Resonance radiation plasma (photoresonance plasma)
`
`l. M. Beterov, A. V. Eletskii, and B. M. Smirnov
`
`Usp. Fiz. Nauk 155, 265—298 (June 1988)
`
`A plasma formed by the action on a gas ofmonochromatic radiation whose frequency
`corresponds to the energy ofa resonance transition in the atom is studied. The elementary
`methods ofcreating and studying a plasma ofthis type are analyzed. The kinetics offormation ofa
`photoresonance plasma is studied, including collision processes with participation of excited
`atoms leading to formation of molecular ions and highly excited atoms, processes of stepwise
`ionization and triple recombination, and radiative processes. A photoresonance plasma is
`characterized by a high electron density with a relatively low electron temperature; for this reason
`the condition ofideality is more easily violated in a plasma ofthis type. Some ways of utilizing a
`photoresonance plasma are presented
`
`TABLE OF CONTENTS
`1. Introduction ............................
`2. Types of photoresonance plasmas and methods of preparing them ................ 536
`2.1. Photoresonance non-laser plasmas. 2.2. Photoresonance laser plasmas.
`2.3. Quasiresonance plasmas. 2.4. Beam and jet photoresonance plasmas.
`................................. 541
`3. Elementary processes in a photoresonance plasma .........
`3.1. Photoprocesses. 3.2. Collision of electrons with excited atoms. 3.3. Ioniza-
`tion with participation of excited atoms.
`..... 545
`4. Properties of photoresonance plasmas ................................
`4.]. Establishment of equilibrium in photoresonance plasmas. 4.2. Nonideal
`photoresonance plasmas.
`5. Optogalvanic spectroscopy ..............................................
`5.]. The optogalvanic efiect. 5.2. Laser isotope analysis.
`
`.............. 552
`6. Conclusion ..................
`.............. 553
`References
`
`.
`
`................................. 551
`
`1. INTRODUCTION
`
`One of the methods of creating a plasma involves the
`action of optical resonance radiation on a gas. This method
`was first realized by Mohler and Boeckner,I who observed
`the formation of ions upon irradiating cesium vapor with
`resonance radiation. Thus they established the possible oc-
`currence in the gas ofthe process ofassociative ionization, in
`which an electron and a molecular ion are formed by colli-
`sion of excited and unexcited atoms, so that the energy need-
`ed for ionizing the atom is released through formation of a
`molecular ion. Studies of phétoresonance plasmas (PRPs)
`began with the study of Morgulis, Korchevoi, and
`Przhonskii2 in 1967. By illuminating cesium vapor with res—
`onance radiation to obtain a gas with a high concentration of
`excited atoms, they found as a result that a plasma is formed
`with a rather high concentration of charged particles. Since
`the ionization energy ofthe cesium atom (3.89 eV) exceeds
`by more than twofold the energy ofa resonance photon ( 1.39
`or 1.45 eV), this result indicated a complex, multistep char-
`acter ofthe kinetics ofthe ionization of cesium atoms under
`the conditions studied. The subsequent detailed studies of
`this kinetics“5 have permitted obtaining rich information on
`the mechanisms and rates of processes involving excited
`atoms.
`
`The formation of a photoresonance plasma is accompa—
`nied by various phenomena that occur in gases. Thus, the
`ionization of a gas under the action of resonance optical radi-
`ation is one ofthe fundamental mechanisms offormation of
`an ionization wave in the gas, which propagates upon apply-
`
`ing an external electric field.° This same mechanism plays
`the decisive role in the phenomenon of ionization of a gas
`ahead of the front ofa strong shock wave in the gas.7 Irradia-
`tion ofa gas with Optical resonance radiation is used as one of
`the methods of preliminary ionization ofthe active medium
`of high-pressure molecular lasers.3 This enables one to cre-
`ate a plasma homogeneous throughout the volume, while
`avoiding the factors that favor the development of instabili-
`ties and Spatial inhomogeneities ofthe active medium.3 The
`stated method of creating a high-density plasma homoge—
`neous throughout the volume has attracted the attention of
`investigators also in connection with the problem ofheating
`thermonuclear targets with beams of light ions.“ In this case
`the ionization of the gas under the action of resonance radi-
`ation enables One to create for a short time an extended plas-
`ma channel, which serves for transport ofthe ion beam to the
`target, while hindering electrostatic repulsion of the ions.9
`The potentialities of study of photoresonance plasmas,
`as well as the set of their applications, have been expanded by
`the invention of frequency-tunable lasers. On the one hand,
`this has enabled considerable increase in the fluxes of reso—
`nance radiation transmitted through the gas, and on the oth—
`er hand, study of the processes that occur upon optical exci—
`tation of various states of the atom. The photoresonance
`plasma formed by using tunable lasers is used as a nonlinear
`element in frequency transformation of laser radiation,m as
`a source ofions ofa given type,’ "'3 etc.
`The set of phenomena that occur in a photoresonance
`plasma is closely bordered by the optogalvanic effect, which
`
`535
`
`Sov. Phys. Usp. 31 (6), June 1988
`
`0038-5670/88 /060535-20$01 .80
`
`© 1989 American Institute of Physics
`
`535
`
`t. u-
`
`ASML 1027
`
`ASML 1027
`
`

`

`consists in a change in the electrical properties of a gas-dis-
`charge plasma or flame (e.g., the volt-ampere characteris-
`tics) when acted on by optical resonance radiation.” The
`optogalvanic effect is used for determining trace impurities
`of elements in a gas, in studying the mechanisms of elemen-
`tary processes in a gas-discharge plasma, in controlling the
`parameters of a gas-discharge plasma for transmission and
`processing of information, and as a detection method in laser
`spectroscopy with superhigh sensitivity.
`The process of multistep ionization of atoms Opens up
`broad possibilities. The use for this purpose simultaneously
`of several frequency-tunable lasers enables one to transfer an
`appreciable number of atoms of a certain type to a given
`highly excited state. This technique, usually based on using
`atomic beams, is applied for detecting individual atoms, for
`detecting submillimeter radiation, for generating coherent
`radiation in the UHF range (maser), and in experiments on
`laser isotope separation.I5 The identification of highly excit-
`ed atoms is performed by ionizing them in an external elec-
`tric field. Here one uses the sharp dependence of the ioniza-
`tion probability in a field of given intensity on the effective
`value n“ of the principal quantum number of a highly excited
`atom. The ions formed by ionization are extracted from the
`system by applied fields.
`The properties and specifics of a photoresonance plas-
`ma are associated with the processes that occur in them.
`Thus a photoresonance plasma whose properties are deter-
`mined by elementary collision-radiation processes, is natu-
`rally distinguished from a laser plasma, in which the trans-
`formation of the energy of the laser radiation into the energy
`of plasma particles results from the excitation of collective
`motions in the plasma.
`At the same time it seems natural to classify as a photo—
`resonance plasma one formed by the action on a gas of radi-
`ation having a frequency that does not necessarily corre-
`spond to a resonance transition, but also to transitions
`between ground and highly excited states, or transitions
`between two excited states. In all objects of this type, the
`fundamental ionization mechanism is collisional pr0cesses
`involving excited atoms (more rarely-mmolecules).
`This review will analyze the current status of studies of
`photoresonance plasmas, information will be presented on
`the properties and parameters of this object, and problems
`will be discussed involving the application of photoreson-
`ance plasmas in the technique of physical experimentation
`and in applied fields.
`
`2. TYPES OF PHOTORESONANCE PLASMAS AND METHODS
`OF PREPARING THEM
`
`2.1. Photoresonance non-laser plasmas
`
`The most convenient method of obtaining a photore—
`sonance plasma not involving the use of laser radiation con-
`sists in irradiating a gaseous substance with a gas-discharge
`lamp filled with the same substance. Here the parameters of
`the photoresonance plasma are determined by the intensity
`of the resonance radiation emitted by the lamp. The most
`interesting results have been obtained in cases in which the
`lamp is characterized by a high coefficient of transformation
`of electrical energy into energy of resonance radiation. Thus,
`in the pioneer study,I a quasistationary plasma with an elec—
`tron density Na ~ 1012 cm‘3 and an electron temperature
`
`536
`
`Sov. Phys. Usp. 31 (6), June 1988
`
`Tc ~ 103 K was formed upon irradiating Cs vapor at a pres-
`sure of 10—2—10“I Torr with a cesium gas-discharge lamp.
`The radiation of the lamp corresponded to the spectral range
`/l>600 nm. The bulk of the energy of the radiation of the
`lamp was contained in the lines at /l = 894.3 and 852.1 nm,
`corresponding to the 6ZS~6 2P resonance transitions. A de-
`tailed mass-spectrometric analysis showed1 that the main
`type of ions in this plasma is the atomic ion Cs+. This indi—
`cates a complex character of the kinetics of ionization in the
`photoresonance cesium plasma; in particular, this implies
`that the process of associative ionization in the collision of
`two resonance-excited Cs atoms (62P) is not the fundamen-
`tal ionization channel.‘ As is implied by the results of de—
`tailed experimental studies of recent years,” the complex
`kinetics of ionization of atoms in the cesium photoresonance
`plasma includes processes of collision of two resonance-
`excited atoms,
`
`203 (6 2P) ~> Cs (62S)+CS (8 213),
`
`(2.1)
`
`processes ofquenching ofthe excited Cs atoms (62F, 82F) by
`electron impact, processes of ionization of excited atoms in
`collisions with fast electrons formed as a result of quenching,
`processes of associative ionization, etc.
`As the source of optical radiation for creating the pho-
`toresonance cesium plasma, not only cesium lamps, but also
`helium gas-discharge lamps have been successfully em-
`ployed. This possibility arises from the coincidence of the
`wavelength of one of the effective transitions in the spectrum
`of He (/1 = 388.8 nm) with the wavelength of the 6ZS—SZP
`transition of the cesium atom. We note that this coincidence
`
`is the basis of one of the first schemes for excitation of a gas
`laser with Optical pumping, which was proposed by F. A.
`Butaeva and V. A. Fabrikant'6 and realized experimentally
`in Ref. 17. Using this scheme, a photoresonance cesium plas-
`ma was obtained with the parameters Ngzp : lO7 cm‘3,
`Ne ~108—10" cm‘3, Tc ~O.3 eV, PCs ~10—3’~-10_2 Torr.18
`In a plasma of this type, processes of stepwise excitation of
`highly excited levels from the BZP level by electron impact
`play an essential role. Figure I shows the dependences of the
`parameters of this plasma on the density of the Cs vapor with
`fixed intensity of resonance irradiation.
`
`lit/15,7“ rel. units
`
`
`
`am
`
`2’5
`If”, 10 3[cm—3
`FIG. 1. Dependence of the parameters ofa cesium plasma on the density
`osz vapor at fixed intensity of resonance irradiation.” [Aensity n‘ of
`excited atoms (Cs, 82F);2#density n: ofeiectrons; 3-temperature T, of
`electrons.
`
`Beterov e! 2/.
`
`536
`
`
`
`

`

`lane,iagcm"“3
`Jay eV
`
`°
`
`A
`
`o
`
`O
`
`A
`
`A
`
`0407,)
`
`13-20;)
`
`o
`
`°
`
`o
`
`A
`
`A
`
`10
`
`a
`
`r
`
`7E
`
`E
`
`UHF diagnostics, at a concentration of Cs ~ 3 X 10'5 cm”,
`Hg~3><10’3 cm“, and a pressure ofbufl‘er gas (Ar) ~ 100
`Torr, the photoresonance plasma was characterized by a
`density NC ~10‘2 cm‘3 and a temperature Tc 22000 K. The
`role of the buffer gas consists in reducing the effectiveness of
`the diffusion losses of charged particles, and hence, in main-
`taining the density of the electrons of the photoresonance
`plasma at a sufficiently high level. An analogous scheme for
`creating a photoresonance plasma was realized in Ref. 23,
`where a mixture ode and Cs vapors was irradiated with the
`resonance light ofa cadmium lamp.
`
`2.2. Photoresonance laser plasmas
`
`The invention and wide spread of frequency-tunable
`narrow-band lasers based on dyes has stimulated to a consid-
`erable degree the study of the properties and possible appli-
`cations of photoresonance plasmas. The set of studied ob—
`jects has considerably expanded to encompass all the alkali
`metals, and also a number of metals ofthe second and third
`groups of the periOdic table. The object of the studies was the
`mechanisms of ionization and recombination of particles of
`a plasma, the elucidation of the role of the buffer gas, the
`possibility of more complete extraction of ions in a photore-
`sonance plasma and identifying them, etc.
`Among the large number of experimental studies (see
`the review ofthe early studies““) on the creation and study of
`photoresonance laser plasmas, primary attention is owed to
`a series of publications reporting the practically 100% ioni-
`zation of metal vapors irradiated with the resonance radi-
`ation of a pulsed laser of relatively low power. Figure 3
`shows a diagram of the first of these experiments, which was
`performed by Lucatorto and McIlrath.24 The radiation ofa
`dye laser pumped with a flash lamp was tuned to a line at
`/l = 589.6 nm, which corresponds to the 328 l ,12—32PU2 tran-
`sition ofthe Na atom, and was focused on a lO—cm column of
`Na vapor with addition of He to a total pressure about
`1
`Torr. The pulses of laser radiation of duration 500 ns had an
`energy of 300 m], which corresponds to a pulse power of 0.6
`MW. The degree of ionization of the vapor was determined
`with a vacuum-ultraviolet spectrograph, which enabled
`measuring the absorption coefficient in the region xi = 15—
`42 nm. Figure 4 shows typical densitograms ofthe spectrum
`obtained without (a) and with (b) laser irradiation. As is
`shown by comparison of the absorption coefficients in the
`region of 1:32.21 nm corresponding to transitions of the
`Na+ ion, the degree of ionization of Na during the laser
`pulse reaches 100%. The practically complete ionization of
`the Na vapor is also indicated by the sharp (by a factor of
`
`FIG. 3. Diagram of an experiment to produce and study a photore-
`sonance sodium plasma with a high degree of ionization. I—radi-
`ation source with a continuous spectrum; 2—anode; J—toroidal
`mirror; 4—capillary rings; 5—vacuum pump; 6—furnace; 7—
`three-meter reflecting spectrograph; 8—-dit’fraction grating; 9—
`photoplate; 10—cylindrical lens; 11-—-laser; 12—delay generator;
`13—pulse shaper.
`
`75:,i—._l_l_l_|_l.u_u.l___l_._.
`dwsch :
`2
`4 E 8/0
`/
`A
`
`FIG. 2. Dependence ofthe concentration (I) and temperature (2) ofthe
`electrons of a photoresonance plasma in Hg vapor on the value of the
`discharge current in the optical-excitation source?”
`
`Mercury-vapor lamps are also an intense source of reso-
`nance radiation, and have been successfully used to create a
`photoresonance plasma in mercury vapor.""10 Investiga-
`tions in this direction were stimulated by practical problems
`of0ptical separation ofmercury isotopes.ZI Upon irradiating
`a gas-discharge mercury lamp with the resonance line corre-
`sponding
`to
`the
`transition
`Hg(63Pfi’ —»6‘SO)
`(xi = 253.7 nm) , an increased concentration and a reduced
`temperature of the electrons was observed. This was asso»
`ciated with an increase in the efficiency of ionization under
`the action of the resonance radiation. The mercury lamp
`used as the source of resonance radiation had the form of a
`jacket arranged coaxially around the cylindrical cell being
`irradiated.20 With a pressure of mercury vapor in the cell of
`~0.05 Torr, the concentration ong atoms in the 63F? state
`reached 101 ' cm”. The electrical characteristics ofthe pho-
`toplasma that was formed are shown in Fig. 2. Owing to the
`coaxial excitation geometry used in this experiment. a high
`degree of homogeneity of the photoresonance plasma was
`attained.
`
`An interesting variety of photoresonance plasma was
`realized in Ref. 22, where a mixture ong and Cs vapors was
`irradiated with the resonance radiation of a mercury lamp
`(xi = 253.7 nm). The ions were formed by the Penning reac-
`tion
`
`Hg(63P,)+Cs -> Hg+Cs*+e.
`
`(22)
`
`According to the measurements performed using probe and
`
`
`
`537
`
`Sov. Phys. Usp. 31 (6), June 1988
`
`Beterov et al.
`
`537
`
`

`

`
`
`
`
`Transmission,rel.units
`
` FIG 4. Densitograms of the absorption Spectrum of sodi-
`
`32
`
`30
`
`28
`
`um ” a—Absorption of sodium vapor without irradiation;
`the dots indicate the emission lines ofthe vacuum arc, and the
`dotted line the absorptionin He b—Absorption ofNa vapor
`irradiated with resonance iaser radiation The solid squares
`indicate the absorption lines of neutral Na.
`
`26 2, nm
`
`pors under the action of resonance laser radiation has been
`observed in subsequent experiments with vapors of Li,26
`Cs,27 Ca,”, Sr,” Baffl‘34 Na,35 and Mg.36 The properties of
`the PRP formed thereby have been studied in greatest detail
`in Ref. 36, where Mg vapor was irradiated with radiation
`corresponding to the resonance transition 3'So—>3‘P,.
`(/1 = 285.2 nm) of the singlet system oflevels. A pulsed liq-
`uid dye laser with frequency doubling pumped with the sec-
`ond harmonic ofa neodymium laser with a pulse repetition
`frequency of 3 Hz was used as the radiation source. The
`pulses of ultraviolet radiation had a line width ~0.l cm‘ ',
`and duration ~ 10 us at a power ~1 kW Another liquid
`laser (/1 = 280. 3 nm) tuned to the transition 32PM~-3 I-“
`of the Mg+ ion was used to measure the concentration of
`Mg+ ions. In addition, the experiment measured the lumi-
`.. _..ln.. ..
`1"
`A __AL
`ML.
`nescence of the Mg vapor and the time-dependence of the
`FILULUUUIJCIIL
`l hC pulac Ul
`lLLC Pl Uue iaurauOu hau a Uelay
`with respect to the pump pulse variable1n the range up to
`100 ps. Figure 7 shows a diagram of the experiment.36
`The experimental chamber, which was made of stain-
`less steel and fitted with quartz windows and internal plane
`electrodes to measure the photocurrent, was filled with an
`inert gas at a pressure ~ 1 Torr. The pressure of metal vapor
`(Mg) was varied in the range 0.1—1.0 Torr. Upon tuning the
`pump radiation to the frequency of the resonance transition
`
`
`
`Time, 113
`
`FIG. 6. Dependences ofthe electron temperature on the time elapsed after
`cessation of the laser radiation pulse.25 The curves are marked with the
`different values ofthe concentration of Na vapor.
`
`10") decline in the absorption coeflicient of the resonance
`laser radiation owing to the formation of the photoresonance
`plasma. Even the first rough estimates of the authors24 indi-
`cated a complex, multistep mechanism of ionization of the
`vapor in the described experiments. Neither three-photon
`ionization nor radiation collision
`
`2Na*(3p)+hw—>Na+(2p°)+Na (SSH-e,
`
`(2.3a)
`
`nor multistep collisional excitation
`
`21Va (3p) —>- Na (55) + Na (35)
`
`(2.3b)
`
`with subsequent photoionization of the excited Na ( 55)
`atoms possess values of the rate constants high enough to
`explain the observed 100% ionization of the atoms.
`Analogous results were obtained in Ref. 25, where a cell
`filled with a mixture of Na vapor and Ar at a pressure of
`several Torr was irradiated with pulses of radiation of a dye
`laser based on rhodamine 6G pumped with a nitrogen laser.
`The radiation pulses of 100 kW power had a duration of 10
`ns and a line width ~0.l nm. The ionization of the gas was
`measured by the magnitude of the photocurrent, for which a
`typical dependence on the Na vapor density is shown in Fig.
`5. Figure 6 shows the dependence of the electron tempera-
`ture on the time elapsed since the end of the laser pulse, as
`obtained by the Langmuir double-probe method. The results
`of the probe measurements of the electron density performed
`at a point separated by 2 mm from the focus of the laser beam
`are shown in Table I.
`The effect described above of intense ionization of va-
`
`700 kW
`g?
`N5
`
`g 70’
`3
`E w 2
`1Em
`E 70
`a:
`
`58‘
`
`7
`5
`E 0
`
`m 75
`7” M
`70 V3
`Density of Na vapor. cm‘3
`
`7g W
`
`FIG. 5. Dependences of the photocurrent on the density of Na vapor
`obtained at different levels oflaser power.25
`
`538
`
`Sov. Phys. Usp. 31 (6), June 1988
`
`Beterov er al.
`
`538
`
`
`
`

`

`TABLE I. Values of the electron density NC in a photoresonance plasma measured at different
`values of the density of Na vapor.25
`}
`1.0.1015
`1
`5.9.10“
`1.8.10”
`i
`5.31015
`i’\’Na,cm"3 i
`l
`l
`[1,0,1013 10.1012
`
`
`6,6-10”
`I
`9,5.1012
`1
`Ne, 0171—3
`
`
`ofMg, an emission from the Mg vapor arose in the region of
`the laser-beam focus at frequencies corresponding to transi—
`tions n lD2—3 lPl (n = 4—10) ofthe singlet system and n 3S—
`3 3P (n = 4, 5) of the triplet system of Mg levels. Here the
`absence was noted of radiation at the strong line 5 ‘S0—3 ‘P,
`(xi. =2 571.1 nm) and several other strong lines. As was
`shown by measurements ofthe density ofMg ions performed
`with probe radiation, the maximum concentration of ions
`(~ 2 X 10'4 cm") was observed about 30 ns after the end of
`the pulse ofpump radiation, while the total time of existence
`of the PRP amounted to ~ 10,115. The maximum degree of
`ionization ofthe plasma reached 5%. The absence ofsatura—
`tion in the dependence of the ion density on the intensity of
`pump radiation allows one to expect an increase in the PRP
`density upon using more intense resonance pump radiation.
`Another detailed study worthy ofattention on the char-
`acter of formation and physical properties of a PRP with a
`rather high degree of ionization was performed by the auth-
`ors of Ref. 35, where a dye laser (rhodamine C) was used as
`the source of resonance radiation, It was pumped with the
`radiation ofthe second harmonic ofa solid-state pulsed laser
`based on yttrium aluminum garnet of type LTIP4—5 with a
`pulse-repetition frequency of 12.5 Hz. To narrow and tune
`smoothly the emission line of the dye laser, a difi‘raction
`grating (12001ines/mm) was used and was set up in a glanc-
`ing-incidence system, The width of the emission line ofthis
`laser amounted to ~ 1 nm, and the emission power was :40
`kW at a pulse duration of ~ 10’ x s. The laser was tuned
`either to the resonance transition of Na (xi = 589.0 nm) or
`to the wavelength 578.7 nm corresponding to two—photon
`absorption to the excited state of Na (4d 2D5/z).
`The Na vapor diluted with inert gases filled the dis-
`charge tube, which was made of Pyrex and had niobium tu-
`bular electrodes. The pressure of the vapor in the tube was
`maintained by using a heating element. Under the conditions
`
`
`
`FIG. 7. Diagram of an experimental arrangement to form a quasireson-
`ance laser plasma in Mg vapor.” I—neodymium garnet laser; 2, 4~fre~
`quency doublers based on KDP; 3—dye laser; 5———delay circuit; 6—cell
`with Mg vapor; 7dmonochromator with photomultiplier (PM); 8—05-
`cillograph.
`
`539
`
`Sov. Phys. Usp. 31 (6), June 1988
`
`of the experiment one pulse of laser radiation contained
`about 5 X lOM photons. About the same number ofNa atoms
`was contained in the irradiated volume. Here conditions
`were selected such that, on the one hand, practically com-
`plete absorption of the laser radiation was attained, and on
`the other hand, the intensity of absorption varied weakly
`along the laser beam. This was made possible by detuning
`the frequency of the laser beam from the center of the ab-
`sorption line by four widths of the Doppler absorption con-
`tour (equal to 0.0024 nm). Whereas at the center of the
`absorption line the optical density of the medium amounted
`to ~ 10“, at the detuned frequency it was close to unity.
`The formation ofa PRP was measured from the change
`in electrical conductivity of the irradiated volume. To do
`this, a voltage was applied to the electrodes ofthe discharge
`tube smaller than the excitation voltage ofthe discharge, and
`the current was measured that arose in the electrical circuit
`under the action of the laser illumination. Figure 8 shows an
`oscillogram ofthis current obtained at a voltage on the elec-
`trodes of 100 V and a vapor pressure of Na of3 X 10'3 Torr.
`The pressure of the bufi‘er gas amounted to 1 Torr. As the
`pressure of Na vapor was increased from 3 X 10’4 Torr to
`0.2 Torr, the magnitude of the signal and its duration in-
`creased by more than an order ofmagnitude. Upon detuning
`from the resonance at 17. = 589 nm, no plasma formation was
`observed.
`
`Plasma formation was also observed in two-photon la-
`ser excitation of the level 4p2D5n. The oseillograms of the
`current obtained here at a pressure of Na vapor ~ 5 X 10—3
`Torr are shown in Fig. 9. The rapidly gr0wing advance front
`of the current pulse was associated35 with the phenomenon
`of three-photon ionization of atoms via a two-photon reso-
`nance, and the subsequent, smoother increase in electrical
`conductivity—with a supplementary mechanism of forma-
`tion of free electrons (heating of electrons by superelastic
`processes and subsequent ionization of atoms by electron
`impact).
`As an analysis of the experiments to create and study a
`laser PRP shows, a plasma of rather high density is formed
`using very-low-power laser radiation. This arises from the
`high absorption power of gases for resonance radiation, and
`also the high efficiency of conversion of the energy of reso-
`nance-excited atoms into ionization energy.
`
`2.3. Ouasiresonance plasmas
`
`As has been established in a number of experiments of
`recent years,”'3”'37 to form a photoresonance plasma one
`need not use radiation whose frequency corresponds to a
`resonance transition between the ground and excited states
`of the atom. Efficient ionization of the atoms ofa metal va-
`por has also been observed using radiation corresponding to
`a transition between two excited states of the atom. Here the
`intensity of the laser radiation was not so great that one
`
`Beterov et al.
`
`539
`
`

`

`
`
`r
`
`-_—-—_...__
`
`[3
`
`FIG. 8. Oscillogram ofthe photocurrent that arises
`upon irradiating adischarge tube containing a mix
`ture of Na + Ne with a pulse of laser radiation
`(1 = 589 nm),” in the presence ofa discharge (a),
`and at a voltage below that for ignition of the dis-
`charge (1)).
`
`could ascribe such an unexpected result to effects of multi-
`photon ionization of atoms, including the nonresonance ex-
`citation ofa real level.3B
`A plasma of the type being studied has been called a
`“quasiresonance laser plasma.” As has been established by
`detailed experimental studies,
`to form a quasiresonance
`plasma one can use radiation at one of many frequencies
`corresponding to transitions between excited states of the
`atom. Thus, a quasiresonance cesium plasma was efficiently
`formed upon irradiating Cs vapor with laser radiation hav-
`ing/i. = 583.9, 601.0, 603.5, and 621.3 nm, corresponding to
`the
`transitions
`25,,210s—~ ZP?,.2 6p,
`2D3,2
`8d—~ 3
`[3/2 6p,
`251,210s—30'3’fl 6p, and 2D3,28d-~2P§’,2 6p, as well as when
`using radiation at several other transitions of the atom.13
`Here the formation of the quasiresonance plasma has a
`threshold character—this phenomenon is observed only
`upon exceeding certain values of the pressure of the vapor
`and intensity of the laser radiation.
`Another interesting feature of a quasiresonance laser
`plasma involves the relatively low level of intensity of radi-
`ation used to maintain it. Thus, in the series of studies cited
`above,‘3 continuous irradiation of cesium vapor of density
`~3 X 10'7 cm'3 with quasiresonance radiation of power up
`to ~ 100 mW yielded a plasma with a degree ofionization of
`10—3. A diagram ofthe experiment is shown in Fig. 10. As is
`shown by analyzing photographs of the plasma column
`formed upon focusing the beam ofa continuous laser of 10
`mW power, the extent ofthe column amounts to about 4 mm
`and the diameter to less than 1 mm. Even weaker radiation
`
`was used in Ref. 37 to maintain a plasma in Na vapor. This
`radiation having a power of ~ 2 mW was tuned in resonance
`with the 3p-4d transitions (xi. = 568.8 or 568.2 nm) of the
`Na atom. The pressure of the Na vapor amounted to ~ 10
`Torr. Detailed spectral studies of the plasma that was
`formed led the authors to conclude that the decisive role in
`forming the plasma was played by the process of associative
`ionization involving a (4d) Na atom.
`The mechanism ofionization of a gas under the action
`of quasiresonance radiation ‘3 includes the process of photo-
`dissociation of dimers ofa metallic vapor, which are always
`present in the system:
`
`A2+fiw—>A +A*.
`
`
`
`One of the atoms formed as a result of photodissociation of
`the dimer exists in the excited state (in the case of Cs this
`state is 6p2PWV3/2 ), and is already capable of resonance ab»
`sorption oflaser radiation. This leads to formation ofhighly
`excited atoms, whose ionization results from subsequent
`collisional processes.
`
`2.4. Beam and jet photoresonance plasmas
`
`The most productive way to study the primary starting
`mechanisms responsible for the formation of a PRP involves
`using atomic beams. In this case one can reduce to a mini-
`mum the role of secondary collisional processes and isolate
`in pure form the process that occurs with participation of
`optically excited atoms and leads to their ionization. A char-
`acteristic example ofsuch an experiment is Ref. 39, where an
`effusion beam of Na atoms was irradiated with a dye laser
`tuned to a resonance transition while pumped with an argon
`ion laser. The density of atoms in the irradiated zone
`amounted to 109 cm”, the intensity of the laser radiation
`was 0.5 W/cmz, and the line width was 20 MHz. The posi-
`tive ions were extracted from the interaction region with a
`pair of electrodes to which a potential was applied, and
`which created an electrical field in the cell. Then the ions
`entered the input of a quadrupole mass spectrometer. This
`made it possible to establish that only molecular Nag+ ions
`are formed under the conditions of the experiment. As was
`shown by comparing the results of the mass-spectrometric
`measurements with those of relative measurements of the
`concentration of resonance-excited atoms based on the flu-
`orescence of the beam, a proportional relation is observed
`between the yield of Na;+ ions and the square ofthe density
`of excited Na atoms. This enabled concluding that the deci-
`sive role is played by the process of associative ionization
`
`2Na(3p)—>Na;+e
`
`(2.4)
`
`and establishing the value ofthe rate constant ofthis process
`(~1.5>< l0“ '3 cm3/s) and its cross section (~0.5><10‘I7
`sz) . I}
`A higher intensity of ionization using resonance radi-
`ation was obtained in ajet experiment,40 in which a beam of
`monoenergetic Cs+ ions was formed in this way. A glass
`
` Continuous
`
`dye laser
` Furnace
`
`FIG. 9. Oscillogram of the photocurrent observed in two-photon laser
`excitation of Na vapor (/l. = 578.9 nm).”
`
`FIG. 10. Diagram of an experiment to form a quasiresonance laser plasma
`using a continuous-wave dye laser. ‘7‘
`
`540
`
`Sov. Phys. Usp. 31 (6), June 1988
`
`Beterov e! a/.
`
`540
`
`
`
`

`

`chamber maintained at a constant temperature in the range
`400—500 K (pressure of Cs vapor ~ 2 X 104—02 Torr) and
`equipped with a nozzle 0.12 mm in diameter was used as the
`source of the cesium vapor. Thus a jet of vapor was formed
`that was concentrated in a solid angle 6 = 2.7 X 10‘2 stera—
`dian and characterized by an intensity of 8.6x 10” atoms/s
`and a mean energy of the atoms of ~0.06 eV. The energy
`spread ofthe atoms in thejet was ofthe same order ofmagni-
`tude. This corresponds to a density of Cs atoms near the
`critical cross section ofthe nozzle of ~10” cm ”3. The iOni-
`zation of the Cs atoms was carried out in two stages. In the
`first stage the resonance radiation ofa semiconductor injec-
`tion CaAlAs laser was used (/l = 852.1 nm), which convert-
`ed the Cs atoms to the (>3P3,2 state. In the second stage the
`radiation of an argon ion laser was used (/1 2 488.0, 496.5,
`and 501.7 nm), which enabled the photoionization of the
`resonancerexcited atoms. Figure 11 shows the dependence of
`the photocurrent on the extracting potential. According to
`the estimates the flux ofions in the jet reached values of 10"
`s“ ‘. A substantial increase in this parameter was obtained in
`a subsequent study by this same author,“ where values were
`reached of ~ 10‘2 so ’ with an energy spread ofthe ions at the
`level of O. 15 eV.
`
`Two-stage and multistage photoionization ofatoms has
`been used successfully to obtain beams ofthe following ions:
`Ca+ ,“ In”r (under the action ofdye—laser radiation),“ Al+
`(using the radiation of an excimer XeCl laser),“ etc.”"5
`The ion beams thus formed were characterized by being
`highly monoenergetic and by a degree ofpurity unattainable
`when using other methods of forming an ion beam. Here, as
`in the studies cited above, the obtainable ion fluxes were
`relatively small, so that problems of extracting the positive
`ions from the photoresonance plasma did not arise. Further
`increase in the efficiency of ionization of atoms in atomic
`beams and jets can lead to the appearance