements for hydrogen. The packing surface was
pickled with nitric acid to develop its roughness and improve wettability by liquid hydro-
gen.
This was the first test of this particular packing in a cryogenic distillation column.
A similar type of packing, but of bigger size (3×3 mm), was tested earlier in PNPI water
distillation columns of 80 mm inner diameter, and exhibited good separation characteristics
(HETP = 2.5-3 cm) [9]. A similar packing made of stainless steel (Heli-Pak and Coil Pack)
tested in a cryogenic distillation column of small diameter gave an HETP value about 5-
6 cm [10]. Thus, the HETP value of 5 cm was put in the design according to a conservative
approach.
 Figure 7. Packing

Operating modes
As mentioned earlier, there are several possible DRU operating modes for producing deute-
rium-depleted hydrogen. Four modes were tested during the experiments: “feedthrough”
with purging of the deuterium-enriched mixture, “feedthrough” without purging, continu-
ous circulation, and so-called Rayleigh depletion.
During Raleigh depletion the column was fully primed with liquefied hydrogen.
After the priming, the hydrogen was released through the product line. Natural hydrogen
was used only for preliminary tests. The accumulation of required product was carried out
from pre-depleted protium.
In the “feedthrough” modes the mass-flow controllers were used to operate the feed
(MFC5) and the product (MFC4) flows. The rate of the purging flow was adjusted with a
manual needle valve using the bubble flow meter.

4. Analysis
Cromatographic analysis of ortho-para composition
A method for quickly determining the deuterium content in deuterium-depleted hydrogen
was not available during the experiment. It was essential, however, to be able to rapidly
estimate column performance so that adjustments could be made. A chromatographic
analysis of ortho-para isomers of hydrogen was proposed as a method of indirect estimation
of the column’s separating power.
A chromatographic device equipped with a Al2O3-filled chromatographic column
immersed in liquid nitrogen was used for the analysis. The low temperature is essential for
separating the isomers, which have temperature-dependent differences in adsorption affin-
ity. Neon was used as the carrier gas, instead of the more conventional helium, in order to
provide a greater difference in thermal conductivity between the separated species and the
carrier gas.
Measurements of the concentration of ortho and para isomers in the top and bottom
regions of the column make it possible to estimate its separation performance, and to calcu-
late the concentration profile for deuterium. An example of a chromatogram for samples
taken from the top and bottom part of the column is shown in Figure 8.

30
Hydrogen composition:
Top: Para=72.5% Ortho=27.5% Top
25 Bottom: Para=5.0% Ortho=95.0% Bottom

20
Signal. mV

15

10

1500 2000
2500 3000 3500
Time,
Figure 8. Chromatogram of separated ortho- andsparahydrogen. (Column pressure: 1.2 bar;
Reboiler power: 10 W; Operating mode: total reflux)
A specially prepared synthetic mixture of H2, HD, and D2 was prepared for the column test.
The mixture was prepared by diluting natural hydrogen with pure deuterium and subse-
quently passing the mixture through the catalytic column under heating conditions. The
column is filled with palladium catalyst immersed in granulated silica gel. The mixture
thus obtained is analyzed by the same chromatographic device as used for ortho-para com-
position measurements, albeit with minor readjusting of the regime [11]. The chroma-
togram of the converted mixture is shown in Figure 9.
 16

14
H2=31.2%
12
Signal, mV

10 HD=47.3%
8
D2=21.5%
6
4
2
0

-2
100 120 140 160 180 200 220 240
Time,mixture
Figure 9. Chromatogram of synthethic s of hydrogen isotopes

Analysis at large tandem accelerator

Direct measurements of low deuterium concentrations in samples were carried out by a
large tandem accelerator group at the Institute of Particle Physics, ETH-Zurich, Switzer-
land. Measurements of the probe samples were performed using a new 200 kV accelerator
designed for isotope analysis in Zurich [12]. Its unique ion source provides extremely low
backgrounds (e.g. hydrogen ions from walls, etc.).

5. Results of tests and production

Pressure drop and the liquid holdup studies
For the liquid holdup measurement, a known quantity of hydrogen was batched into a col-
umn. From the observed liquid level and reboiler dimension, the quantity of hydrogen resi-
dent on the packing was calculated. It was assumed that the liquid film in the overhead
condenser contributed insignifically to the column inventory. Figure 10 shows the relation
between the pressure drop across the column and the vapor flow rate of natural hydrogen at
various pressures. The flooding regime was not achieved due to the relatively large (for
this design) inner column diameter. The relatively low pressure drop can be explained by
the moderate range of vapor flow rate. The packing holdup presented in Figure 11 had an
expected high value in the full range of vapor fluxes.
 1.8
1.2 bar
1.6 1.5 bar
3 bar
1.4 4 bar
Pressure drop, mbar
5 bar
1.2 Linear fit
1.0

0.8

0.6

0.4

0.2

0.0
40 50 60 70 80 90 100 110 120
Vapor flow rate, mol/h
Figure 10. Pressure drop across the column

30

28 Experiment
Linear fit
26
Packing holdup, %

24

22

20
18

16

14

12
40 50 60 70 80 90 100 110 120 130
Vapour flow rate, mol/h
Figure 11. The holdup as a function of the flow rate
The first DRU run was carried out using natural hydrogen with 126.9±1.9 ppm of deute-
rium. The initial gas volume was 181.6 liters. The column was filled by liquid hydrogen in
large excess, and after 1 hour of operation the excess gas (84 liters) was withdrawn from
the column bottom through the discharge line. This explains the low final deuterium con-
centration in the bottom (Table 1). Withdrawn gas was sufficiently enriched to signifi-
cantly change the average concentration in the column. The negative deuterium concentra-
tion in the column’s top (−1.9 ppm) indicates the precision of the measurement resolution.
(The method relies on extrapolating a measurement of deuterium content from a sample of
natural hydrogen, and gives low accuracy for close-to-zero concentrations.) Thus the small
negative value indicates effectively zero value for the HD admixture.

Probe location Ortohydrogen

HD concentration, ppm
concentration, %
Source gas (measured) 126.9±1.9 75
Column top −1.9 14.6
Column middle 32.2
Column bottom 56.2 85
Average concentration calcu-
7.7 53.5
lated from the mass balance
Table 1. Natural hydrogen run results

The experimental data for the ortho-para separation were fitted with the Fenske equation
(Figure 12). The average HETP calculated on the basis of tests with various vapor flow
rates is 2.2 cm (Figure 13), which is comparable with the best world results [12].

70

60 !o-p(P)74.4
Experimental points
Separation ratio

50

40

30

20

10

0
1 2 3 4
Figure 12. Fenske equation fit for total reflux mode
5
Column top pressure, bar
 3.2
1.2 bar
3.0 1.5 bar
2.8 3.0 bar
4.0 bar
2.6
HETP, cm

2.4
2.2
2.0
1.8
1.6
1.4
1.2
30 40 50 60 70 80 90 100 110 120 130
Vapour
Figure 13. HETP flow rate,
at various mol/h
flow rates

Natural gas separation in “feedthrough without purging” mode

During this test, a large amount of natural hydrogen (~ 6000 liters) was passed through the
column to increase the average deuterium concentration and thereby obtain measurable
concentrations in the middle of the column.

Production of deuterium-depleted hydrogen

The production procedure was carried out using all possible operating modes in order to
determine which one was optimal. The results of the experiments are shown in Table 2.
“Circulation” mode corresponds to the “Feedthrough without purging” regime, with the dif-
ference that the gas was continuously circulated through the column system for two days.
The goal of this test was mainly to clean the circulation system of all traces of deuterium.
In addition, a number of deuterium-depleted probes were collectedso-called “zero
protium” samples. The expected HD concentrations (Figure 14) were calculated using the
special simulation program based upon the aforementioned mathematical model with
known HETP and total packing height of the column.
 “Feedthrough” “Feedthrough” Rayleigh’s
Mode Circulation
no purging with purging depletion
Reboiler power, W 20.5 19.0 12.0 20.0
Pressure, bar 1.5 1.5 1.5 2.0
Feed flow, l/min 0.81 0.81 1.25 0.55
Top 67.8 69.4 – 64.8

Ortho-hydrogen, % Middle 69.2 74.6 94.2 69.3

Bottom 96.4 98.5 – 98.2

HD initial, ppm 1.45 ± 0.14 <1.45 ± 0.14 10 <1.45
HD in the product,
< 0.01 < 0.01 0.09 < 0.01
expected, ppm
Amount of gas,
572 630 390 420
Standard Liters
Bottom probe HD
57.5 ± 1.2 76.9 ± 1.6 – 16.2 ± 0.5
measured, ppm
Table 2. Results of cleaning

160

140

120
Height, cm

100

80

60

40

20

0
1E-6 1E-5 1E-4 1E-3 0.01 0.1 1 10 100
HD fraction, ppm
Figure 14. Expected concentration profile for HD. Feed flow = 1 l/min, pressure = 1.5 bar,
purging flow = 0.015 l/min, HETP = 2.2 cm, initial deuterium atomic fraction = 3 ppm.
The column’s deuterium concentration profile, as calculated from the ortho and para hy-
drogen separation, yields a very low value at the top of the column. This low concentration
could not be confirmed by online direct measurement because of insufficient sensitivity in
the mass spectroscopy.

Direct deuterium measurements by the accelerator method

The direct analysis of product samples using accelerator mass spectrometry at the Institute
of Particle Physics, ETH-Zurich, Switzerland [12], indicated an absence of signal at 6 ppb
sensitivity, which means at least better than 6 ppb. This is an order of magnitude better
than the technical requirements of the MuCap experiment, and corresponds to the most iso-
topically pure protium in the world.

Measurements of deuterium concentration in depleted hydrogen

As mentioned earlier, even the accelerator method provides insufficient sensitivity for di-
rect deuterium measurements at extremely low concentrations. Therefore, it was proposed
to use the DRU to concentrate the deuterium from a large amount of depleted hydrogen.
The HD admixture is measured in the enriched “stillage bottoms,” which is hydrogen col-
lected in the reboiler. The final result was obtained by dividng the measured value by the
ratio of the concentrated “stillage bottoms” and the full volume of hydrogen.
The concentration was provided by cleaning the full hydrogen amount in
“feedthrough without purging” mode. The full volume of the gas was precisely measured
by the previously calibrated mass-flow controller MFC4. The calibration was provided by
discharge of hydrogen from the vessel with the well-known volume under control of pres-
sure by a precise manometer.
The entire amount of deuterium-depleted hydrogen available (846.7 liter in total)
was passed through the column. As a result, 80% of deuterium was concentrated in ap-
proximately 15 ml of liquid hydrogen in the reboiler. The coefficient of concentration was
calculated on a basis of the mathematical model.

Direct isotopic measurements via gas chromatography

The gas chromatographic device used in the present work permitted direct measurements of
large (1% and higher) concentration of components (H2, HD or D2) without modifying the
measuring apparatus. The mixture for this test was specially prepared by diluting pure D2
with natural hydrogen, by passing the mixture through a column filled with palladium cata-
lyst immersed in silica gel. The equilibrated composition of the mixture after the conver-
sion had the composition: 31.2% H2, 47.3% HD, rest D2 (Figure 15).
The test of the H2 – HD – D2 mixture was done at higher pressure (5 bar) to reduce
the separation factor. On the basis of an assumed N and observed composition of hydrogen
isotope, the fraction profile of hydrogen isotopes in the column can be calculated.
Finally, the value number of theoretical plates is obtained to give the best-fit values
of calculated compositions with the hydrogen isotope compositions in the top, middle and
bottom of the column. The HETP calculated in this test was 1.95 cm, which conforms with
the previously obtained value.
 1.0
0.9
0.8
0.7
Mole fraction

D2 measured
0.6 D2 calculated
0.5 HD measured
HD calculated
0.4 HD measured
0.3 H2 calculated

0.2
0.1
0.0
0 10 20 30 40 50 60 70 80 90
Figure 15. IsotopeNumber of theoretical
composition plate
along the column

6. Conclusions
The height equivalent to a theoretical plate (HETP) for the column is 2.2 cm, on the basis
of our measurements. It is one of the best results yet for columns of low and medium cryo-
genic power. A new packing design was developed and successfully tested.
The column’s isotopic separation power exceeds the demands of the MuCap ex-
periment. The output deuterium concentration is lower than 6 ppb, and this value is nearly
independent of the initial concentration. The output of the system in the “feedthrough”
modes, which are most appropriate for production, is 500 l/day. A sufficient amount of
deuterium-depleted hydrogen was obtained for use in MuCap experimental runs in 2006
and 2007.
In addition to deuterium-depleted hydrogen, the setup can produce pure ortho-
hydrogen, which is useful for some experimental physics applications.
A method for measuring low deuterium content in hydrogen (1 ppm and less) was
developed and used to measure the deuterium admixture for the 2006 MuCap run.
7. References
[1] V. A. Ganzha et al., A circulating hydrogen ultra-high purification system
for the MuCap experiment. Nuclear Instruments and Methods in Physics Re-
search Section A: Accelerators, Spectrometers, Detectors and Associated
Equipment. Volume 578, Issue 3, 11 August 2007, 485–497.
[2] V. A. Andreev et al., Measurement of the Muon Capture Rate in Hydrogen
Gas and Determination of the Proton’s Pseudoscalar Coupling g P , PRL 99,
032002 (2007).
[3] B. Lauss et al., New Precision Determination of g P and GF : the MuXperi-
ments at PSI, Int. Conference on Exotic Atoms and Related Topics (EXA
2005), Vienna, Austria, Feb. 21-25, 2005, nucl-ex/0601004.
[4] I. J. Halvorsen and S. Skogestad, ”Distillation Theory”, In: Encyclopedia of
Separation Science. Ian D. Wilson (Editor-in-chief), Academic Press, 2000,
pp. 1117-1134.
[5] M. R. Fenske, Ind. Eng. Chem., 24, 1932.
[6] David R. Lide (Editor-in-chief). CRC Handbook of Chemistry and Physics.
87th Edition. Online version. http://www.hbcpnetbase.com/?BookID=34.
Taylor and Francis Group, LLC. 4-18 (2007).
[7] NIST Standard Reference Database Number 69, June 2005 Release.
http://webbook.nist.gov.
[8] I. Alekseev et al., Experimental results of hydrogen distillation at the deute-
rium removal unit of the MuCap experiment. Preprint PNPI-2702, Gatchina
(2006), 26p.
[9] V.D. Trenin, G.A. Sukhorukova, et al., “Fullscale experimental studies of
the various type mass exchange packings by water distillation”, Fusion
Technology, 28, 1579 (1995).
[10] T. Yamanishi and M. Kinoshita. “Preliminary Experimental Study for Cryo-
genic Distillation Column with Small Inner Diameter”, J. Nucl.
Sci.Technol., 21, 853 (1984)
[11] A. Markov, G. Sokolov, V. Medvedev, N. Chernov. The quiantitative analy-
sys of hydrogen stable isotopes and their nuclear magnetic isomers by a gas
chromatography method. Preprint PNPI-724, 22p. (1981).
[12] M. Suter et al., Advances in particle identification in AMS at low energies.
Nuclear Instruments and Methods in Physics Research B 259 (2007) 165–
172.