Ultrafast movies of atomic motion with femtosecond laser-based x-rays

Transcription

1 header for SPIE use Ultrafast movies of atomic motion with femtosecond laser-based x-rays Craig W. Siders a, Andrea Cavalleri a, Klaus Sokolowski-Tinten b, Ting Guo a,, Csaba Töth a, Ralph Jimenez a,, Christoph Rose-Petruck a,, Martin Kammler c, Michael Horn von Hoegen c, Dietrich von der Linde b, Kent R. Wilson a, and Christopher P.J. Barty a a The University of California San Diego, Mail Code 0339, La Jolla, CA b Universität Essen, D Essen, Germany c Universität Hannover, D Hannover, Germany ABSTRACT Using ultrafast x-ray diffraction from a laser-plasma x-ray source, we have observed coherent phonon generation and propagation in bulk (111)-GaAs, (111)-Ge, and thin (111)-Ge-on-Si films. At higher optical pump fluences, ultrafast melting of Ge films is observed. Keywords: x-ray, diffraction, ultrafast, semiconductor, solid-state 1. INTRODUCTION Many fundamental processes in nature, such as chemical reactions or phase transitions, involve changes in the structural properties of matter and atomic rearrangement. These changes usually occur on ultrafast timescales, comparable with the natural oscillation periods of atoms and molecules. Timescales, ranging from a few tens of femtoseconds to picoseconds, have been accessible for more than two decades through the use of ultrashort pulses in the visible frequency range. However, structural properties of matter are only indirectly accessible using these same optical pulses. Light of visible frequencies is in fact only sensitive to the dynamics of the valence electrons, which are delocalized over many atomic sites and therefore carry little information of the actual structural properties of matter. On the other hand, hard X-ray radiation is well suited to measure structural properties of materials, because the wavelength is comparable with the inter-atomic distances and interacts with the electrons residing in core atomic levels, spatially localized around each nucleus. Until recently however, hard X-ray pulses shorter than several tens of picoseconds were not available. Newly developed techniques for ultrafast X-ray diffraction in the picosecond and sub-picosecond regime include: 1. Femtosecond-laser driven plasma sources. The development of chirped pulse amplification 1 has enabled efficient amplification of femtosecond, laser pulses in solid-state materials such as TiAl 2 O and the development of new dispersion and amplitude control technologies has lead to the development of Tabletop multi-terawatt-class lasers 5-9. Femtosecond multi-terawatt lasers, when focused onto solid targets, generate high-density plasmas emitting short-pulses of hard X-ray line radiation 10, 11. Using silicon as target material, a femtosecond laser-driven plasma source has been shown to produce 1-keV X-ray pulses capable of resolving structural disassembly occurring within 300 fs 12. In our laboratory, an 8-keV laser-plasma X-ray source, created by focusing 25-fs, 75-mJ pulses on a moving solid-copper target, has been used to study ultrafast, coherent lattice dynamics in GaAs Thomson scattering sources. 90 o Thomson scattering between infrared Terawatt laser pulses and relativistic electrons from an accelerator 14 has been shown to produce 300-fs, broadband 30-keV pulses. Those pulses have been used to probe structural dynamics in InSb High temporal resolution Synchrotron sources: Taking advantage of ultrafast disordering of semiconductors, 2-ps temporal resolution was achieved by optically gating a 70-ps X-ray pulse diffracted from an optically pumped InSb crystal 16. A jitter-free X-ray streak camera 17 has also been also used to improve the time resolution of 70-ps synchrotron pulses to about 2 ps. In this way coherent acoustic dynamics in bulk InSb could be measured 18. Present address: The University of California Davis, USA (T.G.); The Scripps Research Institute, La Jolla, California, USA (R.J.); Department of Chemistry, Brown University, Providence, Rhode Island, USA (C.R.-P.);

2 For the small laboratory, the first of these is quite attractive for several reasons, most notably cost but also for overall utility in performing optical-pump x-ray probe experiments: the same laser source driving the x-ray pulse generation easily provides a source of intense ultrashort laser pulses both rigorously synchronous and at the same repetition rate. Our laboratory at the Department of Chemistry and Biochemistry of the University of California, San Diego has been involved for several years in the development of 10 to 20-fs, multi-terawatt laser systems 5, 6, 8, 19, 20. The ultrashort pulses from our 20-fs 5-TW system are focused onto solid copper targets at intensities exceeding W/cm 2, which results in the production of ultrashort hard X-ray bursts of CuK α radiation at a wavelength of 1.54 Å. X-ray bursts produced in this manner have been used in a series of optical pump/x-ray probe experiments in semiconductors. In these experiments, rocking curves of the photo-pumped areas of the samples are measured as a function of time delay between optical pump and X-ray probe. Figure 1 shows a schematic layout of the experiment. Figure 1: Schematic of the experimental setup for visible pump X-ray probe experiments at UCSD. Figure 1 shows a schematic of our experiment. The 800-nm laser pulses are generated at 20 Hz by an ultrafast Ti:sapphire laser producing 30-fs pulses of 200-mJ energy 6. The output of the laser is used for both sample excitation and x-ray generation. The latter is done by focusing 30-fs, 75-mJ, laser pulses onto a moving Cu wire in vacuum, resulting in a source with measured diameter of approximately 25 µm. With an average laser power of 1.0 W, approximately 5x10 10 Cu K α photons (4π steradians s) -1 are produced. The emitted K α radiation, which consists of two closely spaced, spin-orbit split, K α1 ( Å) and K α2 ( Å) lines, is diffracted from the optically pumped GaAs and detected by an x-ray CCD camera. The accumulation time is two minutes per delay step. We excite the GaAs crystal at a fluence of 59 mj/cm 2 with separate ~30-fs pulses derived from the same pulse that produces the probe x-rays. The flat-top pump beam size is set to about 4 mm 2 (95 ± 5% of peak) so that it includes the sample area encompassing the Bragg angles of the two K α lines and the auxiliary lines. The crystal is continuously moved after each time step during the experiment in order to minimize the effects of cumulative damage by the optical pump pulse. In addition to the GaAs crystal shown, bulk Ge and Ge/Si heterostructures have been investigated. The ultrashort pulse of laser-generated Cu K α x-rays diffracts in a symmetric Bragg configuration from the 3.26-Å (111) lattice spacing in GaAs, penetrating 2 µm into the bulk along the surface normal. A 30-fs pump pulse with variable time delay generates electron-hole pairs via interband excitation within the sub-µm penetration depth of the 800-nm light. By illuminating with light only a portion of the x-ray probed area, we simultaneously observe the K α lines from both photopumped and unperturbed areas of the semiconductor surface. Two-minute exposure x-ray-ccd images are normalized with respect to the incident x-ray flux and binned within the region of uniform illumination.

3 1. Coherent Lattice Dynamics in Crytals 2. ULTRAFAST X-RAY DIFFRACTION AT UCSD Crystalline GaAs is available in large samples of very high crystalline quality and is an ideal system for a quantitative optical pump, x-ray probe investigation. Its physical parameters are known with great precision 21, and numerous prior studies 22 on its ultrafast electronic properties provide a solid foundation for interpretation, and for testing the potential of ultrafast x-ray diffraction. Indeed, some information on ultrafast lattice dynamics after optical excitation has already been indirectly inferred from a variety of linear and nonlinear optical techniques However, due to the short penetration depth of visible light, no information on bulk dynamics in absorbing materials could be gained. As shown in figure 1, an ultra-short pulse of laser-generated Cu K α x-rays diffracts in a symmetric Bragg configuration from GaAs lattice. An example of our experimental results on GaAs 13 can be seen on the left hand side of Figure 2, which displays the measured time-dependent rocking curve of dynamically strained GaAs after absorption of a 30-fs, 800-nm pulse at a fluence of 20 mj/cm 2. Figure 2: Left: Measured rocking curve of dynamically strained GaAs as a function of optical pump/ X-ray probe time delay. Right: Calculated time dependent rocking curve. Zero delay corresponds to the initial deviation of the diffracted signal. We observe for early times that both of the original K α lines broaden and shift slightly to larger angles ( 20 arcseconds), while two additional transient lines appear, shifted by approximately -150 arcseconds relative to the original Bragg angle. As the pump-probe delay is increased, the new lines decrease in width, increase in intensity, and merge asymptotically with the still perturbed main lines. Finally, the angleintegrated diffraction signal (not shown) increases monotonically with delay, reaching a plateau of twice the unperturbed value. The results can be interpreted as follows. The incident optical energy couples into the material by promoting electrons from the valence to the conduction band. Single and two-photon absorption contributes to interband excitation during absorption of the pump pulse. After absorption, energy is transferred to the LO modes of the lattice via intraband relaxation 22, 30. Most of the energy is efficiently transferred to the lattice within a few picoseconds 31 and it is thermalized within the acoustic phonons

4 modes in less than 10 ps. At this point the crystal has not expanded yet, and the heating process can be regarded as isocoric (i.e. at constant volume), causing increase in the pressure (i.e. stress) but no change in lattice spacing (i.e. strain). Following our line of interpretation and the available parameters, we calculate a 300-nm thermal energy deposition length, with a maximum surface temperature increase of 800 K after 10 ps. Subsequent lattice dynamics and related propagation effects are interpreted solving the elastic equations after impulsively generated thermal stress in an absorbing solid 32. Stress is relieved by lattice expansion at the crystal surface and a travelling compression/expansion strain wave is driven into the bulk at the longitudinal speed of sound (v L = 5397 m/sec in (111)-GaAs 21 ). Figure 3 shows a qualitative sketch of this behavior. Mechanical relaxation of an impulsively stressed lattice layer of thickness d occurs within a time d v L -1, which corresponds to 300 ps for our x-ray probe depth. As both energy deposition and probing occur significantly faster than mechanical relaxation, we are in a regime where the generation and observation of coherent lattice dynamics is possible. This contrasts with many prior experiments 33, 34, where laser heating is slower than mechanical relaxation and the strain is linearly proportional to the temperature. (a) (b) (c) Figure 3: Intuitive representation of the strain profile in GaAs after impulsive heating. The laser pulse impinges upon the crystal surface from the left side of the figure, leading to rapid heating and stress formation followed by surface expansion and formation of a strain pulse. The strain pulse consists of leading compression and trailing expansion. Each lattice plane indicated in the figure corresponds to many thousand real lattice planes. Figure 2 also shows the calculated diffraction profiles, presented as a function of angular deviation from the Bragg angle, θ θ Β, and of pump-probe time delay. We numerically solve the thermo-elastic problem using the known parameters 21 for GaAs with the calculated energy deposition depth and temperature. The calculated temperature increase corresponds to a maximum surface strain of approximately 0.25%, or 8 må of lattice expansion, after the stress is released, in excellent agreement with the experimental 150 arcsecond deviation seen near zero delay in Figure 2a. A 100-ps acoustic pulse propagates into the crystal at v L and has peak bipolar strain of approximately ±0.12%, or ±3 må. The theoretical timeresolved diffraction curves of Figure 2b, corresponding to the calculated strain profile, were calculated using standard Takagi-Taupin 35, 36 techniques and convolved with the spectral and spatial distributions of our x-ray source. They are in excellent agreement with the experimental data of Figure 2a. The results can be intuitively understood as follows. The initially thin but highly strained layer of lattice expansion produces the broad, low intensity lines appearing at early time delays on the left of the main ones. As the strain wave propagates away from the surface, the layer of surface expansion thickens and these lines correspondingly become narrower and more intense, but diffract at smaller angular deviations due to the weaker average strain. In parallel, the compression wave leading the propagation into the bulk contributes to the slight shift to higher angles seen in the main lines. At the longest delays, the strain wave has largely moved beyond the depths probed by the x-ray pulse and we see diffraction lines broadened and shifted slightly to lower angles due to the exponential surface strain of the relaxed lattice. The strain for these late times is simply proportional to the temperature distribution. Finally, the theoretical angle-integrated diffraction signal similarly reproduces the monotonic increase and plateau behavior seen in the experimental data. It is important to note that the penetration depth (probed depth) of the x-rays does not coincide with the absorption depth 1/α = 6 µm. According to dynamic diffraction theory, the x-rays are depleted while propagating into the crystal and the actual extinction depth varies with angle of incidence and strain profile. For rays impinging crystal at the Bragg angle onto an unstrained GaAs the probed depth is minimum, of the order of 2 µm. When the crystal is strained, the lattice constant varies as a function of depth and the

5 Bragg condition is fulfilled over a wider range of angles. Therefore the total probed depth increases with inhomogeneous strain and the total angle-integrated diffraction signal increases. A second experiment extends our study of coherent acoustic dynamics to multi-layered semiconductors. Germanium thin films of various thicknesses are grown on bulk Silicon substrates using novel surfactant-mediated crystal growth technologies 37, 38. Taking advantage of the difference in lattice constant between Germanium and Silicon, we simultaneously measure both rocking curves of the dynamically strained structures at two different diffraction angles. Figure 4 shows the time resolved X-ray diffraction data on bulk Germanium, and 870-nm and 400-nm thick Germanium films. Figure 4: Time resolved X-ray diffraction data on (a) Bulk Ge (b) 870-nm Ge layer on Si (c) 400-nm Ge on Si. Pump fluence: 5 mj/cm 2. Zero time delay has been set arbitrarily at the earliest measured time delay. 4 Buried Si Substrate Centroid Shift [arcsec] Si K a Time Delay [ps] Figure 5: Time resolved X-ray diffraction data on a buried Si substrate, overlaid by 400-nm thick Ge. Plotted is the centroid, or center of mass, of the Si K α1 and K α2 lines as a function of time delay from the initial observation of strain in the Ge overlayer. As the bipolar acoustic pulse enters the Si substrate, a positive-going change of the diffraction angle is seen, corresponsing to a leading compressive feature. K α1 K α2

6 The bulk sample (figure 4a) has been measured as a control and shows the same qualitative behavior as that observed in bulk GaAs. Figures 4b and 4c show the generation and propagation of acoustic excitation in the two Ge films (L 1 = 870 nm and L 2 = 400 nm). The observed effects have a marked dependence on the film thickness. In the case of the 870-nm film, a slow shift to the left of the Bragg line (indicating lattice expansion) is followed by a symmetric recovery towards the right (higher angles) over a time period of about 320 ps. In the case of the 400-nm film an oscillatory pattern is observed with a period of T = 150 ps. The acoustic pulse is generated at the surface of the Germanium layer as discussed for the GaAs case, but the existence of a Ge/Si interface with discontinuous elastic properties perturbs the propagation of the pulse, which is partially reflected and transmitted across the interface. In the Germanium layer, an oscillatory behavior with period given by 2L/c (L is the film thickness and c is the speed of sound) indicates that the acoustic pulse is reflected at the interfaces. Analysis of the time resolved data from the silicon substrate of the 400-nm sample (figure 5) indicates a shift of the line towards higher diffraction angles starting approximately 80 ps after the first observable feature in the Germanium overlayer. Transmission of the acoustic pulse across the buried Ge/Si interface into the substrate causes the observed +3-arcsec shift in the Silicon lines, indicating a compressive leading edge. The peak magnitude of the compressive strain is of the order of 0.006% corresponding to an absolute change in interlayer spacing of approximately 20 femtometers. The magnitude of the measured compression feature is one order of magnitude smaller than in any previous reported data. 2. Ultrafast Melting in Semiconductors For several reasons, semiconductor melting is a natural choice when measuring atomic dynamics during ultrafast phase transitions. First, the quality of available crystalline semiconductors is extremely high and the measurement of a transition between an ordered and a disordered phase fully exploits this advantage. Second, although the physical properties of covalently bonded semiconductors are well characterized, the physics of short pulse laser melting provides a very rich and interesting phenomenology. In brief, ultrafast melting is known to follow two different physical pathways depending on the initial density of photoexcited carriers. In the case of near-threshold femtosecond pulse energies and for pulses that are longer than a few picoseconds, the melting follows a thermal behavior 39, 40. Carriers thermalize with the lattice within a few picoseconds and bring the quasi-equilibrium temperature of the crystal above the equilibrium melting temperature. In this case melting starts at the surface, where defects and inclusions serve as efficient nucleation centers. Melting then proceeds into the bulk of the crystal and a liquid-solid interface propagates at a speed that is limited by the degree of superheating (T solid > T melting ) of the interface 41. Reported interface velocities range between a few tens and a few hundred meters/sec 42, 43, requiring several hundreds of picoseconds to melt a layer a few tens of nanometers thick. Figure 6: Snapshots of a gallium arsenide surface undergoing melting after absorption of a 200-mJ/cm 2 femtosecond pulse.

7 In the case of femtosecond pulses of higher fluence 42, 43, melting proceeds as a non-thermal, ultrafast process 42-46, where lattice destabilization is caused by excitation of a dense electron-hole plasma as originally proposed by van Vechten to explain nanosecond annealing 50. It has been calculated that about one picosecond after lattice destabilization the material reaches a conventional thermodynamic state of the liquid 51. In this case, complete melting occurs on a timescale of a few picoseconds over a layer of several tens of nanometers. Figure 6 shows a time-resolved microscopy measurement reported by Sokolowski-Tinten et al. in GaAs 42. The center of the pumped spot becomes metallic (liquid) and reaches the exact reflectivity of liquid GaAs within about 500 fs, indicating that the phase transformation has occurred homogeneously over at least 20 nm, corresponding to the absorption depth of the probe light. The reflectivity of the outside area, on the other hand, continuously raises toward the reflectivity of the liquid over a time of more than 100 ps, indicating slow interface propagation and inhomogeneous melting. Optical measurements, however, do not resolve atomic dynamics during the phase transition and characterization of the optical properties 42, 43, 45, 46 is still an indirect measurement of the disassembling of the crystal. Substantial effort has been invested recently in the time resolution of femtosecond laser induced melting using ultrafast X-ray diffraction. All previous studies, however, did not clearly resolve the process 15, 16. Two major shortcomings limited the achievable results: First, the experiments were performed in bulk samples. In that case the depth over which the visible pump light induced the melting was approximately one order of magnitude smaller than the penetration depth of the X-rays. Therefore, the thickness of the optically molten surface layer was always a small fraction of the probed depth, causing a small reduction of the overall diffraction yield. Second, due to the high repetition rates, the samples could not be translated after every laser shot, causing cumulative damage and poorly controlled experimental conditions. The availability of large samples of single crystal 155-nm thin film Germanium grown on Silicon allowed us to overcome both experimental problems. We were able to pump and probe similar depths in the Germanium crystal and to translate the sample every shot, thus acquiring data using the same standards that are routinely used in ultrafast optical studies of non-reversible dynamics of solids. The diffraction signal drops significantly in the center after a few picoseconds, where the pump intensity is high enough to cause homogeneous ultrafast melting on a thick layer. On the other hand the outer area melts over the same thickness on a significantly longer time-scale, indicating thermal inhomogeneous melting nm Ge(111) on Si (111) -reflectivity [norm.] integrated K a Delay Time [ps] Figure 7: Normalized integrated diffraction yield as a function of visible pump-x-ray probe delay. Figure 7 shows the integrated diffraction efficiency in the center of the photopumped area (solid line). Also plotting is the integrated diffraction efficiency, calculated from the same x-ray diffraction data, off-center at approximately the half-pumpintensity point (dashed line). The signal has been normalized to the efficiency of the unperturbed crystal. At negative delays, the photo-pumped region of the Bragg line remains unchanged compared to the unpumped regions. A distinct drop of the

8 diffraction efficiency in the central photopumped area is observed for positive delay times, with a magnitude (20-50%) larger than that expected from the Debye-Waller reduction resulting from lattice heating. Additionally, the rapid response visible at early times indicates that a 30 to 40-nm layer has homogeneously lost crystalline order within a few picoseconds. Interface velocities above 10,000 m/sec would be required to explain such an effect in an inhomogeneous thermal process, which can be ruled out by principle considerations 41. The off-center region, with correspondingly lower pump fluence, exhibits a slower response indicative of thermal melting. The observed 20% drop in diffraction efficiency observed over 40 ps would correspond to an interface velocity of 750 m/sec, which would not be unreasonable for thermal melting of a highly superheated solid. It is interesting to note that over longer times, indicated by the point in Figure 7, a significant part of the crystalline order is re-established, indicating subsequent recrystallization of the molten material. 3. ACKNOWLEDGEMENTS The authors would like to acknowledge the significant assistance of Christian Spielmann, Jeff Squier, David Fittinghoff, and Barry Walker. KST acknowledges support by the Deutsche Forschungsgemeinschaft and by the Flughafen Frankfurt/Main Foundation. 4. REFERENCENES [1] D. Strickland and G. Mourou, Compression of Amplified Chirped Optical Pulses, Opt. Commun. 56, pp , [2] J. D. Kmetec, J. J. Macklin, and J. F. Young, 0.5-TW, 125-fs Ti:sapphire laser, Opt. Lett. 16, pp , [3] A. Sullivan, H. Hamster, H. C. Kapteyn, S. Gordon, W. White, H. Nathel, R. J. Blair, and R. W. Falcone, Multiterawatt, 100-fs laser, Opt. Lett. 16, pp , [4] J. Squier, F. Salin, G. Mourou, and D. Harter, 100-fs pulse generation and amplification in Ti:Al2O3, Opt. Lett. 16, pp , [5] C. P. J. Barty, C. L. G. III, and B. E. Lemoff, Multiterawatt 30-fs Ti:sapphire Laser System, Opt. Lett. 19, pp , [6] C. P. J. Barty, T. Guo, C. L. Blanc, F. Ráksi, C. Rose-Petruck, J. Squier, K. R. Wilson, V. V. Yakovlev, and K. Yamakawa, Generation of 18-fs, Multiterawatt Pulses Using Regenerative Pulse Shaping and Chirped Pulse Amplification, Opt. Lett. 21, pp , [7] C. P. J. Barty, G. Korn, F. Raksi, C. Rose-Petruck, J. Squier, A. C. Tien, K. R. Wilson, V. V. Yakovlev, and K. Yamakawa, Regenerative pulse shaping and amplification of ultrabroadband optical pulses, Opt. Lett. 21, pp , [8] K. Yamakawa, M. Aoyama, S. Matsuoka, H. Takuma, C. P. J. Barty, and D. Fittinghoff, Generation of 16-fs, 10- TW pulses at a 10-Hz repetition rate with efficient Ti:sapphire amplifiers, Opt. Lett. 23, pp , [9] A. Antonetti, F. Blasco, J. P. Chambaret, G. Cheriaux, G. Darpentigny, C. Le Blanc, P. Rousseau, S. Ranc, G. Rey, and F. Salin, A laser system producing 5x10 19 W/cm 2 at 10 Hz, Appl. Phys. B B65, pp , [10] J. D. Kmetec, C. L. Gordon, III, J. J. Macklin, B. E. Lemoff, G. S. Brown, and S. E. Harris, MeV X-ray generation with a femtosecond laser, Phys. Rev. Lett. 68, pp , [11] A. Rousse, P. Audebert, J. P. Geindre, F. Fallies, J. C. Gauthier, A. Mysyrowicz, G. Grillon, and A. Antonetti, Efficient K alpha X-ray source from femtosecond laser-produced plasmas, Phys. Rev. E 50, pp , [12] C. Rischel, A. Rousse, I. Uschmann, P. A. Albouy, J. P. Geindre, P. Audebert, J. C. Gauthier, E. Forster, J. L. Martin, and A. Antonetti, Femtosecond time-resolved X-ray diffraction from laser-heated organic films, Nature 390, pp , [13] C. Rose-Petruck, R. Jimenez, T. Guo, A. Cavalleri, C. W. Siders, F. Raksi, J. A. Squier, B. C. Walker, K. R. Wilson, and C. P. J. Barty, Picosecond-milliangstrom lattice dynamics measured by ultrafast X-ray diffraction, Nature 398, pp , [14] R. W. Schoenlein, W. P. Leemans, A. H. Chin, P. Volfbeyn, T. E. Glover, P. Balling, M. Zolotorev, K. J. Kim, S. Chattopadhyay, and C. V. Shank, Femtosecond X-ray pulses at 0.4 AA generated by 90 degrees Thomson scattering: a tool for probing the structural dynamics of materials, Science 274, pp , [15] A. H. Chin, R. W. Schoenlein, T. E. Glover, P. Balling, W. P. Leemans, and C. V. Shank, Ultrafast Structural Dynamics in InSb Probed by Time-Resolved X-ray Diffraction, in Ultrafast Phenomena XI, T. Elsaisser, J. G. Fujimoto, D. A. Wiersma, and W. Zinth, Eds. Berlin: Springer, 1998, pp

9 [16] J. Larsson, P. A. Heimann, A. M. Lindenberg, P. J. Schuck, P. H. Bucksbaum, R. W. Lee, H. A. Padmore, J. S. Wark, and R. W. Falcone, Ultrafast structural changes measured by time-resolved X-ray diffraction, Appl. Phys. A 66, pp , [17] T. Larsson, Z. Chang, E. Judd, P. T. Schuck, R. W. Falcone, P. A. Heimann, H. A. Padmore, H. C. Kapteyn, P. H. Bucksbaum, M. M. Murnane, R. W. Lee, A. Machacek, J. S. Wark, X. Liu, and B. Shan, Ultrafast X-ray diffraction using a streak-camera detector in averaging mode, Opt. Lett. 22, pp , [18] A. M. Lindenberg, I. Kang, S. Johnson, R. W. Falcone, P. A. Heimann, H. A. Padmore, T. Missalla, R. W. Lee, Z. Chang, M. M. Murnane, and H. C. Kapteyn, Ultrafast structural dynamics in InSb measured using time-resolved X- ray diffraction, presented at CLEO/QELS, Baltimore, Maryland, [19] G. A. Mourou, C. P. J. Barty, and M. D. Perry, Ultrahigh-Intensity Lasers: Physics of the Extreme on a Tabletop, Physics Today, pp , [20] D. N. Fittinghoff, B. C. Walker, J. A. Squier, C. S. Toth, C. Rose-Petruck, and C. P. J. Barty, Dispersion considerations in ultrafast CPA systems, IEEE J. Sel. Top. Quantum Electron. 4, pp , [21] J. S. Blakemore, Semiconducting and other major properties of gallium arsenide, J. Appl. Phys. 53, pp. R123-81, [22] J. Shah, Ultrafast Spectroscopy of Semiconductors and Semiconductor Nanostructures. Berlin: Springer Verlag, [23] D. von der Linde, J. Kuhl, and H. Klingenberg, Raman scattering from nonequilibrium LO phonons with picosecond resolution, Phys. Rev. Lett. 44, pp , [24] J. A. Kash, C. Tsang, and J. M. Hvam, Subpicosecond time-resolved Raman spectroscopy of LO phonons in GaAS, Phys. Rev. Lett. 54, pp , [25] C. Thomsen, J. Strait, Z. Vardeny, H. J. Maris, J. Tauc, and J. J. Hauser, Coherent photon generation and detection by picosecond light pulses, Phys. Rev. Lett. 53, pp , [26] G. L. Eesley, B. M. Clemens, and C. A. Paddock, Generation and detection of picosecond acoustic pulses in thin metal films, Appl. Phys. Lett. 50, pp , [27] G. C. Cho, W. Kutt, and H. Kurz, Subpicosecond time-resolved coherent-phonon oscillations in GaAs, Phys. Rev. Lett. 65, pp , [28] Y. M. Chang, L. Xu, and H. W. K. Tom, Observation of coherent surface optical phonon oscillations by timeresolved surface second-harmonic generation, Phys. Rev. Lett. 78, pp , [29] A. Bartels, T. Dekorsy, H. Kurz, and K. Kohler, Coherent zone-folded longitudinal acoustic phonons in semiconductor superlattices: excitation and detection, Phys. Rev. Lett. 82, pp , [30] M. C. Downer and C. V. Shank, Ultrafast heating of silicon on sapphire by femtosecond optical pulses, Phys. Rev. Lett. 56, pp , [31] U. Strauss, W. W. Ruhle, and K. Kohler, Auger recombination in intrinsic GaAs, Appl. Phys. Lett. 62, pp. 55-7, [32] C. Thomsen, H. T. Grahn, H. J. Maris, and J. Tauc, Surface generation and detection of phonons by picosecond light pulses, Phys. Rev. B 34, pp , [33] J. R. Buschert, J. Z. Tischler, D. M. Mills, Q. Zhao, and R. Colella, Time resolved X-ray diffraction study of laser annealing in silicon at grazing incidence, J. Appl. Phys. 66, pp , [34] B. C. Larson, J. Z. Tischler, and D. M. Mills, Nanosecond resolution time-resolved X-ray study of silicon during pulsed-laser irradiation, J. Mat. Res. 1, pp , [35] S. Takagi, A Dynamical Theory of Diffraction for a Distorted Crystal, J. Phys. Soc. Japan 26, pp , [36] D. Taupin, Theorie dynamique de la diffraction des rayon X par les cristaux deformes, Bull. Soc. franc. Miner. Crist. 87, pp , [37] M. Horn-von Hoegen, Surfactants: perfect heteroepitaxy of Ge on Si(111), Appl. Phys. A A59, pp , [38] M. Horn-von Hoegen, M. Copel, J. C. Tsang, M. C. Reuter, and R. M. Tromp, Surfactant-mediated growth of Ge on Si(111), Phys. Rev. B 50, pp , [39] E. Rimini, P. Baeri, and G. Foti, Laser pulse energy dependence of annealing in ion implanted Si and GaAs semiconductors, Phys. Lett. A 65A, pp , [40] D. H. Auston, C. M. Surko, T. N. C. Venkatesan, R. E. Slusher, and J. A. Golovchenko, Time-resolved reflectivity of ion-implanted silicon during laser annealing, Appl. Phys. Lett. 33, pp , [41] F. Spaepen and D. Turnbull, in Laser Annealing of Semiconductors, J. M. Poate and J. W. Mayer, Eds.: Academic Press, [42] K. Sokolowski-Tinten, J. Bialkowski, M. Boing, A. Cavalleri, and D. von der Linde, Thermal and nonthermal melting of gallium arsenide after femtosecond laser excitation, Phys. Rev. B 58, pp. R , 1998.

10 [43] C. V. Shank, R. Yen, and C. Hirlimann, Femtosecond-time-resolved surface structural dynamics of optically excited silicon, Phys. Rev. Lett. 51, pp , [44] K. Sokolowski-Tinten, J. Bialkowski, and D. von der Linde, Ultrafast laser-induced order-disorder transitions in semiconductors, Phys. Rev. B 51, pp , [45] H. W. K. Tom, G. D. Aumiller, and C. H. Brito-Cruz, Time-resolved study of laser-induced disorder of Si surfaces, Phys. Rev. Lett. 60, pp , [46] P. Saeta, J. K. Wang, Y. Siegal, N. Bloembergen, and E. Mazur, Ultrafast electronic disordering during femtosecond laser melting of GaAs, Phys. Rev. Lett. 67, pp , [47] P. Stampfli and K. H. Bennemann, Theory for the instability of the diamond structure of Si, Ge, and C induced by a dense electron-hole plasma, Phys. Rev. B 42, pp , [48] P. Stampfli and K. H. Bennemann, Dynamical theory of the laser-induced lattice instability of silicon, Phys. Rev. B 46, pp , [49] P. Stampfli and K. H. Bennemann, Time dependence of the laser-induced femtosecond lattice instability of Si and GaAs: role of longitudinal optical distortions, Phys. Rev. B 49, pp , [50] J. A. Van Vechten, R. Tsu, and F. W. Saris, Nonthermal pulsed laser annealing of Si, plasma annealing, Phys. Lett. A 74A, pp , [51] P. L. Silvestrelli, A. Alavi, M. Parrinello, and D. Frenkel, Ab initio molecular dynamics simulation of laser melting of silicon, Phys. Rev. Lett. 77, pp , 1996.