What happens when light is pushed to extremes? Focused laser beams are the best way to concentrate and control energy. Lasers in the next decade will have powers exceeding a million billion watts—equivalent to the entire electrical power consumption of Earth concentrated in a single beam of light for a millionth of a billionth of a second. The enormous electric fields present at the focus of one of these beams completely overwhelm the forces that bind electrons in atoms and molecules, leading to exotic states of matter that are usually found only in stars, hydrogen bombs, or particle accelerator collisions. A different kind of extreme is the wavelength of the light. During the next decade, new facilities will use electron beams from accelerators to create x-ray lasers a billion times brighter than our best present sources. These brilliant x-ray lasers will be used to irradiate complex biological molecules with a brief x-ray flash, thereby capturing crucial details about their shape in order to learn what makes them so efficient as they carry out the processes of life. Still a third extreme is the speed of light. New methods have recently been found to reduce the speed of light to be nearly at rest in a material with very little loss in the process (Box 4–1). During the next decade, scientists and engineers will use these new tools to unravel the inner workings of nature and the nanoworld, while also harnessing extreme light for many important technological areas, from more powerful x-ray microscopes to faster drug discovery, and from alternative energy sources to quantum information.
EXTREME X-RAY LASER LIGHT
In the 20th century, short-wavelength light from synchrotrons or lasers in the ultraviolet or x-ray regimes enabled the visualization of the crystalline structure of proteins, the imaging of cells in three dimensions, the study of the electronic structure of superconducting materials, and the creation of highly excited states of atoms, molecules, and clusters. This work included the use of x rays to uncover the double-helix structure of DNA. Indeed, thousands of scientists worldwide
Coherent manipulations of vapors of three-level atoms have led to the observation of remarkable phenomena, such as electromagnetically induced transparency, slow light, and nonlinear optics, at very low (approaching single-photon) light levels. These phenomena originate in the subtle interplay and exchange of optical and atomic coherence. For example, in slow light, optical excitations, which propagate at the speed of light in free space, are reversibly transferred into atomic excitations in a vapor. However, these atomic excitations propagate at a fraction of the speed of light. Once in the vapor, these excitations can be monitored and manipulated. They can be subsequently restored (if desired) to optical excitations. Crucially, the transfer process preserves excitation amplitude and phase information. Envisioned applications include signal processing elements such as delay lines, taps, and bandwidth compressors. As these processes can function at the single-photon level, they also enable new approaches for manipulating and storing quantum information.
Recent work has demonstrated the ability to delay light in optical fibers for applications in fiber-optic communication networks. Slow light would be very useful in all-optical routers, which are used in communication systems to direct information from one point to another. Current routers convert optical information into an electronic form (a so-called communication bottleneck), while an all-optical router would eliminate the optical to electrical to optical conversion and greatly speed up the process. An all-optical router would require an optical buffer—a device that would function as temporary optical storage—to synchronize data packets effectively. A slow light device would accomplish this function. Researchers have recently demonstrated a more than 300-fold reduction in the group velocity of an optical pulse propagating on a silicon chip by using an ultracompact photonic integrated circuit with a silicon photonic crystal waveguide. Many view slow light as an important aspect of our future capability to process and transport information optically: Photonic crystals may be the key to that future.
continue to use short-wavelength light for an astonishing array of applications in basic physics and chemistry, in biology, in materials science and engineering, and in medicine.
New advances in atomic and optical physics are creating brilliant bursts of x-ray beams with laserlike properties. These bright, directed x-ray beams can be focused to the size of a virus and are fast and bright enough to capture the complex dance of atoms within molecules or—even faster—the fleeting motion of electrons within atoms and molecules. These extreme strobe lights, with x-ray vision, will provide a direct view of the electronic and structural changes that govern biology and nanoscience at the molecular level. Scientists have never had such a window through which to explore the nanoworld.
Advanced x-ray sources will be developed through the combined efforts of scientists in universities, national laboratories, and industry. Their scale will range from tabletop systems designed for very short pulses of soft x rays to large national
facilities capable of generating brilliant high-energy pulses of hard x rays. The progress in both pulse duration and pulse brightness is dramatic, both for x rays and for visible lasers, as shown in Figure 4–1.
TABLETOP SOURCES OF X RAYS
Table-sized sources of x rays, such as the ones used for medical or dental x rays, have been around for a century. Although tremendously useful for medical diagnostics, a major disadvantage is that they produce weak pulses of quite long duration (see Figure 4–1 for source brightness comparisons). This combination—dimness and duration—means that the tabletop x-ray machines of the 20th century are useless for viewing very small samples or fast events such as chemical changes in a molecule, which become blurred if the x-ray pulse is longer than a trillionth of a second.
Two new types of table-sized soft x-ray lasers have been developed in the past 10 years. (Box 4–2 discusses the difference between soft and hard x rays.) The first uses ionized atomic plasmas as the lasing medium to generate highly monochromatic and directed laser beams at wavelengths from 11 to 47 nm. Figure 4–2 shows one such laser. These devices are already in use for high-resolution spectroscopy and microscopy, and their ruggedness and relatively low cost may lead to industrial applications in next-generation microchip mask manufacturing or in nanomachining.
Another exotic new source of coherent, laserlike beams at soft x-ray wavelengths uses an extreme version of nonlinear optics (see Box 4–3). By focusing an intense femtosecond laser into a gas, the electrons in the gas atoms are driven so nonlinearly that high harmonics of the fundamental laser are emitted as coherent, laserlike beams at short wavelengths. The photon energies released by this high-harmonic generation process can span the entire ultraviolet and soft x-ray region of the spectrum, up to photon energies of kiloelectronvolts and higher. During the next decade, these tabletop extreme light sources may be used for compact microscopes capable of imaging, with unprecedented time resolution, the complex nanoworld within a single cell or for probing the behavior of materials and interfaces on nanoscale dimensions. Furthermore, the high-harmonic x-ray emission process itself is extremely brief, much shorter than the laser pulse duration. This means that by using very short laser pulses, 5 femtoseconds or less, it is possible to create x-ray beams with ultrashort, subfemtosecond (or attosecond) durations. Pulses generated in this manner are the shortest strobe lights that we can generate to date—short enough to capture some of the fastest events in atomic and molecular science. As we will read in Chapter 5, studies using these pulses will add to our fundamental knowledge of how electrons, atoms, and molecules respond to light.
X Rays—Soft or Hard?
X rays originally referred to radiation that emerged after exciting very tightly bound electrons in atoms. These days, the term “x ray” can mean any source of radiation between ≈250 eV (4 nm) and 50 keV (0.02 nm). Sometimes the definition extends to encompass vacuum ultraviolet radiation at 50 eV (20 nm). X rays are classified as soft or hard depending on the energy of the x-ray light particle (photon); the energy determines how far the photon will penetrate matter. Higher energy photons (harder x rays) travel deeper into matter before they are absorbed. The commonly accepted dividing line between soft and hard is at about 1 kiloelectronvolt of photon energy, or around 1 nm wavelength. To get a feel for what this means in practice: Dental x rays and radiation therapy x rays are hard; solar flares produce both hard and soft x rays; and the gaseous plasmas in arc lamps and welders produce mainly soft x rays. These light sources may be used for next-generation microlithography, where the energy needs to be absorbed in a very precise nanopattern in a very thin layer (see Figure 4–2–1).
In the 20th century, optical scientists harnessed a remarkable property of certain transparent crystals: When laser beams containing photons of one frequency, such as those corresponding to visible or infrared light, pass through these crystals, they fuse together in twos and threes to form new photons whose frequency is the sum of the photons in the original group. This leads to doubled or tripled frequency of the original laser and a correspondingly shorter wavelength. This traditional nonlinear optics extends laser light throughout the visible, infrared, and ultraviolet regions of the spectrum.
Extreme nonlinear optics happens when we turn up the focused laser power and electrons are literally ripped away from atoms by the laser field. These same electrons then gain energy from the laser field and liberate this accumulated energy as a high-energy photon when they are forced to recollide with the atom. This process, called “high-harmonic generation.” is a totally new photon conversion process. Instead of the two or three laser photons that are added together in traditional nonlinear optics, high-harmonic generation combines tens to hundreds of visible laser photons together, to generate laserlike beams with photon energies up to the kiloelectronvolt regime.
Although we can think of high-harmonic generation as an extreme version of nonlinear optics, this process of stripping away electrons from an atom is really a quantum phenomenon and is deeply affected by the quantum wave nature of the electrons as they move under the influence of the laser. Figure 4–3–1 shows the quantum wave of an electron as it is being gradually stripped away from its parent atom and accelerated by a strong laser field.
Many experimental and theoretical challenges will have to be overcome by scientists and engineers during the next decade to reap the full benefit of these tabletop x-ray sources. With further advances in extreme nonlinear optics, these sources have the potential to be much brighter and to span a larger photon energy range while generating designer femtosecond- or attosecond-duration pulses. Hybrid x-ray
sources that combine the benefits of different kinds compact x-ray sources have just been demonstrated, paving the way for further advances in extreme light.
EXTREME X-RAY LIGHT SOURCES AND THE WORLD’S FIRST X-RAY LASER FACILITY
The largest x-ray laser currently under construction in the United States and indeed the world, and the first scientific user facility for x-ray lasers, is the Linac Coherent Light Source (LCLS) x-ray free-electron laser (XFEL) at the Stanford Linear Accelerator Center (SLAC)1 (see Figure 4–3). As described in Box 4–4, the LCLS (and its sister projects being planned in Europe and Japan) will produce an x-ray laser beam of sufficient brightness to illuminate and capture pictures from single biomolecules (see Figure 4–4). The molecular architecture of a complex biomol-
Structural Biology: Novel Approaches to the Study of Macromolecular Structure Using X-Ray Free-Electron Lasers
X-ray photons from synchrotron storage rings have revolutionized structural biology. Today one can use synchrotron-based crystallography to study the intricate details (at the atomic level) of very complex biological assemblies that have been formed into a crystalline sample. These developments have primarily utilized the high average brightness and broad spectral range of current-generation synchrotron x-ray sources. These x-ray sources, however, are based on electron storage rings and are therefore constrained in the type of x-ray radiation that can be produced. For example, x rays from today’s typical third-generation x-ray sources have relatively long pulse durations, ranging from tens to hundreds of picoseconds. Moreover, there are a relatively small number of coherent photons in the hard x-ray regime. Therefore, use of a crystal containing millions of copies of the molecule (thus greatly amplifying the magnitude of the scattering effect) is required to determine the structure. These limitations result in two important scientific barriers. First, with present-day x-ray sources, structure determinations with atomic or near-atomic resolution can only be performed on biomolecules (for example, proteins) that can be crystallized. Second, only static or very slowly evolving structures can be measured; however, the processes underlying biological function involve dynamically evolving molecular structures.
In contrast, the extreme XFEL sources that use linear electron accelerators have the potential to produce much shorter and more brilliant x-ray pulses. These extreme x-ray laser sources take advantage of true laser amplification to generate coherent beams that are exquisitely directed and focusable. The promise of these so-called fourth-generation XFELs, such as the LCLS at SLAC, is to open up a completely new realm of x-ray science, enabling a new era of single biomolecular and nanostructure determination as well as the ability to study structural dynamics in materials and chemical/biological systems.
The challenge of developing completely new approaches for noncrystalline, atomic-level molecular imaging is, however, formidable. Many problems must be overcome for new approaches to succeed and become practical. Therefore, a multidisciplinary approach is mandatory, requiring a collaboration of the finest minds and most talented experimenters in structural biology, AMO physics, mathematics, statistics, laser science, and accelerator physics to accomplish the integration of theory and simulations, novel sample injection schemes, high-speed x-ray detection, and new algorithms for data analysis and visualization. The benefits of success for science and society are enormous. We could rapidly and routinely study all biomolecules, including those that are difficult or impossible to crystallize. Freed from the limitations imposed by crystallization, scientists will be able to study many membrane-bound proteins as well as the large molecular machines responsible for many aspects of cellular function.
ecule can be determined from a series of such pictures. While this alone would be an extraordinary accomplishment, XFELs hold additional promise. Unlike current synchrotron x-ray sources, XFEL light will also produce very short bursts (tens to hundreds of femtoseconds) of brilliant x-ray light. In fact, the bursts can be shorter than the movement of the atoms making up the biomolecule and eventually short enough to capture the molecular structure before the molecule explodes as a result of the bright x-ray flash. Furthermore, if the molecule is involved in some chemical
reaction (for example, a chlorophyll molecule as it harvests light), it maybe possible to chart the course of all of the atoms in the molecule as it changes.
AMO Contributions to Single-Molecule Imaging
Several fundamental challenges of time-resolved, single-molecule imaging involve AMO physics. The very high radiation damage to a single biomolecule from a focused x-ray laser beam of some trillion x rays is far beyond anything known in protein crystallography. Even the fundamental mechanisms of damage at such high intensities are not well understood and relate to basic questions in AMO physics such as these:
Are there important new nonlinear damage mechanisms?
Can the coherence of the x-ray laser change the character of the damage?
Can we find ways to lessen the effects of radiation damage on imaging by shortening the duration of the laser pulse or by changing the properties of the beam?
These issues affect the quality of the x-ray diffraction pattern that will be analyzed to obtain the molecular structure. It is known that the initial interactions between the x-ray beam and the molecule involve ionization (removal of negatively charged electrons) and that the accumulating positive charge generates a tremendous force within the molecule that gives rise to what is termed a “Coulomb explosion” (see Figure 4–5). What has been realized recently is that if extremely short x-ray laser pulses (tens of femtoseconds or less) with sufficient brilliance per pulse (~1012 photons) are used, then an individual x-ray diffraction pattern could be recorded from the macromolecule in the gas phase before radiation damage manifests itself and ultimately destroys the molecule by literally blowing it apart. These results come from theoretical simulation of the complex behavior of the atoms within the molecule before, during, and after absorption and scattering of the x rays and the subsequent Coulomb explosion. There are a number of factors that influence the behavior of molecules under such extreme conditions: Improved theory and simulation techniques need to be developed to better understand them and how they may limit accurate structure determination. The hydrodynamic codes used in understanding plasmas and nuclear events are being adapted to model these “molecular explosions.” These simulations need to be benchmarked against experimental data as soon as such data become available.
To determine the three-dimensional structure of a noncrystalline biomolecule using LCLS, a large number of individual two-dimensional x-ray diffraction pat-
terns must be recorded, classified, and averaged. Owing to the relatively weak signal that will come from the scattering from a single biomolecule, all other background contributions must be minimized. One concept being developed is to inject the biomolecules as a molecular beam into the XFEL beam. This injection process is already used and known to work for determining molecular weights using mass spectrometry. In the bioimaging experiment, thousands of two-dimensional images would be recorded sequentially from individual molecules, their orientations determined, aligned, and averaged to produce a single three-dimensional molecular diffraction pattern. Because the individual two-dimensional patterns will be relatively weak, new reconstruction algorithms must be developed that work at the minimum possible signal-to-noise levels. Ultimately, the three-dimensional diffraction pattern will be converted into the three-dimensional molecular structure.
Single-molecule imaging must meet and overcome some formidable challenges. To reduce background, a single molecule must be held in containerless packaging so that only the sample will be imaged. Techniques developed for electrospray can be adapted to the injection of single biomolecules into vacuum. Equally important are molecular dynamics simulations to study how the biomolecules behave under these high-vacuum conditions and how the water structure on their surfaces or other structural elements are affected. There is also significant advantage if, rather than recording images from randomly oriented biomolecules, one can use physical methods to induce alignment (or partial alignment). For example, as discussed in Chapter 5, very strong laser fields have been used to simultaneously force all three axes of a small molecule to align along given axes fixed in space and inhibit the free rotation. The adaptation of these techniques to macromolecules is an area of strong interest.
TESLA Test Facility Early Results
With the recent start of operations of the soft XFEL in Hamburg, Germany, measurements with lower-energy x-rays will soon become possible. The regime of such high-peak-power x-ray pulses has never been accessed before and, while the physics of models and simulations appears to be correct, such experiments are very interesting as they will provide the first direct experience relevant to the eventual use of LCLS for atomic resolution imaging of nonperiodic materials. Figure 4–6 shows an example of single-pulse imaging using vacuum ultraviolet (VUV) light.
Inner Shell Atomic Multiple Ionization
The LCLS x-ray laser beam will be the first x-ray source in history to be able to generate the same extreme focused powers that can be accessed by current-
generation high-powered lasers. While we can make predictions about what happens to matter as we turn up the x-ray powers to such extreme levels, we also expect to initiate unusual physical phenomena that have not previously been studied, or even observed, anywhere on Earth or in the universe. Chief among these new physical effects is rapid multiple ionization of the most deeply bound electrons. This will lead to the creation of “hollow” atoms and ions—that is, species with two or more electrons missing from the highly bound region at the very center of the atom, next to the atomic nucleus. In the past, such atoms have been produced in high-energy ion-atom collisions, where the ability to study them in detail has been limited because they are created at random times by collisions. Small numbers of hollow atoms have also been produced by an extremely weak process in which a single x-ray photon causes the ejection of the two most tightly bound electrons in the atom, and hollow atoms of lighter elements such as lithium have been produced using light from a combination of lasers and synchrotrons. Such atoms store enormous amounts of potential energy and represent extreme matter in a truly exotic form. The decay of these hollow atoms also provides considerable insight into the correlated motion of the electrons remaining in the atom.
The creation of large numbers of hollow atoms on demand using photons has been impossible to date. Although the intense fields created at the focus of high-powered visible lasers are truly enormous, these lasers tend to preferentially strip away the most weakly bound electrons from the atom—much like peeling an onion. In contrast, the LCLS will produce hollow atoms directly from neutral
atoms through inside-out absorption of multiple x-ray photons. This relies on three properties of the x-ray laser, which exist in combination only in this new tool—namely, the short wavelength of the photons, the brevity of the pulse, and the high pulse intensity. The short-wavelength x rays have a much stronger effect on tightly bound “core” electrons than on the weakly bound outer “valence” electrons in an atom. Almost all of the x rays absorbed by the atom remove the inner shell electrons rather than the valence electrons. This property, of course, is shared with all other x-ray sources—inner shell photoemission is a well-known x-ray phenomenon. However, in all experiments to date, the hole left in the atom by the departing electron is filled in so rapidly that any subsequent probing of the atom will not see the hole but will only see the absence of the outer electron that filled the core hole.
This is where the very high intensity of the LCLS comes into play. The intensity is so high that the atom can absorb a second x-ray photon and eject a second electron in a few femtoseconds, even before relaxation has refilled the core vacancy left by the first departing electron. The result of all this is that photoabsorption should be quite different at the LCLS due to rapid multiple ionization. This phenomenon is not only of purely academic interest. The dynamics of many other processes, including the single-molecule imaging described above, will be different due to the formation and presence of these exotic doubly excited states. The study of these phenomena is a high priority at LCLS.
X-Ray Nonlinear Optics
Coupled with the high intensities of these new extreme light sources of x rays is their high coherence. Coherence, briefly stated, is the property that makes a wave move in a regular, predictable fashion. The wake of a boat is an example of a coherent water wave; so is a tsunami. On the other hand, the chop that occurs on a lake on a gusty day has very low coherence, since the wave crests are neither regular nor predictable. The most incoherent light comes from thermal sources like the sun. The most coherent light comes from lasers, and the new extreme light sources—both tabletop and large facilities such as the LCLS—produce the most coherent x rays ever made.
Coherent waves can drive matter much more effectively than incoherent waves—think about a surfer! We have already explored high-harmonic generation, which relies on the coherence of a visible laser beam to produce very high harmonics at soft-x-ray wavelengths. What if we replace the visible laser with a coherent source of x-rays? What kinds of superextreme nonlinear optics would be possible then?
The first nonlinear optical phenomenon to be studied in isolated atoms will probably be multiphoton ionization. This is different from the multiple ionization discussed in the previous section: Instead of ionizing multiple electrons, multiphoton ionization refers to the pooling of more than one photon to ionize a single electron from an atom. Two-photon ionization has much in common with second-harmonic radiation, discussed in a previous section of this chapter. Its observation will be a major milestone in the new field of x-ray laser science.
Many other exotic phenomena may follow the development of intense x-ray lasers. High-flux, short-wavelength radiation will have dramatic effects on materials, which might even lead to new kinds of x-ray lasing mechanics in solids.
Summary of Extreme X-Ray Light Sources
In summary, during the next decade, the availability of coherent x-rays from XFELs, together with the potential for a real breakthrough in x-ray imaging, is setting the stage for a unique and privileged period of discovery in x-ray science that will impact structural biology and other fields where nonperiodic nanostructured materials are central. Atomic-resolution imaging of noncrystalline biological materials appears to be feasible, with many of the key concepts being demonstrated experimentally or with detailed models where AMO physics has a strong role to play. Tabletop extreme x-ray sources will allow us to capture the fleeting motion of electrons in atoms, molecules, and solids for the first time, challenging theory and opening up new possibilities for manipulating matter on unprecedented subnanoscale levels. These tabletop extreme x-ray sources will also bring the source to the application, enabling widespread use of next-generation microscopes and spectroscopies to probe materials with unprecedented spatial and temporal resolution.
ULTRAINTENSE LASERS: USING EXTREME LIGHT SOURCES TO HARNESS EXTREME STATES OF MATTER
Think of a laser beam focused onto a spot on a solid surface smaller than the diameter of a human hair. As we begin to increase the laser pulse energy, we first vaporize the spot to create a crater. At still higher energies, the laser continues to heat the vapor from the crater until atoms and molecules explode into electrons and ions, forming an ultrahot, ionized plasma with a temperature of millions of degrees, similar to a star’s interior. The laser pulse energy can then be turned up even higher, so that the laser light pushes the electrons and ions around so violently that they accelerate to relativistic velocities close to the speed of light and hardly interact with each other at all. Finally, lasers can now produce pulses of such in-
credibly high energies that empty space can be ripped apart, to form new matter and light where none previously existed. Figure 4–7 illustrates the rapid increase in laser intensities achievable over the last 50 years.
In the 20th century, scientists explored these exotic states of matter by generating enormous focused laser powers to try to understand extreme states of matter. The
challenge in the 21st century is to harness and control such extreme states of light and matter. The biggest accelerator currently being built, the LHC at CERN, requires an accelerator ring 28.5 km in diameter to generate two counter-rotating 7-TV proton beams at great expense and size. Can we design tabletop particle accelerators that can accelerate electrons to gigaelectronvolt energies in a distance no greater than the length of your hand (≈10 cm)? Box 4–5 discusses progress toward this goal. Can we achieve nuclear fusion at the focus of powerful laser beams to harness the atom as a source of clean, abundant energy? Can we use lasers to create energetic x-ray, electron, proton, and neutron beams that will lead to higher-resolution mammograms or allow engineers to predict when an aircraft wing is about to fail?
NIF and Other Large Facilities
At the present time, the National Ignition Facility (NIF) being constructed at Lawrence Livermore National Laboratory is expected to enable scientists to create unique, high-energy plasma conditions in the laboratory that can only be found on Earth during the detonation of a nuclear weapon (see Figure 4–8). The 192-beam, 1.8-million-Joule laser system will address several important scientific questions—some of which are related to producing and understanding basic high-energy-density science, some to the use of fusion as a viable energy source for the world, some to the long-term stability of the nuclear weapons stockpile, and some that will help scientists to understand spectacular astrophysical observations. Lasers with peak power in excess of a thousand trillion watts (1 petawatt) are currently being constructed in the United States, Japan, Britain, France, and Germany.2 Lasers with powers a hundred times higher than these will be possible in the coming decade. The scientific opportunities enabled by ultraintense lasers to understand, control, and use high-energy-density states of matter are diverse and very exciting.
At the ultrahigh intensities now achievable with the current generation of lasers, enormous electric fields can accelerate electrons to very high energy. Ultrafast, ultra-high-intensity laser production of fast electrons is currently a promising candidate to aid in the ignition of an imploded fusion pellet by externally heating the fusion fuel. This could permit a dramatic leap in the technology of controlled nuclear fusion research. Initial results from Japan and elsewhere are promising. While the prospect of achieving fusion gain high enough for viable energy production is challenging with conventional fusion approaches, fast ignition with intense
Using Lasers to Accelerate Electrons
Directed laser beams provide a very powerful way of concentrating energy efficiently—provided this energy can be transferred to a particle such as an electron so that it can be accelerated to teraelectronvolt (1 TeV=1012 eV) energies. An electron in an intense laser beam will behave much like a cork on water as a huge wave passes by, bobbing up and down on the wave but not really going anywhere. More than two decades ago, Tajima and Dawson solved the problem of how to transfer the laser beam energy to an electron: An intense focused laser beam can create a giant fast-moving plasma wave as it passes through a gas and ionizes it. The electron can then surf on the plasma “tsunami” and be accelerated to enormous energies. The laser beam in essence generates a large-amplitude plasma density wave that ripples through the plasma like a wake behind a boat. The longitudinal electric field associated with this density wave can easily be in excess of tens of gigaelectronvolts per meter (1 GeV=109 eV), which is more than three orders of magnitude beyond conventional accelerator technology.
Although experiments that tested this concept produced large gradients for accelerating particles, until 2004 the accelerated electron beams had 100 percent energy spreads with only a small fraction of electrons at high energy. However, in exciting advances independently achieved by three different groups worldwide in the United States, the United Kingdom, and France, the generation of 100-MeV-class electron beams with narrow energy spreads was demonstrated using laser-plasma accelerators. Such high-quality, narrow-energy-spread electron beams are necessary for exploring several scientific frontiers, such as for generating high-brightness x-ray sources, for producing electron and positron beams with energies in excess of 1 TeV, for creating particles from the vacuum, and for testing the fundamentals of quantum and classical electrodynamics.
Figure 4–5–1 shows an experiment in which an electron bunch of around 2×109 electrons was accelerated, and the bunch length was inferred to be near 10 femtoseconds. The limitation on the maximum achieved energy to date stems from the fact that the plasma density used in these experiments was relatively high (1–4 ×1019 cm−3), causing the laser pulse to move relatively slowly through the plasma. Much like a surfer on a wave, accelerating particles move forward on the wave and can ultimately even overtake the laser pulse, thus terminating the acceleration and limiting the energy gain.
The next decade will realize gigaelectronvolt- and teraelectronvolt-class electron beams using next-generation lasers, lower plasma densities, and longer guiding distances for the laser.
lasers is more likely to achieve the high gain needed to realize energy production with fusion.
High Energy Density Science: Laboratory for Extreme Conditions in the Matter-Filled Universe
Although plasmas represent the most abundant form of observable matter in the universe, our understanding of this state of matter is remarkably incomplete. This is particularly true of plasmas at very high energy densities. The high-energy lasers described above will provide a laboratory to study the physics of extreme plasma regimes with the promise of learning more about the physics of some of the most important, but inaccessible, plasma environments in the universe—such as
the cores of neutron stars and white dwarfs or the plasmas near black holes. High energy density (HED) plasmas also exist in nuclear explosions, or in plasmas that might be controlled to produce energy from nuclear fusion. HED science therefore is an important frontier field of modern science. In addition, these HED plasmas may be controlled for application in many important technological problems, ranging from accelerating particles to high velocity to developing new and precise imaging technologies.3
Ultraintense lasers provide a controlled means for creating and studying these
For a review of opportunities in HED physics, see NRC, Frontiers in High Energy Density Physics: The X-Games of Contemporary Science, Washington, D.C.: The National Academies Press (2003), available at <http://fermat.nap.edu/catalog/10544.html>.
unique states of HED matter. Laser technology has advanced in recent years to the point where light pulses with peak powers of tens to thousands of trillions of watts are possible. These ultraintense lasers can essentially concentrate the equivalent power of the entire electrical grid of the United States onto a spot only a tenth of a human hair in diameter (though only for an instant). It is this concentration of such enormous powers in a laboratory setting that now allows the controlled study of matter in HED states.
One of the greatest challenges in the field of HED science is a theoretical one: to develop conceptual models that can describe the behavior of these new exotic states. Although HED matter exhibits characteristics that range from those expected of a plasma to those that are like condensed matter, it frequently behaves like neither. Intense lasers now allow us to study this behavior and to craft theories that describe it.
For example, a high-energy ultrafast laser can heat solid matter on a timescale much faster than the material expands. This heating at high density produces very high pressure states of matter, in some cases with pressure well above 1 billion atmospheres. Matter in these states is normally found only in the interiors of large planets, dense stars, and nuclear detonations. Thus, controlled laboratory experiments that can inform the study of stars or enhance the nation’s security are now possible.
The extremes in temperature that can be accessed with intense lasers now make possible laboratory experiments that could aid in understanding exotic as-
trophysical events. For example, it is believed that plasmas composed of a mixture of matter and antimatter may exist near black holes. Some day, intense lasers may even permit the creation of small amounts of such hyperenergetic matter in the lab. Other scientific frontiers that will be uncovered in astrophysical research are discussed below.
Accelerating Particles with Light
The preceding sections have just described how the enormous electric fields present in the plasmas created by superintense lasers can accelerate electron beams to multi-GeV energies within a few centimeters.
The Energy Frontier
The high-energy frontier for particle physics will require particle energies well in excess of 1 TeV for studies of fundamental properties of matter. The International Linear Collider (ILC), a superconducting radio-frequency accelerator, has been proposed by the international high energy physics community as a primary tool for such studies and may be built in the next two decades. However, laser accelerator science has advanced very rapidly in the past 5 years, and this holds out the prospect that new accelerator technologies based on lasers may play a part in future high-energy accelerators. There is no clear path for this at present, but the technical problems are understood, and some possible avenues for implementing this vision have emerged.
One possible route to achieve this would be to stage a hundred or more smaller acceleration stages (10-GeV “modules”), each one driven by a synchronized petawatt-class laser. The overall length of such an accelerator would be 200–500 m, a fraction of the distance that would be needed to accelerate a particle to these enormous energies using conventional approaches. The cost would also be a fraction of the projected cost of the ILC. However, there are very significant challenges ahead: To achieve the required high luminosity, high-repetition-rate lasers are required, and to achieve high wall-plug efficiency, a revolutionary new approach in the design and implementation of high-energy laser systems will be required. This is one of the grand challenges for laser science and technology.
The Ultrafast Source Frontier
The bright femtosecond electron bunches that are created using laser-based accelerators can be used as new probes of atoms, molecules, and materials. The electron bunches themselves can be used for time-resolved electron beam dif-
fraction, which is complementary to x-ray diffraction. Such experiments require electron beams with energies of 0.1 to 5 MeV, 106 to 107 electrons per bunch, low energy spread, and good directionality (emittance). Experiments are under way to demonstrate such beams. A few electrons present in a plasma created by an intense femtosecond laser can receive a boost in energy from an additional laser pulse that enables them to catch the fast-moving plasma wave. By carefully controlling the wave amplitude to avoid wave breaking (and hence self-trapping), simulations indicate that electron beams suitable for electron diffraction experiments could be produced.
Other sources of radiation can also be generated from energetic electron beams accelerated by lasers. For example, broadband coherent THz fields (1 THz=1012 Hz), corresponding to wavelengths of 100 microns, with high fields in excess of 1 MeV/cm could be generated. Such fields can be used for spectroscopy and for probing and/or exciting various materials, such as superconductors, magnetic materials, and nanostructures. Another exciting and important prospect is to demonstrate the next generation of x-ray tubes by crashing the bright electron beam into a target. Because the number of electrons accelerated by an intense laser could be orders of magnitude brighter than currently used in a conventional x-ray tube, the generated x-rays beam could also be very bright and energetic, with polarized photon energies from keV to multi-MeV by proper choice of electron energy, scattering laser wavelength, and geometry. Looking even further into the future, even more intense x-ray beams can be generated using XFELs. However, in some envisioned configurations, these devices require a seed electron beam. If the experimentally demonstrated low energy spread from laser-accelerated electron beams is maintained as the mean energy increases from 100 MeV to a few GeV, and if the electron beam divergence is maintained, intense 100 nm radiation could be produced (1013 photons/pulse) from a high-gain XFEL driven by a laser accelerator. Several groups around the world are pursuing this goal. As discussed above, such a compact source of ultraintense, femtosecond x rays would enable many experiments involving molecular, atomic, and biological systems on the natural timescales of atomic motion.
The Intensity Frontier—Sparking the Vacuum
A third frontier is the intensity frontier, where electric field strengths in excess of the Schwinger critical field limit are generated. At this field strength of 3×1018V/m, the vacuum is unstable, and quantum electrodynamics predicts that electron-positron pairs can be spontaneously generated from the vacuum. Indirect experimental verification of this prediction was first accomplished in landmark experiments at SLAG in the 1990s, which used the relativistic shift in the intensity and photon energy of a
laser beam colliding with the SLAG relativistic electron beam to reach the required Schwinger limit. The laser intensity required to reach the Schwinger critical field without this trick of relativity is enormous—around 1029 W/cm2. Such high laser intensities are several orders of magnitude beyond the current state of the art. This would require a megajoule-class laser pumping a meter-scale Ti-sapphire crystal amplifier. However, experiments extending the original SLAG results could study this “vacuum boiling” regime in much greater detail. A natural candidate for such an experiment would be a 10-GeV electron beam scattering from a petawatt-class laser beam. This would enable access to the exciting regime of nonlinear quantum electrodynamics by producing effective field strengths that are four to five orders of magnitude greater than those currently achievable directly from lasers. Such petawatt lasers are currently under construction.
High Energy Density Science and XFELs
The XFEL described in the preceding section is expected to have a big impact on the field of HED science. Since x rays can penetrate dense matter, the XFEL can deposit large amounts of energy uniformly over a reasonable volume of material, transforming ordinary matter into warm dense matter similar to the interiors of planets. Other laboratory sources, such as the NIF laser at Livermore, are unable to perform this function well because of the short absorption length for visible or ultraviolet light. The x-ray laser penetration depth also makes it a unique probe of exotic conditions such as those found in the center of superdense plasmas.
The Fastest Pulse: Complementarity Between Extreme Light and Extreme Particle Beam Collisions
A fully stripped uranium nucleus passing rapidly through a high-Z atom at a distance of one hundredth of a typical atomic diameter from the nucleus of the target atom applies to the inner shell electrons of the atom an electric field a million times greater than the field that an electron in the ground state of a hydrogen atom would experience. If this ion is moving at relativistic speeds (see Box 4–6), the field can grow to nearly a billion times the hydrogenic field. Accompanying the electric field is a magnetic field times thousands of times greater than the strongest laboratory DC magnetic field. This electromagnetic pulse completely destroys the normal environment of even the innermost electron. The time duration can be so short that the spectral content of the electromagnetic pulse can extend into the gamma-ray region, sufficient to remove both inner and outer shell electrons as well as to excite the nucleus. The collision can produce a distribution of nearly all excited states of nearly all ionization states of the target. The target responds by
New Opportunities in Collisions with Relativistic Heavy Ion Beams and Other New Ion Facilities
The heavy ion facilities at the Gesellschaft für Schwerionenforschung (GSI), Darmstadt, Germany, are entering a period of major facility construction and development that will place in the hands of AMO physicists an unprecedented array of tools for investigating the interactions of relativistic heavy ions with photons and matter (Figure 4–6–1). These facilities will make available to the experimentalist essentially all charge states of all ions up to fully stripped uranium at energies up to 300 GeV per nucleon (relativistic gamma of 300). The fast ions will be used directly for studies of collision dynamics in strong electromagnetic fields and of fundamental interactions between electrons and heavy nuclei up to bare uranium. Alternatively, they will be slowed, after stripping, and stored in two new heavy ion storage rings for spectroscopy and dynamics studies. Finally it will be possible to bring the ions nearly to rest and to cool them in a heavy ion trap for essentially Doppler-free spectroscopic studies of few-electron systems.
A few of the new opportunities include these:
emitting a radiation spectrum rich with lines that display physical effects specific to these high energies. It is sufficient to exceed the Schwinger limit described in the preceding section, creating electron-positron pairs.
With the rapid progress of laser and XFEL technology toward ever shorter pulses and harder photons, one must ask where the accelerator collisional approach will meet the laser approach and what information can be expected to issue from each? Collisions produce the shorter time pulses and the harder photons, but for a
given collision neither of these parameters is under the control of the experimenter. Both the timescale and the frequency spectrum of the photons depend on how close the projectile passes to the target (the impact parameter), and this cannot be dialed in experimentally. In the case of the laser, a macroscopic sample of atoms is exposed to the same pulse length and frequency spectrum, which are under the control of the experimenter. Thus the technologies are complementary, and both are likely to lead to new insights in high-intensity science.