JOURNAL DE PHYSIQUE,

Colloque C6, supplement au na 10, Tome 47, octobre 1986

RECENT PROGRESS I N SOFT X-RAY LASER DEVELOPMENT AT PRINCETON

S. SUCKEWER, C.H. SKINNER, D. KIM, E. VALEO, D. VOORHEES and A. WOUTERS

Plasma Physics Laboratory, Princeton University, princeton, NJ 08544, U.S.A.

Abstract - Recent advances in research on soft X-ray lasers at Princeton are described. A one-dimensional code has been constructed which is in good agreement with the measured radial dependence of soft X-ray gain at 182 A in a magnetically confined recombining plasma. Multichannel detectors have been installed in the diagnostic spectrometers and spectra of the line emission in axial and transverse directions are presented. Initial measurements of the relative divergence and, very recently, absolute divergence measurements of the axial 182 beam have been made by scanning the axial spectrometer across the beam.The absolute divergence was measured to be in the range 5-10 mrad, depending on experimental conditions and the maximum power of soft X-ray beam was 100 kW. Finally, a new two laser approach to create gain at wavelengths below 100 A is briefly described.

I.

INTRODUCTION

Progress in the development of soft X-ray lasers has been recently reported by several 1aboratories.l At Princeton, an approach hased on a magnetically confined recombining plasma column, cooled by radiation losses, has generated amplification of stimulated emission of a 100 (a one pass gain length of kll%6.5) .2 In this experiment a commercially available 1 kJ TEA COg laser (duration 8n nsec, maximum gain was obtained at an energy 300 J) was incident on a carbon disc target in a strong (90 kG) magnetic field. Rapid recomhination, after the laser pulse, created a population inversion between levels 3 and 2 in hydrogen-like carbon, (cvI). Installation of a soft X-ray mirror3 in a double pass arrangement provided an additional demonstration of the amplification of stimulated emission. With a measured normal incidence reflectivity of 12% at 182 1, a 120% increase in axial stimulated emission In was ob~erved.~ work with axially oriented, thick (35~)carbon fiber targets coated with a t&in layer of aluminum, gains of up to 6 cm-l were also generated on the CVI 182 A transition4. In this article, we will focus on recent progress in the theoretical interpretation of our earlier results, measuremepts of the relatively small divergence of the soft X-ray laser beam at 182 A with power %I00 kW, and on the development of a system (two-laser approach) for the study of conditions for lasing action significantly below In0 8 .
11. COMPARISON OF GAIN MEASUREMENTS WITA A ONE DIMENSIONAL MODEL

A key element in the achievement of relatively large amplification of stimulated emission was the realization, from measurements of the radial profiles of the CVI line radiation, that the most favorable conditions for maximum gain

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1986603

JOURNAL DE PHYSIQUE

should exist in the off-axis regions of the plasma column. Gain was measured hy recording the amplification (enhancement), E, of the axial CVI 182 k emission; that is the ratio of the 182 stimulated plus spontaneous emission in the axial direction to the mostly spontaneous 182 t emission in the transverse f direction. The enhancement is related to the one pass gain, G, by:

By varying the position of the laser focus on the carbon disc target with resDect to the observation volume of the axial and transverse monochromators it was possible to measure the gain as a function of radius r in the plasma column. The results of one such scan were presented earlier.5 Although in that experiment the C 0 2 laser energy was not optimal, there was a rapid rise of the enhancement in the region r = 1.5 to 2.5 mm off axis. Further experiments2 with optimal. plasma conditions led to measurements of an amplification of E %l00 near r = 1.3 mm. A one-dimensional hydrodynamic plus atomic physics model has been developed to aid understanding these result^.^ In this model a single mean flow velocity was used to describe the ion mass motion. After solution of time-dependent equations for the ion density, momentum and electron energy, the gain was calculated by a post processor code from the electron density and temperature and the number density of the ground state populations of fully stripped and hydrogen like carbon. Because the laser pulse length is longer than the compressional Alfven transit time, radial pressure balance is quickly established in the plasma. Strong heating on the cylindrical axis of symmetry (at the laser focus) leads to a centrally peaked temperature profile with a corresponding electron density minimum. On the other hand, off axis, strong radiative cooling by CIV leads to low temperature, high density conditions conducive to a fast recombination rate and high 182 gain. With the introduction of an ion diffusion rate an order of magnitude greater than the classical value, totally stripped ions were transported from the center to the cold off axis region where fast recombinati.on generated high gain.

y~~
\
kmox

r\

Corbon disc 8. Mognetic Field Radiation Cooling 8. Diffusion -

At Time Of

200

7 ;

_r

3-

\ -

'

.z 2 w

\ \ \

kmox

Fa
W
[ L

\ '\\

Diffusion a + Coefficient [L W D =2.5~1~~cm~/sec 2 -too 2

I-

\
OO I 0.5

I 1.0 1.5 2.0 PLASMA RADIUS (mm)

2.5

Fig. 1

Radial profiles of CVI 182 % gain, k , , , and electron temperature, Te, versus radius in the plasma column as predicted by a lD code.

Figure I shows the predicted gain versus radius and it can be seen that high gain occurs in a narrow %50-100p wide annulus at a radius of 1.4 mm. This is in excellent agreement with the experimental results.

111.

AXIAL AND TRANSVERSE EMISSION SPECTRA

Recently, multichannel detectors, based on microchannel/reticon arrays, were installed in the axial and transverse soft X-ray spectrometers. These oermit the recording of emission spectra in the axial and transverse directions in a single laser shot. One example is shown in Fig. 2. In the transverse spectrum, the spontaneous CVI 182 A emission is weak compared to the strongest lineoin the spectrum, OVI 173 L. However in the axial direction, the stimulated 182 A emission dominates the spectrum. Some exciting results ohtained very recently were made by scanning the ~osition of the axial spectrometer in the transverse direction (perpendicular to the axis of the plasma column). The axial emissions are imaged by a grazing incidence mirror onto the entrance slit of the axial spectrometer (see Ref. 2 for experimental arrangement). The mirror is constructed by bending a glass strip therefore the optical quality of the system is not ideal. Hence a transverse scan of the axial spectrometer gives information of the relative divergence of the stimulatedo 182 A emission in comparison to non-lasing lines. As shown in Fig. 3 the 182 A emission was so intense on axis that the detector was saturated whereas the CVI 186 I? and OVI 173 A lines remained weak with a flat spatial profile. The axial spectra presented in Fig. 2 were ohtained at a position of the spectrometer corresponding to x = 2 5 0 ~in Fig. 3. This horizontal position of spectrometer was established for all earlier measurements2 by alignment of the system i.e. C02 laser focusing mirror, target slot, and the spectrometer entrance slit using a He-Ne laser beam. However, from Fig. 3 we can see that maximum gain is near x = 200u with a corresponding amplification of 182 1 radiation ahout factor of 5 larger than presented earlier2. The sensitivity of axial instrument for the data of Fig. 3 was higher than in Fig. 2 in order to measure intensity changes of the non-lasing lines.

Fig. 2

Transverse and axial spectra in the region near 182 A from a carhon disc target with four carbon blades. The laser energy was 500 J.

JOURNAL DE PHYSIQUE

To obtain information ahout the absolute divergence of the soft X-ray laser beam, the grazing incidence bent mirror was removed. In its place a 1 mm wide and 10 mm high collimating slit was installed to block reflections from the walls of the vacuum chamber. Figs. 4 and 5 show the result of a shot-hy-shot horizontal scan of both the collimating slit and the axial soft X-ray spectrometer. The principle of the experiment is presented schematically on the right-hand side of Fig. 4, and on the left-hand side is shown the intensity distribution of the CVI 182 1, CV 186 1, and OVI 173 lines for a magnetic field R = 20 kG. For every shot the intensities of all three lines were recorded simultaneously on the multichannel detector of the axial soft X-ray spectrometer. One may see that the lasing line, CVI 182 W, is strongly peaked on axis with a F W H M % ~ . cm at distance 304 cm from plasma (target). ~ This corresponds to a horizontal divergence of the beam ofo%9 mrad. At the same time the intensities of the non-lasing lines OVI 173 A and CV 186 % are quite constant over the scan region %3 cm. With increasing magnetic field (B = 35 kG and 50 kG) we observed further narrowing of soft X-ray laser beam down to a 5 mrad (Fig. 5 . ) This indicates that with a higher magnetic field maximum gain is created in a more narrow plasma region (less than 50p transversely). Refraction of the 182 emission in the plasma is negligible as the electron density (ne (L l ~ l ~ c m - is) too . ~ low and thus an estimate of the divergence of the stimulated emission in single pass amplification can be obtained from ray tracing. With an annular width of 50p as predicted by the one-dimensional code and a plasma length of 1 cm, the angular range of rays that pass through the gain region is 10 mrad. With a peak in the gain profile at a particular radius the divergence will be even less and thus the 1D calculation provide a good understanding of the remarkably low value of the measured divergence. We also observe a slight shift of the peak intensity of the 182 1 radiation at higher magnetic fields, which may be caused by a small tilt of the magnet at high currents. Another interesting feature is the rise of the 182 8 line intensity near the geometrical limits of scan indicated by open circles and triangles in Fig. 5. The limits are determined by diameter of vacuum tube hetween the target and the soft X-ray spectrometer. In order to decrease the effect of shot-to-shot line intensity fluctuations, the 182 % radiation was normalized in Fig. 5 to OVI 173 8 line intensity (173 if line intensity was quite uniform across the scanning distance, as can be seen in Fig. 4 . Our ) initial interpretation was that this rising intensity of 182 k radiation near the edges of the scan was due to diffraction effects on the edges of target slot. However a more likely explanation may just be the reflection of the 182 1 radiation (grazing-incidence reflection) inside the vacuum tube. In the future, we plan to reinstall the grazing-incidence mirror and relocate the axial soft X-ray spectrometer with the entrance slit in the focal plane of the bent mirror in order to measure the divergence of 182 if radiation in the far-field. Knowledge of divergence allowed us to estimate the total power of soft X-ray beam 100 kW from measurements of the intensity of the 182 1 radiation (a1-3 mJ) and pulse duratipn ($10-30 nsec).
IV. TWO LASER APPROACH TO X-RAY LASER DEVELOPMENT BELOW 100

We consider the lasing actiogs presently achieved by Livermore and Princeton in the spectral region 100-200 A, to be the beginning of the rogd toward the development of lasers in the more important spectral region -10 A. Therefore, as a next step we are proceeding with a program for the development of soft X-ray lasers at wavelengths below 100 i. We have begun constructing an experiment in which a 1.5 kJ CO2 laser generates a highly ionized magnetically confined plasma column in which a powerful (Id 1016 w/cm2) picosecond laser beam will produce a population inversion and gain. The role of the C02 laser is to provide access to the high energy, short wavelength transitions of high Z ions which are then excited by the picosecond laser via multi~hoton processes.

Fig. 3

Transverse scan of CVI 182 1, CV 186 1, and OVI 173 8 emission showing a strong central peak for the CVI 182 2 stimulated emission. In this figure, "Satur " indicates the level above which the detector is saturated. Intensities ahove this level were obtained hy comparing the non-saturated region of the 182 spectral profiles.

Scanning Axia l Spectrometer

Fig. 4
Entrance Slit

Slit 304cm

Absolute divergence measurement: (31,9 mrad) of the 182 A lasing radiation for a magnetic field B = 20 kG. For comparison are shown intensities of nonlasing lineg OVI 173 1 and CV 186 A, recorded simultaneously with 182 8 line.

Plasma

1
0 1 2 3 4 5 HORIZONTAL DISTANCE X (cm)

JOURNAL DE PHYSIQUE

Fig. 5

Element

Fig. 6

Wavelength o f the 4s24p45s2 - 4 ~ 2 4 ~ 5 5 s transition in the KrI isoelectronic sequence (from Ref. 8 ) .

The feasibility of exciting either one subvalence electron or two valence electrons of an ion through multiphoton processes7 was studied for the argon and krypton isoelectronic sequences excited by a powerful picosecond KrF* (2480 i) 1aserb8 The excitation of two valence electrons (e.g. 4s24p45s2 state) is especially attractive because of the faster progression to shorter wavelengths One potential lasing transition with increasin change of the target ion. ~ 4 ~ and 5 s would be 4 ~ 5s2 + 4 ~ ~ 4 ~ ~ Hartee-Fock values of the wavelength scaling of this transition in the Kr isoelectronic sequence are s h o q in Fig. 6 from Ref. 8. The potential lasing wavelength for e.g. cd12+ is 89 A. An additional advantage with the higher ionization stages is that competing processes such as photoionization or autoionization are reduced or eliminated. Multiphoton excitation is also expected to significantly increase the population inversion and gain at 182 in the current experiment. As shown in Fig. 7, the new experimental system incorporates two lasers; a 1.5 kJ COq laser to produce the ionized medium and a powerful picosecond laser to generate the population inversion. The target is placed in a solenoidal magnet which will radially confine the plasma and produce the long, thin geometry suitable for laser action. Primary diagnostics will consist of multichannel soft X-ray spectrometers and an X-ray streak camera. The powerful picosecond laser is expected to generate a 1 J, Ips, 2480 i laser pulse. The main oscillator is a YAG-laser pumped dye laser at 6470 1 which amplified in a three stage dye amplifier producing an energy output of a few mJ. This is then drequency doubled and mixed with the 1.06~ YAG-laser to produce a 1 ps, 2480 A pulse with an energy of a few hundred IJJ. In the final stage, two KrF* amplifiers increase the pulse energy to the joule level with an (unfocussed) power in the terawatt range. Focussing by a suitable lens will produce intensities in excess of 1016 w/cm2 and in addition to short wavelengths laser applications, will produce new insights into the interaction of radiation with ions in a regime where the laser field is comparable to the Coulomb field between the electrons and the nucleus.

TWO LASER APPROACH TO X-RAY LASER DEVELOPMENT SYSTEM DIAGRAM

#85X'5q3

Acquisition
System D.A.S. D.A.S.

Fig. 7

Diagram of new experimental setup "Two-Laser Approach" to X-ray laser development below 100 2.

JOURNAL DE PHYSIQUE

We would like to acknowledge support and encouragement from H. Furth and J.R. Thompson; assistance with the analysis of the multiphoton excitation scheme by C.W. Clark, M.G. Littman, T.J. McIlrath, and R. Miles; significant contributions by C. Keane, L. Meixler, C.H. Nam, J.L. Schwob, T. Srinivasan, and W. Tighe; and technical assistance from L. Guttadora and J. Robinson. This work was made possible by financial support by the U.S. Department of Energy Basic Energy Sciences, Contract No. KC-05-01, and the U.S. Air Force Office of Scientific Research, Contract No. AFOSR-86-0025.

See e.g. other papers in this Proceeding. S. Suckewer, C.H. Skinner, H. Milchberg, C. Keane, and D. Voorhees, Phys. Rev. Lett. 55, 1753 (1985). T.W. Barbee, Jr., S. Mrowka, and M.C. Hettrick, Appl. Opt. 6, (1985). 883 H. Milchberg C.H. Skinner, S. Suckewer, and D. Voorhees, Appl. Phys. Lett. 47, 1151 (1985). S. ~uck&er, C.H. Skinner, H. Milchberg, C. Keane, and D, Voorhees, PPPL Report 2207 (March 1985); also Proc. Int. Conf. "Laser 85", Las Vegas, NA (~ec.1985). E.J. Valeo, C. Keane, and R M . . Kulsrud, Bull. Am. Phys. Soc. 30, 1600 (1985); also S. Suckewer et.al.; Proc. Int. Conf. "Laser 85", Las Vegas, NA (Dec. 1985). Rhodes, T.S. Luk, H. Pummer, K. Boyer, M. Shahadi, H. Egger, and C.K. Phys. Rev. Lett., 2, 110 (1983). C.W. Clark, M.G. Littman, R. Miles, T.J. McIlrath, C.H. Skinner, S. Suckewer, and E. Valeo, J. Opt. Soc. Am. March 1986.