Group leaders: Keith Nelson, Chemistry Dept., MIT, Cambridge MA 02139

Philip Anfinrud, NIDDK, NIH, Bethesda, MD 20892

The availability of a source of intense hard x-ray pulses of short duration opens up many opportunities for time-resolved measurements of molecular and collective structure and dynamics. We consider first the unique capabilities projected for the LCLS output and the measurement methods made possible or enhanced substantially by these capabilities. We then elaborate on some of the most exciting scientific applications that can be pursued.

The main features of the LCLS that impart unique capabilities for time-resolved measurements are the following.

Time-resolved Laue diffraction can be conducted with the incoherent part of the output. While this does not exploit the coherence, it still capitalizes on pulse duration and photon number as well as synchronization with an optical excitation pulse.

Time-resolved monochromatic crystallography may be conducted with the coherent output, using tight focussing to achieve a wide range of incident angles onto the sample.

Diffraction methods can be used to characterize structural rearrangements in molecules, including biological molecules, and in condensed matter including phenomena such as phase transitions and domain growth or reorientation.

Optical initiation of photochemical, charge transfer, and other molecular processes can be followed by x-ray spectroscopic probes of various types. The time resolution is provided by the delay between optical excitation and x-ray probing pulses.

Time-resolved EXAFS/NEXAFS can be conducted using the coherent part of the output, tuning the wavelength through the available range. The tuning range may be broadened by using the third harmonic of the output, with wavelengths as short as 0.5A. The high intensities available may permit sufficiently rapid measurements that dilute samples including gas-phase molecules, molecules on surfaces, and biomolecules can be studied. Since time-resolved measurements will be made on those molecules that have been optically excited, almost all samples are rather "dilute".

Time-resolved atomic x-ray fluorescence can be conducted to gain information about the electronic orbital structure around individual atoms, complementary to EXAFS which provides information about the positions of nearby atoms. X-ray Raman scattering can also be detected, providing additional information about vibronic coupling to the core excited state through measurement at incident x-ray wavelengths variably detuned away from resonance. Raman measurements of low-energy states are possible, since the resolution is determined only by the excitation linewidth and the spectral filtering of the emission. The LCLS output is ideally suited for this application since it has sufficient spectral brightness to permit detection of Raman signals even after filtering of excitation and emission.

Time-resolved x-ray photoelectron spectroscopy can be conducted to examine the evolution of chemical species following optical excitation.

Time-resolved x-ray scattering can be measured to examine localized disturbances resulting from optical excitation. Information about time-dependent sizes and shapes of electronic excitations (e.g. excitons, charge transfer states) and the molecular distortions surrounding them, localized structural responses (e.g. solvent rearrangements around photoexcited species), and collective structural responses (e.g. nucleation or domain evolution) can be extracted and correlations between response time scales and distance scales can be elucidated.

Time-resolved four-wave mixing (or "transient grating" measurements) can be conducted using diffractive beam splitting and recombination methods (permitting heterodyne detection of the signals) or using reflective optics for small excitation angles. Various excitation mechanisms will be at work including absorption and impulsive stimulated Brillouin and Raman scattering, resulting in coherent acoustic phonons, coherent low-frequency optic phonons, electron-hole, and thermal gratings. Time-dependent vibrational responses, hot electron relaxation and diffusion, and thermal diffusion can be examined at selected, high wavevectors up to the Brillouin zone boundary. These measurements exploit all of the main features of the LCLS output, including spatial coherence which is necessary for formation of the excitation optical interference pattern.

5. Single-shot detection methods

Due to problems associated with sample damage, fluctuations in output intensity, and general signal/noise limitations, it is desirable when possible to collect all the time-dependent information in a single shot. This represents a major change from traditional methods in which a delayed probe pulse provides information about just a single point along the time axis, and many repetitions of the measurement are conducted at each delay time (for signal averaging) and with variable delay times. Several methods, demonstrated in the visible regime, should be applicable to time-resolved x-ray methods as well. These include the use of echelons to produce many probe pulses with variable delays that all pass through the sample following a single excitation event, and the use of a large angle between excitation and probe beams to set up a space-to-time correspondence at the sample. In both cases, the transmitted or reflected probe light is imaged at a CCD to provide complete temporal information on each laser shot.

The experimental capabilities elaborated above open up unique opportunities for the study of molecular and collective behavior. Some of the prospects most exciting to subpanel members are discussed here.

With coherent x-rays, transient grating measurements with wavevectors extending throughout the Brillouin zone will be possible. In this manner the dynamics associated with lengths scales on the order of typical correlation lengths can be examined. Most exciting is the prospect of direct examination of acoustic phonons at all wavevectors. In current optical measurements, restricted to near-zero wavevectors, strong dispersion in longitudinal and shear acoustic wave velocities and damping rates are associated with overlap of the acoustic frequency (or period) with characteristic relaxation rates (or times). At high wavevectors, strong dispersion, including the transition from underdamped to overdamped acoustic responses, may be due to overlap of the acoustic wavevector (or wavelength) with correlation wavevectors (lengths). Complete determination of both time and space dependences of density, shear, and other quantities coupled to them (e.g. polymer, glass, or liquid structural features, order parameters in crystals near structural phase transitions, etc.) should be possible. Recent x-ray Brillouin scattering results have demonstrated the value of broadband acoustic phonon characterization, particularly in glasses, polymers, supercooled liquids, and other disordered or partially ordered systems in which the acoustic behavior at high wavevectors is unclear. However the x-ray Brillouin scattering results suffer from the same limitations as their optical-regime analogs, namely that as the overdamped limit is approached and/or as strong central peaks in the light scattering spectrum arise due to structural relaxation processes, the Brillouin peaks broaden and merge with the central peak, making unique determination of phonon characteristics impossible. As in the optical regime, time-domain scattering will circumvent this problem and reveal phonon dynamics even in the overdamped limit. Thus, structural correlation lengths and times both can be probed directly through their influence on acoustic phonons of tunable wavelength (on the correlation length scale) and frequency or oscillation period (on the correlation time scale). X-ray transient grating measurements will also permit examination of thermal diffusion on small distance scales, including those at which heat transport (like density and shear) becomes nonhydrodynamic. Finally, electron or exciton diffusion on short grating distance scales can also be monitored. In resonant experiments, the responses observed may be atom-specific.

If optical rather than x-ray excitation is used, then the excitation region is spatially uniform rather than modulated on the x-ray wavelength scale. Nevertheless localized or mesoscopic-scale responses often result, either through electronic excitations that are localized on impurities or that have limited coherence lengths or through nonuniform structural responses (domain formation or change in morphology, nucleation, etc.). Optical pump/x-ray scattering probe measurements, in which the x-ray scattering pattern is analyzed as a function of time following photoexcitation, can be used to reveal the dynamics associated with different response length scales. For example, in relaxor ferroelectrics, optically induced changes in the morphologies or orientations of polarized microdomains will give rise to scattering features associated with corresponding microdomain sizes, and the dynamics observed should reveal the correlations between microdomain size and time scales. Similarly, the responses of liquids to photoexcited solutes, the formation of charge-transfer or exciton states, and the generation of other localized or partially localized responses to optical excitation all give rise to features of various sizes, shapes, and time dependences which can be monitored through measurement of the x-ray scattering pattern as a function of time. Examination of charge transfer or redistribution in extended donor-acceptor systems such as conjugated polymers and organic superlattices, in semiconductor quantum wells, and in photorefractive materials would be of particular interest.

Finally, the possibility of direct monitoring of crystal structure and its evolution through x-ray diffraction following optical or x-ray excitation has long been of great interest. Time-resolved measurements of melting and recrystallization, photoinduced structural phase transitions, lattice distortions due to acoustic or optic phonon oscillations in the linear or nonlinear lattice dynamical regime, and similar collective responses to either collective or localized excitation (the latter, for example, in impurity-induced structural phase transitions) will answer many questions concerning the microscopic mechanisms of collective structural change. Large-amplitude motion along soft lattice vibrational coordinates could be examined directly, and in those cases in which structural phase transitions or switching between domain orientations could be induced optically, the nature of the motion into the new phase or domain orientation could be elucidated. In single or multiple-component ferroelectrics, for example, time resolution of 250 fs should be sufficient to observe soft mode vibrations following optical excitation, and if current efforts toward optically induced transitions between phases or domain orientations prove successful, the collective, coherent motion between structures could be monitored and intermediates could be characterized.

Time-resolved x-ray fluorescence spectroscopy provides a means for determination of the local electronic structure around a selected atomic site following a photoexcitation event. Selectivity is achieved through the element specificity and also the exploitation of chemical shifts in the excitation process. Furthermore, angle resolved measurements make it possible to distinguish orbitals of different symmetries in oriented samples such as molecules on some surfaces. Thus the distribution of orbitals in space can be directly viewed. The method has been developed at third-generation sources on molecules in the gas phase, in condensed matter, and on surface adsorbates. Molecules on surfaces pose particularly exciting targets due to the importance of surface reactions in biology and technology. Surface chemistry has generally been studied through electron spectroscopies applicable primarily in vacuum, but photon-in/photon-out x-ray methods will permit study of surfaces under liquids or in high-pressure gaseous environments. Biological molecules on surfaces in ambient conditions also can be examined. Following photoexcitation, time-resolved x-ray emission will provide atom-specific information about how the electronic structure evolves during charge transfer or other chemical reactions. Short-lived metastable states will be characterized directly, and activated complexes in biological systems may be accessible to direct observation.

X-ray Raman measurements at or near core resonances can be used to examine further the electronic structure and symmetry of intermediates in reactions and other processes. For example, changes in the allowed Raman transitions are a sensitive measure of symmetry-breaking in molecular intermediates.

Complementary to x-ray fluorescence are EXAFS and NEXAFS which may provide molecular structural information. Interatomic distances can be determined accurately, and together with x-ray emission data both the geometry and electron density distribution of reaction intermediates can be elucidated. In liquid solution, small molecules (e.g. iodine in CCl₄) as well as polyatomics and biological molecules can be examined. Photodissociation and collision with the solvent cage, geminate recombination, electronic and vibrational relaxation, and above all sequential reaction processes in charge transfer or other complex reaction events can be monitored following photoinitiation.

Isolated molecules or mass-selected clusters in the gas phase can also be examined through time-resolved x-ray emission and EXAFS, as well as x-ray photoelectron spectroscopy for determination of chemical species. Characterization of photochemical evolution in metallic and other clusters of various sizes is a particular priority, well suited to x-ray spectroscopic techniques. Finally, time-resolved x-ray diffraction offers prospects for direct measurement of molecular structure.

The use of synchrotron radiation for solution of protein structures has revolutionized the field of structural biology. Such efforts have focussed on static structures and not on the structural evolution that accompanies the chemical time scale of femtoseconds. Macromolecular crystallography on the femtosecond time scale is not feasible on any existing x-ray source. However, the ultrashort pulse duration of the LCLS combined with its spectral brightness and high photon flux provides the x-ray characteristics required to elicit detailed structural information on the primary events that lead to protein function. Time-resolved x-ray crystallography of photoactive proteins such as photoactive yellow protein, bacteriorhodopsin, photosynthetic reaction center, and other phototriggerable proteins such as heme proteins would provide deeper insight into the function of these molecules of life.

In some cases, high resolution structural information can also be extracted on a local level using EXAFS and NEXAFS. The spectral brightness of the LCLS output opens up the possibility of carrying out such studies with subpicosecond time resolution. For example, EXAFS studies on the iron in ligand-binding heme proteins can provide high resolution information on the time-dependent position of a photodetached ligand, complementing the structural information acquired through time-resolved diffraction studies. Another particularly important class of events is biochemical charge transfer, which in some cases is photo-triggerable and can be probed through EXAFS and NEXAFS of the metal ions involved. Determination of atomic charge state and coordination number for nuclei with selectively accessible absorption will be possible.

The possibility of x-ray rather than optical excitation to initiate processes of interest (such as charge transfer) in biological systems that are not normally photo-triggerable should be explored, since this would make possible time-resolved examination of systems that are currently inaccessible.

Optical generation of excited electronic states in biological antannae (e.g. light harvesting systems) and reaction centers could be followed by x-ray scattering measurements to elucidate the size scale of the excitation, i.e. the extent of delocalization. This could provide information about electronic behavior complementary to the structural information extracted through diffraction or EXAFS measurements.

Subpicosecond Time-resolved X-ray Measurements

Photon Correlation Spectroscopy and Holography

Non-Linear X-Ray Optical Processes

High Field Physics and Non-linear Quantum Electrodynamics with the LCLS

Content Owner: J. Arthur
Page Master: L. Dunn
Last Update: 16 Nov 1999