Trends in Quantum Electronics Proceedings of the 2nd Conference, Bucharest, September 2-6, 1985
Edited by A.M. Prokhorov and I. Ursu
Springer-Verlag Berlin Heidelberg GmbH
A.M. Prokhorov
Institute of General Physics Academy of Sciences Vavilov Street 38 SU-117942 Moscow Soviet Union
I. Ursu
ISBN 978-3-662-10626-6 ISBN 978-3-662-10624-2 (eBook) DOI 10.1007/978-3-662-10624-2
This work ist subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically those of translation, reprinting, re-use of illustrations, broadcasting, reproduction by photocopying machine or similar means, and storage in data banks. Under§ 54 of the German Copyright Law where copies are made for other than private use, a fee is payable to "Verwertungsgesellschaft Wort", Munich. ©Springer-Verlag Berlin Heidelberg 1986 Originally published by Springer-Verlag Berlin Heidelberg New York in 1986 Softcover reprint of the hardcover 1st edition 1986
2153/3140-543210
PREFACE
The second International Conference on Trends in Quantum Electron­ ics (TQE'85) was held in Bucharest at the National Centre for Physics in September 1985, and brought together more than 350 scientists from 22 countries.
In accordance with the objectives established at the first confer­ ence, which was held in 1982 in conjunction with the third Interna­ tional Summer School in Coherent Optics, the second conference con­ centrated upon the central topics and chief directions of develop­ ment in quantum electronics - which stands out as an area of science and technology that is currently expanding vigorously. On the other hand, it was also apparent that TQE'85 was primarily influenced by the worldwide celebration, in 1985, of the 25th anniversary of the laser - a moment of prime importance in the development of many frontline fields, including communications, chemistry, biology, health care and materials processing. A special session was devoted to this anniversary.
In keeping with the dynamic spirit of the conference, the fine quality of the invited lectures and the other contributions set a high scientific standard for the proceedings. Mention should be made of the posters that were presented throughout the conference. These, together with the exhibition of books and journals and a display of specialized scientific equipment, did much to create a framework for effective communication and stimulating interaction, to the benefit of all the participants.
Of invaluable help in the preparation of the conference was the collaboration of the International Scientific Advisory Committee. Their suggestions and proposals greatly contributed to the final sci­ entific programme, and we take this opportunity to express our grati­ tude to them once again.
We also thank the European Physical Society, especially its Quan­ tum Electronics Division, the Academy of Sciences of the USSR and the Society for Optical Engineering for sponsoring the conference and for their continuing help and encouragement. TQE'85 benefited from sub­ stantial support and assistance of the Romanian Academy and of the Romanian National Committee for Physics. We gratefully thank them, as we also thank the Central Institute of Physics , and its Director General, Prof. Marin Ivascu, for making it possible to hold this in­ teresting and pleasant conference under such good conditions.
Thanks are due to Springer-Verlag and Dr. H.K.V. Latsch for advice during the prepar~tion of these proceedings. We also wish to express our appreciation to the scientific secretaries and to the Information and Documentation Office of the Central Institute of Physics for their contribution to the organization of the conference, and for the preparation and printing of the proceedings.
Finally, a large share of our gratitude should go to the partici­ pants themselves, for it is their contributions and discussions that finally created this scientific event, which we hope will be a focus of the attention and interest of the whole quantum electronics com­ munity.
A. M. Prokhorov Moscow
I. Ursu Bucharest
TABLE OF CONTENTS
Achievements in the Field of Solid-State Lasers By A.M. Prokhorov . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Recent Advances in Ultraviolet Laser Technology By J.F. Figueira, S.R. Foltyn, L. Rosocha, R.C. Sze, C.R. Tallman, J. Telle, D.E. Watkins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
The Pulse-Repetition Operation Regime of a Chemical Oxygen­ Iodine Laser
By N.G. Basov, P.G. Kryukov, N.N. Yuryshev . . . . . . . . . . . . . . . . . . 23
Experiments on Mode Pulling in a C0 2 Laser By U.P. Oppenheim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Infrared Solid State Lasers By I. Ursu, V.I. Lupei . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Cascade Stimulated Emission in Crystals with Several Meta­ stable States of Ln3+ Ions
By A.A. Kaminskii, V.A. Fedorov . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Recent Developments in Monochromatic Birefringent Filters By M. Francon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Laser Research and Applications in China By Wang Zhijiang . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Helical and Linear Distributed Feedback Lasers By F .K. KneubUhl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Spectral, Luminescence and Lasing Properties of Yttrium, Scandium, Gallium Garnet Crystals with Chromium and Erbium
By E.V. Zharikov, N.N. Il'ichev, S.P. Kalitin, V.V. Laptev, A.A. Malyutin, V.V. Osika, P.P. Pashinin, A.M. Prokhorov, Z.S. Saidov, V.A. Smirnov, A.F. Umyskov, I.A. Scherbakov 115
Tunable Dye Lasers in the Visible and Ultraviolet and their Spectroscopic Applications
By M.L. Pascu, A. Pascu, G. Dumbraveanu, A. Vasile, D. Cristu, M. Munteanu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
"Channeling" of Relativistic Electrons in a Periodic E.M.Potential liy M. Bertolotti, C. Sibilia, Li Fuli . . . . . . . . . . . . . . . . . . . . . . 155
Optoga~vanic Laser Frequency Stabilization By D.C. Dumitras, D.C.A. Dutu, V. Draganescu, N. Comaniciu 161
VII
Thermodynamics of Laser Systems By X. de Hemptinne . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
The Applicability of the Mean-Field Method in Optical Bistability By I.M. Popescu, E.N. Stefanescu, P.E. Sterian . . . . . . . . . . . . . . 173
Nonlinear Guided Waves: Physics and Applications By G.I. Stegeman, C.T. Seaton . . . . . . . . . . . . . . . . . . . . . . • . . . . . 179
Laser-Plasma Detection: Generation of Ultrahigh-Frequency Currents on Solid Surface Exposed to Laser Radiation
By A.A. Antipov, N.G. Basov, A.Z. Grasiuk, V.V. Losev, A.P. Lutsenko, E.A. Meshalkin . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Electron Stimulated Desorption (ESD): Principles, Techniques and its Applications for Surface Molecules Analysis
By J .L. de Segovia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
Intense UV-Laser Radiation Interaction with Surfaces By V.P. Agee~, A.F. Chernishev, A.A. Gorbunov, V.I. .Konov, A.V. Kuzmichov, A.A. Orlikovsky . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Fundamentals and Applications of the Interaction of Picosecond, One Micron Radiation with Crystalline Silicon ·
By I.W. Boyd, T.F. Boggess, A.L. Smirl, S.C. Moss . . . . . . . . . . . . 281
Time-resolved Infrared Spectroscopy on the Picosecond and Sub­ Picosecond Timescale Studying Coherent Pulse Propagation
By A. Laubereau . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Studies of Metal Oxidation under the Action of CW C0 2 Laser Radiation
By I. Ursu, L.C. Nistor, I.N. Mihailescu, V.S. Teodorescu, L. Nanu, A.M. Prokhorov, V.I. Konov, I. N. Chapliev . . . . . . . . . . . . . . . . . . 313
Laser Induced Phase Transformation and Vapour Deposition of Amorphous and Crystalline Carbon Phases
By W. Pompe, H.-J. Scheibe, G. Kessler, A. Richter, H.-J. Weiss 331
Subnatural Linewidth Laser Spectroscopy By H. Takuma, F. Shimizu, K. Shimizu . . . . . . . . . . . . . . . . . . . . . . 357
Analog Optical Computing By A. Herden, T. Tschudi
Picosecond Sp~ctroscopic Investigation on Surfaces By F.R. Aussenegg, S. Draxler, A. Lettner, M.E. Lippitsch,
369
Optogalvanic Spectroscopy By I.I. Popescu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391
Laser Interferometry at 10.6 Jl1TI. for Plasma Diagnostics By I. Ursu, D. Apostol, M. Stoica, I. Apostol, D. Craciun, A. Hening, I.N. Mihailescu, V. Stancalie . . . . . . . . . . . . . . . . . . . 443
VIII
Inertial Confinement Fusion with Lasers or Particle Beams By P. Mulser .......................................... .
Optically Pumped FIR Lasers and Their Application in the Plasma Diagnostics
By J .s. Bakos
Heterogeneous Selective Processes under the Action of Laser Radiation
By N.V. Karlov, A.S. Laguchev, A.N. Orlov, Yu.N. Petrov, A.M. Prokhorov, M.Ya. Yakubova .......................... .
Trends of Laser Technology in Microelectronics By S. Metev ........................................... .
Some Properties of New Single-Frequency Stabilized He-Ne Lasers By F. Petru, Z. Vesela ................................. .
Laser Applications in Ophthalmology By M. Olteanu, B. Carstocea, M.L. Pascu, D. Dutu, v. Draganescu, R. Dabu, A. Pascu ...................................... .
High-Power Fast-Flow C02 Laser Facility By J. Stanco, E. Antropik, P. Grodecki, M. Irczuk, J. Konefal, P. Kozyra, P. Kukiello, W. Mikienko, M. Piskulski, G. Rabczuk, Z. Rozkwitalski, T. Stelter, G. Sliwinski, R. Werden, R. Zaremba
463
481
505
517
531
539
551
A. tot. PRCI<HOROV
1. lntrocb:tion
In 1985 all the developed countries of the world widely celebrated the 25th anniversary
of the creation of the first laser. The advent of laser in 1960 was not unexpected. The
way for it was paved by all the previous development of physics. By the end of the 50's
the most important ideas had been formulated which formed the basis for a new field
of physics - quantum electronics: the ideas of Einstein and Dirac concerning induced
radiation, the conceptions of inverse population of energy levels in quantum systems
and of active media capable of amplifying incident monochromatic radiation,
radiophysics ideas of feed-back, resonators and monochromatic wave gener!!tion.
It is necesary to stress that though the principles of Einstein on induced
radiation had been formulated in the first third of our century, none of those engaged
in optics made attempts to develop a laser. The experiments of those times were made
for the only purpose to prove the existence of the induced radiation. The possibility of
creating monochromatic light oscillators on the basis of induced radiation was not
considered. The problem was put forward only after creating masers. In radiophysics,
unlike in optics, there had already been monochromatic oscillators and the theory of
auto-oscillating systems had been well developed. Thus, it was obvious that in order to
have an amplifying element, it was enough to introduce a feed-back resonator to make
this system produce monochromatic oscillations. In 1954-1955 the first masers on
molecular ammonia beams were developed (J. Gordon, G. Zeiger and C. Townes in the
U.S.A; N.G. Basov and A.M. Prokhorov in U.S.S.R.). After the creation of masers the
question arose how to advance to the optical region. In 1955, N.G. Basov and A.M.
Prokhorov proposed a new method of attaining inverse population, known as the
method of three levels. According to it a system consisting of three energy levels is
aff.ected by a high-power external source-pumping, which excites it. As a result,
population inversion occurs between one of the pairs of levels. A considerable
contribution to the development of this method was made by Bloembergen (U.S.A).
This idea allowed to create in 1957-1958 solid-state masers on paramagnetic crystals,
processing in the limiting low intrinsic noise (H.E.D. Scovit, G.Feher and H. Seidel in
the U.S.A.; G.M. Zverev, L.S. Kornienko, A.A. Manenkov and A.M. Prokhorov in the
U.S.S.R.). This idea also facilitated the development of laser, as it solved the problem
2
of attaining an inversion population in the optical region. Another obstacle in
advancing to the optical region was the cavity. The matter is that for short radio
waves, the cavities have dimensions which are comparable with the wavelength. But
when proceeding towards millimeter and submillimeter waves the cavity dimensions
become so small that they lose sense. In 1958, the author of this paper suggested so
called open cavities whose dimensions are far greater than the wavelength. The next
essential step was made by A.L. Shawlow and C.H. Townes in U.S.A. Open cavities are
now widely used in lasers.
As is known, the ruby laser was the first in the world (1960, T.H.Maiman). It
was followed by the other lasers both on crystals and other types of active media, such
as neon-helium gas mixture (1960), lime-silicate glass doped with neodymium ions
(1961), semiconductors (1962), neodymium solution in inorganic liquid of selenium
oxichloride (1966). At present lasers have been developed using as active media the
substances in all aggregate states: solid, liquid, gaseous and in the plasma state.
During the past 25 years the lasers of different types have been continuously modified,
many of them becoming indispensable for a series of important practical applications,
such as engineering, communication and medicine, technology and location, science
and ecology. In all the pointed fields the application of lasers has led to revolutionary
changes.
It should be noted that the past quarter of the century has been characterized
by a strong competition among different types of lasers. In different periods one or
another type was given preference. Despite the strong. competition the most
practically used became the solid-state laser - on crystals and glasses. For instance,
according to statistics presented in /1/, 42 % of the sum from the sale of lasers on the
world market (without socialit;t countries) falls on solid-state lasers. More than half of
the solid-state lasers sold are Nd:YAG lasers. These data reflect the state of the art in
1982. For the last three years the contribution of solid-state lasers has had a tendency
to further increase.
The reason for such a wide spreading of solid-state lasers is that, though they
are not record-breaking by separate parameters (for example, mean radiation power
gain), the aggregate of the most important parameters makes them undoubted leaders.
The major benefit of solid-state lasers , consists in their combining high energy
parameters (pulse energy is up to 20 KJ, mean power is about 1000 W) with an ability
to perform reliable and long-term operation in various modes from ultrashort
picosecond pulses ( < 1ps) up to c.w. high-power operation, and they are compact.
Even though solid-state lasers have been widely used for solving scientific and
practical problems, further work on improving their characteristics has not been so
intensive due to the development of new-type lasers, such as tunable dye lasers,
excimer lasers, recombinational plasma lasers, metal vapour lasers, etc. Moreover,
3
there was an opinion that further development of solid-state lasers was not promising.
In the General Physics Institute in Moscow much effort has been concentrated
on solid-state lasers research. To improve their characteristics we have studied in
detail active centers /2, 3/, the processes of excitation energy transfer /4, 5-7/, the
properties depending on the composition and structure of the laser matrix /4-6, 7-10/..
As a result of these works a new generation of crystals and glasses has been created.
As is known, a considerable drawback of lasers on crystals and glasses is their
low efficiency. The most widely used solid-state laser employs yttrium-alluminium
garnet with neodymium as an active ion. This laser has an efficiency of 1 t 2 %. This is
due to the fact that only a small portion of the pump radiation gets into the bands of
neodymium absorption. To increase the efficiency of this laser one may increase
neodymium concentration in order to increase the optical density of the active
medium, or dope with the other ion which absorbs well the pump radiation and then
efficiently transfers excitation to neodymium. Though these principles of increasing
the efficiency are simple, their realization encounters many difficulties. Increasing
neodymium concentration in Y AG crystals leads to luminescence quenching. In order
to clarify the opportunity of using this technique, it was necessary to study in detail
the mechanisms of concentrational luminescence quenching, which was done in /20-
29/. As a result, crystals and glasses with high neodymium concentration and low
luminescence quenching have been developed. For example, in our Institute
concentrational neodymium phospate glasses CNPG have been created /10/, followed
by Q-100 glasses /30/, LGH-8 /31/. The efficiency of the lasers based on these glasses
is twice or three times higher than that of Nd:YAG and glass lasers.
Another way which has been used to increase the efficiency of laser crystals is
to dope with ions which absorb well the pump radiation and efficiently transfer
excitation to the active ions. The first attempt to use an effect of sensibilization for
increasing the efficiency of neodymium lasers was undertaken in 1964 /32/. The
Nd:YAG crystal was taken for this purpose, and was doped with cr3+. But no
significant results were obtained. It took us ten years to comprehend the conditions
under which this sensibilization is efficient. The first step was made in /33/ with the
use of a complex chromium-neodymium alumborate crystal. The authors of this paper
attributed the high efficiency of sensibilization attained only to the small distance
between the ions of Cr3+ and Nd3+.
An important step in understanding the mechanism of energy transfer in the
Cr3+ -Nd3+ pair was made in /5,6, 34-36/. It was shown that a decisive role is played by
the energy gap tc. E between 2E and 4T 2 chromium energy levels. This gap depends
upon the composition and structure of the crystal matrix. A rapid energy transfer from
Cr3+ to Nd3+ is possible only if tc. E < kT /1/. The cr3+ levels in Nd:YAG
4
crystal do not satisfy this condition and, therefore, Cr3+ in Nd:YAG does not provide
an essential advantage. A search of crystals and glasses was undertaken which would
make it possible to attain an essential increase in efficiency when doping the crystal
with Cr3+. As a result of this search it turned out that a complex gallium garnet with
gadolinium, scandium, chromium and neodymium is an efficient laser crystal referred
to as GSGG. The luminescence spectrum of Cr3+ in GSGG at 300°K is given by a broad
band with a maximum of 750 nm. Luminescence from the 2E level is absent because of
a very small gap between 2E and 4T 2 levels. In the absence of Nd3+ the 4T 2cr3+ level
has an exponential decay and does not depend on cr3+ concentration up to 6.1020 111, 37 I and the decay time is 120 11s. When doping the crystal with Nd3+ and 4T 2cr3+ level
population sharply decreases pointing that an efficient energy transfer from Cr3+ to
Nd3+ occurs. The first results with the Nd:Cr:GSGG-laser were reported in 111-13, 191.
It is of interest to present the data on Nd:Cr:GSGG - lasers pumped with low
energy 1141. Neodymium and chromium concentrations were 3.5.1020 em -3 and 2.1020
cm-3, respectively. The size of the active element was 3 x 50 mm2• The efficiency of
the pulsed laser was 5.5 % at the pump energy of 3.2 J and 4 % at the pump energy of
1 J. The Q-switch laser (electrooptical modulator) had the efficiency 1.5 % at a 2 J
pump energy.
Pulsed pumped lasers at 50 Hz with active rods 5 x 88 mm produce an average
power of 40 W with an efficiency of 4.4 %. The Nd:Cr: GSGG crystals enable to
provide an efficient lasing at transition to the ground state (transition 4F 312 .- 4I2/3) at ;1.= 0.936 11m.
The appearance of Nd:Cr:GSGG - crystals has stimulated the use of the
sensitization phenomenon with other laser ions. Thus, an efficient energy transfer to 3+ 3+ 3+ 3+ 3+ 3+ J+ 1 1 Pr , Sm , Dy , Ho , Er , Tm , and Yb should be expected 38 • It was
reported in 1151 on fabrication of the Er3+ - glass laser doped with Cr3+ and Yb3+
ions, emitting at 1.5 11m. Doping with Cr3+ ions led to a decreased threshold of
oscillation and a 3-fold increase in the efficiency.
Tunable dye lasers have been widely used. As to solid-state lasers, the efforts
for the development of tunable solid-state lasers have been started only recently,
using mainly colour-center lasers. We shall consider only this type of lasers. The first
laser of this type was fabricated already in 1965 1391 however it is only since 1974
that work in this field was continued 140, 42, 431. The review papers 116-18, 44-461
have considered the physical processes of formation of different colour centers in
tunable laser crystals. These lasers cover a wide range from 0.65 11m up to 3.3 11m and
have a high frequency stability. At present, laser action is attained on colour centers
in the crystals of lithium fluoride, sodium fluorides and chlorides, potassium fluorides
and chlorides, rubidium chlorides, as well as fluorides of calcium, strontium,
magnesium and· a complex fluoride KMgF 3 1471. Reports have appeared on attaining
5
oscillation on colour centers in CaO /48/ and Al2o3 /49/.
In the crystals of LiF and alkali halogenides the most simple colour center is
the F -center, that is, an anion vacancy capturing one electron. Colour centers for
lasers are more complex. For instance, the F 2 center consists of two F -centers located
in the neighbouring lattice sides: Fi is a pair center losing one electron ; F 2 -is a pair
center capturing a third electron. Sometimes the crystal lattice is doped with cation
impurity (for instance u+ instead of K+ in the KCl lattice). In this case, if there is an
impurity cation near the F -center, such a center is denoted by FA. If there are 2
impurites near the center, it is denoted by F 8 .
Scientists in the U.S.A. and F .R.G. /39-41, 44, 46/ wide'ly use FA and F 8 centers for tunable lasers operating below 100°K. Soviet physicists employ Fi and Fz
centers. Such lasers operate at room temperature. To overlap a wide range af tuning
one has to use several colour centers. Fitting approximate crystals with the similat
type of centers it is possible to overlap almost continously the range from 2.2. up to
3.3 ).Lm for the F A(II) and F 8(II) centers and from 0.82 up to 2 ).Lm for the Fi center.
A new group of crystals containing FA centers, where n+ /50, 51/, or Ag +
/51/are used for doping, makes it possible to produce lasers overlapping the range from
1.3 up to 1.8 ).Lm.
To extend the region of tunable oscillation one may use complexes, i.e. quasi­
molecular centers F 2 and Fi associated with the impurities of univalent and bivalent
metals. They are denuted as (F 2) A and (Fi) A and their bands of absorption and luminescence shift by thousands of inverse centimeters, thus providing an opportunity
to extend the region of tunable oscillation /16-18, 45, 46, 52-56, 71/.
Nowadays LiF lasers with F 2, Fi, and Fz centers are the most advanced.
The main shortcoming of the Fi- centers is their low thermal stability. The
problem of thermal stability has been however solved in /53,54/, where Fi-centers are
produced in LiF having at the beginning only F 2-centers. This is attained by converting
F 2 into Fi using a frequency-doubled neodymium laser (0.53 J.Lm). The transfer of F 2 into Fi is a two-stage process /55, 56/. LiF Fi lasers operate efficiently at room
temperature. This crystal was used to produce nano- and pico-second lasers tunable in
the region from 0.8 to 1.1 J.Lm. Using an active element 4 x 8 x 40 mm3 in size and a
flash lamp with a pump energy of 100 J with 25 J.LS pulse duration, a tunable laser was
obtained with a maximum output energy of 14 mJ at ).max = 0.91 J.Lm /61/.
When irradiating F 2 centers in a LiF crystal one can observe not only the
formation of Fi centers, but also the excitation of F 2 centers capable to produce laser
action in the region from 0.65 to 0.75 ).Lm with an efficiency up to 10 % in the
maximum of the tuning curve. It is possible to decrease the transition rate from F 2 to
Fi, which gives an opportunity to work with F 2 centers in LiF with 106 t 10 7 pulses
/55, 56/.
6
The colour centers Fz in UF are promising for the development of tunable
lasers operating at room temperature. When pumping these centers with a nanosecond
pulse from a Nd:YAG - laser a tuning range from 1.08 to 1.25 j.lm with a maximum
efficiency of 25 % at A = 1.17 j.lm was obtained /55, 56/. max A tunable laser has been produced using a large-size LiF -Fz element
(220 x 80 x 40 mm in size), which operates at room temperature with a maximum peak
power of 103 MW with a pulse energy of 100 J and 100 ns pulse duration. Pumping was
carried out with a neodymium glass laser with the pump energy of 700 J /65/.
It was shown in /69/ that Fz centers exhibit nonlinear absorption at A= 1.06
j.lm. This permits to employ them as passive Q-switches /70/. The CNPG laser with
such a Q-swithch emitted 8 mJ at 3 J pump energy /66, 67/. These elements can also
be used in more powerful laser systems. A neodymium glass Jaser with a 40 x 24 x
720 mm2 active element, using LiF -Fz as a passive Q-switch, produced an output
energy of 150 J with 150 ns pulse duration. The pump energy was 7.5.104 J.
In conclusion I would like to say thet solid-state lasers have obtained a second
breath, since their efficiency can be increased 2-3 times and they are promising for
creating tunable lasers covering a wide spectral range with high output energy and
good efficiency.
7 REFERENCES
1 2
26
27
28
29
35
36
37
3a
B.I.Denker, V.V.Osiko, A.M. Prokhorov, I.A.Scherbakov: Kvantovaya electronika 5, 849 (1978) I.A. Bondar, - A.I.Burshtein, A. V .Krutikov, L.P .Mezentseva, V. V .Osika, V.P. Sakun, V.A.Smirnov, I.A.Scherbakov: Soviet JETP §.!., 96 (1981) A.G. Avanesov, T.T. Basiev, Yu.K. Voronko, B.I. Denker I.Ya. Karasik, G.V. Maximova, V.F. Pisarenko, A.M. Prokhorov: Soviet JETP lL 1771 (1979) A.G. Avanesov, T.T. Basiev, Yu.K. Voronko, B.I. Denker, G.V. Maximova, V.A.Myzina, V.V. Osika, V.S. Fedorov: Soviet JETP 84, 1028 (1983) Prospect of the Firm Kigre Inc. "Q-1000 a super-gainlaser glass" Prospect of the Firm Hoya Corporation, May, 1982 Z.J. Kiss, R.C. Duncan: Appl.Phys.Lett. 5, 200 (1964) H.D. Hattendoeff, G. Huber, M.C. Danielmayer: J. Phys. C _g. 2399 (1978) E.M. Dianov, M.V.Dimitruk, A.Ya.Karasik, E.O.Kirpichenkova, V.V.Osiko, V.G.Ostroumov, M.I.Timoshechkin, I.A.Shcherbakov: Kvantovaya electronika 7, 2105 (19aO) - A.G. Avanesov, B.I. Denker, V.V.Osiko, V.G.Ostroumov, V.P.Sapun: Kvantovaya electronika 2, 682 (19a2) .E.V.Zharikov, S.V.Lavrischev, V.V.Laptev, V.G.Ostroumov, Z.S.Saidov, V.A.Smirnov, I.A.Scherbakov: Kvantovaya electronika!!. 487 (l9a4) E.V. Zharikov, V.V. Laptev, V.G. Ostroumov, Yu.S.Prisvis, V.A. Smirnov, I.A.Scherbakov: Kvantovaya electronika:!!. 1567 (1984) E. V .Zharikov, S.P .Kalitin, V. V .Laptev, V. V .Osika, A.M.Prokhorov, V .A.Smirnov, I.A.Shcherbakov: Doklady Akademii Nauk (to be published) B. Fritz, E.Menke: Solid State Comm. 3, 61 (1965) L.F.Mollenbauer, D.H.Olson: Appl. Phys.Lett. 24, 3a6 (1974) R.Beigang, G.Litfin, H.Welling: Appi.Phys.LetL31, 3al (1977) Yu.L.Gusev, S.N.Konoplin, S.I.Marennikov: Kvantovaya electronika !, 2024 (1977) Yu.L.Gusev, S.I.Marennikov, V.P.Chebotaev: Soviet JETP Letters t No.7, 305 (1977) H.Welling, D. Frlilich: "Pro ress in Tunable Lasers" - in Festkorperprobleme, VXIX, pp.403-425, J.Treusch ed., Vieweg, Braunschweig, 1979 V.A.Arkhangelskaya, P.P.Feofilov: Kvantovaya electronika ?_,No.6, 1141 (19aO) L.F .Mollenbauer: "Progress in color center laser" - in Defects in insulating crystals, Proceedings of the International Conference, Riga, Phyllics Institute, 524-541, May la-23, 19al A.P.Shkadarevich, A.R.Yarmolkovich: Novye lasersnye sredy na centrach okraski v slozhnykh phtoridach, Preprint of IPhAN of the Byelorussian SSR, No.361, 24, 19a5 B. Henderson: Optics Lett.~ No.9, 437-439 (19al) S.A.Mikhnov, A.N.Voytovich, V.A.Kononov, V.N.Uskov, V.E. Greenkovich: Generatsiya perestraivaemogo po chastote izluchenya na crystalle saphira s centrami okraski. Preprint of IPhAN of the Byelorussian SSR, No.367, a, 19a5 W.Gellermann, F.Luty, C.R.Pollock: Optics Comm. 39, No.6, 391-395 (1981) W.Gellermann, K.P.Koch, F.Luty: Laser Focus, April,71-75 (1982) Yu.L.Gusev, A.V.Kirpichnikov, S.N.Konoplin, S.I.Marenninkov: Kvantovaya electronika a, No.6, 1376-1377 (1981) T.T.Basiev, S.B.Mirov, A.M. Prokhorov: Doclay AN SSSR 246, 72 (1979) T.T.Basiev, Yu.K.Voronko, S.B.Mirov, V.V.Osiko, A.M. Prokhorov: Soviet JETPh Letters~ 661 (1979) T. T .Basiev, Yu.K. Voronko, E.O.Kirpichnikova, S.B.Mirov, V. V .Osika: Kratkie soobscheniya po Physike, No.3, 3-9, 19a2 T.T.Badiev, F.A.Vakhidov, Yu.K.Voronko, S.B.Mirov, V.V.Osiko: Kvantovaya electronika, v.l3 (l9a5) T. T .Basiev, Yu.K. Voronko, E.O.Kirpichnikova, S.B.Mirov, V. V .Osiko, M.S. Soskin, V .B. Taranenko: Kvantovaya electronika !!_. No.2, 419-421 (1981) T.T.Basiev, N.S. Vorob'ev, S.B.Mirov, V.V.Osiko, P.P.Pashinin, E.V.Postavalov, A.M. Prokhorov: Soviet JETP Letters 31, No.5, 316-320 (19aO)
9
59 T. T .Basiev, V .D.Lokhygin, S.B.Mirov, G.I.Onischukov, A.A.F omichev: In collection "Prestraivaemya po chastote lasery", ITP So, Akademia Nauk SSSR, Ed. by V.P. Chebotaev, 399-402, 1984
60 T. T .Basiev, F. V .Karpushko, S.B.Mirov, V .P .Morozov, N.A.Saskovich, G.V.Sinitsin, V.B.Taranenko, A.P.Shkadarevich: In -collection "Prestraivaemya ~o chastote lasery''. ITP So, Akademia Nauk SSSR, Ed. V.P. Chebotaev, BJ-90, 984
61 F.V.Karpushko, N.A.Saekevich, G.V.Sinitsyn: Laser na crystalle LiF-Fz s lampovoy, Preprint IPhAN BSSR, 1985
62 T. T .Basiev, Yu.K. Voronko, S.B.Mirov, V. V .Osjko, A.M. Prokhorov, M.S.Soskin, V.B.Taranenko: Kvantovaya electronika ~No.8, 1771-174~. (1982)
63 T.T.Basiev, V.S.Burakov, D.V.Kovalev, S.B.Mirov, A.N.Morozov, A.M. Prokhorov, G.V.Sipinitsin, A.P.Shkadarenvich: Kvantovaya electronika !Q, 1919-1922 (1983)
64 F.V.Karpushko, V.P.Morozov, C.V.Sinitsyn: Proceedings of the third International Conference of Infrared Physics ETH Zurich, ed. Rnegsegger W. and Kneubuhl F.K., 427, 1984
65 T.T.Basiev, B.V.Ershev, S.B.Kravchenko, S.B. Mirov, V.A.Spiridonov, V.B.Fredov: Kvantovaya electronika 12, No.6 (1985)
66 T.T.Basiev, B.I.Denker, N.N.Il'ichev, A.A.Malyutin, S.V.Mirov, v;v.Osiko, P.P.Pashinin: Kvantovaya electronika ~.No.8, 1536-1542 (1982)
67 T.T.Basiev, S.A.Boldyrev, B.I.Denker, N.N.Il'ichev, G.S.Lenov, A.A.Malyutin S.V.Mirov, P.P.Pashinin: Kvantovaya electronika !!, No.8, 1671-1674 (1984)
68 T. T .Basiev, Yu.K. Voronko, B. V .Ershov, S.B.Kravtsov, S.B.Mirov, V. V .Osika, A.M. Prokhorov, V.A.Spiridonov, V.B.Feodorov: Kratkie soobscheniya po physike, No.2, 36-39 (1984)
69 Yu.L.Gusev, S.N.Konoplin, A.V.Kirpichnikov, S.I.Marenninkov: In "Laseri s prestraivaemoy chastoty", Novosibirsk, ITP SO AN SSSR, 116-118, 1980
70 T.T.Basiev, Yu.K.Voronko, S.B.Mirov, V.V.Osiko, A.M. Prokhorov: Kvantovaya electronika ~ No.4, 837-839 (1982)
7l N.P.Erminova, N.I.lvanov, A.A.Mikhalenko, V.I.Hulugurov, V.A.Chupurnoy: Proceedings of the VI National Conference of young scientists, Research in the field of spectroscopy and quantum electronics, Vilnyus, State University, 114 (1983)
Recent Advances in Ultraviolet Laser Technology JF. FIGlEIRA, S.R. FU. TYN, L. ROSOCHA, R.C. SZE, C.R. TALLMAN, J. T£LLE, D.E. WATKINS
Los Alamos National Laboratory, Los Alamos, New Mexico, U.S.A.
I . INTRODUCTION
The successful development of reliable UV lasers based on electrical pumping of rare gas halogen mixtures has lead to an increasing array of commercial and research applications that has served to further speed the development of these devices. Commercial applications which range from semiconductor processing to arterial reconstruction have driven the development of more efficiel')t, more reliable, more versatile and eventually less costly laser devices. Research applications in atomic physics and plasma physics have led to interests in aperture scaling and short pulse operation of excimer lasers. At Los Alamos, development programs address high power discharge lasers, high energy E­ beam lasers and efficient methods of achieving tunability of the laser output.
II. DISCHARGE LASERS
Population inversi.on in an excimer discharge laser is achieved by separately initiating and maintaining a Townsend avalanche in a high pressure qas mixture. Because of the electro-negativity of the gas~s used, recombination of the initial preionization occurs very quickly and the entire discharge can collapse into a destructive arc if no precautions are taken. Advances in high speed pulse power and novel redesigns of the discharge electrode structures have been used to produce stabilized discharges at pulse repetition rates approaching 1000 Hz.
A series of laser systems have been designed and built to explore the laser science and technology required to generate high average power from XeCl lasers. These devices have specifically addressed the areas of magnetic pulse compression, prepulsing techniques, x-ray preionization, advanced thyratrons gas cleanup techniques, inert construction materials and optically uniform gas flow. Prototype devices have achieved powers approaching 1KW at repetition rates of 1 KHz. Optical quality of the discharge medium is well controlled with RMS wavefront errors ot less than 0.07 waves at 308 nm. These initial prototype d•vices have led to the development of advanced designs where compactness and reliability have been emphasized. Figure 1 shows a view of such a system currently under development. Design goals of this system are 250 W of average power operating at 308 nm.
12
The above devices typically operate with conventional pulse durations in the 20 50 ns range. A second type of discharge device has been developed that extends the output pulse duration into the 200 ns range while achieving a compaction of the physical design. In these devices the
usual solid cathode of the discharge volume is replaced by an inductively stabilized electrode. The electrode is composed of a segmented array of conductors each of which is connected to the main ground return by a separate inductor. If the discharge starts to become unstable, the current rises through one portion of the electrode, and the induced voltage drop in the ballasting inductor reduces the current thereby stabilizing the discharge. These devices have proven to be very robust in their operation. Because of the inherent stability against arc formation these lasers can operate at longer pulse durations and at reduced flow requirements in high pulse repetition rate applications. Figure 2 shows a cross sectional view of an inductively stabilized XeCl laser, referred to as the STABLEX design. Prototype devices have produced 165 mJjpulse and 1000 Hz operation in separate devices. The long pulse output of these devices makes the use of modelocking a practical means of producing subnanosecond pulses of 308 nm radiation. Actively modelocked cavities have produced pulses in the 300 - 500 ps range that approach the transient limit of 160 ps. Tests are currently in progress to demonstrate 200 W of average power from an extended cavity version of the STABLEX design and to extend the modelocking studies by the use of passive modulators to the sub-100 ps regime.
III. ELECTRON-BEAM PUMPED LASERS
In addition to the small apertur~, high average power devices described above, work is 1n progress to explore aperture scaling of large excimer lasers that are pumped by direct electron beam excitation. The Aurora laser is a large aperture long-pulse KrF laser specifically designed for the inertial confinement fusion program. It will be optically multiplexed to produce a single pulse 5 to 10 ns in duration for target irradiation experi~ents. The Aurora amplifier fe~tur~s a pumped volume of 2 m with an output aperture of 10 em . Electrical power is provided by four pulse-charged water lines that drive opposed electron guns to achieve a uniform gain distribution in the lasing media. Figure 3 shows a view of the Large Aperture Module of the Aurora laser system. Two of the water storage lines are shown in the foreground attached to the electron gun and laser pumping chamber. The 10.8 m long water lines each store 150 KJ of energy and deliver approximately 31% of this energy to the laser gas at a cathode voltage of 675 KV. The carbon felt electron gun produces 25 A/cm2 into a lOOcm x 200cm emitting area at an impedance of 2.7 ohm. Preliminary testing of the device in an unstable resonator configuration has been completed and output energies in excess of 10 KJ have been obtained in pulse durations of 500 ns at an electrical pumping rate of 133 KW/cm3 to the laser gas.
13
IV. WAVELENGTH CONVERSION
Although efficient excimer lasers have been developed to cover the wavelength range from 190 nm (ArF) to 350 nm (XeF) some applications require wavelength tunability throughout this range. The use of organic dyes pumped by excimer lasers have provided complete tunability while the use of stimulated Raman scattering has provided efficient stepwise tunability ~n the near ultraviolet.
* Laser-pumped dye lasers
The commercial availability of reliable, high average power excimer lasers has led to the development of high repetition rate dye lasers operating in the ultraviolet. In order to achieve large pulse energies at high PRFs, master oscillator/power amplifier configurations have been used. Master oscillators have been designed and tested that produce 375 mW of power in a 3X diffraction limited beam operating at 250 Hz. Power amplif~ers have been designed for pumping volumes of 3.5x3.5x70 mm and intrinsic conversion efficiencies in the range of 18% to 22% have been measured for a variety of dyes. For saturated MOPA amplifier chains, system efficiencies in the range of 14% have been achieved for specific operating ranges. Figure 4 shows the saturation characteristics of a such a power amplifier operating in the near uv.
* Raman conversion
Although the dye lasers possess inhere~t tunability they suffer from being complex devices with limited efficiency. An alternate method of obtaining UV radiation is through the use of stimulated Raman scattering. This technique provides only stepwise tunability but has the potential advantage of increased conversion efficiency. Experiments with single focus Raman generators have demonstrated intrinsic photon conversion efficiencies of 85% in single shifts in methane (CH4) driven with XeF discharge lasers. Figure 5 shows the measured photon conversion efficiency as a function of input pump energy. First Stokes generation (open circles) reaches 85% before the first indication of second Stokes generation (closed circles) is experimentally observed. These high efficiencies are due in part to the increased suppression of second Stokes generation that have been observed. The large f/No. geometries used in these experiments tend to suppress the four wave mixing contribution to Stokes generation thereby increasing the second Stokes threshold.
V. LASER INDUCED DAMAGE
The ability of optical components to survive and maintain their performance in the presence of high radiation fields is of important fundamental and practical concern to the designers and builders of laser devices. The current programs at Los Alamos are directed toward quantifying and
14
understanding the damage processes that occur for the ultraviolet wavelength range. Of particular concern is the variation of damage thresholds with wavelength and optical pulse duration. Figure 6 shows a compilation of experi­ mental results for several different materials for three different laser wavelengths, 248nm, 308nm and 35lnm. Laser pulse duration was approximately 15 ns for all these tests and the damage threshold is defined as the fluence producing a zero probability of damage independent of laser spot size. For the data shown the d~age threshold appears to scale at least as (wavelength) over the wavelength range investigated with an upper limit of (wavelength) 4 . The specific trends are also seen to be material dependent, e.g. the two different polishing techniques on Suprasil.
A second area of interest is the scaling of optical damage with pulse duration. At present no complete, self-consistent data is available that covers the entire range of interest. However by comparing the existing data we can obtain an indication of the trends to be expected. In figure 7 we have plotted data from three different experiments as noted, where the absolute damage thresholds have been normalized at the pulse duration boundaries for each experiment. The data of Newnam and Gill and Rainer were taken with tripled Nd:YAG at 353nm, whereas the rest of the data was obtained with a discharge XeF laser (Foltyn) and a frequency doubled dye laser (Foltyn and Marrs) at 351 nm. For short duration pulses below 1 ns we see that damage is no longer dependent on the fluence of the laser radiation but rather becomes linearly dependent on the laser intensity reminiscent of an electron avalanche initiation process. At the longer pulse durations, up to 450 ns, the scaling of the damage threshold becomes a very weak function of the pulse durat~og, scaling as (t ) 0 •2. This is much slower than the (t ) · expected for '/' completely thermal process and at p~esent is not understood.
VI. CONCLUSIONS
The science and technology of ultraviolet lasers using electrical pumping of excimer gases has developed rapidly in recent years. The results achieved to date have been quite impressive for the short time involved. Average powers in excess of 1 KW have been demonstrated and commercial devices exist with output powers approaching 300 watts. Single pulse energies in excess of 10 KJ have been demonstrated in laboratory devices and efforts are in progress to scale these single pulse energies into the 100 kJ range.
Future developments in laser devices will emphasize aperture scaling of discharge and e-beam pumped devices, approaches to cw operation, improved efficiency and lifetime and the development of very high brightness laser sources.
Methods of wavelength conversion will continue to be refined in the ultraviolet. Improved uv dyes, with enhanced lifetime and efficiency will be developed, increased emphasis will be placed on the development of efficient and scalable Raman converters, and the potential of optically pumped x-ray lasers will be explored.
Fig.L
Fig.2
Fig.3
Fig.4
Fig.5
Fig.6
Fig.7
15
FIGURE CAPTIONS
High average power prototype XeCl laser. Optical axis, gas flow and discharge current are mutually orthogonal. Discharge length is 60 em.
High average power Stablex laser. Laser con~ists of four separate discharges, each lx1x40 em , controlled by a thyratron switch. System includes axial fan and heat exchanger. Scale is noted in figure.
View of the Large Aperture Module of the Aurora laser fusion facility. Water lines are 91 em in diameter and the laser volume (loc~ted between the Helmholtz coils) is 100x100x200 em •
Amplifier saturation curve for XeCl laser pumped dye laser amplifier using solution of TBS in cyclohexane. For this double sided pumped amplifier, 50mJjpulse was extracted for a pump energy of 200mJjside in a 22ns pulse.
Measured Raman conversion efficiency for first Stokes (open circles) and second stokes (closed circles) for 110 atm cell of CH4 pumped by a 20 ns pulse of 351nm radiation from a XeF laser. Threshold for second stokes is nearly x6 that for the first Stokes emission.
Wavelength scaling of laser induced damage for several uv materials and coating combinat.ions for KrF, XeCl and XeF lasers. Solid curves are contours for (wavelength) 2 scaling of the damage threshold. Data taken with approximately 15 ns pulses.
Normalized pulse duration scaling of damage threshold. Open triangles are data from B.E.Newnam and D.H.Gill, "Ultraviolet damage resistance of laser coatings", NBS Special Publication 541, p.190, 1978. Closed squares are taken from Foltyn and Marrs(unpublished). Closed circles are from F.Rainer, C.L.Vercimak, D.Milam, C.K.Carniglia, T.Tuttle Hart, " Measurements of the dependence of damage thresholds on laser wavelength, pulse du€Rtion and film thickness", presented at the 15 Annual Symposium on Optical Materials for High Power Lasers, Boulder, co, Nov. 14-16, 1983, in press.
16
DISCHARGE VOLUME----~~~--~--
FLOW SYSTEM
60~--~----~--~----~----~--~----~
-J 50 E ->- (!) 40 a:: w z w 30 0 w ~ (.) 20 <t a:: ~
~ 10
PUMP ENERGY= 400 mJ
VOSCILLATOR I PREAMP OUTPUT
DYE INPUT FLUENCE (mJ/cm2) Fig .4
70
20
100 I I I I
>- 0 u z 80 - - w
Sl u 0 LL. 60 - XeF BANDWIDTH: 160 cm- 1
- LL. w CONVERTER PRESSURE : 110 atm.
z BEAM DIAMETER : 4 mm 0 40 r 0
-.,_ 0 I a..
0 20 !"""' - en
01 52 0 I I I • 0 5 10 15 20 25
PUMP ENERGY (mJ)
R E
F LE
C T
O R
LONG-RANGE PULSELENGTH SCALING FOR REFLECTORS AT 350 nm
5 I I I I Q _J 6 NEWNAM & GILL (Zr02/Si02) 0 • FOL TYN & MARRS (AI203/Si02) I 4 t- e RAINER, et al. .-en w (average of several materials)
Q a: I w ..... 3 N t- -
_J w <(!) ~ < !::. a: ~ 2 t- • -o< zO ••
w {.) z 1 t- -w • ::> ~ _J
u.. 0 I I l l 0.01 0.1 1 10 100 1000
PULSELENGTH (ns)
N.G. BASOV, P.G. KRYU<OV, N.N. YURYSI-EV
Institute of Physics, Moscow, USSR
A chemical oxygen-iodine laser (COIL) generates the emission of radiation at
A = 1.315 f.Lm from the atomic iodine transition 2P112_2P312• The upper laser level is
pllmped as a result of a fast quasi-resonant energy transfer from the oxygen molecules
in the state o2<1 ~ ) to the iodine atom:
1 Z Kl 3 2 02 ( A ) +I( p3/2 --• 02( l:) +I( p1/2), (1)
-11 3 -1 k1 = 7.7xl0 em s
The iodine atoms are formed in the mixing of singlet oxygen with molecular iodine
during the process
(2)
Efficient cw operation of a chemical oxygen-iodine laser has been achieved
with an output power of the order of 2 kW /1/.
Such a laser appears t? have considerable promise when used in a pulse­
repetition (PR) regime with high energy characteristics. So, there arises ·the problem
of increasing the peak power of a COIL with high average power and energy yield~ In
this paper we consider different solutions to this problem. Note that all our estimates
have been made for a laser operating at a temperature T = 300 K.
One of the methods by which PR operation is achieved is quite obvious for gas­
flow lasers. In our case, for a COIL, it involves preliminary mixing of a flow containing
singlet oxygen with molecular iodine, during continuous pumping of the active medium
through the laser cell. When the cell is filled, the optical feedback is switched (in the
case of an oscillator), or the input signal is applied (in the case of an amplifier).
Another process can occur as well as the processes (1) and (2) in the system
o2<1 A) + 12 - the relaxation of the electronic excitation o2C1 A ):
(3)
This process limits the characteristic dimension of the active medium, along the flow,
to tl.e value L = v • T rel' where v is the flow velocity and T rel = 1/k3
24 2
[I( P l/2)] "'1/k3 [I ] is the characteristic time of process (3). The energy E stored in
this active medium is
(4)
where o2 <1 6) is the concentration of the o2<1 A) molecules in mol.cm-3, 5 is the
area of the flow· and h v L is the energy of the laser quantum. The release time of the
stored energy is determined by the time of excitation transfer from o2 (1A ) to I
[according to (1)], i.e., 1/k1 [1]. This value determines either the laser pulse duration or
the necessary input pulse duration in the amplification regime. Thus, the pulse power
achievable in this case is
(5)
By t!lking into account that h v L [oz<1 A)] Sv is the power limit for cw operation, we
find that the maximum gain in power is, m this case, krfk3, or 7.7xl02• Actually, this
gain is less, because L is taken to be less than v rei in order to provide sufficient
homogeneity of the active medium.
So, according to this method, the maximum pulsed power is connected
unambiguously with the power produced in continuous operation, and there is no
possibility of increasing the pulsed power for a given average power. This is connected
with the fact that the concentration of the atomic iodine determines both the pulse
energy E and the lasing pulse duration, and in general does not affect the value of the
peak power. It would be desirable to eliminate the dependence of either E or Tlas on
the iodine concentration. Then it would be possible to increase the pulsed power by
varying the iodine concentration. This can hardly be done with respect to las' but
appears quite possible with respect to E. Thus, we approach the second method of
realization of pulsed operation in a COIL.
The point of the method is that the laser volume should be filled with singlet
oxygen in the absence of iodine. After the filling, it is necessary to provide, rather
quickly, a uniform concentration of atomic iodine. This can be done by adding the
singlet oxygen with iodine-containing substances which do not react with the singlet
oxygen and do not deactivate it. The mixture so produced yields, after it is acted upon
in some way (viz., photolysis, decomposition by an e-beam or by other types of
radiation), the atomic iodine.
At the stage when the active volume is filling, the excitation relaxation is
determined by the following process K
02 (1 A) + 02 (1 6 ) --1. 02 (1l: ) + 02 (3 l: ), (6)
-17 3 -1 k4 = 2x10 em s
2.5
The time in which the concentration of o2(1 1l ) decreases by e in (6) is
(2) 1 T rel = l. 7 lk4 [02 ( ll )] in it (7)
In the first method, the time taken for the active volume to fill is determined by the
atomic iodine concentration, and is T r~~) = 1lk~ [ I]. But when the COIL is
characterized by the relation [I] I [02(1 1l)] "' 10- , the second method gives the
greater filling time by an amount l. 7 k3 [I] lk4 x [02(1 /l )] "' 85 times. Thus, the
volume of the active medium in this case, and hence the stored energy, can be
increased by almost 85 times as compared to the first method. Note that the optimal
relation [I] I [02] in the PR regime can be greater than in continuous operation; hence
the gain of the hybrid version would also be higher. Since the pulse duration in this
case is 11[1] kl' the maximum power would also increase by a factor of 85. The atomic
iodine concentration can be changed for a fixed active volume. This means that the
pulse duration can also be changed at the given energy, that is, the pulsed power can
be varied within a wide range for a fixed average power.
The advantage of this method is the possibility it presents of increasing the
pressure of the active medium. The necessary condition for the mixing laser, i.e., that
the mixing time should be less than the relaxation time, can be provided for higher
presaures. With it, some other methods for producing the singlet oxygen in a COIL can
be used, for example, ozone photolysis which needs an increased pressure to operate
121. lkllik~ the first method, the second method needs additional sources of energy
for the decomposition of the iodine-containing substances. Let us evalupte the ratio
las of the energy spent in the preparation of the atomic iodine to the laser energy.
The specific .energy E that can be extracted from the active medium is
[02(1 fl )] h v L n , where n is the extraction efficiency coefficient of the singlet
oxygen. The "expenses" for the production of atomic iodine, in the case of iodide
photodiuociation by uv-radiation, are [I] h VfDI n FD' where h v FD is the uv quantum
from a aource of efficiency Fo· Let us take typical values realized in a cw COIL
11,21 and in the photodissociation iodine laser with flash-lamp pumping 131, i.e.,
[IJI[02(1 1l)] = 10-2, n = 0.4, h v Folhv L = 5.2, and n FD = 0.1. Then, for the value
we are interested in, we obtain nlas !!_80%. The efficiency can be increased with the
help of more efficient photodissociation sources. In particular, quite attractive are the
so-called excimer KrF ( ). = 249 nm) and Xel ( ). = 253 nm) lamps whose radiation
wavelengths are contained in the dissociation band of a great number of iodine­
containing substances; their efficiency can reach 30% 141. The idea of using a COIL as a driver (i.e., a high-power laser which heats up
the target) for the initiation of thermonuclear fusion looks very promising 151. This
26
follows from the ability of a COIL to operate with high efficiency in a closed cycle.
The es~mates made in /5/ show that when the conversion efficiency of o2 excitation
energy into laser energy is 30%, the full efficiency of a laser system, allowing for the
er;utrgy required for reagent reproduction, would achieve 6%, even with today's level of
technology. Moreover, lasing in a COIL occurs at the same transition of an iodine atom
as in the iodine photodissociation lasers. It is, therefore, quite possible to construct a
hybrid system including a photodissociation laser; this has been developed and
advanced in a number of laboratories /6-8/. To make the COIL usable in controlled
fusion research one should not only achieve the PR regime, but obtain quite a short
(about 1 ns) pulse with an energy up to a hundred kilojoules. Let us see how it can be
done.
Since the pulse duration is shorter than the characteristic transfer time of the
reaction 0 2 (1 t.) + I, it is quite evident that an amplified pulse can take, in one pass,
only part of the energy, about equal to the ratio [1]/ [02], from the active medium.
To obtain the total energy yield one must repeat the amplification process many
times. This can be realized in a circular scheme (Fig.1) over a perimeter c T 1 where c
is the light velocity and T 1 the characteristic time of transfer T 1 "' 1/k1 [02(1 t. )].
The required number of passes is of order of magnitude N = [02] /[I]. It is evident that
the scheme must include polarization elements and optical shutters in order to lead
the radiation out of the circle and to preveht self-excitation of the system.
To lead the radiation out of the circle one can use second-harmonic conversion
of noncollinear beams (Fig.1). The beams of two circular schemes intersect in a
nonlinear crystal at an angle exactly equal to the angle of noncollinear synchronism.
The exact lengths of perimeters are chosen so that the pulses propagating in the first
and second rings coincide in the nonlinear crystal only after a certain number of
passages. At this coincidence second-harmonic conversion takes place, producing a
pulse which radiates along the bisector of the intersection angle.
Under certain conditions coherent amplification is quite achievable in oxygen­
iodine lasers operating at low pressure, less than 100 Torr, with a pulse duration of
"'1 ns, for which the condition T pulse< 1/ t.w is valid, where t.w is the radiation
linewidth of the iodine atom. With each pass, the energy yielded would, in this case, be
twice as high. This would allow one to reduce the number of passes.
To achieve these proposals one should obtain, first of all, an active medium
with the required stored energy. In /5/ it is suggested that an active medium should be
used with dimensions 1.7 m x 3m with 10 Torr of o2[ o2(1 t.) = 80%] and 0.3 Torr of
12• Even when the flow has the velocity of sound, 300 ms-1, one can, perhaps, obtain an
active medium of length 18 em, essentially shorter than required.
However, when the second method described above is used, the admissible
time is increased 500 times for a given 0 2 pressure, and the preparation of an active
medium of the dimensions needed becomes quite feasible.
27 REFERENCES
2 A.I.Didiukov, Yu.I.Krasnoshchekov, Yu.A.Kulagin, V.A.Morozov, S.A.Reshetnyak, L.A.Shelepin: Kvantovaya Elektron. 9, 731 (1982)
3 R.Gross, J.Bott, Eds.: Handbook of Olemicallalers,-J.Wiley & Sons (N.Y.) 4 G.H.Miley, M.Zediker, H.Elsayed-Ali, D.Shannor: Paper TuM-16, CLE0-83,
May 17-20, 1983 5 G.N.Hays, G.A.Fisk: IEEE J.Quant.Electron. QE-17, 1823-1826 (1981) 6 N.G.Basov, V .S.Zuev, V .A.Katulin, A. Yu.Liubchenko, V. Yu.Nosach, A.P .Petrov:
Kvantovaya Elektron. §., 311 (1979) 7 A.I.Zaretsky, G.A.Kirilov, S.B.Kormer, G.G.Kochemasov, V.M.Murugov,
S.A.Sukharev: Kvantovaya Elektron. !Q, 756 (1983) 8 G.Brederlow, E.Fill, V.Fuss, K.Hohla, R.Folk, K.Witte: Kvantovaya Elektron.
~ 906 (1976)
URI P. OPPENHEIM
Department of Physics Technion -Israel Institute of Technology Haifa 3.2000, Israel
Beat frequencies between the transverse modes of an unstabilized cw co2 laser were
studied by detecting the homodyne signals emitted by the laser. Frequency pulling was
observed and a resulting mode spacing reduction of up to 13% was measured. The
dependence of mode pulling on total gas pressure was measured and was found to be in
fair agreement with theory. The observed effect is important in some laser tuning
techniques where it leads to a reduction of the tuning range and a nonlinear
dependence of the laser frequency on cavity length.
1. INTRODUCTION
When the output of a cw co2 laser was measured with a fast detector a strong rf
signal was observed~ It was concluded that the cw laser operated in a multimode
regime and that the signal was caused by interference (homodyne effect) between the
modes. The exceptional stability of the signal was not affected by variations of cavity
length. This could be explained by the fact that, as the length of the cavity d changed
by t.d, the frequency spacing between the modes changed by t.d/d - 10-5 and
therefore remained effectively constant. However, neither the observed frequencies of
the homodyne signals nor the spatial mode patterns were consistent with simple
theory.
A study of the observed homodyne beats in relation to the cavity resonances
was carried out to clarify this problem.
2. MODE THEORY
The resonant frequencies of a optical cavity consisting of two spherical mirrors with
radii of curvature R1 and R2 and separated by a distance d are /1/
v = fa {q + ( 1 +m+n) } , (1)
(2)
30
The transverse mode numbers m and n can take values 0,1,2, ... ; the longitudinal mode
number q is equal to the number of half-wavelengths contained in the distance d. For
stable resonators, the factor f can vary between 0 and 1/2.
The spatial distribution of modes is determined by the set of orthogonal
functions
where Hi are the Hermite polynomials. These functions obey the orthogonal relations
f E (x,y) E (x·;y) dxdy = o (m-p) o (n-q). _.., mn pq
(3)
The orthogonality of these eigenfunctions implies that the net signal resulting from
superposition of any two modes will be zero.
If the laser oscillates in more than one mode, a beat signal will appear equal to
the frequency difference between the modes, that is,
(4)
The orthogonality of the eigenfunctions in Eq. (3) means that in order to
observe these homodyne beats it is necessary either to block off part of the spot size.
In this study it was our purpose to investigate the discrepancy between the
observed homodyne frequencies and those calculated from eqs. (2) and (4).
The intensity of the homodyne signal generated by two oscillating modes
having frequencies '1 and v2 is given by the expression /2/
I - ~ cos(211(v 1-v 2 )),
where Ii stands for the intensity of the ith mode. It is seen that the intensities of the
modes enter symmetrically in this expression. This means that even a weakly
oscillating mode will give rise to a significant homodyne signal when it beats against a
strong mode. The homodyne technique, therefore, is very sensitive to weak modes. On
the other hand, visual observation of the spatial mode pattern can detect only the
strongest modes. Thus, the homodyne signal appears even when the spatial mode
pattern does not indicate that a certain (weak) modes is oscillating.
3. MODE PULLING THEORY
Mode pulling theory effects arise from the anomalous dispersion of the refractive
index in the lasing medium. We shall use the description of this effect given by
Siegman /3/. In an active cavity each resonant frequency is shifted in the direction of
the line center as a result of mode pulling. For a predominantly collision broadened
31
transition this displacement is directly proportional to the initial distance of the
resonant frequency from the line center. The proportionality factor is (1-Qm/Qc),
where
(5)
is the molecular Q-factor (equal to the line frequency divided by the linewidth). The
cavity Q-factor Qc is given by the expression
2~n d{ 1 }-1 Q = ~ 2a 0 d + ln(--) , c h r 1r 2
(6)
where ri and r~ are the intensity reflectivities of the two mirrors, and a 0 describes
the absorption losses.
As a result of mode pulling, the frequency spacing between adjacent modes,
although independent of the frequency itself, is reduced by the factor
(7)
It follows that the ratio of mode spacing in an active cavity, 6., to that in a passive
one, A 0 , is given by
A Qm A= 1 - ~·
(8)
The observation of beats between adjacent modes is thus a simple and
sensitive way to measure the mode pulling factor. Other methods involve an absolute
frequency measurement, usually obtained by heterodyning the laser under study
against a highly s~abilized standard laser operating at line center /4/. The present
method has the advantage of requiring one stabilized laser only.
4. EXPERIMENTAL SETUP
Figure 1 shows the experimental setup for homodyne measurements. A HgCdTe
detector (made by Santa Barbara Research Center) was used, having a sensitive area of
4 mm2• Mirror M alloed translation of the beam spot, which had an area of 1 cm2,
across the detector. A powermeter was used to monitor the power output for different
alignments of the cavity mirrors. The plasma tube was 1.5 m long and had NaCl
Brewster windows at each end. The cavity mirror M1 was a concave front surface
mirror with a radius of curvature of 2.30 m and a reflectivity of over 99 percent. The
other cavity mirror, M2, was a flat Ge mirror with a reflectivity of 65%. This mirror
served as the output coupling mirror. The cavity length was 1. 73 m.
According to eqs. (2) and (4), the longitudinal mode spacing of this cavity was
32
86.7 lVI Hz and the transverse mode spacing [ /::, (m+n)=1 J was 28.5 lVI Hz. It was assumed
that the laser oscillated at 10.6 jlm.
The intensity of the radiation falling on the detector was- 0.1 W /cm2 which
was below the damage threshold of the detector. Since only the frequency and not the
amplitude of the signals was of interest in this experiment, saturation of the detector,
if it occurred, was not important. While the manufacturer of the HgCdTe detector
gave its bandwidth as 10 MHz, the level of radiation was so high that it allowed
observation of signals of up to 35 MHz on a spectrum analyzer connected directly to the detector. However, when it was attempted to observe the homodyne signals on an
oscilloscope it was necessary to use a 5-500-MHz amplifier. The accuracy of the frequency measurement (on the spectrum analyzer) in the relevant range was 0.05 MHz with a resolution of 0.1 MHz.
The spatial mode patterns were observed by burning an image on white paper.
5. EXPERIMENTAL RESULTS
As mentioned above, in the first experiments neither the homodyne frequencies nor
the spatial pattern of the modes conformed to theory. A large number of frequencies
appeared on the spectrum analyzer in a seemingly random fashion and these could not
be reconciled with eq. (4). At the same time, the spatial mode patterns were distorted and irregular.
It was assumed that small irregularities in the optical components of the laser
cavity were the cause of this multitude of frequencies and also gave rise to the
distortion of the spatial mode patterns. replacing the Brewster windows by new ones
removed most of the multiple frequencies and spatial distortions. It was noted that to
the naked eye the old windows revealed no defects.
Precise alignment of the cavity mirrors, using a diaphragm of 8-mm diam at
the flat mirror, finally produced a single beat signal. This precise alignment and the
diaphragm were essential to obtaining a single frequency signal, since otherwise
satellite frequencies appeared in addition to the central one, s;imilar to those described
by Goldsborough /5/ for a He-Ne laser. The frequency of the homodyne peak was 26.2 MHz (at 25 Torr) compared with 28.5 MHz calculated from eqs. (2) and (4). This
discrepancy was ascribed to mode pulling.
To study the dependence of mode pulling on Qm (eq.(5)) we varied the total
pressure in the plasma tube, thereby altering the linewidth. The current was kept
constant at 30 rnA. According to eq. (7), the factor Qm/Glc depends inversely on
linewidth and hence decreases with increasing pressure. Therefore, the frequency interval between consecutive modes grows as the pressure increases, approach the cold
cavity value given by eqs. (2) and (4).
For this experiment, the mirrors were aligned in such a way that the TEM00
33
and TEM10 modes oscillated. The passive frequency spacing between these modes,
calculated from eqs. (2) and (4), was 28.5 MHz.
Table 1 shows the measured frequency of the homodyne signal as a function of
pressure. Since the frequency spacing between adjacent modes is known for the pa88ive
cavity from eq. (4) and for the active cavity from experiment, it is possible to
calculate Qm/Qc from eq. (8) and then to proceed to calculate the linewidth A-> from
eq. (7). In these calculations we used the values d = 150 em, n0 = 1, a0 = 0 and
r1 r 2 =JO:l;'$. The results of the calculations appear in Table 1 in the third col1,1mn.
Above 35 Torr the linewidth is seen to depend linearly on pressure, with a slope of 4.2
MHz/Torr. This is in good agreement with the calculated linewidth obtained for
collision broadening.
We may compare these- results with a theoretical calculation of collision and
Doppler linewidths. In the high-pre88ure region the line is predominantly collision
broadened and the linewidth (FWHM) may be calculated from /4/
(9)
where fx is the fraction of the gas x, P is the total pressure in torr, and T is the
absolute temperature of the gas (350 K). For the mixture used in this study
(co2:N2:He = 1:1:4), eq. (9) gives a slope of- 5 MHz/Torr. In the low-pressure region
(below 10 Torr) the line is predominantly Doppler broadened and has a calculated width
(FWHM) of 57 MHz. The conditions in this experiment were such that plasma
broadening could be neglected.
It may be shown from the data in Table 1 that the asymptote to the high
pressure (35-60 Torr) points passed through the origin, in agreement with pressure
broadening theory. However, at low pressures the points lie above the aymptote,
indicating that Doppler broadening becomes dominant. While qualitatively there is
agreement between theory and experiment, the exact functional behavior of the points
differs considerably. The source of this discrepancy may be twofold. First, the
estimated .value of Qc' based on the known reflectivities of the two mirrors, could be
too low since it does not take into account losses due to diffraction and reflection
from the windows. Second, the expression for mode pulling (eq. (7)), while accurate for
a homogeneously broadened line, is only a first-order approximation in the case of a
line with combined Doppler and pressure broadening.
6. CONCLUSIONS
Mode pulling effects were observed in a single-mode cw co2 laser, resulting in a
reduction of up to 13% in the mode spacing. The shift in the mode frequency that
accompanies reduction of the mode spacing has serious consequences in off-line center
34
tuning experiments in which tuning is accomplished by varying the cavity length. In
such cases mode pulling may lead to significant errors in the calculated value of the
wavelength. Mode pulling is especially prominent in narrow linewidth high-loss lasers.
Another observation of this study is that imperfect Brewster windows give rise
to serious distortions of the mode structure. These distortions manifest themselves in
irregularly shaped mode patterns and, in frequency space, in multiple resonances which
appear at essentially random frequencies.
The help of Mira Naftaly in carrying out the experimental and theoretical
work is gratefully acknowledged.
/1/ G.O.Boyd, H.Kogelnik: Bell Syst. Tech. J. 41, 1)47 (1962)
/2/ W.W.Duley: "C02 Lasers: Effects and Applications", chap. 3.4. 7 (Academic,
New York, 1976)
/3/ A.E.Siegman, "An Introduction to Lasers and Masers'", chap. 10.3 (McGraw­
Hill, New York 1971)
/4/ A. Van Leberghe, S.Avrillier, C.J.Borde: IEEE J. Quantum Electron. QE-14,
481 (1978)
LEGEND FOR TABLE 1
Values of observed homodyne frequency 6. and calculated linewidth (FWHM) 6. v as a
function of total gas pressure in the laser.
LEGEND FOR FIGURE 1
Schematic diagram of experimental arrangement to observe homodyne signals in a
C02 laser.
Total gas Observed homodyne Calculated pressure frequency (MHz) linewidth (Torr) A 6'1) (HHz)
10 2&J.7 77.1
u M
E T
E R
S P
E C
TR U
Central Institute for Physics, 1FT AR, Bucharest, Romania
1. INTRODUCTION
Research and development in the field of solid state lasers, oased upon discovery and use of new laser active media or the relevance of new properties and an improved use of traditional media is, at present, one of the most dynamic directions of activity in the whole field of lasers. The main goals of this activity can be summarized as: (i) extension of wavelength range for efficient direct generation, (ii) improved tunability, (iii) improved efficiency, (iv) increas~ of average power. Tnough these goals have been traditional in solid state lasers R & D, a complex interplay of factors makes possible a qualitatively new approach to the problem. The mai[l factors are: (i) new practical requirements (as, for instance, the need for compact, reliable, efficient lasers generating in a given wavelength range, in a given regime of operation); (ii) theoretical developments and new experimental techniques of investigation for solid state laser active media (such as deactivation processes, energy transfer, high-resolution site-selective or temporally resolved spectroscopy and so on); (iii) new technological possibilities (improved crystal growth technologies of high productivity for production of large, high-quality, reproaucible laser active media); (iv) technical development in the field of lasers (new pumping schemes, conceptual developments ana so on).
In this paper we review some of the main problems of current interest in the field of infrared solid state lasers.
2. INFRARED TUNABLE SOLID STATE LASERS
Tunaole solid state lasers are very useful sources of coherent raaiation for laser spectroscopy, remote sensing, laser driving of chemical reactions and so on. Over other tunable sources, they offer increased efficiency, reliability and compactness; at the same tirn they snow good prospects for extens~on of the wavelength range into the infrared where strong and efficient sources do not exist.
Two classes ·of tunable solid state lasers are descrioed, based on transition­ ion-doped crystals, and on crystals with color centers, respectively; we will concentrate only on the first class. Solid state laser tunability is possible with active media which show broad band luminescence due to transitions terminated on empty vibrational satellites of the electronic grouna state.· Thus such a phonon-terminated laser operates in a highly efficient four-level laser scheme.
Some of.the main requirements imposea upon the active media for tunable solid state lasers are: (i) the emission should be broadly tunable; this implies broaa emission bands ana absence of parasitic reaosorption of emission through ground state reabsorption (GSA) or excited state aosorption (ESA); (ii} the
38
laser should operate at room temperature at a high output level. This implies a hiyh emission cross-section (above lo-20 cm2) ana a high, almost temperature-independent, quantum efficiency; (iii) high total efficiency, which is assured by a high quantum efficiency and broad and strong pump banos situated in a wavelength region where strong ana efficient pump sources exist; (iv) large laser active crystals should be available, of high optical qua.hty and mechanical and optical stability, with no (or very small) non-linear behavior and with a high thermal conductivity.
The features of optical absorption and luminescence of the ions from the transitional groups in crystals are determined oy a very complex interplay of static and dynamic interactions. The optical transitions for laser emission take place usually between energy levels which belong to the ground configuration of the ion. The energy-level diagram which determines the wavelength range for optical absorption and emission, and finally the range of stimulated emission is determined by tne interactions of the electrons from the incomplete shells inside the ion and with the neighbouring ions (the crystal field). The strength of interactions with the crystal f~eld depends on the character of the electrons in the incomplete electronic shell of the ion. Thus, in the case of iron (n = 3), palladium (n = 4) and platinum (n = 5J groups the electrons of the ground configuration ndN are submitted directly to the action of the crystal field which is stronger than the spin-orbit coupling; it can oe smalle:r (intermediate fiela case, 3dN ions) or stronger (strong field case, 4dN and soN ions and some 3d complexes) than the electron-electron interaction. In the case of lanthanides (n = 4) or actinides (n=S) the ions usually have the grouno configuration nfN (n+l)s2(n+l)p6; because of the screening effect of the s and p shells the interaction of the f-electrons with ti1e crystal field is weak (except the case of the sfl configuration in octahedral crystal fields).
Along with the interaction with the static crystal fielo, the interaction of electrons with the vibrations of the crystalline ~attice (the electron-phonon interaction) is very important in determimng the wavelength range and tne dynamical features of the laser emission. Due to the electron-phonon coupling the optical spectra of tl1e transition ions in crystals appear as sharp zero-phonon lines accompanied by resolved progressions of phonon sidebands or broad bands formed by superposition of phonon sioebands. These broad lines are very important for tunable laser generation. The width and relative intensity of this band witl1 respect to tne sharp zero-phonon line aepends on the difference in the strength of the electron-phonon coupling (charactt:!rized by the so-called Huang-Rhys factor S) for the two electronic states connected oy the optical transition. This implies that the two states would have very different equilibrium coordinates in the configurational coordinate diagram. A large S factor implies strong and broad vibronic bands in the optical spectra of tnese centers: in absorption they appear mainly on the high energy side of the zero-phonon line whereas in luminescence they are placed toward the low energy side. Since the zero-phonon lines in absorption and luminescence are resonant, the broad phonon sidebands in luminescence are shifted toward lower energies as compared witn those appearing in absorption by the so-called Stokes shift which depends on S. Large Stokes snifts are desirable in order to avoid ground state reaosorption of emitteo radiation.
The phonon sidebands borrow intensity from the zero-phonon line; thus the peak cross-section of emission for these broaa bands is weak. However, the coupling. of electrons with some vibration mooes which lack inversion symmetry leads to a dynamical m~x~ng witn the ground configuration of some high-lying configurations of opposite parity and tne phonon sideband becomes electric­ dipole-allowed .reven for complexes which w a static equilibrium situation have inversion symmetry. The intensity of these bands increases with temperature.
39 The strength of the electron-phonon coupling depends on the nature of the electrons from the incomplete transition shell. Thus in the case of d.-electrons the S factor is large ( » 1) while in the case of f-electron systems (with the exception of 5fl in octaheoral crystal fields) it is small ( < 1). As a consequence, the optical spectra of the d.-electron ions in crystals show both sharp zero-phonon lines and broad strong phonon sidebands while in the case of f-electron systems the spectra are dominated by sharp zero-phonon lines and the phonon sidebands are usually very weak.
A very important consequence of the electron-phonon interaction is the occurrence of non-radiative transitions between energy levels, which depopulate the excited electronic levels through multiphonon emission. The probability of such processes depends on the factor s and on the relative value of the energy gap between levels compared with the energy of the phonon which is most effectively coupled to the excited electronic state. In order to get a high quantum efficiency it is necessary that the probability of the radiative transition, Trl, be much larger than that of the non-radiative process, Tn~• in a temperature range that includes the room temperature.
We see then that a strong electron-phonon coupling has contradictory effects on the properties of the phonon-terminated lasers: on the one hand it increases the range of tunability; on the other hand it reduces the quantum efficiency of emission, especially when the energy gap between the excited and ground levels is small (in the infrared). These contradictory effects restrict the wavelength range (toward the infrared) of tunable emission at room temperature for solid state lasers.
Besides the transition inside the ground electronic configuration, the optical spectra of transition ions in crystals may sometimes show very strong bands, usually situated in the high energy range, . due to transitions between the ground configurations and other configurations of opposite parity (charge transfer - CT - bands). If from these bands the excitation is transferred rapidly and efficier.tly to the excited states of interest for laser emission, the CT bands could be used as pump bands.
2.1. Lasers with d.-electron systems
From these facts we see that the d-electron systems in crystals show good prospects as active ions for tunable solid state laser emission. Their optical spectra contain usually sharp zero-phonon transitions accompanied by strong and broad phonon sidebands inside the ground dN configuration, as well as strong charge transfer bands. The d-d transitions could oe either spin-allowed or not. The lifetime of electronic charge transfer bands is of the order of 10-8 - lo-9 · s; for spin-allowed d-d luminescence transitions it is of the order lo-5 - 10-6 s while for spin-forbidden d-d luminescence it is of the order of lo-3 s. Thus in order to have a high peak cross-section it is recommended _one use spin-allowed broad sidebands of d-d transitions with high quantum efficiency, provided a pump source with suitable pulse duration is employed. As pump bands, the CT bands could be used provided they show a good spectral match with the pump source and transfer efficiency from the excitation to the emitting level. Unfortunately, for many of the d.-electron ions in the condensed state the spectroscopic information is restricted only to some chemical compounds or solutions and few data are available on ions ·in crystals. Moreover, in many of the existing optical studies on d-ions in crystals useful information for laser work, such as details of the luninescence spectrum, excitation spectra, lifetime, ESA studies, effect of concentration or temperature on luminescent properties and so on is missing. Because of this, the search for active media for tunable solid state lasers
40
determined a new field of if'lterest [1] in the spectroscopy of the d-ions in crystal~ and in understanding the complex interactions of these electrons with the static and dynamic crystal field. A very sustained effort was directed toward development of methods for rapid growth of crystals of suitable quality for spectroscopic and stimulated emission studies or improvement of the crystal growth methods to produce crystals of suitable size and quality for lasers. It is now thought that suitable systems with d-electrons may be found to cover the wavelength range from 0.5 to 4 or even 5· microns.
2.1.1. Ions of 3d group
Usually the 3d ions come in crystals as substitutional ions surrounded either by six anions placed in the corners of a regular or slightly distorted octahedron or four (tetrahedron) or -eight (cube) anions. In 1110st cases- the crystal field can be regarded as having a strong cubic component (which determines most of the splitting) and a weak axial or lower symmetry component whose effect is of the same order of magnitude as the spin-orbit coupling. Sometimes the Jahn-Teller effect leads (or contributes together with other effects such as, for instance, a low-symmetry crystal field component) to the splitting of some levels left degenerate by the cubic crystal field. Thus the crystal field splitting in the case of 3d ions in crystals can be taken into account in two stages: first, the effect of the cubic part of the crystal field (which is of the order of 103-1o4 cm-1) upon the spectroscopic terms of the free ion is considered and subsequently the lower symmetry part (of the order of 102 cm-1) is taken into account together with the spin-orbit coupling ( - 102 cm-1) and/or the Jahn-Teller splitting (102-1o3 cm-1). The distance between the· energy levels, and consequently the range of emission, is mainly determined by the cubic part of the crystal field. Thus the positions of the energy levels inside a 3dN configuration in intermediate crystal fields are determined by the Racah parameters B and C, the cubic crystal field parameter o4, the spin-orbit parameter tcJ, and the low-symmetry crystal field parameters B'2 and s: . Since the ratio of the parameters 8 ano C is almost constant in a given transition series, the main parameters which determine the splitting of the energy levels of a 3dN configuration under the action of the cuoic components of the crystal field are the Racah parameter B and the cubic crystal field parameter.
For a given 3dN configuration the Racah parameter is larger for ions in higher valence states; a reduction of this parameter in crystals was observed, stronger for ions in lower valence states. For a·characterization of the cubic crystal field, instead of the 84 parameter, another parameter, Dq, is usually used, so chosen that the cubic crystal field splitting of the dl configuration (into a triplet and a doublet) be equal to lOOq = 12004. The effect of the cubic crystal field upon the free-ion spectroscopic terms versus the ratio Dq/8 can be represented as diagrams (the so-called Tanabe-Sugano diagrams) for each of the 3dN ions.
The cubic crystal field parameters can be calculated for specific models of interaction. Thus in a point charge approximation 84 is negative for coordinations eight and four and positive for coordination six, and we have 84 (coord 6) = -9/4 84(4) = -9/8 84(8). The ions with ground configuration 3dl0-N may be regarded as ions which have N positive holes in a complete )dlO electronic shell; thus the crystal field parameter 84 for these ions has the same value, but with opposite sign, as for the ions JdN.
We note that the crystal field analysis gives the positions of the levels which correspond to the zero-phonon lines in optical absorption or
41 luminescence spectra. The maximum of the broad phonon sidebands may be shifted from the zero-phonon lines by energies as large as 103 cm-1.
Since the elements from the 3d transition group can form in condensed state a variety of ions, a given 3dN configuration can be formed by ions of different valence states. Table I gives the ions of the 3d-group for each of the 3dN ground configurations.
Table I. Ions from 3d-group
Sc2+ n2+ y2+ cr2+ Cr+ Fe2+ eo2+ co+ Ni+
n3+ v3+ cr3+ ,.,3+ ,.,2+ eo3+ Ni3+ Ni2+ eu2+
v4+ cr4+ ,.,4+ Fe3+ eu3+
cr5+ ,.,s+
An analysis of existing spectroscopic data on these ions in the condensed state (doped crystals or glasses, solutions, chemical compounds) shows that for a given 3dN" configuration the crystal field splitting is larger for the ions in the higher valence state. At the same time, the crystal field splitting for a given ion in sites of the same coordination depends on the nature of the ligands, i.e. is ordered according to the series er- < cl- < F- < HzO < NH3 < CN-. Thus, by selecting carefully the ionic species and the host, systems with a suitable energy-level scheme for generation ir the desired wavelength range may be found.
A • .canfiguratign 3dl. This configuration is the siqllest of the series; it has only one spectral term, 2o which, under the action of a cubic crystal field, splits into two levels -- a doublet 2E and a triplet 2T2• In an octahedral field the triplet is lowest while in tetrahedral or eight-fold cubic fields the doublet is th