Bigger, better, brighter

Stephen Mounsey takes a look at the world's largest lasers, those already built and those planned for the future

In January, the USA National Nuclear Security Administration announced that the National Ignition Facility (NIF), at the USA’s Lawrence Livermore National Laboratory, had successfully delivered a pulse of a laser radiation with an unprecedented energy of one megajoule. The facility is built around the largest laser in the world, with a designed pulse energy capability of up to 1.8 megajoules. The aim of the facility is to demonstrate laser inertial confinement nuclear fusion, in which a millimetre-sized target containing a mixed nuclear fuel consisting of deuterium and tritium is compressed and heated with laser energy to the point at which nuclear fusion ‘ignites’ within the fuel, releasing massive amounts of energy. The hope is that the technique may extract more energy than it requires.

From a business perspective, there are many ways in which enterprising developers of photonic technology can take advantage of the unusual markets opened up by large projects such as the NIF. Construction of the facility officially began in 1997, and by its eventual completion in March 2009 it was five years behind schedule and had cost four times its projected budget. During that time, the developers of the facility purchased the equipment for 192 identical beam lines, each of which was at a scale seldom seen in laser engineering. The laser is based on flash lamp-pumped neodymium-doped glass gain medium, and uses KDP (potassium dihydrogen phosphate) and KD*P (potassium dideuterium phosphate) crystals as a non-linear crystal used to convert the frequency of the laser light.

The KDP and KD*P crystals were grown by Ohio-based Cleveland Crystals (now part of UK crystal specialist Gooch and Housego). Dr Andrew Robertson, vice president of engineering at Gooch and Housego, states that each of the facility’s 192 beamlines has its own KD*P frequency doubler and frequency tripler optics, as well as a KDP plasma-electrode Pockels cell (PEPC, sometimes pronounced ‘pepsi’), which, according to Robertson, is effectively an optical switch. A total of 672 optics were produced for the project, cut from 77 crystals, each of which had to be produced to a very high level of crystal perfection. Additionally, the optics had to be large in order to avoid localised heat build-up, as Robertson explains: ‘Because of the high energy of the beam, they have to expand the beam out and send it through the very large KD*P crystals. The crystals are therefore grown to 150-170cm in diagonal each, and they’re very high quality too – there can be no flaws in these crystals,’ he says. Advancements in growth techniques meant that a majority of these huge crystals took around two months to grow, but some crystals had to be grown by more conventional methods, and these took up to two years to reach the required size.

The size of the crystals required for NIF necessitated that they be produced from KDP because the material’s solubility allows crystals to be grown in a liquid environment. ‘Many other types of non-linear or optical crystals are grown in high-temperature environments, where size is limited,’ explains Robertson. That aside, Robertson observes that the optical properties of KDP and KD*P are not ideal: ‘If you were building a small frequency-doubled laser, you would not use KDP; its non-linear coefficients are modest, and therefore nonlinear interactions are fairly limited,’ he says. For this reason, other large lasers around the world are being planned and built with different materials in mind.

At a laser facility near Bordeaux, the French Commission for Atomic Energy (CEA) has demonstrated a laser capable of 200 joules at 527nm (red). The non-linear optics for this laser were composed of LBO (lithium triborate) rather than the KDP used in the NIF. Marc-Andre Herrmann, general manager of Crystal Laser, the company that supplied the optics, outlines some of the advantages of the material: ‘It is more efficient, especially when the incoming infrared beam is not perfectly homogeneous. The incoming beam can be distorted for a variety of reasons. Typically, the beams are subject to very large thermal effects; Nd:Glass is often used as the gain medium, and glasses tend to have very poor thermal behaviour,’ he says. The effects of thermal distortions become more pronounced if the laser is used heavily: ‘If the laser shoots repeatedly, then we tend to get a poor beam profile. LBO is much better in terms of angular and thermal tolerance [than K*DP], and this allows [frequency conversion] to be less affected by deteriorating beam quality.’

Engineers inspect the target chamber at the National Ignition Facility in the Lawrence Livermore National Laboratories. Image credit is given to Lawrence Livermore National Security, LLC, Lawrence Livermore National Laboratory, and the Department of Energy under whose auspices this work was performed.

Currently, however, manufacturers such as Crystal Laser are not able to produce LBO crystals large enough for the biggest, hottest lasers. ‘We managed to produce a 65mm diameter LBO crystal, but of course this was a prototype – a one-off crystal,’ says Herrmann, adding that typically the company manufactures crystals with apertures up to 40mm. At the same facility as the 200-joule laser, the CEA is developing the Laser Mégajoule (LMJ), a high power laser, which aims to achieve nuclear ignition in 2012. Similar to the NIF, the LMJ depends on KDP to achieve the large crystals required. LMJ’s crystals are grown by Saint-Gobain, a French optics specialist. However, KDP will not be suitable for the next generation of super lasers, as the UK-based Hiper project is discovering.


The UK is in the early stages of commissioning its own giant laser for laser inertial confinement nuclear fusion. Unlike the US and French projects, both of which were designed at least in part to study the conditions that arise during a nuclear detonation, the Hiper project will focus specifically on the feasibility of energy generation from nuclear fusion.

Chris Edwards, project manager for the project, perceived the recent news of NIF’s early success as ‘very encouraging’ for the future of the Hiper project. ‘Clearly, with a project like that, nothing is 100 per cent guaranteed when you start to build it,’ he says. ‘The physics is not 100 per cent, and the technology is not 100 per cent either. The results to date, however, show that the technology is pretty much there, and some of the minor question marks hanging over the physics seem to be in retreat.’ He adds that the NIF’s preliminary firings represent a ‘double first’, in that not only was a pulse of unprecedented energy produced (1MJ), but 95 per cent of that energy was shown to have been absorbed by the target. ‘As you can imagine, there are all sorts of unhelpful processes that can go on to stop that light getting absorbed, and a lot of work (computer simulations in particular) has gone into looking at ways of controlling the processes that we can’t stop, and stopping the processes that we can stop. The results to date show that that work was at least as successful as anyone had dared to hope it would be,’ he explains.

A project such as Hiper is not cheap. Edwards describes some of the requirements to securing funding for the giant laser: ‘Three things have to be demonstrated before we would be able to make a very convincing case for construction of Hiper: the first is NIF ignition; in a sense, that’s all NIF has to do. Second, the laser has to be efficient. Laser ignition and recycling the power to fire the next laser shot has to be sufficiently efficient to make sense on an energy basis. NIF itself is 0.1 per cent efficient, which is far too low to make commercial power generation feasible. We therefore need a laser driver that is between 10 and 20 per cent efficient,’ he says. The final component is repetition rate. ‘If NIF doesn’t ignite a fuel target, it can fire a couple of times a day,’ says Edwards, ‘and if it does ignite the fuel, then it has to be left alone for five or six days for radiological effects to subside. In contrast, a commercially viable laser would need to fire at 10Hz or more.’

The next phase of the Hiper project will aim to demonstrate that these demands can be met on the scale of a component beam line, the energy of which will be in the order of 10kJ. However, if this beamline is to achieve 10Hz pulse repetition rates at efficiencies of 10-20 per cent, then some innovative technology is required. Both NIF’s and the LMJ’s lasers are pumped using a series of xenon flash-lamps. In order to increase the efficiency of the systems, engineers working on Hiper plan to use laser diodes to pump the gain medium, as these are far more efficient. German laser diode manufacture Dilas has supplied high power laser diodes for pumping of various high energy class lasers for fusion and research applications. Dr Joerg Neukum, a director at the company, explains why diodes are preferable in large lasers: ‘Flash lamps deliver a broad spectrum of light, but only a small fraction of the spectrum is absorbed by the dopant ions in the gain material. The larger part of the spectrum turns into heat causing a poor efficiency of lamp-pumped lasers. In addition the heat deposited causes thermal lensing, hence such lamp pumped systems give a sub-optimal beam.’

Edwards also believes that Hiper must be diode-pumped in order to meet efficiency targets, although he hopes that the cost of the components reduces before construction of the facility begins: ‘The costs of building Hiper at current diode prices would be high, but there’s good reason to believe that the cost of laser diodes will reduce by factors – by a factor of 10, at least. The reason for this is that, currently, these diodes are made on a small scale with lots of human intervention – but Edwards hopes that, once the number required is sufficient, increased demand will justify automated production, leading to reduced cost, increased reliability, and longer lifetimes. A project such as Hiper would certainly increase demand for these components dramatically.

While efficiency is vital for Hiper, constructors of high-energy laser projects are also looking for reliability. German firm Lastronics supplies diode-based pump modules to large laser projects, including the experimental Polaris laser (Polaris being an acronym for petawatt optical laser amplifier for radiation intensive experiments), at Jena University in Germany. According to Dr Thomas Töpfer, general manager of the company, constructors of these giant lasers now look for reliability over innovation when it comes to their components. ‘You can no longer earn scientific paper writing about a petawatt-class laser amplifier,’ he says. ‘Nowadays, people see these projects as tools. In the early days, the lasers themselves were a novelty, but now users are more concerned with uptime.’

While uptime is of chief importance in a ‘tool’ laser, a scientific proof-of-concept such as the Hiper beamline precursor will require many innovative components in order to meet its targets of high repetition rate and high efficiency. Robertson, from Gooch & Housego, has attended planning meetings for the project, and believes that all aspects of the technology are ‘up in the air,’ – nothing is decided. ‘Hiper will probably need a gain material that doesn’t heat up as much as Nd:Glass,’ he says, by way of example, although he adds that Nd:Glass may be suitable at higher repetition rates.

There are certainly many challenges inherent in developing the materials and components for a project on this scale, but the manufacturers are all eager to be involved at every step of the process, as Robertson explains: ‘We want to ensure that, going forward, we are involved in Hiper. If the project needs novel materials, we will be in a position to grow them.’

Edwards believes the effort will be more than worth the promise of clean, renewable power: ‘There’s the joke about how nuclear fusion is 50 years away, and it always will be 50 years away. Well I believe that once NIF achieves fusion, those 50 years will begin to count down.’