X-Ray Lasing: The Novelle Although the Novette laser system served primarily as a test bed for advanced concepts of targets and for the system, its flexibility enabled us to use it for research that led to the world's first demonstration of x-ray lasing.

ver the past decade, we have harmonic conversion for producing For further information contact built a series of ever more light of higher frequencies (that is, Kenneth R. Manes (415) 423-6207. O powerful and complex laser shorter wavelengths). With Novette, systems Ganus, Argus, Shiva, and now we demonstrated a system for Nova) devoted to research on inertial routinely generating powerful beams confinement fusion (ICF) and weapon of green (0.53-pm) and ultraviolet physics. However, the Novette laser (0.35- and 0.26-pm) light, greatly Fig. 1 system (Fig. 1) did not fit neatly into broadening the range of feasible Artist's rendering of the business end of this series. Although it shared physics experiments. We confirmed the Novette laser system, showing the target characteristics of the other lasers, such that laser- coupling, and chamber, the harmonic-conversion arrays, and the extensive diagnostic instrumentation. as master-oscillator power-amplifier therefore the performance of the rCF The laser amplifier chains are out of the (MOPA) architecture, ­ capsules, is greatly enhanced if green picture, folded trombone style to fit into the glass amplifiers, and a 1.053-pm or ultraviolet light is used instead of available space. fundamental wavelength, in many ways Novette was considerably less complex than its predecessor, the . The primary function of the Novette laser was to serve as a test bed for the Nova laser system, and for this neither extreme energy nor great complexity was required. Once the Nova design was proved, the Novette laser would be dismantled and its parts incorporated into the Nova system. In its short two-year life, however, Novette was very productive. It enabled us to conduct one series of experiments that greatly enlarged our understanding of laser­ plasma coupling and another that led to the first-ever demonstration of soft-x-ray lasing. Novette was the first LLNL laser system to incorporate full-power

15 infrared light. Using light of shorter amplifier stages with increasingly wavelengths, we investigated in detail larger apertures. The spatial filters the mechanisms by which laser light minimize local intensity fluctations interacts with a plasma; we also were and keep the maximum laser flux able to produce the conditions below optical-damage limits. The peak necessary to demonstrate x-ray lasing. output flux from each amplifier stage is near its operating limit. By ovette Design maximizing the efficiency with which The Novette laser consisted each amplifier's aperture is used N of two identical power (the fill factor), we minimize the peak amplifier chains, fed by a single laser flux that each optical component master oscillator, and terminating at a must tolerate to produce a given common focal point. Figure 2 shows output. the layout of one of the amplifier The major design contraint on chains and indicates the generic origin the maximum energy delivered by of the various components. The beam each amplifier stage is the damage leaves the master oscillator with a threshold of its optical components, diameter of about 1.5 mm and which in turn depends on the purity emerges from the amplifier chain at of the optical medium, the optical an intensity of about 2 GW/cm2 and surface, the pulse duration, and the enlarged to a diameter of 74 cm. wavelength of the laser light. To raise At the master oscillator, the laser the damage threshold in the Novette beam is spatially Gaussian. As laser, optical components, including the beam proceedes through the spatial-filter lenses, windows, and preamplifier, it is expanded and then target-focusing lenses, are treated passed through an aperture that by a neutral-solution process that allows only the smoothest central produces an antireflection surface. l region to exit the preamplifier and Such treated optics display excellent enter each of the power-amplifier transmission (near 100%) and have arms. In each arm, seven beam­ surface-damage thresholds expanding spatial filters separate approaching those of the bulk material (12 to 16 J/cm2 for 1-ns pulses). 9.4 Janus The components most vulnerable size to laser-induced damage are the input lenses to the last two spatial filters. Argus·Shiva The neutral-solution antireflection size treatment on these surfaces survived unharmed after being subjected to Novette­ 74cm more than 100 shots of 1-ns duration {''''~il\ /\ /\ 1\ 1\ 11\1 1\ 1\ 1\. Nova 2 .J L.J with average fluxes of 5 to 7 J/cm size 31.S-cm disk amplifiers 46-cm disk amplifiers and peak fluxes twice as large. In fact, the practical flux . limit for these lenses Box is now bulk damage (due to ~ amplifier microscopic metallic particles in the glass) rather than surface damage. t>

16 INERTIAL FUSION

birefringent material. With Novette, however, it usually takes considerably we demonstrated that we could scale more than two hours to set up and up this process to the high beam align the target and the necessary powers and large apertures required diagnostic instruments. for the Nova system. Figure 4 is a chart of the high­ In doing so, we had to overcome energy laser shots made with Novette. the problem presented by the Those in blue-gray were diagnostic- enormous size of the Novette optics. We needed to convert beams 74 cm in Fig. 3 diameter, but the largest KDP crystals An array of potassium dihydrogen phosphate available were less than 30 cm across (KDP) crystals for converting infrared laser and it takes a year or more to grow light to visible (green) light. By placing another KDP crystals even to this size. similar array in series with this one, it is Therefore, we pieced the available possible to produce light in the ultraviolet region. smaller crystals together into 74-cm­ diameter arrays; we built 2S-element arrays of IS-cm crystals and 9-element arrays of 27-cm crystals (Fig. 3). To ensure that the crystals are held in proper alignment, they and their supporting framework are sandwiched between flat windows (treated by the neutral-solution process for maximum transmission). A selective optical filter is placed directly after the KDP array to pass green light but stop any unconverted infrared light. Handling and focusing the green light is relatively straightforward because we can use the same optical I I materials and coating technologies July I I developed for infrared ligh t. I I Ultraviolet light, however, poses a Fig. 4 I I number of problems since most Diagram showing April I I optical glasses have low damage how the rate of use thresholds, high nonlinear refractive I 1 of the Novette laser indices, and a tendency to darken I I improved with prac­ (solarize) in this spectral range. We January I I tice. The blue-gray bars represent laser­ I I have found solutions to these system diagnostic problems for the Nova laser and I I calibrations and have proved them out in Novette October 1 1 laser-related experi­ experiments. However, because they I I ments, the gray bars could not be implemented rapidly represent x-ray laser I I experiments, and the enough to meet the schedule for our M July blue bars represent co t=J x-ray laser studies, we were restricted a> I 1 all other plasma­ to green light for these experiments. I I physics and ICF ex­ periments. The April I 1 steady improvement I I ser Operation in shot rate in the I I x-ray laser experi­ It takes about two hours to ments was made January I I cool Novette's amplifiers after possible by the use t N co I of a standard target a shot and to realign the laser arms a> for the next shot. Theoreticall y then, I configuration and a coherent, well­ I I I I I we should be able to fire eight shots planned program for in a 16-hour working day, or 160 o 10 20 30 40 50 60 exploring the param- shots each month. In practice, Number of shots eter space.

17 system calibrations and laser-related incorporated in the Nova laser. experiments, those in gray were These included modifications to the x-ray laser experiments, and those in amplifiers, especially the larger ones, blue were all other plasma-physics insertion of an apodizer plate to experiments. The striking feature of protect the frequency conversion this chart is the high shot rate attained array, monitoring and specification in the x-ray laser campaign, in of beam parameters with detailed contrast to the performance in the computer modeling, and the other plasma-physics experiments. introduction of astigmatism to create This was possible because, in the the elongated focal spot required for x-ray laser experiments, we were able the x-ray laser experiments. to use a single target configuration Several of the design changes we and a standard arrangement of were testing on the Novette system diagnostic instruments for all the involved the box amplifiers, which shots. As a result, everyone connected use flashlamp energy more efficiently with the experiments and the laser than do the Shiva-type cylindrical operation was able to settle into an amplifiers. Each box amplifier (Fig. 5) efficient routine. Extrapolating the contains two or three elliptical disks of learning curve in the final x-ray laser neodymium-glass set at Brewster's series, one can predict that in angle to the beam axis (the angle at prolonged, routine operation with a which there is zero reflection). standard target configuration, we Flashlamps line the two sides of should be able to make more than the box that view the faces of the 60 shots per month. amplifier disks, but not the top and bottom (which see the disks edge on). erformance All four sides of the box are lined With the Novette laser, we with silver-plated reflectors. Pwere testing a number of In the past, amplified spontaneous design innovations that would be emission (ASE) has limited the size

Fig. 5 One of the largest (46-cm aperture) box amplifiers during assembly. The two neodymium-glass disks are set at Brewster's angle to eliminate unwanted reflections from their surfaces and are elliptical to present a circular aperture to the beam. The dark vertical band in each disk absorbs unwanted laser light travelling lengthwise through the disks (out of the beam) to improve the amplifier's efficiency by suppressing a principal loss mechanism.

18 INERTIAL FUSION

of disk amplifiers. ASE wastes the conversion array, for a 10-TW pulse. energy stored in the disk by Apart from the vertical shadow of amplifying light traveling in unwanted the absorbing strips in the largest directions. Since amplification is amplifier disks, the beam was very exponential, ASE is worst along the uniform. MALAPROP calculations major axis of the disk and is especially enabled us to determine that severe if the disk edges can reflect operation with 1-ns pulses was safe light back into the disk. In previous at 10 TW but not at 12 TW. Above amplifiers, we clad the edges of the 10 TW, operation was unacceptably disks with a material that absorbs risky because of the possibility of stray laser light without reflection. In small-scale, self-focusing effects that the large sizes needed for Novette and could cause damaging hot spots. Nova, we carried this idea one step The x-ray laser experiments further by inserting a strip of similar required yet a further innovation. material across the middle of each Most Novette target designs called disk (Fig. 5). Our experiments with for a circular focal spot for the laser Novette proved that this solves the beams measuring 0.2 to 1 mm in problem of ASE propagating along diameter. However, the x-ray laser the major axis, virtually removing the experiments require a long and limit on amplifier size. We should narrow focal spot, 10 mm long by now be able to make lasers as large as 0.1 mm wide. To achieve this, we necessary for any desired application. introduced astigmatism. By adding a In another design innovation tested weak cylindrical lens after the final on Novette, we placed an apodizer focusing lens in each beam, we can plate directly in front of the harmonic produce the desired horizontal line conversion system to protect the KDP array. In an unprotected converter (a) array, the harsh contrast between the 70 Fig. 6 clear KDP crystals and the opaque Uniformity of the output infrared beam as "egg-crate" frames supporting the 60 measured just before the KDP crystal array. crystals would cause severe diffraction (a) Beam photograph showing the distribution ripples on the otherwise uniform 50 of intensity variations and the vertical shadow E caused by the absorbing strips in the amplifier beam. To prevent this, we insert CJ ai 40 disks. (b) Densitometer scan along the cross­ CJ an apodizer plate with bead-blasted c: hatched strip in (a) showing the amplitudes of bands to shadow the edges of the ~ 30 the intensity ripples. crystals and soften the contrast, thus is smoothing the beam and preventing 20 potentially damaging hot spots. 10 Although this technique costs 10 to 15% of the beam area, it protects the o KDP crystals and the final focusing (b) lenses. 4r------, Computer modeling of beam propagation through the laser chains 3 contributed greatly to our success N E with Novette, in particular to the CJ achievement of very high beam ~ CJ uniformity. With the MALAPROP ;!- 2 ·iii code/ we modeled all the components c: of the laser system and included such ! effects as diffraction, nonlinear -= changes in refractive index with beam intensity, and dust motes or damage sites on surfaces. In Fig. 6, we show OL-~-~-~~~--L-~~~ a typical image of the infrared beam, o 10 20 30 40 50 60 70 just before it enters the frequency Distance, cm

19 focus several centimetres ahead of the about 200 ps long. By the end of our former focus. These cylindrical lenses x-ray laser study, we were using were available only in glass, not in pulses up to 500 ps long. The average fused quartz with its higher damage intensities we used ranged from threshold. Because of the inability of about 10 TW/ cm2 to several times these glass lenses to pass ultraviolet 100 TW/ cm2 at the target. light without suffering severe damage, the x-ray laser experiments were ummary limited to green light. In its short life, the Novette Another requirement for the Slaser was an outstanding x-ray laser experiments was that success. In routine operation, this laser the elongated focal spot be precisely system achieved as many as six full­ aligned with both the x-ray laser power shots in a single day, more target's axis and with the various than achieved by any other large diagnostic instruments. We developed laser. As a test bed for the Nova laser, special techniques to achieve the Novette provided the first operational necessary precise alignments. In experience with: addition, although the line focus • Segmented laser-amplifier disks, presented to the x-ray laser target was which can now be scaled up to almost neither uniform nor diffraction limited arbitrarily large apertures. (Fig. 7), the target had been designed • Neutral-solution antireflection to tolerate large intensity variations. coatings on laser optics, which have At the beginning of our x-ray laser more than tripled the nominal damage experiments, we set the master threshold for 1-ns pulses (now 2 oscillator to produce short pulses 15 J/ cm ). • Nova-style digital control and (a) diagnostic techniques (coupled with optical-fiber communication links). 500 • Full-power harmonic conversion with large-aperture arrays of KDP 400 crystals. [ In addition to completing the ICF ff 300 ~----' experiments planned for it, the c: Novette laser enabled us to perform ~ C 200 target experiments that greatly clarified the physics of laser-plasma 100 coupling and that led to the world's first demonstration of x-ray lasing. ~

100 200 300 400 (b) 0.25 r------, Key Words: inertial confinemen t fusion (lCF); laser-Argus, Janus, Nova, Novette, Shiva, x- ray. 0.20 i-

NE Fig. 7 ~ 0.15 i- ll. The x-ray laser targets were irradiated by a ;!o line focus 0.1 mm wide and 10 mm long, pre­ . ~ 0. 10 - Notes and References cisely aligned with the x-ray laser and its as­ S 1. G. R. Wirtenson, N. J. Brown. and L. M. sociated diagnostics. (a) A small segment of a .E Cook, "Scaling Up the Neu tral Solution line focus showing the local intensity varia­ 0.05 - Antire fl ecti on Process," Optica l Eng. 22 (4). tions tolerated by the x-ray laser targets. 450 (1983). (b) Intensity as a function of position across 2. W. W. Simmons, J. T. Hunt, and W. E. the line focus, showing that the laser power O~_~I ~~I ~I --~I --~I ~ Warren, " Light Propagati on Through Large was concentrated within a band 0. 1 mm wide. o 100 200 300 400 500 Laser Systems," j. Quantum Elect. QE-17 (1 PW equals 1015 W.) Distal1ce, 11m (9), 172 (1981).

20