IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. 27, NO. 12. DECEMBER 1991 2691 Applications of Infrared Free-Electron Lasers : Basic Research on the Dynamics of Molecular Systems Dana D. Dlott and Michael D. Fayer Invited Paper Abstract-Applications of tunable infrared (IR) picosecond An FEL application may involve an experimental appa- (ps) pulses generated by free-electron lasers (FEL’s) are con- ratus with complexity approaching that of the FEL. sidered from several perspectives. First, general considerations In recent years, laser spectroscopy using conventional are presented that discuss characteristics that an IR ps FEL must have to be successfully used as an optical source in order lasers has become very sophisticated [5], [6], and with to perform complex ultrafast experiments commonly under- this sophistication, the demands routinely placed on these taken with conventional ps lasers. The features that FEL’s can lasers have grown. FEL operation at a wide variety of have that will make them superior to conventional sources for wavelengths has been demonstrated [I], [4], but the ca- the investigation of complex molecular systems in condensed and gas phases are pointed out, and the difficulties of combin- pability of simultaneously and reliably fulfilling the nec- ing an FEL with an intricate experimental apparatus are essary demanding operational specifications to conduct brought forward. Next, ps photon echo experiments on a near- sophisticated experiments is not straightforward. It is a IR dye molecule in a low-temperature glass, performed using challenge to the FEL community to harness the unique the Stanford superconducting linear accelerator-pumped FEL, capabilities of FEL’s to realize the vast potential of sug- are discussed to illustrate the ability of FEL’s to act as light sources for sophisticated applications. The methods employed gested applications and to make FEL’s into potent tools and the problems encountered in the echo experiments are de- for basic research. In order to attain this goal, FEL de- scribed. Finally, the applications of IR ps FEL’s to the study velopers and users must develop a relationship that can of molecular vibrational dynamics are discussed, including de- ultimately lead to the realization of these applications and tailed numerical feasibility calculations. It is shown how to de- to the development of better FEL’s. termine the extent of vibrational excitation for a given system, In this paper, we focus on applications of tunable in- given readily determined molecular and FEL parameters; sub- stantial excited vibrational populations can be generated in the frared (IR) picosecond (ps) pulses in the study of dynam- examples considered. Once a vibration is excited, a variety of ics of molecular systems. Infrared FEL’s are at this time methods of probing the subsequent dynamics are discussed. the most highly developed and reliable systems that offer These include absorption probing (pump-probe experiment), clear advantages over available benchtop lasers. The ex- probing by anti-Stokes Raman scattering, and probing by hy- per-Raman scattering. A class of experiments referred to as amples presented here include the dynamics of condensed nanoheating, i.e., selective heating on a nanometer length scale, matter (e.g., solids and surfaces), gas-phase dynamics, is described. Biomedical and biophysical applications of nano- and medical and biological applications. All these sys- heating are suggested. tems involve the interactions of many bodies, and there- fore they present a rich variety of dynamical phenomena occurring on the ultrafast time scale. I. INTRODUCTION There are two additional reasons for our focus on in- REE-ELECTRON lasers (FEL’s) have developed frared applications. IR spectroscopy [7], [8] is intrinsi- Frapidly in the last 15 years [ 11, and many applications cally interesting and valuable, and it seems hardly worth- for them have been envisioned [2]-[4]. To use the output while to consider FEL applications in the visible and near of an FEL for fundamental studies of condensed-phases UV, where an enormous variety of high-power benchtop and gas-phase molecular systems requires both an optical laser systems are available from commercial sources. The source that meets stringent requirements and the mating most important spectral region for chemical applications of that source with an experimental apparatus. A laser by is the vibrational IR (3-30 pm), although many interest- itself is not capable of making meaningful measurements. ing applications involve the near IR (1-3 pm). The dy- namics of transitions in the far IR (30-200 pm) are of Manuscript received March 27, 1991; revised July 13, 1991. This work undisputable importance but are barely explored. was supported by the National Science Foundation, Division of Materials Research by Grant DMR 87-21253, the U.S. Army Research Office by Intense IR pulses can be applied in three canonical ap- Grant DAALO-90-G-0030 (to D.D.D.), by the MFEL program through the plications. The first is the study of electronic transitions Office of Naval Research by Grant N00014-89-KO154, and by National of materials with small energy gaps. These include a va- Science Foundation, Division of Materials Research by Grant DMR 87- 18959 (to M.D.F). riety of semiconductors, particularly those like HgCdTe, D. D. Dlott is with the School of Chemical Sciences, University of 11- which are used in IR detectors, high T, superconductors, linois, Urbana, IL 61801. and large organic and inorganic molecules with low-lying M. D. Fayer is with the Department of Chemistry, Stanford University, Stanford, CA 94305. electronic states. The second is the study of mechanical IEEE Log Number 9103642. excitations. These include phonons and vibrons, which 0018-9197/91$01.00 0 1991 IEEE ~~~~ 2698 IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. 21, NO. 12, DECEMBER 1991 are collective modes of crystalline solids, and vibrations plications. These applications were chosen using three localized on individual molecules. Molecular vibrations criteria. They are expected to provide scientifically inter- are sensitive to the details of the structure of a molecule, esting and unique information about important materials, and therefore provide fingerprint selectivity in molecular they cannot reasonably be performed using available identification [7], [8]. Selective excitation of a specific benchtop laser systems, and there is a high probability of molecular vibration may provide a route to the control [9] success using existing FEL sources. of the chemical reactivity of that molecule. Molecular vi- brational spectroscopy can be used to investigate crystal- 11. FREE-ELECTRONLASER APPLICATIONS: line and amorphous solids, liquids, molecules on sur- SPECIFICATIONSFOR APPLICATIONS faces, and systems of biological or biomedical interest. The third type of application involves nanoheating, i.e., In this section we investigate the performance specifi- transient heating of nanoscale structures. The FEL differs cations an FEL must achieve to be useful in sophisticated from ordinary heat sources in its ability to excite one com- applications. In many cases it has been demonstrated that ponent of a complex mixture via vibrational selectivity to a particular FEL can deliver a substantial (time averaged) thereby produce a substantial jump in temperature of ex- power at a convenient center wavelength. This type of ceedingly brief duration. measurement does not characterize FEL performance in Ultrashort light pulses are of fundamental importance enough detail to determine whether an FEL is capable of in probing the dynamics of complicated systems [SI, [6]. performing a specific application. An elementary argument in statistical mechanics can be The most important attribute of an FEL is reliable per- used to show that the limiting time scale of most mechan- formance. At this time the greatest obstacle to all FEL ical processes is of order h/k,T = s at ambient applications is the lack of availability of FEL photons. It temperature. Practically every mechanical and chemical is also important to keep in mind that an FEL is simply a process of interest, including relaxation processes, energy laser. There is a fundamental difference between an FEL transfer, and chemical reactions, occurs slower than 100 facility and a synchrotron light source in the FEL’s in- fs. The pulse duration produced by most FEL’s is only ability to service many users. The power supplied by a one order of magnitude greater than this characteristic synchrotron light source is independent of the number of time, thus opening up a window to a substantial variety user ports. Furthermore, a synchrotron radiation light of dynamical processes. A number of proposals have been source provides polychromatic radiation so each user has advanced to achieve the remaining order of magnitude, the ability to select from a wide range of wavelengths. including optical pulse compression [ 101 using IR-trans- Experience acquired using conventional lasers indi- mitting optical fibers. cates that sharing a laser among many users can be quite Because reliability of existing FEL’s is always a press- difficult. Benchtop lasers generally service a single ex- ing issue, we believe there is a tendency in the FEL com- periment at one time. Sharing is further inhibited by tun- munity to imagine that compared to operating the FEL, able lasers as different users will typically need pulses at FEL applications are straightforward. We wish to stress different wavelengths. Although in some situations it may the rigorous and demanding nature of the application, i.e., be possible, it is generally unlikely that two or more ex- the experimental setup, which exploits the unique prop- periments using an FEL can be performed simulta- erties of FEL’s. A successful application will not only neously. advance our understanding of the material at hand but also However, an FEL facility can service more users than provide insight into the detailed behavior of the FEL by a benchtop laser. Because an FEL typically costs 10-100 demanding a high level of performance from the FEL.