More than 5000 low-energy photons have been converted into one high-energy x-ray photon. The coherent, directed light beam was generated for the first time using a device small enough to fit on a single laboratory table. “This is the broadest spectral-bandwidth, coherent-light source ever generated,” said Henry Kapteyn, an engineering and physics professor at JILA at the University of Colorado at Boulder. He led the study with JILA colleagues in collaboration with scientists from Cornell University, the University of Salamanca and Vienna University of Technology. “It definitely opens up the possibility to prove the shortest space and time scales relevant to any process in our natural world other than nuclear or fundamental particle interactions.” This art represents an electron being ripped from an atom by a strong laser field that stretches its quantum wave function over hundreds of atomic sizes. Just as electrons accelerated in an x-ray tube emit bremsstrahlung (braking) radiation, those accelerated by a laser can emit rainbows of coherent x-rays in a laserlike beam. Such light, invisible to the naked eye, is important for being able to ‘see’ the fine details and fastest motions of the nanoworld. (Image: Tenio Popmintchev, JILA, University of Colorado at Boulder) An international team, led by engineers from the National Science Foundation’s Engineering Research Center (ERC) for EUV Science and Technology, focused intense pulses of infrared light — each just a few optical cycles in duration — into a high-pressure gas cell and converted part of the original laser energy into a coherent supercontinuum of light that extends well into the x-ray region. The emerging x-ray burst had much shorter wavelengths than the original laser pulse, making it possible to follow the tiniest, fastest physical processes in nature, including the coupling of electrons and ions in molecules as they undergo chemical reactions or the flow of charges and spins in materials. An actual image of a coherent (laserlike) x-ray beam. In contrast to the incoherent (lightbulblike) light emitted in all directions from a roentgen x-ray tube, the x-rays produced by high-harmonic generation emerge as well-directed laserlike beams. (Image: Tenio Popmintchev, JILA, University of Colorado at Boulder) “Thirty years ago, people were saying we could make a coherent x-ray source, but it would have to be an x-ray laser, and we’d need an atomic bomb as the energy source to pump it,” said Deborah Jackson, the program officer who oversees the ERC’s grant. “Now, we have these guys who understand the science fundamentals well enough to introduce new tricks for efficiently extracting energetic photons, pulling them out at x-ray wavelengths … and it’s all done on a tabletop!” The researchers used high-harmonic generation (HHG) in their experiment — a method discovered in the late 1980s when scientists focused a powerful, ultrashort laser beam into a spray of gas — and discovered that the output beam contained not only the original laser wavelength, but also a small amount of many different wavelengths in the UV region. Gas atoms ionized by the laser created the new UV wavelengths. “Just as a violin or guitar string will emit harmonics of its fundamental sound tone when plucked strongly, an atom can also emit harmonics of light when plucked violently by a laser pulse,” said JILA professor Margaret Murnane. “The laser pulse first plucks electrons from the atoms, before driving them back again where they can collide with the atoms from which they came. Any excess energy is emitted as high-energy ultraviolet photons.” The experimental setup used to create a coherent version of the roentgen tube in the soft x-ray region of the spectrum. When a long-wavelength, femtosecond laser is focused into this hollow waveguide filled with high-pressure helium gas, part of the laser is converted into an ultrafast, laserlike x-ray beam. (Image: Tenio Popmintchev, JILA, University of Colorado at Boulder) There was little science to explain HHG when it was first discovered. After years of work, scientists eventually understood how very high harmonics are emitted, but one significant challenge remained: The output HHG beams were extremely weak for most wavelengths in the x-ray region. To turn HHG into a useful x-ray source, Murnane, Kapteyn and their students created a tabletop-scale x-ray laser that had been a goal for laser science since shortly after the first laser was demonstrated by Theodore Maiman in 1960. “This was not an easy task,” Murnane said. “Unlike a laser — which gets more intense as more energy is pumped into the system — in HHG, if the laser hits the atoms too hard, too many electrons are liberated from the gas atoms, and those electrons cause the laser light to speed up. If the speed of the laser and x-rays do not match, there is no way to combine the many x-ray waves together to create a bright output beam, since the x-ray waves from different gas atoms will interfere destructively.” This art represents a coherent (laserlike) x-ray pulse with the largest color spread generated to date. Such rainbows of colors can support extremely short, few-attosecond light pulses — one attosecond is the time it takes for light to travel the length of three hydrogen atoms. (Image: Tenio Popmintchev and Brad Baxley, JILA, University of Colorado) To get x-ray waves from many atoms in the gas to interfere constructively, the scientists used a relatively long-wavelength, mid-infrared laser and a high-pressure gas cell that also guides the laser light. It resulted in bright, x-ray beams that maintain the coherent, directed beam qualities of the laser that drives the process. The HHG process works only when the atoms are hit “hard and fast” by the laser pulses, with durations nearing 10 to 14 s — a fundamental limit that represents just a few oscillations of the electromagnetic fields. This technology was developed in 1990 by Murnane and Kapteyn, using lasers that developed HHG-based light sources in the extreme-UV region in the 2000s. Although many scientists were using this technology to measure ever-shorter-duration light pulses, the two JILA researchers were hung up on how to make coherent light at shorter wavelengths in the more penetrating x-ray region. “We would have never found this if we hadn’t sat down and thought about what happens overall during HHG, when we change the wavelength of the laser driving it, what parameters have to be changed to make it work,” Kapteyn said. “The amazing thing is that the physics seem to be panning out, even over a very broad range of parameters. Usually in science you find a scaling rule that prevents you from making a dramatic jump, but in this case, we were able to generate 1.6 keV — each x-ray photon was generated from more than 5000 infrared photons.” This art shows data from the Young’s double-slit interference pattern, generating laserlike, directed x-ray beams. (Image: Tenio Popmintchev and Brad Baxley, JILA, University of Colorado) To truly control the beam of photons, the researchers needed to understand the HHG process at the atomic level and how x-rays emitted from individual atoms combine to form a coherent beam of light. That knowledge combines microscopic and macroscopic models of the HHG process with the fact that those interactions occur at very high intensity in a dynamically changing medium. The development of such a conceptual understanding took the last decade to develop. The result was the realization that there is no fundamental limit to the energy of the photons that can be generated using the HHG process. To obtain higher-energy photons, the system paradoxically begins with laser light using lower-energy photons — specifically, mid-infrared lasers. Besides achieving the high energy, the increasingly broad spectrum opens a range of new applications. “In an experiment using such a source, one energy region from the beam will correspond with one element, another with another element, and so on to simultaneously look at atoms across entire molecules, and that will allow us to see how charge moves from one part of a molecule to another as a chemical reaction is happening,” Kapteyn said. “It’ll take us a while to learn how to use this, but it’s very exciting.” The research will appear in the June 8 issue of Science. For more information, visit: www.nsf.gov