3 FAR-INFRARED REGION:1000 to 10 µm

The scientific case for free electron lasers is particularly strong in the far infrared, because in this region nonlaser sources are weak and the use of conventional laser sources is limited by a lack of nonlinear crystals. Moreover, the technical requirements of the electron accelerator used in an FEL relax as the wavelength increases, and therefore far-infrared FELs are cheaper and have lower technical risk than FELs operating at shorter wavelengths. Thus there are a number of opportunities for cost-effective scientific applicaitons of FELs in the far-infrared region.

PRESENT CAPABILITIES

Conventional continuous-wave nonlaser sources are thermal in nature. Two broad band sources are conventionally used. The hot silicon carbide source used in commercial infrared spectrometers, commonly called the “glow bar,” is limited to wavelengths shorter than 100 µm by the λ−2 dependence of the intensity of the light, as well as by the fact that its emissivity drops at long wavelengths, reducing the intensity still further. At wavelengths longer than 100 µm, the mercury discharge lamp is used and has an intensity that drops as λ−1. The long wavelength emission of synchrotrons has also been used as a source, since it decays with a weaker power of the wavelength than does the glow bar. Nevertheless, the low intensities, combined with the relative lack of sensitive detectors, often lead to marginal signals with concomitant long data-acquisition times. Stimulated spin-flip Raman scattering has produced tunable coherent infrared radiation in the 5- to 20-µm regime but has not been developed into a commercially viable technology.

The extension of the operation of solid-state, laser-based optical parametric oscillators to the region from 10 to 22 µm has been considered, using as nonlinear materials AgGaSe2and CdSe, which have progressively longer wavelength limits. At present these materials are available, but surface damage by the pump laser pulse is an unsolved problem. Difference frequency generation in cryogenically cooled GaAs over the region from 67 to 200 µm has been demonstrated using pulsed 10-µm CO2 lasers. No materials generally suitable for use in parametric oscillators operating at wavelengths longer than about 20 µm have been proposed.

Resonantly enhanced Raman scattering has been used to obtain tunable farinfrared (FIR) emission over limited wavelength regions. For example, CH3F pumped by a tunable CO2 laser has been used to obtain 100-ns pulses of < 1-mJ energy over the wavelength range from 250 to 300 µm. However, the wavelength coverage



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 24
FREE ELECTRON LASERS AND OTHER ADVANCED SOURCES OF LIGHT: Scientific Research Opportunities 3 FAR-INFRARED REGION:1000 to 10 µm The scientific case for free electron lasers is particularly strong in the far infrared, because in this region nonlaser sources are weak and the use of conventional laser sources is limited by a lack of nonlinear crystals. Moreover, the technical requirements of the electron accelerator used in an FEL relax as the wavelength increases, and therefore far-infrared FELs are cheaper and have lower technical risk than FELs operating at shorter wavelengths. Thus there are a number of opportunities for cost-effective scientific applicaitons of FELs in the far-infrared region. PRESENT CAPABILITIES Conventional continuous-wave nonlaser sources are thermal in nature. Two broad band sources are conventionally used. The hot silicon carbide source used in commercial infrared spectrometers, commonly called the “glow bar,” is limited to wavelengths shorter than 100 µm by the λ−2 dependence of the intensity of the light, as well as by the fact that its emissivity drops at long wavelengths, reducing the intensity still further. At wavelengths longer than 100 µm, the mercury discharge lamp is used and has an intensity that drops as λ−1. The long wavelength emission of synchrotrons has also been used as a source, since it decays with a weaker power of the wavelength than does the glow bar. Nevertheless, the low intensities, combined with the relative lack of sensitive detectors, often lead to marginal signals with concomitant long data-acquisition times. Stimulated spin-flip Raman scattering has produced tunable coherent infrared radiation in the 5- to 20-µm regime but has not been developed into a commercially viable technology. The extension of the operation of solid-state, laser-based optical parametric oscillators to the region from 10 to 22 µm has been considered, using as nonlinear materials AgGaSe2and CdSe, which have progressively longer wavelength limits. At present these materials are available, but surface damage by the pump laser pulse is an unsolved problem. Difference frequency generation in cryogenically cooled GaAs over the region from 67 to 200 µm has been demonstrated using pulsed 10-µm CO2 lasers. No materials generally suitable for use in parametric oscillators operating at wavelengths longer than about 20 µm have been proposed. Resonantly enhanced Raman scattering has been used to obtain tunable farinfrared (FIR) emission over limited wavelength regions. For example, CH3F pumped by a tunable CO2 laser has been used to obtain 100-ns pulses of < 1-mJ energy over the wavelength range from 250 to 300 µm. However, the wavelength coverage

OCR for page 24
FREE ELECTRON LASERS AND OTHER ADVANCED SOURCES OF LIGHT: Scientific Research Opportunities obtained by these schemes is limited for each gas and laser combination and is not continuously tunable. Molecular gas lasers using gases such as HCN, CH3F, or methanol provide a number of discrete wavelengths in the wavelength region beyond 30 µm and extending to longer than 700 µm. Most of these lasers are optically pumped by pulsed CO2 lasers and emit pulses in the microsecond time domain; some, such as HCN, can be excited by a gas discharge. They have not found extensive use as sources for far-infrared spectroscopy because of the limited line tuning that is possible. In summary, there is a lack of intense, tunable sources in the far-infrared region. In particular, there are no tunable sources producing the intense picosecond pulses needed for some of the potential experiments discussed below. The FELs that are currently operational or under construction are summarized in Chapter 2. The Free Electron Laser for Infrared Experiments (FELIX) located in the Netherlands is of interest as an example of an operational user facility, while the Collaboration for an Infrared Laser at Orsay (CLIO) in France is a user facility that has begun to produce a number of scientific results. ENVISIONED DEVELOPMENT OF FREE ELECTRON LASERS In contrast to FELs proposed for the ultraviolet and x-ray regions, the development of FELs for the far-infrared wavelength region poses a relatively low technological risk. A number of operational systems, as well as systems nearing operation, have provided experience with the technology and are a guide to the cost of future machines. It is important to note that the electron-beam energies required for operation of an FEL in the far infrared are substantially lower than those needed for operation at 2 to 5 µm, and this has a number of favorable consequences. Typical near-infrared FELs use beam energies of the order of 40 MeV. The beam energy for a longer-wavelength machine is in the 10- to 15-MeV range. Lowering the beam energy significantly below 15 MeV can reduce the shielding requirements, and thus the cost and the space requirements. It is a goal of some current projects to develop a low-cost far-infrared FEL that could be supported as a departmental or individual investigator machine. Some technology needs for further development of FIR FELs can be specified now. Stability of the RF power source has been shown to be a problem in some machines and can lead to undesirable and uncontrollable frequency drifting. The use of photocathodes activated by a pulsed laser (e.g., the 355-nm third harmonic of a Nd:YLF [neodymium yttrium lithium fluoride] laser) or the use of RF subharmonic bunching allows the time between micropulses to be made long enough, typically a few nanoseconds, for pump-probe measurements. The Nd:YLF laser harmonics can also be used to pump a tunable dye laser or may themselves serve as

OCR for page 24
FREE ELECTRON LASERS AND OTHER ADVANCED SOURCES OF LIGHT: Scientific Research Opportunities an ultraviolet or visible laser source for two-color experiments. Sum-frequency experiments, described below, are one example of such two-laser studies. The existing FELs provide a good guide to potential costs. The FELIX FEL in the Netherlands covers the wavelength regions from 5 to 30 µm and 16 to 110 µm using two separate beamlines and was designed to operate as a user facility. FELIX cost approximately $8.5M to build and install in an existing building. However, it is important to distinguish the cost of the machine from that of the complete user facility. The latter entailed remodeling and partially equipping five user rooms and cost an additional $3.5M. The 1994 operating costs for FELIX are budgeted at $1.75M and are expected to provide for 2000 hours of beam time; this translates to about $900/hr. SCIENTIFIC OPPORTUNITIES Not only have experiments using FEL radiation in the spectral region from 10 to 1000 µm been proposed, but recent results from several of the currently operational devices also give some sense of what can be achieved. Some of the areas in which experiments have been proposed or conducted are surveyed below. Surface Science Much has been learned about molecules and atoms adsorbed onto surfaces through the accessibility of tunable infrared sources, both coherent and incoherent, that can measure intramolecular vibrations in the infrared between 1 and 10 µm. Fourier-transform infrared and Raman scattering have extended the range of measurement of intramolecular vibrations down to about 400 cm−1 (25 µm). There is substantial interest in extending this low-frequency limit even further, because one could then access the regime where most frustrated translational and rotational modes (i.e., modes associated with adsorbate-surface bonds) lie. Internal torsional modes of molecules are also located here, and these have never been seriously studied in chemisorption systems for solid, liquid, or polymeric surfaces. The situation is similar for measurements of the intermolecular modes in van der Waals clusters, which provide information about the structure and dynamics of these systems. If one wants to understand the molecule-surface bond and how it is modified, for example, by adsorption of other species during a chemical reaction, then this region of the spectrum is of crucial interest. Moreover, measuring the energy distributions and line shapes of these modes is important for understanding how such modes couple to and dissipate their energy into the electron or phonon continuum of the substrate. Such experiments could be conducted either in a reflection or absorption mode or by resonant nonlinear optical studies such as sum-difference frequency generation and second harmonic generation. The latter requires synchronization of

OCR for page 24
FREE ELECTRON LASERS AND OTHER ADVANCED SOURCES OF LIGHT: Scientific Research Opportunities the FEL output with a visible laser but offers the advantage of detecting the signal in the visible region. The strongest motivation for using these nonlinear techniques would be to characterize the molecular structure and coupling of energy from the molecule to the substrate or surface. There is also some interest in measuring surface and interfacial electronic properties in this region by nonlinear optical methods. There are some initial results from these types of experiments using FELs, albeit at shorter wavelengths. The French group using CLIO has reported on the use of sum-frequency generation to study the Pt-methanol interface using 5-µm radiation (Peremans and Tadjeddine, 1994; Peremans et al., 1993). Surface science experiments would use several unique properties of FEL radiation: (1) continuously tunable high-intensity radiation, which would allow probing of specific vibrational frequencies with good signal-to-noise ratio and (2) picosecond pulses, which would allow pump-probe techniques to be used to study energy transfer processes at the surface. Chemistry The study of energy transfer in molecules in the gas and liquid phase would be enhanced by the availability of pulsed sources that could access lower-frequency molecular modes. A recent report from the CLIO group discusses studies of the kinetics of d2ethane isomerization using FEL radiation (Roubin et al., 1994). The experiment used radiation of wavelengths no longer than 10 µm and was not time resolved. Therefore, it might have been carried out using ordinary lasers. However, it suggests a class of experiments to study mode-selective chemistry that would require high-powered, short-pulsed sources capable of exciting the full range of molecular vibrations, and these future experiments would require a far-infrared FEL. The Stanford FEL was used in studies of vibrational infrared photon echoes in a liquid and glass (Zimdars et al., 1993). Once again, since this particular experiment was done using 5.1-µm radiation (although here picosecond pulses were required), it could have been done with sophisticated laboratory lasers. However, an important extension of this type of work would be to investigate lower-frequency modes closer in frequency to vibrations of the surrounding medium in order to study the transfer of energy from the molecule to the medium. These longer-wavelength experiments would require far-infrared picosecond pulses and could only be done using an FEL. Although these experiments used wavelengths at the short-wavelength end of the region of interest, they are examples of the kind of chemical studies that can be performed. Solid-State Physics For bulk, homogeneous materials the photon energies accessible to the infrared-FEL overlap essentially completely the principal excitations in condensed

OCR for page 24
FREE ELECTRON LASERS AND OTHER ADVANCED SOURCES OF LIGHT: Scientific Research Opportunities matter: phonons, plasmons, magnons, and inter-sub-band transitions. An infrared-FEL would permit direct, linear probing of both defect modes and buried interfaces with bond specificity. For time-independent linear spectroscopy, however, the FEL offers few advantages over other spectroscopic approaches. In combination with another source (e.g., a fixed-frequency laser), the FEL also permits both space- and time-resolved studies of similar systems via nonlinear spectroscopy— in particular sum-frequency generation. Even more importantly, it would open to study new classes of problems that require probing of mode-mode interactions, relaxation pathways from deliberately prepared excited states, spatial and temporal selectivity, and inhomogeneous materials (defects, buried interfaces, artificially structured superlattices, and quantum wells). A particularly compelling opportunity for FELs exploits their high-energy and high-frequency attributes. Together these could modify the system under study by driving it into specific preselected strongly nonequilibrium states. The flows of energy and relaxation from such states can then be probed. This approach is especially important in semiconductors because quantum optical and electronic devices operate under such nonequilibrium conditions. The field amplitudes in an FEL laser pulse can be of the same or higher magnitude than those inherent in the structures or devices themselves (on the order of kilovolts per centimeter). Furthermore, they can be deliberately reconfigured on essentially any time scale—including faster-than-typical structural or carrier relaxation times. The FEL thus provides opportunities to observe in semiconductors nonlinear and coherent effects (like resonant tunneling) previously seen only in superconductors at microwave frequencies. One might envision other exploitation of the high field effects in ferroelectric, piezoelectric, and Jahn-Teller solids as well. Biophysics Low-frequency modes of large biomolecules such as nucleic acids and proteins are excited by frequencies in the far infrared. The committee heard several suggestions of experiments that would attempt to produce a transient alteration of some properties of large biomolecules by exciting these modes. If such experiments could be performed, they would be very interesting and important. However, fast relaxation of the excited modes might eliminate any mode-specific effect. If these experiments were to be done, they would require a far-infrared FEL to produce the necessary energy in the long-wavelength region. Plasma Physics Applications of FELs to plasma physics fall into two main categories: plasma heating and plasma diagnostics. Heating of controlled fusion plasmas can be achieved using high-power microwave radiation. Powers in the tens-of-megawatt

OCR for page 24
FREE ELECTRON LASERS AND OTHER ADVANCED SOURCES OF LIGHT: Scientific Research Opportunities range are needed. Energy absorption is achieved by electron cyclotron resonance, which requires wavelengths in the range of 0.5 to 2 µ m. In some current experiments, it would be desirable to be able to tune the frequency within this range of wavelengths so as to heat plasmas to fusion temperatures, for different magnetic fields, or to move the absorption position from place to place inside the plasma. Sources used for heating should be quite efficient (about 50%) and able to sustain long pulses (several seconds). Proposals have been made for using localized heating as a means of controlling plasma instabilities. This approach would probably require the ability to change frequency rapidly so as to heat the critical region of the plasma. The requirements for control of plasma instability would involve the ability to deliver high peak power quickly and for a short time. FELs are one of the few sources with the potential to meet the above requirements. Plasma diagnostics is the second major use of far-infrared radiation in plasma physics. Such diagnostics involve either measuring the phase shifts of waves sent through the plasma or measuring the reflection of waves from critical density regions. For typical fusion plasmas the wavelengths required for phase shift measurements are in the range of centimeters to millimeters. More information about the state of a plasma can be obtained if it can be probed at two or more frequencies. For reflection measurements of plasma properties, the wavelength must be roughly in the same wavelength range as that required for transmission diagnostics. In some reflection experiments, multiple frequencies are used to map the plasma density using reflections from regions of different density. For this application, having a single source whose frequency could be changed on a millisecond time scale, such as a far-infrared FEL, would be quite valuable. Studies of low-temperature plasmas such as are used in plasma processing (plasma etching, plasma chemistry, and so on) would benefit greatly from tunable sources in the infrared for detecting the various molecular species and radicals that exist in a discharge. Since the chemical species give a signature of the processes taking place and also play a central role in those processes, such information is vital to understanding and controlling the chemical species. FINDINGS The scientific case for a tunable, short-pulse (picosecond) source in the far infrared is compelling, but at present there are no picosecond far-infrared FEL user facilities in the United States. This is the spectral region where molecule-surface vibrations, intermolecular cluster vibrations, and transitions in semiconductor quantum wells can be excited. It is also the region for probing transitions between adjacent high-lying Rydberg states of atoms and low-frequency motions in large biomolecules. There is sufficient scientific interest to efficiently use a far-infrared user facility capable of producing picosecond pulses, and the committee believes that the scientific opportunities justify the establishment of such a facility. Operation of

OCR for page 24
FREE ELECTRON LASERS AND OTHER ADVANCED SOURCES OF LIGHT: Scientific Research Opportunities such a facility would provide information on whether additional user facilities might be needed subsequently. Two relevant issues that should then be considered are the quality of the proposed science that cannot be accommodated by a single facility and the existence of alternative infrared sources such as laboratory-sized far-infrared FELs. Free electron lasers already exist in the far-infrared region, even as user facilities, and therefore uncertainty about the technical requirements and cost of establishing a far-infrared free electron laser facility is relatively low. The relatively modest accelerator and undulator requirements mean that a user facility might be built for several million dollars using existing technology. There is some promise that further development will lead to a significantly less expensive free electron laser that could be used by a single academic department or perhaps an individual investigator. A photocathode electron source may increase the utility of a user facility both because of the free electron laser's improved performance and because the laser used to excite the photocathode could also serve as a second, synchronous photon source for pump-probe experiments. RECOMMENDATIONS A far-infrared free electron laser user facility capable of producing picosecond pulses should be established. Such a facility could be a new construction or a modification of an existing facility. The committee believes that the scientific opportunity justifies the establishment of such a facility. Such a facility should have a means of synchronizing a second tunable laser with the free electron laser for use in pump-probe experiments as well as a means for selecting single picosecond pulses from the free electron laser. One method of doing this involves using a photocathode electron gun. The current research and development directed toward the production of a compact free electron laser that could be purchased and operated by a single academic department or individual investigator should be continued. REFERENCES Zimdars, D., A. Tokmakoff, S. Chen, S.R. Greenfield, M.D. Fayer, T.I. Smith, and H.A. Schwettman. 1993. “Picosecond Infrared Vibrational Photon Echoes in a Liquid and Glass Using a Free Electron Laser,” Phys. Rev. Lett. 70:2718. Peremans, A., and A. Tadjeddine. 1994. “Spectroscopic Investigation of Pt-Methanol Interface in Perchloric Acid Medium by Sum-Frequency Generation,” Chem. Phys. Lett. 220:481.

OCR for page 24
FREE ELECTRON LASERS AND OTHER ADVANCED SOURCES OF LIGHT: Scientific Research Opportunities Peremans, A., A. Tadjeddine, M. Suhren, R. Prazeres, F. Glotin, D. Jaroszynski, and J.M. Ortega. 1993. “CO Species from CH3OH Electrochemical Decomposition on Platinum Studied by Visible-IR Sum-Frequency Generation Using CLIO Infrared Free Electron Laser, ” J. Electron Spectrosc. Relat. Phenom. 64-65, 391-393. Roubin, P., S. Cardonatto, H. Dubost, and J.M. Ortega. 1994. “Free Electron Laser Infrared Radiation of Ethane Isolated in a Xenon Matrix,” in press.