Diode-pumped Nd:YAG laser with 38 W average powerand user-selectable, flat-in-time subnanosecond pulses

John Honig,1,* John Halpin,1 Don Browning,1 John Crane,1 Richard Hackel,1 Mark Henesian,1

John Peterson,1 Doug Ravizza,1 Tim Wennberg,1 Harry Rieger,2 and John Marciante3

1Lawrence Livermore National Laboratory, P.O. Box 808, L-462, Livermore, California 94550, USA2JMAR Research, 10905 Technology Place, San Diego, California 92127, USA3Laboratory for Laser Energetics and Institute of Optics, 250 East River Road,

Rochester, New York 14623, USA

*Corresponding author: honig1@llnl.gov

Received 5 January 2007; accepted 16 February 2007;posted 1 March 2007 (Doc. ID 77651); published 15 May 2007

A diode-pumped injection-seeded Nd:YAG laser system with an average output power of 38 W is de-scribed. The laser operates at 300 Hz with pulse energies up to 130 mJ. The temporal pulse shape isnominally flat in time and the pulse width is user selectable from 350 to 600 ps. In addition, the spatialprofile of the beam is near top hat with contrast �10%. © 2007 Optical Society of America

OCIS codes: 140.0140, 140.3280, 140.3480, 140.3530, 140.3580.

1. Introduction

High-peak power ��10 MW�, high-average power��10 W� laser operation with Nd:YAG is typicallyachieved by power oscillator [1,2,3,4,5] operationor by MOPA (master-oscillator power amplifier)[6,7,8,9,10,11] operation. In the case of q-switchedpower oscillators, pulse widths typically range from7–100 ns. The minimum pulse widths for q-switchedpower oscillators are limited by the required build-uptimes and the achieveable gain-to-loss ratio [12,13,14].Mode-locking, either as a power oscillator or as aMOPA, is another option and this technique can gen-erate pulses from 2–100 ps [15,16,17,18]. In this case,the minimum pulse width depends upon the recipro-cal of the gain linewidth. To obtain HPP (high-peakpower), HAP (high-average power) laser operationwith pulse widths in the 300 ps to 1 ns regime usingNd:YAG generally requires a MOPA [19].

HPP-HAP laser operation using MOPAs has beendemonstrated in Nd:YAG with pulse widths rang-ing from 7–100 ns. These lasers typically employ aq-switched or mode-locked laser as the master oscil-lator whose temporal characteristics ultimately de-termine the pulse width. Work by Schiemann et al.

[20] showed that HPP-HAP laser pulses can be ob-tained by pulse compression but the pulse-to-pulseenergy and temporal stability were less than idealdue to the stochastic nature of the compression pro-cess. The primary focus in most HPP-HAP lasers ispeak power, average power, beam quality, energy ef-ficiency or some combination of the above. The actualtemporal pulse shape is rarely of concern. The effectsof gain saturation in the amplifier modify the oscil-lator temporal pulse shape by steepening the leadingedge and sometimes reducing the overall pulse width.

In fusion-class lasers [21,22], the temporal pulseshape can be very important as it drives the dynamicsof the target interaction [23]. Fusion-class lasers arecurrently large beams ��1000 cm2� with very highpeak power but very low repetition rate. Optics dam-age due to these very high peak power lasers isalways of concern [24]. Studies [25–28] for develop-ing scaling laws relating laser fluence, pulse width,and pulse shape to damage probability exist but aredifficult to test since most lasers do not have well-defined temporal and spatial pulse shapes. A HHP-HAP laser that has a FIT (flat in time) pulse shapewith user-selectable pulse width as well as a spa-tially flat beam profile can be an excellent tool toraster scan optics and acquire large amounts ofstatistical data to test the scaling laws.

0003-6935/07/163269-07$15.00/0© 2007 Optical Society of America

1 June 2007 � Vol. 46, No. 16 � APPLIED OPTICS 3269

2. Experimental System

The MOPA laser system described in this paper con-sists of three major components, an oscillator, a re-generative amplifier, and a power amplifier.

A. Oscillator

The oscillator for this laser system is modeled afterthe NIF (National Ignition Facility) [21] and OMEGA[29] oscillators. It is an all-fiber-based system withcomponents as shown in Fig. 1. A 10-mW cw single-frequency fiber oscillator [30] is injected into an AO(acousto-optic) modulator that carves out a 30-nspulse. This 30-ns pulse is injected into a fiber ampli-fier [31] and amplified to �10 nJ. The 10-nJ pulse isinjected into a two-channel amplitude modulator asdescribed in Ref. [29]. One channel of the modulatoris driven by a PSPL (Picosecond Pulse Laboratory)pulse generator that creates a 0 to 10V square pulsewith user-selectable pulse width. The other channelis driven by an ACSL (aperture-coupled strip line)[29], that generates a 3V to 6V linear ramp voltageover 600 ps. The resultant light pulse exiting themodulator is a shaped pulse of �300 pJ as shown inFig. 2. The pulse shape can be varied by adjusting therelative timing between the PSPL pulse generatorand the ASCL.

B. Regenerative Amplifier

The regen (regenerative amplifier) for this laser isa simple modified-V-shaped oscillator cavity shown

schematically in Fig. 3. The regenerative cavity isrelatively long �1.5 m� and allows for easy injectionand extraction of pulses with widths up to 3 ns. Thegain is provided by a JMAR-designed [32] end-pumped Nd:YAG rod. The rod is 3 mm in dia., has again length of 20 mm and an atomic doping of 1%.In addition, the rod has a piece of un-doped YAG [33],(3-mm dia., 5-mm long) diffusion-bonded [34] to thediode-pump end. This un-doped piece serves two pur-poses: 1) it helps to homogenize the pump light; 2) ittakes some of the thermal load away from the di-chroic coating on the rod end, minimizing wavefrontaberrations. The pump light is provided by a 4-barSpectra Physics diode array and is focused into therod with a simple, spherical plano-convex lens.

The narrow V-shaped cavity allows the use of 0°angle-of-incidence mirrors and reduces the overallfootprint to fit on a single 30.5-cm by 122-cm bread-board. The 5-m radius of curvature end mirror makesthe cavity passively stable and, in conjunction withthe thermal lensing of the end-doped rod, places thebeam waist near the quarter wave-plate.

The 300-pJ signal oscillator pulse exits the single-mode fiber and is mode-matched to the regenerativeamplifier cavity using a single commercial asphericlens (focal length 8 mm). This expanded and nearlycollimated oscillator beam passes through a tandemFaraday isolator to provide more than 70 db isolation.The injected oscillator pulse is trapped in the regener-ative amplifier cavity using a FastPulse 5046E pulseslicer. After approximately 30 round trips, the 2.3-mJamplified oscillator pulse is switched out. A photo-graph of the oscillator fiber launch, tandem isolator,and regenerative amplifier on the breadboard is shownin Fig. 4.

C. Power Amplifier

The power amplifier for this laser is a modified JMARBrightLight four-pass amplifier, shown schemati-cally in Fig. 5.

The amplifier heads are diode-pumped Nd:YAGrods, 8-mm diameter, 72-mm long. The atomic dopingconcentration is 0.8%. Each head is pumped by five14-bar diode arrays arranged in pentagonal symme-try. A 90° quartz rotator and the clocking of the twoidentical amplifier heads at 36° with respect to eachother assist in birefringence compensation. With again of �2.1 per head, birefringence compensation iscrucial to prevent the amplifier from sponataneouslylasing.

The vacuum telescope has a single f � 408 mmplano-convex lens on each end. In combination withthe thermal lensing of each amplifier head �f ��1600 mm�, one obtains �1500 mm between relayplanes. This layout is similar to [35] except in thiscase the relay planes are not in the rods but at theend mirror. This arrangement allows us to maintainrelay for four passes instead of two. Each of the vac-uum telescope lenses is mounted on a bellows to per-mit fine-tuning of the relay. A 2-mm aperture at thefocus of the vacuum telescope eliminates parasiticoscillation and helps to reduce diffraction effects.

Fig. 1. (Color online) Schematic of the all-fiber-based masteroscillator.

Fig. 2. (Color online) Some temporal pulse shapes of the fiber-based oscillator are shown. The pulse shape can be varied byadjusting the relative timing between the pulse generator and theASCL. These data show timing adjustments of 0, 100 ps, 200 ps,and 300 ps.

3270 APPLIED OPTICS � Vol. 46, No. 16 � 1 June 2007

The beam from the regenerative amplifier is ex-panded and collimated using a pair of plano-convexlenses. A custom-made chrome-on-glass serratedaperture (140 teeth, 1.25-mm tooth length, 4.5-mm-dia. clear aperture) creates a beam with a roundedtophat spatial profile. The serrated aperture planebecomes the reference plane for the relay imaging.The reference plane is relay-imaged onto M2, back toM4, back again to M2, and finally onto a plane�15 cm to the right of input polarizing beam splitter,PB. The quarter wave plate (QWP) near M2 allowsthe user to select two-pass or four-pass amplification.We only discuss four-pass operation in this paper.The entire assembly fits on a single 30-cm by 140-cm

breadboard. A photograph of the power amplifier isshown in Fig. 6.

3. Results

The laser system normally runs at 300 Hz. The pumpdiodes (for the regenerative amplifier and for thepower amplifier) are pulsed for 200 �s. Near-fieldbeam images (at the output relay plane) and pulseshape measurements are monitored continuously.The pulse shape measurements are obtained using aNew Focus 1011 45-GHz detector into a Tektronix8200 sampling oscilloscope. The amplifier pulseshapes for 350, 400, 500, and 600 ps are shown in Fig.7 below. The FWHM (full width half maximum) is

Fig. 3. (Color online) Schematic layout of the regenerative amplifier. The regenerative amplifier consists of an end-diode-pumped Nd:YAGrod, a 0° incidence flat high reflector (HR), a polarizing beam-splitter cube (PB), a quarter wave plate (QWP), a quarter-wave voltagepulse-slicing Pockels cell (PC), and a 0° incidence high reflector with 5 m radius of curvature (R5). The input and output beams areseparated by a Faraday rotator (FR) half wave plate (HWP) polarizing beam splitter (PB) combination.

Fig. 4. (Color online) Photograph of the fiber launch, tandem isolator, and regenerative amplifier on the 30.5-cm by 122-cm breadboard.The beam path is drawn for reference.

Fig. 5. (Color online) Schematic of the power amplifier. Two identical diode-pumped Nd:YAG amp heads are separated by a vacuumtelescope that act to relay image the serrated aperture (SA) onto high reflector (M2), then onto high reflector (M4), back onto M2 and thenonto a plane to the right of the right-most polarizing beam splitter (PB). A quartz rotator (QR) assists in birefringence compensation anda quarter wave plate (QWP) allows four-pass operation of the amplifier. A Faraday isolator (FI) protects the regenerative amplifier anda Faraday rotator (FR) half wave plate (HWP) combination separates the input and output beams.

1 June 2007 � Vol. 46, No. 16 � APPLIED OPTICS 3271

used to define pulse width. The relatively long tails ofthe pulse shapes are artifacts of the detector and thelong cable length to the oscilloscope. In each case, theoutput energy of the amplifier is 128 mJ which, at300 Hz, corresponds to an average output power of�38 W. At each pulse width, the pulse-to-pulse en-ergy stability is �3%. Because of gain saturation inthe amplifier, the required pulse shapes from theoscillator and the regenerative amplifier are notFIT. Each successive amplification stage tends tosharpen the leading edge of the pulse. The pulseshapes from the oscillator and the regenerative am-plifier that result in FIT amplifier pulses are shownin Figs. 8 and 9. The output energy of the oscillator is400 pJ with pulse-to-pulse energy stability of �1%.The output energy of the regenerative amplifier is2.3 mJ with pulse-to-pulse energy stability of �2%.

The amplifier output beam image is continuouslymonitored using a Cohu 4800 camera and CoherentBeamview software. The camera is located at anequivalent output relay plane and a beam image canbe recorded on any shot. Output beam images for fourdifferent FIT pulse widths are shown in Fig. 10. Themethod for determining the statistics for a givenbeam image is discussed in Section 4.

4. Discussion

One of the goals of this laser system is to generateboth a FIT and a tophat spatial profile pulse. A math-ematical way of representing a FIT or tophat spatialprofile pulse is with a super-Gaussian shape, given byan equation of the form

f�t� � Ae�2��t�to�2

�2 �n

(1)

where A is an arbitrary constant, � is the Gaussianhalf width, to is the center location of the super-Gaussian, and n is the order of the super-Gaussian.The 500-ps pulse shape and a second-order super-Gaussian fit with FWHM � 500 ps is shown in Fig. 11.All four of the FIT pulses described in this paperare well approximated by a second-order super-Gaussian.

As we see in Fig. 8, the oscillator pulse shapesrequired for a FIT amplifier becomes less linear andmore exponential as the desired pulse width in-creases. This temporal curvature is obtained simplyby changing the relative timing between the PSPLpulse generator and the ASCL, as described above.No other adjustments are necessary.

The spatial profiles can be fitted by a 5th-ordersuper-Gaussian. Horizontal (x) and vertical (y) line-

Fig. 6. (Color online) Photograph of the power amplifier. The vacuum telescope and diode-pumped amplifiers are outlined.

Fig. 7. Amplifier output pulse shapes are shown as a function ofFWHM pulse width. The pulse shape amplitudes are scaled to 1and the pulse shapes are time shifted such that the 50% amplitudeof the leading edge corresponds to time � 0.

Fig. 8. (Color online) Oscillator pulse shapes that generate FITamplifier pulses are shown as a function of pulse width.

3272 APPLIED OPTICS � Vol. 46, No. 16 � 1 June 2007

out profiles taken through the 500-ps beam centroidand the corresponding super-Gaussian fits are shownin Fig. 12.

The beam statistics shown in Fig. 10 are calculatedas follows. First, the captured beam image is fit to atwo-dimensional super-Gaussian of the form

f�x, y� � Ae�2��x�xo�2

a2�

�y�yo�2

b2 �n

(2)

where A is an arbitrary constant; a, b are the Gauss-ian half widths in x and y, respectively; xo, yo repre-

sent the center location of the super-Gaussian in xand y, respectively; and n is the order of the super-Gaussian.

The binary beam image file has a one-to-one corre-spondence between relative fluence per pixel andpixel value. The 90%-amplitude ellipse of this super-Gaussian is defined as the ROI (region of interest).The normalized mean beam fluence is calculated bydetermining the average pixel value of all pixelswithin the ROI. At the same time, a histogram of allthe pixels within this ROI is generated. The beamcontrast is then defined as (the standard deviation ofall the pixels within the ROI) � (the mean pixelvalue). The 10% energy fraction is defined as (thenumber of pixels whose pixel value falls within �5%of the mean pixel value) � (the number of pixelswithin the ROI). Similarly, the 20% energy fraction isdefined as (the number of pixels whose pixel valuefalls within �10% of the mean pixel value) � (thenumber of pixels within the ROI). Finally, the peak-to-mean value is defined as (the maximum pixelvalue anywhere in the image) � (the mean pixelvalue within the ROI).

We see that as the FIT pulse width becomesshorter, the beam contrast and peak-to-mean ratiosincrease. While the beam contrast remains below10%, the beam images and beam statistics clearlyshow that the beam is degrading with decreasingpulse width. Using our energy and pulse width num-bers in a plane-wave Frantz-Nodvik model [36] indi-cates that the B-integral ranges from 1.5 (600-pspulse width) to 2.7 (350-ps pulse width). The 350-pspulse width is probably the lower limit for this laserbecause of B-integral concerns. In all likelihood, weare already beginning to see self-focusing effects.

Beam quality M2 measurements were not per-formed. Since we relay image the near field spatialprofile onto the sample plane, the near-field beamquality is of primary interest. We plan to performbeam quality experiments and report on them in thefuture.

Fig. 9. (Color online) Regenerative amplifier pulse shapes thatgenerate FIT amplifier pulses are shown as a function of pulsewidth.

Fig. 11. (Color online) The 500-ps pulse, as measured with a45-GHz photodetector into a sampling oscilloscope, is shown witha second-order super-Gaussian fit.

Fig. 10. (Color online) Images of the amplifier beam taken at anequivalent output relay plane are shown as a function of pulsewidth. The white reference circle is 5 mm in diameter. The“bulls-eye” pattern in the upper-right quadrant is a damage spoton the camera neutral density filter. The statistics are discussedin the section below.

1 June 2007 � Vol. 46, No. 16 � APPLIED OPTICS 3273

5. Summary

We have demonstrated the operation of a Nd:YAG-based MOPA generating 127-mJ pulses at 300 Hzwith an average power of �38 W. The MOPA inte-grates a 400-pJ fiber-based oscillator with a 2.4-mJregenerative amplifier and a relay-imaged 4-passdiode-pumped Nd:YAG amplifier to generate a300-Hz train of output pulses with �3% pulse-to-pulse energy stability. This laser generates both FITpulses as well as a tophat spatial beam. The FWHMtemporal pulse width is continuously user adjustablefrom 350 ps to 600 ps by two simple electronic ad-justments. Longer and different pulse shapes couldeasily be generated by designing a different ASCL.We have begun frequency conversion experiments us-ing this laser and we plan to report on these resultsin the near future.

We dedicate this work to our friend and colleague,Jerry Auerbach. This research was performed underthe auspices of the U.S. Departments of Energy byLawrence Livermore National Laboratory under con-tract W-7405-ENG-48.

References1. O. Mehl, J. Bourderionnet, A. Brignon, E. Lallier, L. Boudje-

maa, and C. Simon-Boisson, “Compact 300-W diode-pumpedoscillator with 500 kW pulse peak power and external fre-quency doubling,” in Conference on Lasers and Electro-Optics(CLEO 2001), Postconference Digest, Vol. 56 of OSA Trends inOptics and Photonics (Optical Society of America Washington,D.C., 2001), CThI6.

2. Y. Bo, A. Geng, Y. Bi, Z. Sun, X. Yang, Q. Peng, H. Li, R. Li, D.Cui, and Z. Xu, “High-power and high-quality, green-beamgeneration by employing a thermally near-unstable resonatordesign,” Appl. Opt. 45, 2499–2503 (2006).

3. B. Le Garrec, G. Razé, P. Thro, and M. Gilbert, “High-average-power diode-array-pumped frequency-doubled YAG laser,”Opt. Lett. 21, 1990–1992 (1996).

4. E. Honea, C. Ebbers, R. Beach, J. Speth, J. Skidmore, M.Emanuel, and S. Payne, “Analysis of an intracavity-doubled

diode-pumped q-switched Nd:YAG laser producing more than100 W of power at 0.532 �m,” Opt. Lett. 23, 1203–1205 (1998).

5. S. Konno, T. Kojima, S. Fujikawa, and K. Yasui, “High-brightness 138-W green laser based on an intracvity-frequency-doubled diode-side-pumped Q-switched Nd:YAG laser,” Opt.Lett. 25, 105–107 (2000).

6. S. Seidel and N. Kugler, “Nd:YAG 200-W average-poweroscillator-amplifier system with stimulated-Brillouin-scatteringphase conjugation and depolarization compensation,” J. Opt.Soc. Am. B 14, 1885–1888 (1997).

7. H. Kiriyama, K. Yamakawa, T. Nagai, N. Kageyama, H. Miya-jima, H. Kan, H. Yoshida, and M. Nakatsuka, “360-W averagepower operation with a single-stage diode-pumped Nd:YAGamplifier at a 1-kHz repetition rate,” Opt. Lett. 28, 1671–1673(2003).

8. K. Furuta, T. Kojima, S. Fujikawa, and J. Nishimae, “Diode-pumped 1kW Q-switched Nd:YAG rod laser with high peakpower and high beam quality,” Appl. Opt. 44, 4119–4122(2005).

9. M. Ostermeyer, P. Kappe, R. Menzel, and V. Wulfmeyer,“Diode-pumped Nd:YAG master oscillator power amplifierwith high pulse energy, excellent beam quality, and frequency-stabilized master oscillator as a basis for a next-generationlidar system,” Appl. Opt. 44, 582–590 (2005).

10. S. Amano and T. Mochizuki, “High average and high peakbrightness slab laser,” IEEE J. Quantum Electron. 37, 296–303 (2001).

11. A. Binder, H. Jaber, D. Ashkenasi, T. Riesbeck, and H. Eichler,“High-power and high-brightness solid state laser systems forprecise and fast micromachining,” Proc. SPIE 5339, 500–508(2004).

12. J. Degnan, “Theory of the optimally coupled Q-switched laser,”IEEE J. Quantum Electron. QE-25, 214–220 (1989).

13. J. Zayhowski and C. Dill, “Diode-pumped passively q-switchedpicosecond microchip lasers,” Opt. Lett. 19, 1427–1429 (1994).

14. J. Zayhowski and P. Kelley, “Optimization of Q-switched la-sers,” IEEE J. Quantum Electron. 27, 2220–2224 (1991).

15. W. Koechner, Solid State Laser Engineering (Springer-Verlag,Berlin, 1993).

16. M. Dymott and K. Weingarten, “Picosecond diode-pumped la-ser system with 9.3-W average power and 2.3-mJ pulse en-ergy,” Appl. Opt. 40, 3042–3045 (2001).

17. D. Walker, C. Flood, H. van Driel, U. Greiner, and H. Klin-

Fig. 12. (Color online) Horizontal (x) and vertical (y) profiles through the 500-ps beam centroid are shown. In addition, the 90% amplitudepoints of the fitted profiles are indicated by vertical dashed lines.

3274 APPLIED OPTICS � Vol. 46, No. 16 � 1 June 2007

genberg, “High power diode-pumped Nd:YAG regenerativeamplifier for picosecond pulses,” Appl. Phys. Lett. 65, 1992–1994 (1994).

18. M. Dawson, W. Schroeder, D. Norwood, A. Smirl, J. Weston, R.Ettelbrick, and R. Aubert, “Characterization of a high-gainpicosecond flash-lamp-pumped Nd:YAG regenerative ampli-fier,” Opt. Lett. 13, 990–992 (1998).

19. C. Gaeta, H. Rieger, I. Turcu, R. Forber, S. Campeau, K.Cassidy, M. Powers, A. Stone, J. Maldonado, S. Mrowka, G.French, J. Naungayan, C. Kelsy, P. Hark, J. Morris, R. Foster,J. Carosella, D. Fleming, R. Selzer, H. Siegert, H. Smith, M.Lim, Z. Cheng, J. Burdett, D. Gibson, R. Whitlock, C. Dozier,and D. Newman, “High-power collimated laser-plasma sourcefor proximity x-ray nanolithography,” J. Vac. Sci. Technol. B21, 280–287 (2003).

20. S. Schiemann, W. Hogervorst, and W. Ubachs, “Fourier-transform-limited laser pulses tunable in wavelength and induration (400–2000 ps),” IEEE J. Quantum Electron. 34, 407–412 (1998).

21. E. Moses, “The National Ignition Facility: Status and Plans forthe Experimental Program,” Fusion Sci. Technol. 44, 11–18(2003).

22. A. Jolly, J. Gleyze, J. Luce, H. Coic, and G. Deschaseaux,“Front-end sources of the LIL-LMJ fusion lasers: progress re-port and prospects,” Opt. Eng. 42, 1427–1438 (2003).

23. D. Hinkel, S. Haan, A. Langdon, T. Dittrich, C. Still, and M.Marinak, “National Ignition Facility targets driven at highradiation temperature: ignition, hydrodynamic stability, andlaser-plasma interaction,” Phys. Plasmas 11, 1128–1144 (2004).

24. J. Honig, “Cleanliness improvements of National Ignition Fa-cility amplifiers as compared to previous large-scale lasers,”Opt. Eng. 43, 2904–2911 (2004).

25. E. Bliss, “Pulse duration dependence of laser damage mecha-nisms,” Opto-electronics 3, 99–101 (1971).

26. B. Stuart, M. Feit, A. Rubenchik, B. Shore, and M. Perry,“Laser-induced damage in dielectrics with nanosecond to sub-picosecond pulses,” Phys. Rev. Lett. 74, 2248–2251 (1995).

27. K. Starke, T. Gross, D. Ristau, W. Riggers, and J. Ebert,“Laser-induced damage threshold of optical components forhigh repetition rate Nd:YAG lasers,” Proc. SPIE 3578, 584–593 (1998).

28. C. Carr, J. Trenholme, and M. Spaeth, “The effect of temporalpulse shape on optical damage,” Appl. Phys. Lett. 90, 041110(2007).

29. M. D. Skeldon, “Optical pulse-shaping system based on anelectronic optic modulator driven by an aperture-coupled-stripline electrical-waveform generator,” J. Opt. Soc. Am. B 19,2423–2426 (2002).

30. Koheras A�S, Blokken 84, DK-3460 Birkerod, Denmark.31. Keopsys SA, 21 Rue Louis de Broglie, 22300 Lannion, France.32. JMAR Technologies, Inc., 10905 Technology Place, San Diego,

Calif. 92127, USA.33. R. J. Beach, E. C. Honea, S. B. Sutton, C. M. Bibeau, J. A.

Skidmore, M. A. Emanuel, S. A. Payne, P. V. Avizonis, R. S.Monroe, and D. G. Harris, “High-average-power diode-pumpedYb:YAG lasers,” Proc. SPIE 3889, 246–260 (2000).

34. ONYX Optics Corporation, 6551 Sierra Lane, Dublin, Calif.94568, USA.

35. Q. Lü, N. Kugler, H. Weber, S. Dong, N. Müller, and U. Wit-trock, “A novel approach for compensation of birefringence incylindrical Nd:YAG rods,” Opt. Quantum Electron. 28, 57–69(1996).

36. L. Frantz and J. Nodvik, “Theory of pulse propagation in alaser amplifier,” J. Appl. Phys. 34, 2346–2349 (1963).

1 June 2007 � Vol. 46, No. 16 � APPLIED OPTICS 3275