Free-Electron Lasers as Pumps for High-Energy Solid-State Lasers

The Concept

http://pbpl.physics.ucla.edu/Work supported by DOE BES grant DE-FG03-98ER45693

Abstract

Parameter Value

Pump Wavelength [Ti:S] 490 nm

Macrobunch Length [Ti:S] 3.5 µs

Macrobunch Energy 500 J

Microbunches 2000(1 in 5 RF buckets)

Beam Energy [LCLS] 250 MeV

Peak Current [LCLS] 500 A

Undulator Period 5 cm

Undulator Parameter 2.5

Undulator length(Un-optimized; depends on seed)

≈ 20 m

FEL efficiency 5%

Optical energy per pulse 12.5 mJ

ReferencesConclusions

Acknowledgments

The authors thank James Rosenzweig, Sven Reiche, Nick Barov, Alex Murokh andBill Krupke for useful discussions.

QuickTime™ and aTIFF (Uncompressed) decompressorare needed to see this picture.

Work supported by ONR grant N00014-02-1-0911

This work was performed under the auspices of the U.S. Department of Energy by the University of California Lawrence Livermore National Laboratory under contract No. W-7405-Eng-48.

QuickTime™ and aTIFF (Uncompressed) decompressorare needed to see this picture. G. Travish1, J. K. Crane2 and A. Tremaine2

(1) UCLA Dept. of Physics & Astronomy, Los Angeles, CA 90095. USA. (2) Lawrence Livermore National Laboratory, Livermore, CA 94551. USA.

High average-power free-electron lasers may be useful for pumping high peak-power solid-state laser-amplifiers. At very high peak-powers, the pump source for solid-state lasers is non-trivial: flash lamps produce thermal problems and are unsuitable for materials with short florescence times, while diodes can be expensive and are only available at select wavelengths. FELs can provide pulse trains of light tuned to a laser material’s absorption peak, and florescence lifetime. An FEL pump can thus minimize thermal effects and potentially allow for new laser materials to be used.

This paper examines the design of a high average-power, efficient high-gain FEL for use as pump source. Specifically, the case of a 100 J class pump for a 100 TW class laser is considered. FEL design goals, laser-material selection-guidelines, and specific examples are discussed. The modification and use of planned fourth-generation light-source infrastructure to also act as high-energy pumps is considered.

100TW at LCLS A MERCURY-like Pump

Pump Source

Flashlamp Diode Laser FEL

Avg. Energy Very high High Low High

Peak Energy Medium Low High Very High

Heat Load High Low Low Very Low

Wavelength VIS IR-VIS IR-UV IR-UV

Use a high average power FEL to pump a conventional laser

QuickTime™ and aTIFF (Uncompressed) decompressorare needed to see this picture.

High brightness injectorHigh average power acceleratorCompressorSeed laserLong tapered undulatorConventional laser amplifier

Components

>10J or >100TW laser hard to makePumps only available for some wavelengthsLarge diode array only good for 1 laserCan use existing FEL facilityCan synchronize big laser to beam & FELNew materials, new power formats

What & Why

A comparison of existing laser pump sources with the FEL based pump. The FEL is suited to high energy and short wavelength applications.

High Energy Laser Applications

High-field physicsNuclear physicsFusion sciencesProton beam generationRadiography

Ideal Pump SourceMatched to Gain Medium:

WavelengthBandwidthTime structureSize

And:StableEfficientLow cost per watt

Not a lot of pumps to choose from…

[1] T. Tajima and J. M. Dawson, Phy. Rev. Lett. 43 267 (1979).[2] M. D. Perry and G. Mourou, Science 264, 917 (1994).[3] T.E. Cowan, et. al., Laser and Particle Beams 17, 773 (1999).[4] M. H. Key, Nature 412, 775 (2001).[5] Y. Sentoku, T. E. Cowan, A. Kemp, and H. Ruhl, Phys. Of Plasmas 10, 2009, (2003).[6] M.D. Perry, et. al., Rev. Sci. Instr. 70, 265-269 part 2, (1999).[7] J. A. Paisner et. al, SPIE Proceedings Series 2633, p. 2, Bellingham, WA (1995).[8] W. Koechner, Solid-State Laser Engineering, Springer (1999), pp312.[9] J. T. Weir, et al., Proc. SPIE 1133, pp.97-101 (1989).[10] A. J. Bayramian, et. al., Proc. Adv. Solid State Photonics 83, 268 (2003).[11] V. Ayvazyan et al., Phys. Ref. Lett. 88 (2002) 104802.[12] J. Lewellen et al., Proc 1998 Linac Conf., ANL-98/28, 863-865 (1999).[13] Linac Coherent Light Source (LCLS), SLAC-R-521, UC-414 (1998).[14] J. Als-Nielsen, Proc. Workshop on 4th Gen. Light Sources, ESRF Report, Grenoble (1996).[15] I. B. Vasserman, et al., Proc. Part. Accel. Conf. (1999).

The use of an FEL as a pump for a solid-state lasers may find application in existing facilities as well as purpose built machines. A high energy, high-efficiency FEL has yet to be demonstrated experimentally, but appears achievable. Ultimately, the practicality of such a system may be an economic decision as diodes become more affordable. However, the flexibility of the FEL to pump at multiple wavelengths and to act as a useful source in its own right may prevail over a simple cost-analysis.

Work remains to find materials better suited to the FEL based pump-source. Optimization of the FEL design as well as a realistic accelerator design also remain to be done. Finally, accelerator-based alternatives to FEL pumping need to be considered such as direct electron-beam excitation of a gain material, optical pumping of laser diodes, and FEL assisted mixing using an optical parametric amplifier (OPA).

Consider a high power FEL that pumps a Ti:S amplifier:Use front end of LCLSAssume a multibunch photoinjector (i.e. TTF or AFEL)Compress beam in BC-1Send all but head and tail bunches to long tapered undulatorProduce 25J or 490nm pump light over 3µsObtain > 10J at < 100fs (> 100 TW) of 800 nm lightCan do this at 120 Hz!

Challenges:Prove high efficiency for visible FELBeam loading compensation

(more linac sections?)Syncrhonization of light to x-ray due to BC-2, etc

(use head pulse to measure phase error?)High energy seed laser

(Multiple diode pumped YLF? OPA?)

Match the macrobunch length to the florescence lifetime of Ti:SMatch the FEL wavelength to the absorption peak of Ti:S

Goal: Produce a high peak power laser using the LCLS front end

QuickTime™ and aTIFF (Uncompressed) decompressorare needed to see this picture.

Consider a MERCURY-class pump that can deliver 1 KJ of 905 nm light in 1.1 ms:

What is MERCURY?

State of the art diode pumped solid state laser (DPSSL)Designed as a scalable direct drive fusion laserGoals: 100J, 10% efficiency, 10HzPulse length is 5ns, but compressibleSystem uses over 6000 diodes producing 60kW peak!Yb:S-FAP disks are the final amplifierHypersonic gas cooling of crystals

Superconducting linac is selected to take advantage of the long fluorescence-time.

Assume a TTF based linac Need 3x105 bunches of 1 nC

eachFilling 1 in 10 RF bucketsRun at about 60 MeV

FEL-wavelength is long:RF thermionic-gun with a

compression alpha-magnet may work

Though a long undulator (≈20 m)A 5%-efficient FEL.Each pulse is 3 mJ of optical energyYielding 1 KJ of optical power

Can an FEL do this?

100J is a lot, but Yb:S-FAB has a 1.1 ms florescence time!

So, yes, you can do it.

Why use an FEL for this?

Problem with diodes:100J class laser costs about $10M

That’s on the order of the FEL6000+ diodes cost about $3M

Diodes only work for one arrangementDiodes have 108 shot lifetime.

That’s 1 year at 10HzAdvantages of FEL

Can pump many different lasersCan run at much more than 10HzOptically superior — easier to couple to crystal

Parameter Value

Pump Wavelength [Yb:S-FAP]

905 nm

Macrobunch Length [Yb:S-FAP]

1.1 ms

Macrobunch Energy 20 k J

Microbunches 300,000(1 in 10 RF buckets)

Beam Energy 60 MeV

Peak Current 500 A

Undulator Period 2 cm

Undulator Parameter 1

Undulator length(Un-optimized; depends on seed)

< 20 m

FEL efficiency 5%

Optical energy per pulse 3 mJ