We suggest a state-of-the art x-ray source using a compact electron storage ring with modest energy (less than 1 GeV) and a high power mm-wave as an undulator. A source of this type has x-ray energies and brightness comparable with third generation synchrotron light sources while it can be very compact and fit in a small university or industrial laboratory or hospital. We propose to operate an isochronous mm-wave FEL and a hard x-ray inverse Compton source at the Duke storage ring to test this concept. Resonant FEL conditions for the mm- wave will be provided by the off-axis interaction with an electromagnetic wave. A special optical resonator with holes for the e-beam is proposed for pumping a hard x-ray inverse Compton source with very high brightness. Simulation results of mm-wave FEL operation of the Duke storage ring are discussed. Expected performance of mm-wave FEL and hard x-ray inverse Compton source are presented.

We present some of our experimental results for the generation of picosecond x-ray pulses by means of a diode driven by ultrashort ArF laser pulses. Hard x-ray pulses with duration in the range 1-100 ps have been generated at a repetition rate of 300 Hz. Some applications of this system to picosecond time resolved x-ray diffraction from laser heated crystals are reported.

We apply the reflectron principle to design an electron gun capable of delivering femtosecond duration electron pulses. The gun is based on photoemission and uses a high repetition rate femtosecond laser system. We show that even at low laser power electrons are generated in a multiphoton process, generating an electron beam with a significant energy spread. The temporal broadening resulting from this distribution can be partially compensated by a reflective electron mirror. Our reflectron features a gridless mirror that focuses the electron pulses in space and time. Computer simulations highlight the important design parameters and demonstrate that pulse durations of hundreds of femtoseconds are achievable, even in cases of broad kinetic energy distributions. The reflectron is very accommodating toward space charge effects, making it an attractive choice in applications requiring high electron densities.

The production of high photoelectron densities from metallic photocathodes under ultrashort visible and UV-laser finds applications in various fields. Many workers tried, by the use of different techniques, to improve the photoelectric performances of pure metals, essentially W and Mo, but all these prepared photocathodes were not able to withstand very high laser intensities, i.e., >= 1GW/cm². We demonstrate that ion implantation is a valuable new tool to improve the photoelectric emission yield of pure metals. An enhancement of the sensitivity by a factor 5 to 50 was measured in ion potassium implanted tungsten irradiated by subpicosecond laser pulses with peak intensity of a few GW/cm². Data on tungsten implanted by calcium and barium ions is also reported. An investigation of the chemical and physical properties at the origin of this enhancement are also reported. Optical and structural properties of these new materials were investigated by various techniques like reflectometry, ellipsometry, and grazing incidence x-ray diffraction, while the chemical structure of surfaces was determined by AES and XPS. The measurable changes on reflectivity and ellipsometric measurements can be interpreted both by the interface layer due to the radiation induced damage and by a modification of the dielectric function of W in the presence of ion-impurity.

The transport of electronic carriers in gold films excited by femtosecond laser pulses is observed to contain both a ballistic and an interactive component and is, accordingly, well outside thermal equilibrium. The ballistic component is observed to traverse up to 400 nm through the film at near the Fermi velocity. The interactive component undergoes various scattering events and requires a correspondingly longer time to traverse the film. The carriers which originate through laser excitation at the 'front' surface of the film arrive at the 'back' surface and influence the reflectivity. The relative reflectivity, (Delta) R/R, at the back surface is measured with a time delayed probe beam which produces information on the temporal evolution of the coupled dynamics and transport of the carriers. Experimental results are compared to a theoretical basis in terms of the Fermi-liquid theory and 1D carrier transport. Agreement between experiment and theory is remarkably good. Electron transport is also influenced by the presence of multiple metallic layers, for example, the carrier arrival profile at the back surface in gold differs from that of a corresponding Au-Ti-Au multilayer. The difference permits observation of an 'image' of the gold and multilayer interface.

A 1.1 GeV electron storage ring dedicated for UV-VUV FEL operation was commissioned last year at the Duke University Free Electron Laser Laboratory (DFELL). The UV-VUV OK-4 FEL project, based on the collaboration of the Duke FEL Laboratory and Budker Institute for Nuclear Physics (BINP, Novosibirsk, Russia) is under way. The OK-4 FEL has arrived at the Duke FEL Laboratory from Novosibirsk and is in the process of installation. High average intracavity power and natural synchronization of electron and optical pulses in the OK-4 FEL allow the production of intense inverse Compton (gamma) -rays (5-150 MeV) on the return pass of the optical pulse. The projected intensity of this (gamma) -ray source allows high energy resolution with simple geometric collimation of the (gamma) -rays. The wide tunability of the OK-4 FEL also allows us to control the (gamma) -ray energy. In this paper we discuss the processes involved in (gamma) -ray production, the influence of beam parameters and geometry on (gamma) -ray energy spread, and present projected performance of the Duke/OK-4 inverse Compton (gamma) -ray source for two simple cases. The studies reported in this paper were performed in 1993. Results were presented at the 1994 Free Electron Laser Conference (22-26 August 1994, Stanford, CA) but were published only as internal DFELL reports. A group of scientists from the Triangle University Nuclear Laboratory (TUNL) also ran independent simulations in 1994 which confirm our predictions. A workshop on the DFELL-TUNL (gamma) -ray facility was held in Durham from December 16-17, 1994 to discuss unique features of this facility and its utilization for nuclear physics and pion spectroscopy.

A 1 GeV electron storage ring dedicated for UV-VUV FEL operation was commissioned last year at the Duke University Free Electron Laser Laboratory. The XUV FEL project, based on the collaboration of the Duke FEL Laboratory and Budker Insititute for Nuclear Physics (Novosibirsk, Russia) is described. The OK-4 UV FEL has arrived from Novosibirsk at the Duke FEL laboratory and is in the process of installation. The main parameters of the DFELL storage ring, the OK-4 optical klystron, and the experimental set-up are presented. The parameters of the UV-VUV FEL are given and the possible future upgrades to this system are discussed. We have studied the dynamics of giant pulse generation in the Duke/OK-4 UV FEL. We have developed a new macro-particle code for giant pulse simulation including all known mechanisms of storage ring FEL interaction. Results of these giant pulse simulations are presented in the paper. A new mechanism of 'super-pulse' generation was discovered during these studies. It allows the generation of peak power up to 10 gigawatts using 'phase- space' refreshment of the electron beam caused by synchrotron motion.

We review time-resolved reflection high-energy electron diffraction studies of 200-ps laser heated metal surfaces. The dynamic structural behavior is shown to be strongly dependent on surface orientation and is different from that observed under slow heating conditions. Single crystals of lead and bismuth are used as examples.

We present a detailed description of ultrafast electron diffraction and its applications to study photoinduced molecular dynamics at single crystal surfaces. Experimental investigations for a new design of an ultrashort pulsed laser activated electron gun ((tau) < 5 ps) for time- resolved surface analysis are described. In addition, a novel electron detection and image analysis system, as it applies specifically to time-resolved reflection high-energy electron diffraction in the multiple-shot operation, are reviewed. The total experimental temporal resolution is discussed in terms of the electron pulse width and the time difference between an electron scattered at the front edge of the sample to an electron scattered at the trailing edge of the sample.

We present a time-resolved pump-probe low energy electron diffraction experiment to study the dynamics of surface phase transitions. An ultrashort laser pulse heats a surface on a picosecond time scale, and a time-delayed, photogenerated electron pulse probes the resultant surface dynamics by diffraction. The diffracted electrons are detected with a position-sensitive microchannel plate detector. We discuss the limits of the achievable time resolution as given by the time-dependent surface temperature jump and the temporal broadening of the electron pulse due to the kinetic energy distribution of the electrons, the focusing of the electron beam, and other effects. The instrument function of the experiment, obtained by monitoring integer- order diffraction spot intensities as a function of surface temperature (the Debye-Waller effect), may yield the electron pulse duration. A model for the surface temperature profile based on a solution to the heat conduction equation, and a simulation of the electron trajectories in the electron gun, show that the instrument function is on a picosecond time scale. Initial experiments focus on the dynamics of reconstruction on the clean, single-crystal W(001) surface.

Recent contributions by our group to the development of the experimental and theoretical bases of time-resolved studies by pulsed-beam gas electron diffraction (GED) are described. Time-resolved, stroboscopic GED experiments of laser-excited species were made possible by the development of nonphotographic, online data recording techniques. The current Arkansas detection system is characterized by an intervening light-stage that is fiber-optically coupled to either a PDA (photodiode array) or a CCD. Compared to systems in which the detector chip is directly exposed to the scattered electrons and the reactive gases inside the vacuum tank, this system is extremely durable. In experiments with a continuous electron beam, PDA detection was recently found to be one of the most precise structural tools, affording limits of precision of a few hundredths of a picometer. In kinetic studies of photo-excitation processes the traditional GED data analysis methods are not applicable because they are restricted to randomly oriented molecules with small amplitude vibrational motions in thermal equilibrium. Since laser-excited systems in general exist in nonequilibrium distributions, and the interaction with polarized laser irradiation may lead to spatially anisotropic distributions, new intensity expressions were derived for stroboscopic GED which are applicable to systems with large amplitudes, to dissociative and pre-dissociative electronic states, inverted vibrational distributions, and nonrandomly oriented molecules.

Multiphoton emission and space-charge effects require the generation of electron pulses for ultrafast time-resolved electron diffraction experiments to be performed with low power lasers. Using mode-locked laser sources it is possible to generate a train of electron pulses containing several nanoamperes, while simultaneously minimizing the number of electrons per pulse. Experiments with CCl₄ demonstrates that this electron beam current is sufficient to record diffraction patterns from gas phase samples in modecular beams. To observe time dependent structural data one must subtract the diffraction signal obtained when the pump laser is on from that with the pump laser off. This imposes stringent requirements on the system noise and drift. We present a scheme to stabilize the electron current to 0.1%, and discuss a fast-gated detection scheme to subtract diffraction patterns while maintaining single particle counting efficiency.

Numerical solutions of the time-dependent Schroedinger equation for the molecular ion H₂⁺ in intense laser fields, I approximately equals 10¹⁴ W/cm² and long wavelengths, (lambda) equals 1064 nm lead to the conclusion that electron motion can be controlled at these high intensities as a function of laser phase difference for two color photoionization, either in (omega) + 3(omega) or (omega) + 2(omega) combinations. A new phenomenon, charge resonance enhanced ionization, CREI, has been discovered and can be rationalized in terms of a tunneling model. Finally, the ionized electron is rescattered by the molecular ion core at long laser wavelengths, thus suggesting the use of such rescattering to study laser- induced electron diffraction, LIED, in order to 'clock' time-dependent nuclear dynamics.

Time-resolved macromolecular x-ray crystallography is a new capability for structural analysis driven by continuing improvements in synchrotron x-ray sources, optics, and detectors (image plates and CCDs). Protein crystal Laue data (stationary crystal and polychromatic x-rays) were recorded at SRS Daresbury station 9.5 and ESRF Grenoble beamline 3, and processed with the Daresbury Laue software package. The Laue method allows exposure times set by the synchrotron electron bunch width, e.g. 50 picoseconds. The instruments and methods developments widen opportunities for perturbation chemical crystallography studies too. A temperature dependent phase transition of a liquid crystal nickel-octahexylphthalocyanine is studied with a rapid readout CCD detector. Structure solution by molecular replacement methods with Laue data is reported for orthorhombic lysozyme. By use of tetragonal lysozyme as a test case it is shown that with fine angular intervals, wide total angular coverage of Laue exposures and the deconvolution of multiples, good connectivity of electron density maps can be realized. The monochromatic rotating crystal method offers possibilites of extremely fast rotations which allow a complete data set to be recorded onto a single image--large-angle oscillation technique (LOT). the processed LOT data looks promising. LOT electron density maps are presented.

There exists a variety of methods allowing to slow down reaction rates in proteins in order to match the lifetime of intermediates with the data collection time from standard X-ray diffraction techniques: slow substrates or less reactive mutants can be produced and cryocrystallography has proved extremely useful. Unfortunately, these methods run the risk of trapping an enzymatic reaction in the abnormal state. The development of synchrotron sources, allowing the simultaneous use of a broad range of X rays, led to a renewal of the old Laue technique. In this technique, diffraction occurs simultaneously from all lattices planes fulfilling the Bragg condition for some wavelength present in the incident X-ray beam, resulting in a considerable gain in acquisition speed. Although involving serious complication in data processing (spatial and harmonic overlaps, wavelength normalization, low resultion hole, high sensitivity to crystal disorder), the Laue method has been used successfully. With the third generation os synchrotron radiation sources like the ESRF, providing a spectacular increase in brilliance, the accessible timescales can be lowered down to the lifetime of transient species involved in genuine enzymatic reactions. Taking advantage of the low divergence, high flux, wide bandpath and temporal structure of X-ray beams delivered by these sources, one can attack the challenging task of solving the 3D structure of short lived intermediates building up on the milli, micro and even 100 psec timescale, and tertiary structural changes in proteins can be captured. The White Beam Station (BL3) at the ESRF uses the X-rays emitted by a single electron bunch (50 psec rms, ESRF single bunch mode, 5-15 mA) or by an electron "superbunch" (1 usec pulse, ESRF standard mode, 150 mA) to record Laue diffraction patterns from protein crystals on the usec to 100 psec timescales, and thus open the possibility of following conformational changes in these range of timescales. The potential of time resolved macromolecular crystallography not only depends on the existence of a high quality X-ray source, but also on the parallel development of new instrumental methods, detectors and data processing techniques. Furthermore, its success primarily relies on the availability of biochemical techniques allowing the simultaneous and uniform triggering of a transient process ini a molecules of a crystal. Temperature, pressure or pH jumps might be used to start a reaction, but the fastest timescales seem to be reachable only by photochemical activation. BL3 provides a YAG/DYE laser system synchronized witht eESRF bunchclock to trigger photodissociation of substrates or caged compounds on the nanosecond timescale.

Successful time-resolved x-ray crystallography depends on the rapid, uniform, non-damaging initiation of a structural reaction in a single crystal of excellent diffraction quality; on the ability to monitor the course of the reaction by the change in x-ray structure amplitudes in real time; and on the analysis of the changes in electron density in terms of time-independent structures of reactants, intermediates, and products. The process is greatly aided if there is an independent measure of the reaction coordinate afforded by, for example, the optical density of the crustal; and may be hindered by artifacts, some of which are thermal in origin, which degrade the x-ray diffraction patterns. Recent results which utilize multiple bunches emitted by a synchrotron source on the millisecond time domain, or a single bunch on the sub-nanosecond time domain, will be presented for several protein crystals in which the structural reaction can be initiated by light.

Topological transitions in membrane liquid crystals formed by biological lipid-water systems have been the subject of much recent interest. We have developed an x-ray diffraction system capable of initiating pressure jumps of up to 3 kbar in about 5 ms. Time-resolved x-ray diffraction patterns were obtained (approximately 9 ms each) at the National Synchrotron Light Source using two state-of-the-art CCD based detectors developed at Princeton. Numerous Bragg diffraction patterns were obtained in studying the effect of pressure on the simplest topological transitions in membranes, the lamellar to hexagonal phase transition. The patterns from one of the detectors were recorded with a signal-to-noise sufficient to measure peak positions, peak widths, and integrated areas to an accuracy adequate to test models and mechanisms of phase transition kinetics. Additional longer time-scale studies were performed using optical turbidity measurements and were found to be consistent with x-ray studies. Transition rates were found to vary by nearly 5 orders of magnitude as the difference between the final pressure and the equilibrium transition pressure was varied. As the magnitude of the pressure jump in these lyotropic systems is increased, the transition mechanism is determined not only by the rate at which water and lipid molecules transform from one phase to the new emerging phase, but also by the need for water transport. Finally, it was found that the lamellar phase acts as an intermediate phase in transitions between the gel phase and the hexagonal phase, induced by very large pressure jumps (> 2 kbar).

The focused white beam on beamline 3 is produced by the joint action of a windulator insertion device coupled to a tunable toroidal mirror. It is shown that this combination gives the highest possible flux density at the sample position. The intensity is sufficient to record single bunch Laue pattern from smaller proteins containing up to 1200 reflections per image. The possibilities of increasing the flux further by going to mini gap wundulators are also discussed. A comparison between a low (beta) and a high (beta) undulator is discussed in an attempt to boost the time resolving power in monochromatic experiments.

The high brightness, high intensity, and pulsed time-structure of synchrotron sources provide new opportunities for time-resolved x-ray diffraction investigations. With third generation synchrotron sources coming online, high brilliance and high brightness are now available in x- ray beams with the highest flux. In addition to the high average flux, the instantaneous flux available in synchrotron beams is greatly enhanced by the pulsed time structure, which consists of short bursts of x-rays that are separated by approximately tens to hundreds of nanoseconds. Time-resolved 1D and 2D position sensitive detection techniques that take advantage of synchrotron radiation for materials science x-ray diffraction ivestigations are presented, and time resolved materials science applications are discussed in terms of recent diffraction and spectroscopy results and materials research opportunities.

X-ray diffraction during transient events requires a high x-ray brightness source which is spatially collimated or spectrally concentrated, as well as synchronizable to the event. Perhaps the most demanding transient event to study is the shockwave, because the event moves at high speed and the sample possesses a high density of mechanical energy which can be hazardous to the measuring apparatus. The properties of diffraction--narrow angular acceptance, spectral requirements, shallow penetration depths, line-of-sight integration--are both enabling and limiting. This paper will discuss the factors involved in transient x-ray diffraction experiments of shocks, to include a summary of past work, and an orientation to the use of laser plasmas for both x-ray pulse production and shock generation. We have diffractively probed laser shocks in the launching of elastic compression waves and their reflection from a free surface, and have probed orthogonal lattice planes simultaneously to reveal directional differences in compression. Diffraction imaging (topography) with approximately 50 micrometers resolution has revealed microstructural effects. A focusing powder spectrum has been acquired in a static experiment. We have used x-ray streak cameras to record diffraction patterns with 50 ps resolution.

X-ray diffraction from dynamically compressed solids has been an area of active research for more than half a century. As early as 1950, Schall obtained submicrosecond, single-shot x-ray diffraction patterns of single crystals under dynamic deformation. Almost two decades later Q. Johnson and coworkers succeeded in obtaining diffraction patterns with an exposure time of tens of nanoseconds from an explosively shocked crystal, and were the first to demonstrate diffraction evidence for a shock induced phase transition. Over the past few years we have shown that even shorter exposure times can be achieved by using a laser-plasma as the source of x-rays, synchronous to a laser driven shock. In this paper we will review the progress made in this field, emphasising the potential applications fo time-resolved x-ray diffraction for addressing some of the fundamental problems of shock wave physics.

Our goal is to watch the evolution of matter on the atomic length scale and on the time scale on which elementary chemical reactions take place. We present initial experiments made in collaboration between UCSD and the INRS laboratory in Canada, on time-resolved ultrafast, 3 ps temporal resolution, near-edge x-ray absorption of gas phase SF₆ at 2.4 keV (4.89 A). We can see both the initial presence of the F atoms around the S and their absence after photodissociation produced by pumping with an intense optical pulse. Simulations of ultrafast EXAFS and diffraction experiments are presented. We are constructing an ultrahigh intensity laser to generate ultrafast x-ray pulses from laser-produced plasmas. This laser is especially designed to achieve high average power, short pulse duration and high intensity to produce very high temperature solid density plasmas and ultrahot electrons for ultrafast hard x-ray production at high x-ray photon flux, which should enable us to perform a variety of ultrafast x-ray absorption and diffraction experiments. Finally, we discuss several means to measure the duration of subpicosecond x-ray pulses.

In this paper, we present a general theory of time-resolved x-ray diffraction. To show the application of the theory, we calculate the time-resolved x-ray diffraction patterns for the laser-induced vaporization kinetics of a finite system. Effect of the system size and laser- intensity will be examined.

In a current study being conducted by the author, microcrystallines were observed to exist in amorphous, short order, structures of several metallic glasses. The observation is based on x- ray diffraction (XRD) and electron diffraction (ED) and transmission electron microscopy (TEM). The data of the x-ray diffraction show that the metallic glasses have typical amorphous structures. However, the data of the electron diffraction indicates that the metallic glasses possess polycrystalline structures. This discrepancy between the XRD and ED data can be interpreted and explained by the diffraction theory with the aid of the transmission electron microscopy. In fact results in the present work show that with a mathematical relationship originally derived by Sherrer, one can determine the boundary line between microcrystallines in amorphous, short order, structures and crystalline, long order, structures. The boundary line of microcrystallines is defined with the aid of the transmission electron microscopy in which the size of subgrains, of the metallic glasses, was determiend for the mathematical relationship.

It is shown that contribution to the integrated intensity of x-ray anomalous transmission from the elastic and plastic lattice distortions can be separated. The results of the x-ray investigations allows to determine the dislocation density, relative level and sign of the elastic distortion (lattice curvature) along selected crystallographic directions in a GaAs crystal slab. More detailed confirmation of the method for separation of the lattice distortion and dislocations density contributions to the intensity of x-ray Laue diffraction in high-absorption case was made by investigation of x-ray diffraction integrated intensity as a function of crystal thickness.

The small GTP-binding protein p21^H-ras was one of the first systems whose GTPase reaction was monitored by time-resolved crystallography. Numerous experiments have been carried out to establish the experimental design. The different parameters which needed to be optimized range from the stereochemistry of the photolabile protecting group of GTP, mutants of p21^H-ras, crystallization conditions, crystal size and photolysis conditions. Using this approach we were able to determine two intermediate structures of unstable p21^{H- ras}:GTP. In addition, we could map regions of p21^H-ras which display high dynamic behavior during the reaction. This short report will summarize the time-resolved diffraction experiments on p21^H-ras and the results that have been obtained so far. I would like to introduce this report with general aspects of time-resolved x-ray experiments and highlight these with a few recent developments on this field (without the intention of complete coverage) which provide important criteria for the design and the strategy of new experiments.

We present the first results from two highly parallel detector systems designed for fast time resolved x-ray diffraction experiments. The readout systems have been designed to give throughputs well in excess of 10⁷ events per second. The detector systems have been designed to allow high flux diffraction patterns to be collected with very much reduced rate effects when compared with previous designs. This has been achieved using wire microgap proportional counters coupled to multi-channel data acquisition systems. The efficiency and low noise of the detector coupled to the speed of the readout has produced a detector system capable of more fully exploiting the time resolved diffraction stations on the UK Synchrotron Radiation Source at the Daresbury Laboratory. Modifications to the design will be presented which will allow the system to cope with even higher count rates in the future.

A large-area pixel x-ray detector is being developed to collect eight successive frames of wide dynamic 2D images at 200kHz rates. Such a detector, to conjunction with a synchrotron radiation x-ray source, will enable time-resolved x-ray studies of proteins and other materials on time scales which have previously been inaccessible. The detector will consist of an array of fully-depleted 150 micron square diodes connected to a CMOS integrated electronics layer with solder bump-bonding. During each framing period, the current resulting from the x-rays stopped in the diodes is integrated in the electronics layer, and then strored in one of eight storage capacitors underneath the pixel. After the last frame, the capacitors are read out at standard data transmission rates. The detector has been designed for well-depth of at least 10,000 x-rays (at 12 keV), and a noise level of one x-ray. Ultimately, we intend to construct a detector with over one million pixels (1024 by 1024). We present the result of our development effort and various features of the design. The electronics design is discussed, with special attention to the performance requirements. The choice and design of the detective diodes, as they relate to x-ray stopping power and charge collection, are presented. An analysis of various methods of bump bonding is also presented. Finally, we discuss the possible need for a radiation-blocking layer, to be placed between the electronics and the detective layer, and various methods we have pursued in the construction of such a layer.