This chapter is a summary of the technology of high-powered pulsed lasers for high-intensity laser science. The first section describes current laser sources operating at the petawatt (PW) level. The next looks toward the future, both at technologies under development and complete systems under various stages of construction, proposal, or development. The final section provides a comparison of the technologies, present and future, as well as issues related to achieving the next levels of high intensities needed for scientific breakthroughs. This includes non-standard technologies such as X-ray free-electron lasers (FELs) and fiber lasers, which utilize wholly different technical concepts to get to future higher intensities.
The report includes extensive tutorial material in several appendixes for readers who need to know more about the technologies discussed. Appendix B1 provides a background for solid-state lasers. Appendix B2 covers the nonlinear optics related to optical parametric amplifiers, particularly the broadband optical parametric chirped-pulse amplifiers (OPCPAs) used in some intense sources. Appendix B3 provides details of some enabling technologies. Appendix B4 describes systems under construction or in planning stages, including X-ray FELs. These appendixes include much background material that readers familiar with laser technology may bypass to reach discussions of specific PW-class systems.
Solid-state lasers are the general technology that has enabled demonstration and operation of PW-class systems and generation of intensities up to 1022 W/cm2. Solid-state lasers are either the basic source or the drive source for nonlinear amplifying media, in the case of OPCPA systems. To date, PW levels of output have been achieved directly from two solid-state laser media, neodymium-doped glass
(Nd:glass) and titanium-doped sapphire, (Ti:sapphire) as well as OPCPAs pumped by Nd:glass lasers. All of them are described here. The chapter concludes with a discussion of advanced technologies and X-ray FELs. The latter achieve similar high intensities to PW lasers but with much lower power and far shorter pulse durations. The committee also refers the reader to appendixes that describe various specific systems, including brief summaries of critical supporting technologies that are used to facilitate or enhance system operation.
Table B.1 in Appendix B1 lists important characteristics of many common laser materials. While there are many different combinations of demonstrated laser-active ions and host materials, only two solid-state lasers have been operated with peak power outputs at the PW level. The two materials (Nd:glass and Ti:sapphire) that, to date, have led to PW-class scaling are related in that the technology of both are often incorporated in complete systems. The OPCPA-based sources to date have relied on Nd:glass systems as pump lasers.
A common element in all PW-class lasers is a simple U.S.-based invention, chirped-pulse amplification (CPA). This first allowed existing nominally low-power systems, based on Nd:glass as the laser gain medium and traditionally operating at the sub terawatt (TW) level, to be configured for much higher power operation at minimal cost. The adaptation of this CPA technique applied to these existing Nd:glass systems formed the basis of most activity at the onset of the PW era, most notably pioneered at the Lawrence Livermore National Laboratory (LLNL) where the first PW laser pulse was produced in 1996 by converting a beamline of the exiting Nova system.1 Nova PW was closed very shortly after this first demonstration, and the mantle of driving PW-scale science predominately fell to the Rutherford Appleton Laboratory in the UK, with the conversion of the Vulcan system2 to the PW level, and the Japanese Gekko PW system.3 Over that time new, original systems, based on Ti:sapphire as the gain medium, began to emerge, which owing to its far larger gain bandwidth (compared to Nd:glass) significantly reduced the physical scale and thus entry-level cost to the field. The subsequent emergence of several commercial organizations offering cost-effective products based on Ti:sapphire
1 M.D. Perry et al., “Petawatt Laser Pulses,” Optics Letters 24, no. 3 (February 1, 1999): 160, doi:10.1364/OL.24.000160.
2 C.N. Danson, P.A. Brummitt, R.J. Clarke, J.L. Collier, B. Fell, A.J. Frackiewicz, S. Hancock, S. Hawkes, C. Hernandez-Gomez, and P. Holligan, 2004, Vulcan Petawatt—an ultra-high-intensity interaction facility, Nuclear Fusion 44(12): 239–246.
3 Y. Kitagawa et al., “Prepulse-Free Petawatt Laser for a Fast Ignitor,” IEEE Journal of Quantum Electronics 40, no. 3 (March 2004): 281–93, doi:10.1109/JQE.2003.823043.
TABLE 3.1 Summary of the Key Parameters of the Four Major Technologies Used in Petawatt-Class Lasers and X-ray FEL High-Intensity Lasers
|PW technology||Energy||Pulse length||Repetition rate||Comments|
(0.15 ps to ns)
|Established technology; large scale facilities; low efficiency|
(10 – 100 J)
(> 10 fs)
|Low - High
(0.01 Hz – 10 Hz)
|Established and new technology; medium scale; scope to improve repetition rate, energy|
|OPCPA||Medium – Very High
(5 J – kJ)
(> 10 fs)
|Low - High
(0.01 Hz – 10 Hz)
|New technology; medium scale; scope to develop; high risk development|
(100 Hz – 1MHz)
|Ultrahigh intensity comes from tighter focus and shorter pulse. Not a Petawatt peak power laser. Low risk but high capital and operating cost.|
has really propelled the global uptake of relevant technology and consequently has taken the field forward. In recent years, advanced developments such as optical parametric chirped-pulse amplification (OPCPA) and diode-pumped solid-state lasers (DPSSL) driven CPA have further diversified the technological basis of the field and opened up new scientific and application environments.
Table 3.1 provides a summary of the characteristics of three techniques used to date for PW-class outputs. These include the energy, pulse length, and repetition rate of the system. It should be noted that this table includes only the range of parameters associated with PW-class systems and not the wide range available from these technologies in general.
The first laser to deliver petawatt performance was the “Nova Petawatt,” based at the Nova Facility at the Lawrence Livermore National Laboratory (LLNL).4 This system used Nd:glass as its main amplification medium, which produced an amplified pulse of 660 J in a 440 fs pulse, resulting in 1.5 PW incident on the target. This result is indicative of how glass-based systems operate; they rely on very high-energy pulses (hundreds of J and above) contained in a long pulse (~picoseconds – nanoseconds). The use of glass as the main amplifier medium
4 M.D. Perry et al., “Petawatt Laser Pulses.”
restricts these systems to low repetition rates (< 1 Shot/min); due to the thermal properties of glass, sufficient time must be left between shots to allow the glass to cool in order to prevent damage to the system or adversely affect the beam properties. For example, the Vulcan laser at the UK’s Central Laser Facility (CLF) has a repetition rate of 1 shot every 20 minutes. Research is currently underway to increase this to 1 shot every minute through enhanced thermal management.
Nd:glass-based systems have sufficient bandwidth to generate sub-ps pulsewidths and a low enough saturation fluence for efficient extraction, typically in multi-pass designs, but not so low as to make amplified spontaneous emission (ASE) an unmanageable challenge. (In contrast, and ignoring the issue of scaling material sizes, the other common Nd-doped materials in Table B.1 have too high a gain and hence ASE issues at large stored energies, along with too narrow a linewidth for efficient sub-ps-pulse amplification.) Nd:glass can be pumped by flashlamps due to the overlap of lamp emission and the higher-lying levels of the Nd ion, and can be readily fabricated in very large sizes, (Figure 3.1) as developed through years of work related to the quest for laser-driven inertial confinement fusion (ICF).
Two important examples of Nd:glass PW systems are the original PW system, which was built using an arm of the LLNL NOVA fusion research laser, and the Texas Petawatt, which has attempted to produce the shortest pulses that can be supported in Nd:glass. Both of these are described in Appendix B1. Their designs underscore three limitations to this technology: (1) The gain bandwidth limits output pulses to 0.5 ps in single-glass amplifiers in Nd:glass; therefore, large pulse energies are required and large diffractive optics must be designed to handle the energy (Figure 3.2). (2) Mixed-glass amplifiers can expand the bandwidth leading to shorter pulses and lower overall pulse energies, but this is ultimately limited by the properties of Nd:glass to pulse durations of about 0.15 ps. For details, consult Appendix B1. (3) A significant limitation of glass as a laser host is poor thermal conductivity compared to crystalline hosts such as sapphire or YAG. Thus, the pulse repetition rate in the largest system is about one shot per hour or lower, and experiments must be designed for this.
Ti:Sapphire has been used as the primary amplifier medium for ultrafast lasers for decades due to its inherent broad bandwidth that is necessary to amplify short pulses. The lasers operate around a central wavelength of 800 nm and are typically pumped using green (515 nm – 532 nm) lasers. The green pump light can be generated by frequency doubling a number of different infrared laser sources including flash-lamp-pumped Nd:glass lasers (high energy, low rep rate), diode-pumped Nd:YAG lasers (lower energy, high rep rate), or diode-pumped Yb:YAG lasers (high energy, medium rep rate). The former is the typical method adopted for PW-class lasers in this category to date, which results in a limited repetition rate.
The Ti:sapphire gain medium has the largest linewidth of all the materials in common use, allowing amplifier systems to operate with 15-30-fs-duration pulses. This reduces the pulse energy for PW operation to well under 100 J. The gain cross section is high and saturation fluence low, and the host material has excellent thermal and mechanical properties. The most evident challenge is the short storage time of 3.2 µs, reflecting the relation in Eq. B.2 in Appendix B1, which shows both a large linewidth and cross section arise at the expense of a long storage time. Fortunately, the absorption band (Figure B.4 in Appendix B1) overlaps with the second-harmonic of Nd-doped (or Yb-doped) solid-state lasers, which can pump the upper laser level. Thus, one can employ pump sources based on conventional ns-pulsewidth, flashlamp-pumped, Nd-doped solid-state lasers with long upper-state lifetimes and thus good energy storage. As long as the energy put into the Ti:sapphire upper energy level is extracted within a short (< 1 µs) period, the stored energy from the pump laser can be efficiently extracted by a large-bandwidth, stretched pulse. An important point is that multiple pump lasers can be used, as long their beams all overlap inside the active volume of the laser material, and thus energy scaling of the Ti:sapphire laser is not limited by the energy available from a single pump source.
In terms of technology interrelationships, to date, PW-class Ti:sapphire lasers have most often employed high-energy Nd:glass lasers as pumps, while Ti:sapphire mode-locked lasers and often low-energy amplifiers are employed in PW-class Nd:glass lasers.
The OPCPA concept for large-aperture systems was developed at the CLF to further increase the energy of high-power laser facilities,5 with the first practical demonstration on the Vulcan laser at the CLF.6 In this technique (see Figure 3.3) the frequency-doubled light from a high-energy laser facility is transferred to a chirped short pulse laser via parametric amplification in a nonlinear optical material, typically potassium dihydrogen phosphate (KDP) or lithium triborate (LBO) crystals. The parameter space available to the OPCPA technique is greater than its alternatives, as is shown in Table 3.1. The potential pulse energy for PW-class systems is 5J – kJ, made possible by the extremely broad bandwidth and large aperture crystals available to support the parametric process. The technology is well developed in the low energy, high repetition rate regime, and significant developmental work has been conducted at the high energy, low repetition rate regime.
At present, a number of advanced high-peak-power systems use lower energy OPCPAs in the initial stages of systems, where the high and relative freedom from amplified spontaneous emission (ASE) are major advantages. Their use is described in more detail is different sections of Appendix B.
Provided one can find large-enough-aperture nonlinear crystals, sufficiently broad-bandwidth chirped signal pulses, and high-energy pump lasers, OPCPA technology can be scaled up in energy to reach the PW peak-power level. At present there are three published examples of such systems, all employing frequency-doubled Nd:glass lasers as pump sources. Their deployment worldwide as petawatt sources is described in section 4.5.3 of the Internaional Landscape chapter of this study, and additional details are in Appendix B.
This section briefly assesses and summarizes the state of current intense sources.
- Current approaches to PW-class systems involve CPA technology and, to an almost exclusive degree, employ some combination of Nd:glass lasers, Ti:sapphire lasers, and OPCPAs.
- Nd:glass laser amplifiers can be direct sources of PW-level pulses, or provide pump energy, when frequency doubled, to Ti:sapphire lasers or OPCPAs.
5 Ross et al., “The prospects for ultrashort pulse duration and ultrahigh intensity using optical parametric chirped pulse amplifiers”, Optics Communications, Volume 144, Issue 1, 1997, Pages 125-133, ISSN 0030-4018.
6 Chekhlov et al., “35 J broadband femtosecond optical parametric chirped pulse amplification system”, Optics Letters, 31, 24, 2665-3667 (2006).
7 R. Li, L. Yu, Z. Gan, C. Wang, S. Li, Y. Liu, X. Liang, et al., 2016, “Development of a Super Intense Laser Facility at Shanghai,” presentation at IZEST Conference Extreme Light Scientific and SocioEconomic Outlook, Paris, Nov. 25-29.
- Ti:sapphire laser amplifiers provide shorter pulses than Nd:glass laser direct sources (130-440 fs vs. 20-30 ps), and thus require less pulse energy to reach the PW level.
- At present, PW-class OPCPAs used as the final stage in systems provide pulsewidths in the 32-70-fs range, thus also requiring lower pulse energies than Nd:glass lasers.
- The highest peak power system in operation is at the SIOM in Shanghai, and it uses Nd:glass-laser-pumped Ti:sapphire technology to generate a 5.3- PW, 127 J, 24-fs pulse.7
- The highest pulse rate of current PW-level systems is 1 pulse per second, for the BELLA Ti:sapphire system at LBNL, which uses flashlamp-pumped Nd:YAG lasers as pumps. All the other systems, with Nd:glass-laser-based pumps, operate at several orders-of-magnitude lower rates.
All current PW systems utilize power-amplifier technologies such as flashlamp pumping that are both well-developed and, at the industrial level, close to being obsolete. Nd:glass lasers, because of their limited average power, find little use in applications other than high-energy systems for ICF research or intense sources
for science. Even the lamp-pumped Nd:YAG lasers used in the higher repetition rate projects such as the Berkeley Lab Laser Accelerator (BELLA) system at LBNL (see Figure 3.4) are being replaced for materials processing applications by other types of lasers (see Sec. 5.2).
In Appendix B3, the committee describes some technologies, such as diode pumping, that are already being explored for application to intense sources and can be the basis for future systems with performance that goes beyond the limits currently set by flashlamp-pumped, Nd-doped lasers. The committee also considers technologies that will facilitate pushing the upper boundaries of peak power and lower boundaries of pulsewidth, to enable reaching focused intensities of 1022 W/cm2 and higher.
In Appendix B4, the committee reviews future intense sources at their current stages of development, some under construction, some at the early demonstration stages, and others that have been proposed. One common theme is to increase the peak power to the 10-PW level, with some systems under construction, and to go beyond that, still at the conceptual stage.
Beyond mere power scaling, another path to better exploration of high intensity applications is to increase the pulse rate beyond the very limited range of present Nd:glass lasers and even the 1 Hz rate of Nd:YAG-pumped, Ti:sapphire lasers. For scientific studies, higher rates allow an improvement in the productivity of experiments and in their statistical validity. From a technology standpoint, higher rates can allow a much improved means of reaching high intensities, through active feedback from measurements of the focused spot from the system.
Yet a third scaling parameter besides repetition rate and peak power is laser wavelength. Since the focused intensity based on the diffraction limit scales as the inverse-square of the laser wavelength, some of the most intense sources are Angstrom-class X-ray free electron lasers (see Figure 3.5 for two examples). This is also reviewed in Appendix B4. Since the gain medium in an FEL is the relativistic electron beam, there are some unique features compared to conventional gain media. The gain bandwidth is limited only by the radiation per undulator period, since the electrons slip one X-ray cycle on every undulator wiggle. The more conventional limitations of material properties of the gain medium are absent in an FEL. Current FELs are capable of 100 attosecond pulses, but there are designs for sub-attosecond pulses (zeptosecond range). In addition, relativistic electron beams produced in continuous superconducting radio frequency structures can produce X-rays with kilowatt average powers. Such applications are described more fully in Ch. 4.
126.96.36.199 Futuristic Technologies: ICAN
There have been proposals for systems based on coherent combination of large numbers of fiber lasers. The intention is to produce a system capable of high average powers, combined with high wall plug efficiency. These aims would be accomplished by amplifying a matrix of lasers through Yb-doped fiber. The International Coherent Amplification Network (ICAN) (see Figure 3.6), part of the International Center on ZettaExawatt Science and Technology (IZEST), aims to do just this.8 The laser has begun the combination of a small amount of fibers, with more planned in the future. The proposed final specification of ICAN is to provide 1 KJ energy ultra-short pulses (of under 10 fs) at a high repetition rate of 10 KHz, but present work has much more modest goals. Funding and construction of a completed facility based on this technology is considered wholly aspirational at this point; therefore, while fiber-based systems should be considered for a long-term roadmap, they are not technologies that can be considered viable in the near- to medium-term. Appendix B3 provides more details on the technology of fiber lasers, along with examples of the current state of the art.
This section briefly assesses and summarizes where future source technology is headed.
- The technology of diode-laser pumping, which has revolutionized the operation of lower-power solid-state lasers, is emerging for use in all the stages of
8 W. S. Brocklesby et al., “ICAN as a New Laser Paradigm for High Energy, High Average Power Femtosecond Pulses,” The European Physical Journal Special Topics 223, no. 6 (May 1, 2014): 1189–95, doi:10.1140/epjst/e2014-02172-4.
- PW-class lasers, including diode-pumped, high-energy Nd:glass lasers used to drive final-stage Ti:sapphire amplifiers and possibly next as drivers for the final stages in OPCPA systems.
- Diode pumping has enabled a new class of solid-state lasers, based on Yb-doped materials, which, compared to Nd-doped materials, feature a relatively low level of material heating per W of output power and, for some materials, such as Yb doped mixed oxide ceramics, a much larger bandwidth.
- More advanced cooling of large-aperture laser materials is being employed to increase the pulse rate of high-peak-power systems, featuring either gas or liquid flow between relatively thin disks of gain media.
- Fiber-geometry solid-state lasers, primarily based on Yb-doped silica glass media, have revolutionized the technology of high-cw-power lasers, providing powers on the order of 10 kW in near-diffraction-limited beams. They are severely limited in generation of high peak powers by a number of issues, primarily due to nonlinear effects and optical damage in the small-aperture, long-interaction-length media. The highest single-fiber peak power generated to date has been 3.8 GW (2.2 mJ in 415 fs) from a CPA-based system.9
- Active efforts are underway to scale cw and peak powers of fiber-laser-based sources through beam and/or pulse-combination schemes, with the highest peak power to date of 35 GW (12 mJ in a 262-fs pulse) achieved by a system employing both spatial and temporal beam combining. Techniques are under investigation to extend the number of combined fiber lasers from the current 10s to several orders-of-magnitude higher.
- Coherent spectral combination techniques, primarily with OPCPAs, are being developed to permit generation of pulses in the several-fs region, greatly reducing the pulse energy needed to reach the PW peak-power level.
- Several CPA pulse-compression schemes to either replace or enhance present grating technology for PW-class systems are under investigation. They are at an early stage of development, with performance either orders-of-magnitude below desired levels, or yet-to-be-demonstrated.
This section briefly assesses and summarizes where future intense source systems are headed.
9 T. Eidam, J. Rothhardt, F. Stutzki, F. Jansen, S. Hädrich, H. Carstens, C. Jauregui, J. Limpert, and A. Tünnermann, 2011, Fiber chirped-pulse amplification system emitting 3.8 GW peak power, Opt. Express 19(1): 255-260.
- All of the technologies in current systems also feature in planned 10-PW-class lasers, but, to date, higher-peak-power sources, at the planning level, propose to use OPCPAs, since the aperture of available nonlinear crystals such as KD*P exceeds that of present Ti:sapphire crystals.
- Many of the higher-peak-power systems continue to employ flashlamp-pumped Nd:glass, with higher pulse rates (1 per minute) enabled by active cooling of large discs of glass.
- A diode pumped Nd:glass laser, HAPLS laser, (see Chapter 6), is used as a pump for a Ti:sapphire laser under construction at LLNL and will enable operation at 1-PW level with a 10-Hz pulse rate and 300 W of compressed, average power.
- Many of the “front ends” of the systems under construction make use of broadband parametric amplifiers and pulse-contrast enhancement techniques such as crossed polarized wave (XPW) to yield high levels of pulse-contract ratio. This property becomes increasingly important as the pulse energies increase to reach 10 PW and higher powers.
- Two systems under development feature direct generation of near PW-peak power pulses from diode-pumped Yb-doped media. They feature optical efficiencies below the 10 percent level, with pulsewidths in the 100-fs region.
- Other work in Yb-doped materials seeks to obtain ns-duration pulses for applications as pumps for OPCPAs. With cryogenic cooling, Yb:YAG lasers have generated 100-J-level pulses with >20 percent optical efficiency. One system, in the planning stage, seeks to generate 25 kW of average power (250-J pulses) and drive a PW-level OPCPA at a 100-Hz rate.
- The most ambitious system proposed (ICAN) in terms of average power would do coherent beam-combining of > 10,000 fiber lasers to produce > 0.1 PW-peak-power pulses, at a > 10-kHz rate, for > 100 kW of average power. Those numbers are many orders-of-magnitude greater than present fiber-laser technology and issues such as phasing control of such a larger number of fibers and obtaining a high pulse-contrast ratio remain as significant challenges that are well beyond current or near-future technology.
- Free-electron lasers now operate at much shorter wavelengths (soft to hard X-rays) than the solid-state lasers now used to reach PW peak powers. On that basis, they can, in principle, reach the same intensity levels now possible with PW-class lasers but with much lower pulse energies. To date, they have reached intensities of 1020W/cm2 but in the future may reach and could exceed the intensities of much-longer-wavelength sources.
Table 3.2 provides a comparison of existing, planned, and proposed sources discussed in this chapter, emphasizing the highest levels of various measures of performance (e.g., power, pulse rate) attained in existing facilities and the specified or proposed measures otherwise. In the prior sections and appendixes to this chapter, the report has covered current technologies and the systems using them to reach the PW level, and described systems that are under construction, planned, or conceptualized.
In building facilities for intense sources one has a variety of options for the source configuration; the choices are driven by numbers of factors including the science/engineering to be done, the system initial size and cost, the operating cost, and the reliability of the source. Concerns such as the pulse rate are important in terms of the productivity of the facility. In the facilities under construction in Europe and, to a lesser extent, in Asia, multiple sources based on different technologies are being deployed to, in principle, increase the breadth of science to be done and also increase productivity.
Below the committee discusses some other more qualitative properties of the various technologies.
TABLE 3.2 A Comparison of Existing, Planned, and Proposed Sources Discussed in This Chapter
|Type||Pump||Status||Peak pwr. (PW)||Pulse energy (J)||Pulsewidth (fs)||Pulse rate||Notes|
|Nd:glass||FL||OP||1||130||130||1/hour||Texas Petawatt (U. Texas, Austin)|
|Nd:glass||FL||UC||10||1500||150||1/min.||By National Energetics, US, for L4 at ELI-Beamlines|
|Ti:sapphire||FL-Nd:YAG||OP||1.3||40||30||1 Hz||BELLA at LBNL, Berkeley, CA|
|Ti:sapphire||FL-glass||OP||5.3||127||24||Low||At SIOM, Shanghai, China|
|Ti:sapphire||DP-glass||UC||1||30||30||10 Hz||HAPLS, by LLNL, for L3 at ELI-Beamlines|
|Type||Pump||Status||Peak pwr. (PW)||Pulse energy (J)||Pulsewidth (fs)||Pulse rate||Notes|
|Ti:sapphire||FL-glass||UC||10||210||21||1/min.||By Thales, France for ELI-NP (2 systems)|
|Ti:sapphire||FL-glass||UC||10||150||15||1/min.||APOLLON system in France|
|OPCPA||FL-glass||OP||1||32.6||32||Low||At SIOM, Shanghai, China|
|OPCPA||FL-glass||UC||20||600||30||Low||VULCAN 20 PW at CLF, Rutherford Laboratories, UK|
|OPCPA||FL-glass||Prop||75||1500||20||1/105 min.||EP-OPAL at LLE, Rochester, NY|
|OPCPA||DP-Yb:YAG||Prop||1||10||10||100 Hz||Part of GEKKO-EXA, ILE, Osaka, Japan|
|Yb:CaF2||DP||UC||1||150||150||1 Hz||PENELOPE, HZDR, Dresden, Germany|
|Yb:fiber||DP||Prop||0.1||> 10||100-200||> 10 kHz||100 kW ave. power, ICAN, now XCAN, possibly in France|
|X-ray FEL with high intenstity focus||2-16 GeV Cu linac||OP||6×10-4||0.003||5-50||120Hz||Wavelength 0.11-4.4nm LCLS at SLAC in the US. Other FELs with high intensity capabilities: SACLA (Japan). In construction: European X-ray FEL (Hamburg), PAL (S. Korea), SwissFEL (Switz.), LCLS-II (SLAC).a|
NOTE: FL, Flashlamps; DP, Diode lasers; FL-Nd:YAG, Flashlamp-pumped Nd:YAG; FL-Glass, Flashlamp-pumped Nd:glass; DP-Glass, Diode-pumped Nd:glass; DP-Yb:YAG, Diode-pumped Yb:YAG; OP, Operational; UC, Under construction; Prop, Proposed or notional.
a C. Pellegrini, A. Marinelli, and S. Reiche, 2016, The physics of X-ray free-electron lasers, Reviews of Modern Physics 88(1): 015006.
Nd:glass systems are the oldest direct PW-class technology and remain the established pump source for high-peak-power Ti:sapphire and OPCPA systems. As a direct source, advances in technology have enabled the pulsewidth to reduce from 500 fs to about 150 fs, which is key to enabling the construction of a 10-PW source for ELI-Beamlines in the Czech Republic. Nd:glass systems are well suited for science applications where high pulse energy as well as peak power are important. Recent advances in liquid or gas cooling of glass disks will increase the pulse rate to 1/minute with flashlamp pumping and to the 10-Hz region with diode pumping. The latter also facilitates an increase in conversion of pump light from the 1-2 percent possible with flashlamps to about 20 percent with diode lasers.
The positive aspects of Nd:glass technology include its relative maturity and low cost, at least for flashlamp–pumped lasers, and this has allowed worldwide proliferation of systems. It is, for example, used for shot peening turbine blades in commercial jet aircraft engines.10 Flashlamp pumping is giving way to diode-pumping as the commercial cost of diodes comes down. LLNL has shown ICF power-plant designs that could be based on diode-pumped Nd:glass, in the event that ICF can eventually prove to be feasible.11 There could be a problem with future viability of the technology, however, if key commercial suppliers of large glass slabs and flashlamps decide to exit the business.
At this writing, Ti:sapphire technology has enabled the highest-peak-power laser system yet demonstrated (5.3 PW) and also features in systems under construction to reach the 10-PW level at ELI-NP (Romania), in France and possibly in China.
In common with OPCPAs, it requires a laser as a pump source, as direct diode pumping looks to be a very remote future possibility. Compared to OPCPAs, Ti:sapphire to date has shown higher efficiency in converting pump energy to output energy, about 50 percent before compression, double that of present OPCPAs.
10 Graham Hammersley, Lloyd A. Hackel, and Fritz Harris, “Surface Prestressing to Improve Fatigue Strength of Components by Laser Shot Peening,” Optics and Lasers in Engineering, Laser Material Processing, 34, no. 4 (October 1, 2000): 327–37, doi:10.1016/S0143-8166(00)00083-X.
11 C. D. Orth, S. A. Payne, and W. F. Krupke, “A Diode Pumped Solid State Laser Driver for Inertial Fusion Energy,” Nuclear Fusion 36, no. 1 (1996): 75, doi:10.1088/0029-5515/36/1/I06; A. C. Erlandson et al., “Comparison of Nd:Phosphate Glass, Yb:YAG and Yb:S-FAP Laser Beamlines for Laser Inertial Fusion Energy (LIFE) [Invited],” Optical Materials Express 1, no. 7 (November 1, 2011): 1341, doi:10.1364/OME.1.001341; “Design and Performance of a Diode-Pumped Nd:Silica-Phosphate Glass Zig-Zag Slab Laser Amplifier for Inertial Fusion Energy,” Japanese Journal of Applied Physics 40, no. 11R (November 2001): 6415, doi:10.1143/JJAP.40.6415.
Also in contrast with OPCPAs, Ti:sapphire technology does not connect pump phase to output phase and is capable of accumulating energy from multiple, multimode and random-phased pumps, with only modest pump-timing requirements. This enables future high-energy and high-average power systems that can make use of multiple, relatively low-energy, diode-pumped pump lasers.
Unfavorably in comparison with OPCPAs, heat dissipation in Ti:sapphire crystals is inherent to laser operation, although cryogenic cooling can greatly reduce the effects of the heat. Also unfavorable is the need to counter the effects of transverse ASE at high stored energies, although approaches using time-separated pump pulses (extract during pumping) have so far allowed scaling beyond that possible with a single pump pulse. Ultimate energy limits may come about from the available size of crystals, which cannot equal the aperture of OPCPA crystals such as KD*P.
Ti:sapphire technology still has widespread commercial viability, but at present there is only one commercial source of the large-aperture crystals employed in the highest-energy systems, with a concern that the limited market for these could threaten future supplies.
As with Ti:sapphire, OPCPAs are laser-pumped systems. Following on the prior discussions and details in the appendixes, in comparison to Ti:sapphire, the lack of transverse ASE and the ability to obtain very large-aperture KD*P nonlinear crystals enables scaling, on paper so far, to energies well above 10 PW. Under the proper conditions, OPCPAs can also produce shorter pulses through spectral beam combining. The lack of inherent heating in the OPCPA process promises future scaling to high average powers provided appropriate pump lasers can be developed.
On the other hand, besides the need for precise control of the pulsewidth and timing, OPCPAs do place a burden on the pump laser in terms of its beam quality. In theory, the OPCPA idler wave can carry off pump beam phase inhomogeneity, but in practice back-conversion that occurs as the pump-signal conversion increases to 25 percent and beyond does require good phase properties. Other issues include the coupling of pump intensity to signal phase, requiring relatively uniform pump intensity. These sensitivities to pump beam quality are a challenge to average-power scaling for high-energy OPCPAs. As with Ti:sapphire, alternative nonlinear crystals, such as LBO, which have more desirable thermos-mechanical properties compared to KD*P, require further development of large-aperture growth techniques.
As this chapter has described, OPCPAs in combination with Nd:glass and Ti:sapphire final amplifiers have proven to be an excellent combination to providing short output pulses with high temporal contrast. Much remains to be done in
operational systems to advance OPCPAs also as final-stage amplifiers, as the great majority of present 10-PW-class systems under construction still rely on Nd:glass or Ti:sapphire final amplifiers.
The appeal of direct diode-pumped, Yb-doped bulk lasers is simplicity, ultimate electrical efficiency, and the chance for average-power scaling enabled by the favorable properties of crystals as active media. To date, the major problem is the high saturation fluence of available Yb-doped crystals, which has led to long pulsewidths and low final-stage efficiencies (below 10 percent) in converting diode pump energy to output pulse energy for systems attempting to reach the PW peak-power level. This technology would benefit from further development of host media with higher gain cross sections with appropriate bandwidth.
A more promising approach to the use of Yb-doped crystals is their application as a ns-duration, high-energy, efficient pump sources for either Ti:sapphire or OPCPA final stages. With 100-W-average-power-systems now operational, and kW-average-power sources under construction or planning, the Yb-doped devices can provide some of the first tests to determine the actual average-power capabilities of OPCPA-based, high-energy, and peak-power sources.
Yb-doped fiber systems are the furthest away from PW-class capability and, barring a major technical breakthrough, combined with a substantial increase in development funding, will not reach that capability for a considerable time. They do provide the highest average powers for ultrafast systems, at very limited pulse energies.
The major driving force for the technology is the promise of high efficiency combined with a high average power and pulse rate, a requirement for future applications for intense sources, such as drivers for advanced particle accelerators.
Conventional radiofrequency (RF) and superconducting RF electron linear accelerators offer two advanced means of reaching high intensities: FELs and relativistic scattering. The machines are in the billion-dollar range and so only exist at national laboratories.
Current focused X-ray intensities from FELs are comparable to the performance of petawatt lasers and are limited by the quality of sub-micron focusing technologies, which may improve in the future.
The intensity in the center of momentum frame for a laser backscattering from a relativistic electron beam enters a new, much higher regime than anything envisioned for lasers alone. If a PW-class focused laser is directed at a >10 GeV electron beam, the resulting intensity exceeds the Schwinger limit. A summary of these comparisons is captured in Table 3.3.
The discussions have generally emphasized source performance in terms of peak power, with the pulsewidth and pulse energy as parameters. In many of the science and engineering applications the focused intensity of the source is the key performance parameter, and, for most if not all the sources the committee discusses, data on this is limited or a goal for the system. Intensity is a function not only of the peak power, but also of the output beam quality and the beam focusing optics used in the system.
TABLE 3.3 Comparison of Technologies for High-Intensity Lasers
High energy pulses
Relative low cost
Low pulse rate
Highest peak power
Good conversion of pump to output
Flexibility, simplicity of pump lasers
Ultimate energy limit from crystal size
No inherent heating
Potential for fs-duration pulses
Burdensome pump requirements
Crystal size limits for non-KD*P systems
Untested at highest energies, rates
Simplicity for direct lasers
Potential for high efficiency
High average powers as pumps
Not yet at PW level as direct sources
Low efficiency to date as direct sources
Promise of high efficiency
Promise of high average power, PW
Now high average ultrafast source
Many orders of magnitude from PW
Technology of massive beam combination not at hand
|Linac-based light sources||
Shortest wavelength (X-ray FELs), and tunable over many decades of wavelength.
Tightest focus (sub-micron)
Intensities exceed Schwinger limit for PW lasers scattering from >10 GeV beams
Linacs are extremely expensive (~$1B)
Science access is limited to a few places in the world.
Highest intensities are only reached in a relativistic reference frame, limiting the general utility.
Measurements of beam quality have tended to be a challenge (even definitions can vary), especially for sources with a pulsed output and even more so for sources with a limited pulse rate. The use of CPA in intense lasers leads to an additional complication, as all the wavelengths present in the beam may not have the same beam properties, and thus the focused beam may contain some form of spectral distortion that leads to a reduced peak intensity. In general, nonlinear effects in the gain media, especially in the final amplifier stages, can lead to spatial and temporal phase distortion in the output beam. There are some special concerns about OPCPA systems where pump spatial or temporal intensity fluctuations may also lead to complex phase irregularities in the beam.
For beam focusing optics, the challenge as the pulses get shorter is to accommodate the broad spectral linewidth, which rules out the use of refractive optics. Mirrors are an option, but dielectric-coated mirrors face limits on obtaining a constant-phase, high-reflectivity over a large wavelength region, and metallic mirrors have inherent absorption that can limit operation at high pulse rates. Discussions of future spectrally combined sources that produce several-fs-duration pulses require accompanying ultra-broadband focusing optics to reach high intensities. Finally, X-ray FELs can reach very high intensities only if focusing optics with sufficient spatial accuracy can be fabricated. 100-nm focal spot sizes have been demonstrated, but this is far from the diffraction limit. Diffraction-limited optics for X-ray sources require absolute accuracies about three-orders-of-magnitude better than those for the near-infrared. However, given the very short wavelengths involved, even if the focusing optics are not diffraction-limited, the focal spot area will still be many orders-of-magnitude smaller than that for solid-state laser sources.
Unfortunately, there is no simple way to obtain a good measure of intensity, and for many systems this number must be inferred from the physical results of experiments conducted with the beam. In principal, the development of adjustable, phase-correcting optics, both temporal and spatial, can allow a substantial enhancement ability of sources to focus to a high intensity. As LLNL has pointed out to the community, for low pulse-rate sources these corrections are “feed-forward” in that the corrections are applied based on a measurement of the last beam, with iterations necessarily somewhat of a “cut-and-try” nature.12 As pulse rates increase (above about 5 Hz), one can apply active “feedback” control using goal-seeking algorithms that apply corrections without requiring physical understanding of what the corrections actually do. Technologies where this high rate is possible will almost exclusively involve diode-pumped lasers. Future development of intense
12 Lawrence Livermore National Laboratory (LLNL), 2016, High-power high-intensity lasers for science and society, white paper submitted to the National Academy of Sciences Committee on the Opportunities in the Science, Applications, and Technology of Intense Ultrafast Lasers, LLNL-TR-704407.
sources may seek to trade high peak power at low pulse rates for lower peak powers at higher rates, if the latter sources can be operated with active feedback to generate higher intensities.
The discussion in this chapter (and appendixes) has presented some mention of the national origins and commercial status of various technologies. Below is a concise summary of that information with some additional commentary, which will serve as an introduction to the next chapter of the study, the International Landscape. This also forms part of the basis for Conclusion 5 described in Chapter 7 of this report.
In general, the United States has originated key technologies employed for PW-class lasers and FELs capable of high intensities. The list includes the following:
- Nearly all of the solid-state lasers that have been applied, including Nd:glass, Nd:YAG, Yb:YAG, and Ti:sapphire;
- The CPA technique that was critical to the overall approach;
- The lamp-pumped, high-energy Nd:glass systems that were used in the first PW system and are still being used to scale to higher powers;
- Large-aperture gratings that also enabled the first PW system;
- High-power diode lasers and diode-pumped solid-state lasers, including low-heat-generation Yb-doped lasers;
- Nonlinear optics, notably harmonic generation and parametric devices;
- Cladding-pumped fiber lasers that have allowed scaling of fiber-geometry systems to high average powers; and
- The original free-electron laser concept and the extension to self-amplified spontaneous emission (SASE) that made X-ray FELs possible.
The origins of other technologies are diverse. The OPCPA concept originated in Lithuania in the 1990s and was first scaled to near-PW-class peak powers in the UK. The rapid growth of large-aperture KDP-family crystals originated in Russia but was subsequently scaled further in the United States for the ICF program, which employs large-aperture crystals to generate the ultraviolet (UV) energy now employed in experimental work. The nonlinear materials beta barium borate (BBO) and LBO, now used extensively as high-gain OPAs in the first stages of PW-class lasers and now also as the final stage (for LBO) in an OPCPA system, were first grown in China. The XPW technique for pulse-contrast enhancement was developed in France.
In terms of current trends, this chapter illustrates that the United States continues to develop advanced PW-class lasers, with the mixed-Nd:glass, Texas Petawatt
laser, which had origins in the ICF effort, leading to the ongoing development of a 10-PW system being built in the United States by National Energetics for ELI-Beamlines in the Czech Republic. The HAPLS system, under construction at LLNL, also for delivery to ELI Beamlines, represents a significant milestone in diode-pumped sources and also in high-average-power Ti:sapphire lasers. However, the construction of higher-peak-power Ti:sapphire lasers is now being done in China and by two French companies, Thales and Amplitude Technologies, for delivery to ELI-NP in Romania and ELI-ALPS in Hungary. High-power OPCPAs have been built in Russia and China, with even higher-power units under development in the UK. Recent work to develop broad-bandwidth, diode-pumped, Yb-doped crystals has been primarily in France, and construction of high-peak-power systems is in progress in Germany. Efforts on high-pulse-energy Yb:YAG lasers as pump sources for Ti:sapphire and OPCPA systems are to be found in the UK, in cooperation with the Czech Republic, and in Japan.
Although the United States has extensive ongoing efforts on beam-combined fiber lasers to generate high cw powers, beam-combined, CPA-architecture fiber-laser work is centered in Germany and to a lesser extent in France, with limited work in the United States.
The U.S. program to develop a high-energy Nd:glass laser for ICF at LLNL and also LLE has had a major role in the early development of PW-class lasers, and much of the technology for scaling the energy of Nd:glass lasers has originated from this program. Considerable expertise in mega-Joule class Nd:glass lasers and CPA technology can be found at LLNL and LLE. At LLNL there is a CPA-based advanced radiographic capability (ARC) system, which is a PW-class CPA source with a long pulsewidth of about 30 ps. ARC has a specialized application to ICF diagnostics. Beyond the LLNL-developed HAPLS laser, which will be shipped to the Czech Republic, at present, there is no new construction of PW-class lasers at either LLNL or LLE. LLE does have a proposal to build the EP-OPAL, OPCPA-based, 75 PW source but no significant funding to do so.
In terms of commercial expertise, there is one company in the United States, National Energetics, active in building PW-class, Nd:glass lasers. Another company, Continuum, now owned by Amplitude Technologies in France, does have background in high-energy Nd:glass and Nd:YAG lasers. Otherwise, companies in the United States, such as Coherent, Spectra-Physics, and KMLabs, have concentrated on lower-power ultrafast sources, primarily based on Ti:sapphire, to address the still-large scientific market for these devices. It is to be expected that the ELI program in the UK, besides providing business to the French companies Thales and Amplitude Technologies, will create EU-based spin-off companies that specialize in high-peak-power laser systems.