Single-Shot Electron Diffraction using a Cold Atom Electron Source
Rory W. Speirs, Corey T. Putkunz, Andrew J. McCulloch, Keith A. Nugent, Benjamin M. Sparkes, Robert E. Scholten
SSingle-Shot Electron Diffraction using a Cold Atom Electron Source
Rory W. Speirs, Corey T. Putkunz, Andrew J. McCulloch, Keith A. Nugent, Benjamin M. Sparkes, andRobert E. Scholten a) School of Physics, The University of Melbourne, Victoria 3010, Australia ARC Centre of Excellence for Advanced Molecular Imaging, Department of Physics, La Trobe University,Victoria 3086, Australia (Dated: 23 July 2018)
Cold atom electron sources are a promising alternative to traditional photocathode sources for use in ul-trafast electron diffraction due to greatly reduced electron temperature at creation, and the potential for acorresponding increase in brightness. Here we demonstrate single-shot, nanosecond electron diffraction frommonocrystalline gold using cold electron bunches generated in a cold atom electron source. The diffractionpatterns have sufficient signal to allow registration of multiple single-shot images, generating an averagedimage with significantly higher signal-to-noise ratio than obtained with unregistered averaging. Reflectionhigh-energy electron diffraction (RHEED) was also demonstrated, showing that cold atom electron sourcesmay be useful in resolving nanosecond dynamics of nanometre scale near-surface structures.Keywords: laser cooled atoms, cold electron source, ultrafast electron diffraction, single-shot diffraction
I. INTRODUCTION
Ultrafast single-shot diffraction is revolutionising ourunderstanding of materials science, chemistry, and biol-ogy, by imaging objects on atomic length and time scalessimultaneously . X-ray free electron lasers (XFELs)have been used to perform single-shot coherent diffractiveimaging on micro- and nano-metre scale targets , wherethe imaging pulse is briefer than the time scale of dam-age to the object . An alternative and complementaryapproach is ultrafast electron diffraction, which benefitsnot only from much stronger scattering of electrons rela-tive to X-rays , but also significantly reduced damage perelastic scattering event . To enable single-shot diffractionstudies, the number of electrons per pulse must be of or-der 10 or greater to have sufficient signal per pixel at thedetector . This number is easily achievable with pho-tocathode sources, and when combined with RF bunchcompression, sub 100 fs pulses have been achieved . Thebrightness of photocathode sources is limited by the highinitial temperature of the extracted electrons (10 K),leading to a high transverse emittance . The emittancerequired for single-shot imaging depends on the size ofthe object being imaged: larger object sizes or crystalperiods require lower emittance to generate useful coher-ent diffraction patterns. Ultrafast single-shot electrondiffraction has been achieved from large single crystaland polycrystalline samples using a variety of photocath-ode based sources , but insufficient source brightnesshas prevented demonstration for micro or nanocrystals,or single molecules.Cold electron sources are a promising alternative tosolid state photocathodes, producing electrons by nearthreshold photoionisation of laser cooled atoms. Elec-trons from these sources have an intrinsic temperature a) Electronic mail: [email protected] as low as 10 K, which for a given flux leads to severalorders of magnitude increase in brightness . The lowelectron temperature, together with the ability to spa-tially shape the beam , should allow these sources toproduce uniformly filled ellipsoid bunches, which do notsuffer emittance degradation resulting from non-linear in-ternal Coulomb forces .A cold electron source was recently used for thefirst time to generate a transmission electron diffractionpattern . In that experiment, cold, sub-picosecond elec-tron pulses containing a few hundred electrons were scat-tered by graphite. To produce diffraction patterns withclearly discernible Bragg reflections, several thousand in-dividual shots were integrated.Here we present the first single-shot electron diffrac-tion patterns obtained using a cold electron source. Thepatterns were obtained from a monocrystalline gold foilusing a single 5 ns bunch of 5 × electrons. No electronaperture was required due to the high spatial coherenceof the electrons at the source. This allowed all gener-ated electrons to contribute to the image, resulting in asingle shot with sufficiently high signal-to-noise ratio foridentification of the sample and registration of succes-sive images. Single-shot reflection high-energy electrondiffraction (RHEED) was also demonstrated from a waferof monocrystalline silicon. II. THE COLD ATOM ELECTRON SOURCE
The cold atom electron source (CAES) generates elec-trons by photoionisation of rubidium-85 atoms in amagneto-optical trap, which is positioned between twoaccelerating electrodes as shown in figure 1b.The photoionisation is a two-stage process (figure 1a).The atoms are excited from the 5 S / ( F = 3) groundstate to the 5 P / ( F = 4) excited state using a 100 nspulse of laser light of wavelength 780 nm. A 5 ns pulsefrom the ionisation laser (wavelength 480 nm) then drives a r X i v : . [ phy s i c s . acc - ph ] S e p (a) Rb + (b) Beam block10 (i) (ii) n m FIG. 1. (a) Rubidium atoms are ionised in a two step process: 780 nm laser light drives them to the first excited state wherethey are ionised by a 5 ns pulse of 480 nm light. (b) The electrons produced are accelerated by a static electric field, focused,and scattered off a sample, either in transmission (i) or reflection (ii) geometries. Distances are in millimetres. the atoms either to a Rydberg level, or directly to thecontinuum.The excitation laser illuminates the atom cloud alongthe axis of electron propagation (longitudinal direction),and the focused intensity profile can be changed arbitrar-ily using a liquid crystal spatial light modulator, whichdefines the shape of the electron bunch in the two di-mensions transverse to propagation . The blue ionisa-tion laser illuminates the atom cloud transversely to theelectron propagation axis, defining the longitudinal pro-file of the electron bunch which is generated in the regionof overlap of the two laser beams such that the bunch isshaped in all three dimensions.The ionisation time is determined by the temporal pro-file of the blue tunable dye laser pulse, with full widthhalf maximum (FWHM) duration of 5 ns, pulse energyof 5 mJ at the cloud, and repetition rate of 10 Hz. Theblue laser is focused with a cylindrical lens onto the atomcloud, so that the ionisation region is defined by a ribbonof light with a FWHM width of 30 µ m in the longitudinaldirection.Before ionisation the atom cloud has a peak den-sity of approximately 10 atoms cm − , and temperature100 µ K. The quadrupole magnetic field is switched offand allowed to decay for 3 . . − and ablue beam width of 30 µ m results in an electron energyspread of 8 eV. This is a relatively high energy spreadcompared to sources used in conventional electron micro- scopes, where chromatic aberration in the strong objec-tive lens drives the need for low energy dispersion. Thecontribution to the point spread function due to chro-matic aberration is proportional to the beam semi-angleaccepted into the lens, and for the single weak condenserlens used in our setup, this contribution is significantlysmaller than the detector resolution. Polychromaticityalso results in a spread of diffraction angles for any givensample spatial frequency, limiting the achievable resolu-tion in coherent diffractive imaging. We typically use anelectron energy of 8 keV for diffraction experiments, giv-ing a fractional energy spread of ∆ E/E < . . The energy spread could be reduced fur-ther if required by tailoring the extraction field strengthor reducing the focal spot size of the blue laser beam,though the latter method would also reduce the numberof electrons generated.After the electron bunch leaves the accelerator, it tra-verses a solenoid magnetic lens at a distance of 225 mm,before drifting 323 mm to the sample. The low numeri-cal aperture of the lens limits the ability to create verysmall beam sizes at the sample, but results in a highlycollimated beam without the need to introduce furtherelectron optics. After the target sample the diffractedelectrons propagate 77 mm to a phosphor-coupled mi-crochannel plate (MCP) detector which is imaged with acamera. III. BEAM PARAMETERS
We used a Gaussian excitation laser beam with aFWHM width of 80 µ m at the focus. However, as ourintension was to ionise as many electrons as possible, weused a very high power in the beam (with a peak in-tensity of thousands of times the saturation intensity).This resulted in significant excitation even a long wayfrom the centre of the beam, as well as significant fluo-rescence and reabsorption, which both have the effect ofincreasing the excitation area. The excited area was ulti-mately determined by measuring the unfocused electronbunch size at the detector, along with the known magni-fication of the beam path. Using this method, the elec-tron bunch at the source was determined to be a peakedshape, with a FWHM width of 1 . , electrons generated in this method areknown to have a source divergence of σ θ x = 0 . µ rad,resulting in a source emittance for the bunch generatedhere of (cid:15) x = 50 nm rad. The electrons were focused to aminimum spot size at the MCP as shown in figure 1, re-sulting in a beam width at the sample of approximately300 µ m, with a corresponding coherence length at thispoint of (cid:96) c = 2 nm.Using a Faraday cup, the number of electrons per pulsewas measured to be 5 × (80 fC), corresponding to anionisation fraction of approximately 50% within the ioni-sation region of the atom cloud when taking into accountthe density of the cloud, and the volume of the illumi-nated region.The bunch temporal profile was estimated to be Gaus-sian with a FWHM duration of 5 ns based on the pulselength of the blue laser. While the actual electron pro-file may differ from this significantly due to effects likea rapid depletion of excited state atom population, oran extended laser tail, 5 ns is expected to offer a fairlygood characteristic timescale over which the bunch isproduced. Combining the measured electron numberper pulse, and the estimated temporal profile, yieldsa peak beam current of 20 µ A. When combined withthe estimated emittance, this gives a peak brightness of B = 3 × A m − sr − , using the same conventions asin ref. .The emittance and bunch charge for the CAES aretherefore approaching the values required for single-shotdiffraction of microcrystals , but the pulse duration ispossibly still three orders of magnitude too long to avoiddegradation of the diffraction pattern due to beam in-duced damage of such small samples. However recentstudies have suggested the constraints on pulse dura-tion due to damage could be relaxed for electrons com-pared with X-rays, because of the differences between thescattering and damage-inducing processes . To achievesub-picosecond ultrafast electron diffraction, the ionisa-tion process can be modified to use femtosecond ratherthan nanosecond laser pulses . Picosecond to fem-tosecond duration electron bunches containing the samecharge will require spatial beam shaping in order toavoid brightness degradation caused by the otherwisenon-linear space charge expansion of the bunch . IV. SINGLE-SHOT ELECTRON DIFFRACTION
We demonstrated diffraction of electron bunches gener-ated in the CAES from an 11 nm thick gold foil mountedon a 3 mm transmission electron microscopy grid, with an electron energy of 8 keV. Figure 2 shows the diffractionpattern from a single 5 ns electron bunch. The resultingBragg reflections are clearly visible out to the (240) spotat a resolution of 1 . − , where the crystallographic con-vention has been adopted for reciprocal lattice vectors,such that | g hkl | = 1 /d hkl , where d hkl is distance betweenatomic planes in real space.The reflections show the four-fold symmetry of the goldface-centered cubic (fcc) lattice, and the { } and { } reflections visible on the sides and corners of the innersquare confirm the unit cell parameter as 0 .
407 nm, con-sistent with the agreed value for gold of 0 . ◦ C . To obtain a higher signal-to-noise ratio, 2000shots were averaged. The result (figure 3) allows Braggspots to be identified out to the (660) reflection, with aneffective resolution of 2 .
08 ˚A − , limited by the size of thedetector.Closer inspection reveals that the Bragg spots havebeen significantly broadened when compared to thesingle-shot case, indicating that the transverse coher-ence of the time-averaged electron beam is reduced com-pared to any single constituent bunch. This loss of co-herence stems from a slight beam wobble due to smallvariations in the decay of the quadrupole magnetic fieldafter it is switched off. The effect of the beam drifton the diffraction pattern is analogous to the vary-ing shot-to-shot diffraction patterns obtained in XFELnanocrystal diffraction experiments, where diffractionpatterns are obtained from millions of randomly alignednanocrystals .To compensate for the beam wobble, successive single-shot images were registered. The eleven brightest spotswere used to adjust the alignment by performing a crosscorrelation of the individual images and the unregisteredaverage in the region surrounding the spots. The result-ing registered average can be seen in figure 4, showingnotably sharper Bragg spots.A lineout (figure 5) of the (200) Bragg reflection inthe b ∗ direction shows how the direct average reducesthe noise level compared to a single shot, at the expenseof increasing peak width. The registered average main-tains the low noise level of the direct average, while fullyrecovering the peak resolution. These results emphasisethat the imaging is indeed effectively single-shot, withfeatures clearly visible above the noise out to a resolu-tion to 1 . − .Due to the structure amplitude of fcc gold, the only al-lowed reflections are those where the Miller indices h, k, l ,are all even or all odd. The diffraction images of goldwere taken along the (cid:104) (cid:105) zone axis, where the latticeamplitude dictates that reflections are only allowed if h and k are even. While at low diffraction angles theserules are obeyed, it can be seen in figures 3 and 4 thatsome kinematically disallowed reflections are visible athigher angles. This is caused by a departure from thesingle scattering kinematic approximation, where bothelastically, and inelastically scattered electrons are thenre-scattered into directions which are forbidden to the -1 camera signal (arb.) FIG. 2. Single-shot transmission electron diffraction fromgold, formed from a 5 ns pulse of cold electrons. Main im-age is logarithmically scaled, inset is linearly scaled. camera signal (arb.) -1 FIG. 3. 2000 diffraction shots from gold, directly averaged.Averaging results in a higher signal-to-noise ratio, but shot-to-shot beam instabilities lead to a broadening of the Braggpeaks. Main image is logarithmically scaled, inset is linearlyscaled. unscattered incident beam. It can also be seen from fig-ure 6 that the Bragg reflections are accompanied by twoor more satellites, offset from the main reflection at 45 ◦ from the direction of the reciprocal basis vectors.These satellites are the result of { } crystal twin-ning, which can form when (100) gold films are prepared -1 a * b * camera signal (arb.) FIG. 4. 2000 diffraction shots from gold averaged by regis-tering individual images. Registration is possible due to thehigh signal-to-noise ratio in the single shots, and recovers thesharpness in Bragg peaks lost in direct averaging. The recip-rocal lattice vectors a ∗ , b ∗ are drawn to scale. Main image islogarithmically scaled, inset is linearly scaled. Single shotDirect averageRegistered –1.0 –0.5 0.0 0.5 1.01.00.80.60.40.20.0 Position (mm) R e l a t i v e i n t e n s i t y FIG. 5. A lineout of the (200) Bragg reflection in the b ∗ direction. Registering the single shots averages out the noisewithout resulting in Bragg spot broadening, as happens whenshots are directly averaged. by evaporation . There are two different mechanismsbehind the satellite creation. The two relatively strongsatellites that can be seen around the (200) and (020)Bragg reflections are created directly by diffraction fromthe crystal twins. This also produces the rightmost spotaround the (220) reflection. The other satellites aroundthe (220) reflection arise from double diffraction, where -1 a * b * camera signal (arb.) FIG. 6. The two satellite spots around the (200) and (020)reflections, and the rightmost spot around the (220) reflection,are due to diffraction from crystal twins. The other spotsvisible around the (220) reflection are due to the electrondiffracting twice: once each from two different twins. Thesmall arrows indicate the positions of the faint satellite spots.The reciprocal lattice vectors are not to scale. electrons are first diffracted by a (100) oriented domain,and then diffracted again by one of the twins. Dynami-cal effects would not normally be seen with such a thinsample because electron energies used in a transmissionelectron microscope are typically ten times greater thanused here, resulting in a much lower scattering cross sec-tion and interactions that more closely match the simplekinematic theory.Reflection high-energy electron diffraction (RHEED)is a surface sensitive diffraction technique routinely usedto monitor crystal surface quality and epitaxial crys-tal growth , and has also been used to observe sur-face dynamics resulting from illumination by ultrafastlasers . RHEED is a useful technique to further demon-strate diffraction from our source, both because the elec-tron energies typically required fall into the range wecan easily generate, and because high quality single crys-tals are more readily available as bulk wafers than asthe nanometre-thick foils needed for transmission elec-tron diffraction. To adjust the system for RHEED, allthat was required was to rotate the sample through 90degrees as shown in figure 1b(ii). Figure 7 shows RHEEDpatterns from a (cid:104) (cid:105) silicon wafer, which was HF etchedto remove the native oxide layer immediately prior totransfer to the vacuum chamber. The beam was nomi-nally incident on the crystal from the [110] direction at 0 ◦ polar angle. Since the sample stage could not be rotatedin the azimuthal direction, it is likely that the wafer wasslightly misaligned, which would account for the appar-ent horizontal asymmetry of the Bragg reflections at anyparticular polar angle. For clarity the RHEED patternsshown are 100 shot averages, however Bragg reflectionswere easily visible from a single shot as shown by theinset.Cold electron RHEED offers a promising opportunityto investigate near-surface dynamics on nanosecond timescales. The high transverse coherence of the beam shouldalso allow coherent scattering to be observed from struc-tures tens of nanometres wide, such as quantum dots camera signal (arb.) FIG. 7. 100 shot averages of RHEED from silicon (cid:104) (cid:105) ata range of polar angles. Inset: a single shot of the regionindicated, clearly showing a Bragg reflection. and optical metamaterials. Cold electron sources withbunch shaping to control space-charge induced bright-ness degradation are perhaps uniquely placed to performthese studies, due to their potential to deliver high bunchcharges and high coherence at relatively low electron en-ergies. Using very high electron energies to mitigatespace-charge effects is not an option for RHEED, sincevery energetic electrons penetrate too deeply to accu-rately probe surface structure.
V. SUMMARY
We have demonstrated single-shot electron diffractionusing fast electron bunches produced with a cold atomelectron source. The 5 ns bunches contained around5 × electrons, and because of their low temperatureand high coherence, no beam aperture was required, al-lowing all generated electrons to contribute to imaging.When scattered by a single crystal of gold, the resultingsingle-shot diffraction pattern contained sufficient signalto give information about the crystal structure withoutaveraging. The large signal-to-noise ratio allowed sub-sequent shots to be merged through image registration,which compensated for shot-to-shot beam drift that de-graded the image quality when directly summing. Single-shot diffraction pattens have also been obtained in reflec-tion mode, which may prove useful for investigating thedynamics of nanometre scale surface structures, wherehigh beam coherence is necessary. Demonstrating single-shot diffraction is a significant step forward for cold atomelectron sources, and supports the promise that theycould complement solid state photocathode sources foruse in ultrafast single-shot electron diffraction experi-ments. ACKNOWLEDGEMENTS
We thank L. J. Allen for helpful discussions and advice.BMS gratefully acknowledges the support of a Univer-sity of Melbourne McKenzie Fellowship. This work wassupported by the Australian Research Council DiscoveryProject DP140102102.
REFERENCES A. Barty, S. Boutet, M. J. Bogan, S. Hau-Riege, S. March-esini, K. Sokolowski-Tinten, N. Stojanovic, R. Tobey, H. Ehrke,A. Cavalleri, S. Dusterer, M. Frank, S. Bajt, B. W. Woods, M. M.Seibert, J. Hajdu, R. Treusch, and H. N. Chapman, “Ultrafastsingle-shot diffraction imaging of nanoscale dynamics,”
Nat Pho-ton , vol. 2, pp. 415–419, 07 2008. J. R. Dwyer, C. T. Hebeisen, R. Ernstorfer, M. Harb, V. B.Deyirmenjian, R. E. Jordan, and R. J. D. Miller, “Femtosecondelectron diffraction: ’making the molecular movie’,”
Philosophi-cal Transactions: Mathematical, Physical and Engineering Sci-ences , vol. 364, no. 1840, pp. pp. 741–778, 2006. A. H. Zewail, “4d ultrafast electron diffraction, crystallography,and microscopy,”
Annual Review of Physical Chemistry , vol. 57,no. 1, pp. 65–103, 2006. PMID: 16599805. M. M. Seibert, T. Ekeberg, F. R. N. C. Maia, M. Svenda, J. An-dreasson, O. Jonsson, D. Odic, B. Iwan, A. Rocker, D. Westphal,M. Hantke, D. P. DePonte, A. Barty, J. Schulz, L. Gumprecht,N. Coppola, A. Aquila, M. Liang, T. A. White, A. Martin,C. Caleman, S. Stern, C. Abergel, V. Seltzer, J.-M. Claverie,C. Bostedt, J. D. Bozek, S. Boutet, A. A. Miahnahri, M. Messer-schmidt, J. Krzywinski, G. Williams, K. O. Hodgson, M. J.Bogan, C. Y. Hampton, R. G. Sierra, D. Starodub, I. Ander-sson, S. Bajt, M. Barthelmess, J. C. H. Spence, P. Fromme,U. Weierstall, R. Kirian, M. Hunter, R. B. Doak, S. Marchesini,S. P. Hau-Riege, M. Frank, R. L. Shoeman, L. Lomb, S. W.Epp, R. Hartmann, D. Rolles, A. Rudenko, C. Schmidt, L. Fou-car, N. Kimmel, P. Holl, B. Rudek, B. Erk, A. Homke, C. Re-ich, D. Pietschner, G. Weidenspointner, L. Struder, G. Hauser,H. Gorke, J. Ullrich, I. Schlichting, S. Herrmann, G. Schaller,F. Schopper, H. Soltau, K.-U. Kuhnel, R. Andritschke, C.-D.Schroter, F. Krasniqi, M. Bott, S. Schorb, D. Rupp, M. Adolph,T. Gorkhover, H. Hirsemann, G. Potdevin, H. Graafsma, B. Nils-son, H. N. Chapman, and J. Hajdu, “Single mimivirus particlesintercepted and imaged with an x-ray laser,”
Nature , vol. 470,pp. 78–81, 02 2011. H. N. Chapman, A. Barty, M. J. Bogan, S. Boutet, M. Frank,S. P. Hau-Riege, S. Marchesini, B. W. Woods, S. Bajt, W. H.Benner, R. A. London, E. Plonjes, M. Kuhlmann, R. Treusch,S. Dusterer, T. Tschentscher, J. R. Schneider, E. Spiller,T. Moller, C. Bostedt, M. Hoener, D. A. Shapiro, K. O. Hodg-son, D. van der Spoel, F. Burmeister, M. Bergh, C. Caleman,G. Huldt, M. M. Seibert, F. R. N. C. Maia, R. W. Lee, A. Szoke,N. Timneanu, and J. Hajdu, “Femtosecond diffractive imagingwith a soft-x-ray free-electron laser,”
Nature Physics , vol. 2,pp. 839–842, 2006. R. Neutze, R. Wouts, D. van der Spoel, E. Weckert, and J. Ha-jdu, “Potential for biomolecular imaging with femtosecond x-raypulses,”
Nature , vol. 406, pp. 752–757, 08 2000. H. M. Quiney and K. A. Nugent, “Biomolecular imaging andelectronic damage using x-ray free-electron lasers,”
Nat Phys ,vol. 7, pp. 142–146, 02 2011. F. Carbone, P. Musumeci, O. Luiten, and C. Hebert, “A per-spective on novel sources of ultrashort electron and x-ray pulses,”
Chemical Physics , vol. 392, no. 1, pp. 1 – 9, 2012. R. Henderson, “The potential and limitations of neutrons, elec-trons and x-rays for atomic resolution microscopy of unstained biological molecules,”
Quarterly Reviews of Biophysics , vol. 28,pp. 171–193, 5 1995. O. Krivanek and P. Mooney, “Applications of slow-scan { CCD } cameras in transmission electron microscopy,” Ultramicroscopy ,vol. 49, no. 14, pp. 95 – 108, 1993. T. van Oudheusden, P. L. E. M. Pasmans, S. B. van der Geer,M. J. de Loos, M. J. van der Wiel, and O. J. Luiten, “Compres-sion of subrelativistic space-charge-dominated electron bunchesfor single-shot femtosecond electron diffraction,”
Phys. Rev.Lett. , vol. 105, p. 264801, Dec 2010. B. J. Claessens, S. B. van der Geer, G. Taban, E. J. D. Vreden-bregt, and O. J. Luiten, “Ultracold electron source,”
Phys. Rev.Lett. , vol. 95, p. 164801, Oct 2005. S. Tokita, S. Inoue, S. Masuno, M. Hashida, and S. Sakabe,“Single-shot ultrafast electron diffraction with a laser-acceleratedsub-mev electron pulse,”
Applied Physics Letters , vol. 95, no. 11,pp. –, 2009. R. Li, W. Huang, Y. Du, L. Yan, Q. Du, J. Shi, J. Hua, H. Chen,T. Du, H. Xu, and C. Tang, “Note: Single-shot continuouslytime-resolved mev ultrafast electron diffraction,”
Review of Sci-entific Instruments , vol. 81, no. 3, pp. –, 2010. P. Musumeci, J. T. Moody, C. M. Scoby, M. S. Gutierrez, H. A.Bender, and N. S. Wilcox, “High quality single shot diffractionpatterns using ultrashort megaelectron volt electron beams froma radio frequency photoinjector,”
Review of Scientific Instru-ments , vol. 81, no. 1, pp. –, 2010. G. Sciaini, M. Harb, S. G. Kruglik, T. Payer, C. T. Hebeisen,F.-J. M. z. Heringdorf, M. Yamaguchi, M. H.-v. Hoegen, R. Ern-storfer, and R. J. D. Miller, “Electronic acceleration of atomicmotions and disordering in bismuth,”
Nature , vol. 458, pp. 56–59,03 2009. B. J. Siwick, J. R. Dwyer, R. E. Jordan, and R. J. D. Miller, “Anatomic-level view of melting using femtosecond electron diffrac-tion,”
Science , vol. 302, no. 5649, pp. 1382–1385, 2003. A. J. McCulloch, D. V. Sheludko, M. Junker, and R. E. Scholten,“High-coherence picosecond electron bunches from cold atoms,”
Nat Commun , vol. 4, p. 1692, 04 2013. W. J. Engelen, M. A. van der Heijden, D. J. Bakker, E. J. D.Vredenbregt, and O. J. Luiten, “High-coherence electron bunchesproduced by femtosecond photoionization,”
Nat Commun , vol. 4,p. 1693, 04 2013. A. J. McCulloch, D. V. Sheludko, S. D. Saliba, S. C. Bell,M. Junker, K. A. Nugent, and R. E. Scholten, “Arbitrarilyshaped high-coherence electron bunches from cold atoms,”
NatPhys , vol. 7, pp. 785–788, 10 2011. O. J. Luiten, S. B. van der Geer, M. J. de Loos, F. B. Kiewiet, andM. J. van der Wiel, “How to realize uniform three-dimensionalellipsoidal electron bunches,”
Phys. Rev. Lett. , vol. 93, p. 094802,Aug 2004. M. W. van Mourik, W. J. Engelen, E. J. D. Vredenbregt, andO. J. Luiten, “Ultrafast electron diffraction using an ultracoldsource,”
Structural Dynamics , vol. 1, no. 3, pp. 1 – 7, 2014. M. Zurch, J. Rothhardt, S. Hadrich, S. Demmler, M. Krebs,J. Limpert, A. Tunnermann, A. Guggenmos, U. Kleineberg, andC. Spielmann, “Real-time and sub-wavelength ultrafast coherentdiffraction imaging in the extreme ultraviolet,”
Sci. Rep. , vol. 4,12 2014. T. van Oudheusden, E. F. de Jong, S. B. van der Geer, W. P. E.M. O. t Root, O. J. Luiten, and B. J. Siwick, “Electron sourceconcept for single-shot sub-100 fs electron diffraction in the 100kev range,”
Journal of Applied Physics , vol. 102, no. 9, pp. –,2007. R. F. Egerton, “Outrun radiation damage with electrons?,”
Ad-vanced Structural and Chemical Imaging , vol. 1, no. 1, 2015. O. J. Luiten, B. J. Claessens, S. B. van der Geer, M. P. Reijn-ders, G. Taban, and E. J. D. Vredenbregt, “Ultracold electronsources,”
International Journal of Modern Physics A , vol. 22,no. 22, pp. 3882–3897, 2007. A. Maeland and T. B. Flanagan, “Lattice Spacings of GOLD-PALLADIUM Alloys,”
Canadian Journal of Physics , vol. 42, pp. 2364–2366, Nov. 1964. H. N. Chapman, P. Fromme, A. Barty, T. A. White, R. A. Kirian,A. Aquila, M. S. Hunter, J. Schulz, D. P. DePonte, U. Weier-stall, R. B. Doak, F. R. N. C. Maia, A. V. Martin, I. Schlichting,L. Lomb, N. Coppola, R. L. Shoeman, S. W. Epp, R. Hart-mann, D. Rolles, A. Rudenko, L. Foucar, N. Kimmel, G. Wei-denspointner, P. Holl, M. Liang, M. Barthelmess, C. Caleman,S. Boutet, M. J. Bogan, J. Krzywinski, C. Bostedt, S. Bajt,L. Gumprecht, B. Rudek, B. Erk, C. Schmidt, A. Homke, C. Re-ich, D. Pietschner, L. Struder, G. Hauser, H. Gorke, J. Ull-rich, S. Herrmann, G. Schaller, F. Schopper, H. Soltau, K.-U. Kuhnel, M. Messerschmidt, J. D. Bozek, S. P. Hau-Riege,M. Frank, C. Y. Hampton, R. G. Sierra, D. Starodub, G. J.Williams, J. Hajdu, N. Timneanu, M. M. Seibert, J. Andreasson,A. Rocker, O. Jonsson, M. Svenda, S. Stern, K. Nass, R. An- dritschke, C.-D. Schroter, F. Krasniqi, M. Bott, K. E. Schmidt,X. Wang, I. Grotjohann, J. M. Holton, T. R. M. Barends,R. Neutze, S. Marchesini, R. Fromme, S. Schorb, D. Rupp,M. Adolph, T. Gorkhover, I. Andersson, H. Hirsemann, G. Pot-devin, H. Graafsma, B. Nilsson, and J. C. H. Spence, “Fem-tosecond x-ray protein nanocrystallography,”
Nature , vol. 470,pp. 73–77, 02 2011. D. W. Pashley and M. J. Stowell, “Electron microscopy anddiffraction of twinned structures in evaporated films of gold,”
Philosophical Magazine , vol. 8, no. 94, pp. 1605–1632, 1963. J. Neave, B. Joyce, P. Dobson, and N. Norton, “Dynamics offilm growth of gaas by mbe from rheed observations,”
AppliedPhysics A , vol. 31, no. 1, pp. 1–8, 1983. H. E. ElsayedAli and J. W. Herman, “Picosecond time-resolvedsurface-lattice temperature probe,”