Instant Access to Ceramic Lasers 1st Edition Akio Ikesue ebook Full Chapters
Instant Access to Ceramic Lasers 1st Edition Akio Ikesue ebook Full Chapters
Instant Access to Ceramic Lasers 1st Edition Akio Ikesue ebook Full Chapters
https://ebookfinal.com/download/epigenetic-risks-of-cloning-1st-
edition-akio-inui-editor/
ebookfinal.com
https://ebookfinal.com/download/progress-in-nanotechnology-processing-
progress-in-ceramic-technology-1st-edition-the-american-ceramic-
society/
ebookfinal.com
https://ebookfinal.com/download/advances-in-ceramic-armor-vii-ceramic-
engineering-and-science-proceedings-1st-edition-jeffrey-j-swab/
ebookfinal.com
https://ebookfinal.com/download/advanced-ceramic-processing-1st-
edition-adel-mohamed/
ebookfinal.com
https://ebookfinal.com/download/soft-tissue-lasers-in-dental-
hygiene-1st-edition-jessica-blayden/
ebookfinal.com
https://ebookfinal.com/download/glass-ceramic-technology-second-
edition-wolfram-holand/
ebookfinal.com
https://ebookfinal.com/download/advances-in-ceramic-matrix-
composites-1st-edition-jim-low/
ebookfinal.com
Ceramic Lasers 1st Edition Akio Ikesue Digital Instant
Download
Author(s): Akio Ikesue, Yan Lin Aung, Voicu Lupei
ISBN(s): 9780521114080, 052111408X
Edition: 1
File Details: PDF, 21.57 MB
Year: 2013
Language: english
more information - www.cambridge.org/9780521114080
CERAMIC LASERS
Until recently, ceramic materials were considered unsuitable for optics due to the numerous
scattering sources, such as grain boundaries and residual pores. However, in the 1990s the
technology to generate a coherent beam from ceramic materials was developed, and highly
efficient laser oscillation was realized. In the future, the technology derived from the
development of the ceramic laser could be used to develop new functional passive and
active optics.
Co-authored by one of the pioneers of this field, the book describes the fabrication tech-
nology and theoretical characterization of ceramic material properties. It describes novel
types of solid lasers and other optics using ceramic materials to demonstrate the application
of ceramic gain media in the generation of coherent beams and light amplification. This
is an invaluable guide for physicists, materials scientists and engineers working on laser
ceramics.
akio ikesue is the President of World-Lab. Co., Ltd. He is also an executive scientist at
SCHOTT AG, Germany, and an invited professor at ENSCP (Ecole Nationale Supérieure
de Chimie de Paris).
yan lin aung is a Senior Researcher at World-Lab. Co., Ltd. His research focuses on
opto-ceramics.
voicu lupei is a Professor at the National Institute for Laser, Plasma and Radiation
Physics in Bucharest, Romania. His research interests are in photonic materials, quantum
electronics processes in doped laser materials, and applications of solid-state lasers.
CERAMIC LASERS
A K I O IK E S U E
World-Lab. Co., Ltd, Japan
YA N L IN AUNG
World-Lab. Co., Ltd, Japan
VO I C U L U P E I
National Institute for Laser, Plasma and Radiation Physics,
Romania
cambridge university press
Cambridge, New York, Melbourne, Madrid, Cape Town,
Singapore, São Paulo, Delhi, Mexico City
Cambridge University Press
The Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521114080
c A. Ikesue, Y. L. Aung, V. Lupei 2013
Printed and bound in the United Kingdom by the MPG Books Group
A catalog record for this publication is available from the British Library
Preface page ix
Acknowledgement xii
1 Introduction 1
1.1 Research background 1
1.2 Technical problems of melt-growth single crystals 2
1.3 Technical problems of ceramics 7
1.4 Purpose of this research 11
1.5 Outline of the book 12
References 16
2 Solid-state laser processes and active materials 18
2.1 Interaction of quantum systems with electromagnetic radiation
(radiation absorption and emission processes in quantum systems) 18
2.2 Solid-state lasers 25
2.3 The flow of excitation inside the laser material 63
2.4 Laser emission processes 67
2.5 The spatial distribution of the de-excitation processes 82
2.6 Thermal field inside the pumped laser material and thermal effects 85
2.7 Performance scaling of solid-state lasers 91
2.8 The laser material 93
References 115
3 Experimental technique: powder characteristics and the synthesis of optical
grade ceramics, effects of sintering aids 121
3.1 Introduction 121
3.2 Microstructure and optical characteristics of Nd:YAG processed by HIP
(hot isostatic pressing) 142
References 158
4 Synthesis of polycrystalline ceramic lasers (RE-doped sesquioxides) 160
4.1 Current status of single crystal technology 160
4.2 Requirements for sesquioxide ceramic lasers 162
v
vi Contents
Research and development on various types of solid-state laser has been carried out since
the creation of the first ruby laser by Dr. T. H. Maiman in 1960 [1]. Examples of laser
technologies include continuous wave (CW) lasers using Nd:YAG (neodymium-doped
yttrium aluminum garnet) single crystals at room temperature developed by Dr. Geusic
in 1964 [2], and tunable and ultra-short pulse lasers using Ti:sapphire single crystals
developed by P. F. Moulton in 1982 [3]. These lasers are still being applied in industrial
applications.
On the other hand, it is said that people started to use porcelain (ceramic ware) as standard
tableware in the days before Christ. In the USA, modernization of ceramic technology
started in the 1950s. Ceramics have now become an essential part of technologies that
support industries such as electronics and engineering etc., and their industrial use has
expanded greatly from year to year.
The interaction of ceramics with optics originated from the success that was achieved by
Dr. R. L. Coble of MIT (Massachusetts Institute of Technology) in the late 1950s [4]. He
succeeded in fabricating alumina ceramics including fewer residual pores by controlling the
microstructure. This material exhibited translucency (i.e., diffused transmission, not in-line
transmission throughout the specimen), and was applied as a discharge tube in a high-
pressure sodium vapor lamp, thus contributing to the advancement of lighting technology.
Research and development of various types of translucent and transparent ceramics then
occurred around the world, but the optical quality of the translucent ceramics was much
inferior to that of commercial single crystal and glass materials. In addition, there were no
practical applications of these ceramics at that time, and most of the translucent ceramics
were only of interest for laboratory research.
The development of transparent ceramic materials for applications in laser oscillation
is an ultimate aim of materials technology. It is also a dream of materials scientists in
the field of ceramics to develop ceramic (polycrystalline) materials which can be used
for laser oscillation with a performance equal to or better than that of conventional single
crystal materials. The first two attempts were from the USA, where laser oscillation using
Dy:CaF2 ceramics under cryostatic conditions was reported in 1964 [5], and pulse laser
oscillation using Nd:ThO2 –Y2 O3 ceramics was reported in 1973 [6]. Both succeeded in
producing laser oscillation using ceramic materials, but the laser performance was very
ix
x Preface
poor even though the ceramic samples were fabricated using the most advanced technology
at that time. Almost two decades passed until the next attempt was made in 1995 (by
the author) to produce efficient laser oscillation using ceramic materials [7], selecting
Nd:YAG, which is a technologically and industrially important material in the field of
solid-state lasers. It was demonstrated that polycrystalline Nd:YAG ceramics can be used
for CW laser generation with an efficiency comparable to that of commercial single crystal
lasers. Microchip lasers [8] and single mode laser oscillation [9] were achieved using
heavily doped Nd:YAG ceramic, and the excellent performance of ceramic lasers was
demonstrated. The importance of ceramic lasers began to achieve recognition worldwide
from 2000, and nowadays research and development of ceramic lasers is being carried
out around the world. Moreover, the development of new optical materials in addition to
Nd:YAG ceramics was initiated by applying the technology to make highly transparent
ceramics.
The ceramic laser changed completely the common concept of the conventional solid-
state laser. It was possible to fabricate polycrystalline ceramic laser gain media with
extremely low scattering and optical homogeneity, which cannot be obtained with commer-
cial single crystals. In addition, ceramic technology permits the fabrication of composite
laser elements with extremely complicated structures to improve the laser performance
[10], and the fabrication of fiber laser elements to generate high beam quality lasers. In this
sense, ceramic lasers can provide a broad range of design flexibility for the enhancement of
laser performance compared to traditional single crystal lasers. New laser materials, which
are difficult to obtain using the conventional melt-growth technology, such as Sc2 O3 and
Y2 O3 , can be fabricated by advanced ceramic processing, and high laser performance from
such sesquioxide ceramics has been confirmed [11]. It is predicted that ceramic lasers will
play a major role in solid-state lasers in the future, and will have applications in many
industrial fields.
It is interesting that the author successfully produced a highly efficient ceramic laser
even though he was not an expert in the field of ceramics or lasers at that time, but was
engaged in the research and development of refractory materials. The readers may question
(1) why the author focused on the development of ceramic lasers and (2) how the author
managed to achieve technological success. Regarding the first question, it is normally the
dream of a researcher to make a breakthrough in the development of the most difficult
materials theoretically and technologically, and this was the best opportunity for the author
to realize a contribution to ceramic technology. Regarding the second question, to be
honest, at that time the author did not actually know about the special (strict) requirements
(demand characteristics) of laser materials, and that is why he could perform such a reckless
exploration of laser generation using ceramic materials. Therefore, even an engineer who
does not know the materials technology may well open the door to new technologies if he
or she has a challenging spirit and can keep the dream alive as an engineer. “Challenging
nothing for fear of failure might be the biggest risk for a scientist.”
In this preface, the development of ceramic lasers has been described briefly. In this
book, a wide range of topics is described, from the fabrication technologies of ceramic
Preface xi
lasers, which are expected to be used in the development of future solid-state lasers, to their
performance.
References
[1] T. H. Maiman, Stimulated optical radiation in ruby, Nature (London) 187 (1960)
493–494.
[2] J. E. Geusic, H. M. Marcos, and L. G. van Uitert, Appl. Phys. Lett. 4 (10) (1964)
182–184.
[3] P. F. Moulton, Opt. News 8 (6) (1982) 9.
[4] R. L. Coble, Am. Ceram. Soc. Bull. 38 (10) (1959) 501.
[5] S. E. Hatch, W. F. Parson, and R. J. Weagley, Appl. Phys. Lett. 5 (1964) 153.
[6] C. Greskovich and J. P. Chernoch, J. Appl. Phys. 44 (10) (1973) 4599–4605.
[7] A. Ikesue, T. Kinoshita, K. Kamata, and K. Yoshida, J. Am. Ceram. Soc. 78 (4) (1995)
1033–1040.
[8] I. Shoji, Y. Sato, T. Taira, A. Ikesue, and K. Yoshida, Highly Nd3+ -doped YAG
ceramic for high power microchip, CGCT-1, 2000/Abstract Book, T-C-05 (2000),
pp. 345–355.
[9] A. Ikesue, Yan Lin Aung, T. Taira, T. Kamimura, K. Yoshida, and G. L. Messing,
Progress in ceramic lasers, Mater. Res. Annu. Rev. 36 (2006) 397–429.
[10] A. Ikesue and Yan Lin Aung, Synthesis and performance of advanced ceramic lasers,
J. Am. Ceram. Soc. 89 (6) (2006) 1936–1944.
[11] A. Ikesue, K. Kamata, and K. Yoshida. Synthesis of transparent Nd-doped HfO2 –Y2 O3
ceramics using HIP, J. Am. Ceram. Soc. 79 (2) (1996) 359–364.
Acknowledgement
This research and development effort was started when the author was working in a refrac-
tory manufacturing industry (Kurosaki Harima Corporation) as an engineer. Support from
the members of Kurosaki Harima Co. for research up to the present date is greatly appre-
ciated. The authors are very thankful to collaborators around the world for their supportive
collaboration: Professor Y. Iwamoto (Nagoya Institute of Technology), Associate Profes-
sor T. Kamimura (Osaka Institute of Technology), Associate Professor T. Taira (Institute
of Molecular Science) and Associate Professor T. Yamamoto (Tokyo University) from
Japan, and Dr. A. Lupei (Romania) and Dr. Yvonne Menke (SCHOTT AG, Germany) from
Europe, and Dr. R. Shori (UCLA), Dr. J. Sanghera (Naval Research Laboratory) and Dr. M.
Dubinsky (Army Research Laboratory) from the USA. Also, the authors greatly appreciate
support and useful advice from Dr. Joan Fuller, Dr. K. Jata and Dr. K. Goretta of AFOSR
(Air Force Office of Scientific Research)/AOARD (Asian Office of Aerospace Research
and Development) for the development of the “high power density laser.”
xii
1
Introduction
1
2 Introduction
energy conversion efficiency. The material that satisfies these demand characteristics is
YAG, which occupies the majority of the solid-state laser market. It is not likely that there
will be new materials in the future that can exceed the performance of YAG materials in
terms of physical properties and laser characteristics. However, almost all of the materials
used for solid-state lasers, including YAG, are single crystals produced by the melt-growth
process. Initially, single crystals were produced by the Verneuil process, also called flame
fusion [2], but because of quality issues in terms of laser amplification efficiency and
laser beam quality, crystals grown by the Cz (Czochralski) process are now being used
in industrial applications [32]. However, it takes about one month to grow an undoped
YAG single crystal with a relatively large diameter, and it takes longer to grow an Nd:YAG
single crystal. In addition, the optically homogeneous portion of the single crystal ingot
which is suitable for laser oscillation is very limited because of the intrinsic core and facets.
Therefore, the productivity and output characteristics are still unsatisfied from an industrial
point of view.
Therefore, it was considered that if it were possible to produce YAG laser materials using
ceramic technology, the above technical issues of single crystal YAG materials could be
solved, but a weak point was that the grain boundaries in ceramics can disturb the efficient
amplification of the coherence beam in the materials. This weak point was reflected in
the results achieved from Dy:CaF2 ceramics in 1964 and from Nd:ThO2 –Y2 O3 ceramics
in 1974. Although laser oscillation was demonstrated using these ceramic materials, the
resulting beam quality and lasing efficiency were not sufficient for the development of
ceramic materials as laser gain media in the future. However, this negative opinion changed
abruptly with the successful demonstration of a highly efficient polycrystalline Nd:YAG
ceramic laser at room temperature by the author in 1995. The author demonstrated that
ceramic materials can show laser performance equivalent to that demonstrated for single
crystal laser gain media and, moreover, ceramic lasers have potential which cannot be
realized with conventional single crystal lasers. Nowadays, the optical quality of polycrys-
talline ceramic laser gain media is comparable to that of the high quality commercial single
crystal counterparts, and further technological progress is expected in the future.
20
Exciting Level
16
R2
11502cm-1 Upper Level
Energy [×103 cm-1]
4F
3/2 11414
4F R1
4F 5/2 Laser Transition
12 3/2
1.6 μm 4I ~6000
Laser Transition 15/2
8
4I ~4000
4I 13/2
15/2
2526
4I 2473
4 13/2 4I 2146
11/2 2111
2029 Lower Level
4I 2001
11/2
4I 848
9/2 4I
0 Ground Level 9/2 311
197
134
0
normally Nd:YAG crystals are used for 1.06 µm generation. To obtain laser action in
Nd:YAG materials, first the electrons of Nd ions in the ground state are excited with a
xenon or krypton lamp (white light source from a discharge flashlamp) so that the electrons
are pumped up to energy levels higher than 4 F3/2 . More recently, a semiconductor laser (LD,
laser diode) with a certain wavelength has been used to pump the ground state electrons
directly to the 4 F5/2 band. Accordingly, these excited electrons which are in the upper levels
decay rapidly to the 4 F3/2 level (non-radiative transition), and then when these electrons are
transferred to 4 F13/2 , 4 F11/2 and 4 F9/2 , three fluorescent emissions (radiative transitions) (λ =
0.9, 1.06 and 1.3 µm) occur. Then the fluorescent emission is amplified in a set of mirrors
(the resonator), and finally an intensified light with a single wavelength (monochromatic
light) is emitted. In the Nd system, when electron transitions occur from the upper levels
to the 4 F3/2 level on exciting with a lamp source, the transitions are non-radiative and the
energy is released as heat. Therefore, effective laser oscillation cannot be achieved using
a lamp excitation system. (However, a high output power can be achieved using a lamp
excitation system because the lamp can provide very high input power to the laser gain
medium.) In the LD excitation system, laser diodes of wavelengths 808 nm or 885 nm
are used which can pump the ground state electrons directly to upper levels, 4 F5/2 or 4 H3/2
levels. Therefore, there are no excited electrons which undergo non-radiative transitions as
seen in the lamp excitation system, and efficient laser oscillation can be achieved. However,
4 Introduction
Gas
Ir Crucible
Induction Coil
Melted YAG
Zirconia Refractory
Zirconia Powder
Figure 1.2 Schematic diagram of a YAG single crystal grown by the Czochralski method.
it was difficult to obtain high power LDs, and low power models were also very expensive.
Therefore, the lamp excitation system was common for solid-state lasers. In the twenty-first
century, the price of high performance LDs is becoming more reasonable, and recently a
highly effective compact and high power (kilowatt range) Nd:YAG laser system with LD
pumping has been commercialized. But, a technical issue of the LD excitation source is
that the wavelength of the LD is shifted when its temperature is increased during laser
operation. It is necessary to control the temperature of the main body of the LD and the
spectral width of the emission band of the LD.
As described in Section 1.1, the Cz method is commonly used to produce YAG single
crystals (see Figure 1.2). High purity raw materials (over 4N, 99.99 mass%) of Y2 O3 , Al2 O3
and Nd2 O3 are used. They are weighed in the YAG stoichiometric ratio, and blended. Then
the blended powders are pressed into powder compacts and sintered. This relatively dense
Nd:YAG sintered body is put into an Ir (iridium) crucible, and melted by high frequency
induction heating at over 1950 ◦ C. Then a YAG seed crystal is placed on top of the YAG
melt, and the growing crystal is continuously pulled very slowly at a rate of 0.2 mm h−1 with
a rotation speed of 10–30 rpm. Normally, a YAG seed crystal with an orientation of 111
is used because it has the largest surface energy, but in some cases, 110 and 100 seeds
are also used. When Nd ions are doped into a YAG crystal, Nd replaces Y sites in the garnet
structure. However, the ionic radius of the Nd ion is much larger than that of the Y ion, and it
is well known that Nd hardly dissolves in YAG crystal. The segregation coefficient of the Nd
ion in YAG crystal (the ratio of the Nd concentration in the crystal to the Nd concentration
in the melt) is very small, about 0.2 [35]. Therefore, a highly doped Nd:YAG sintered body
is prepared in advance to produce a YAG melt with 2–3 times higher Nd concentration
than the target Nd concentration. (For instance, suppose the target composition is 1 at.%
1.2 Technical problems of melt-growth single crystals 5
Nd:YAG, then sintered bodies of 2 or 3 at.% Nd:YAG are prepared in advance, and they
are melted and grown to obtain a crystal of 1 at.% Nd:YAG.) Although, high concentration
Nd:YAG melts are prepared in advance, it is not certain that the grown crystal will have a
high Nd concentration with homogeneous doping. When the Nd concentration in YAG is
higher than 1 at.%, many inclusions (light scattering sources) are generated in the crystal,
and it cannot be used as a laser crystal. Therefore, the concentration of Nd ions is limited
to 1 at.% in YAG single crystals for laser applications. Even a 1 at.% Nd:YAG crystal
ingot includes optically inhomogeneous parts such as the core and (211) facets from the
pulling axis towards the rim area [35]. Accordingly, only a part (the optically homogeneous
part) of the grown crystal can be used as a laser gain medium. Another technical issue
with melt-growth crystals is that the Nd concentration varies slightly in the longitudinal
direction of the ingot. This is because when the crystal is pulled, the concentration of Nd
ions in the grown crystal is significantly lower than that in the melt. Thus, Nd ions are
concentrated in the melt as the crystal growth progresses. Therefore, the Nd concentration
in the grown crystal at the early stage of crystal growth (near the seed crystal) is slightly
lower compared to the middle stage and terminal stage (the end of the ingot) along the
longitudinal direction of the crystal ingot. As a result, this process produces a crystal rod
which has a change in refractive index at each end. This is the technical limitation of the
conventional melt-growth method.
A polarized image of approximately 1 at.% Nd:YAG single crystal (ingot), a Schlieren
image and X-ray tomography image are shown in Figure 1.3(a), (b) and (c), respectively
[35]. All of the images (especially the cross-sectional images) show crystal defects (such
as core and facets) in the crystal ingot. A commercial Nd:YAG slab and polarized images
(cross-nicol) are shown in Figure 1.3(d) and (e, f ), respectively. The commercial Nd:YAG
single crystal has very high transparency, and it looks optically very homogeneous to the
naked eye (under natural light). However, when it is observed under a polarizer, lamella
shaped facets can be seen across the length direction of the slab, suggesting that even the
commercial high quality single crystal materials include such optically inhomogeneous
components.
Any optical defects remaining in a laser gain medium can significantly reduce the light
amplification efficiency and beam quality. Therefore, the demand characteristics for laser
materials are necessarily very strict. Regarding quality control in commercial single crystal
production, the thermal stress and refractive index distribution of the grown Nd:YAG ingot
are inspected using a polarizer and interferometry. Optically homogeneous parts of the
ingot are cut off and polished, and then the numbers of optical scattering centers in the
samples are detected using a laser system. Finally, the most optically homogeneous part is
selected as a high quality laser gain medium.
The optical loss of commercial high quality Nd:YAG single crystal is in the region of
0.2% cm−1 , and the quality has improved significantly compared with the crystal grown
by the Bernoulli process in 1964. In fact, the melt-growth single crystal technology has
reached its limits, and it is extremely difficult to improve further the quality of current
single crystal materials.
6 Introduction
(d)
(e) (f)
Figure 1.3 Cross-section of an Nd:YAG single crystal imaged by (a) polarizing light, (b) the Schlieren
method (interferometry image), and (c) X-ray tomography. (d) Appearance of a commercial single
crystal Nd:YAG slab, and (e, f ) polarized images (open and cross-nicol) under a polarizer.
To grow a YAG crystal, a very expensive Ir crucible and growth unit is necessary. What
is worse is that the growth rate is very slow. Normally it takes about one month to grow
an ingot. Other economic issues are the very high initial costs of the fabrication equipment
and running costs for a long delivery time (electricity, and recovery costs for the Ir crucible
etc.). In addition, the yield percentage of high quality laser crystal is very low. There are
also technological issues. Only a small cross-section rod can be achieved. The diameter
of the ingot is limited to 5–6 inches, and a laser rod with approximately 10 mm diameter
by 150 mm length or a laser slab of approximately 6 × 20 × 150 mm3 can be removed
1.3 Technical problems of ceramics 7
from the ingot. Therefore, large gain media for the future development of high power lasers
cannot be produced by the current melt-growth crystal technology.
As described above, generally YAG has many advantages compared to other laser mate-
rials. However, the most important part in the laser system is the Nd:YAG material, and
there are still many unsolved economic issues (including yield percentage) and technolog-
ical issues. It is technologically almost impossible to achieve a breakthrough in the current
single crystal growth process, and even a partial solution to the problems appears far
away.
100
Al2O3 single crystal
80
Transmittance (%)
60
Al2O3 ceramics (Hot press)
40
Al2O3 ceramics
20
0
0.4 0.6 0.8 1.0 1.5 2.0 2.5
Wavelength (m)
Figure 1.4 Relationship between measured wavelength and in-line transmittance for sapphire single
crystal, hot-pressed Al2 O3 ceramic, and Al2 O3 ceramic produced by normal pressure sintering.
Figure 1.5 Illustration of optical scattering in polycrystalline ceramics due to various microstructural
defects: (a) grain boundary, (b) residual pore, (c) secondary phase, (d) birefringence, and (e) surface
roughness.
ceramics were mainly used for applications requiring a thin slice shape in the technology
of the twentieth century.
Figure 1.5 illustrates the microstructure and scattering sources of normal translucent
ceramics [44]. The ceramic is composed of many fine grains (crystallites) with randomized
crystal orientations. Scattering sources in normal ceramics are (a) grain boundaries,
(b) residual pores, (c) inclusions and secondary phases, (d) birefringence and double
reflections, and (e) surface roughness. However, the fundamental differences in microstruc-
ture between single crystal materials and ceramic materials which affect the scattering
1.3 Technical problems of ceramics 9
Nd:YAG rod
Window
Chamber (water cooling)
Reflection mirror (upper)
Reflection mirror (lower)
Figure 1.6 Schematic diagram of a commercial laser oscillator using a single crystal gain medium
and flashlamp excitation.
characteristics are (a) the present of grain boundaries, and (b) the quantity by volume
of residual pores. Accordingly, it is considered that these scattering sources significantly
influence the optical quality of ceramic materials.
On the other hand, it is true that laser crystals with very low scattering loss can have a
very high laser amplification efficiency. In fact, ultra low loss single crystals still cannot be
produced using the current crystal technology. The optical loss of high quality commercial
single crystals is around 0.3–0.2% cm−1 (some latest data show less than 0.2% cm−1 ), and
this quality can offer relatively efficient laser generation from commercial single crystals.
A schematic diagram of a commercial laser oscillator using a single crystal gain medium
and flashlamp excitation is shown in Figure 1.6 [45]. External energy (from a flashlamp
or semiconductor laser) is input into the laser gain medium (doped with fluorescent ions)
in order to excite the electrons continuously from the ground state to laser levels. Then,
the emitted fluorescent beams are reflected and amplified between two mirrors. The laser
beam quality and amplification efficiency are severely affected by the numbers of scattering
sources in the laser gain medium.
Previously reported papers on translucent ceramics have discussed few details of the
optical characteristics (optical constant), and there have been almost no reported papers
which have described the scattering loss, condition of double refraction and homogeneity
of the refractive index in ceramic materials. It can be considered that the scattering loss of
the translucent or transparent ceramics developed in the past was very high, and they could
not meet the requirements of laser materials.
However, one researcher, Dr. Greskovich of GE Co., developed a transparent Nd-
doped 10%ThO2 –Y2 O3 ceramic, and succeeded in producing laser oscillation at room
10 Introduction
Figure 1.7 Picture of Dr. C. D. Greskovich and microstructure of Nd:ThO2 –Y2 O3 ceramic.
temperature for the first time using a polycrystalline ceramic gain medium in 1973. (About
10 years before this, successful laser oscillation was achieved using Dy:CaF2 ceramic
under cryostatic conditions, but the details of the physical properties of the materials
were not described.) Dr. Greskovich succeeded in fabricating transparent Nd:ThO2 –Y2 O3
ceramic by sintering low sinterability Y2 O3 powder, with 10% ThO2 as a sintering aid, at
2200 ◦ C for about 100 hours.
Figure 1.7 shows reflection microscopy and transmission microscopy images of the 1%
Nd:ThO2 –Y2 O3 ceramic together with a picture of Dr. Greskovich [6, 46, 47]. Although
only a few residual pores are observed on the surface in the reflection micrograph, many
internal defects (such as voids and grain boundary phases) can be observed in the transmis-
sion micrograph. An optically inhomogeneous structure, called an “orange peel” structure,
can be seen with the naked eye. Although it was reported that the number of residual pores
in the Nd:ThO2 –Y2 O3 ceramic reached 10−5 –10−7 (10–0.1 ppm) in 1973, the optical loss
of the material was lower than several percent per centimeter. Because of this optical qual-
ity, pulse laser oscillation could be produced by flashlamp excitation at room temperature,
but the slope efficiency was as low as 0.1%. This technology was the best achieved in the
1.4 Purpose of this research 11
1970s, and the success of Dr. Greskovich encouraged materials scientists. However, there
was no challenger who could compete with this achievement for about 20 years.
Compared to normal optics for active and passive applications, the optical quality of the
single crystals required for laser application is significantly higher because light amplifi-
cation occurs in the gain medium, and the amplified light is passed repeatedly through the
gain medium. Therefore, even a small amount of optical loss of the material may have a
significant effect on laser oscillation. In the past, various types of translucent or transparent
ceramic materials have been developed. But it can be assumed that the scattering losses
of those materials were very large compared to the materials reported by Dr. Greskovich.
It was difficult to produce single crystal material with optical quality sufficient for laser
applications, and it was far too difficult to produce conventional translucent ceramics, which
include many crystal defects such as grain boundaries, residual pores and inclusions etc.,
which could meet the requirements for laser applications. For that reason, conventionally
it was doubted that ceramic materials could be used as laser gain media in the future of
solid-state lasers. If a technology existed which could remove microstructural defects per-
fectly (except grain boundaries) from ceramics, perhaps ceramic materials could be used as
optical materials in laser applications. This problem was not solved until the author demon-
strated high efficiency laser oscillation from a polycrystalline ceramic material. Details of
this development are described in this book.
at the formation of the host composition (e.g., garnet) step and calcination, in advance of
the sintering process. In principle, therefore, it is very effective for improving homogeneity
during the sintering process.
However, from another technical point of view, it can be considered that the wet process
was invented to avoid the above described problems of the dry process because the properties
of raw powder materials for the conventional dry process (solid-state reaction method) were
not good enough for the production of optical quality ceramics. Whether this speculation is
true or not hinges upon the fact that high optical quality ceramics can be produced by the dry
process, using good quality raw powder materials. In this research, therefore, we focused
on the solid-state reaction method without using synthesized raw powders prepared by the
wet process. If the dry process can overcome the above mentioned general technical issues,
it will become a very important industrial process for the economic production of optical
ceramics because the powder preparation process is very simple and the preparation of
powders with various compositions is very easy. Accordingly, the dry process was selected
for this study. The main objectives of this study are as follows:
Chapter 5 describes the synthesis of Nd:YAG ceramics heavily doped with 0.3–4.8 at.%
Nd ions, their optical characteristics and their laser performance. Generally, only about
1 at.% Nd ions can be dissolved homogeneously in YAG single crystal grown by the Cz
method. If the doping concentration of Nd ions exceeds 1 at.%, segregation of Nd occurs,
the optical quality decreases, and the material is not suitable for use as a laser gain medium.
In the case of polycrystalline YAG ceramics, it was confirmed that a greater amount of Nd
ions dissolved homogeneously compared to the melt-growth Nd:YAG single crystals. From
the spectroscopic point of view, heavily doped polycrystalline Nd:YAG ceramic (more than
1 at.% Nd ions) has a new functionality as an advanced laser material. One of its applications
is as a microchip laser and its laser performance (slope efficiency) was better than that of
commercial Nd:YAG single crystals. In addition, single longitudinal mode laser oscillation
was realized without using an etalon, mode selector, from a microchip ceramic laser.
In this chapter, the effect of impurities (Si added as a sintering aid) on the dissolution
of Nd ions in YAG ceramics is described, and we discuss why a large amount of Nd ions
can be dissolved homogeneously in YAG ceramics compared with single crystal YAG.
The main difference between single crystals and ceramics in terms of microstructure is the
presence or absence of grain boundaries, but a small amount of Si is also added in Nd:YAG
ceramics. It was identified for the first time that the addition of a small amount of Si leads
to more homogeneous dissolution of Nd ions in the YAG ceramics. The traditional theory
says that Nd can hardly replace Y in the YAG lattice because of the large difference in ionic
radius of the atoms. In this study, it was assumed that the lattice distortion, occurring when
Nd3+ replaces Y3+ in the YAG lattice, was compensated for by the addition of Si, and as
a result, a large amount of Nd could be doped homogeneously into the YAG lattice. In an
actual experiment, it was confirmed that 7.2 at.% Nd was doped homogeneously into YAG
ceramic using Si as a sintering additive.
In Chapter 6 we investigate the sources of optical scattering in polycrystalline ceramics,
and summarize the results of fabrication techniques for very reliable high quality Nd:YAG
ceramics. Single crystals have been utilized traditionally as the laser gain medium because
they do not have the problem of grain boundary scattering. This advantage has led to the
use of single crystals for optical applications. In this chapter, we focus on the pore volume
of Nd:YAG ceramics and the amount of grain boundaries, and evaluate the amount of
optical scattering quantitatively. In addition, the correlation between laser characteristics
and optical scattering was investigated. It was verified that the number of residual pores
and grain boundary phases greatly affects the laser oscillation characteristics. It was also
realized that the laser oscillation characteristics do not depend upon the number of grain
boundaries. Recent data demonstrated that the scattering coefficient of high quality laser
ceramics at the laser oscillation wavelength regions is lower than that of commercial laser
single crystals, and a breakthrough occurred in the optical performance of ceramics. The
laser performance of ceramics with extremely low scattering has reached very close to the
quantum oscillation efficiency, and ceramic materials have almost reached the ideal state
for application in optics. In addition, using laser tomography, the scattering conditions of
ceramics and commercial laser single crystals were investigated. Results of laser damage
tests, which can be used in the future development of high power lasers, are also described.
1.5 Outline of the book 15
defect-free microstructure. This basic idea will be applicable in other technological fields
(e.g., electronic materials and structural materials etc.). Application technologies developed
by the author, and their possible development in the future, are also described in this chapter.
Chapter 10 describes the main results in the spectroscopic investigation of doped trans-
parent ceramics. It is shown that spectroscopic investigation has a double role, as a structural
method for describing the variety, nature and structure of the various structural centers of
the doping ions and the distribution of the doping ions in these centers, and as a method
of characterizing the optical processes that determine laser emission. Thus, spectroscopic
investigation provides the necessary information on the energy level structure and transition
probabilities to enable mathematical modeling of the laser process and selection of suitable
parameters for the laser material, pumping system and resonator design. Data on a large
variety of spectroscopic studies on doped ceramics are presented and their implication in
the structural properties of the ceramic and in modeling of the laser processes is discussed.
Chapter 11 describes the main factors that influence laser emission and heat generation
in doped ceramics. It is shown that a proper exploitation of the properties of laser ceramics
requires optimization of the pumping scheme for optimum utilization of the assorted exci-
tations for laser emission and for a reduction of heat generation. The structure–properties–
functionality relation for ceramic material is discussed and modalities to upgrade traditional
laser materials or to tailor new materials are discussed, together with new schemes for power
and energy scaling.
References
[1] T. H. Maiman, Stimulated optical radiation in ruby, Nature (London) 187 (1960)
493–494.
[2] J. E. Geusic, H. M. Marcos, and L. G. van Uitert, Appl. Phys. Lett. 4 (10) (1964)
182–184.
[3] P. F. Moulton, Opt. News 8 (6) (1982) 9.
[4] R. L. Coble, Am. Ceram. Soc. Bull. 38 (10) (1959) 501.
[5] S. E. Hatch, W. F. Parson, and R. J. Weagley, Appl. Phys. Lett. 5 (1964) 153.
[6] C. Greskovich and J. P. Chernoch, J. Appl. Phys. 44 (10) (1973) 4599–4605.
[7] A. Ikesue, T. Kinoshita, K. Kamata, and K. Yoshida, J. Am. Ceram. Soc. 78 (4) (1995)
1033–1040.
[8] I. Shoji, Y. Sato, T. Taira, A. Ikesue, and K. Yoshida, Highly Nd3+ -doped YAG
ceramic for high power microchip, CGCT-1, 2000/Abstract Book, T-C-05 (2000),
pp. 345–355.
[9] A. Ikesue, Yan Lin Aung, T. Taira, T. Kamimura, K. Yoshida, and G. L. Messing,
Progress in ceramic lasers, Mater. Res. Annu. Rev. 36 (2006) 397–429.
[10] A. Ikesue and Yan Lin Aung, Synthesis and performance of advanced ceramic lasers,
J. Am. Ceram. Soc. 89 (6) (2006) 1936–1944.
[11] A. Ikesue, K. Kamata, and K. Yoshida. Synthesis of transparent Nd-doped HfO2 –Y2 O3
ceramics using HIP, J. Am. Ceram. Soc. 79 (2) (1996) 359–364.
[12] T. H. Maiman, Phys. Rev. Lett. 4 (11) (1960) 564–566.
[13] K. Maeda and Y. Fujii, Oyo Butsuri (Bull. Appl. Phys. Jpn.) 56 (1987) 1651.
[14] B. Struve and G. Huber, J. Appl. Phys. 57 (1988) 45.
References 17
Lasers constitute the main component (the optical range) of quantum electronics, the science
of generation, amplification, the control of the properties (wavelength, temporal regime,
power range) of radiation by the processes of interaction of electromagnetic radiation with
quantum systems. The word LASER is an acronym standing for Light Amplification by
Stimulated Emission of Radiation. The driving mechanism that determines the amplifica-
tion of radiation in a laser is the stimulated emission in an ensemble of identical quantum
systems (the active medium) with inversion of populations between a high energy level
and a lower energy level that act as emitting and terminal energy levels for the emission,
in a resonator that grants the necessary conditions for amplification and extraction of the
amplified radiation beam and gives it a modal structure. A very important class of lasers
is the solid-state lasers, where the quantum systems are embedded in a transparent solid
material. The first laser was a solid-state laser based on the ruby crystal [1] and during their
longer than fifty years history lasers have undergone tremendous scientific and technolog-
ical development. Many textbooks discuss the basic physical properties and processes that
determine the characteristics of laser materials and processes and the physics and engineer-
ing of solid-state lasers [2–20]. They contain extensive lists of relevant literature which will
not be repeated here. Based on these books and on other relevant literature, this chapter
discusses some of the major characteristics of laser materials and processes based on doped
transparent polycrystalline materials produced by ceramic techniques.
18
2.1 Interaction of quantum systems with radiation 19
Figure 2.1 Absorption and emission processes in a two-level quantum system: (a) stimulated absorp-
tion, (b) stimulated emission and (c) spontaneous emission.
related by the Boltzmann distribution law n2 /n1 = (g2 /g1 ) exp[−(E2 − E1 )/kT ], where g1
and g2 are the degeneracies of the two energy levels. Interaction with the electromagnetic
field induces variation of the population densities described by the rate equation
dn2 dn1 (st) (sp)
=− = W12 n1 − W21 n2 − W21 n2 , (2.1)
dt dt
such that at any moment of time the sum of the populations remains equal to the total
(st) (sp)
population n0 , i.e. n1 + n2 = n0 . In Eq. (2.1), W12 , W21 and W21 denote the rates of
absorption, stimulated emission and spontaneous emission, respectively. If the action of
the external electromagnetic field is suddenly stopped, the variation of the population n2
is determined by the spontaneous emission only and is described by an exponential decay
law with the spontaneous radiative emission (luminescence) lifetime τrad = (W21 )−1 ,
(sp)
i.e.
The rates of the stimulated (induced) processes depend on the energy density ρ(ν) of the
(st)
electromagnetic field, W12 = B12 ρ(ν) and W21 = B21 ρ(ν), while the rate of spontaneous
emission W21 = A21 is constant and τrad = (A21 )−1 . The Einstein coefficients A21 , B12
(sp)
the coherence; however, if the incident beam is non-coherent, the beam of stimulated
photons will have a very low degree of coherence.
At low incident energy density ρ(ν) the de-excitation of the level E2 by spontaneous
emission (luminescence) dominates completely the stimulated emission. The spontaneous
emission lines of the isolated quantum systems have small but finite width ν determined
by the lifetime broadening (the natural linewidth), such that τrad ν = (2π )−1 , which influ-
ences uniformly all the identical quantum systems from the ensemble and the lines have
Lorentzian shape. However, in reality these quantum systems are not completely free and
isolated and thus their quantum states are influenced by additional interactions that induce
broadening of the lines. The broadening of the lines can be homogeneous, when it influences
all the quantum systems from the ensemble identically, or it can be inhomogeneous, when
these additional interactions induce different shifts of the energy levels of the various quan-
tum systems. If these shifts are not resolved spectrally, they result in broadening of the lines,
although large perturbations could induce resolved effects. The homogeneously broadened
optical lines (absorption and emission) are symmetric around the central frequency ν0 and
have Lorentz shapes
−1
2 4(ν − ν0 )2
gL (ν) = 1+ , (2.3)
π ν (ν)2
where ν is the full width at half maximum (FWHM). In the case of inhomogeneous
broadening the lines can be symmetric, with Gaussian shapes
ln 2 1 4(ν − ν0 )2
gG (ν) = 2 exp − ln 2 , (2.4)
π ν (ν)2
or asymmetric. Generally these two broadening mechanisms act together and sometimes
symmetric broadening combining the Lorentzian and Gaussian shapes (Voight profiles)
can be observed. Since the lineshapes are normalized, the linewidth ν and the peak value
g(ν0 ) are related, gL (ν0 )ν = 2/π ≈ 0.637, and gG (ν0 )ν = 2(ln 2/π )1/2 ≈ 0.939. Under
the action of narrow-band radiation all members of the ensembles with homogeneously
broadened lines respond identically, whereas in the case of absorption lines with dominant
inhomogeneous broadening the excitation is selective and involves systems whose peak
frequency is near the exciting frequency and are confined in the spectral region delineated
by the homogeneous broadening.
In real systems the characteristics of transitions are determined by the quantum states
(energy levels, wave functions) and by interaction with the electromagnetic field (selection
rules and transition probabilities for the various components – electric multipole and mag-
netic dipole – of the field). The optical transitions are investigated experimentally by static
(absorption, emission spectra) or dynamic (emission decay) or combined (time-resolved
spectroscopy) spectroscopic methods. The Einstein coefficients are seldom used in practice,
and parameters more relevant to collection or interpretation of the experimental results are
employed.
2.1 Interaction of quantum systems with radiation 21
r The line strength S characterizes the intensity of the spectral lines and is related to the
transitions between the quantum states for the various components i of the electromag-
netic field (electric dipole, electric quadrupole, magnetic dipole),
S= |ϕ1 |Pi |ϕ2 |2 , (2.5)
i
where ϕi are the wavefunctions of the states 1 and 2, and Pi is the operator of interaction.
The line strength can be measured directly from the experimental absorption spectrum.
r The oscillator strength f is a dimensionless parameter that establishes the relation
between the strength of transitions in a quantum system and the theoretical strength
of a single electron in a harmonic oscillator model and is related to the line strength by
8π 2 me ν
f12 = S12 , (2.6)
3e2 hg1
where me and e are the electron mass and charge and ν is the frequency of transi-
tion. The oscillator strengths for the downward and upward transitions are related by
f21 = (g1 /g2 )f12 . The line strength and oscillator strength can be related to the Einstein
coefficients,
8π 2 e2 ν 2 (8π 2 )2 ν 3
A21 = 3
f12 = S12 . (2.7)
mc 3hc3 g1
r The cross-section expresses the ability of an individual quantum system to absorb or
emit radiation and its dimension is that of an area (usually cm2 ) per quantum system.
The cross-section reflects the spectral composition of the optical process, σ (ν, ν0 ) =
σ0 g(ν, ν0 ), where the peak cross-sectionσ0 corresponds to the maximum (ν = ν0 ) of line
and, since the lineshape is normalized, σ (ν, ν0 )dν = σ0 . The absorption and emission
cross-sections are connected to the Einstein coefficients by the Fuchtbauer–Ladenburg
relations
1
σa ≡ σ12 (ν, ν0 ) = hνB12 g(ν, ν0 ) (2.8)
c
1 c2 c2
σe ≡ σ21 (ν, ν0 ) = hνB21 g(ν, ν0 ) = A21 g(ν, ν0 ) = g(ν, ν0 ) (2.9)
c 8π ν 2 8π ν 2 τrad
and a relation of reciprocity between these two cross-section holds,
σ12 (ν, ν0 )dν = (g2 /g1 ) σ21 (ν, ν0 )dν. (2.10)
The emission cross-sections corresponding to the center of the emission line (the peak
cross-sections) are related to the characteristics of the line and to the radiation lifetime: for
the Lorentz lineshapes
c2 c2 1
σ21 (ν0 , ν0 ) = 2 2
A21 = 2 2
, (2.11)
4π ν0 ν 4π ν0 ν τrad
22 Solid-state laser processes and active materials
E3(Ep ) E4(Ep )
E2(Eem ) E3(Eem )
Pump
Pump
Laser
Laser
E2(Et)
E1(Et = Eg ) E1(Eg )
(a) (b)
Figure 2.2 The three-level (a) and four-level (b) laser schemes.
are called quasi-three-level (Q3L) schemes due to the partial reabsorption of emission,
although several authors prefer the name quasi-four-level (Q4L) scheme since the terminal
level is not the ground state. In order to avoid stimulated emission on other transitions, in
all these schemes the reflectivities of the laser mirrors must correspond to the wavelength
of the desired laser emission.
A quantum generator of radiation using stimulated emission is then composed of three
major parts: (i) the active material that contains the ensemble of quantum systems whose
quantum properties determine the ability to create inversion of the population and to
amplify the stimulated emission, as well as the major characteristics of this emission
(wavelength, gain spectrum and so on); (ii) the system which produces the inversion of
population between the emitting and the terminal energy levels (the pumping system);
(iii) the resonator which provides the conditions for feedback and outcouples part of the
beam of radiation amplified by stimulated emission. It is obvious that the starting point
in designing a quantum generator is the active material. The other two parts, the pumping
system and the laser resonator, must enable the most efficient exploitation of the properties
of this material; moreover, optimization of the quantum generator of the radiation will
require correlated optimization of these three parts. Fundamentals and engineering of laser
emission are described in many textbooks, such as [2–11]; these textbooks also provide
lists of the relevant literature, which will not be repeated here.
The stimulated emission can be considered as a negative absorption that increases when
the beam travels repeatedly inside the pumped laser material due to the reflections on the
mirrors with reflectivity R2 ∼ = 1 and R1 < 1. In order to reach gain at any given geometrical
point inside the pumped laser material placed in the resonator, the stimulated emission beam
must travel a round trip in the resonator. During this round trip the coherent radiation is
amplified with a gain dependent on the length of the laser rod according to exp(2glrod ).
However, taking into account that part of the laser beam is outcoupled through the mirror
R1 , the global round-trip gain is reduced to G = R1 exp(2glrod ). Similar to the absorption
coefficient α, the emission coefficient equals n2 σe ; however, taking into account the possible
reabsorption of the amplified beam due to a residual population of the terminal level, the net
effect on the amplified beam can be characterized by an effective gain coefficient g = nσe
with n = n2 − n1 .
The laser material and the resonator can introduce additional losses at the wavelength
of laser radiation, such as parasitic absorption, scattering, diffraction, residual Fresnel
reflection and so on, and these are coined in a loss coefficient L. The amplified beam
becomes intense enough to overcome the residual losses and the outcoupling loss when
R1 exp(2gl − L) ≥ 1; (2.18)
the strict equality in this equation defines a threshold condition for laser emission,
2gl = −ln R1 + L; (2.19)
usually the reflectivity R1 is large enough and ln R1 = ln(1 − T ) ≈ −T where T is the
transmission of the exit mirror, and
2gl ≈ T + L. (2.20)
2.2 Solid-state lasers 25
Below the threshold, the losses block the stimulated emission flux and a small-signal
gain coefficient g0 can be defined; immediately above the threshold, the stimulated emission
is amplified according to the gain g0 but for higher excitation the amplification of the light
beam continues until the gain saturates for the inversion of the population determined by
the actual pump intensity. The value of the saturated gain is determined by the pumping
conditions and by the characteristics of the laser material and the design of the resonator
and thus Eq. (2.18) establishes the connection between the output and input power by
taking into account all these characteristics. The gain threshold condition defines the pop-
ulation inversion and the emission cross-section necessary to overcome the actual losses
of the resonator: a larger emission cross-section determines a smaller threshold population
inversion and thus the laser starts to oscillate for frequencies around the central value ν0 of
the function g(ν, ν0 ) and the gain profile will be considerably narrower than the width of
this function. Nevertheless, pumping high above the threshold determines a broader gain
profile. Since the peak value g(ν0 , ν0 ) of the function g(ν, ν0 ) is inversely proportional to
the linewidth ν, and the Einstein coefficient A21 is inversely proportional to the emission
lifetime, the quantum systems with narrow emission lines and long lifetime will show a
lower laser threshold.
Although the first laser was a three-level system, the ruby laser [1], very soon the advan-
tages of four-level lasers became evident. The quantum systems with discrete energy levels
able to give emission in the optical range (ultraviolet, visible, infrared) with wavelengths
from hundreds of nanometers to tens of micrometers can be free atoms (such as in the
He–Ne laser) or ions (such as Ar+ ), free molecules (CO2 , CO, N2 and so on), excimer
noble gas halide molecules (ArF, KrF, XeF, KrCl, XeCl), ions of elements from the transi-
tional groups (elements with incomplete inner electronic shells) doped in transparent solids,
optically active defects in solids (color centers), solutions of organic molecules (dyes), and
a special category is the semiconductors, where the optical transitions take place between
collective electronic states (the conduction and valence bands). Efficient laser emission
was demonstrated in all types of laser materials; the present book refers to lasers based on
transparent solids doped with laser active ions, called solid-state lasers.
The phase coincidence of the photons in the stimulated emission process gives high
temporal and spatial coherence to the amplified beam: it is monochromatic and propagates
with reduced divergence. These properties open a very broad area of applications, based on
the quality of laser radiation as a carrier of energy or information, as a highly monochromatic
source for excitation or as a frequency etalon, as well as primary radiation for non-linear
optical processes.
La nuit…
La nuit, lâche ennemie des faibles : elle abuse contre eux de sa
force obscure, insaisissable. Les enfants, les sans-asile, les malades
s’épouvantent à sentir qu’ils sont sa proie désarmée : les malades
surtout, que leur impotence lui livre enchaînés. Alors le monde,
autour de leur lit chétif, s’espace, se vide, les abandonne à leur
souffrance et à leur angoisse. Alors ils pensent : « Nous sommes un
déchet, un rebut ; on nous tolère ; on ne nous achève point ; mais, le
matin venu, le monde ne s’apercevra même pas si nous ne nous
réveillons pas avec lui. »
L’ombre de la nuit passagère rejoint ainsi, pour consommer leur
détresse, la grande ténèbre éternelle.
Heureux, parmi ces désolés, ceux que des mains pleines de
santé retiennent au bord du ravin d’épouvante. Ils se raccrochent à
cette chose vivante, capable de résister à l’ennemi. Par la vue, par le
contact, par l’avide aspiration des paroles et du bruit, des pas qui
rompent le silence hostile, ils s’incorporent à un organisme sain et
fort, ils s’abritent contre la nuit… Les mains de Madeleine, ses yeux
de lumineuse poussière, son visage rayonnant de jeunesse et de
vigueur paysanne, l’articulation lente et sûre des mots qu’elle
prononce : quel bienfait vaudrait celui-là pour le débris humain dont
elle veille l’insomnie… Il l’a appelée au secours dès que l’autre
femme l’a laissé seul après l’avoir frappé au cœur. Sa détresse
égoïste n’a pas eu un mot de repentir ou de pitié pour le mal que
l’humble gardienne a enduré, qu’il sait qu’elle a enduré. Il a dit
brièvement, sèchement :
— Aide-moi à me coucher. Et que personne n’entre, personne…
tu entends ?
Mais, dans la calme tiédeur du lit, sa colère s’est peu à peu
détendue. Il s’est laissé bercer, caresser par l’enfant maternelle,
comme un enfant plus débile. Devant elle, si humble, si discrète,
pourquoi se contraindre ? Il a laissé couler des larmes qui
l’oppressaient et dont le cours l’a soulagé. Il s’est plaint puérilement
du mal qu’il a souffert. Il a confessé la peur atroce qui l’étreint depuis
que des mots irréparables ont été prononcés : la peur de mourir.
Oui, ce même prince Paul qui, dans les combats du front oriental,
s’est fait gourmander et punir par ses chefs pour sa témérité
maladive, la peur de mourir le mue en un chiffon humain. Il s’agrippe
aux poignets de Madeleine.
— Dis-moi la vérité. Pas de dérobades ni de défaites. Est-ce que
je suis perdu ? Je veux savoir !
Ah ! le solide appui, l’efficace réconfort — ce regard, inflexible
sous l’attaque de son regard et la réponse de cette bouche qui ne
peut pas mentir :
— Jamais le médecin ne m’a dit pareille chose. Vous avez été en
danger, alors que vous ne pouviez pas vous lever. Mais quelle
maladie ne comporte pas de danger ? Une des novices de la
Quarantaine est morte à vingt ans sous nos yeux pour une piqûre de
mouche.
Il pense :
« Comme c’est vrai, ce qu’elle dit ! On ne sait rien à l’avance…
J’ai été déjà plus malade que je ne suis, et j’ai guéri. »
— Si tu me soignes bien, fait-il, je ne mourrai pas. J’ai confiance.
— Je vous soignerai de toutes mes forces.
Les mains de l’ange, à force d’être serrées et retenues, sont
presque douloureuses. Et il ne lui permet pas non plus de se taire ; il
veut l’entendre parler, car cette voix l’empêche de penser par lui-
même. S’il était capable, dans sa détresse, de comparer ce qu’elle
dit avec ce qu’a dit Stéphanie, comme il admirerait le don inné, chez
la paysanne inspirée, de l’apostolat consolateur ! Pourtant, elle ose
lui parler de cela même qui l’épouvante, de la Visiteuse suprême
dont il croyait tout à l’heure entendre le pas derrière le seuil. Mais
dans les propos de Madeleine, il ne s’agit plus d’un danger qui le
menace, lui, le malade. Elle raconte avec simplicité comment elle-
même envisage le terme de sa propre vie ; combien il est aisé,
quand le cœur est paisible et net, de vivre amicalement avec une
telle pensée. « Je serais bien malheureuse, dit-elle, si je devais
échapper au sort commun, et voir éternellement les jours et les nuits
se succéder… » Il écoute avidement. Il lui semble que le péril n’est
plus pour lui seul ; que c’est comme au front d’Orient : une menace
confuse, aussi probable pour n’importe qui que pour soi. On ne sait
où l’obus éclatera ; on se fie à sa chance. Et la Visiteuse effrayante
finit par se muer en une compagne douce et pitoyable qui nous suit
tout le long de la vie et se rapproche enfin de vous pour soutenir et
recueillir les derniers pas.
Madeleine sentait mollir peu à peu l’étreinte angoissée de ses
mains et leur fièvre fléchir. Les paupières du malade commencèrent
de battre, puis s’abaissèrent. Cependant Madeleine ne cessa point
de parler, connaissant le pouvoir de sa voix… Mais, peu à peu, au
lieu de l’exhorter, elle murmura simplement des prières. Non pas
uniquement des prières apprises, mais ses prières « à elle », ces
conversations implorantes qu’on ne lui avait jamais enseignées, et
qu’elle n’interrompait guère, à travers les travaux de sa vie… « Ma
chère patronne, je vous supplie de réconcilier ce pécheur, et je
m’offre à être sa rançon. Le Sauveur s’est bien chargé des péchés
du monde : une pauvre petite chose comme moi ne peut-elle pas
être sacrifiée au salut d’une âme ? Ma chère sainte patronne, je
vous en prie, je vous en prie… »
Il reposait à présent. Elle mit silencieusement le lit bien en ordre ;
puis elle traça du pouce droit, sur le front du malade, les deux traits
perpendiculaires de la croix. Ensuite, elle éteignit la lumière des
ampoules, et, à la lueur de la veilleuse, gagna sa couche, où elle
s’étendit, parée à toute alerte.
Le lendemain, au cours de la matinée, il attira contre lui la tête de
la jeune fille et l’étreignit tendrement, sans l’ombre de perversité.
— Hier, lui dit-il, tu m’as fait beaucoup de bien, et moi, je t’avais
fait du mal. Pardonne-moi. Je ne vaux rien… et puis, je suis malade
et persécuté. Je n’ai plus que toi au monde. Je ne veux plus voir que
toi. Ne laisse entrer personne !
Elle eut assez de peine à le convaincre qu’il ne devait pas fermer
sa porte au comte Osterrek.
— Celui-là vous aime à sa manière, mais il vous aime fidèlement.
Maintenant elle le gouvernait. Il consentit à le recevoir quelques
instants, après la visite du médecin : son humeur était adoucie parce
que Burcart, le trouvant en meilleur état, lui avait permis de se lever.
Le prince s’amusa de l’air embarrassé de son camarade. Il
s’obstina à lui parler de tout, sauf de Stéphanie, comme s’il tenait
pour inexistante la rencontre de la veille. Et le comte lui demandant
ses ordres pour la journée :
— Mon vieux, lui dit-il, je veux qu’on me laisse vivre aujourd’hui
une vie végétative ; je ne veux parler à personne ; je ne veux voir
personne que l’ange. Si tu as une communication à me faire,
demande l’ange.
La figure bilieuse du comte se crispait, et son maître vit dans le
coin de ses paupières fripées quelque chose d’humide qui reflétait la
lumière.
— Ne te chagrine pas, vieux camarade, lui dit-il. Je te connais, et
je sais que toi, au moins, tu n’essayeras pas de me faire du mal
exprès.
Il lui serra la main fortement. Le comte sortit sans pouvoir
prononcer un mot.
Ces choses se passaient aux environs de midi. Un brouillard
lucide, mais pourtant impénétrable au regard, s’exhalait du lac et
tendait une gaze claire derrière les vitres. Tout d’un coup cette gaze
se déchira et, par la déchirure jaillit, à la manière des projections de
théâtre, un faisceau de clarté jaune et chaude. Puis le voile entier se
partagea en lambeaux, s’émietta, se volatilisa, et le paysage du lac
redevint éblouissant. Aussitôt le prince voulut se lever. Son valet de
chambre et Madeleine hâtèrent sa toilette. Il avait faim.
— Qu’on serve le lunch dans le petit salon, dit-il. Toi, Madeleine,
tu déjeuneras avec moi. Ah ! pas d’objection, n’est-ce pas ? Tu
déjeuneras avec moi.
Elle y eut moins d’embarras qu’on ne l’aurait supposé : rien ne la
troublait, des événements de la vie matérielle, parce qu’elle ne leur
demandait ni joie ni profit. La gaîté du prince, pendant ce repas tête
à tête, la surprit et la ravit. Gaîté un peu nerveuse, réaction contre la
mortelle angoisse de la veille : on eût dit qu’il prenait sa revanche
sur Stéphanie et voulait se prouver à lui-même qu’il renaissait, qu’il
vivait. Il ordonna à Madeleine de goûter au champagne sec qui était
sa boisson ordinaire et que le médecin lui permettait à doses
modérées ; il s’amusa de la grimace qu’elle fit, buveuse d’eau et de
bière légère depuis l’enfance ; sept ou huit fois dans sa vie (la
dernière au Café franco-suisse) elle avait goûté à du vin rouge…
Mais comment eût-elle résisté au bonheur de voir son malade
rasséréné, et dépouillant pour elle tout ce qu’elle redoutait de lui : la
perversité et l’ironie ? Il était sincère quand il lui dit :
— Je ne vaux rien, et malgré deux rudes coups que j’ai reçus ici,
je n’ai pas changé. On ne change pas !… Si je redeviens solide et
libre, j’ai bien peur de recommencer ma mauvaise vie. Mais j’ai
changé pour toi, et aucune femme n’aura connu l’homme que je suis
avec toi. Je suis désarmé contre toi, comprends-tu ? parce que j’ai
besoin de toi, telle que tu es, petite sainte, et que j’ai une peur
superstitieuse de te défaire… Tu ne peux pas comprendre !
En effet, elle ne comprenait pas : mais les mots qu’elle écoutait la
berçaient, et elle laissait dire. S’il l’attirait contre lui et lui baisait les
yeux et les joues, elle ne résistait pas : elle sentait que d’un mot elle
pouvait l’arrêter. Ces caresses lui étaient douces comme celles d’un
frère chéri ; elle n’en était point troublée. Un contact, une pression de
main, un baiser ne risquaient pas de l’émouvoir : c’était du fond de
son âme, de ce mystérieux inconscient où sa féminité demeurait
tapie, intacte et pressante, que pouvait monter la révélation de
l’amour.
Quand cette dînette, où le prince s’enchanta lui-même à une
façon d’aimer qu’il n’avait plus pratiquée depuis ses sorties de
collégien chez le général Delenca, s’oubliant exprès et cherchant le
bonheur de l’autre, il exigea que l’ange s’étendît sur le divan du petit
salon et se reposât.
— Mais je ne suis point lasse !…
— C’est un ordre. Tu as beau être en acier, on ne résiste pas
indéfiniment à la vie que je te fais mener. Étends-toi ici.
Lui-même traîna près du divan un gros coussin formant tabouret
et s’y assit ; puis il posa sa tête sur la poitrine de la jeune fille. Elle
s’y prêta sans hésitation ni méfiance. Le malade plongea d’ailleurs
tout de suite dans le plus profond repos, corrigeant l’insomnie initiale
de la nuit précédente. Elle, bien qu’elle ne redoutât rien de lui, ne put
s’assoupir, même un instant. Elle était trop heureuse : il lui semblait
que sa jeune santé enveloppait le dormeur, dans cette maternelle
étreinte, et le pénétrait. Le cou du prince, qu’elle entourait de son
bras, s’appuyait à son sein gauche ; elle percevait à la fois le rythme
de leurs deux vies. D’abord l’artère du malade battit en désordre,
tantôt active à l’excès, tantôt presque défaillante, tandis qu’au flanc
de la paysanne les pulsations se succédaient à intervalles courts,
mais égaux. Puis, comme par l’effet d’une mystérieuse endosmose,
les deux rythmes s’harmonisèrent. Vint un moment où l’oreille de
Madeleine n’entendit plus qu’un seul choc géminé. Elle en fut
puérilement joyeuse. Le trésor de maternité que recèle l’âme de
toute vierge sage, elle l’épancha sur ce viveur dont la débilité
douloureuse refaisait un enfant. Furtive et ardente, elle posa ses
lèvres sur sa tempe et les y laissa appuyées, guettant, pour qu’il ne
surprît pas le secret de ce baiser, le plus léger mouvement de celui
qui reposait sur son cœur.
Aussitôt réveillé, le prince exigea qu’on le descendît sur la
terrasse. Il voulait y poursuivre les dernières clartés du jour. A
mesure que déclinait le soleil, son alacrité maladive s’exténuait : le
souvenir de l’atroce soirée de la veille le harcelait, et il s’effarait
devant l’angoisse nocturne. Vainement sa compagne essaya de le
distraire : il lui souriait, il lui parlait avec tendresse : mais elle le
sentait inquiet. Il prolongea sa station sur la terrasse jusqu’à l’heure
où le brouillard montant du lac commença de l’oppresser. Revenu
dans sa chambre, il resta longtemps silencieux. Madeleine avait pris
un ouvrage et travaillait, silencieuse aussi, tout près de lui. Quand
elle levait les yeux sur lui, il souriait encore ; parfois il paraissait sur
le point de lui parler. Finalement, il continuait de se taire. Son lit, où il
avait senti, la veille, perler sur son visage la sueur mortelle, lui faisait
peur ; il refusa de se recoucher selon l’habitude, pour prendre son
repas du soir. Il dîna comme il avait déjeuné, tête à tête avec l’ange,
dans le petit salon. Là, il s’égaya un peu : cette séduction du geste,
de la voix et des mots à laquelle si peu de femmes avaient su
résister, il la déploya pour l’humble fille qui était toute conquise, prête
à lui sacrifier sa vie, et plus que sa vie. Mais il ne disait toujours pas
ce qui le hantait, et Madeleine n’arrivait pas à le deviner. Il se décida
pourtant :
— La nuit dernière, dit-il, si je n’avais pas eu tes mains et ta voix,
je crois que j’aurais étouffé de désespoir. Mais sais-tu le plus grand
bien que tu m’aies fait ?
— Non…
— Eh bien… C’est tantôt… quand tu as mis tes lèvres sur mon
front.
Elle devint pourpre.
— Oh ! vous ne dormiez pas ! c’est mal.
Mais comme elle disait cela, elle sentit poindre en elle un trouble
singulier, où il y avait de l’inquiétude et du bien-être.
D’une voix qui tremblait et se trouait comme celle d’un
adolescent épris, il chuchota, près d’elle :
— Alors ?… Tout le ravage que la souffrance a creusé dans mon
pauvre visage ne t’éloigne pas de moi ?
Elle répondit :
— Bien avant de vous connaître, j’ai aimé votre visage.
Il ne comprit pas : il ne pouvait pas comprendre. Que de paroles
prononcées par cette enfant lui demeuraient inexplicables et dont le
sens mystérieux lui imposait cependant ! Il l’attira contre lui et, de
ses lèvres fermées, effleura sa bouche close ; un grand
frémissement le convulsa. Elle, au contraire, soudain anxieuse,
sentit s’évanouir l’émoi de tout à l’heure.
— Allons, dit-il, appelle pour qu’on me déshabille et qu’on me
couche. Tu ne me quitteras pas ? Tu resteras près de moi comme
hier ?
Elle eut un rire qui l’enchanta, en disant :
— Vous le savez bien !
Depuis que la blessure était cicatrisée, c’était le valet de chambre
qui le dévêtait, aidait à sa toilette et lui passait son pyjama de nuit. Il
fit rappeler Madeleine quand il fut couché. Elle eut la surprise de le
trouver calme ; elle ne se doutait pas qu’elle avait cicatrisé d’un mot,
tout à l’heure, une blessure plus douloureuse que celle du stylet : la
blessure faite au prince par Stéphanie quand, au contact de son
visage, elle avait laissé percer sa répugnance… A présent, le prince
se complaisait à penser : « Un jeune être sain et sincère comme
cette enfant a du plaisir à me regarder, à mettre ses lèvres sur ma
tempe. » Il fut tendre et simple avec elle ; il s’inquiéta de sa fatigue,
et comme, tout en lui tenant les mains, il la voyait fléchir sous le
sommeil, il lui dit :
— Je me sens bien et je suis sûr de m’endormir. J’exige que tu
ailles te reposer dans ta chambre et dans ton lit. Je ne veux pas
abîmer mon ange.
Elle résista, mais il tint bon. Elle le signa de son pouce sur le
front, puis, brisée et comme grisée de fatigue, elle gagna docilement
sa chambre et son lit.
Elle dormit enfin ; elle dormit longtemps. D’abord d’une torpeur
où se dissolvait l’extrême lassitude de son corps, puis d’un sommeil
moins opaque, où transparurent les souvenirs de cette émouvante
journée. Elle revécut le plus doux de tous, quand la tête de son
maître, de son enfant, reposait sur elle. De nouveau, elle osait lui
presser le front de ses lèvres, et son bonheur grandissait à mesure
que leurs deux cœurs s’approchaient de battre d’accord… Quel
sommeil pourrait résister à tant de joie ! Le voile peu à peu s’amincit.
Est-ce qu’elle dort ? Elle a peine à ressaisir la réalité des choses.
Non, elle ne dort plus. Elle est toujours étendue dans son lit, et
pourtant sa bouche est sur le front du maître, et le rythme de leurs
deux vies palpite à l’unisson. Il dit tout bas :
— Si tu m’ordonnes de partir, je partirai.
Elle le serre plus tendrement encore ; il est immobile ; il ne parle
plus… Voici qu’il s’assoupit de nouveau, et qu’elle-même, une douce
torpeur la reprend…
Comment son innocence pourrait-elle pressentir et craindre la
puissance magnétique du sommeil à deux et le déchaînement des
forces de l’amour, alors que la conscience ne les contrôle plus ?
Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.
ebookfinal.com