The free electron laser (FEL) has endured wild levels of optimism about its capabilities in the last three decades as well as near elimination because of pessimism about its cost and potential.
Today, however, the FEL has emerged as a vibrant research tool with an exciting future of delivering tunable coherent light to resolve some of the most important outstanding questions in materials science, biology, and condensed matter physics.
Behind George Neil is a computer simulation image of electromagnetic fields accelerating electrons in the superconducting resonators of the European XFEL, a research facility under construction in Germany that will produce high-intensity, ultra-short x-ray flashes with the properties of laser light. (Photo courtesy DESY/European XFEL)
In 1976, when free electron laser oscillation was first demonstrated by John Madey and co-workers at Stanford University (USA), the initial excitement was on the potential of this light source to produce exceptionally high average powers for defense and other applications. In subsequent years, funding and expectations for the FEL's use in the U.S. Strategic Defense Initiative inflated rapidly but burst with the demise of the Soviet Union and the results of meager experimental demonstrations.
In the United States, the fallout from that disillusionment lasted more than a decade with little support from funding agencies and the general belief that FELs were little more than an expensive, large, laboratory curiosity with capabilities that were exceeded by simpler, cheaper, conventional laser systems.
There was broader support for research on FELs worldwide, with more than 35 operational systems by 2000. However, even there, the FEL existed in a niche market and seemed more a source in search of an application. Even as late as 2001, reports for the U.S. Department of Energy (DOE) and the National Science Foundation (NSF) suggested little practical use for such systems except in special wavelength ranges.
That situation has now changed with considerable interest from researchers worldwide, substantial funding for national and international research facilities, and exciting growth opportunities which point to a vibrant foreseeable future in R&D, especially in materials science. Industrial and other applications remain unfulfilled, but they, too, may emerge as the technology catches pace with requirements.
What has changed in the R&D landscape that makes the FEL community so excited today?
The enabling drivers for this momentum are advances in the accelerator technology that powers the FEL. It is worth taking a moment to recall the components of a FEL and how they interrelate.
The FEL converts a small fraction of the power of a relativistic electron beam to tunable coherent light using a device called an undulator or wiggler. The undulator is an alternating magnetic field of fixed wavelength which gets one Doppler shift and one relativistic Lorentz contraction to shorter wavelength. Roughly speaking, the output wavelength is the undulator wavelength divided by four times the electron beam energy squared in MeV. Thus a 3 centimeter undulator wavelength will yield infrared output for 100 MeV electron beams and 1 Angstrom output for 15 GeV electrons.
High peak currents are required for high gain; perhaps even over 1 kA, but these currents are only provided for very short periods of 100 femtoseconds or less. If high average power is desired, the electron pulses must be repeated at high repetition rates in the MHz region and above. Gain falls with electron beam energy, so very long undulators are required for ultra-short wavelengths: 10m to even 100m.
The Second Law of Thermodynamics (via Liouville's Theorem) says you can't make a photon beam brighter than the electron beam it is drawn from. Therefore, there is a high priority on creating and maintaining very high brightness electron beams.
The electron beam is formed in an injector generally from a photocathode pulsed by a mode-locked drive laser in the visible or UV. Substantial advances in the performance of such injectors have been achieved in the last decade due to improved handling of space charge forces.
An electron beam is formed in an injector as illustrated above.
Another accelerator advance has been in the use of superconducting cavities as a means for accelerating the electrons. These virtually eliminate ohmic losses of the microwave accelerating fields and permit efficient continuous beam production. They also permit the recovery of the spent electron beam power after lasing.
Superconducting cavities made of niobium are assembled in ultraclean environments to avoid dust contamination which could lead to breakdown in the high electric fields. Photo courtesy of Robert Rimmer, Jefferson Lab.
This energy recovery operation substantially reduces microwave power requirements and provides significant system level benefits.
New Laser Systems
Two systems illustrate the range of impact that these technologies have had on the FEL field and point toward future developments and applications.
The Jefferson Lab FEL is an example of the application of superconducting radio-frequency technology. The accelerator technology used for this machine was developed to support nuclear physics research, specifically the Continuous Electron Beam Accelerator Facility (CEBAF), a 5 GeV machine used to explore our understanding of the atomic nucleus, nucleon binding, and how quarks pair up to make up those nucleons.
The electron accelerator of the JLab FEL produces a 10 mA average current beam at up to 120 MeV in energy. The JLab FEL has been used to produce up to 14.3 kW of average power at 1.6 micron wavelength in a continuous train of 150 fs bunches at 75 MHz. The system has lased from 0.7 to 11 microns and can tune over a factor of 6 or more in wavelength in a few seconds, although doing this at full power is problematic due to mirror coating technology.
Applications of such high average power UV to near infrared include not only basic science studies but possible industrial applications. For example, work at Vanderbilt University (USA) has shown that complex organics can be deposited on substrates through resonant laser ablation, a process that excites a molecular resonance of the molecule gently enough that it doesn't crack, as would be the case with conventional pulsed laser deposition (PLD). This opens the possibility of high-volume PLD production of organic video display circuitry.
At the UV end of the spectrum, a new laser micro-engineering station at Jefferson Lab is getting ready to produce picosats: 50-gram satellites with thrusters, telemetry, etc., out of machined ceramics with the physical properties (hardness, etc.) determined by the UV exposure. The FEL's high power in a tunable train of subpicosecond pulses to prevent heat stress cracks makes this the only possible laser source capable of meeting the technical need.
After converting up to 2% of the electron beam power to light, the electrons are sent through the accelerator a second time at 180 degrees out of rf phase. Their energy is converted back to microwaves at high efficiency, and when the electrons have decelerated to below 10 MeV, they are dumped in a water-cooled copper block.
This energy recovery operation is serving as a prototype for systems such as the Energy Recovery Linac (ERL) being developed at Cornell University (USA), which may accelerate beams of several hundred milliamps up to 2500 MeV to produce short, ultra-bright x-ray synchrotron light pulses for materials research. Dozens of undulators would be provided on Cornell's ERL to allow many researchers to access radiation simultaneously.
At the short end of the wavelength spectrum is the Linac Coherent Light Source (LCLS) at the SLAC National Accelerator Laboratory at Stanford University (USA) (formerly the Stanford Linear Accelerator Center). Using the last third of the SLAC linac to accelerate a high charge pulse to 15 GeV, the LCLS recently reached saturation to become the world's first 1.5 Angstrom laser.
The brightness provided by this source is extraordinary: 1033 photons/sec/mrad2/mm2/0.1% BW, eight orders of magnitude brighter than anything humankind has previously produced at this wavelength. The pulses are so intense that a substantial distance between the undulator output and the first optic must be provided to allow the beam to diverge, else the optic be destroyed by the 1 mJ pulse.
Rich Research Program
Although only a single portal of the LCLS wiggler is in use at a time, a rich research program is planned of condensed matter physics, warm dense plasmas, biological imaging (holograms of previously living cells which blow apart within 30 fs of being imaged), high field studies, etc. Despite the very aggressive specifications, the system quickly achieved its operating goals and is proving to be robust in its initial performance tests.
An aerial view of the LCLS x-ray laser. Photo courtesy of John Galayda, SLAC National Accelerator Lab.
4th Generation Light Sources
These FELs and ERLs span the range of ideas incorporated in what has come to be known in the x-ray community as Fourth Generation Light Sources (4GLS).
Over the past several decades, scientific research on storage rings (third-generation light sources) has produced greatly improved capabilities for photon beams. For example, much of our knowledge of protein structure has come from x-ray scattering measurements on second- and third-generation sources.
Storage ring synchrotron radiation sources are, however, near the end of a development path, having nearly reached ultimate performance limits for systems that recycle the electrons.
Fourth-generation light sources are not expected to displace third-generation rings but rather to augment their capability. Because the electrons do not remain in 4GLS systems long enough to reach thermodynamic equilibrium, higher beam brightnesses are achievable.
Substantially shorter pulses are also possible in linear-accelerator-driven systems so that x-ray studies of materials (such as protein) and molecular dynamics are envisioned on 10 fs timescales. Advantageously, free electron lasers on 4GLS sources can produce higher brightness than synchrotron radiation from the same beam because the electrons bunch at optical wavelengths and emit their radiation collectively (that is, the electron micro-bunches oscillate in phase so their electric fields add coherently and the radiation is proportional to the square rather than linearly with the number of electrons).
Both the NSF and the Office of Basic Energy Sciences in DOE are developing plans to exploit these new capabilities. DOE has issued a document detailing the photon characteristics desired for answering a set of grand challenges in science for the 21st Century, Next Generation Photon Sources for Grand Challenges in Science and Energy. These grand challenges were identified in a series of workshops as
How do we control materials processes at the level of the electron?
How do we design and perfect atom- and energy-efficient synthesis of new forms of matter with tailored properties?
How do remarkable properties of matter emerge from the complex correlations of atomic and electronic constituents, and how can we control these properties?
How can we master energy and information on the nanoscale to create new technologies with capabilities rivaling those of living things?
How do we characterize and control matter away fromand especially very far away fromequilibrium?
The tools required to answer these questions will involve the construction of new research facilities costing as much as $1 billion. Already such facilities are under construction in Japan and Germany to supplement the capability of the LCLS at Stanford.
It is a formidable challenge to the light source community to achieve the characteristics required, but answering these questions will occupy the best minds in materials science for many decades. The benefits that will accrue from that achievement will benefit future society in ways that cannot now be imagined.
It is an exciting time to be involved in this development.
More Laser News
The Energy Recovery Linac, or ERL, is a fourth-generation light source being developed at Cornell with support from the National Science Foundation. ERLs have the potential to generate synchrotron radiation with brightness about 1000 times greater than that of third-generation storage rings, resulting in highly coherent x-radiation.
Nikolay Zheludev and a team at the University of Southampton (UK) have reported the first proof-of-concept demonstration of a tunable, electron-beam-driven, nanoscale radiation source in which light is generated as free-electrons travel down a 'light-well,' a nano-hole through a stack of alternating metal and dielectric layers. Their research demonstrates near-infrared emission and the possibility that the concept can be scaled to other wavelength ranges by varying the periodicity of the structure. More at: arxiv.org/abs/0907.2143
- The SLAC National Accelerator Laboratory (www.slac.stanford.edu) in Menlo Park, CA, is home to a two-mile-long linear accelerator, the longest in the world. Originally a particle physics research center, SLAC is now a multipurpose lab for astrophysics, photon science, accelerator, and particle physics research operated by Stanford University for the U.S. Department of Energy. Six scientists have been awarded the Nobel Prize for work carried out at SLAC.
The SPIE Newsroom
has a 2008 video interview with Fred Dylla, former associate director of the Jefferson Lab FEL, in which Dylla explains how U.S. government investment in the FEL as a user facility has led to important commercial and medical applications. See: spie.org/dyllainterview
The Grand Challenges adopted as scientific goals by the DOE's Office of Science are general enough that the impact on society would be dramatic from answering even one of them.
For example, discovering how we can tailor atoms and energy-efficient forms of matter will directly lead to more efficient and lower cost photovoltaics.
The tailoring of material hardness through laser exposure of ceramics has direct application in the production of turbine blades where a hard exterior reduces corrosion while a softer but tough inner core gives the strength to prevent shattering of the blade at high speeds.
In the area of understanding how complex correlations result in materials properties, a whole new field of spintronics, controlling information flow and storage through the correlated spins of electrons, has developed.
In terms of producing new materials, work on photosynthesis by catalyzed chemistry on artificial systems is already underway. Why shouldn't we be able to harness solar power for the production of synthetic fuels more efficiently than living systems?
George R. Neil
George R. Neil is associate director of Free Electron Lasers at the Thomas Jefferson National Accelerator Facility in Virginia (www.jlab.org/fel). His PhD from the University of Wisconsin is in nuclear engineering. He joined Jefferson Labs to manage the CEBAF Linac construction in 1990 after 13 years at TRW Defense and Space Systems Group (now part of Northrop Grumman) in Redondo Beach, CA. He has two grown children working in software engineering in California. His wife Doreen Osowski Neil is a senior atmospheric scientist at NASA Langley Research Center. On weekends they both enjoy kayaking Virginia's coastal rivers, and he is an avid runner with 50 marathons to his credit including at least one on every continent including Antarctica
Have a question or comment about this article? Write to us at firstname.lastname@example.org