May, 2008

Samuel Zagala's laser laboratory at Stanford University is an improbable setting to break the longest standing record in the history of laser science. Three optical tables comprise Zagala's small laser, each table crammed to capacity with optics and amplifiers that scatter second-harmonic green light at ten times a second. A "mere" table-top laser, but this system has recently matched the largest lasers in the world for peak power of a single beam.

Much like peak computer power, laser power rose exponentially between the 1960's and 1990's, doubling every year or two. Unlike computers, however, peak laser power leveled off at the end of the century, just above the quadrillion-watt mark. And without funding for multi-million dollar short-pulse systems, it seemed the record might remain there for decades to come.

But now, thanks to a deceptively simple idea that Zagala came up with as a graduate student, that exponential pace is about to begin again. And this time, there's no telling where it might stop.

The Petawatt

For twelve years now, the world's most powerful laser has been an enormous, complex device with a simple name: the Petawatt¹. Originally built at Lawrence Livermore National Laboratory in 1996, the machine has been moved not once, but twice. Following the shutdown of LLNL's Nova laser facility in 1999, the Petawatt was moved fifteen miles southeast to LLNL-operated Site 300, where for six years it was hidden from the public by a veil of secrecy. Then, emerging from classification, it was brought back to LLNL proper in late 2005, where it has been undergoing an extensive upgrade to a 20-Petawatt system for use in conjunction with the oft-delayed National Ignition Facility (NIF).

"The reason that no one has beaten our 5-Petawatt record," remarked Perry Johnson, associate director of the laser program at LLNL, "is that no one thinks it would be worth the effort." Looking beyond the front end of the laser system, it is easy to see why. From the small oscillator room, the laser passes into the main NIF building and is injected into four of the 192 beamlines that make up the gargantuan billion-dollar laser.

These immense glass amplifiers chains, of course, are already obsolete. Smaller and more efficient quantum-dot systems could be used for the final amplification of another Petawatt laser. But the real challenge lies past the laser amplifiers, past the jungle of beams in a room called the "switchyard", and into the laser target bay itself. There--just beyond the primary 20m sphere where Livermore scientists hope to ignite the world's first confined, self-sustaining fusion reaction--lies the true heart of the Petawatt: the compression chamber.

There is a rumor at Livermore that the Petawatt compression chamber was once the hull of a U.S. Navy submarine. The rumor is false, but it is easy to see a resemblance. The cylindrical chamber is 18m long, 5m in diameter, and vacuum rated. Four pairs of meter-sized diffraction gratings are mounted in the chamber, each one the largest in the world.

"It's the gratings that have been the stumbling block," explained Zagala, the head of Stanford's Plasma Research Group. "All of the other laser technologies have made enormous strides in the last ten years. [Quantum] dot amplifiers can increase bandwidth, energy, efficiency, you name it. But without meter-class, high damage-threshold gratings, you can't translate that technology into an ultra-high power laser."

High-Power Lasers

The theoretical peak final power per unit area for any lasing medium is given by the equation:
P_peak = 2Dn hn/s

Here Dn is the bandwidth of the lasing medium, equal to half the inverse of the minimum possible pulse length; s is the stimulated-transition cross section of the laser medium. For the most advanced quantum dot amplifiers operating at hn = 1eV, s = 2 x 10^-21 cm² and Dn = 10¹⁴ sec^-1, corresponding to a peak theoretical power of over 10¹⁶ W/cm². According to the above Equation, it would seem that a modest 100 cm² beam could surpass an exawatt with today's amplifiers. This, of course, is not the case.

The primary reason these theoretical power levels remain unrealized is that any amplifier would be destroyed by the corresponding intensities. For the last twenty years this problem has been circumvented using a technique known as chirped-pulse amplification². It works as follows: starting with a high-bandwidth short-pulse laser, small gratings are used to 'chirp' the beam, separating out the different frequency components to form a much longer pulse. The laser can then be safely amplified to high energies, because of the lower intensity of the long-pulse beam. After gaining energy in this manner, the short-pulse can be recovered by recompressing the laser with another pair of gratings, thus creating the final high-power beam.

But it is the final grating that must withstand the full power of the laser, up to several kilojoules of energy in less than a picosecond. To avoid damage, this final grating in the laser system must be extremely large. And because the short pulse must propagate in vacuum, the gratings must be enclosed by an enormous compression chamber and the associated vacuum hardware. The end result is anything but cheap.

Plasma Compression Grating

Zagala set out not to break any records, but to make high-power lasers more affordable. "We knew we had to make gratings that could withstand higher intensities," said Zagala, "but an incremental improvement wasn't the answer. We wanted a grating that could withstand anything we could throw at it." Drawing upon the concept of plasma mirrors³--devices that reflect high-power beams with a solid density plasma created by the laser itself--Zagala developed the concept of a plasma compression grating.

The idea is simple: the long-pulse laser is split into two equal beams, each reflected off of a small conventional grating, and then recombined in a short-pulse interference pattern on a small piece of fused silica. Partially focused, the beams instantly form a solid-density plasma on the surface, and the interference structure is imprinted on the plasma itself. Then, with the plasma acting as a regular grating, the various frequencies that make up the laser's bandwidth are reflected in different directions. The different path lengths for the separate frequency components compress the beam into a single short pulse.

Simple in principle, perhaps, but in practice the plasma grating was a difficult project. "We had to clean up the [laser's] phase fronts by an order of magnitude," explained Cameron Randall, a postdoctoral researcher at SPRG. "And getting the beams back together after the plasma grating was a real pain. Just getting 20% compressed took two years."

But this week, Zagala announced that his group has improved the compression fraction to 50%, and, combined with their new quantum dot amplifiers, their laser has surpassed the petawatt level for the first time.

"It's not clear whether or not we've technically set the power record yet," says Randall. But the consensus is that the record is only a matter of time. And with their relatively tiny compression chamber (three orders of magnitude smaller than Livermore's) their entire system is much smaller--and cheaper--than the huge Petawatt at LLNL.

The shot rate is still low, as the fused silica that forms the plasma grating must be replaced after every shot. But Zagala's group is working on a way to replace it while still under vacuum, as well as the possibility of shooting the same piece of fused silica more than once.

"This is going to open so many doors," Zagala says. "So much laser technology has just been waiting for this breakthrough. We're talking petawatt, 10-petawatt table-top systems, something people would have never thought possible just five years ago."

At LLNL, Johnson sounded similarly upbeat. "This is a tremendous piece of work," he said. "We've already begun discussion of an exawatt laser system, based on [Zagala's] work." When asked about possible applications, he immediately mentioned Livermore's fusion program. "We knew from the start that twenty petawatts was cutting it close for Fast Ignition," referring to the fusion scheme in which a high-power laser beam ignites a small area in a compressed DT fuel pellet⁴. "But with 100 petawatts we'd have no problem. NIF will ignite by itself, of course, but with Fast Ignition we can get fusion energy gains that are a hundred times higher."

Zagala spoke of rather different applications. "Chan Joshi's half-GeV plasma accelerator at UCLA⁵ is a great example of what you can do with a high-power laser." Most estimates agree that a 10-petawatt laser could drive a 20GeV electron accelerator, and thanks to Zagala's work both the laser and accelerator might fit in a single room. Hospitals and industries could be only some of the beneficiaries. For example, particle physicists might be able to use steep plasma acceleration gradients to get muons to relativistic speeds before they decay. Turning Fermilab into a muon-antimuon collider is just one of the possibilities that this breakthrough could allow.

Other fundamental physics could be explored by increasing the peak laser power by a few more orders of magnitude. Even at terawatt-scale intensities, electrons in the laser focus oscillate at relativistic velocities. A well-focused multi-exawatt beam would increase an electron's acceleration to what a particle might experience at the event horizon of a black hole. And according to Einstein's equivalence principle, the result could very well be lab-produced Hawking-radiation, providing the first experimental verification of the concept.

How high of a laser power will this breakthrough eventually permit? Zagala is not venturing a guess, but the next obvious stumbling block is the theoretical peak power discussed above; parameters entirely determined by the new quantum dot amplifiers. But if the bandwidth and peak-energy requirements can continue to increase, there may be no end in sight. "Perhaps one day the only fundamental limit on laser intensity will be loss of energy to electron-positron production when we focus the beam [to 10²⁸ W/cm²]," says Randall. "But first we'll have to come up with some new metric prefixes to describe the laser."

REFERENCES:

1. M.D. Perry et al., Opt. Letters 24, 63 (1999)
2. R.A. Fisher and W.K. Bischel, IEEE J. Quantum Electron. QE-11, 46 (1975); D. Strickland and G. Mourou, Opt. Comm., 56, 219 (1985)
3. M.D. Perry, V. Yanovsky, M.D. Feit and A. Rubnechik, J. of Appl. Phys., March 3 (1999)
4. M. Tabak, J. Hammer, M.E. Glinsky, W.L. Kruer, S.C. Wilks, J. Woodworth, E.M. Campbell, and M.D. Perry, Phys. Plasmas 1, 1626 (1994)
5. C. Clayton, R. Narang, K. Marsh, C. Joshi, Phys. Rev. Lett. 89, 325 (2007).