United States Patent
`
`H91
`
`{Ill
`
`3,900,803
`
`[45] Aug. 19, 1975
`Silfvast et al.
`__________________________.__—————-——————
`
`l54l LASERS OPTICALLY PUMPED BY
`LASER-PRODUCED PLASMA
`
`[75}
`
`Inventors: William Thomas Siltvast: ()bert
`ReevesVVoodII.b0d10flI0hndel
`Twp., Monmouth County. NJ.
`
`[73] Assignee: Bell Telephone Laboratories.
`Incorporated, Murray Hill. NJ.
`
`122|
`
`Filed:
`
`Apr. 24. 1974
`
`[21 1 Appl. No.1463.616
`
`
`331/945 I’: 330/43
`[52] US. Cl. ......................
`
`Int. Cl.‘-’ ................................. H018 3/09
`[51 ]
`Field of Search...................... 33 ”94.5: 330/43
`[58]
`
`[56]
`
`3.l(w4.782
`IREI .hol)
`
`References Cited
`UNITED STATES PATENTS
`
`l/l‘fl‘iS Ordwtiy. Jr Ill/94.5 P
`6/1974
`Shang etal Ill/94,5 P
`OTHER PUBLICATIONS
`
`Furumoto ct al.. Optical Pumps for Organic Dye La-
`sers. Applied Optics. Vol. 8, No, 8. (August 1969).
`pp.
`lfil3—lbl3.
`
`Primary lirumim'r—William L. Sikes
`Attorney. Agent. or Firm—Wilford L. Wisner
`
`[57 1
`
`ABSTRACT
`
`Laser solids. liquids and gases are pumped by a new
`technique in which the output from an efficient mo-
`lecular laser. such as a CO._. laser.
`ionizes a medium.
`such as xenon. into a generally cylindrical plasma vol—
`ume.
`in proximity to the pumped laser body. Break‘
`down yields a visible and ultraviolet-radiation-emitting
`plasma in that volume to pump the laser body. The
`spectral radiance of the plasma is significantly higher
`than that produced by a d-c—dischargc-heated plasma
`at nearly all wavelengths in the plasma spectrum. The
`risetime of radiation from the laser-produced plasma
`can also be significantly shorter than that of a d—e
`heated plasma. A further advantage resides in the fact
`that in some applications the attenuating walls needed
`by llashlamps may be eliminated with the result that
`laser
`threshold is more readily reached. Trawling
`wave excitation may he provided by oblique incidence
`of the pumping laser beam through the ionizable me-
`dium to create sequential ionization of portions of that
`medium along the length of the pumped laser body.
`
`5 Claims. l2 Drawing Figures
`
`
`
`DETECTOR
`
`MASKING
`
`72 MATERIAL
`FLASHLAMP
`
`
`
`
`
`ASML 1015
`ASML 1015
`
`

`

`PATENTEU AUG 1 9 I975
`
`3.900.803
`
`Fr-
`ILL.
`Ski. .
`
`IIGFY
`
`5088
`
`mm
`
`
`
`AEcVIhozw4w><>>
`
`con0?00¢comcomcomcom
`
`
`
` _<22:82“:fun/h,
`
`5/,
`
`(LUU —Ja;s— ZWUJ/suDM ) HDNVIGVH WHlDEdS
`
`3Gxu.
`
`

`

`PATENTEU AUG 1 91975
`
`3.900.803
`
`u! M.»
`
`Li:
`
`0.3fl
`
`00¢
`
`comcom
`
`
`
`:5:505.9%
`
`
`
`
`
`.3233;8039.5-mmm<.._
`
`0.m
`
`O.N
`
`o._
`
`(UJU
`
`~345—
`
`zww/salno!’ ,1) mama almosgv
`
`fix6x14.
`
`
`
`

`

`PATENTEDAUG‘9'975
`
`3. 900, 803
`
`mm3m7
`
`23
`
`7 l
`
`co
`
`FIG. 2A
`
`VACUUM CHAMBER
`2|
`
`EEi5agg\
`
`
`KC} SURROUNDED
`
`“‘“\\\“L\‘\\\
`
`BY Xe
`
`
`
`|6
`METAL PLATE
`ON WHICH RADIATION
`IS FOCUSED
`
`
`
`3
`5
`,1
`
`II
`
`23
`
`l7
`
`BUILT INTO SEALED
`CHAMBER WALL
`
`FIGZB
`
`r
`
`l7
`
`l5
`
`V”IIANA“N'IIIIIIIIIII
`
`
`ialI5i4
`
`
`
`l“‘\‘“\“‘
`
`
`\“\\“‘
`
`
` "'Illlllll A
`
`
` 20
`
`l6
`
`55v
`
`’Illllnfl
`
`
`
`23
`
`T
`
`VACUUM
`PUMP
`
`

`

`PATENTEDAUG! 91975
`
`3 . 900 . 8 03
`
`3:113 u a; 7
`
`FIG. 3
`
`PLASMA
`
`31
`
`°
`
`(1-: HEATED PLASMA
`
`
`
`
`
`SPECTRALRADIANCE(“HS/mmz—sfer-nm)
`
`1.0
`
`0.5
`
`A.
`
` LASER'PRODUCED
`
`A
`
`‘
`
`A
`
`600
`
`‘
`
`700
`
`ICC
`
`200
`
`500
`400
`300
`XENON PRESSURE (TORR)
`
`FIG. 4
`
`
`
`
`
`SPECTRALRADIANGE("ans/mmZJter—nm)
`
`LASER-PRODUCED
`PLASMA
`
`0.5
`
`.2
`
`
`l.2
`LO
`B.
`.6
`.4
`INPUT ENERGYtJOULES/cm)
`
`

`

`PATENTEDAUBISIBIS
`
`3,900,803
`
`312m 5 OF 7
`
`5|
`
`73
`
` I” DETECTOR
`
`"
`
`F762 5
`
`MASKI NG
`MATERIAL
`
`FLASHLAMP
`
`57
`
`C02 —
`LASER
`————
`
`53
`
`F761 5
`
`ALUMINUM SHIELD
`
`62'7km
`
`TUBE FILLED
`WITH C3F7]
`
`
`
`\
`FOCUSED \
`RADIATION FOR ——-
`LASER PLASMA / .
`/ 57
`
`
`
`'
`
`‘
`I
`
`POSITION OF
`LASER PLASMA
`
`

`

`PATENTED AUG 1 9 [975
`
`3.900.803
`
`01‘VIA—u
`'-'T 5 DF?
`
`FIG. 7
`
`XENON PRESSURE
`5,175 TORR C3F7I)
`
`I
`
`c351, PRESSURE
`(so TORR XENON)
`
`
`
`552:was>._._mzwhz_mmm<..
`
`50
`
`100
`
`I50
`
`200
`
`250
`
`GAS PRESSURE (TORR)
`
`alinlilll.
`aJillian!nEEK—«III.
`llnlnlll.
` lilimlllll
`linlilll.
`liliilll.
`lillilll.
`lillnflul.
`mammal-l.Illumwll.
`
`
`
`
`
`

`

`PATENTED AUG 1 9 INS
`
`.m
`
`LE
`
`war";
`
`3.900.803
`
`S93ox/Urxf.L
`k.. 5.58.
`
`.\
`
`
`
`/.?.?.>(..(.>(..(\.{>.\,.\
`
`
`
`mzfi45.2235\
`a.\\
`
`2.4.059:
`
`2232
`
`bxka
`
`
`

`

`l
`LASERS OPTICALLY PUMPED BY
`LASER-PRODUCED PLASMA
`
`BACKGROUND OF THE INVENTION
`
`This invention relates to techniques for optically
`pumping laser media.
`Much effort has been exerted in the laser art to pro-
`duce more efficient flashlamps. Any potential laser me-
`dia, especially solids and liquids are most readily tested
`for laser action by optically pumping. The broad band
`pumping light from a flashlamp improves the prospects
`that the experimenter is able to excite the pumped me-
`dium via one or more optical absorptions of the me-
`dium. If a suitable optical absorption is sufficiently ex-
`cited, then a population inversion can be obtained for
`generation of laser radiation.
`Nevertheless, typical flashlamps have been a disap—
`pointment in many cases in trying to achieve laser ac-
`tion in new media or in testing new media for the possi-
`bility of laser action. In general, it has been very diffi-
`cult to reach laser threshold with flashlamps. Even so,
`many new solid and liquid lasers have first been suc—
`cessfully reduced to practice by being pumped by an-
`other, typically higher frequency, laser. Such a pump-
`ing laser must be carefully selected in order to excite
`an appropriate absorption in the pumped medium.
`Many other potential lasers media have never been suc-
`cessfully optically pumped, either by a flashlamp or an-
`other laser.
`
`Still, persistent efforts have been made to improve
`flashlamps, as shown by the article by H. W. Furumoto
`et 3]., Applied Optics, Vol. 8, page 1616 August 1969.
`We have recognized that still further improvement in
`optical pumping techniques is desirable.
`SUMMARY OF THE INVENTION
`
`Our invention is based on our discovery that the
`spectral radiance, that is the total light output, from a
`laser—produced plasma is two to three times greater in
`the ultraviolet region (200 nanometers to 300 nanome-
`ters) than is the spectral radiance from a xenon flash—
`lamp of comparable size and input energy. While we do
`not wish to limit our discovery by any theoretical expla-
`nation, the observed results appear to be attributable to
`the different energy absorption mechanisms present in
`the ate-heated plasma (laser—produced plasma) and in
`the prior art dc—heated plasmas.
`According to the principal feature of our invention,
`laser solids, liquids and gases are pumped by a gener-
`ally cylindrical plasma in a medium such as xenon
`which is broken down by radiation from an efficient
`molecular laser, such as a C02 laser. Breakdown yields
`a visible and ultraviolet-radiation—emitting plasma in a
`generally cylindrical volume in proximity to the laser
`body to pump that body.
`According to another feature of the invention. light
`confinement aids overall efficiency of our new pump-
`ing technique.
`According to still another feature of our invention, in
`many applications the attenuating walls needed by
`flashlamps may be eliminated, so that laser threshold is
`more readily reached.
`Advantages of reduced radiation risetime and higher
`peak powers at shorter wavelengths as compared to
`flashlamps are obtained by the embodiments of our in-
`vention.
`
`3 ,900 ,803
`
`2
`
`5
`
`10
`
`15
`
`20
`
`30
`
`35
`
`45
`
`50
`
`55
`
`60
`
`65
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`Further features and advantages of our invention will
`become apparent from the following detailed descrip—
`tion. taken together with the drawings, in which:
`FIGS. 1A and IB show spectral radiance curves com-
`paring the laser—produced plasma with a dc—heated
`plasma;
`FIGS. 2A and 28 show respectively a pictorial per-
`spective view and an end elevation of the simplest em-
`bodiment for pumping the new laser medium with the
`laser-produced plasma;
`FIGS. 3 and 4 show curves that point out the differing
`spectral radiance characteristics of the different plas-
`mas in terms of gas pressure and input energy;
`FIGS. 5 and 6 show a partially pictorial and partially
`block diagrammatic illustration of a more complex em—
`bodiment of the invention for pumping a photodisso-
`ciation laser in iodine;
`FIGS. 7, 8 and 9 show curves that are useful in ex—
`plaining the operation of the embodiment of FIGS. 5
`and 6', and
`FIG. IO shows a modified embodiment of the inven-
`tion for traveling wave excitation.
`DESCRIPTION OF ILLUSTRATIVE
`EMBODIMENTS
`
`The curves of FIGS. 1A and IB compare the Spectral
`radiance for laser produced plasmas such as are used
`according to our invention and d—c—heated plasmas
`such as are ordinarily present in flashlamps. The com-
`parison is made under comparable conditions with like
`presence or absence of attenuating walls between the
`plasma and the detector. In both FIGS. IA and IB the
`abscissa or horizontal axis is in nanometers a unit of
`wavelength (one namometer = 1 X 10‘9 meters).
`In FIG. IA the ordinate or vertical axis for spectral
`radiance is logarithmic in terms of watts per square mil—
`limeter-steradian-nanometer which is the power re-
`lated and most typical measure of spectral radiance.
`The units of spectral radiance differ from those for ra-
`diation intensity or power because it is an attempt to
`define the total light output from the source as deter—
`mined from a carefully calibrated detector and a care-
`fully calibrated position.
`The ordinate or vertical axis of FIG. [8 is linear in
`terms of absolute energy,
`that
`is, microloules per
`square millimeter-steradian-nanometer which is a mea-
`sure of total energy output, that is, pulse power times
`pulse length rather than being in terms of power. as in
`the ordinate values of FIG. IA.
`In FIG. 1A the solid curve 12 shows the results for
`the laser—produced plasma when the output from a l0.6
`micrometer C0; laser breaks down a cylindrical xenon
`gaseous volume. The dotted curve II shows the results
`for a similar cylindrical plasma produced by a d-c dis-
`charge. Overlying curves 25 through 29 are black body
`radiation curves for constant absolute temperature
`range from 800()°K to 20.000°I(.
`In the ultraviolet region from 200 to 300 nanometers,
`the spectral radiance from the laser-produced plasma,
`e. g. curve 12, is two to three times greater than that
`produced by a xenon flashlamp of comparable size and
`input energy, as shown in curve 11. In addition, the rate
`of increase of ultraviolet emission with increasing input
`energy (not shown in the curves) is much greater for
`the lasenproduced plasma than for the cl cheated
`
`

`

`3
`
`3 £00,803
`
`4
`
`plasma for input energies up to 0.8 Joules per centime—
`ter. which was the maximum laser energy density used
`to produce the plasma.
`As a specific example, a spectral radiance of two
`watts per mmz-steradian-nanometer at 230 nanometers
`from an elongated xenon plasma can be produced by
`focused 10.6 micrometers laser radiation from a C02
`molecular laser providing input energies per unit length
`to the cylindrical volume of the plasma of about 08
`Joules per centimeter.
`The differences of overall shape of curves l3 and 14
`of FIG. 18 with respect to the shape of the analagous
`curves 11 and 12 respectively of FIG. 1A, aside from
`the accentuation of differences in height due to the lin-
`ear scale, are due to the fact that the plasma pulses tend
`to become shorter at shorter wavelengths, for example,
`in the 250 to 300 nanometer range, than at the wave-
`lengths longer than 300 nanometers, This fact may
`prove to be a positive advantage when using the laser—
`produced plasma in pumping some new laser media.
`The relative available radiant energy at
`the longer
`wavelengths may be deduced from the shape of curve
`14 for the laser-produced plasma in the 300 to 700
`nanometer wavelength range.
`The laser-produced plasma was initiated by focused
`10.6 micrometer radiation in the manner shown in
`FIGS. 2A and 2B. Illustratively, an arbitrary solid me—
`dium to be optically pumped such as potassium Chlo—
`ride (KCI) crystal 15 has attached thereto along one
`side an aluminum plate 16 and attached to one end
`thereof the crystal [7 selected to be highly transparent
`at the probable laser wavelength of the crystal 15. The
`assembly of crystal 15. plate 16 and crystal 17 is im-
`mersed in xenon gas from a compressed xenon gas
`source (not shown) in a vacuum chamber 21 which has
`a rock salt window 22 through which 10.6 micrometers
`radiation from CO2 laser 18 is admitted. The vacuum
`chamber may have the xenon circulated therethrough
`by a vacuum pump 23 so that plasma products may be
`eliminated after each pulse.
`While the embodiment of FIGS. 2A and ZB is shown
`in an arrangement for practical application to pumping
`a proposed laser crystal 15, it should be understood
`that.
`in our actual measurements for the purpose of
`comparison with d c-heated plasmas, crystal l5 and 17
`were replaced by our spectrometer and the plate 16
`was oriented so that the laser energy was incident
`thereon at an angle of 45° with respect to the normal
`to the plate.
`In addition, the spectrometer was posi-
`tioned to observe the plasma radiation at an angle of
`45° from thc normal to plate 16 and thereby at an angle
`of 90° with respect to the incident laser radiation.
`In contrast with most studies of optical breakdown.
`which use a spherical lens to form a localizes plasma,
`this investigation uses a 7.5 cm focal length cylindrical
`lens 24 (10mm X 0.4 mm focal region) to focus the
`laser beam onto a 1 mm wide aluminum target 16 en-
`closed in a cell 2] filled with xenon gas. An approxi-
`mately cylindrically—shaped xenon plasma ~| cm long
`and 3—4 mm diameter was formed directly in front of
`the target 16. The presence of a target reduced the
`breakdown threshold of the gas to a level such that sig-
`nificant plasma radiation was produced in the gas for
`laser intensities as low as 107 W/cm2. Aluminum was
`chosen as the target material for reasons of maximum
`reproducibility and minimum vaporization for the laser
`intensities used.
`
`For higher input laser intensities or for different tar-
`gets, target vaporization can advantageously be pro-
`duced deliberately; and the vapor can yield the radia—
`tive plasma.
`The 10.6 um laser pulse was provided by a one Joule
`double-discharge TEA C0, laser 18. The output from
`this laser consisted of a 50 nsec gain-switched spike
`containing one-third of the energy and a 500 nsec pulse
`tail containing two—thirds of the energy. The laser out—
`put energy was incident on the target at an angle of 45°
`with respect to the target normal. Measurements of the
`angular dependence of the spectral radiance showed
`that the measurements taken at 45° were similar to
`those taken at other angles and indicated the produc-
`tion of a relatively uniform plasma.
`For the d-c heated plasma, an EG&G FX38 flash-
`lamp (not shown) (length ~7.5 cm between electrodes
`and bore diameter ~4 mm) was chosen in order to pro-
`vide approximately the same diameter plasma as that
`produced by the focused laser radiation. The flashlamp
`current was provided by a low inductance circuit that
`included a 0.02 uF capacitor charged to voltages up to
`30 KV and a triggered spark gap. A 2500 amp, 0.5 nsec
`duration current pulse was obtained at 30 KV charging
`voltage (current density ~20,000 amps/cm?)
`The flashlamp output consisted of a 0.5 nsec dura-
`tion pulse, except at the longer wavelengths (>400
`nm) for which this pulse was followed by one or more
`subsidiary pulses. Radiation from the laser-produced
`plasma occurred with a pulse duration 1.0 to 1.3 nsec,
`for wavelengths longer than 300 nm. And, in the region
`between 250 nm to 300 nm, the pulse duration was 0.5
`nsec. In xenon, the risetime for radiation from the las-
`er—produced plasma is as short as 30—40 nsec in the
`250—300 nm spectral region.
`Absolute intensities (averaged over the width of the
`plasma) were determined by first viewing a standard
`lamp through welLdefined optics and a grating mono-
`chromator. and then replacing the standard by the plas-
`mas. using a calibrated set of filters to determine inten—
`sity ratios. The standard lamp was a tungsten ribbon
`lamp calibrated against an NBS standard over the
`wavelength range from 250 to 500 nm.
`In FIG. IA. the spectral radiance as a function of
`wavelength produced by 0.8 Joule/cm input energy is
`shown for a laser-produced plasma in xenon at 50 Torr
`(near
`the optimum value for
`the laser-produced
`plasma) and for the flashlamp filled with xenon at 400
`Torr. (This typical fill pressure for commercial lamps
`is a compromise between high output intensity and long
`lamp life). Between 200 nm and 250 nm. the absolute
`value of the spectral radiance is somewhat less certain
`since it was obtained by extrapolation from the cali-
`brated value at 250 nm using the known response
`curves of the grating and detector. The wavelength re-
`gion from 200 nm to 500 nm was measured at 2.5 nm
`intervals except at line peaks where an accuracy of 0.5
`nm was obtained. Although the laser-produced plasma
`and the flashlamp provided similar intensities, in units
`of power, for wavelengths longer than 350 nm, the in—
`tensity. in units of power. of the laser—produced plasma
`was much higher in the 200 nm to 300 nm spectral re-
`gion. FIG. 1, in curves 11 and 12. clearly illustrates that
`the radiation from the laser-produced plasma exhibits
`more pronounced line structure. All of these lines can
`be attributed to either singly or double ionized xenon.
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`55
`
`6t)
`
`65
`
`

`

`3 300,803
`
`6
`
`5
`Also shown in FIG. 1A are curves 25—29 for spectral
`radiance of a blackbody radiator calculated for a few
`representative temperatures labeled on the curves. Ex-
`cept for the pronounced dip near 350 nm and the
`prominent line structure over the entire spectral
`re—
`gion, the flashlamp yields the same spectral radiance as
`would a I0,000°K blackbody. In contrast, in the region
`below 300 nm the spectral radiance from the laser—
`produced plasma is more nearly reproduced by a black-
`body at I2.000°K.
`In the flashlamp. the capacitor energy is mostly dissi—
`pated in ohmic heating, and only enough ionization oc-
`curs to sustain the discharge. The peak electron density
`was estimated (from voltage and current measure-
`ments) to be 10‘7 cm‘3 for an input energy of 0.8 Jou-
`les/cm. In the laser-produced plasma, the laser energy
`is absorbed by the inverse bremsstrahlung process, re—
`sulting in an electron density only limited by the critical
`density at which the plasma becomes highly reflecting
`(2 X l019 cm“’). Since significant energy at 10.6 am is
`not absorbed by the inverse bremsstrahlung process
`until a plasma density of ~10” cm‘3 is reached. the
`electron density and. hence. the xenon ion density is
`possibly as much as an order of magnitude higher in the
`laser-produced plasma than in the d c heated plasma,
`This higher ion density is consistent with the domi—
`nance of emission lines from ionized xenon in the radia—
`tion from the laser-produced plasma.
`The variation of spectral radiance at a constant input
`energy (0.8 Joules/cm) and at a fixed wavelength
`(282.5 nm) as a function of xenon pressure is shown in
`curves 31 and 32 in FIG. 3. Differences in the radiation
`characteristics of the laser—produced plasma (curve 32)
`and the d-c-heated plasma (curve 3]) are clearly illus—
`trated. If the spectral radiance from both kinds of plas-
`mas is proportional to the rate of energy absorption
`(the loss mechanism for both plasmas should be similar
`for similar pressures), then the energy absorption for
`the laser—produced plasma is greatest at 50 Torr,
`whereas, in the dc heated plasma, this rate maximizes
`at a much higher pressure.
`In the laser-produced
`plasma (at 106 am), the input energy is expected to be
`predominantly absorbed by electrons during electron-
`ion collisions (at the electron densities under consider-
`ation) at a rate which is proportional to the electron-
`ion collision frequency, 12.5,. In the d-c case, the energy
`is absorbed between these collisions at a rate which is
`proportional to l/vm. In general V3, is proportional to
`n,./Tg"’2 (n). is the electron density and T,, is the electron
`temperature.) In the dc heated arcs, T), is known to be
`relatively high at low pressures. decreases at some in—
`termediate pressure (due to thermalization with ions)
`and then rises again at high pressures. If such a varia-
`tion of T,. with pressure holds for both plasmas, then it
`explains the observed maximum of VE, at 50 Torr in the
`xenon laser—produced plasma (where Te presumably is
`a minimum) and the corresponding minimum at a simi-
`lar pressure for the d-c heated plasma (provided that n,.
`does not exhibit a strong pressure dependence over the
`pressures studied). Experimental confirmation of this
`explanation will await measurements of the electron
`density and temperature of the plasma. We do not wish
`our invention to be limited to this explanation.
`The variation of spectral radiance at a constant pres—
`sure and for a fixed wavelength (282.5 nm) as a func—
`tion of the input energy is shown in curves 4] and 42
`in FIG. 4.
`
`The flashlamp output intensity increases linearly with
`increasing input energy up to the maximum input en-
`ergy of 1.2 Joules/cm. This energy (9 Joules into the
`flashlamp in a 0.5 usec duration pulse) approaches the
`manufacturer's maximum recommended power input
`to the lamp. The intensity from the laser-produced
`plasma in xenon gas at 50 Torr increases nonlinearly
`with increasing input energy over most of the range of
`energy available. The output from the laser-produced
`plasma equals that of the electrical flashlamp at 0.35
`Joules/cm input and is nearly a factor of four greater
`than the flashlamp at 0.8 Joules/cm input (correspond-
`ing to 8 X l07 W/cmz.) While this behavior is character—
`istic of emission in the 200—300 nm region, the slope of
`the curve 42 describing the laser-produced plasma de—
`creases for longer wavelengths (not shown) and be»
`comes approximately equal to that (curve 41 ) describ-
`ing the d—c heated plasma in the 450—500 nm region.
`This comparison indicates the stronger heating effect
`of the laser at higher laser intensities and results in
`much higher peak—power emission than the d-c heated
`plasma at shorter wavelengths. A similar increase in ra‘
`diation from a laser-produced plasma with input energy
`has previously been observed at a wavelength of |20
`nm using a 1.06 pm neodymium laser focused onto a
`tantalum target. The results in FIG. 4 strongly suggest
`that the use of higher laser input intensities will pro-
`duce much higher peak power radiation in the UV and
`vacuum UV.
`
`The foregoing data shows that the spectral radiance
`of a cylindrically-shaped laser-produced plasma is
`greater in the ultraviolet than a geometrically similar
`d—c heated plasma for equivalent input energies/length.
`Such a cylindrically—shaped plasma, as a source is
`readily adapted to optical pumping of a crystal such as
`KCI crystal IS in FIG. 2A and 28. since an extended
`source is more easily matched to an elongated region
`(for gain enhancement) than a point source.
`The specific arrangement shown in FIGS. 2A and 2B
`also eliminates the radiation absorbing envelope that
`surrounds typical flashlamps. produces much shorter
`risetimes, and therefore could more easily pump short—
`lived, high energy excited states in gases. liquids and
`solids.
`Furthermore, the recent development of large aper-
`ture TEA lasers in efficient gaseous molecular laser
`media should make possible the generations of cylindri-
`cal plasmas having much larger dimensions, higher in-
`tensities and possibly much shorter risetimes than those
`produced in this study. Such extensions would make
`optical pumping of lasers with radiation from these
`plasmas not only possible but also relatively efficient.
`The feasibility and the practical utilization of our in-
`vention in pumping lasers has been actually demon—
`strated with the specific arrangement of FIGS. 5 and 6.
`In FIGS. 5 and 6 an iodine photodissociation laser is
`optically pumped by ultraviolet radiation from a laser-
`produced plasma in xenon gas initiated by a 10.6 mi-
`crometer transverse excitation atmospheric pressure
`(TEA) C02 laser 53. The power output of the iodine
`laser at l .3 l5 micrometers, which oscillated well above
`threshold when pumped in this manner, was found to
`be a factor of 5 greater than the power output resulting
`from linear flashlamp pumping.
`In FIG. 5 either CF31 or C3F7I is contained in a quartz
`tube 51 inside of an aluminum collecting shield 52 as
`best seen in cross section in FIG. 6. The 10.6 microme-
`
`LII
`
`IO
`
`30
`
`4O
`
`45
`
`55
`
`65
`
`

`

`
`
`7
`
`3 ,900 ,803
`
`8
`
`ter laser beam from C02 laser 53 as above described
`was focused by a cylindrical lens 54 through the win-
`dow 55 of a vacuum chamber 56 so that it just passes
`through the opening 57 of the aluminum collecting
`shield 52. Again, xenon gas is circulated within the vac—
`uum chamber 56 so that a xenon plasma occurs inside
`aluminum shield 52 in a volume extending from just in—
`side opening 57 to the proximity of the back wall of
`shield 52 facing opening 57. The long dimension of the
`cylindrical plasma volume is along the line of sight, or-
`thogonal to the paper. The radiation from the plasma
`was detected by a detecting system including the mirror
`57, lens 58 and monochrometer 59 including a photo-
`tube 60. The output of phototube 60 is displayed on
`scope 6|.
`The laser resonator for the 1.315 micrometer iodine
`laser is formed by laser mirror 62 and 63 at the oppo-
`site ends of tube 51 which has Brewster’s-angle end
`windows 64 and 65.
`The additional structural complication of the appara-
`tus of FIG. 5 is attributable to the experimental appara-
`tus needed to compare the output of the laser to that
`obtained with flashlamp pumping. In other words, the
`comparison is made to the laser output obtained when
`the optical pumping means is a flashlamp 66 and the
`laser active medium is enclosed partially in an alumi-
`num collecting shield 72 identical to shield 52. Shield
`72 also encloses tube 51, which is the same quartz tube
`filled with the same amount of CFal or Cal-7,1 as was
`used within shield 52. The laser path lengths for both
`lasers are comparable. In other words. the illustrative
`structures enable the only variable to be the pumping
`source.
`
`Notice that a major difference of the pump sources
`in FIG. 5 is that the laser-produced plasma is unim-
`peded by any tube walls, such as the flashlamp 66 has.
`so that the laser-produced plasma pumping technique
`according to our invention achieves the additional ad-
`vantage attributable to lack of such walls and the elimi-
`nation of the optical loss caused by them. In other
`words, as shown in FIG. 6 inside shield 52 there is only
`one tube, the tube 51, in which the pump medium is
`contained. In contrast. inside shield 72 there are two
`tubes 66 and 51 and, thereby, an additional attenuation
`for the d-c-heated plasma in tube 66. The laser radia-
`tion is detected by 1.315 micrometer detector 73 and
`may also be displayed on scope 61 to show its time rela-
`tionship to the radiation from either pumping source
`within shield 52 or within shield 72. When source 66 is
`used mirror 57 is reoriented so that monochrometer 59
`may monitor its radiation.
`While prior workers have established photodisso—
`ciation of CF31 or C3F7l by ultraviolet light in the 2500
`A. to 3,000 A. spectral region to produce a laser, none
`of these prior lasers employed the radiation from a las—
`er—produced plasma.
`In particular, in our experiment, an EG&G FX38C
`linear flashlamp 66 was mounted adjacent to the laser
`tube 51 as shown in FIG. 5 within collecting shield 72.
`The light output from both the laser—produced plasma
`and the flashlamp occurred within similar regions of
`their respective collecting shields. To provide a reason-
`able comparison with the laserproduced plasma, the
`flashlamp was masked-off to a 3 cm length.
`low-
`The flashlamp was excited by a 0.02 [1.1:
`inductance capacitor (not shown) charged to voltages
`up to 30 kV. This allowed up to 9 joules of energy to
`
`be dissipated in the 7.5 cm long, 4 mm flashlamp or
`~3.6 joules within the 3 cm optical pumping length of
`the unmasked region of the flashlamp. The peak cur-
`rent within the flashlamp was 2470 amperes with a
`pulse width of 0.8 usec, resulting in a peak current den—
`sity of 19,600 amperes/cmz.
`The 4 mm bore quartz laser tube 51 shown in FIG.
`5 was 30 cm long and was enclosed with quartz Brew-
`ster's angle windows on each end. High reflectivity di—
`electric mirrors 62 and 63 were used to form a 45 cm
`optical cavity for the 1.315 um laser light. The laser
`produced plasma and the flashlamp were positioned
`symmetrically within the optical cavity 62, 63 to pump
`equivalent regions within the mode volume of the cav-
`lty.
`FIG. 8 shows oscilloscope traces of the iodine laser
`pulse 81 and the UV pumping pulse 82 produced by the
`laser produced plasma.
`In FIG. 9, curves 91 and 92 show the like pulses for
`the flashlamp.
`In FIG. 8, the lower trace 82 shows the light output
`from the laser produced plasma at 2700 A. produced
`by a 2.75 joule C02 laser pulse having a duration of
`~50nsec (gain—switched spike) and 1 usec (pulse tail).
`The upper trace 8] shows the resulting iodine laser out-
`put.
`In FIG. 9, the lower trace 92 shows the flashlamp out-
`put at 2700A. The upper trace 91 shows the weaker io-
`dine laser output from the flashlamp 66. The intensity
`scale for this trace is five times smaller than the corre-
`sponding trace in FIG. 8. The intensity from the flash-
`lamp cannot be directly compared to that of the laser-
`produced plasma in this experimental arrangement be-
`cause of difficulties in obtaining equivalent
`light-
`collecting geometries.
`FIG. 7 shows the variation of laser output with C3F-,I
`pressure in the laser tube with an optimum at 175 Torr
`(Curve 71). This optimum is consistent with previous
`measurements made by Hohla for a 4 mm bore, flash-
`lamp—pumped C3F7I iodine laser. The optimum is deter—
`mined by the absorption length for the UV light within
`the active mode volume of the laser tube. Curve 72 of
`FIG. 7 also shows, for the laser-produced plasma the
`variation of laser output with xenon pressure with an
`optimum at 50 Torr. This is consistent with the fact that
`the UV peak power output from the plasma in the
`2500A. — 3000A. spectral region (corresponding to the
`absorption band of C3F7l) is a maximum at this pres—
`sure, as explained above.
`It may also be advantageous to produce the radiative
`plasma by directly focusing the radiation of the pump-
`ing laser on the body to be pumped through the mate-
`rial to be ionized, rather than focusing it on an adjacent
`target. The risetime of plasma radiation nearest the re-
`gion of initial ionization is shortest and is needed to
`pump lasers with very short upper state lifetimes.
`If a potential laser medium possesses a very high gain
`or a very short lifetime or has a population inversion
`between two energy levels that would correspond to a
`very short wavelength, than traveling-wave excitation
`would be highly advantageous. The advantage will par-
`ticularly be strong if the build up of radiation from the
`laser produced plasma otherwise occurs along the path
`of travel of the laser light to be stimulated in a time
`short compared to the travel time of the light to be
`stimulated, that is, in a time short compared to the ratio
`
`10
`
`IS
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`

`

`9
`of the length of the laser medium divided by the veloc-
`ity of light therein.
`Traveling-wave excitation would inhibit uncontrolled
`stimulated emission of radiation in directions other
`than the one desired, it would permit extraction of use-
`ful energy from laser systems with the very short life—
`times, and it would make the presence of the stimulated
`emission of radiation more apparent
`in very short
`wavelengths where suitable laser resonators do not now
`exist.
`As show in FIG. 10, a laser-produced plasma pump-
`ing source can easily provide traveling-wave optical ex-
`citation of a laser gain medium 120. This objective is
`achieved in marked contrast to a flashlamp which is not
`as readily adapted to such a purpose. In FIG. 10 it is
`merely necessary to orient the direction of the beam of
`the ionizing input laser 118, illustratively a 10.6 mi-
`crometer carbon dioxide molecular laser, so that its
`beam propagates at an angle 0 less than 90° with re-
`spect to the axis of the laser gain medium 120. Source
`118 is adapted to provide its beam with a substantial
`width illuminating the entire length of a cylindrical lens
`[24 parallel to laser gain medium 120. Lens 124 fo—
`cuses the light passing through it to a focus along a line
`parallel to the axis of medium 120 and ionizes the
`xenon gas in a cylindrical volume that parallels medium
`120. The laser light arrives first at the left-hand end of
`the intended plasma region because of the shorter
`travel distance of the light propagating along that edge
`of the beam from laser 118. The laser light at the oppo-
`site edge of the beam from laser 118 arrives last at the
`focus