Crane et al.

Vol. 13, No. 1 / January 1996 / J. Opt. Soc. Am. B

89

Strong anomalous emission from He-like and H-like Ne in short-pulse laser-driven plasmas
J. K. Crane, H. Nguyen, S. C. Wilks, T. Ditmire, C. A. Coverdale, T. E. Glover, M. D. Perry, and Y. Zakharenkov
Lawrence Livermore National Laboratory, L-493, P.O. Box 808, Livermore, California 94550 Received March 15, 1995; revised manuscript received June 12, 1995 We observe strong emission from the 1s 2 1snp Rydberg series in He-like Ne and from the Lyman-a transition in H-like Ne. These emissions are observed when 1.05-mm light in a 600-fs laser pulse is focused into the dense, localized output of a pulsed supersonic nozzle. The maximum focal irradiance of our laser was measured at full power in vacuum to be 1.1 3 1018 W cm2 . Although emissions from lower charge states such as Ne6+ and Ne7+ closely follow rates predicted by tunneling theory, emissions from Ne8+ and Ne9+ are observed at irradiances 2 orders of magnitude below tunneling theory estimates. We discuss the origins of these anomalously high charge states and the implications to the development of short-pulse, ultrahighbrightness x-ray sources. 1996 Optical Society of America

1.

INTRODUCTION

Much of the research on short-pulse laser-driven plasmas in gas targets has been motivated by interest in recombination and collisionally excited x-ray lasers.1 3 In the case of recombination-pumped x-ray lasers, very low electron temperatures of a few electron volts at relatively high electron densities ( 1020 cm3 ) are required for producing adequate gain.1 As a result of these requirements several authors1,4 6 have studied the feasibility of producing these conditions to determine the ideal laser specifications and have concluded that short wavelengths, 248 nm, and short pulses, 100 fs, must be used to minimize rapid electron heating that is due to a variety of mechanisms. In addition to the interest in x-ray lasers, there is a more general interest in the interaction of short-pulse high-irradiance lasers with different forms of matter, spanning many orders of magnitude in density from collisionless, low-density gas to solid state. This new, broad arena of high-field physics is driven by several areas of application, including high-brightness x-ray sources for medical imaging and lithography,7,8 high-gradient electron and plasma accelerators,9,10 and advanced schemes for inertial confinement fusion.11 Of particular interest is the density range from 1018 to 1021 cm3 (1021 cm3 is critical density for l 1 mm). In this density regime there is a transition from purely field-driven phenomena such as above-threshold ionization (ATI) and tunneling ionization to collective, plasma effects such as stimulated Raman scattering (SRS) and stimulated Brillouin scattering. These collective plasma effects are responsible for producing energetic particles or photons that fuel interest in the several potential applications. In the past year there have been several reports12 14 of anomalous x-ray emission from the interaction of subpicosecond laser pulses with the dense gas output of a pulsed gas jet. In these experiments the laser irradiance is typically greater than 1017 W cm2 , and gas densities vary from 1018 to 1021 cm3 . Coverdale et al.15 recently reported million-electron-volt electrons correlated with the observation of SRS in the forward direc0740-3224/96/010089-06$06.00

tion in a He plasma produced by a 600-fs, l 1 mm laser focused at the output of a gas jet. Blyth et al.14 used Thomson scattering to measure electron temperatures in similar plasmas excited by a 248-nm 350-fs pulse from a KrF laser. They also observed a strong Raman signal, indicating that a significant portion of the incident laser energy is coupled into the plasma wave. The plasma wave heats the electrons, which can excite charge states that are inaccessible by tunneling ionization at these laser irradiances. In a series of papers McPherson and co-workers16 report anomalous x-ray emission from laserirradiated rare-gas clusters produced in a gas jet under conditions similar to those described above. The large clusters (several hundred atoms) are formed from one of the heavier rare gases, Ar, Kr, or Xe, and the anomalous x-ray emission comes from inner-shell transitions excited by the coherently driven electrons stripped from the clusters. In this paper we report bright, anomalous x-ray emission from a Ne plasma driven by a 600-fs 1.05-mm laser pulse that is linearly polarized. We observe a strong correlation between these x-ray emissions and the exponential growth of SRS. Particle-in-cell calculations based on our experimental conditions indicate that electron temperatures of several thousand electron volts can be generated in the hot tail of the electron distribution. Applying these electron temperatures to a rate-equation model, we simulate the experimentally observed rapid and efficient emission from these transitions.

2.

DESCRIPTION OF THE EXPERIMENT

Our experiments were performed with a Nd:glass laser system that produces a 600-fs pulse at 1.05-mm with as much as 8 J of energy.17 We made a thorough investigation of the output characteristics of the laser to aid in interpreting the results of our experiments with both solid and gas phase targets. Normal shot diagnostics include the near-field image and the frequency spectrum at the laser output. We routinely take autocorrelations of the laser pulse to determine its duration. We measure
1996 Optical Society of America

90

J. Opt. Soc. Am. B / Vol. 13, No. 1 / January 1996

Crane et al.

the far-field image of the beam at full power, which enabled us to determine accurately the peak irradiance of the focused laser in vacuum. The far-field image of the beam at the focus of our f 8.6 lens in vacuum shows a slightly elliptical beam (ratio of axes 1.5) with a Gaussian intensity profile. B-integral effects in the air path to the experiment and through the final focusing lens cause 35% of the energy to be scattered away from the central spot at 6 J, resulting in a peak irradiance at the focus of 1.1 3 1018 W cm2 . The full width at half-maximum of the focal spot is 16 mm 3 24 mm. The 1.05-mm light used in our experiments is linearly polarized in the horizontal orientation. Figure 1 shows the experimental arrangement for the laser plasma experiments. The laser pulse is focused into a large vacuum vessel by an f 8.6 aspheric lens. At the center of the vacuum vessel is a pulsed valve equipped with a supersonic nozzle that can produce Mach 8 flow.18 The laser is focused approximately a millimeter below the nozzle orifice where particle densities are 1019 atoms cm3 .19 Emission from the laser-driven plasma is collected with a grazing-incidence extreme ultraviolet spectrometer, whereupon the light is dispersed and is reimaged onto the detector. The spectrometer detector combination can cover a wavelength range spanning 5 1000 A by use of different diffraction gratings. For time-integrated experiments we use a microchannel-plate detector coupled to a visible-light CCD array. For time-resolved measurements we employ an x-ray streak camera with resolution up to 10 ps. Visible light that is scattered back in the direction of the incoming laser light is collected by the input focusing lens and is sent to a 0.25-m spectrometer equipped with a 16bit CCD camera. This diagnostic is used to determine the electron density in the plasma by measurement of the Raman shift of the scattered laser light,19 or it is used to measure the quantity of light that is backscattered to determine the Raman heating in the plasma. Finally, using two different techniques, we seek information about the laser pulse as it focuses and propagates through the plasma. We split a small fraction of the incoming laser light into a second optical leg that contains a variable delay to overlap the probe pulse with the laser pulse in time. The probe pulse is sent through the plasma perpendicular to the laser, and the scattered light is collected with a small CCD camera. This arrangement allows us to perform shadowgraphy on the plasma at various times after the laser pulse and to look for refraction or filamentation of the pulse as it propagates through the plasma. We can also monitor the central portion of the laser beam as it emerges from the x-ray spectrometer in zero order. By reimaging this beam onto a camera we can compare the beam that propagates through the plasma with the same beam in vacuum to gain qualitative information about the presence of significant filamentation or other extreme refraction effects. We observed no signs of beam filamentation or selffocusing with the techniques described above. In these measurements the plasma length was 1.0 mm, which is comparable with the confocal parameter of the focused laser, and the plasma density was increased to 0.2ncrit . In addition to propagation effects we measured Mie scattering20,21 in an identical gas jet, using a 100-fs

Cr-doped LiSAF laser that was frequency doubled to produce light at 400 nm. We observed no scattering with Ne or He but significant scattering with Ar and Kr in the gas jet. This result implies that Ne clusters are ,50 A.

3.

X-RAY EMISSION

Figure 2(a) shows a time-integrated x-ray spectrum in Ne at a laser irradiance of 1 3 1017 W cm2 and a particle density of 2 3 1018 atoms cm3 . This spectrum shows several strong series of lines in the Be-like and Li-like Ne

Fig. 1.

Experimental layout.

Fig. 2. (a) Spectrum showing line emission from Ne VII and Ne VIII transitions (Be-like and Li-like). Laser irradiance I 1.0 3 1017 W cm2 , electron density ne 1.5 3 1019 electrons cm3 . (b) Spectrum showing line emission from Ne IX and Ne X transitions (He-like and H-like). I 5.0 3 1017 W cm2 .

Crane et al.

Vol. 13, No. 1 / January 1996 / J. Opt. Soc. Am. B

91

ions (Ne VII and Ne VIII transitions), including the 2s np, 2p nd, and the 2p ns series in Ne VIII. The 74- and 98-A lines in Ne VIII, which stand out distinctly in this spectrum, have been suggested as possible candidates for recombination-pumped laser schemes.1 At higher particle densities and irradiances most of the line emissions shown in Fig. 2(a) become sufficiently bright to saturate the microchannel-plate CCD detector. Figure 2(b), a similar spectrum taken at a higher laser power, shows a shorter-wavelength region. The prominent line at 13.4 A is the He-a line and heads the series of 1s2 np transi tions in Ne IX that converge to the series limit at 10.37 A. In addition, we can see the Lyman-a line of Ne X at 12.1 A, which is relatively weak in this spectrum, where the laser irradiance was 5 3 1017 W cm2 , but this line increases to saturate the detector at higher laser power. Using the x-ray streak camera, we studied the time behavior of these x-ray transitions. In Fig. 3 we show the time decay of the 13.4-A transition, Ne IX 1s2 1s2p, and the 66-A transition, Ne VIII 1s2 2p 1s2 5d. At the operating speed used to take these images the camera has a resolution of 20 ps. The data show that the 13.4-A emission is short-lived, comparable in decay time with the camera resolution, whereas the emission from the Ne VIII transition (as well as other transitions from this ionization stage) decays at a rate of 1010 s1 , indicating that the upper level continues to be populated long after the passage of the laser pulse. Earlier 22 we compared emission decay rates for the Ne VIII transitions excited by green light (l 0.526 mm) with a simple recombination model. The measured decay rates agreed with three-body recombination rates for temperatures of a few electron volts. In the data shown here the Ne VIII emission increases in intensity from t 0 (coincident with the laser pulse), reaching a peak 1 ns after the laser pulse, and then decays with a time constant of 1 ns, consistent with our earlier measurements. If recombination is populating the upper level, then we would expect the rate to increase and reach a maximum only after the electron temperature has cooled from its initial elevated temperature at the end of the laser pulse. In contrast, the Ne IX emission peak nearly coincides with the laser pulse, which suggests that sources other than recombination, such as direct collisional excitation, are responsible for pumping the more energetic He-like and H-like levels. All the Ne IX and the Ne X transitions shown in Fig. 2(b) are anomalous in the sense that they appear at values of laser irradiance that are more than 2 orders of magnitude below the appearance intensity predicted by Ammosov et al.23 (ADK) or barrier-suppression24 models for tunneling ionization; consequently they must originate from a collisional mechanism that is capable of heating these plasmas on time scales comparable with the laser pulse duration. In this paper we consider mechanisms that may be responsible for the excitation of these anomalous He-like and H-like Ne lines in short-pulse laserdriven plasmas.

that evolves as the short, intense incoming laser pulse rapidly ionizes a dense gas. We routinely employ this backscattered Raman signal as a technique for measuring the electron density in the interaction region of our gas jet.19 Figure 4(a) is a low-intensity backscattered spectrum in He showing the laser light and a weaker, Stokes peak that is red shifted from the laser frequency by an amount equal to the plasma frequency, vp v0 2 vs . From this frequency shift we derive a value for the electron density, ne vp e 2 me0 , where m and e are the electron mass and charge, respectively, and e0 is the vacuum permittivity. The plasma wave, the incoming electromagnetic wave, and the scattered Raman wave can couple together parametrically. In this case the Stokes wave (at vs ) can beat with the incoming laser wave v0 at the plasma frev0 2 vs , to drive the plasma wave resoquency, vp nantly, which in turn feeds back to drive the Raman wave. The amount of Raman signal grows exponentially with laser irradiance until large fractions of the incoming laser light are backscattered. Figure 4(b) shows the Raman signal scattered from the Ne plasma at different laser intensities. As the laser intensity is increased the Raman signal goes from a single narrow peak to a broad, extended spectrum of light. In some cases we observe that the backscattered light extends even to the blue side of the laser wavelength, an effect indicative of strongly coupled Raman scattering.26

5.

DISCUSSION

Various authors4 6 have considered different heating mechanisms and their relative importance in ultrashortpulse laser-driven plasmas. Rae and Burnett5 modeled ATI and collisional heating (also known as inverse bremsstrahlung) under conditions relevant to recombination-pumped x-ray lasers. They report electron temperatures of a kilovolt for a 1.0-mm 100-fs laser (I 2.5 3 1017 W cm2 ) exciting a plasma in Ne at an atom

4. BACKWARD-SCATTERED RAMAN EMISSION

Several authors14,15,25 have observed that large quantities of light are scattered backward from the plasma wave

Fig. 3. Time-resolved emission from the Ne IX 1s 2 1s2p tran sition at 13.4 A (heavier curve) and the Ne VIII 1s 2 2p 1s2 5d transition at 66 A (lighter curve).

92

J. Opt. Soc. Am. B / Vol. 13, No. 1 / January 1996

Crane et al.

level. The growth rate g0 in the region of normal growth is proportional to the quiver velocity of the electrons, which is a function of the laser irradiance I: p vosc 25.6l I W cm2 cm s, and l is the laser wavelength in micrometers. The rate of increase of the Raman Is and plasma Ip waves from noise is given by6 Is Ip Is0 exp kg0 tp , Ip0 exp kg0 tp , (1a) (1b)

where k is a constant that depends on the temporal pulse shape, k 1 2, and g0 is given by g0 1 2 vosc c vp v0 2 vp
1/2

(2)

Fig. 4. (a) Backward-scattered Raman emission in He at low intensity. ne 4 3 1018 electrons cm3 . (b) The Raman signal in Ne grows exponentially with laser irradiance. I 8.0 3 1017 W cm2 (filled squares), I 4.40 3 1017 W cm2 (lighter curve), I 2.6 3 1017 W cm2 (darker curve).

density of 2.5 3 1018 cm3 . Penetrante and Bardsley4 also studied ATI and collisional heating in the context of recombination-pumped x-ray lasers in Ne. They showed how, through space-charge effects, one can reduce ATI by choosing a laser pulse width, tpulse , that meets the condition tpulse vp p 2. They predicted that an electron temperature of 400 eV would be produced by a 100-fs 1.0-mm laser with an irradiance of 1018 W cm2 in Ne plasma with an ion density of 1019 cm3 . Recent papers by Blyth et al.14 and Coverdale et al.15 give experimental evidence for the dominant role that SRS plays in producing very hot electrons in short-pulse laser-driven plasmas. Coverdale et al.15 observed a correlation between the production of 2-MeV electrons and forward-scattered anti-Stokes radiation in He plasmas produced from a gas jet (identical to the one used in these experiments) operating at densities of 1019 cm3 . Blyth et al.14 also observed K-shell emission from Ne, which they attributed to excitation by electrons heated by SRS. Wilks et al.6 showed that large electron temperatures can be produced from collective plasma heating by SRS, even with 100-fs laser pulses at 248 nm. Both the backscattered electromagnetic wave Is and the plasma wave Ip can be expressed in terms of an exponential growth starting from a background-noise

As the growth rate continues to increase with laser intensity it eventually surpasses the plasma frequency, g0 . vp , the Raman wave can become strongly coupled,25 and p the growth rate is given by g 3 vp 2 v0 vosc 2 16c2 1/3 . Therefore, for normal growth, the rate is proportional to I 1/2 , whereas for strongly coupled Raman g I 1/3 . We made simultaneous measurements of x-ray emission and backscattered energy as a function of laser irradiance at a fixed atom density in the gas jet. Figure 5 plots the integrated Raman signal and fluorescence yields on the 13.4-A transition versus laser irradiance. For both sets of signals we observe a steep rate of rise with laser power to a point where both signals start to roll over and saturate at a value near 5 3 1017 W cm2 . In Fig. 6 we plot the same integrated Raman signal versus the growth rate given by Eq. (2) for the measured electron density, ne 0.015 ncrit . At lower laser intensity the Raman signal shows rapid growth, as predicted by Eq. (1). At a point where the growth rate approaches the plasma frequency, the rate of increase slows and eventually saturates at the highest intensities. This change of slope at g vp indicates that the plasma instability has become strongly coupled. If we take a cylinder of volume 2p 2 w0 4 l filled with electrons at a density of 1.5 3 1019 electrons cm3 and calculate the energy per electron assuming that 2% of the 5-J input goes into plasma heating by means of SRS,27 we obtain a value of 14 keV per plasma electron. Treating SRS as the principal source of electron heating in our experiment, we model the x-ray emission for Ne VII Ne X and compare the results with data taken of line emission versus laser irradiance. We first estimate an electron temperature, using the particle-in-cell code ZOHAR28 for the conditions of our experiment. An example electron energy distribution for our parameters is shown in Fig. 7. For this one-dimensional simulation a laser pulse with a peak intensity of 5 3 1017 W cm2 and a full width at half-maximum of 600 fs was injected into a 0.01ncrit plasma with an initial electron tempera25 eV. Once the laser pulse passed through ture Te the 300-mm slab of plasma, the electron distribution was found to be that shown in Fig. 7. We assume that the distributions are Maxwellian, based on the rapid electron electron collision rate, 1013 s. The background electron temperature has heated up to roughly 6 keV, and a large heated tail component exceeds 10 keV. Because this simulation is one spatial dimension and three

Crane et al.

Vol. 13, No. 1 / January 1996 / J. Opt. Soc. Am. B

93

temperatures seen in one dimension, typically by factors of 2, because there would be more degrees of freedom in the problem. Nonetheless, this simulation illustrates the fact that substantial electron heating can occur as a result of the SRS backscatter and that this must be taken into account when one estimates x-ray yield in this parameter regime. In fact, when the laser traverses a long ( 1 mm) region of plasma, the Raman forward-scatter instability29 can generate energetic electrons. This prediction was experimentally corroborated by Coverdale et al.16 with the measurement of 2-MeV electrons in a similar laser-driven, gas jet plasma in He. However, these electrons are too energetic to contribute significantly to the excitation of He-like and H-like Ne excited states. Using the electron distribution function f and classical collisional ionization cross sections si , we determine

Fig. 5. (a) Fluorescence yield versus laser irradiance. ne 0.015ncrit . (b) Integrated Raman signal versus laser irradiance.

Fig. 7. Electron energy distribution from particle-in-cell calculations.

Fig. 6. Integrated Raman signal versus growth rate. 0.015ncrit .

ne

velocities, two- and three-dimensional effects are not included. These effects will reduce the effective heated

Fig. 8. X-ray emission versus laser irradiance: comparison with the rate-equation model. Triangles with white dots, Ne VII, 2s 3d transition at 106.1 A. Open boxes with xs, Ne VIII, 2s 5p transition at 60.8 A. Circles with white grids, Ne IX, 1s2 1s2p transition 13.4 A. Diamonds with white dots, Ne X, 1s 2p transition 12.1 A. The solid curves represent the predictions of the model that is described in the text.

94

J. Opt. Soc. Am. B / Vol. 13, No. 1 / January 1996

Crane et al.

rates that are used in a set of coupled rate equations for each of the ion states in Ne. These equations are of the form dni11 dt Wi,i11 ni 1 ni ncoll , (3)

the auspices of the U.S. Department of Energy by the Lawrence Livermore National Laboratory under contract W-7405-ENG-48.

Z i, Wi is the ADK tunwhere ni is the Ne ion with R neling rate, and ncoll ne df si v is the collisional ionization rate averaged over the distribution function. To obtain an integrated emission signal we integrate the equations for a time determined by the electron temperature decay rate from the particle-in-cell simulations. Figure 8 shows data and model results. The emission data have the usual corrections for spectrometer and detector efficiencies at the different wavelengths. In addition, we divided the integrated emission yields by the gf values of their respective transitions so that the plotted quantities would be proportional to an upper-level density, more in line with the model that calculates an ion density. To compare the model results with the data, we used a scale factor to overlap the two sets in the region in which the Ne VII data saturate. In this irradiance regime the ionization rate is the same with or without collisions. The comparison between data and model reveals good agreement for the Ne VIII and the Ne IX states, but emission from the Ne X state is higher than predicted. This lack of agreement may be due to the simplicity of our model. We are continuing to model the experimental results by using a more sophisticated hydrodynamic code that also includes details of the atomic-level populations.

REFERENCES AND NOTES

1. P. Amendt, D. C. Eder, and S. C. Wilks, Phys. Rev. Lett. 66, 2589 2592 (1991). 2. N. H. Burnett and P. B. Corkum, J. Opt. Soc. Am. B 6, 1195 1199 (1989). 3. B. E. Lemoff, C. P. J. Barty, and S. E. Harris, Opt. Lett. 19, 569 571 (1994). 4. B. M. Penetrante and J. N. Bardsley, Phys. Rev. A 43, 3100 3113 (1991). 5. S. C. Rae and K. Burnett, Phys. Rev. A 46, 2077 2083 (1992). 6. S. C. Wilks, W. L. Kruer, E. A. Williams, P. Amendt, and D. C. Eder, Phys. Plasmas 2, 274 279 (1995). 7. C. Tillman, B. Erlandson, K. Herrlin, C. Olsson, H. Petterson, G. Svahn, S. Svanberg, J. Larsson, A. Persson, and C. G. Wahlstrom, Phys. Scr. 49, 187 197 (1994). 8. I. Turcu, I. Ross, A. Hening, and M. Steyer, Microelectron. Eng. 21, 95 98 (1993). 9. T. Tajima and J. M. Dawson, Phys. Rev. Lett. 43, 267 270 (1979); J. Krall, A. Ting, E. Esarey, and P. Sprangle, Phys. Rev. E 48, 2157 2161 (1993). 10. C. Joshi, W. B. Mori, T. Katsouleas, J. M. Dawson, J. M. Kindel, and D. W. Forslund, Nature (London) 311, 525 529 (1984). 11. M. Taback, J. Hammer, M. E. Glinsky, W. L. Kruer, S. C. Wilks, J. Woodworth, E. M. Campbell, M. D. Perry, and R. J. Mason, Phys. Plasmas 1, 1626 1634 (1994). 12. P. Audebert, J. P. Giendre, J. C. Gauthier, F. Amiranoff, V. Malka, E. De Wispeleare, C. Stenz, R. Brukner, F. Blasco, A. Antonetti, Dos Santos, G. Rey, X-ray emission from an Ar gas jet induced by an intense femtosecond laser pulse, presented at the meeting on generation and application of ultrashort x-ray pulses, Salamanca, Spain, 1994. 13. J. K. Crane, T. Ditmire, H. Nguyen, and M. D. Perry, in 4th International Colloquium on X-Ray Lasers, D. C. Eder and D. L. Matthews, eds. (American Institute of Physics, Woodburg, N.Y., 1994), pp. 253 257. 14. W. J. Blyth, S. G. Preston, A. A. Offenberger, M. H. Key, J. S. Wark, Z. Najmudin, A. Modena, A. Djaoui, and A. E. Dangor, Phys. Rev. Lett. 74, 554 557 (1995). 15. C. A. Coverdale, C. B. Darrow, C. D. Decker, W. B. Mori, K.-C. Tzeng, K. A. Marsh, C. E. Clayton, and C. Joshi, Phys. Rev. Lett. 74, 4659 4662 (1995). 16. A. McPherson, T. S. Luk, B. D. Thompson, A. B. Borisov, O. B. Shiryaev, X. Chen, K. Boyer, and C. K. Rhodes, Phys. Rev. Lett. 72, 1810 1813 (1994); A. McPherson, B. D. Thompson, A. B. Borisov, K. Boyer, and C. K. Rhodes, Nature (London) 370, 631 634 (1994). 17. F. G. Patterson, M. D. Perry, and J. T. Hunt, J. Opt. Soc. B 8, 2384 2391 (1991). 18. J. G. Pronko, D. Kohler, I. V. Chapman, T. T. Bardin, P. C. Filbert, and J. D. Hawley, Rev. Sci. Instrum. 64, 1744 1747 (1993). 19. M. D. Perry, C. Darrow, C. Coverdale, and J. K. Crane, Opt. Lett. 17, 523 525 (1992). 20. T. Ditmire, T. Donnelly, R. W. Falcone, and M. D. Perry, Strong X-ray emission from hot plasmas produced by intense irradiation of clusters, Phys. Rev. Lett. (to be published). 21. O. Abraham, S. S. Kim, and G. D. Stein, J. Chem. Phys. 75, 402 411 (1981). 22. J. K. Crane, M. D. Perry, D. Strickland, S. Herman, and R. W. Falcone, IEEE Trans. Plasma Sci. 21, 82 89 (1993). 23. M. V. Ammosov, N. B. Delone, and V. P. Krainov, Sov. Phys. JETP 64, 1191 1194 (1986). 24. S. Augst, D. D. Meyerhofer, D. Strickland, and S. L. Chin, J. Opt. Soc. Am. B 8, 858 867 (1991). 25. W. P. Leemans, C. E. Clayton, W. B. Mori, K. A. Marsh, P. K. Kaw, A. Dyson, and C. Joshi, Phys. Rev. A 46, 1091 1105 (1992).

6.

SUMMARY

We observe strong anomalous emission from He-like and H-like transitions observed in an intense, subpicosecond laser-driven Ne plasma produced in a gas jet. These emissions are observed at irradiances that are 2 orders of magnitude below the appearance intensities of Ne9+ and Ne10+ , as predicted by tunneling theory. In the same experiments we see intense SRS signals in the backward direction from the incoming laser pulse. These SRS signals grow exponentially with laser irradiance and finally saturate at the highest values. Particle-in-cell simulations of SRS under our experimental conditions predict electron temperatures of 6 keV. We used this electron distribution function to calculate an average collisional ionization rate and included this term in a rate-equation model to predict formation rates for Ne7 10+ . The model produced good qualitative agreement with the data, underestimating the production rate for Ne10+ . Although other mechanisms such as inverse bremsstrahlung and ATI undoubtedly play a role in heating the electrons, the strong correlation between the SRS backscattered light and the anomalous x-ray emission observed in the Ne plasmas produced in our experiments indicates that SRS heating dominates under these conditions (l 1.053 mm, I 2 3 1017 1018 W cm2 , and ne . 0.01ncrit ).

ACKNOWLEDGMENTS
Discussions with C. B. Darrow, D. Eder, P. Amendt, and M. H. Key were useful in the preparation of the manuscript of this paper. This research was performed under

Crane et al. 26. C. B. Darrow, C. Coverdale, M. D. Perry, W. B. Mori, C. Clayton, K. Marsh, and C. Joshi, Phys. Rev. Lett. 69, 442 445 (1992). 27. Coverdale et al.15 measured 2% reflected energy in the backscattered Raman wave in a He plasma under similar conditions to those reported here. Blyth et al.15 also report a maximum scattered fraction of 2% from a Ne plasma.

Vol. 13, No. 1 / January 1996 / J. Opt. Soc. Am. B

95

28. A. B. Langdon and B. F. Lasinski, in Methods in Computational Physics, J. Killeen, ed. (Academic, New York, 1976), Vol. 16, p. 327. 29. W. B. Mori, C. D. Decker, D. E. Hinkel, and T. Katsouleas, Phy. Rev. Lett. 72, 1482 1485 (1994).