thelausannecentrefor ultrafastscience(lacus)the lausanne centre for ultrafast science (lacus) chimia...

The Lausanne CenTre for uLTrafasT sCienCe (LaCus) CHIMIA 2017, 71, No. 5 A265

The Lausanne Centre forUltrafast Science (LACUS)

Fundamental and practical challenges facing our society can be addressed by new methods and thusapproached from a new perspective. Examples of present day challenges are energy conversion, biology,medicine, new materials, etc.

Over the last decade, ultrafast technology has made enormous progress, opening a large variety of newresearch fields and applications. Examples include table-top high-harmonic generation sources of vacuumultraviolet to soft X-ray radiation that allow new forms of spectroscopy and diffraction, lab-based sources ofultrashort electron pulses and new sources of intense terahertz radiation, etc. All of these tools have openednew directions in materials science, chemistry, and biology.

The EPFL has a large community of ultrafast scientists, most of them participating in the NCCR:MUST(Molecular Ultrafast Science and Technology) and it includes both experimental and theoretical teams fromChemistry, Physics, Life Sciences, and Engineering, developing new methods as well as using them in variousapplications. The complementarity of these various groups in terms of methods, technology and science andthe availability of world-class facilities have led to the creation of the Lausanne Centre for Ultrafast Science, orLACUS (which means “lake” in Latin). It was launched on June 3, 2016 at the Rolex centre of the EPFL, withthe participation of distinguished international speakers such as Theodor Hänsch (München), Shaul Mukamel(UC-Irvine), and Villy Sundström (Lund).

The research areas covered by LACUS are very diverse, spanning from fundamental to applied research.In addition, several EPFL groups have been pioneers in ultrafast science as witnessed by the number of ‘first’achievements. LACUSpools in the expertise in the development and the use of advancedultrafast laser, X-ray andelectron technology, and associated methods, along with the EPFL theory groups. It also aims at complementingand strengthening existing Swiss scientific infrastructures. The quickly evolving field of X-ray science with theadvent of free-electron lasers (FEL) calls for the development of new ultrafast laser techniques. LACUS offers anideal platform for such developments, while complementing SwissFEL by covering lower photon energies andofferingmore flexibility in terms of pulse duration, repetition, flux, as well as allowing complementary experimentsusing ultrashort pulses of Electrons.

Oneof themissionsof LACUS is to enable cutting-edgeultrafast scienceand technologyand their applicationsby means of world class experimental set-ups that are out of reach of the resources of a single group. LACUSprovides to the EPFL community frontier electron and light sources of various pulse duration, going from theTHz range to the vacuum ultraviolet and the soft X-ray range. It also envisions unusual combinations of tools:e.g. THz pump/XUV probe or IR pump/XUV, photons with electron pulses, photons and scanning tunnellingmicroscopy (STM), etc.

The present edition of CHIMIA samples a selection of articles showing the capabilities of LACUS such as:the new Harmonium facility for ultrafast photoelectron spectroscopy of gases, liquid solutions and solids (time-and angle-resolved photoelectron spectroscopy or tr-ARPES), the swissFEL facility, the deep-UV spectroscopycapabilities at the LOUVRE lab, the sum-frequency methods to probe interfacial structure and dynamics, thefemtosecond laser machining of materials and theoretical methods. These papers exemplify the capabilities ofthe methods by recent results of studies on molecules, solid materials and proteins. This selection of articles isfar from exhaustive as LACUS includes several other groups, such as, unique in Switzerland, those developingand using ultrafast electron diffraction and microscopy (F. Carbone and U. Lorenz), in addition to the groupsusing pulsed Terahertz and optical domain radiation for probing the charge carrier dynamics in solar materials(J. E. Moser and A. Hagfeldt), ARPES (H. Dil), frequency combs (T. Kippenberg) and theory (U. Röthlisberger).

In summary, LACUS represents a truly interdisciplinary centre dedicated to the development of newmethodsof ultrafast science methods and techniques and their applications in chemistry, photonics, materials science,and engineering. It also bridges the lab-based activities in X-ray science with those at the Paul-Scherrer-Institutusing the Swiss Light Source synchrotron or soon, the swissFEL free electron laser. Finally, LACUS is open tothe Swiss ultrafast science community within the NCCR: MUST and beyond.

Professor Majed CherguiLaboratoire de Spectroscopie Ultrarapide (LSU) http://lsu.epfl.ch/ andLausanne Centre for Ultrafast Science (LACUS) http://lacus.epfl.ch/Ecole Polytechnique Fédérale de LausanneISIC, FSB, Station 6CH-1015 Lausanne, SwitzerlandE-mail: [email protected]

The Editorial Board of CHIMIA expresses its deep gratitude to Prof. Majed Chergui for organizing this fascinatingissue on LACUS, introducing a pioneering facility at EPFL, enabling cutting-edge, interdisciplinary science opento the Swiss ultrafast science community.

Majed Chergui

266 CHIMIA 2017, 71, No. 5 Contents Inhalt sommaIre

Editorial

265 M. Chergui

268 Harmonium: An Ultrafast Vacuum Ultraviolet FacilityC. A. Arrell*, J. Ojeda, L. Longetti, A. Crepaldi, S. Roth, G. Gatti, A. Clark, F. van Mourik,M. Drabbels, M. Grioni, M. Chergui

273 Time-resolved ARPES at LACUS: Band Structure and Ultrafast Electron Dynamics of SolidsA. Crepaldi, S. Roth, G. Gatti, C. A. Arrell, J. Ojeda, F. van Mourik, P. Bugnon, A. Magrez,H. Berger, M. Chergui, M. Grioni*

278 Aqueous Nanoscale SystemsS. Roke

283 Several Semiclassical Approaches to Time-resolved SpectroscopyJ. Vanícek

288 The LOUVRE Laboratory: State-of-the-Art Ultrafast Ultraviolet Spectroscopiesfor Molecular and Materials ScienceM. Oppermann, N. S. Nagornova, A. Oriana, E. Baldini, L. Mewes, B. Bauer, T. Palmieri, T. Rossi,F. van Mourik, M. Chergui*

295 Ultrafast Laser to Tailor Material Properties: An Enabling Tool in Advanced Three-dimensional MicromanufacturingY. Bellouard

299 Opportunities for Chemistry at the SwissFEL X-ray Free Electron LaserC. J. Milne*, P. Beaud, Y. Deng, C. Erny, R. Follath, U. Flechsig, C. P. Hauri, G. Ingold,P. Juranic, G. Knopp, H. Lemke, B. Pedrini, P. Radi, L. Patthey

308 Corrigendum: CHIMIA 2017, 71, 92–102Mechanistic Insights into Gold Organometallic Compounds and their Biomedical ApplicationsS. Jürgens, A. Casini*

The Lausanne Centre forUltrafast Science (LACUS)

Contents Inhalt sommaIre CHIMIA 2017, 71, No. 5 267

Columns, ConfErEnCE rEports

309 swIss sCIenCe ConCentrates

C. D. Bösch, M. Kownacki, Y. Vyborna, S. M. Langenegger, R. Häner*

310 hIghlIghts of analytICal sCIenCes In swItzerland

How Do Plants Know when to Let Go?S. Augustin, J. Santiago*

311 Polymer and ColloId hIghlIghts

Nanofiber-based AerogelsT. Burger, F. Deuber, M. Merk, S. Mousavi, L. Vejsadová, C. Adlhart*

312 BIoteChnet swItzerland

Conference Report on TEDD & 3R Workshop at ZHAW WaedenswilE. Heinzelmann

315 ConferenCe rePort

Report on 17th International Symposium on Solubility Phenomena and RelatedEquilibrium Processes (ISSP17)M. Filella*, W. Hummel*

316 Swiss Chemical Society Spring Meeting, University of Bern, 21st April 2017J. P. Byrne, P. Melle, M. Valencia, Á. Vivancos*

Community nEws

318 Swiss Chemical Society News

321 Honors and Awards

322 Industrial News

EvEnts

324 SCS Conferences and Symposia 2017

325 Non-SCS Conferences in Switzerland

326 Lectures

327 Education Courses in Analytical Chemistry

CHimia rEport/Company nEws

331 Markt, Apparate, Chemikalien, Firmenportraits und Dienstleistungen

268 CHIMIA 2017, 71, No. 5 The Lausanne CenTre for uLTrafasT sCienCe (LaCus)doi:10.2533/chimia.2017.268 Chimia 71 (2017) 268–272 © Swiss Chemical Society

*Correspondence: Dr. C. A. Arrella

E-mail: [email protected] of Ultrafast Spectroscopy and LausanneCentre for Ultrafast Science (LACUS)ISIC Station 6Ecole Polytechnique Federale de Lausanne (EPFL)CH-1015 LausannebLaboratory for Electron Spectroscopy and LausanneCentre for Ultrafast Science (LACUS)ICMP, Station 3EPFL, CH-1015 LausannecLaboratory for Molecular Nanodynamics andLausanne Centre for Ultrafast Science (LACUS)ISIC, Station 6EPFL, CH-1015 Lausanne

Harmonium: An Ultrafast VacuumUltraviolet Facility

Christopher A. Arrell*a, Jose Ojedaa, Luca Longettia, Alberto Crepaldib, Silvan Rothb, GianmarcoGattib, Andrew Clarkc, Frank van Mourika, Marcel Drabbelsc, Marco Grionib, and Majed Cherguia

Abstract: Harmonium is a vacuum ultraviolet (VUV) photon source built within the Lausanne Centre for UltrafastScience (LACUS). Utilising high harmonic generation, photons from 20–110 eV are available to conduct steady-state or ultrafast photoelectron and photoion spectroscopies (PES and PIS). A pulse preserving monochromatorprovides either high energy resolution (70 meV) or high temporal resolution (40 fs). Three endstations have beencommissioned for: a) PES of liquids; b) angular resolved PES (ARPES) of solids and; c) coincidence PES and PISof gas phase molecules or clusters. The source has several key advantages: high repetition rate (up to 15 kHz)and high photon flux (1011 photons per second at 38 eV). The capabilities of the facility complement the Swissultrafast and X-ray community (SwissFEL, SLS, NCCRMUST, etc.) helping to maintain Switzerland’s leading rolein ultrafast science in the world.

Keywords: Photoelectron spectroscopy · Photoion spectroscopy · Ultrafast science

1. Introduction

During the last decades the abilities ofultrafast sciencehave increaseddramatical-ly. Early techniques, while still providinga plethora of information to physicists andchemists, are often unable to follow struc-tural dynamics as they remain sensitive tovalence electrons which are usually delo-calized over the molecule. Ultrafast tech-niques using higher photon energies havebeen developed using synchrotron sourc-es,[1] free electron lasers[2] and lab-basedlasers.[3] These advanced techniques haveallowed ultrafast structural dynamics to befollowed by accessing inner-shell and corelevel electrons, but are limited by either:temporal resolution, beam-time access orchoice of photon energies (and resolution)available. To increase the range of toolsavailable to scientists, the Harmoniumultrafast vacuum ultraviolet facility has

been constructed within the LausanneCentre for Ultrafast Science (LACUS) atEPFL.

The standard technique for ultrafastmeasurements is pump-probe where apumppulse perturbs the system and a probepulse follows the evolution of the systemback to equilibrium. The photon energiesof Harmonium are ideal for photoelectronand photoion spectroscopy (PES and PIS)and as such three endstations have beenconstructed for PES of liquids, and solidscalled ASTRA, and of clusters and heliumdroplets (CD) (Fig. 1). PES has the advan-tage of directly mapping the electron statesof a system and is not blind to ‘dark states’as can be the case for all-optical spectros-copies. Furthermore, with sufficiently highphoton energies, inner-shell and core lev-els are accessible giving access to elementspecific bands and hence chemical markersof a system.

The Harmonium facility is a largesource built over the past five years. It in-corporates three amplified laser systems,a VUV source and three endstations. It isover twenty metres long and involves fourdifferent research groups from the chem-istry and physics institutes of EPFL. Aschematic of the system is given in Fig. 2and a description of the photon source andcapabilities of the endstations is given inthe following sections.

2. Light Source

2.1 Vacuum Ultraviolet ProbeAs Harmonium has multiple users, var-

ious probe configurations are required. Forexample a high temporal resolution (sub-50 fs) is required to follow the dynamics ofmolecular systems, high energy resolution(sub-100 meV) to follow electronic band

Fig. 1. View of theVUV source cham-bers of Harmonium.Behind the photo-grapher are the lasersources, to the farleft of the picture theLiquid endstationand behind the labwall the Solid and CDendstations.

The Lausanne CenTre for uLTrafasT sCienCe (LaCus) CHIMIA 2017, 71, No. 5 269

preferential to have a similarly sized fo-cus to reduce photoelectron contributionsfrom gasmolecules (see section 3.3), whilea larger focal spot is desired for theARPESand CD to reduce space charge effects andto better match the collection volume fortheir respective detectors. A gold-coatedellipsoidalmirror (ZeissAG) is used for theLiquid phase endstation to reduce the ~100µm source size to ~ 25 µm with a 2000mm entrance arm and a 500 mm exit arm(6.0° angle of incidence). For the ASTRAand CD endstations toroidal mirrors areused in a 2f–2f configuration, providinga ~100 µm focus. Given the large endsta-tion footprint, the entrance/exit arm of themirror for ASTRA is 5100 mm (4.5° angleof incidence) and 3900 mm (4.0° angle ofincidence) for the CDmirror. To ensure thefull beam is collected the carbon-coatedmirrors (Pilz Optics) dimensions are 440 ×60 × 50 mm3. Pointing control of the VUVspot to accurately position and optimisethe respective foci is critical. To achievethis the ellipsoidal mirror uses a six-axishexapod positioner (SpaceFab, PI GmbH)while the toroidal mirrors have actuatorson the three principle mirror axes (AlcaTechnology Srl). Precise alignment ofthe VUV focus is achieved in all casesby imaging the VUV focus on a YAG:Cecrystal.

2.2 Optical PumpTo minimise disruption when switch-

ing between endstations and given the dif-ferent temporal/energy resolution require-ments of the optical pump laser systems,two separated laser systems have been in-stalled to provide the optical pump.

For the Liquid station where typical-ly higher temporal resolution is required(<50 fs) and a pump in the UV to visiblerange is desirable, a booster stage has beenconstructed for the principal laser systemdriving the HHG. A fraction (~10%) ofthe uncompressed 6 kHz output of the re-generative amplifier is subsequently dou-ble passed in a home-built cryogenicallycooled Ti-Sapphire multi-pass amplifier toprovide 12 W at 6 kHz to pump an opticalparametric amplifier (OPA) dedicated tothe Liquid endstation.

A separate regenerative amplifier(Coherent Astrella) producing 6 W at 6kHz, pumps an OPA providing the opticalpump (200 nm – 12 µm) for eitherASTRAor the CD endstation. This oscillator of theAstrella is electronically synchronisedto the oscillator seeding the principal re-generative amplifier of the system. Thereis typically a sub-100 fs jitter in-betweenthe two outputs, which is acceptable forthe trARPRES and CD measurements.Furthermore provision is provided to pumpwith the intrinsically synchronised pumparm for the Liquid endstation if required.

the monochromator is shown in Fig. 4 anda full report of the system is given in ref.[6]. For photon energies up to 40 eV, argonwas used as the generation medium in theHHG process and neon was used for ener-gies up to 110 eV, typically providing 1 ×1011 photons/second (measured at Liquidendstation target) at 30–40 eV and 2 × 108

photons/second (measured at Liquid end-station target) between 50–100 eV.

The optics of the monochromator im-ages the effective entrance slit of the in-strument to the exit slit in a 1:1 configura-tion. As it is not feasible to use an entranceslit as this would be close to the fundamen-tal IR focus, the VUV source spot is usedwhich is imaged to a variable exit slit ofthe monochromater. Typically the VUVsource diameter is 100 µm FWHM, but ishowever highly dependent upon the focalparameters of the fundamental laser andphase matching. The control of the VUVsource spot size including the divergenceof the VUV gives further control over thetemporal response of the system and willbe the subject of a future publication.

The VUV focal requirements differfor the Liquid endstation as compared tothose required forASTRA and the CD end-stations. For the former, due to the smalldiameter (~25 µm) of the liquid jet it is

dispersion, a high photon (100 eV) energyto access Auger states or a photon energybelow 21 eV to avoid ionisation from he-lium.

By using an intense laser pulse to tun-nel ionise a target (typically a noble gas)high harmonic generation (HHG) produc-es a spectrum extending into the VUV.HHG intrinsically yields a spectrum ofharmonics (Fig. 3) extending over many10’s of eV with a short pulse structure,[4]which requires subsequent monochroma-tisation for PES.

To maintain flexibility, a grating-basedmonochromator design was chosen in col-laboration with CNR-IFN in Padova. Thedesign, with the grooves running parallelto the VUV propagation direction, limitsthe temporal broadening introduced by apath length difference from the grating. Inso doing and by using a collimated VUVbeam, the number of grooves illuminatedis controlled and hence also the temporalbroadening of the pulse. This technique isreviewed in ref. [5].

To provide either high energy (~70meV) or high temporal resolution (~40fs) over the full energy range accessibleby Harmonium (30–110 eV) one of fourdifferent gratings with different groovedensities can be used. The performance of

Fig. 2. Schematic of the Harmonium facility. The laser systems are shown in red, the purple linerepresents the VUV source and monochromatisation, the orange line the Liquid endstation, thegreen the Solid endstation and the blue the CD endstation. The beam path of the lasers is omittedfor clarity.

Fig. 3. A spectrumof VUV produced byHarmonium.The HHG yields aspectrum that re-quired monochroma-tisation.

270 CHIMIA 2017, 71, No. 5 The Lausanne CenTre for uLTrafasT sCienCe (LaCus)

key to understanding biological processesand Harmonium is an ideal tool to advancethis research.

3.2 Cluster and Droplet EndstationThe evolution of molecular systems

is governed by the potential energy land-scape describing the interactions betweenits constituents and the dynamics takingplace on this potential energy landscape.For isolated molecules, both aspects cannowadays be investigated in great detailand precision. The situation is much morecomplex for molecules in solution, due toenvironmental interactions. Studying theeffect of microsolvation can form a link be-tween isolated systems and solution phasesamples, as it allows for a systematic sep-aration of environmental interactions fromintrinsic sample effects.

An endstation for microsolvation stud-ies based on time-resolved photoelectronspectroscopy (trPES) is currently un-der construction at Harmonium (Fig. 6).Size-controlled clusters or nanodropletscontaining a single solute molecule areformed using a combination of molec-ular beam and pickup techniques. Theenergy redistribution in the system afterphotoexcitation of the solute will be ex-plored using trPES. The VUV radiationprovided by Harmonium allows probingof energy levels all the way down to theground state of the system. The detectionsystem of the setup is based on a doubleimaging photoelectron photoion coinci-dence (i2-PEPICO) spectrometer. Uponphotoionisation of an analyte in a time-re-solved pump-probe process, photoelec-trons and photoions are extracted in oppo-site directions by electric fields in a doublevelocity map imaging (VMI) configurationand detected in coincidence on two-dimen-sional spatial detectors with a high timeresolution. The i2-PEPICO technique[17–19]allows for an extremely detailed, multi-

inner shell to study both the electronic andstructure dynamics following charge trans-fer by measuring both the valence and in-ner-shell dynamics.

Recent work at the Liquid endstationhas focused on several areas. As LPES isa relatively new technique,[12] methods fortrLPES are still under development.[9,13–15]Importantly to aid the progress we haverecently identified and characterized thelaser-assisted photoelectric effect (LAPE)from liquids, a strong field process oc-curring within the cross-correlation ofthe pump and probe fields.[16] Followingthis development, we have been able toutilise the high temporal resolution to fol-low the sub-100 fs reduction dynamics of[Fe(CN)

6]3– following a photoinduced li-

gand-to-metal charge transfer (LMCT).An important feature of Harmonium

that contrasts with higher photon energysources (e.g. synchrotrons and free electronlasers) is that LPES at Harmonium is sensi-tive to the surface of the liquid jet. Owingto the photon energy range of Harmonium(20–110 eV) the inelastic mean free pathof the photoelectron gives sensitivity to thesurface or near surface in contrast to thebulk sensitivity of UPS and XPS. The sig-nificance is clear, interfacial chemistry is

3. Endstations

3.1 Liquid EndstationFormany photochemistry processes the

solvent has a large effect on the energeticsand temporal outcomes of photoexcitation.Therefore, to fully apply time-resolvedtechniques to chemistry, these techniquesneed to be implemented in solution. Forthese reasons we have developed an end-station at Harmonium to conduct time-re-solved liquid photoelectron spectroscopy(trLPES).

LPES was first introduced by Siegbahnand co-workers in 1973,[7] but it was notuntil the development of the liquid micro-jet technology by Faubel and co-workersin 1988[8] that the technique became morewidely implemented. The Faubel designuses a small glass capillary (20 µm indiameter) to inject a continuous flow ofliquid in vacuum. When the velocity issufficiently high (30 ms–1) a region of lam-inar flow is formed extending for severalmillimetres. This region provides a liquid–vacuum interface where photoelectronscan be directly collected from the liquid.This technique combined with cryo-pump-ing and a specialist electron spectrometerdesign[9] allows photoelectron spectra tobe collected from a range of different sol-vents and respective solutes. The LPESspectrum of NaI in an aqueous solution isshown in Fig. 5, along with a pure waterspectrum. Due to electrokinetic charging itis not possible to measure a PES from purewater and a small concentration of salt isadded, here 25 mM NaCl, see ref. [10] formore details. This data, taken at 83 eV,exploits the high photon energy availablefrom Harmonium to access the inner-shellI 4d shell, a first for a VUV source basedon HHG. The spectra contain contributionsfrom both the liquid jet and surroundingevaporated gas. The relative binding ener-gies of gas and liquid phase are separatedby ~1.5 eV allowing the gas contributionsto be subtracted if required. Examples ofthis can be found in ref. [11].

Work is currently underway at Har-monium to utilize this ability to access the

Fig. 4. Gratingresponse ofthe Harmoniummonochromator.Depending on theselected grating (andhence groove density)either high temporalresolution or highenergy resolution isprovided. Figure fromref. [6].

Fig. 5. Photoelectron spectra of water and an aqueous solution of NaI. Measured with photonenergy of 83 eV.


spectrometer providing 0.05 Å–1 momen-tum resolution.

Fig. 7 summarizes a time-resolvedARPES (trARPES) study of Bi

2Se

3, a lay-

ered compound from the family of topo-logical insulators.[31,32] The stacking of thelayers is visualized in Fig. 7a. These ma-terials are promising candidates for futurenanotechnology applications such as quan-tum computing,[33] spintronics,[34,35] highperformance field effect transistors[36] andflexible electrodes in solar cells.[37] Theseeffects rely on the excitation of electronsfrom the ground state into unoccupied partof the materials bandstructure.

Therefore, Harmoniumwith itsARPESendstation is ideal to follow the dynamicsof such excitation. The curve shown inFig. 7 shows the population dynamics ofthe electronic states above the Fermi level(E

F) following an optical pump (800 nm).

The ARPES spectrum shows the conduc-tion band minimum slightly below theE

F, while the top of the valance band is at

–0.4 eV. In between the two bands is thetopologically protected surface state. Thefirst spectrum shows the bandstructure atthe point of the highest excitation. Withincreasing time between the pump and theprobe pulse, relaxation processes lead to acontinuous relaxation of the states popu-lated above E

F. The second and the third

image show the bandstructure 4 ps and8 ps after the maximum excitation. Withthe addition of the dedicated optical pump-ing arm for the ASTRA and CD endsta-tions providing a pump between 200 nm –12 μm inter-band transitions can be excitedresonantly and studied.

4. Conclusions

We believe the Harmonium facilityrepresents a significant benefit to the ul-trafast community within Switzerland.The range of endstations covers samplesfrom gas phase molecules through to clus-ters, liquids and solids providing a levelof functionality more commonly foundat a national level such as Artemis in theUK.[38] The ASTRA endstation also hasthe capability to probe molecules on sur-faces. Through our associate membershipof LaserLab and collaborative networkswithin Switzerland (e.g. NCCR MUST)the source is also available for externalusers, either to use the current endstationsor to bring their own experiment. With thecurrent opening of SwissFEL, Harmoniumis operational at an opportune moment toact as a testing ground and complementa-ry source to the larger ultrafast facilitieswhere beamtime access is limited.

Received: April 25, 2017

lecting cluster/droplet sizes will allow aclear picture of solvent-induced dynamicsto be established.

3.3 Solid Phase EndstationA detailed description of the solid

phase endstation can be found in the ar-ticle ‘Time-resolved ARPES at LACUS:Band Structure and Ultrafast ElectronDynamics of Solids’ in the same issueof this journal,[29] a summary is present-ed in the following. Angular ResolvedPhotoelectron Spectroscopy (ARPES) is apowerful tool to investigate the electronicstructure of a solid.[30] Making use of thephotoelectric effect, electrons are emit-ted from the illuminated sample surfaceinto vacuum. The conservation of ener-gy and momentum parallel to the surfaceduring the emission process allows us totrace back the electron to its initial stateinside the solid. Combining the ASTRAARPES endstation with the capabilities ofHarmonium to probe ultrafast phenomenaprovides new insight into the electronicbandstructure, which are not possible withconventional photoelectron spectroscopy.By means of pump-probe techniques, elec-trons can be excited (pumped) from belowthe Fermi level into the unoccupied part ofband-structure, and their subsequent relax-ation followed by aVUV probe. Other thantwo-photon photoemission experimentswhere the probe consists of two times thefrequency doubled or tripled fundamental(2 × 3.1 eV or 2 × 4.6 eV), HHG photonsup to several tens of eV, allow us to probelarger areas of the k-space e.g. the entireBrillouin zone of a material. Typically forsuch processes a high energy resolution(below 100 meV) and a good momentumresolution (better than 0.1 Å-1) is required.Harmonium is optimised for these require-ments with the variable grating configura-tion of the monochromator providing ~75meV energy resolution and the focusingparameters combined with the ARPES

plexed view into excited state dynamicalprocesses. Photoinduced fragmentationscan be readily identified by time-of-flightmeasurements and, by recording electronsand ion fragments in coincidence, photo-electron spectra for each photofragmentcan be recorded. Coincidence detection al-so gives angular distributions of both pho-toelectrons and photofragments, allowingnear-laboratory frame photoelectron angu-lar distributions to be recorded,[20] whichserve as an extremely effective probe ofthe symmetry of the electronic states be-ing ionized.[21] Utilizing each of these var-ious sources of information in 2-PEPICOmakes it possible to disentangle differentcontributions to the dynamics, a necessitywhen considering systems with multiplerelaxation channels.[22]

A prototypical example for such stud-ies is theDNAbase adenine. Experimentaland theoretical studies on gas-phase ad-enine have identified several efficientnon-radiative decay processes proceedingvia conical intersections[23–25] that allowfor a fast conversion of the electronic en-ergy to heat. It is hypothesized that thisinfluences adenine’s biological role asa DNA base by serving as a protectivemechanism for DNA from solar UV light.However, there is still debate whether thesame behaviour is exhibited once adenineis hydrogen-bonded and base-stacked insolution with DNA.[26,27] Previous studiesshowed a marked decrease in the 1L

b(ππ*)

excited electronic state lifetime inducedby a superfluid helium solvent.[28] Thiswas tentatively attributed to a lowering ofthe potential energy barrier blocking ac-cess to a conical intersection. The degreeof change is significant given that super-fluid helium is ostensibly an extremelynon-perturbative solvent. Aqueous envi-ronments would induce markedly largerchanges, so much so that individual con-tributions are prone to be washed out inbulk phase. Controlling solvation by se-

Fig. 6. The ASTRAendstation ofHarmonium oftrARPES, beamsenter the chamberfrom the left.


[1] C. J. Milne, T. J. Penfold, M. Chergui, Coord.Chem. Rev. 2014, 277, 44.

[2] J. Feldhaus, J. Phys. B At. Mol. Opt. 2010, 43,194002.

[3] M. Drescher, M. Hentschel, R. Kienberger,M. Uiberacker, V. Yakovlev, A. Scrinzi,T. Westerwalbesloh, U. Kleineberg, U.Heinzmann, F. Krausz, Nature 2002, 419, 803.

[4] F. Frank, C. Arrell, T. Witting, W. A. Okell, J.McKenna, J. S. Robinson, C. A. Haworth, D.Austin, H. Teng, I. A.Walmsley, J. P. Marangos,J. W. G. Tisch, Rev. Sci. Instrum. 2012, 83,071101.

[5] F. Frassetto, N. Fabris, P. Miotti, L. Poletto,Photonics 2017, 4, 14.

[6] J. Ojeda, C. A. Arrell, J. Grilj, F. Frassetto, L.Mewes, H. Zhang, F. van Mourik, L. Poletto, M.Chergui, Struct. Dyn. 2016, 3, 023602.

[7] H. Siegbahn, K. Siegbahn, J. Electron.Spectrosc. 1973, 2, 319.

[8] M. Faubel, S. Schlemmer, J. P. Toennies, Z.Phys. D Atom. Mol. Cl. 1988, 10, 269.

[9] C. A. Arrell, J. Ojeda, M. Sabbar, W. A. Okell,T. Witting, T. Siegel, Z. Diveki, S. Hutchinson,L. Gallmann, U. Keller, F. van Mourik, R.T. Chapman, C. Cacho, N. Rodrigues, I. C.E. Turcu, J. W. G. Tisch, E. Springate, J. P.Marangos, M. Chergui, Rev. Sci. Instrum. 2014,85, 103117.

[10] N. Preissler, F. Buchner, T. Schultz, A. Lubcke,J. Phys. Chem. B 2013, 117, 2422.

[11] B.Winter, R.Weber,W.Widdra, M. Dittmar, M.Faubel, I. V. Hertel, J. Phys. Chem. A 2004, 108,2625.

[12] B. Winter, M. Faubel, Chem. Rev. 2006, 106,1176.

[13] I. Jordan, M. Huppert, M. A. Brown, J. A. van

Bokhoven, H. J. Worner, Rev. Sci. Instrum.2015, 86, 123905.

[14] M. Borgwardt, M.Wilke, I.Y. Kiyan, E. F.Aziz,Phys. Chem. Chem. Phys. 2016, 18, 28893.

[15] J. Nishitani, C. W. West, T. Suzuki, Struct. Dyn.2017, 4, 044014.

[16] C. A. Arrell, J. Ojeda, L. Mewes, J. Grilj,F. Frassetto, L. Poletto, F. van Mourik, M.Chergui, Phys. Rev. Lett. 2016, 117, 143001.

[17] A. Vredenborg, W. G. Roeterdink, M. H. M.Janssen, Rev. Sci. Instrum. 2008, 79, 063108.

[18] A. Bodi, P. Hemberger, T. Gerber, B. Sztaray,Rev. Sci. Instrum. 2012, 83, 083105.

[19] G. A. Garcia, B. K. C. de Miranda, M. Tia, S.Daly, L. Nahon, Rev. Sci. Instrum. 2013, 84,053112.

[20] O. Gessner, A. M. D. Lee, J. P. Shaffer, H.Reisler, S. V. Levchenko, A. I. Krylov, J. G.Underwood, H. Shi, A. L. L. East, D. M.Wardlaw, E. T. H. Chrysostom, C. C. Hayden,A. Stolow, Science 2006, 311, 219.

[21] K. L. Reid, Ann. Rev. Phys. Chem. 2003, 54, 397.[22] P. Maierhofer, M. Bainschab, B. Thaler, P.

Heim, W. E. Ernst, M. Koch, J. Phys. Chem. A2016, 120, 6418.

[23] L. Blancafort, J. Am. Chem. Soc. 2006, 128, 210.[24] C. Plutzer, K. Kleinermanns, Phys. Chem.

Chem. Phys. 2002, 4, 4877.[25] C. M. Marian, J. Chem. Phys. 2005, 122,

104314.[26] J. Luo,Y. Liu, S.Yang, J. Photochem. Photobio.

A 2017, 337, 1.[27] N. K. Schwalb, F. Temps, Science 2008, 322,

243.[28] S. Smolarek, A. M. Rijs, W. J. Buma, M.

Drabbels, Phys. Chem. Chem. Phys. 2010, 12,15600.

[29] A. Crepaldi, S. Roth, G. Gatti, C. A. Arrell, J.Ojeda, F. van Mourik, P. Bugnon,A. Magrez, H.

Berger, M. Chergui, M. Grioni, Chimia 2017,71, 273.

[30] S. Hüfner, ‘Photoelectron Spectroscopy’,Springer, 2003.

[31] Y. Xia, D. Qian, D. Hseih, L. Wray, A. Pal, H.Lin, A. Bansil, D. Grauer, Y. S. Hor, R. J. Cava,M. Z. Hasan, Nat. Phys. 2009, 5, 398.

[32] H. J. Zhang, C.-X. Liu, X.-L. Qi, X. Dai, Z.Fang, S.-C. Zhang, Nat. Phys. 2009, 5, 438.

[33] M. Z. Hasan, C. L. Kane, Rev. Mod. Phys. 2010,82, 3045.

[34] X. B. Xiao, S.Y. A.Yang, Z. F. Liu, H. L. Li, G.H. Zhou, Sci. Rep. 2015, 5, 7898.

[35] Y. Cao, J. A. Waugh, X.-W. Zhang, J.-W. Luo,Q.Wang, T. J. Reber, S. K. Mo, Z.Wu,A.Yang,J. Schneeloch, G. D. Gu, M. Brahlek, N. Bansal,S. Oh, A. Zunger, D. S. Dessau, Nat. Phys.2013, 9, 499.

[36] H. Zhu, C. A. Richter, E. Zhao, J. E. Bonevich,W. A. Kimes, H.-J. Jang, H. Yuan, H. Li, A.Arab, O. Kirillov, J. E. Maslar, D. E. Ioannou,Q. Li, Sci. Rep. 2013, 3, 1757.

[37] H. Peng, W. Dang, J. Cao, Y. Chen, D. Wu, W.Zheng, H. Li, Z. X. Shen, Z. Liu, Nat. Chem.2012, 4, 281.

[38] I. C. E. Turcu, E. Springate, C. A. Froud, C.M. Cacho, J. L. Collier, W. A. Bryan, G. R. A.J. Nemeth, J. P. Marangos, J. W. G.Tisch, R.Torres, T. Siegel, L. Brugnera, J. G. Underwood,I. Procino, W. R. Newell, C. Altucci, R. Velotta,R. B. King, J. D. Alexander, C. R. Calvert,O. Kelly, J. B. Greenwood, I. D. Williams, A.Cavalleri, J. C. Petersen, N. Dean, S. S. Dhesi,L. Poletto, P. Villoresi, F. Frassetto, S. Bonora,M. D. Roper, Proc. Spie. 2010, 7469, 746902.

Fig. 7. Time-resolved ARPES study of Bi2Se3measured at Harmonium.

The Lausanne CenTre for uLTrafasT sCienCe (LaCus) CHIMIA 2017, 71, No. 5 273doi:10.2533/chimia.2017.273 Chimia 71 (2017) 273–277 © Swiss Chemical Society

*Correspondence: Prof. Dr. M. Grionia

E-mail: [email protected] of PhysicsEcole Polytechnique Fédérale de Lausanne (EPFL)CH-1015 LausannebLaboratory of Ultrafast Spectroscopy, ISICand Lausanne Centre for Ultrafast Science (LACUS)Ecole Polytechnique Fédérale de Lausanne (EPFL)CH-1015 Lausanne

Time-resolved ARPES at LACUS: BandStructure and Ultrafast Electron Dynamicsof Solids

Alberto Crepaldia, Silvan Rotha, Gianmarco Gattia, Christopher A. Arrellb, José Ojedab, Frank vanMourikb, Philippe Bugnona, Arnaud Magreza, Helmuth Bergera, Majed Cherguib, and Marco Grioni*a

Abstract: The manipulation of the electronic properties of solids by light is an exciting goal, which requiresknowledge of the electronic structure with energy, momentum and temporal resolution. Time- and angle-resolvedphotoemission spectroscopy (tr-ARPES) is the most direct probe of the effects of an optical excitation on theband structure of a material. In particular, tr-ARPES in the extreme ultraviolet (VUV) range gives access to theultrafast dynamics over the entire Brillouin zone. VUV tr-ARPES experiments can now be performed at the ASTRA(ARPES Spectrometer for Time-Resolved Applications) end station of Harmonium, at LACUS. Its capabilities areillustrated by measurements of the ultrafast electronic response of ZrSiTe, a novel topological semimetal char-acterized by linearly dispersing states located at the Brillouin zone boundary.

Keywords: Angle-resolved photo emission spectroscopy (ARPES) · LACUS · ZrSiTe

Angle-resolved photoemission spec-troscopy (ARPES) is a powerful probe ofthe electronic structure of solids.[1] It yieldsthe energy-momentum band dispersion,and provides fundamental microscopicinformation on the electronic correlationsand electron-lattice interactions that shapethe properties of interesting materials.[2]Moreover, in pump-probe experiments itgives access to the electron dynamics in thetime domain.[3] In ARPES, the dispersionof the electronic states E(k) is encoded inthe angular dependence of the kinetic en-ergy E

kinof the photoelectrons, through the

equations that express the conservation ofenergy and momentum:

EB = hν − Ekin − φWF ; (

k =1

h2meEkin sin θ .

(1)

(2)

EBand m

eare the electron binding en-

ergy and mass, hν is the photon energy, φWF

is the material’s work function, and θ theemission angle with respect to the sample’ssurface normal. k|| = (k

x; k

y) is the surface

projection of the electron’s momentumwhich, unlike the third component k

z, is

conserved in the photoemission process.Parallel momentum conservation, and a

large surface sensitivity, makes ARPES anideal technique to study the band structureof surfaces and, more generally, of two-dimensional (2D) materials.[4,5] ARPES ex-periments were for instance instrumental toverify theoretical predictions of linearly dis-

persing electronic states (Dirac cones) at theFermi level (E

F) of graphene. These peculiar

states are solutions of the Dirac equation[6]with the speed of light being replaced by theFermi velocity v

F≈ 0:003 c.[7,8] The linear

crossing of the bands, without the formationof energy gaps, reflects the symmetry of thegraphene honeycomb lattice (Fig. 1(a)),where the rhombic unit cell contains twononequivalent carbon atoms C

Aand C

B.[4]

The according reciprocal space picture isshown in Fig. 1(b). ARPES measurementsreveal the linearly dispersing Dirac conesat the K and K’ corners of the hexagonalBrillouin zone (BZ), as shown in the inset.

reciprocal spacereal space

CA CB

a1

a2

Kk1

k2

Γ

Å

K’

37.5

37.0

36.5

36.0

35.5

1050-5-10angle (deg)

35.0

34.5

kinet

icen

ergy

(eV)

3.02.01.00.0

a) b)

Fig. 1. (a) Real space sketch of graphene. The unit cell (central rhomb) contains two carbon atomsCA, and CB, one from each sublattice. (b) Corresponding reciprocal space picture with the firstBrillouin zone (BZ) in the center. The electronic band structure around EF consists of six doublecones touching at the K and K’ points of the BZ. ARPES measurement of a p-doped graphenesample showing a branch of one of the Dirac cones.


of the HHG pulses was estimated fromthe total experimental broadening of theFermi-Dirac distribution in low tempera-ture experiments to be < 100 meV. Thisenergy broadening corresponds well tothe temporal duration of 240 fs, measuredfor the grating used in this work (900 gr/mm).[49]TheUHVphotoemission chamberis also equipped with a monochromatizedGammadata VUV5000 He electron cyclo-tron resonance (ECR) source, providinghigh-brilliance He Iα (21.2 eV) and HeIIα (40.8 eV) emission lines. This sourceis used in high-energy resolution staticARPES measurements to determine theband structure prior to the tr-ARPES ex-periments. The pump at 800 nm was Hpolarized and the fluence on the samplesurface was ≈ 0.5mJcm–1. High qualitysingle crystals of ZrSiTe were produced bythe chemical vapor transport method andcleaved in situ to expose atomically cleansurfaces.

In the followingwe present preliminaryresults on the novel topological semimetalZrSiTe, which illustrate the capabilities ofASTRA for studies of the equilibrium andthe ultrafast electron dynamics of solids.The crystal structure of ZrSiTe (Fig. 2(a))belongs to the tetragonal P4/nmm spacegroup,[52] with lattice parameters a = 3.7Å and c = 9.5 Å.[54] It results from the or-dered stacking of Te (brown), Zr (green)and Si (blue) planes in quintuple layers.The weak van der Waals interaction be-tween two adjacent Te layers determinesthe most favorable cleavage plane. Fig.2(b) shows the corresponding 3D BZ withthe high-symmetry points, and the SBZ forthe (001) cleavage surface.

ZrSiTe and its sister compounds ZrSiSand ZrSiSe have recently attracted consid-erable attention as members of the noveltopological family of nodal line Diracsemimetals (NSMs).[50–52,55,56] The hall-mark of these materials is schematicallyillustrated in Fig. 2(c). In 2D and 3D Diracsemimetals, such as graphene and Cd

3As

2,

the linear dispersion gives rise to point-like Fermi surfaces, in correspondence ofthe Dirac points, DP (top). By contrast, inNSMs the linearly dispersing states crossnot only at isolated points, but along a con-tinuous line in reciprocal space. Hence, fora suitable doping level, the Fermi surfaceis formed by a closed line resulting fromthe crossing of the inverted valence (green)and conduction (yellow) bands (bottom).The presence of a nodal line of Diracpoints has important consequences on thetransport properties, such as the emergenceof a large non-saturating magnetoresis-tance.[51,55,56]

Fig. 3 presents an overview of the bandstructure of ZrSiTe obtained by means ofstatic ARPES. Fig. 3(a) shows the Fermisurface measured with the He I emission

ics of quasi-free standing single-layer[38,39]and bilayer[40] graphene could only be ex-plored using high harmonics generation(HHG) in a gas. Those studies have clari-fied the complex relaxation mechanismsincluding scattering with electrons, withhot phonons[41] and with lattice defects.[38]They also showed that multiple hot elec-trons can be generated in the unoccupiedpart of the Dirac cone as a result of chargecarrier multiplication, an effect of great in-terest in the perspective of an efficient con-version of light into electrical current.[42]

The capability of tr-ARPES in the ex-treme ultraviolet (VUV) photon energyrange, to map the out-of-equilibrium elec-tron dynamics over the entire BZ is ofpivotal importance for many interestingsystems, such as the transition metal di-chalcogenides[43,44] and perovskites.[45–47]It is also crucial to resolve the antinodalexcitations in the cuprate high-tempera-ture superconductors.[48] However, owingto the technical difficulty of achievinghigh flux and stable harmonic generation,only a few VUV tr-ARPES instrumentsare operational worldwide. In this articlewe present the results of the first ARPESand tr-ARPES experiments carried outat ASTRA (ARPES Spectrometer forTime-Resolved Applications) end stationof Harmonium,[49] at the Lausanne centrefor ultrafast science (LACUS), EPFL. Weshow that the large momentum range madeaccessible by the HHG source allows us todetermine the dispersion of the spin-orbitsplit Dirac particles at the zone boundaryof the novel nodal line Dirac semimetalZrSiTe.[50–52] We also show preliminarydata on the ultrafast out-of-equilibriumdynamics in this material.

A detailed description of the lasersource and of the HHG setup can be foundin ref. [49] and in the accompanying ar-ticle ‘Harmonium: An Ultrafast VacuumUltraviolet Facility’[53] in this issue. In thefollowing we provide more technical in-formation about the ASTRA end station.In the present study we have employedthe 17th to 21st harmonics generated in ar-gon, covering the 26–33 eV photon energyrange. The linear polarization of the HHGcan be switched between vertical (V) andhorizontal (H), which allows us to take ad-vantage of polarization-dependent matrixelements in the photoemission process, asshown later. The HHG probe beam is fo-cused on the surface of the sample, placedin an ultrahigh vacuum (UHV) environ-ment, to a spot size smaller than 100 x100 µm2 . Photoelectrons are collected andanalyzed by a Specs Phoibos 150 electro-static hemispherical analyzer, with an an-gular resolution ∆θ = 0.3° and an energyresolution that can be varied between ∆E= 10 meV for static ARPES and 120 meVfor tr-ARPES experiments. The bandwidth

The concept of Dirac fermions, as lowenergy quasiparticle excitations in sol-ids, extends far beyond graphene. Spin-polarized Dirac particles have been ob-served at the surface of topological insula-tors (TIs).[9]Three-dimensional (3D) Diracparticles have been discovered in bulkcrystals, where the linear dispersion in allthree momentum directions is topological-ly protected by the presence of rotationalsymmetries.[10,11] The interest in these nov-el states of quantum matter is not purelyacademic. It also stems from the prospectof achieving defect-tolerant spin currentsin the former,[12] and ultrahigh mobilitiesin the latter.[11] Besides the transport prop-erties, the linear dispersion of the 2D and3D Dirac particles accounts also for someunconventional optical properties.[13,14]Light-harvesting[15] and optically inducedspin-[16] and valley-polarized[17,18] excita-tions are amongst the perspectives openedby the capability of manipulating the bandstructure by an optical excitation.

ARPES has recently experienced afast development towards experimentsenabling simultaneous energy, momen-tum and temporal resolution. The goal isto track transient changes of the electronicpopulation induced by an optical excita-tion. Time-resolved ARPES (tr-ARPES)requires pulsed laser sources providingultrashort pulses with duration ≈ 100 fs(1 fs = 10–15 s). The temporal evolution ofthe band structure is mapped in a strobo-scopic pump-probe experimental scheme,by varying the difference in optical pathbetween the two pulses. The optical ex-citation (pump) often exploits the funda-mental wavelength of the laser, whereasthe photon energy of the probe pulse mustbe larger than the work function φ

WF. The

possibility of optically generating spin cur-rents in the surface Dirac particle of TIshas also motivated tr-ARPES studies,[19–33]also with spin resolution[34–37] aimed at un-derstanding the out-of-equilibrium scatter-ing mechanisms.

The momentum window of an ARPESmeasurement is determined by the photonenergy. According to Eqn. (2) the largestaccessible electron wave vector k

max(for E

B= 0; θ = 90°) is proportional to ℎ𝑣𝑣 − 𝜙𝜙�� .For low photon energies (5.9–6.3 eV) cor-responding to the 4th harmonic generated innon-linear crystals, and for typical valuesφWF

4–5 eV, this yields kmax

< 0.6 Å–1. Yetsmaller values are obtained for more realis-tic emission angles (θ

max≈ 60°). In TIs, the

Dirac cone is located at the center of thesurface-projected Brillouin zone (SBZ),and the relatively small Fermi wavevector(k

F≈ 0.2° Å–1) is well within this window.

By contrast, the Dirac cones of graphene(Fig. 1) are centered at k

F= 1.7 Å–1, which

requires photon energies larger than hν ≥16 eV. Therefore the hot carrier dynam-


the SBZ boundary. Notice that the intensi-ties at symmetric points on opposite sidesof Χ in the first and second SBZs are notidentical, even if these points are formallyequivalent, because the ARPES signal ismodulated by k-dependent matrix element.Fig. 3(b–d) show the corresponding banddispersion along Μ − Χ − Μ (light blueline in Fig. 3(a)), measured with the He Iline (a) and with HHG photons at 33 eV(c,d), respectively. Interestingly, the datareveal that also this band disperses linearlyover a wide energy range of almost 2 eV.Switching the light polarization from hori-

loids of Fig. 2(c) the two contours are notsimply concentric circles. Nevertheless ourobservations are in good agreement with arecent investigation of the band structureof ZrSiTe,[52] where these states were alsoattributed to the nodal line Dirac particles.

The two Fermi surface contours do notmerge along the four equivalent Γ − Χ di-rections, where hybridization opens smallenergy gaps. Their touching points corre-spond to the region of high intensity, at (k

x= 0.15 Å–1, k

y= 0.55 Å–1). An additional

state originating from this point, and indi-cated by a blue arrow, disperses towards

line ΓΧ = 0.85 Å–1. We identify severalfeatures, in particular two states, indicatedby red arrows, forming diamond-like con-tours centered at Γ. The origin of thesetwo states can be understood by consider-ing the cartoon at the bottom of Fig. 2(c).Owing to a slight doping E

Fintersects the

band structure far from the nodal line (red),which breaks up into two distinct contours(blue dashed lines). The inner one derivesfrom the conduction band and the outerone from the valence band. Since the realbands are more complex than the parabo-

(a) (b)

Z R

AX

M

X

M

Quintuplelayer

TeZr

SiZr

Te

DP

(c)

Nodalline

kx

kyE

kx

kyE

Fig. 2. (a) Crystal structure of ZrSiTe. The unit cell consists in the ordered stacking of Te (brown), Zr (green) and Si (blue) planes in quintuple layers.(b) Bulk Brillouin zone and the projected Brillouin zone on the (001) surface (SBZ), subject of our study. The high symmetry points are indicated, aswell. (c) Schematics of the band dispersion in a 2D or 3D Dirac semimetal (top) and in a nodal line Dirac semimetal (bottom). In the former the DiracFermions form a point-like Fermi surface (DP) while in the latter the linearly dispersing states form a close-contour nodal line.

-0.4 -0.2 0 0.2 0.4k (A )x

°-1-0.4 -0.2 0 0.2 0.4

-2.0

-1.5

-1.0

-0.5

0

0.5

-0.4 -0.2 0 0.2 0.4

1.0

-0.5

0

0.5

k (A )x°-1

-0.4 -0.2 0 0.2 0.4k (A )x

°-1k (A )x

°-1

k(A

)y

°-1

E-

E(e

V)

F

a) b) c) d)hv = 21.2eV He Iαhv = 21.2eV He Iα hv = 33eV H - pol hv = 33eV V - pol

Fig. 3. (a) Fermi surface of ZrSiTe, as obtained by static ARPES mapping with the He Iα emission line (hν = 21.2 eV. The high-symmetry points Γ, Χof the SBZ are shown, along with the directions where the data of Fig. 3 (blue line) and Fig. 4 (green dashed line) are taken. A dashed black circleindicates the maximum momentum window accessible at low photon energy (5.9 eV – 6.2 eV) with a realistic maximum emission angle θmax = 60°.(b–d) ARPES image of the band dispersion along the Μ − Χ − Μ high-symmetry directions acquired with He Iα emission line (a), and with HHGphotons at hν = 33 eV, with horizontal (c) and vertical polarization (d).


surveys of the equilibrium band structurecan be performed with a high-brilliancemonochromatized He source. They aresupplemented by measurements that ex-ploit the energy- and polarization tunabili-ty of the HHG laser-based source.We havealso shown that the large-energy photonsgenerated by Harmonium enable mappingthe out-of-equilibrium dynamics in mo-mentum regions precluded to conventionallow photon energy tr-ARPES. Further in-vestigations of the relaxation dynamics ofZrSITe will follow, with the aim to exploitcircularly polarized pump pulses to opti-cally induce an asymmetric charge popu-lation both in the bulk and surface Diracparticles.

AcknowledgementsWe acknowledge financial support from

the Swiss NSF via the NCCR:MUST and thecontracts No. 206021-157773, and 407040-154056 (PNR 70), the European ResearchCouncil Advanced Grant H2020 ERCEA695197 DYNAMOX.


[1] S. Hüfner, ‘Photoelectron Spectroscopy: Prin-ciples and Applications’, Springer, 2003.

[2] A. Damascelli, Z. Hussain, Z.-X. Shen, Rev.Mod. Phys. 2003, 75, 473.

[3] W. S. Fann, R. Storz, H.W. K. Tom, J. Bokor,Phys. Rev. B 1992, 46, 13592.

[4] A. H. Castro Neto, F. Guinea, N. M. R. Peres,K. S. Novoselov, A. K. Geim, Rev. Mod. Phys.2009, 81, 109.

and 500 fs before the arrival of the pumppulse. Blue and red indicate a reductionand, respectively, an increase of intensity.Interestingly, the optical excitation mainlyaffects the surface state, whereas intensitychanges in the bulk states are below ourpresent sensitivity. Fig. 4(d,e) shows thetemporal evolution of the intensity aver-aged inside the two regions defined byred and blue rectangles in panel (b). Thedynamics is fast and after 1 ps the sys-tem is back to equilibrium. This indicatesthe presence of a very efficient relaxationmechanism, which might go beyond thesimple electron–phonon coupling. Furtherexperiments will be necessary to elucidateit, in particular by tracking the dispersionover larger regions of the SBZ. Finally,Fig. 4(f) provides a hint of the fact thatthe optical excitation might not be simplydescribed in terms of an increased elec-tronic temperature. Energy distributioncurves were integrated over a 0.2 Å–1 widemomentum window centered at k

F, for

negative (black) and positive (green) delaytimes. The depletion area below E

Fexceeds

the positive increase above EF. This obser-

vation suggests that the rapid diffusion ofthe optically excited carriers may yield anasymmetric distribution of hot electronsand holes.[25,33] Further experiments are inprogress to clarify this point.

In summary, we have shown that theASTRA setup enables static and VUVtime-resolved ARPES studies of solids.We illustrated its present performanceswith data collected on the novel Diracsemimetal compound ZrSiTe. Detailed

zontal (c) to vertical (d) provides a clearerpicture of the dispersion of the Dirac cone,which is split in two, very likely under theaction of spin-orbit coupling. In a centro-symmetric structure no spin splitting ispossible in the bulk states. Therefore thelifting of the spin degeneracy suggests thatthe state at Χhas a surface character.[57]Thisis confirmed by ab initio fully relativisticcalculations.[52] We stress again that thesespinsplit surface Dirac particles are onlyvisible near Χ, at ky = 0.85 Å–1. Such a mo-mentum region is precluded to tr-ARPESmeasurements with low-energy photons,as achieved by 4th harmonic generation,which can only probe the momentum re-gion inside the dashed black circle centeredat Γ (kmax

≈ 0.57Å–1). Hence, a comparisonbetween the out-of-equilibrium dynamicsof the bulk nodal line Dirac particles andthe spin-split surface Dirac particles inZrSiTe requires VUV tr-ARPES.

During the commissioning phase ofASTRA we have performed preliminarytr-ARPES measurements of ZrSiTe in theregion where both the bulk and surfacestates are present. Fig. 4(a,b) show theband dispersion for k

y= 0.65 Å–1, along

the green dashed line in Fig. 3(a). The in-tensity contrast in the VUV measurementat 27 eV (b) helps to identify the contri-butions of the various electronic states. Inparticular, we attribute the intense shallowband just below E

Fto the surface state. The

other weak spectral features arise from thebulk bands. Fig. 4(c) shows the differen-tial ARPES image, obtained as a differ-ence between data collected 100 fs after

-1.0

-0.5

0

0.5

1.0

-0.4 -0.2 0 0.2 0.4 -0.4 -0.2 0 0.2 0.4

1

2

region 1

region 2

1200

1000

800

600

400

200

-0.3 -0.2 -0.1 0.0 0.1 0.2 0.3

-500 fs100 fs

EDCs

200

100

0

-100

-1200

-800

-400

0

40003000200010000-1000

E - E (eV)F

delay time (fs)

Inte

nsi

ty(a

rb.

un

its)

De

lta

Inte

nsi

ty(a

rb.

un

its)

a) b) c) d)

e)

f)

E-

E(e

V)

F

-0.4 -0.2 0 0.2 0.4

1.5-1.54 3 2 16 4 2 0

hv = 21.2eV He Iα hv = 27eV H - pol

t = -500 fs t = 100 fs

k (A )x°-1 k (A )x

°-1 k (A )x°-1

Fig. 4. (a) Band dispersion along kx for ky = 0.65 Å–1 (green line in Fig. 3 (a)). (b) same as in panel (a), but measured with the HHG at 27 eV. (c)Differential ARPES image obtained as difference between the data 500 fs before and 100 fs after the arrival of the pump pulse. (d, e) Temporal dy-namics of the intensity integrated in the two region within the red and blue rectangles of panel (b), respectively. (f) Comparison between the energydistribution curves measured at –500 fs and 100 fs.


[5] M. Chhowalla, H. S. Shin, G. Eda, L.-J. Li, K. P.Loh, H. Zhang, Nat. Chem. 2013, 5, 263.

[6] P. R. Wallace, Phys. Rev. 1947, 71, 622.[7] S.Y. Zhou, G.-H. Gweon, J. Graf,A.V. Fedorov,

C. D. Spataru, R. D. Diehl,Y. Kopelevich, D.-H.Lee, S. G. Louie, A. Lanzara, Nat. Phys. 2006,2, 595.

[8] A. Bostwick, T. Ohta, T. Seyller, K. Horn, E.Rotenerg, Nat. Phys. 2007, 3, 36.

[9] M. Z. Hasan, C. L. Kane, Rev. Mod. Phys. 2010,82, 3046.

[10] B.-J.Yang, N. Nagaosa, Nat. Commun. 2014, 5,4898.

[11] T. Liang, Q. Gibson, M. N. Ali, M. Liu, R. J.Cava, N. P. Ong, Nat. Mater. 2015, 14, 280.

[12] P. Roushan, J. Seo, C. V. Parker, Y. S. Hor, D.Hsieh, D. Qian, A. Richardella, M. Z. Hasan, R.J. Cava, A.Yazdani, Nature 2009, 460, 1106.

[13] F. Bonaccorso, Z. Sun, T. Hasan, A. C. Ferrari,Nat. Photonics 2010, 4, 611.

[14] A. N. Grigorenko, M. Polini, K. S. Novoselov,Nat. Photonics 2012, 6, 749.

[15] C. X. Guo, H. B. Yang, Z. M. Sheng, Z. S. Lu,Q. L. Song, C. M. Li, Angew. Chem. Int. Ed.2010, 49, 3014.

[16] J. W. McIver, D. Hsieh, H. Steinberg, P. Jarillo-Herrero, N. Gedik, Nature Nanotech. 2012, 7,96.

[17] H. Zeng, J. Dai, W. Yao, D. Xiao, X. Cui, Nat.Nanotech. 2012, 7, 490.

[18] Q. H. Wang, K. Kalantar-Zadeh, A. Kis, J. N.Coleman, M. S. Strano, Nat. Nanotech. 2012, 7,699.

[19] J. A. Sobota, S.Yang, J. G. Analytis,Y. L. Chen,I. R. Fisher, P. S. Kirchmann, Z.-X. Shen, Phys.Rev. Lett. 2012, 108, 117403.

[20] M. Hajlaoui, E. Papalazarou, J. Mauchain, G.Lantz, N. Moisan, D. Boschetto, Z. Jiang, I.Miotkowski, Y. P. Chen, A. Taleb-Ibrahimi, L.Perfetti, M. Marsi, Nano Lett. 2012, 12, 3532.

[21] Y. H. Wang, D. Hsieh, E. J. Sie, H. Steinberg,D. R. Gardner, Y. S. Lee, P. Jarillo-Herrero, N.Gedik, Phys. Rev. Lett. 2012, 109, 127401.

[22] A. Crepaldi, B. Ressel, F. Cilento, M.Zacchigna, C. Grazioli, H. Berger, P. Bugnon,K. Kern, M. Grioni, F. Parmigiani, Phys. Rev. B2012, 86, 205133.

[23] M. Hajlaoui, E. Papalazarou, J. Mauchain,Z. Jiang, I. Miotkowski, Y. Chen, A. Taleb-Ibrahimi, L. Perfetti, M. Marsi, Eur. Phys. J.Special Topics 2013, 222, 1271.

[24] A. Crepaldi, F. Cilento, B. Ressel, C. Cacho,J. C. Johannsen, M. Zacchigna, H. Berger,P. Bugnon, C. Grazioli, I. C. E. Turcu, E.Springate, K. Kern, M. Grioni, F. Parmigiani,Phys. Rev. B 2013, 88, 121404.

[25] M. Hajlaoui, E. Papalazarou, J. Mauchain,L. Perfetti, A. Taleb-Ibrahimi, F. Navarin, M.Monteverde, P. Auban-Senzier, C. Pasquier,N. Moisan, D. Boschetto, M. Neupane, M.Z. Hazan, T. Durakiewicz, Z. Jiang, Y. Xu, I.Miotkowski, Y. P. CHen, S. Jia, H. W. Ji, R. J.

Cava, M. Marsi, Nat. Commun. 2014, 5, 3003.[26] J. Sobota, S.-L. Yang, D. Leuenberger, A.

Kemper, J. Analytis, I. Fisher, P. Kirchmann,T. Devereaux, Z.-X. Shen, J. Electr. Spectrosc.Rel. Phenom. 2014, 195, 249.

[27] M. Neupane, S.-Y. Xu, Y. Ishida, S. Jia, B. M.Fregoso, C. Liu, I. Belopolski, G. Bian, N.Alidoust, T. Durakiewicz, V. Galitski, S. Shin,R. J. Cava, M. Z. Hasan, Phys. Rev. Lett. 2015,115, 116801.

[28] J. C. Johannsen, G. Autès, A. Crepaldi, S.Moser, B. Casarin, F. Cilento, M. Zacchigna,H. Berger, A. Magrez, P. Bugnon, J. Avila, M.C. Asensio, F. Parmigiani, O. V. Yazyev, M.Grioni, Phys. Rev. B 2015, 91, 201101.

[29] S. Zhu,Y. Ishida, K. Kuroda, K. Sumida, M.Ye,J.Wang, H. Pan, M. Taniguchi, S. Qiao, S. Shin,A. Kimura, Sci. Report 2015, 5, 13213.

[30] K. Kuroda, J. Reimann, J. Güdde, U. Höfer,Phys. Rev. Lett. 2016, 116, 076801.

[31] A. Sterzi, A. Crepaldi, F. Cilento, G. Manzoni,E. Frantzeskakis,M. Zacchigna, E. vanHeumen,Y. K. Huang, M. S. Golden, F. Parmigiani, Phys.Rev. B 2016, 94, 081111.

[32] L. Khalil, E. Papalazarou, M. Caputo, N.Nilforoushan, L. Perfetti, A. Taleb-Ibrahimi, V.Kandyba, A. Barinov, Q. D. Gibson, R. J. Cava,M. Marsi, Phys. Rev. B 2017, 95, 085118.

[33] A. Sterzi, G. Manzoni, L. Sbuelz, F. Cilento, M.Zacchigna, P. Bugnon,A. Magrez, H. Berger,A.Crepaldi, F. Parmigiani, Phys. Rev. B 2017, 95,115431.

[34] C. Cacho, A. Crepaldi, M. Battiato, J. Braun,F. Cilento, M. Zacchigna, M. C. Richter, O.Heckmann, E. Springate, Y. Liu, S. S. Dhesi,H. Berger, P. Bugnon, K. Held, M. Grioni, H.Ebert, K. Hricovini, J. Minár, F. Parmigiani,Phys. Rev. Lett. 2015, 114, 097401.

[35] J. Sánchez-Barriga, E. Golias, A. Varykhalov,J. Braun, L. V.Yashina, R. Schumann, J. Minár,H. Ebert, O. Kornilov, O. Rader, Phys. Rev. B2016, 93, 155426.

[36] C. Jozwiak, J. A. Sobota, K. Gotlieb, A. F.Kemper, C. R. Rotund, R. J. Birgeneau, Z.Hussain, D.-H. Lee, Z.-X. Shen, A. Lanzara,Nat. Commun. 2016, 7, 13143.

[37] J. Sánchez-Barriga, M. Battiato, M. Krivenkov,E. Golias, A. Varykhalov, A. Romualdi, L. V.Yashina, J. Minár, O. Kornilov, H. Ebert, K.Held, J. Braun, Phys. Rev. B 2017, 95, 125405.

[38] J. C. Johannsen, S. Ulstrup, F. Cilento, A.Crepaldi, M. Zacchigna, C. Cacho, I. C. E.Turcu, E. Springate, F. Fromm, C. Raidel, T.Seyller, F. Parmigiani, M. Grioni, P. Hofmann,Phys. Rev. Lett. 2013, 111, 027403.

[39] I. Gierz, J. C. Petersen, M. Mitrano, C. Cacho,I. C. E. Turcu, E. Springate, A. Stöhr, A. Köhler,U. Starke, A. Cavalleri, Nat. Mater. 2013, 12,1119.

[40] S. Ulstrup, J. C. Johannsen, F. Cilento, J. A.Miwa, A. Crepaldi, M. Zacchigna, C. Cacho,R. Chapman, E. Springate, S. Mammadov, F.

Fromm, C. Raidel, T. Seyller, F. Parmigiani, M.Grioni, P. D. C. King, P. Hofmann, Phys. Rev.Lett. 2014, 112, 257401.

[41] A. Stange, C. Sohrt, L. X. Yang, G. Rohde,K. Janssen, P. Hein, L.-P. Oloff, K. Hanff, K.Rossnagel, M. Bauer, Phys. Rev. B 2015, 92,184303.

[42] J. C. Johannsen, S. Ulstrup, A. Crepaldi, F.Cilento, M. Zacchigna, J. A. Miwa, C. Cacho,R. T. Chapman, E. Springate, F. Fromm, C.Raidel, T. Seyller, P. D. C. King, F. Parmigiani,M. Grioni, P. Hofmann, Nano Lett. 2015, 15,326.

[43] R. Bertoni, C. W. Nicholson, L. Waldecker,H. Hübener, C. Monney, U. De Giovannini,M. Puppin, M. Hoesch, E. Springate, R. T.Chapman, C. Cacho, M. Wolf, A. Rubio, R.Ernstorfer, Phys. Rev. Lett. 2016, 117, 277201.

[44] S. Ulstrup, A. G. Cabo, D. Biswas, J. M.Riley, M. Dendzik, C. E. Sanders, M. Bianchi,C. Cacho, D. Matselyukh, R. T. Chapman,E. Springate, P. D. C. King, J. A. Miwa, P.Hofmann, Phys. Rev. B 2017, 95, 041405.

[45] M. Liu, M. B. Johnston, H. J. Snaith, Nature2013, 395, 501.

[46] G. Hodes, Science 2013, 342, 317.[47] N. J. Jeon, J. H. Noh, Y. C. Kim, W. S. Yang, S.

Ryu, S. I. Seok, Nat. Mater. 2014, 13, 897.[48] F. Cilento, G. Manzoni, A. Sterzi, S. Peli, A.

Ronchi, A. Crepaldi, F. Boschini, C. Cacho, R.Chapman, E. Springate, M. Capone, M. Berciu,A. F. Kemper, A. Damascelli, C. Giannetti, F.Parmigiani, arXiv:1703.03877v1, 2017.

[49] J. Ojeda, C. A. Arrell, J. Grilj, F. Frassetto, L.Mewes, H. Zhang, F. van Mourik, L. Poletto, M.Chergui, Struct. Dyn. 2016, 3, 023602.

[50] L. M. Schoop, M. N. Ali, C. Strasser, A. Topp,A. Varykhalov, D. Marchenko, V. Duppel, S. S.Parkin, B. V. Lotsch, C. R. Ast, Nat. Commun.2016, 7, 11696.

[51] J. Hu, Z. Tang, J. Liu, X. Liu, Y. Zhu, D. Graf,K. Myhro, S. Tran, C. N. Lau, J. Wei, Z. Mao,Phys. Rev. Lett. 2016, 117, 016602.

[52] A. Topp, J. M. Lippmann, A. Varykhalov, V.Duppel, B. Lotsch, C. R.Ast, L. Schoop, New J.Phys. 2016, 18, 125014.

[53] C. A. Arrell, J. Ojeda, L. Longetti, A. Crepaldi,S. Roth, G. Gatti, A. Clark, F. van Mourik, M.Drabbel, M. Grioni, M. Chergui, Chimia 2017,71, 268.

[54] A. J. K. Haneveld, F. Jellinek, Rec. Trav. Chim.Pays-Bas 1964, 83, 776.

[55] M. N. Ali, L. M. Schoop, C. Garg, J. M.Lippmann, E. Lara, B. Lotsch, S. S. P. Parkin,Sci. Adv. 2016, 2, 1.

[56] X. Wang, X. Pan, M. Gao, J. Yu, J. Jiang, J.Zhang, H. Zuo, M. Zhang, Z. Wei, W. Niu, Z.Xia, X. Wan,Y. Chen, F. Song,Y. Xu, B. Wang,G.Wang, R. Zhang, Adv. Electron. Mater. 2016,2, 1600228.

[57] Y. A. Bychkov, E. I. Rashba, JETP Lett. 1984,39, 78.


*Correspondence: Prof. Dr. S. RokeLaboratory for fundamental BioPhotonics (LBP)Institute of Bioengineering (IBI)and Institute of Materials Science (IMX)School of Engineering (STI)and Lausanne Centre for Ultrafast Science (LACUS)École Polytechnique Fédérale de Lausanne (EPFL)CH-1015 LausanneE-mail: [email protected]

Aqueous Nanoscale Systems

Sylvie Roke*

Abstract: In the past five years the Laboratory for fundamental BioPhotonics (LBP) has worked on developing newtechnology that can access the molecular structure and nanoscale properties of buried aqueous interfaces andaqueous solutions. Using these methods a better understanding of the important role that water plays in (nano-scale and interfacial) processes can be obtained. These processes include the long-range interaction of ions withwater, structural and charge anomalies of the hydrophobic/aqueous interface, the formation and stabilizationof amphiphilic aqueous droplet interfaces, the formation and molecular properties of the electric double layer,as well as membrane structure and hydration. The result of our work on these themes is summarized for thisspecial issue article.

Keywords: Droplets · Interfaces · Nonlinear imaging · Nonlinear light scattering · Membranes · Water

Introduction

Water. No substance on earth is so in-timately linked to our well-being. Withoutit, we die. On a more scientific level, with-out water, membranes – the structures thatprovide the architecture of our cells andorganelles – cannot function. Charges andcharged groups cannot be dissolved, self-assembly cannot occur, and proteins can-not fold. Apart from the intimate link withlife, water also shapes the earth and ourclimate. Our landscape is formed by sloweroding/dissolving processes of rocks inriver and sea water; aerosols and rain dropsprovide a means of transport of water. Oursociety depends on products that all relateto water and aqueous systems, such as foodproducts, medicine, and consumergoods.

Inmost of the abovementioned systemsit is the interfacial region (of themembrane,the droplet, or the particle) that determinesmuch of the physical, chemical, biologi-cal, and geological properties.[1,2] Interfa-cial water is often considered in one of twoways: As a background, describable by asingle parameter, or simply omitted.[3,4]Al-ternatively, it is studied in great detail in an

environment or condition that is preciselydefined but so oversimplified that it has notmuch to do with the real world. Aqueousinterfaces are mostly studied in vacuo, oras a planar water/air interface.[5,6]

However, interfacial water occurs ondifferent length scales, from sub-nano-meter to micron sized (corrugations, or-ganelles, membranes, liposomes), and isoften buried inside another solid or liquidenvironment that is not at all comparable tovacuum or air. This absence of molecularknowledge of realistic interfaces is due toa lack of tools that can access buried nano-or microscopic interfaces in liquids andsolids.

In thepastyears,wehaveobtainedabet-ter understanding of the important role thatwater plays in interfacial processes, suchas the long-range interaction of ions withwater, structural and charge anomalies ofthe hydrophobic/aqueous interface, theformation and stabilization of amphiphi-lic aqueous interfaces, the formation andmolecular properties of the electric doublelayer, as well as membrane structure andhydration. We have achieved these new in-sights by developing methods that can ac-cess molecular and nanoscale properties ofaqueous systems and interfaces more accu-rately, as well as probing multiple time andlength scales simultaneously. In what fol-lows we will first outline our strategy, andthen consider the above mentioned pro-cesses in more detail. Finally, we highlightseveral differences between nanoscopicand macroscopic aqueous interfaces.

Probes for Nanoscopic AqueousSystems

Probing aqueous systems on differentlength scales requires methods that coverdifferent length scales, such as linear andnonlinear light scattering measurements:

dynamic light scattering (~mm,s), fem-tosecond (fs) second harmonic scattering(sensitive to nanoscale information fromthe scattering pattern), vibrational sumfrequency scattering (spectral information,with sub-ps dynamical information andnanoscale information from the scatteringpattern), and multiphoton imaging (with a~200 nm resolution and 500 µm field ofview; and µs acquisition times). A largepart of our research effort has been aimedat developing those methods.We have alsodeveloped the necessary nonlinear opticalmodels and theories that allow us to accessdetailed molecular level information aboutinterfacial processes. Fig. 1 shows an illus-tration of the experimental methods withenergy level schemes for second harmonic(SH) and sum frequency (SF) generation.In both methods the production of a coher-ent SH or SF photon only occurs whennon-centrosymmetric molecules are spa-tially distributed in a non-centrosymmetricway. This allows one to selectively probevarious specific structures, such as po-lar fibrils (e.g. microtubules or collagen),aqueous interfaces, electric field inducedorientation of water molecules or the con-formation of moleculargroups.

In conjunctionwith the opticalmethodswe use a nanoparticle/droplet platform thatallows probing solid and liquid interfacesaround micron or nanoscale particles/droplets in solution. Nanodroplets havea surface to volume ratio that is ~4 ordersof magnitude larger than that of a planarmacroscopic interface. Thus, by using na-no-interfaces we can dramatically increasethe efficiency and accuracy of an interfacemeasurement. Furthermore, preparationprocedures can be done entirely in the bulkphase, and require a small sample volumeof typically ~50–100 microliter. This al-lows for a dramatic reduction of impurityissues and unwanted oxidation induced byambient air. It also reduces restrictions for


values from different surface models. Incontrast to other methods, no mean-fieldassumptions about the interfacial structureare needed. The portable table top methodcan be applied to any type of particle in anyliquid or solid medium.

In addition, the principles of nonlinearlight scatteringcanbe transferred tomicros-copy, allowing for high-throughput widefield multiphoton microscopy.[24–26] Ourapproach relies on reducing the repetitionrate of the femtosecond laser pulse train(from the GHz range), while optimizingthe pulse energy, so that the throughput canbe optimized.As the number of emitted SHphotons scales quadratically with the pulseenergy and linearly with the repetition rate,and the lower repetition rate ensures a re-duced heat load of the aqueous system, ahigher signal to noise level is achieved,[24]and photodamage effects are reduced sig-nificantly in living systems (allowing usto increase the dwell time by a factor of~106[26]). Fig. 3 shows the improvementin throughput compared to scanning con-focal second harmonic imaging. Singleliving mammalian neurons can now beimaged with second harmonic generationon the sub-cellular level with sub-secondacquisition times,[24] and the translationaland rotational diffusing of particles can bemeasured inside living cells with singleshot (5 microseconds)accuracy.[25]

Aqueous Systems

To advance our molecular level un-derstanding of complex aqueous systemswe have studied several phenomena thatcapture essential elements of biochemicalprocesses. Organized in order of increasingcomplexity these are: aqueous electrolytesolutions, hydrophobic aqueous interfac-es, charged amphiphilic interfaces, elec-tric double layer and hydrophobicity, andmembrane interfaces.

Aqueous electrolyte solutions form thematrix of life. Ions interact with water in

tional levels, Fig. 1b). This has led to in-strumentation with an increased through-put by three orders of magnitude.[15] Theenhancement enables the probing of theorientational ordering of water at nano-scopic aqueous interfaces on millisecondtime scales.

The technological developments gohandinhandwith theoreticalwork:Wehavepublished a series of papers laying downfundamental scattering theorems[16–20] thatcan be combined with molecular dynamicssimulations.[21,22] Our models are availableas freeware.

One recent example that stands out isthe measurement of unique values of thesurface potential from angle resolved po-larimetric SHS.[23] The method is illustrat-ed in Fig. 2. Fig. 2 shows two SH scatteringpatterns of 100 nm diameter liposomes indilute aqueous solution. The two indepen-dent scattering patterns are recorded withthe optical fields polarized in differentdirections. The patterns can be describedby an optical model that uses the surfacepotential as one of two fit parameters.The obtained surface potential values areplotted in the right panel and compared to

optical probes as the scattering experimentscan be performedwith the incoming beamsin transmission geometry as opposed tomore complex reflection schemes. In addi-tion, although many of the nanoscopic sys-tems studied can be considered as modelsystems, they also occur in living systemswith important biological functions (thinkof lipid droplets and liposomes).

In what follows, some exiting advance-ments that were achieved in the Roke labat EPFL are outlined, first regarding ex-perimental and theoreticalmethoddevelop-ment, and then regarding the advancementsin understanding the molecular architec-ture of aqueous systems.

Method Development

We have laid a technological founda-tion by developing time- and frequencyresolved femtosecond (fs) vibrational sumfrequency scattering (SFS), a combinationof light scattering and nonlinear spectros-copy in aqueous solutions (see refs [7–11],as illustrated in Fig. 1a). Thanks to thesymmetry selection rules for second-ordernonlinear optical processes, SFS allows tomeasure the molecular surface structure,morphology and chirality of nano- andmicroscopic objects in solution. A varietyof systems, such as polymer particles in asolid matrix,[9] particles in solution,[12] oildroplets in water,[7]water droplets,[13] a mi-cro-jet[11]and liposomes[14] in aqueous solu-tion were characterized, often with surpris-ing outcomes.Nanoscale curved interfacesdo not always behave in the same way asextended planar interfaces.

More recently, we have developed anew instrumental approach for fs-secondharmonic scattering (SHS, a non-resonantform of SFS, employing two identical fre-quencies, and addressing the electronicstates of a molecule rather than the vibra-

Fig. 1. Illustration of methods. (a). Sum frequency scattering. Sketch of the beam geometry andenergy level scheme. A combined IR and Raman transition occurs, which is only allowed in anon-centrosymmetric environment. (b) Photo (Alain Herzog, EPFL) of the second harmonic micro-scope and energy level scheme for non-resonant second harmonic generation. This process iselastic and has the same symmetry selection rules as sum frequency generation.

Fig. 2. Surface poten-tial measurements.(a). SHS patterns ofliposomes in twodifferent polarizationcombinations. Usingthe framework of refs[19,23], the two scat-tering patterns can bedescribed theoreti-cally using the surfacepotential and thesurface susceptibilityas sole unknowns.The procedure allowsfor the extraction ofunique surface poten-tial values.


ference between the surface and the bulkof the electrolyte solution.[27]The observedphenomena offer unique insights intothe nanoscale properties of liquid waterand electrolyte solutions that is exploredfurther in studies that involve both theo-ry[22,28,29] and experiment.[30]

In a second set of experiments we ad-dressed the question: “Does water inter-act differently with positive or negativelycharged ions?” This question is motivatedby the observation that in biological sys-tems interfaces are generally negativelycharged or charge neutral. In addition,hydrophobic interfaces carry an apparentnegative charge. This suggests that there isa difference between the hydration and sur-face structure of cations and anions. Com-bining Raman hydration shell spectro-scopy with SHS and SFS to investigate themolecular interactions with water in so-lution and at oil nanodroplet/water inter-faces, we mapped the interaction of tetra-phenyl ions with water. The chosen ions,sketched in Fig. 5a, that are either cationic,with anAs+ ion as the core of the moleculeor anionic, with a B– ion as the core of themolecule have virtually identical sizes,structures and polarizabilities. As such,the electrostatics of the interaction withwater and an aqueous interface should beidentical. The spectroscopic data, how-ever, shows remarkable differences, as canbe seen in Fig. 5b. Comparing molecularstructures and the vibrational signaturesof the hydrogen bonding in the spectra,we derived that the solvation of anions isslightly more energetically favorable thanthe solvation of cations. This differenceis caused by the different interactions atwork. Water–ion interactions can either bedipole-charge interactions, or comprisedof hydrogen bonds between the ions andthe water molecules. The first interactionshave a different orientation, depending onthe charge sign of the ion, while the hydro-gen bonds are always directed in the sameway, with the hydrogen donor of a watermolecule directed towards the phenyl ringof the molecular ion. As such, for a nega-tive ion both interactions will be coopera-tive, while for the cation they will be anti-cooperative, leading to a slightly less fa-vorable solvation energy of the anion overthe cation. At the interface, this results inthe cations being more readily solvated bythe oil phase.

Hydrophobic aqueous interfaces arekey to understanding interactions betweenwater and macromolecular systems andcan be prepared by dispersing pure hy-drophobic oil nanodroplets in water[31–35]or water droplets in oil.[13] Electrokineticmobility measurements show that hydro-phobic droplets in water are negativelycharged and that the charge increases dra-matically when the pH of the solution is

ic length scales, we have started to probe fsnanoscale structural correlations in liquidwater and electrolyte solutions. We found,surprisingly, that electrolytes induce long-range structural perturbations that appearat micromolar concentrations, equivalentto ion–ion separations of ~77 hydrationshells. Two examples of these experimentsare shown in Fig. 4. The change in the rela-tive SHS intensity is caused by a responseof the water–water hydrogen-bond interac-tions to the combined electric field of theions in the solutions. It can be seen thatthe experiment is particularly sensitive toquantum effects. In addition, the observedchanges in the SHS response correlate withmeasurable changes in the free energy dif-

many ways, changing dipole orientation,inducing charge transfer, and distorting thehydrogen-bonding network. These effectshave been studied in experiments prob-ing e.g. vibrational dynamics, dielectricresponses, infrared and Raman signaturesand computer simulations. All these stud-ies have shown that ion–water interactionsare short-range affecting the structure ofwater in the first, second and at most thethird hydration shell. However, such exper-iments and simulations are biased towardsdetecting short-range perturbations. Usingthe unique sensitivity of the fs-SHS instru-mentation and the property of nonlinearlight scattering experiments to be uniquelysensitive to tiny perturbations on nanoscop-

Fig. 3. Measured second harmonic imaging throughput. Measured Michelson contrast in imagesrecorded from the same position of the same 100 nm BaTiO3 particle sample in four differentsystems: wide-field (200 kHz, gated detection as proposed here, blue and red curves, usingdifferent camera settings), a scanning microscope (Leica TCS SP5 with 1028 nm, 88 MHz, 190 fslaser pulses illumination, a 1.2 NA 20x water immersion objective, a scanning rate of 1000 Hz/line,image size of 256 x 256 pixels, and collecting NA of 0.9), and a wide-field 1 kHz geometry with anormal CCD camera. The used pulse power and repetition rate are given in the legend. The insetshows an image of the nanoparticle sample corresponding to the red data point with the largestcontrast.

Fig. 4. Electrolytes in water: Left: fs-SHS data of aqueous solutions, showing changes in theorientational order at concentrations <10 µM. Right: Large differences are observed between lightand heavy water indicating the importance of nuclear quantumeffects.


that indicate the opposite are known.[40–42]These effects have not been quantified,however.Wehave started to do thiswith ourmultiscale toolbox and find that the nano-meter length scale and the hydrophobic-ity of the cations is crucial in determiningwhether the ions only influence the waterstructure in the interfacial region or alsothe surface itself.[43] In addition, the abilityto determine the average value of the sur-face potential is a great help.[23,44]

Membrane interfaces: Lipid dropletsand liposomes. The interfacial environ-ment of membranes is crucial for transport,signaling and the function of organelles.This comprises the molecular structure ofthe lipids aswell as the hydratingwater.Wehave developed a lipid droplet-like systemcomprised of oil droplets surrounded by alipid monolayer that both mimics the natu-rally occurring adiposomes and at the sametime presents a tunable, large surface tovolume ratio (but small volume) system formolecular studies of membrane systems(Fig. 6).[45,46] These droplets were used tostudy the membrane structure of liposomesin solution and it was found, surprisingly,that the inner and outer leaflets of lipo-somes have identical numbers of lipids.The significant difference in area (~15%,which was always assumed to be filledwith lipids) is filled upwith hydratingwatermolecules.[14] Specific head group interac-tions may cause transmembrane asymme-tries to occur. In addition surface potentialsof liposomes were determined, and it wasconcluded that the commonly employedmean field model cannot be used to de-scribe the electrostatic properties of theseinterfaces.[23] Instead charge condensationplays a very important role.[44]

rection. For the positive ones these interac-tions are anti-cooperative, similar to whatis illustrated in Fig. 5a. This finding is inclear contrast to the general understandingof amphiphilic interfaces.[38] Furthermore,planar interfaces and nanoscopic interfacesmade of the same chemicals do not displaythe same chemistry[36] when charges areinvolved,[34,39] which justifies the need forconsidering different length scales.

The electric double layer is a layerthat surrounds any aqueous charged in-terface and is thus a determining factor ininterface stability and nearly all forms ofbiochemical change. Although the layeris usually modelled with a mean field con-tinuummodel (meaning that all ions behaveas point charges and that water is a passivedielectric), a plethora of specific effects

increased.[31–33] This unusual behavior iscommonly explained by the presence ofhydroxyl ions at the interface. We observe,however, no pH dependent accumulationof OH– at the interface[33] and can explic-itly exclude surface active impurities as themajor driving force for charge accumula-tion.[32] By analyzing our data with nonlin-ear light scattering theory in combinationwith molecular dynamics simulations thatinclude charge transfer effects, we suggestthat rather than ionic adsorption, chargetransfer between water molecules is re-sponsible for the observed phenomenon.[31]

Charged amphiphilic interfaces areaqueous interfaces that contain both hydro-phobic and hydrophilic groups. We haveprepared droplets inwater that contain vari-ous amphiphilic molecules, such as neutraland ionic surfactants, as well as phospho-lipids. Bymeasuring themolecular compo-sition of the droplet oil surface, the surfaceamphiphiles and the adjacent water as wellas the interfacial charge, we obtain infor-mation of different length scales that can becombined in a unified interfacial structuralpicture. With this approach we found thatlong chain alkane (>C

8) oil droplets mini-

mize their interfacial free energy with sur-face oilmolecules that lie flat on the surfaceof the oil droplets.[21,36] Positively and neg-atively charged ionic amphiphiles interactremarkably different with hydrophobic/water interfaces.[37] Studies of the interfa-cial structure of sodium dodecylsulfate anddodecyltrimethylammonium, showed dis-tinct behaviors for negatively charged am-phiphiles and positively charged ones. Thenegatively charged amphiphiles are morehydrophilic than the positively chargedones, as the dipole-charge interaction be-tween the negative amphiphile and the wa-ter and the hydrogen bonding of water withthe head group are pointing in the same di-

Fig. 5. Charge asymmetric behavior of water. a) Cartoons illustrating the solvation of moleculartetraphenyl ions that have a virtually identical chemical structure and polarizability but a differentcharge. For the anions hydrogen bonds and dipole–charge interactions can be optimized withwater oriented in the same direction. For cations each interaction energy minimum requires adifferent water orientation. b) SFS spectra of the molecular anion (black) and the molecular cation(red) on an oil nanodroplet in water. The spectral region below 2960 cm–1 represents thevibrational modes of the oil, while the spectral region above 2960 cm–1 displays the structureof the molecular ions, which are seen to be remarkably different.

Fig. 6. Liposome transmembrane asymmetry. a) SFS spectra of ~50 nm radius DPPS (blue), DOPC(green), and DPPC (red) liposomes in D2O, probed in the P−O stretch region together with an SFSspectrum of hexadecane oil droplets covered with a DPPCmonolayer (top trace). The SFS dataare offset vertically for clarity. It can be seen that, in contrast to the oil droplet covered with lipids,there is no detectable transmembrane asymmetry of lipids. b) SHS patterns of the same liposomesin pure H2O. The scattering pattern originates from the overall transmembrane asymmetry in theorientational distribution of water molecules around the lipids (as illustrated in the cartoons).


Conclusions

In summary nonlinear light scatter-ing and imaging experiments offer a greatmultidimensional toolbox to investigateaqueous nanoscale systems and interfaces.Future studies aimed at characterizing wa-ter droplets, and their supercooling/freez-ing behavior are possible. More complexmembrane and lipid droplet studies can beinvestigated as well as more fundamentalenergy transfer processes (by for exampleimplementing a pump beam). With a size-able amount of information available on theinterfacial structure of nanoscale dropletsystems it becomes possible to more accu-rately determine the physics and chemistrybehind the difference in nanoscopic andmacroscopic systems. The unprecedentedsensitivity of second harmonic scatter-ing to the orientational order of water isa great tool to investigate the active rolewater plays in stabilizing interfaces and so-lutions. It will be used in the future to fur-ther investigate biological systems as wellas other fundamental surface processes.Especially the ability to determine surfacepotentials has great applications in biol-ogy, chemistry, and physics. In addition,many chemical reactions depend on it, forexample the transfer of electrons acrossa surface. A combination with imagingalso has promising applications, as manymembrane processes depend on surfacepotentials (for example (action) poten-tials).

AcknowledgementsThis work is supported by the Julia

Jacobi Foundation, the Swiss National ScienceFoundation (grant numbers 200021_140472,200021-146884, 200021_163210), the Euro-pean Research Council (grant numbers 240556and 616305), and the European Commission,Research Executive Agency Marie CurieActions ‘FINON’ (ITN-2013-607842).

Received: March 30, 2017

[1] A. W. Adamson, A. P. Gast, ‘Physical chemistryof surfaces’Wiley‐interscience: Sidney, 1997.

[2] R. J. Hunter, ‘Foundations of Colloid Science’,Oxford University Press: Sidney, 2002.

[3] P. Ball, Chem. Rev. 2008, 108, 74.[4] ‘Membrane Hydration: The Role of Water

in the Structure and Function of BiologicalMembranes’, Ed. E. A. Disalvo, Springer, 2015.

[5] E. Bjornehohn, M. H. Hansen, A. Hodgson,L. M. Liu, D. T: Limmer, A. Michaelides, P.Pedevilla, J. Rossmeisl, H. Shen, G. Tocci,E. Tyrode, M. M. Walz, J. Werner, H. Bluhm,Chem. Rev. 2016, 116, 7698.

[6] L. G. M. Pettersson, R. H. Henchman, A.Nilsson, Chem. Rev. 2016, 116, 7459.

[7] H. B. de Aguiar, J. S. Samson, S. Roke, Chem.Phys. Lett. 2011, 512, 76.

[8] H. B. de Aguiar, R. Scheu, K. C. Jena, A. G.F. de Beer, S. Roke, Phys. Chem. Chem. Phys.2012, 14, 6826.

[9] H. B. de Aguiar, A. G. F. de Beer, S. Roke, J.Phys. Chem. B 2013, 117, 8906.

[10] R. Scheu, S. Roke, J. Phys. Chem. B 2014, 118,3366.

[11] N. Smolentsev, Y. X. Chen, K. C. Jena, M. A.Brown, S. Roke, J. Chem. Phys. 2014, 141,18C524.

[12] S. Roke,O.Berg, J.Buitenhuis,A.vanBlaaderen,M. Bonn, Proc. Nat. Acad. Sci. 2006, 103,13310.

[13] N. Smolentsev, W. J. Smit, H. J. Bakker, S.Roke, submitted2017.

[14] N. Smolentsev, C. Lutgebaucks, H. I. Okur, A.G. F. de Beer, S. Roke, J. Am. Chem. Soc. 2016,138, 4053.

[15] N. Gomopoulos, C. Lutgebaucks, Q. C. Sun, C.Macias‐Romero, S. Roke, Opt. Express 2013,21, 815.

[16] A. G. F. de Beer, J. S. Samson, W. Hua, Z.Huang, X. Chen, H. C. Allen, S. Roke, J. Chem.Phys. 2011, 135, 224701.

[17] S. Roke, G. Gonella, Annu. Rev. Phys. Chem.2012, 63, 353.

[18] A. G. F. de Beer, Y. Chen, R. Scheu, J. C.Conboy, S. Roke, S. J. Phys. Chem. C 2013,117, 26582.

[19] G. Gonella, C. Lutgebaucks, A. G. F. deBeer, S . Roke, J. Phys. Chem. C 2016, 120,9165.

[20] A. G. F. de Beer, S. Roke, J. Chem. Phys. 2016,145, 044705.

[21] R. Vacha, S. Roke, J. Phys. Chem. B 2012, 116,11936.

[22] G. Tocci, C. Liang, D. M. Wilkins, S. Roke, M.Ceriottit, J. Phys. Chem. Lett. 2016, 7, 4311.

[23] C. Lutgebaucks, G. Gonella, S. Roke, Phys.Rev. B 2016, 94, 195410.

[24] C. Macias‐Romero, M. E. P. Didier, P. Jourdain,P. Marquet, P. Magistretti, O. B. Tarun, V.Zubkovs, A. Radenovic, S. Roke, Opt. Express2014, 22, 31102.

[25] C. Macias‐Romero, M. E. P. Didier,V. Zubkovs,L. Delannoy, F. Dutto, A. Radenovic, S. Roke,Nano Lett. 2014, 14, 2552.

[26] C. Macias‐Romero, V. Zubkovs, S. Wang, S.Roke, Biomed. Opt. Express 2016, 7, 1458.

[27] Y. Chen, H. I. Okur, N. Gomopoulos, C.Macias‐Romero, P. S. Cremer, P. B. Petersen, G. Tocci,D. M. Wilkins, C. W. Liang, M. Ceriotti, S.Roke, Sci. Adv. 2016, 2, e1501891.

[28] C. Liang, G. Tocci, D. M.Wilkins, A. Grisafi, S.Roke, M. Ceriotti, Phys.Rev.B2017, submitted.

[29] D. M. Wilkins, D. E. Manolopoulos, S. Roke,M. Ceriotti, J.Chem.Phys.2017, in press.

[30] H. I. Okur, Y. Chen, D. M. Wilkins S. Roke,Chem. Phys. Lett. (Frontiers) 2017, submitted.

[31] R. Vacha, S. W. Rick, P. Jungwirth, A. G. F. deBeer, H. B. de Aguiar, J. S. Samson, S. Roke, J.Am. Chem. Soc. 2011, 133, 10204.

[32] K. C. Jena, R. Scheu, S. Roke, Angew. Chem.Int. Ed. 2012, 51, 12938.

[33] J. S. Samson, R. Scheu, N. Smolentsev, S. W.Rick, S. Roke, Chem. Phys. Lett. 2014, 615,124.

[34] Y. Chen, K. C. Jena, S. Roke, J. Phys. Chem. C2015, 119,17725.

[35] W. J. Smit, N. Smolentsev, J. Versluis, S. Roke,H. J. Bakker, J. Chem. Phys. 2016, 145, 044706.

[36] H. B. deAguiar, M. L. Strader, A. G. F. de Beer,S. Roke, J. Phys. Chem. B 2011, 115, 2970.

[37] R. Scheu, Y. X. Chen, H. B. de Aguiar, B. M.Rankin, D. Ben‐Amotz, S. Roke, J. Am. Chem.Soc. 2014, 136, 2040.

[38] J. N. Israelachvili, ‘Intermolecular and surfaceforces’, Academic Press, 1991.

[39] R. Scheu, Y. Chen, H. B. de Aguiar, B. M.Rankin, D. Ben‐Amotz, S. Roke, J. Am. Chem.Soc. 2014, 136, 2040.

[40] W. Kunz, Curr. Opin. Colloid Interface Sci.2010, 15, 34.

[41] W. Kunz, Pure Appl. Chem. 2006, 78, 1611.[42] P. Jungwirth, P. S. Cremer, Nature Chem. 2014,

6, 261.[43] R. Scheu,Y. Chen, M. Subinya, S. Roke, J. Am.

Chem. Soc. 2013, 135, 19330.[44] C. Lutgebaucks, C. Macias‐Romero, S. Roke, J.

Chem. Phys. 2017, 146, 044701.[45] Y. Chen, K. C. Jena, C. Lutgebaucks, H. I. Okur,

S. Roke, Nano Lett. 2015, 15, 5558.[46] H. I. Okur,Y. Chen, N. Smolentsev, E. Zdrali, S.

Roke, J. Phys. Chem. B 2017,121, 2808.


*Correspondence: Prof. Dr. J. VanícekEcole Polytechnique Federale de Lausanne (EPFL)Institut des Sciences et Ingenierie ChimiquesLaboratory of Theoretical Physical ChemistryEPFL SB ISIC LCPT, BCH 3110CH-1015 LausanneE-mail: [email protected]

Several Semiclassical Approaches toTime-resolved Spectroscopy

Jirí Vanícek*

Abstract: Ultrafast spectroscopy allows molecular dynamics to be resolved on the femtosecond time scale.Whereas such short time scales obviously pose many experimental challenges, they provide an opportunity forsemiclassical methods, which are naturally suited for short time dynamics. Here we review several semiclassi-cal approaches for evaluating vibrationally resolved electronic pump-probe spectra, starting with the simplest,‘phase averaging’ or ‘dephasing representation’. We continue by discussing several methods developed in ourgroup that allow increasing the efficiency (the cellular dephasing representation) and accuracy (cellular dephas-ing representation with a prefactor) and end with the Gaussian dephasing representation, which, despite itssemiclassical origins, converges to the exact quantum result. The merits as well as shortcomings of the differentapproaches are demonstrated on time-resolved stimulated emission spectra of NCO and pyrazine.

Keywords: Dephasing representation · Fidelity amplitude · Phase averaging · Semiclassical approximation ·Time correlation function · Time-resolved stimulated emission

1. Introduction

High time resolution (such as 10–12 sor even 10–15 s) is important for under-standing physical and chemical processesinduced by the interactions of moleculeswith light; indeed, the femtosecond timeresolution has been the main challengeof ultrafast spectroscopy for almost threedecades. In contrast to the experimentaldifficulties, one expects that the short timescales should simplify theoretical simula-tions by requiring shorter propagation ofthe molecular wavepacket. Yet, whenevernuclear quantum effects play an essentialrole, even short-time simulations of thetime-dependent Schrödinger equation aredifficult because of their exponential scal-ing with the number of degrees of freedom.

To make such calculations practical, itis necessary to develop approximate dy-namical methods, which are feasible withthe computational resources available and,

at the same time, sufficiently accurate toanswer the questions of interest. In the caseof continuous-wave spectroscopy, wherethe light is not pulsed, but its coupling tothe molecular motion is moderately weakso that the first-order time-dependent per-turbation theory is valid, a very useful pic-ture of molecule–light interaction in termsof wavepacket autocorrelation functionshas been developed already in the 1970sand 1980s, especially by Heller,[1] whoalso suggested a very simple semiclassicalapproximation,[2] now called the thawedGaussian approximation, to evaluate vari-ous types of electronic spectra.

In the field of ultrafast spectroscopy,one must invoke higher orders of the time-dependent perturbation theory, a rewardfor this effort being an even richer varietyof phenomena. A systematic analysis ofcorrelation functions and response func-tions contributing to various types of time-resolved spectra has been developed in the1980s and 1990s, and is summarized in acomprehensive book[3] by Mukamel, whoalso proposed a very simple semiclassicalmethod, called phase averaging,[4] allow-ing the evaluation of various types of time-resolved spectra.

In this article, we review several recent-ly developed semiclassical methods forevaluating time-resolved electronic spec-tra that can be thought of as extensions ofMukamel’s phase averaging. Starting froman alternative presentation of linear spec-troscopy that makes the analogy to nonlin-ear spectroscopy obvious, we describe thephase averaging, the dephasing representa-tion and its several variants that can makethe method more efficient, more accurate,and sometimes even exact.

2. Time-dependent Approach toSpectroscopy

2.1 Linear Spectra: AutocorrelationFunction vs. Fidelity Amplitude

In the time-dependent approach tospectroscopy, pioneered by Heller,[1] thelinear electronic absorption spectrumσ (ω) of a molecule can be computed asthe Fourier transform

(1)

of the wavepacket autocorrelation function

(2)

of an initial state ψ〉 ≡ 0,0〉 given by thevibrational ground state of the electronicground state potential energy surface,moving on the excited state potential en-ergy surface described by the Hamiltonianoperator 1H . Here µ

01is the transition di-

pole moment between the ground and ex-cited electronic states, ω is the frequencyof the electromagnetic radiation, and E

0,0denotes the zero point vibrational energyin the ground electronic state. Eqn. (1) as-sumes the validity of the electric dipoleapproximation (requiring the wavelengthof the electromagnetic field to be muchlarger than the size of the molecule), thefirst-order time-dependent perturbationtheory(restricting the strength of the elec-tromagnetic field), Condon approxima-tion (requiring the transition dipole to beindependent of the nuclear coordinates),and low temperature approximation (im-plying that only the ground vibrational

0,0 /201 0

4( ) Re ( )3

i E tC t e dtc

tdtt01Re01c

1ˆ /( ) | ( ) | |iH tC t t e ||


(5)

of the wavepacket correlation func-tion[6,20,21]

(6)

where τ is the time delay between thepump and probe pulses, t denotes the timeelapsed after the probe pulse, and

(7)

stands for the initial state evolved forthe delay time τ with the excited stateHamiltonian and subsequently for timet with either the ground or excited stateHamiltonian (j = 0, 1).

As written, the correlation functionf(t,τ) from Eqn. (6) has an immediate in-terpretation as a quantum fidelity ampli-tude between states ψ

0(t,τ) and ψ

1(t,τ).

This fidelity amplitude now correspondsto evolutions for time t + τ of the sameinitial state ψ with two Hamiltonians, oneHamiltonian being time-independent andequal to 1H , the other becoming time-de-pendent and given by 1H until time τ, andby 0H at later times. Note that the correla-tion function f(t,τ) can be also interpretedas a correlation function (4) from linearspectroscopy, but applied to a nonstation-ary initial state 𝑒𝑒𝑒𝑒𝑒𝑒(−𝑖𝑖𝑖𝑖𝐻𝐻/ℏ) 𝜓𝜓ñ preparedby the pump pulse.[6,7]

3. Several SemiclassicalApproaches to Time-resolvedElectronic Spectra

Correlation functions C(t), f(t), orf(t,τ) can of course be evaluated exactlyquantum mechanically, but this requiresthe exact solution of the time-dependentSchrödinger equation, which can be pro-hibitively expensive in many dimensions.A nice feature of electronic spectra, andultrafast spectra in particular, is theirshort-time nature, which offers itself toapproximative treatment of dynamics.Semiclassical approximation providesa perfect candidate since it is typicallyexact for short times while its accuracydeteriorates at longer times. As the sim-plest starting point, Heller originallyused a single Gaussian as an ansatz forthe wavefunction, which is exact in up toquadratic potentials and yields the thawedGaussian approximation in general po-tentials.[2] Alternatively, one may employmultiple trajectory-based methods suchas the frozen Gaussian approximation,[22]initial value representation,[23,24] or their

0( , ) Re ( , ) i tf t e dt

1 0( , ) ( , ) | ( , ) ,f t t t

1ˆ ˆ/ /| ( , ) : |jiH t iH

j t e e 1ˆ /iH1e 1 |

applications: Outside of electronic spec-troscopy,[6–12] it has proved useful, e.g. inNMR spin echo experiments[13] and theo-ries of quantum computation,[5] decoher-ence,[5] and inelastic neutron scattering.[14]In chemical physics, the fidelity amplitudewas also used as a measure of the dynami-cal importance of diabatic,[15] nonadiabat-ic,[16] or spin-orbit couplings,[17] and of theaccuracy of quantum molecular dynamicson an approximate potential energy sur-face.[18,19]

2.2 Time-resolved ElectronicSpectra

In the case of nonlinear spectra, theautocorrelation picture is no longer valid,yet, as we now show, the more general pic-ture using fidelity amplitude remains ap-plicable. A wide variety of nonlinear time-resolved spectra belong to the pump-probescheme, in which an ultrashort pump pulseprepares a nonstationary nuclear wave-packet in an excited electronic state, andanother ultrashort pulse probes the dynam-ics of this wavepacket after a certain timedelay τ. There are many possible experi-mental setups depending on the polariza-tion and mutual orientation of the pumpand probe laser beams and on the directionin which the signal is detected,[3] but for thesake of clarity we will only consider time-resolved stimulated emission (TRSE) here(see Fig. 1 (b)).

Besides the assumptions used for linearspectra, a simplified picture of TRSE takesadvantage of the nonoverlapping pulsesapproximation (i.e. the pump and probepulses can be treated independently) andthe ultrashort pulse approximation, whichassumes that both the pump and probepulses are short compared to the nucleartime scale but long on the electronic timescale.

Assuming, for simplicity and as be-fore, the zero temperature approximation,electric dipole approximation, and time-dependent perturbation theory (of whichthe third order is now required), the dif-ferential TRSE spectrum at frequency ωand time delay τ can be computed as theFourier transform

state 0,0〉 of the ground surface be occu-pied initially).

The beauty of Eqns. (1) and (2) lies intheir simple interpretation (see also Fig. 1(a)): theabsorptionofaphotonof frequencyω instantaneously promotes the stationaryvibrational ground state of the ground sur-face to the excited potential energy surface,where this, now nonstationary state startsmoving under the influence of the excitedstate Hamiltonian alone. In particular, theexplicit form of the electromagnetic fielddoes not play any role and the linear ab-sorption spectrum is determined solely bythe field-free dynamics of the wavepacketψ (t) on the excited surface. Indeed, this isthe content of the linear response theory,which is here equivalent to the first-ordertime-dependent perturbation theory.

Note that since ψ〉 ≡ 0,0〉 is an ei-genstate of 0H , the spectrum can be alsowritten as

(3)

where

(4)

is a correlation function, called the fidelityamplitude,[5] between two states ψ

0(t) and

ψ1(t), both starting from the same initial

state ψ , but one evolved with 0H and theother with 1H . As the name suggests, thefidelity amplitude measures the similaritybetween the quantum evolutions on theground and excited surfaces. This alterna-tive expression for an electronic spectrumis not a mathematical curiosity; indeed, itis the direct outcome of the derivation ofthe spectrum using the first-order time-de-pendent perturbation theory, and it is onlydue to the additional assumption thatψ is avibrational ground state (oranother eigen-state) of 0H that one obtains the much bet-ter known expression (1) for the spectrumin terms of the wavepacket autocorrelationfunction (2).

The less often used correlation func-tion (4) has, nonetheless, many important

201 0

4( ) Re ( ) ,3

i tf t e dtc 01Re01c

0 1ˆ ˆ/ /

0 1( ) ( ) | ( ) | |iH t iH tf t t t e e ˆ /1 /iH t1 |

Fig. 1. Schematic representation of physical processes underlying two types of vibrationally re-solved electronic spectra. (a) Linear absorption. (b) Time-resolved stimulated emission.

a) b)


and Filinov filtering[33–35] used for othersemiclassical methods. In the cellular de-phasing representation (CDR),[21,36] theinitial conditions are grouped into cells ofneighboring trajectories, and contributionsof all trajectories within a cell are evalu-ated approximately analytically using theinformation collected along the centraltrajectory of the cell. This leads only to aminormodification of the numerical algo-rithm (12), which becomes

(13)

where ACDR

(x0,τ,t) is a prefactor[36] captur-

ing the contribution of neighboring trajec-tories and ρ

IWTis the inverse Weierstrass

transform of ρW(in other words, it is the

expansion coefficient of ρWin a basis of

phase space Gaussians).[36]Fig. 2 shows that in a two-dimensional

model of NCO,[21] which is only slightlyanharmonic, the cellular dephasing repre-sentation based on a single trajectory canyield a time-resolved stimulated emissionspectrum that agrees very closely with thefully converged dephasing representation(using over sixteen thousand trajectories).

Although the dephasing representationis exact for displaced harmonic oscilla-tors[3] and remarkably accurate in chaoticsystems,[26] its efficiency, of course, doesnot come for free. Due to its relation to thesemiclassical perturbation approximation,the dephasing representation breaks downeven for only quadratic perturbations en-counteredalreadyinharmonicsystemswithdifferent force constants. Unfortunately,these are important in so-called ‘silentmodes’ in electronic spectroscopy, whichare modes that are not vibrationally ex-cited by the electronic excitation; if there

0

IWT 0

( , , )/CDR CDR ( )

( ,, ) i S x t

xf t A e

(11)

and the trajectory xt’follows the excited

state Hamiltonian H1for 0 < t' < τ and the

average Hamiltonian H for t' > τ. (Fromnow on, for brevity we shall only presentexpressions for the TRSE spectra, and nomore for linear absorption spectra.)

Expression (10) immediately suggestsa numerical recipe for its evaluation: 1)sample initial conditions x

0from the phase

space density ρW(x

0), 2) run classical tra-

jectories with these initial conditions, 3)evaluate the phase (11) due to the differ-ence of the potential energies along eachof these trajectories, and 4) average overthese trajectories. In a more compactform,this recipe can be expressed as

(12)

where 〈A(x0)〉ρ(x0) denotes, more generally,

an average of an observable A(x0) over ini-

tial conditions x0sampled from the density

ρ(x0).There are many ways to derive the

dephasing representation: it can be ob-tained,[26,27] e.g. by linearizing the semi-classical propagator, which is a procedureinspired by the semiclassical perturbationtheory.[29,30] Shi and Geva[12] derived thesame approximation (but referred to it as thelinearized semiclassical initial value repre-sentation) without invoking the semiclassi-cal propagator – by linearizing directly thepath integral quantum propagator. Amongthe appeals of the dephasing representationis the ease with which it is numerically eval-uated: We have shown[31] that the expectednumber of trajectories required for conver-gence of the dephasing representation isindependent of dimensionality, time, andnature of the dynamics, and depends explic-itly only on the magnitude of the correlationfunction one wants to simulate.

3.2 Increasing the Efficiency andAccuracy: Cellular DephasingRepresentation with a Prefactor

Unlike other semiclassical methods,which typically require a Hessian of thepotential energy and thousands or millionsof trajectories for convergence, the dephas-ing representation only needs the energygradient, and about hundred to thousandtrajectories for full convergence. This ex-traordinary efficiency of the dephasingrepresentation makes it a promising candi-date for on-the-fly ab initio evaluation oftime-resolved spectra.

Yet, for large systems, even a thousandtrajectories may be too much to ask for,and hence we have attempted to reduce thenumber of trajectories by so-called cellu-larization, inspired by cellular dynamics[32]

0'( , , ) ' ( ),

t

tS x t dt V xt

tt

+D = Dò

0

W 0

( , , )/DR ( )( , ) ,i S x t

xf t e

combination giving the Herman-Klukpropagator,[25] but these methods becomequickly expensive since they requireboth the Hessian of the potential energysurface and very large numbers of tra-jectories for convergence. Here we willreview several variants of an alternative,very simple semiclassical approximation,called phase averaging, dephasing repre-sentation, or Wigner averaged classicallimit, which require only the gradient ofthe potential energy, a rather small num-ber of trajectories for convergence, andtake advantage of the specific form of thecorrelation functions corresponding toelectronic spectra.

3.1 Phase-averaging/DephasingRepresentation

A remarkably simple approximationfor the correlation function (or fidelity am-plitude) f(t) is given by the so-called phaseaveraging, dephasing representation, orWigner averaged classical limit[3,4,7–12,26,27]:

(8)

where D is the number of degrees of free-dom, ρ

W(x

0) is theWigner phase-space rep-

resentation of the initial state ψ and

(9)

is the action due to the difference ∆V :=V

1– V

0between the two potential energy

surfaces along the classical trajectory xt

≡ (qt, p

t) driven by the average[3,4,20,28]

Hamiltonian H≡ (H0+ H

1)/2.

In the original phase averaging,[4] theweight function in Eqn. (8) was a classi-cal density ρ(x

0), and three options for

the Hamiltonian used for driving the tra-jectories were considered: besides H, onecould use H

1(suitable for absorption spec-

tra) or H0(suitable for emission spectra).

Replacement of the classical density withthe more accurate Wigner function gaverise to the name ‘Wigner averaged classi-cal limit’.[10,11]The name ‘dephasing repre-sentation’, on the other hand, suggests thatthe overlap of ψ

0and ψ

1in this approxima-

tion decays only due to dephasing, i.e. a de-structive interference,whereas the classicaloverlap is assumed to be constant and fixedat 1.Thedephasing representationbecomesapplicable to ultrafast spectra after an ap-propriate generalization of the fidelity am-plitude. In the case of TRSE spectra (Eqns.(5) and (6)), the fidelity amplitude can beapproximated as

(10)

where the action difference is given by

0( , )/DR 0 W 0( ) ( ) ,i S x tDf t h dx x e

0

0 '( , ) ' ( )t

tS x t dt V x

0( , , )/DR 0 W 0( , ) ( ) ,i S x tDf t h dx x e

-0.4

-0.2

0

0.2

0.4

0.6

0.8

0.08 0.10 0.12 0.14

σ(arb.units)

ω [a.u.]

quantum

DR (N = 16 384)

CDR (N = 1)

Fig. 2. Time-resolved stimulated emissionspectrum of a collinear model of NCO from ref.[21] for a delay time τ = 500 a.u. ≈12 fs. Thespectrum obtained with the cellular dephasingrepresentation (CDR) using a single trajectory(!) is in remarkable agreement with the fullyconverged spectrum obtained with the originaldephasing representation (DR, computed usingN = 16384 trajectories), which, in turn, repro-duces the main qualitative features of the exactquantum spectrum. (Adapted from ref. [21].)


dent Schrödinger equations (17), and fi-nally, the fidelity amplitude is evaluated as

(19)

Note that if the Gaussian basis gα(t,τ)

is large enough, the Gaussian dephasingrepresentation is not a semiclassical ap-proximation; indeed, it should converge tothe exact quantum answer as the Gaussianbasis approaches completeness. (Beware,however, of various numerical issues dueto nonorthogonality of the basis, etc.[37–40])

Fig. 4 demonstrates the accuracy ofthe Gaussian dephasing representationon the time-resolved stimulated emissionspectrum of pyrazine, showing that theGaussian dephasing representation with576 basis functions yields a correlationfunction and spectrum which are basicallyindistinguishable from the exact quantumanalogs, unlike the fully converged de-phasing representation, which does con-tain a semiclassical error.

As for the computational cost of theGaussian dephasing representation, thenumber of trajectories required is muchsmaller than in typical semiclassical meth-ods, and can be even smaller than in theoriginal dephasing representation since inthe GDR, the trajectories carry with themGaussian basis functions, which play asmoothing role similar to the cells in thecellular dephasing representation. Themost expensive part per trajectory of theGDR is the evaluation of the Hamiltonianmatrix elements Hαβ: using, e.g. a localharmonic approximation for the poten-tial requires the Hessian of the potentialenergy, but one can often get away onlywith a linear expansion that only neces-sitates the gradient, which is alreadyneeded for propagating the trajectories.The final contribution to the cost is solv-ing the time-dependent Schrödinger equa-tion, which scales as O(N3), i.e. cubicallywith the number of trajectories, which is

†GDR 1 0( ) ( , ) , ( , ).f t t t tc S c

3.3 Making the DephasingRepresentation Exact: GaussianDephasing Representation

An alternative approach for improvingthe accuracy of the dephasing representa-tion replaces the swarm of N independentsemiclassical trajectories with a swarm ofN ‘communicating’ Gaussian basis func-tions moving along corresponding classi-cal trajectories. This trick is closely relat-ed to the basic idea employed in multiplespawning,[37] variational Gaussian wave-packets,[38] coupled coherent states, andmulticonfigurational Ehrenfest method.[39]In particular, the states |ψ

j(t,τ)〉 are expand-

ed as

(16)

where gα(t,τ)〉 is a Gaussian wavepacketwhose center moves along the classical tra-jectory ofH

1until the delay time τ andwith

the average Hamiltonian H from then on.The expansion coefficients c

j,α(t,τ) satisfy

the time-dependent Schrödinger equation

(17)

where Hjis the Hamiltonian matrix, S is

the overlap matrix, and D the nonadiabaticcoupling matrix defined by their matrix el-ements in the Gaussian basis:

(18)

In the Gaussian dephasing representa-tion (GDR),[40] one runs classical trajec-tories as in the original dephasing repre-sentation, but uses them only to guide theGaussian basis functions g

α(t,τ). The time

dependence of the expansion coefficientscj,αare obtained by solving the time-depen-

,1

| ( , ) ( , ) | ( , ) ,N

j jt c t g t

1( ) , 0 '( ,

,) ' +

jj

j j

i ti

i t tH D c

ScH D c

) ,j , 0) 00, 0))))))((

Sc(jc ,j ,,

j

ˆ( , ) : ( , ) | | ( , ) ,( , ) : ( , ) | ( , ) ,( , ) : ( , ) | ( , ) .

jH t g t H g tS t g t g tD t g t g t( , ) .(g (( ,(

are many such ‘boring’ modes, the dephas-ing representation breaks down completelydue to an artificially fast decay of the cor-relation function. Zambrano and Ozorio deAlmeida proposed[28] a simple recipe forpartially correcting this inaccuracy by in-cluding a (different) prefactor A

DRP(x

0,τ,t)

in the dephasing representation, resultingin the dephasing representation with aprefactor (DRP):

(14)

It turns out that the two prefactors inEqns (13) and (14) can be easily combined,yielding the cellular dephasing represen-tation with a prefactor (CDRP),[36] whichmay be both more accurate and more ef-ficient than the original dephasing repre-sentation:

(15)

Indeed, Fig. 3 shows on the time-resolved stimulated emission of pyrazinethat the cellular dephasing representationwith a prefactor is not only more accuratebut also requires fewer trajectories forconvergence than does the original de-phasing representation. (Note, however,that this property is not universal, and thatin strongly chaotic systems, such as thequartic oscillator, the original dephasingrepresentation can converge faster than theCDRP, since a few chaotic trajectories canresult in large prefactors that require manywell-behaved trajectories to compensatethis blowup in the final result.[36]) Even ifthe number of trajectories is reduced by thecellularization, the cost of each trajectoryincreases significantly since the prefactorsA

CDRand A

DRPrequire the evaluation of the

Hessian or even the third derivatives of thepotential energy, unlike the dephasing rep-resentation for which the force is all thatis needed.

0

W 0

( , , )/DRP DRP ( )

( , ) .i S x t

xf t A e

0

IWT 0

( , , )/CDRP CDR DRP ( )

( , ) .i S x t

xf t A A e

Fig. 3. Time-resolved stimulated emission in the pyrazine S0/S1 model from ref. [36]. Initial state is the ground state of the S0 surface and the delaytime between pump and probe pulses is τ = 2×103 a.u. ≈48 fs. Comparison of the exact quantum result, the original dephasing representation (DR),and the cellular DR with a prefactor (CDRP). (a) Time correlation function (already multiplied by a damping function indicated by a dash-dotted line).(b) Corresponding spectrum. (c) Convergence error (relative L2 norm error) of the damped correlation function as a function of the number of trajec-tories N. (Adapted from ref. [36].)

0.0

0.2

0.4

0.6

0.8

1.0

0 4 8 12

(a)

|f|·χ

t/103 [a.u.]

damping

quantum

DR

CDRP

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

0.12 0.14 0.16

(b)

σ(arb.units)

ω [a.u.]

quantumDR

CDRP

10-2

10-1

100

101

101 102 103 104

∝ 1/√N

(c)

convergence

error

N

DR

CDRP


much worse than the linear O(N) scalingof the dephasing representation, whose Ntrajectories are independent. On the otherhand, and in particular in on-the-fly ab ini-tio applications, the cost of the electronicstructure, especially the Hessian, becomeseasily so high that the cubic scaling of thesolution of the Schrödinger equation maystill be negligible if the number of trajec-tories is below a few hundred.

4. Conclusions and Outlook

All methods that we have discussed sofar express correlation functions needed intime-resolved spectra calculations in termsof interfering contributions from classi-cal trajectories. This common feature ofall the discussed methods makes possibletheir implementation together with an on-the-fly ab initio evaluation of electronicstructure, possibly turning them into anautomated and almost ‘black box’ tool foranalyzing ultrafast spectra without the ne-cessity of a tedious construction of globalor semi-global potential energy surfaces.

The accuracy of the approximate meth-ods, can be, of course, improved in otherways than using Gaussian basis methods.Recently,[41] we have described an arbi-trary-order expansion of the Feynman pathintegral representation of the correlationfunctions (4) or (6), of which the first-orderexpansion yields the dephasing represen-tation, while the zeroth-order expansiongives the static classical limit,[10] which it-self is very useful. It turns out that alreadythe second order expansion would correctmost of the shortcomings of the dephasingrepresentation in typical chemical systems(which are neither exactly harmonic, norchaotic); unfortunately, the most straight-forward implementation is very inefficient.

Another important aspect of ultrafastspectra not captured by the methods re-viewed here is the common presence ofnonadiabatic and spin-orbit couplingsbetween various electronic states contrib-uting to the spectra, which give rise to in-ternal conversion or intersystem crossingsbetween various states. If weak, these pro-cesses lead only to the broadening of thespectra, but, if strong, they can completelychange the spectral line shapes. To addressthis issue, we have generalized the dephas-ing representation to the setting of coupledelectronic states and obtained themultiple-surface dephasing representation[42] thatcan capture the major consequences of thenonadiabatic or spin-orbit couplings on ul-trafast electronic spectra. This method canand has been combined with an on-the-flyab initio evaluation of energies, forces, andnonadiabatic couplings.

Besides the efficiency and ease of on-the-fly evaluation of electronic structure,the trajectory-based methods for evalua-ting ultrafast spectra have another advan-tage, probably the most important of all.Theyprovide an intuitive picture of the dy-namics, which is much easier to decipherthan, e.g. a 30-dimensional wavefunctionin the case of pyrazine. While the simplerpicture of course does not have to be quan-titatively correct 100% of the time, wehave been time and again surprised by thequalitative correctness of the semiclassicalresult.

AcknowledgementsThis research was supported by the Swiss

NSF with the grant number 200020 150098and within the NCCR Molecular UltrafastScience and Technology (MUST), and by theERC Consolidator Grant Project No. 683069MOLEQULE. The author would like to thankall the former and current members of theLaboratory of Theoretical Physical Chemistryat EPFL for their contributions to the researchreviewed in this article, Julien Roulet for trans-lating the manuscript from Latex to MicrosoftOffice Word, and Sergey Antipov for helpfulcomments.


[1] E. J. Heller, Acc. Chem. Res. 1981, 14, 368.[2] E. J. Heller, J. Chem. Phys. 1975, 62, 1544.[3] S. Mukamel, ‘Principles of nonlinear optical

spectroscopy’, 1st ed., Oxford University Press,NewYork, 1999.

[4] S. Mukamel, J. Chem. Phys. 1982, 77, 173.[5] T. Gorin, T. Prosen, T. H. Seligman, M.

Žnidaric, Phys. Rep. 2006, 435, 33.[6] W. T. Pollard, S.-Y. Lee, R. A. Mathies, J.

Chem. Phys. 1990, 92, 4012.[7] N. E. Shemetulskis, R. F. Loring, J. Chem.

Phys. 1992, 97, 1217.[8] J. M. Rost, J. Phys. B 1995, 28, L601.[9] Z. Li, J.-Y. Fang, C. C. Martens, J. Chem. Phys.

1996, 104, 6919.[10] S. A. Egorov, E. Rabani, B. J. Berne, J. Chem.

Phys. 1998, 108, 1407.[11] S. A. Egorov, E. Rabani, B. J. Berne, J. Chem.

Phys. 1999, 110, 5238.[12] Q. Shi, E. Geva, J. Chem. Phys. 2005, 122,

064506.[13] H. M. Pastawski, P. R. Levstein, G. Usaj, J.

Raya, J. Hirschinger, Physica A 2000, 283, 166.[14] C. Petitjean, D. V. Bevilaqua, E. J. Heller, P.

Jacquod, Phys. Rev. Lett. 2007, 98, 164101.[15] T. Zimmermann, J. Vanícek, J. Chem. Phys.

2010, 132, 241101.[16] T. Zimmermann, J. Vanícek, J. Chem. Phys.


2012, 137, 22A516.[18] B. Li, C. Mollica, J. Vanícek, J. Chem. Phys.

2009, 131, 041101.[19] T. Zimmermann, J. Ruppen, B. Li, J. Vanícek,

Int. J. Quant. Chem. 2010, 110, 2426.[20] M. Wehrle, M. Šulc, J. Vanícek, Chimia 2011,

65, 334.[21] M. Šulc, J. Vanícek,Mol. Phys. 2012, 110, 945.[22] E. J. Heller, J. Chem. Phys. 1981, 75, 2923.[23] W. H. Miller, J. Chem. Phys. 1970, 53, 3578.[24] W. H. Miller, J. Phys. Chem. 2001, 105, 2942.[25] M. F. Herman, E. Kluk, Chem. Phys. 1984, 91,

27.[26] J. Vanícek, Phys. Rev. E 2004, 70, 055201.[27] J. Vanícek, Phys. Rev. E 2006, 73, 046204.[28] E. Zambrano, A. M. Ozorio de Almeida, Phys.

Rev. E 2011, 84, 045201(R).[29] W. H. Miller, F. T. Smith, Phys. Rev. A 1978, 17,

939.[30] L. M. Hubbard, W. H. Miller, J. Chem. Phys.

1983, 78, 1801.[31] C. Mollica, J. Vanícek, Phys. Rev. Lett. 2011,

107, 214101.[32] E. J. Heller, J. Chem. Phys. 1991, 94, 2723.[33] V. S. Filinov, Nucl. Phys. B 1986, 271, 717.[34] N. Makri, W. H. Miller, Chem. Phys. Lett. 1987,

139, 10.[35] A. R. Walton, D. E. Manolopoulos, Mol. Phys.

1996, 87, 961.[36] E. Zambrano, M. Šulc, J. Vanícek, J. Chem.

Phys. 2013, 139, 054109.[37] M. Ben-Nun, T. J. Martínez, Adv. Chem. Phys.

2002, 121, 439.[38] G.A.Worth, M.A. Robb, I. Burghardt, Faraday

Discuss. 2004, 127, 307.[39] D. V. Shalashilin, J. Chem. Phys. 2010, 132,

244111[40] M. Šulc, H. Hernandez, T. J. MartÍnez, J.

Vanícek, J. Chem. Phys. 2013, 139, 034112.[41] J. Vanícek, D. Cohen, Phil. Trans. R. Soc. A


2014, 141, 134102.

0.0

0.2

0.4

0.6

0.8

1.0

0 2 4 6 8 10 12 14 16

f

t/103 [a.u.]

damping

quantumGDR (N = 576)

(a) DR (converged)

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

0.12 0.14 0.16

Reσ(arb.units)

ω [a.u.]

quantum

GDR (N = 576)(b) DR (converged)

Fig. 4. Time-resolved stimulated emission inthe same pyrazine S0/S1model as in Fig. 3.Comparison of the exact quantum result, theoriginal dephasing representation (DR), andthe Gaussian DR (GDR) based on 576 trajec-tories only. (a) Time correlation function. (b)Corresponding spectrum obtained as a Fouriertransform of the correlation function multipliedwith a damping function indicated in panel (a)by a gray dash-double-dotted line. (Adaptedfrom ref. [40].)


*Correspondence: Prof. Dr. M. CherguiLab. de Spectroscopie Ultrarapide (LSU) andLausanne Centre for Ultrafast Science (LACUS)Ecole Polytechnique Fédérale de LausanneISIC, FSB, Station 6CH-1015 LausanneE-mail: [email protected]

The LOUVRE Laboratory: State-of-the-ArtUltrafast Ultraviolet Spectroscopies forMolecular and Materials Science

Malte Oppermann, Natalia S. Nagornova, Aurelio Oriana, Edoardo Baldini, Lars Mewes, BenjaminBauer, Tania Palmieri, Thomas Rossi, Frank van Mourik, and Majed Chergui*

Abstract:We describe the facilities for ultraviolet studies in the femtosecond to nanosecond time domain. Thesefacilities consist of: i) a set-up for deep-ultraviolet spectroscopy in the 260–380 nm range in both pump and probepulses for transient absorption/reflectivity or two-dimensional spectroscopy studies; ii) a set-up for ultrafastfluorescence measurements with detection down to 300 nm. The capabilities of these set-ups are demonstratedby examples on molecular systems, biosystems, nanoparticles and solid materials.

Keywords: Absorption · Charge carriers dynamics · Charge transfer · Deep-UV · Energy relaxation ·Fluorescence · Materials · Molecules · Proteins · Reflectivity · Solvation dynamics · Two-dimensional spectros-copy · Ultrafast · Ultraviolet

1. Introduction

Driven by the need to describe non-equilibrium phenomena, steady-state spec-troscopic tools for the investigation ofequilibrium structures continue to be trans-lated into their time-domain analoguesin the femtosecond-picosecond regime.During the past three decades, the growingtoolbox of ultrafast spectroscopies has suc-cessfully established the fields of femto-chemistry and femto-biology,[1–6] with aremarkable ability to unravel in real timethe structural and relaxation dynamics ofelectronic and vibrational excited states ofmolecules, proteins and materials. Pulsedcoherent sources from X-rays to terahertzradiation have allowed researchers to in-vestigate ultrafast dynamics via the asso-ciated light–matter interactions and gainaccess to electronic, spin and structuralchanges from the site-specific atomic tothe supra-molecular scale.[7]

At the LSUwithin the Lausanne Centrefor Ultrafast Science (LACUS), we haveconcentrated part of our research effortson the 250–400 nm spectral range, i.e. theultraviolet (UV) to deep-UV region. Thereare several reasons for targeting this range:

i) the absorption bands of amino-acid res-idues such as tryptophan (Trp), tyrosine(Tyr) and phenylalanine (Phe), and ofnucleobases such as guanine, cytosine,adenine, etc. lie in this region; ii) the lat-ter is also where several small molecules,organic or inorganic, absorb strongly; iii)it corresponds to the bandgap of semicon-ductors transition metal oxides (TMO).The latter has attracted much interest inthe past twenty years due to their appli-cations in solar energy conversion and incatalysis. While these systems have beeninvestigated in the THz,[8,9] infrared[10] andvisible[11–14] regions there have never beenstudies with deep-UV continuum probes,which can cover the region both above andbelow the band gap.

The development of ultrafast UVspectroscopic tools has resulted in theLOUVRE (Lots Of UV Radiation for yourExperiments) laboratory: a unique univer-sity-based facility that comprises varioustechniques for investigating a wide rangeof liquid and solid samples. In this article,we present this toolbox of methods anddiscuss its capabilities by highlightingsome of our most recent results. We hopeto illustrate the rich contributions of ultra-fast UV spectroscopy to a wide range of re-search fields such as biology, coordinationchemistry and materials science.

2. Experimental Techniques andSet-ups

The LOUVRE consists of two mainset-ups: (1) an ultra-stable, high repetitionrate femtosecond laser source for tuneabledeep-UV pump and broadband UV con-

tinuum probe pulses that are employed intransient absorption (TA) measurements(Fig. 1) or two-dimensional (2D) spectros-copy, and; (2) a high repetition rate, hightime-resolution fluorescence up-conver-sion set-up for ultrafast emission meas-urements (Fig. 2). Each of these set-upshas its characteristic advantages and offershighly flexible experimental target stationsdedicated to different sample types. Mostimportantly, these two approaches are fullycomplementary.

In most ultrafast spectroscopic meas-urements, a first (pump) pulse triggers aphotophysical or photochemical processby excitation of the sample, while a second(probe) pulse, whose time delay is tuneablewith respect to the pump pulse, using op-tical delay lines, is used to interrogate thesystem. In broadband transient absorption(TA), the probe pulse monitors the ab-sorption change of the sample induced bythe pump. In the simplest case, the signal(Fig. 1) consists of an induced transpar-ency or ground state bleach (GSB), dueto the fact that the ground state has beendepopulated, an excited state absorption(ESA) or a stimulated emission (SE).Since the latter and the GSB imply morephotons reaching the detector, they haveequal signs, while the ESA has an oppo-site sign (Fig. 1c).

In fluorescence up-conversion (Fig.2a), the second (gate) pulse does not inter-fere with the sample but it opens a gate ata given time delay in a nonlinear mediumwhich permits the detection of the spon-taneous fluorescence, by mixing the gatepulse with the fluorescence and detectingthe sum (up-) or difference (down-) fre-quency of the two.


nonlinear signal.[19,20] The main technicalchallenge of this method is the generationof phase-locked, temporally compressedpulses, required to perform a clean FT ofthe nonlinear signal without distorting thetwo-dimensional maps. Despite these chal-lenges, the access to phase information andthus to vibrational and electronic coher-ences makes it a particularly rewarding ap-proach.[21–25] However, as UV absorptionbands of chromophores are often broad,covering their entire profile of typically50 nm requires a ~3 fs pulse at 300 nm,which is quite challenging. The second ap-proach is usually referred to as 2DUV TA,and it uses a single tuneable narrowbandpump pulse to sequentially acquire TAspectra with a broadband UV continuumprobe. Whilst lacking phase information,this technique neither requires transformlimited pulses nor their phase control. Itsmain advantages are a much simpler andmore robust set-up at the cost of a reducedtime resolution. However, the latter is not

shapes, such that their temporal evolutioncan be used to study the interaction be-tween a system and its environment, e.g. asolute inside a solvent.

The interest in 2DUV spectroscopy isobvious for the study of the structure anddynamics of biosystems with a complexmolecular architecture.[6,16,17] Extendingnonlinear optical spectroscopy techniquesinto the UV region is a particularly chal-lenging technological task, which orig-inates from the increased requirementsregarding phase- and dispersion-control atshort wavelengths. Nevertheless, through-out the past decade impressive progresshas been made, which is presented in theexcellent review by Cannizzo.[18]There aretwo approaches to this type of spectrosco-py: the Fourier transform (FT or coherent)one and the transient absorption (or inco-herent) one. In the former, the excitationfrequency is resolved by varying the delaybetween two ultrashort pump pulses andperforming the FT of the time-dependent

2.1 Two-dimensional UVSpectroscopy

In the past twenty years, there has beena drive aimed at extending the methods ofmultidimensional Fourier transform NMRspectroscopy into the shorter wavelengthrange.[4,5,15] This was achieved first in theinfrared,[6] then in the visible region.[3]The aim of such developments is to revealcorrelations and interactions between thevibrational or electronic dipoles probedas cross-peaks in a two-dimensional map,which typically plots the probe frequencyas a function of the excitation frequency.In this way, one may obtain information onelectron and/or energy transfer processesand conformational changes encoded inthe electronic dipole couplings. This is il-lustrated in Fig. 3, where the idealized caseof two coupled electronic transitions andthe associated two-dimensional spectrumis explained. In addition, homogeneousand inhomogeneous contributions can beextracted from the two-dimensional line

Fig. 1. Schematic illustration of a transient absorption (TA) experiment. (a) typical pump-probe arrangement, where a UV-pump pulse interacts with asample and the transmitted intensity of a broadband Visible- or UV-probe pulse is measured using a detector (a photomultiplier, PMT, for example).(b) processes contributing to the idealised TA spectrum displayed in (c). When a UV-pump transfers population from a ground state (S0) to an excitedstate (S1), the absorption corresponding to the S0–S1 transition is reduced. This leads to a negative ground state bleach (GSB) feature in the transientabsorption spectrum. Analogously, excited state absorption (ESA) and stimulated emission (SE) may take place from the transiently populated state S1.

Fig. 2. Schematic illustration of a fluorescence up-conversion (FlUC) experiment. (a) A simplified experimental set-up, where a UV-pump inducesfluorescence in a sample. The corresponding energy level scheme is displayed in (b). The emission is collected and focused into a nonlinear medium,where it interacts with a gating pulse. The up-converted signal is then detected by a photomultiplier (PMT) or a CCD camera. (c) Principle of afluorescence time-gating: at t=0 the pump pulse (purple) excites the sample and triggers a spontaneous fluorescence of the sample with a giventime profile (green trace). This fluorescence is mixed within the gate pulse (orange) whose time delay with respect to the pump is tuneable. At eachtime delay the sum frequency of the fluorescence with the gate pulse yields the up-converted light (blue arrows) whose time profile reflects that ofthe fluorescence.


20 kHz.[27]As a result, absorbance changes<1 mOD can be resolved after compara-tively short measurement times.

The flexibility of this set-up is exempli-fied by the large diversity of systems andprocesses we investigated. Here, the use offlow cells or wire-guided liquid jet allowsthe investigation of liquid samples, while atransient reflectivity configuration enablesthe study of single crystals.

2.2 Time-resolved FluorescenceFemtosecond fluorescence up-conver-

sion (FlUC) spectroscopy is a widely usedtechnique for studying intramolecularelectronic relaxation processes, vibration-al cooling and solvation dynamics of mo-lecular systems. It is especially used whenthe fluorescence decay times are less than afew ps, a time scale that cannot be resolvedwith detectors such as multichannel platesor streak cameras cannot resolve the decay.This approach represents a simplificationcompared to TA because the sample inter-acts with the excitation pulse only (Fig.2). The operating principle of our set-upis shown in Fig. 5. The fluorescence is col-lected by wide-angle optics and focusedonto a nonlinear sum-frequency (SF) crys-tal in which it is mixed with the gate pulse,whose time delay with respect to the pumppulse is controlled by an optical delaystage. The intensity of the sum-frequency(up-converted) light is then recorded as afunction of delay time between pump andgate pulses. With reflective optics to mini-

cluding its third and second harmonic)or to generate a white-light continuumprobe for visible TA. For the detection,a fibre-coupled imaging spectrograph isused in conjunction with a fast, double-ar-ray multichannel detector. This allowsshot-to-shot recording of the probe pulsespectrum together with a reference beamat an exceptional data-acquisition rate of

a stringent requirement, especially in bio-logical systems where interactions be-tween chromophores can be long-rangeand therefore, occur over rather long timescales.

The latter approach was implementedat the LOUVRE lab. Despite not being asingle-shot technique, due to the requiredpump frequency scan, we circumvent longdata acquisition times through the use ofan ultra-stable, high repetition rate femto-second UV pulse source. The associatedbroadband TA set-up (Fig. 4) accommo-dates probe pulse configurations both inthe visible and deep-UV regions and hasbeen described in detail elsewhere.[26,27]It delivers broadband visible femtosecondpump pulses (520–740 nm bandwidth,15 µJ per pulse) that can be used as suchor be frequency doubled to generate deep-UV narrowband (1.5 nm) pump pulsesvia second harmonic generation (SHG)in different β-barium borate (BBO) crys-tals. For the probe, a so-called achromaticdoubling scheme is employed,[28] wherebroadband phase matching is achieved byspatially chirping the beam before tightlyfocussing it into a thin BBO crystal. Thisroutinely provides pulses with bandwidths>100 nm in the deep-UV. This configura-tion represents the optimal conditions forpump-probe 2DUV spectroscopy withan instrument response function (IRF) of150 fs. Nevertheless, improved time res-olution down to 50 fs has been achievedby narrowing the visible NOPA outputspectrum or by separately compressingthe visible pump pulses before SHG.[29]In addition, part of the 800 nm outputcan be used as a pre-pulse or pump (in-

Fig. 3. Principle of multidimensional spectroscopies. (a) Energy level scheme of two coupledoscillators and (b) relative two-dimensional map of the differential transmission signal. In (a) solidred arrows correspond to a ground-state bleach (GSB), dashed red arrows to stimulated emission(SE) and blue arrows to excited state absorption (ESA). In (b) red signals are positive, blue signalsare negative. Homogeneous broadening (HB) and inhomogeneous broadening (IB) are highlightedin the peak along the diagonal. Figure reproduced from ref. [62].

Fig. 4. Optical beamline (a) and performance of the probe pulse generation via achromaticdoubling (AD) displayed in (b). The pump pulse generation via tuneable narrowband doubling isshown in (c). The probe pulses (dark blue line) are focussed via a parabolic mirror (PM) into thesample flow cell (FC) and the transmitted intensity is collected via a fibre-coupled (F) spectro-graph. The pump pulses from the NOPA (orange line) are optically chopped by a chopper (C) and fre-quency doubled via a tuneable BBO (light blue line). The remaining visible light is filtered via dichroicmirrors (DM) and focussed on the sample via a spherical mirror (SM). Reproduced from ref. [26].


20 to 30 ps for Trp14 and 110 to 140 ps forTrp7.[43–45] These have been attributed toFRET to the heme.[33,45] Using our 2DUVset-up, we revealed that in ferric myoglob-ins MbCN and metMb,[39] Trp14 undergoesa hitherto unsuspected partial electrontransfer to the heme, while relaxation of themore distant Trp7 is indeed due to FRET tothe heme (Fig. 6). The electron transferredfrom Trp14 ends up on the iron atom form-ing a ferrous porphyrin. We then extend-ed these studies to the case of the ferrous(unligated) deoxy form,[40] and found thesame Trp14-to-heme electron transfer, thistime leading to the formation of a low-va-lence porphyrin anion radical. More recent(unpublished) work on ligated ferrousMbsreveals the same pattern of Trp-to-hemeelectron transfer. In ref. [40], we suggestedthat the pathway for the electron transferproceeds via the leucine 69 (Leu69) and va-line 68 (Val68) residues, which are in vander Waals contact with each other, whileLeu69 is in van der Waals contact withTrp14. In a recent theoretical modelling ofthe Trp-heme electron transfer Suess et al.confirmed our hypothesis.[46] The aboveresults of a Trp-mediated electron transfercompeting with FRET are especially im-portant for the fundamental understandingof electron transfer in biological systems,but also show a limitation of Trp as a spec-troscopic ruler in FRET studies of proteindynamics. Indeed, evidence that FRET istaking place requires not only the decay ofthe donor fluorescence, but also the detec-tion of the acceptor luminescence, a crite-rion that is often overlooked in studies ofprotein dynamics.

3.2 Coordination ChemistryTransition metal complexes possess

different types of states such as: met-al-to-ligand charge transfer (MLCT), met-al centred (MC), ligand centred (LC) andligand-to-metal charge transfer (LMCT).The MLCT state is often the lowest di-pole-allowed electronic transition. Thesecomplexes are involved in several appli-cations such as solar cells, OLEDs, mo-lecular electronics, biology, magnetic datastorage, etc., for which the understandingof the underlying ultrafast photophysics isessential.

Fluorescence up-conversion is ideal inthis respect and as an example we showhere the case of the cascade of electronicstates in an iridium complex, Ir(ppy)

3, up-

on 266 nm excitation of its high-lying li-gand-centred (LC) electronic state.[47] Theluminescence of Ir(ppy)

3in DMSO upon

266 nm excitation of its LC state is shownin Fig. 7 as a function of wavelength andtime delay. The LC fluorescence, centred at330 nm, decays in 70 ± 10 fs, which is sig-nificantly shorter than the lifetime of ppyligands in solution (∼1 ns).[48] The LC state

redox states.A single Trp residue, which islocated very close to the heme and knownto undergo very efficient energy transferto it, makes Cyt c an ideal model systemto study the photodynamics of this pro-tein class in its two different redox states.Indeed, we showed, using time-resolvedfluorescence, that excited Trp decays[36]with a much shorter time constant (350 fs)in the ferrous state than in the ferric case(770 fs). This implies a more efficient en-ergy transfer in the former case, whichmaybe due to the significantly different rela-tive orientations and distances between theTrp and the heme groups in the two redoxstates, which is supported by computationsand (CD) measurements.[41,42] Excitingthe heme did not show a response of theTrp contrary to what we had observed inbacteriorhodopsin.[34,35] This result is sur-prising and it calls for further investigationas the Trp decay times reflect a strong in-teraction with the heme. In addition, themuch shorter Trp decay times in Cyt c areto be contrasted to the case of myoglobins,which we discuss hereafter.

Myoglobin (Mb) belongs to the fam-ily of heme proteins and consists of 153amino acids and an active iron porphyrin(heme) centre. It contains two tryptophanresidues: Trp7 and Trp14, which are locatedin the α-helix A with respective distancesof 21.2 Å and 15.2 Å (centre to centre) tothe heme group. In both the ferrous (Fe2+)and the ferric state (Fe3+), the metal atomcan bind a variety of diatomic ligands (in-cluding O

2, CO, NO and CN), which de-

termines its function. Photoexcitation ofthe Trp residues in Mb leads to its fluores-cence, with much shortened decay timescompared to the isolated case (ca. 3 ns):

mize frequency dispersion effects after thesample, a time resolution of 100–130 fscan be reached.[30] To accomplish a broad-band detection, the up-converted signal, atfixed time delay, is detected with a spec-trograph equipped with a charge coupleddevice (CCD) camera, and the SF crystalis rotated during the integration time at aconstant angular speed to cover the spec-tral range of interest. The peculiarity ofour set-up lies in the fact that it extendsfrom the IR (between 1 and 2 µm)[31] to theUV[30] (above 300 nm) spectral domains.

3. Research Highlights

3.1 Biological SystemsUV-chromophores are site-specific

probes of their local environment.[32] Thisis especially the case for tryptophan (Trp),whose absorption and emission bands arecentred at about 280 nm and 360 nm, re-spectively. For example, as part of a donor–acceptor pair undergoing FluorescenceResonance Energy Transfer (FRET), itsfluorescence decay is routinely used todetermine the geometrical arrangementof the chromophore pair with nm resolu-tion.[33] In addition to its capacity to act asa probe for energy and transfer processes,Trp is also a sensor of local electric fieldsinside a protein.[34,35] In this spirit, wehave applied our UV-spectroscopy toolsto probe the energy and charge transferprocesses in hemoproteins, such as horseheart cytochrome c (Cyt c)[36,37] and myo-globins.[38–40]

In refs [36] and [37] we investigated theinteraction between the heme and a singleTrp residue in Cyt c in its ferric and ferrous

Fig. 5. Schematics of the time-gated fluorescence up-conversion experimental set-up. Afterexcitation of the sample (S), spontaneous emission is collected and focused by two parabolicmirrors onto a sum-frequency (SF) crystal. The broad emission spectrum is then time-gated bythe delayed pulses, throughout the continuous rotation of the SF crystal. After spatial filtering ofscattered light, the up-converted signal is dispersed by a grating inside of a monochromator andfinally detected by a liquid nitrogen cooled, UV-enhanced CCD camera. For the sake of clarity, theSFG process and the signal filtering parts are shown from the top and form the side. Reproducedfrom ref. [30].


then decays to the manifold of close-lyingspin-mixed MLCT and MC states whichundergo an ultrafast (<10 fs) intramolecu-lar electronic-vibrational relaxation, lead-ing to population of the high vibrationallevels of the lowest 3MLCT state, whosepopulation rises in 70 fs. We are thereforeable to observe the departure, the interme-diate steps, and the arrival of the relaxationcascade that spans more than 1.6 eV, fromthe 1LC state to the lowest 3MLCT state,yielding the long-lived luminescence of themolecule. These results represent the firstmeasurement of the complete relaxation

time over an entire cascade of electronicstates in a polyatomic molecule. The mostremarkable result is that intramolecularelectronic relaxation proceeds at sub-vi-brational time scales (<10 fs) over a largeenergy gap. This was already hinted at inour previous studies on organic and inor-ganic molecular complexes,[49,50] but theoccurrence of LC emission in the presentcase provides a ‘clock’ of the relaxationdynamics. Such ultrafast electronic ‘cool-ing’ does not mean that the molecule is‘cold’ in absolute terms. The excess elec-tronic energy is impulsively converted into

low frequency vibrational modes of themolecule, which are often optically silentones. The most outstanding outcome ofthese studies is that electronic ‘cooling’ oflarge energy gaps occurs at sub-vibration-al time scales. This issue calls for furtherinvestigations with methods at even highertime resolution, eventually going into theattosecond regime.

3.3 Materials Science: TransitionMetal Oxides

2DUVTA spectroscopy offers new per-spectives in the field of materials science,shedding light on the elusive physics ofstrongly interacting and correlated quan-tum systems. Using this non-equilibriumspectroscopic approach, one can aim at:1) discovering new hidden or metastablehigh-energy excitations in a given materi-al; 2) revealing how the low-energy excita-tions affect specific high-energy collectivemodes (e.g. excitons…) of the system. Inthis respect, ultrafast broadband UV spec-troscopy becomes a superior tool for un-ravelling complex phenomena in TMOs, aclass of materials in which a plethora ofeffects emerges due to the non-trivial in-terplay between low- and high-energy de-grees of freedom.

Anatase TiO2is a prototypical TMO

used in a number of applications (e.g. pho-tocatalysis, photovoltaics, etc.). Recently,the application of equilibrium spectrosco-pies and many-body perturbation theorycalculations revealed that the absorptionthreshold of this material is dominatedby strongly bound direct excitons risingover the continuum of indirect interbandtransitions.[51] These excitons are con-fined on a two-dimensional (2D) planeof the three-dimensional crystal latticeand remain stable at room temperaturein the case of single crystals. However,in all applications, highly defective sam-ples (e.g. nanoparticles, NPs) are used atroom temperature and ambient pressure.One may therefore question the existenceof these excitons for the actual systemsused in technology. Indeed, the equilib-rium absorption spectrum of colloidalanatase TiO

2NPs does not show obvious

signatures of excitonic transitions, and israther featureless (see black dotted tracein Fig. 8a). A powerful approach to ad-dress the existence of excitonic transitionsin a material is to interrogate the systemout-of-equilibrium via ultrafast broadbandoptical spectroscopy. Typically, the excitonline shapes can be identified through thepump-induced transparency of the exciton-ic peak, referred to as ‘exciton bleaching’.To clearly observe this effect in colloidalsolutions of anatase TiO

2NPs, ultrafast TA

UV spectroscopy is a superior techniquebecause it subtracts the scattered light andprovides a better contrast for resolving

Fig. 6. 2DUV transientabsorption of myo-globins revealed thatwhile Trp-7 decaysvia FRET to the heme,Trp14 undergoeselectron transfer witha yield of ca. 50%while the rest decaysby FRET (see refs[39,40]).

Fig. 7. Absorption spectrum of fac-tris(2-phenylpyridine)-iridium(iii) or Ir(ppy)3 in a DMSO solventand time zero fluorescence spectrum (black) upon excitation of the LC band at 266 nm. The redtrace represents the steady-state fluorescence spectrum of the ppy ligand and the blue tracethe steady-state absorption spectrum. The inset compares the cross-correlation (red Gaussian),which is the Raman signal of the solvent, with the time profile measured on the emission at 340nm. Reproduced from ref. [47].


1 ps relaxation of the solvent cage aroundthe aqueous electron[58,59] but the strongabsorption (in TA) or PE signal of the sol-vated electron have hindered the observa-tion of concurrent dynamics of the solute.Fluorescence of the CTTS states wouldprovide a direct measurement of the elec-tron departure to the solvent since photonsare emitted during the time the ground andthe excited (CTTS) state wave functions ofthe solute overlap.We reported for the firsttime the observation of CTTS fluorescenceusing deep-UV FlUC in the case of aque-ous I–.[60]

The fluorescence spectrum of aqueousiodide, measured upon excitation at 266nm, is shown in Fig. 9. An emission span-ning from the UV (approx. 300 nm) to thevisible region (approx. 670 nm) appearspromptly at time zero. Its decay is stronglywavelength-dependent going from ca. 60fs at λ <330 nm to ∼50 fs at 650 nm. Theseresults reveal the very large inhomogeneityof the excited centres with each iodide hav-ing a somewhat different solvent shell att=0. The starting solvent configuration willdetermine the subsequent dynamics wherethe solvent accommodates to the excitedspecies giving rise to very different Stokesshifts spanning over 1 eV. The redder theemission, the larger the Stokes shift and,therefore, the more stable the configura-tion, which is reflected in a longer decaytime. The various time scales, in particularin the redder part of the spectrum couldnot be observed by TA or other methodsbecause the signal of the solvated electronovershadows these dynamics.

4. Conclusions

The present contribution reported onthe applications of ultrafast deep UV spec-troscopies, either in transient absorption,reflectivity, and in fluorescence. The avail-ability of high performance set-ups allowsthe study of a wide range of systems, forwhich prominent deep-UV bands are pres-ent. The examples of biological systemsand transition metal oxides presented hereunderscore the importance of coveringthis spectral range. In the case of transitionmetal oxides, the ability to excite either thefirst excitons or higher into the conductionband allows for a distinction between thedynamics of excitons as opposed to thatof free carriers. Concerning molecularsystems, several studies are now possi-ble for which the photoinduced processesare excitation wavelength dependent, e.g.the recently investigated case of aqueous[Fe(CN)

6]4–.[61] Developments are under-

going aimed at time-resolved deep-UVcircular dichroism, which can be appliedto studies of chirality or conformation-al changes in proteins. Finally, pushing

ety of quantum phenomena in complex ox-ides. Extensions of this technique involvethe study of direct excitons in ZnO,[52] themapping of the charge dynamics across thecharge-transfer gap in NiO,[53] and the un-ravelling of spin polarons in LaSrMnO

4.[54]

3.4 Solvation DynamicsThe first step of an intermolecular

charge transfer (CT) reaction (e.g. electronor proton transfer) between a donor andan acceptor in solution is a charge transferto the solvent. Charge-transfer-to-solvent(CTTS) states, which are quasi-boundstates of the solute–solvent system with noequivalent for the isolated ions, are idealobjects to investigate this first step. CTTSstates are common in the case of aqueoushalides and, since the latter lack internal(nuclear) degrees of freedom, the CTTSdynamics is entirely dependent on thestructure andmotion of the solvent species.CTTS states are therefore ideal probes ofelectronic solvation dynamics upon elec-tronic excitation of a solute.

Ultrafast TA and photoemission (PE)spectroscopies have been employed to ob-serve the early-time dynamics of aqueousCTTS states.[55–59] It has been shown thatin water, the electron is detached from theiodide in ∼0.2 ps followed by an approx.

possibly hidden features. By applying thistechnique on single crystals (in reflectiv-ity mode) and nanoparticles (in transmis-sion mode), the excitonic species could beclearly revealed (Fig. 8).[51] Fig. 8b showsthe results of transient reflectivity meas-urements of single crystals after convert-ing them to absorption changes. The data isdependent on the crystal orientation, show-ing that the first peak is dominant along thea-axis, while the second band is dominantalong the c-axis. More details are found inref. [51]. These results show that the ex-citons are very robust against defects, anaspect that has strong consequences in thefield of applied research, since the excitonscan store the incoming energy in the formof light and guide it at the nanoscale in aselective way.

These measurements have helped clar-ify the single-particle and many-body ef-fects leading to the exciton bleach and totrack the ultrafast intra-band electron dy-namics.[29] So far, they had also been elu-sive to experimental probes and becomeaccessible only when the broadband probeis tuned to cover a broad UV spectral rangebetween 280 and 360 nm.

These first results demonstrate thepower and versatility of ultrafast 2DUVspectroscopy for the investigation of a vari-

Fig. 8. (a) Normalizedtransient absorp-tion (ΔA) spectraat room temperatureof a colloidal solutionof anatase TiO2 NPsat a fixed time-delayof 1 ps and for dif-ferent pump photonenergies (indicatedin the figure). Eachtrace is normalizedwith respect to theminimum of the mainfeature at 3.88 eV.For comparison, theblack trace shows theinverted steady-stateabsorption spectrum.(b) ΔA spectra (con-verted from transientreflectivity meas-urements) of roomtemperature anataseTiO2 single crystalsalong the a- andc-axis at a fixed timedelay of 1 ps. For thisexperiment, the pumpphoton energy is 4.40eV. Reproduced fromref. [51].


[41] R. Schweitzer-Stenner, J. Phys. Chem. B 2008,112, 10358.

[42] D. Eden, J. B. Matthew, J. J. Rosa, F. M.Richards, Proc. Natl. Acad. Sci. U.S.A. 1982,79, 815.

[43] G. Weber, F. J. W. Teale, Discuss. Faraday Soc.1959, 27, 134.

[44] K. J. Willis, A. G. Szabo, M. Zuker, J. M.Ridgeway, B. Alpert, Biochemistry 1990, 29,5270.

[45] R. M. Hochstrasser, D. K. Negus, Proc. Natl.Acad. Sci. U.S.A. 1984, 81, 4399.

[46] C. J. Suess, J. D. Hirst, N.A. Besley, J. Comput.Chem. 2017, DOI: 10.1002/jcc.24793.

[47] F. Messina, E. Pomarico, M. Silatani, E.Baranoff, M. Chergui, J. Phys. Chem. Lett.2015, 6, 4475.

[48] A. Sarkar, S. Chakravorti, Chem. Phys. Lett.1995, 235, 195.

[49] O. Bräm, F. Messina, A. M. El-Zohry, A.Cannizzo, M. Chergui, Chem. Phys. 2012, 393,51.

[50] M. Chergui, Acc. Chem. Res. 2015, 48, 801.[51] E. Baldini, L. Chiodo, A. Dominguez, M.

Palummo, S. Moser, M.Yazdi, G. Auböck, B. P.P. Mallett, H. Berger, A. Magrez, C. Bernhard,M. Grioni,A. Rubio,M. Chergui,Nat. Commun.2017, 8, 13.

[52] P. Zu, Z. K. Tang, G. K. L.Wong, M. Kawasaki,A. Ohtomo, H. Koinuma,Y. Segawa, Solid StateCommun. 1997, 103, 459.

[53] R. Newmann, R. M. Chrenko, Phys. Rev. 1959,114, 1507.

[54] C. Taranto, G. Sangiovanni, K. Held, M.Capone, A. Georges, A. Toschi, Phys. Rev. B2012, 85, 085124.

[55] A. E. Bragg, M. C. Cavanagh, B. J. Schwartz,Science 2008, 321, 1817.

[56] X. Chen, S. E. Bradforth, Annu. Rev. Phys.Chem. 2008, 59, 203.

[57] J. A. Kloepfer, V. H. Vilchiz, V. A. Lenchenkov,S. E. Bradforth, Chem. Phys. Lett. 1998, 298,120.

[58] A. Lübcke, F. Buchner, N. Heine, I. V. Hertel, I.V. Hertel, T. Schultz, Phys. Chem. Chem. Phys.2010, 12, 14629.

[59] Y. Tang, H. Shen, K. Sekiguchi, N. Kurahashi,T. Mizuno,Y.-I. Suzuki, T. Suzuki, Phys. Chem.Chem. Phys. 2010, 12, 3653.

[60] F. Messina, O. Bräm, A. Cannizzo, M. Chergui,Nature Comm. 2013, 4, 2119.

[61] M. Reinhard, G. Auböck, N. A. Besley, I. P.Clark, M. W. D. Hanson-Heine, R. Horvath,T. S. Murphy, T. J. Penfold, M. Towrie, M. W.George, M. Chergui, J. Am. Chem. Soc. 2017,under review.

[62] A. Picchiotti, V. I. Prokhorenko, R. J. D. Miller,Rev. Sci. Instrum. 2015, 86, 093105.

[17] M. Cho, ‘Two-Dimensional Optical Spectro-scopy’, CRC Press, 2009.

[18] A. Cannizzo, Phys. Chem. Chem. Phys. 2012,14, 11205.

[19] W. P. de Boeij, M. S. Pshenichnikov, D. A.Wiersma, Chem. Phys. Lett. 1995, 238, 1.

[20] L. P. DeFlores, R.A. Nicodemus,A. Tokmakoff,Opt. Lett. 2007, 32, 2966.

[21] B. A. West, A. M. Moran, J. Phys. Chem. Lett.2012, 3, 2575.

[22] C. Tseng, S. Matsika, T. C. Weinacht, Opt.Express 2009, 17, 18788.

[23] U. Selig, C.-F. Schleussner, M. Foerster, F.Langhojer, P. Nuernberger, T. Brixner, Opt.Lett. 2010, 35, 4178.

[24] N. Krebs, I. Pugliesi, J. Hauer, E. Riedle, New J.Phys. 2013, 15, 085016.

[25] R. Borrego Varillas, A. Oriana, L. Ganzer, A.Trifonov, I. Buchvarov, C. Manzoni, G. Cerullo,Opt. Express 2016, 24, 28491.

[26] G. Auböck, C. Consani, F. van Mourik, M.Chergui, Opt. Lett. 2012, 37, 2337.

[27] G.Auböck, C. Consani, R. Monni,A. Cannizzo,F. van Mourik, M. Chergui, Rev. Sci. Instrum.2012, 83, 093105.

[28] P. Baum, S. Lochbrunner, E. Riedle, Opt. Lett.2004, 29, 1686.

[29] E. Baldini, T. Palmieri, E. Pomarico, G.Auböck,M. Chergui, arXiv:1703.07818, submitted toNano Lett.

[30] A. Cannizzo, O. Bräm, G. Zgrablic, A.Tortschanoff, A. Ajdarzadeh Oskouei, F. vanMourik, M. Chergui, Opt. Lett. 2007, 32, 3555.

[31] C. Bonati, A. Cannizzo, D. Tonti, A.Tortschanoff, F. van Mourik, M. Chergui, Phys.Rev. B 2007, 76, 033304.

[32] J. R. Lakowicz, ‘Principles of FluorescenceSpectroscopy’, Kluwer Academic/Plenum,NewYork, 1999.

[33] J. A. Stevens, J. J. Link,Y. T. Kao, C. Zang, L. J.Wang, D. P. Zhong, J. Phys. Chem. B 2010, 114,1498.

[34] S. Schenkl, F. van Mourik, G. van der Zwan, S.Haacke, M. Chergui, Science 2005, 309, 917.

[35] J. Leonard, E. Portuondo-Campa, A. Cannizzo,F. van Mourik, G. van der Zwan, J. Tittor, S.Haacke, M. Chergui, Proc. Natl. Acad. Sci.U.S.A. 2009, 106, 7718.

[36] O. Bräm, C. Consani, A. Cannizzo, M. Chergui,J. Phys. Chem. B 2011, 115, 13723.

[37] C. Consani, O. Bräm, F. vanMourik,A. Cannizzo,M. Chergui, Chem. Phys. 2012, 396, 108.

[38] C. Consani, G. Auböck, O. Bräm, F. vanMourik, M. Chergui, J. Chem. Phys. 2014, 140,025103.

[39] C. Consani, G. Auböck, F. van Mourik, M.Chergui, Science 2013, 339, 1586.

[40] R. Monni, A. Al Haddad, F. van Mourik, G.Auböck, M. Chergui, Proc. Natl. Acad. Sci.U.S.A 2015, 112, 5602.

further into the shorter wavelength rangeopens the door to the exploitation of thestrong peptide absorptions in proteins.

AcknowledgementsWe acknowledge financial support

from the Swiss NSF via the NCCR:MUSTand contracts No. 206021_157773 and20020_153660 as well as the EuropeanResearch CouncilAdvanced Grants H2020ERCEA 695197 DYNAMOX.


[1] A. H. Zewail, J. Phys. Chem. A 2000, 104,5660.

[2] S. K. Pal, A. H. Zewail, Chem. Rev. 2004, 104,2099.

[3] Y. C. Cheng, G. R. Fleming, Annu. Rev. Phys.Chem. 2009, 60, 241.

[4] M. D. Fayer, Annu. Rev. Phys. Chem. 2009, 60,21.

[5] R. J. D. Miller, A. Paarmann, V. I. Prokhorenko,Acc. Chem. Res. 2009, 42, 1442.

[6] P. Hamm, M. T. Zanni, ‘Concepts and Methodsof 2D Infrared Spectroscopy’, CambridgeUniversity Press, 2011.

[7] M. Chergui, ‘Comprehensive Biophysics,Vol. 1, Biophysical Techniques for StructuralCharacterization of Macromolecules’, Eds E.H. Egelman, H. J. Dyson, Oxford, AcademicPress, 2012, pp. 398-424.

[8] E. Hendry, M. Koeberg, J. Pijpers, M. Bonn,Phys. Rev. B 2007, 75, 233202.

[9] H.-K. Nienhuys, V. Sundström, Appl. Phys.Lett. 2005, 87, 012101.

[10] J. B. Asbury, E. Hao, Y. Wang, H. N. Ghosh, T.Lian, J. Phys. Chem. B 2001, 105, 4545.

[11] Y. Tamaki, K. Hara, R. Katoh, M. Tachiya, A.Furube, J. Phys. Chem. C 2009, 113, 11741.

[12] M. Rasmusson, A. N. Tarnovsky, T. Pascher, V.Sundström, E. Åkesson, J. Phys. Chem. A 2002,106, 7090.

[13] A. N. Tarnovsky, M. Wall, M. Gustafsson, N.Lascoux, V. Sundström, E. Åkesson, J. Phys.Chem. A 2002, 106, 5999.

[14] G. Benko, J. Kallioinen, J. E. Korppi-Tommola,A. P. Yartsev, V. Sundström, J. Am. Chem. Soc.2002, 124, 489.

[15] R. R. Ernst, G. Bodenhausen, A. Wokaun,‘Principles of Nuclear Magnetic Resonance inOne and Two Dimensions’, Clarendon Press,Oxford University Press, Oxford, New York,1987.

[16] D. M. Jonas, Science 2003, 300, 1515.

Fig. 9. Femtosecond fluorescence of aqueous iodine. (a) Fluorescence of 1 M NaI dissolved in water upon 266nm excitation. The Raman signal fromwater was removed from the plot. (b) Normalized kinetic traces at different wavelengths with their representative fits (continuous lines), comparedwith the Raman signal from water at 293 nm, whose temporal width gives the IRF of the set-up. Reproduced from ref. [60].

http://dx.doi.org/10.1002/jcc.24793

http://dx.doi.org/10.1002/jcc.24793


*Correspondence: Prof. Dr. Y. BellouardEPFL STI/IMT, Galatea LabRue de la Maladière 71b, CP526CH-2002 Neuchâtel 2E-mail: [email protected]

Ultrafast Laser to Tailor MaterialProperties: An Enabling Tool in AdvancedThree-dimensional Micromanufacturing

Yves Bellouard*

Abstract: The progress made in ultrafast laser technology towards high repetition rate systems have openednew opportunities in micromanufacturing. Non-linear absorption phenomena triggered by femtosecond pulsesinteracting with transparent materials allowmaterial properties to be tailored locally and in three dimensions, withresolution beyond the diffraction limits and at rate compatible with fabrication process requirements. In this shortarticle, we illustrate the potential of this technology for manufacturing, and more specifically micro-engineering,with a few examples taken from our own research and beyond.

Keywords: Micro-engineering · Ultrafast laser

Introduction: Femtosecond LaserProcessing on Dielectrics

Tightly focused femtosecond laserpulses of ultrashort duration (10–13–10–15 s)applied to transparent substrates triggernon-linear absorption phenomena such asmultiphoton processes, ionizing the fo-cal volume, creating a plasma that oncerecombining with the material structurelead to a locally modified structure. Thislaser–matter interaction, complex and di-verse, differs from one material system toanother.[1–8]

Among the various dielectric sub-strates, fused silica combines an ensembleof unique characteristics that make it,when exposed to femtosecond laser radia-tion, a particularly interesting candidatefor multifunctional integration in mono-lithic substrates.

In these materials, one can locally in-crease the refractive index,[9,10] enhance theetching rate,[11] introduce sub-wavelengthpatterns[12] with associated form birefrin-gence,[13,14] create voids[15] or change thethermal properties of fused silica.[16] Byscanning the laser through the specimenvolume,[17] one can distribute, combineand organize these material modificationsto form complex patterns to be used for in-stance as waveguides or fluidic channels.

The basic working principle of the fem-tosecond laser direct-writing process is il-lustrated in Fig. 1.

Three-dimensional Micromachining

Using a two-step process that combineslaser exposure and chemcial etching, com-plex parts with three-dimensional (3D)geometries can be efficiently produced. Aswill be seen later on, the chemical step hasa beneficial effect on the mechanical prop-erties of the fabricated microstructures byenhancing the surface quality and reducingthe occurrence of possible micro-cracks.The steps toward micromachining of 3Dparts are further described below and illus-trated in Fig. 1:

Step 1: The material is selectively ex-posed by rasterizing a pattern according toa technique described in ref. [17]. As anillustration of the laser systems used inthe examples presented in this paper, weused mainly ytterbium-based lasers oper-ating around 1030 nm with typical pulseduration around 270-fs and repetition ratesranging from 400 to 800 kHz or above.Average power is typically in the range ofa few to hundreds of milliwatts with pulseenergies of a few tens of nano-Joules.Microscope objectives with moderate NA(typically around 0.4 to 0.6 NA) are oftenused. Writing speeds (that vary dependingon the type of pattern) range from a fewtens of microns per seconds to tens of mmper seconds. Higher writing speeds havebeen reported[18] and essentially depend

Fig. 1. Illustration of the processing steps for three-dimensional manufacturing using ultra-fastlasers,[17] illustrated here in the case of a micro-channel microfabrication. Step 1: The material islocally exposed to femtosecond laser irradiation, for instance, by locally moving the substrateunder the laser beam or, by scanning the laser beam on a static substrate. The inner structure ofthe material is modified, but no ablation is taking place. As the process is non-linear, the laserlight is locally absorbed only at the focal point. Modifications can therefore be three-dimensionaland can occur anywhere within the volume of the glass and under the surface in particular. Thenature of the structural changes induced by the laser can be diverse, ranging from continuouslymodified zones to self-organized patterns such as nanogratings. Furthermore, and as anotherconsequence of the non-linearity of the interaction, the modified zone can be smaller than thespot-size itself and therefore, beyond the diffraction limit. Step 2: To remove material, the laserexposed region is etched away for instance with hydrofluoric (HF) acid in the case of silica. Thissecond step is omitted when the laser modified zones are directly used for their properties (suchas in the case of waveguides).


for identifying micro-algae[35,36] (illustratedin Fig. 3). The main advantage compared toother technology is that it is single monolith,optically transparent for a broad optical spec-trum and compact.

Other examples of devices combiningwaveguides and channels can be found forinstance in refs [37,38].

Micromechanics

Femtosecond lasers provide a way tocarve out any shape from a glass materialas long as the acid can penetrate throughthe laser exposed regions. One can there-fore make fluidic structures as describedabove, but also flexures – a deformableelement used in precision mechanics, bydefining thin-beam-like structures that candeform elastically. Based on this principle,all-optical mechanical sensor[39] where theflexures and the waveguides are all part ofthe same material can be made. There, thestructure withstands stress in excess of 500MPa. Stress in excess of 2 GPa can even bereached in a single flexure.[40,41] The me-chanical performance is strongly affectedby the etching rate.

Controlled Stress State

The femtosecond laser–matter interac-tion in the non-ablative regime induces lo-

at 1550 nm, as reported by several authors.Complex optical devices can be formed bycarefully combiningwaveguides splitting orseparating them.[26,27,29]

Combining Laser-affected Zones ofDifferent Kinds to Form IntegratedDevices

An interesting aspect of this manufac-turing technique is the fact that the samelaser can produce more than one type ofmicrostructure having different functions.In particular, waveguides and microstruc-tures can be combined to form integrateddevices such as optomechanical and opto-fluidics devices. As the same laser is usedto produce all the features, such devicesare intrinsically accurate as each elementis produced at the same time.

Optofluidic Devices

To fabricate an optofluidic device, oneneeds to carefully control the etching processso that waveguides are preserved.[33,34]Usingthe technology platform introduced before,one can create a variety of small optofluidicsinstruments, for example where light propa-gating in waveguides is used to interrogatecells passing through a fluidic channel. Theexample below illustrates a fluidicmicrochip

on the strategy used to stir the beam or tomove the specimen under it.

Step 2: After laser-exposure, the partis etched in a low-concentration hydroflu-oric bath. Concentrations between 2.5%and 5% are typically used.With these con-centrations, the etching rate enhancementis 60–120 times faster[17] in the exposedregion compared to the pristine material.This may result in a total etching timethat can vary from a few tens of minutesto several hours, depending on the com-plexity of the patterns. Noticeably, the la-ser polarization has a strong effect on theetching efficiency.[19] Recent work[20] hasshown that KOH also provides an efficientand more selective means for structuringlaser-exposed patterns with the aspect ra-tio exceeding thousand in certain cases.However, the etching mechanisms are dif-ferent and lead to different surface mor-phologies. The two methods can eventu-ally be combined, taking advantage of thecharacteristics of both.[21]

The mechanism leading to an etchingrate enhancement is complex and not fullyunderstood, but seems to be mainly drivenby the localized densification and porosityintroduced in the material as well as thepresence of stress that is known to have aneffect on the etching rate.[22]

Fig. 2 illustrates a monolithic three-dimensional flexure cross-pivot[23] entirelymanufactured out of glass. Such flexureshave superior performances compared tomore traditional flexures like the notchhinge: the stiffness contrast between thepivot axis and the other degrees of free-dom is significantly higher and betterdefined, making it a true one-degree offreedom pivot. As the figure shows, thesecross-pivots can themselves be combinedin more complex structures, e.g. mechani-cal guidance like the Hoecken mechanismillustrated below. This example illustrateshow the process described above can beused to ‘print’ three-dimensional parts.

Waveguide Writing

Waveguide writing has been one of themost studied applications of femtosecondlaser processing of dielectrics and numer-ous examples can be found in the literature(a few references are provided for fused sil-ica as well as other glass.[9,10,24–31] The topichas recently been reviewed in ref. [31]. Thevolume locally hit by the femtosecond laserexperiences a slight increase of refractiveindex (highest reported Δn are in the or-der of ~0.5×10–3). The laser affected zone(LAZ) shape and size can be determined us-ing either a refractive index map techniqueor more recently a novel technique basedon Scanning Thermal Imaging.[32] Lossesin waveguides can be less than 0.4 dB/cm

Fig. 2. Scanning Electron Microscope image (top left) and microscope image (top right) of across-pivot hinge as part of a mechanism. The images show the three crossed beams and therounded corners at the location where the beams are connected to the main body. Note the highaspect ratio of the micromachining process. A monolithic glass Hoecken linear guidance (left) anda sequence of merged images to illustrate the function of the mechanism (right). The end point,i.e. where the linear motion is produced, is the cross in the upper part. The mechanism is activat-ed by moving the rectangular window (left side of the images) along a circle. The linearity is within±0.5 µm over a displacement of 600 µm. Figure taken from ref. [23].


fromthe lowest to thehighestpulseenergiesand for pulse energies below ~200 fs,[54]the first type of modification (Regime I)corresponds to a localized densification.This laser-induced densification mecha-nism appears to be different than densifi-cation induced by mechanical means as in-dicated by Raman observations.[32] Whilecommon features are observed such asthe narrowing and shift of the main band,femtosecond laser-induced densificationis characterized by an increase of the D2-peak amplitude, suggesting the formationof silica tetrahedral rings of lower order.

At higher pulse energy levels, one ob-serves intriguing self-organization phe-nomena consisting of nanoplanes formingparallel one to another,[12] the so-calledRegime II. Contrary to the lower pulseenergy regime, these nanostructures, dueto their periodicity, induce phenomena ofform birefringence (i.e. a periodic modu-lation of the refractive index[13,14]) that isalso characterized by a localized increaseof volume. As shown in ref. [55], thesenanoplanes contain porous materials intowhich the presence of molecular oxygen

elements. This define is completely trans-parent to light. The electrodes consist ofan indium-tin-oxide simply deposited onthe material.

Thanks to the three-dimensional capa-bilities of micro-manufacturing process,new types of actuating principles thatcould not be easily implemented at the mi-croscale are studied. Illustrations are forinstance dielectrophoretic actuators[46] thatbenefit from a three-dimensional electrodedesign that is able to generate a electro-static field gradient.

The Structure of Laser-affectedZones in Silica

The structural changes causing thechange of properties, like the increasedrefractive index or the localized enhancedetching rate have been analyzed using mul-tiple spectroscopy techniques (see for in-stance refs [26,47–50] including Raman[51]as well as with indirect methods.[52,53] Sofar, it is generally admitted that two maintypes of modifications are found. Starting

calized volume changes in the material. Asa consequence, these density changes gen-erate stress in the material that can be con-trolled both in intensity and direction.[42]By carefully arranging these stressedzones, one can create wave plates[43] or useit for deforming the material locally likefor instance for adjusting the position ofan element.[44]

Micro-actuators andOptomechanical Sensors

The same process can also be used formanufacturing devices like micro-actua-tors. An illustration of a comb-array linearactuator is illustrated in Fig. 4.[45] Here,the high-aspect ratio of the micromachin-ing technique is used to achieve large ac-tuation forces by creating a difference ofelectrical potential between parallel platesarranged in a comb. The gap being shorteron one side than on the other, a net forcepulling the structure in one direction isgenerated. The pull-back force is providedby the stored elastic energy in the guiding

Fig. 4. Left: Working principle for a comb-array electrostatic glass actuator. Right: Scanning electron microscope close-up views of the fabricatedcomb-array actuator (adapted from ref. [45], © American Institute of Physics).

100μm

Fig. 3. Left: Channel and waveguides written in the bulk of glass (adapted from ref. [34], © SPIE). Center: Examples of wavelets signals obtained forspecific algae species. Right: Images captured on the fly by a microscope camera overlooking the fluidic channel (center and right images takenfrom ref. [35], © Optical Society of America).


[30] P. Bado,A.A. Said, M. Dugan, T. Sosnowski, in‘NFOEC’, Vol. 2, Dallas (TX), 2002, pp. 1153.

[31] T. Meany, M. Gräfe, R. Heilmann, A. Perez-Leija, S. Gross, M. J. Steel, M. J. Withford, A.Szameit, Laser Photon. Rev. 2015, 9, 363.

[32] Y. Bellouard, E. Barthel, A. A. Said, M. Dugan,P. Bado, Opt. Express 2008, 16, 19520.

[33] Y. Bellouard, A. Said, M. Dugan, P. Bado, in‘Materials Research Society Symposium -Proceedings’, 2003, Vol. 782, pp. 63.

[34] A. A. Said, M. Dugan, P. Bado, Y. Bellouard,A. Scott, J. R. Mabesa Jr., in ‘Proceedings ofSPIE - The International Society for OpticalEngineering’, 2004, Vol. 5339, pp. 194.

[35] A. Schaap, Y. Bellouard, T. Rohrlack, Biomed.Opt. Express 2011, 2, 658.

[36] A. Schaap, T. Rohrlack, Y. Bellouard, J.Biophoton. 2012, 5, 661.

[37] R. W. J. Applegate, J. Squier, T. Vestad, J.Oakey, D. W. M. Marr, P. Bado, M. A. Dugan,A. A. Said, Lab Chip 2006, 6, 422.

[38] R. M. Vazquez, R. Osellame, D. Nolli, C.Dongre, H. van den Vlekkert, R. Ramponi, M.Pollnau, G. Cerullo, Lab Chip 2009, 9, 91.

[39] Y. Bellouard, A. Said, P. Bado, Opt. Express2005, 13, 6635.

[40] Y. Bellouard, Opt. Mater. Express 2011, 1, 816.[41] A. Champion, M. Beresna, P. Kazansky, Y.

Bellouard, Opt. Express 2013, 21, 24942.[42] C.-E. Athanasiou, Y. Bellouard, Micromachines

2015, 6, 1365.[43] B. McMillen, C. Athanasiou,Y. Bellouard, Opt.

Express 2016, 24, 27239.[44] Y. Bellouard, Opt. Express 2015, 23, 29258.[45] B. Lenssen, Y. Bellouard, Appl. Phys. Lett.

2012, 101, 103503.[46] T. Yang, Y. Bellouard, J. Micromech. Microeng.

2015, 25, 105009.[47] K. Hirao, K. Miura, J. Non-Crystal. Solids

1998, 239, 91.[48] L. Skuja, J. Non-Crystal. Solids 1998, 239, 16.[49] H.-B. Sun, S. Juodkazis, M. Watanabe, S.

Matsuo, H. Misawa, J. Nishii, J. Phys. Chem. B2000, 104, 3450.

[50] A. Zoubir, C. Rivero, R. Grodsky, K.Richardson, M. Richardson, T. Cardinal, M.Couzi, Phys. Rev. B 2006, 73, 224117.

[51] J. W. Chan, T. Huser, S. Risbud, D. M. Krol,Opt. Lett. 2001, 26, 1726.

[52] Y. Bellouard, T. Colomb, C. Depeursinge, M.Dugan, A. A. Said, P. Bado, Opt. Express 2006,14, 8360.

[53] A. Champion,Y. Bellouard,Opt. Mater. Express2012, 2, 789.

[54] C. Hnatovsky, R. S. Taylor, P. P. Rajeev, E.Simova, V. R. Bhardwaj, D. M. Rayner, P. B.Corkum, Appl. Phys. Lett. 2005, 87, 014104.

[55] J. Canning, M. Lancry, K. Cook, A. Weickman,F. Brisset, B. Poumellec, Opt. Mater. Express2011, 1, 998.

[56] M. Lancry, B. Poumellec, J. Canning, K. Cook,J.-C. Poulin, F. Brisset, Laser Photon. Rev.2013, 7, 953.

[57] Y. Bellouard, A. Champion, B. McMillen, S.Mukherjee, R. R. Thomson, C. Pépin, P. Gillet,Y. Cheng, Optica 2016, 3, 1285.

[58] V. R. Bhardwaj, E. Simova, P. P. Rajeev, C.Hnatovsky, R. S. Taylor, D. M. Rayner, P. B.Corkum, Phys. Rev. Lett. 2006, 96, 057404.

[59] Y. Liao, J. Ni, L. Qiao, M. Huang,Y. Bellouard,K. Sugioka,Y. Cheng, Optica 2015, 2, 329.

[60] A. Rudenko, J.-P. Colombier, T. E. Itina, Phys.Rev. B 2016, 93, 075427.

[61] C. B. Schaffer, J. F. García, E. Mazur, Appl.Phys. A: Mater. Sci. Proc. 2003, 76, 351.

[62] S. Eaton, H. Zhang, P. Herman, F. Yoshino, L.Shah, J. Bovatsek, A. Arai, Opt. Express 2005,13, 4708.

[63] Y. Bellouard, M.-O. Hongler, Opt. Express2011, 19, 6807.

[64] Y. Bellouard, SPIE Proceedings 2015, 9351,93510G, DOI: 10.1117/12.2086709.

AcknowledgementsPart of this publication has been pre-

viously reported in Y. Bellouard, SPIEProceedings 2015, 9351, 93510G, DOI:10.1117/12.2086709.[64]


[1] D. Du, X. Liu, G. Korn, J. Squier, G. Mourou,Appl. Phys. Lett. 1994, 64, 3071.

[2] B. C. Stuart, M. D. Feit, A. M. Rubenchik, B.W. Shore, M. D. Perry, Phys. Rev. Lett. 1995,74, 2248.

[3] M. D. Perry, B. C. Stuart, P. S. Banks, M. D.Feit, V. Yanovsky, A. M. Rubenchik, J. Appl.Phys. 1999, 85, 6803.

[4] A.-C. Tien, S. Backus, H. Kapteyn, M.Murnane, G. Mourou, Phys. Rev. Lett. 1999, 82,3883.

[5] S. S. Mao, F. Quéré, S. Guizard, X. Mao, R.E. Russo, G. Petite, P. Martin, Appl. Phys. A:Mater. Sci. Proc. 2004, 79, 1695.

[6] J. B. Ashcom, R. R. Gattass, C. B. Schaffer, E.Mazur, JOSA B 2006, 23, 2317.

[7] R. R. Gattass, E. Mazur, Nat. Photon. 2008, 2,219.

[8] M. Beresna, M. Gecevicius, P. G. Kazansky,Adv. Opt. Photon. 2014, 6, 293.

[9] K. M. Davis, K. Miura, N. Sugimoto, K. Hirao,Opt. Lett. 1996, 21, 1729.

[10] K. Miura, J. Qiu, H. Inouye, T. Mitsuyu, K.Hirao, Appl. Phys. Lett. 1997, 71, 3329.

[11] A. Marcinkevicius, S. Juodkazis, M. Watanabe,M. Miwa, S. Matsuo, H. Misawa, J. Nishii, Opt.Lett. 2001, 26, 277.

[12] Y. Shimotsuma, P. G. Kazansky, J. Qiu, K.Hirao, Phys. Rev. Lett. 2003, 91, 247405.

[13] E. Bricchi, J. D. Mills, P. G. Kazansky, B. G.Klappauf, J. J. Baumberg, Opt. Lett. 2002, 27,2200.

[14] E. Bricchi, B. G. Klappauf, P. G. Kazansky,Opt. Lett. 2004, 29, 119.

[15] E. N. Glezer, M. Milosavljevic, L. Huang, R. J.Finlay, T.-H. Her, J. P. Callan, E. Mazur, Opt.Lett. 1996, 21, 2023.

[16] Y. Bellouard, M. Dugan, A. A. Said, P. Bado,Appl. Phys. Lett. 2006, 89, 161911.

[17] Y. Bellouard, A. Said, M. Dugan, P. Bado, Opt.Express 2004, 12, 2120.

[18] M. Hermans, J. Laser Micro/Nanoeng. 2014, 9,126.

[19] C. Hnatovsky, R. S. Taylor, E. Simova, V. R.Bhardwaj, D. M. Rayner, P. B. Corkum, Opt.Lett. 2005, 30, 1867.

[20] S. Kiyama, S. Matsuo, S. Hashimoto, Y.Morihira, J. Phys. Chem. C 2009, 113, 11560.

[21] S. LoTurco, R. Osellame, R. Ramponi, K. C.Vishnubhatla, J. Micromech. Microeng. 2013,23, 085002.

[22] A. Agarwal, M. Tomozawa, J. Non-Cryst.Solids 1997, 209, 166.

[23] V. Tielen,Y. Bellouard,Micromachines 2014, 5,697.

[24] Y. Sikorski, A. A. Said, P. Bado, R. Maynard, C.Florea, K. A. Winick, Electron. Lett. 2000, 36,226.

[25] C. B. Schaffer, A. Brodeur, J. F. García, E.Mazur, Opt. Lett. 2001, 26, 93.

[26] A. M. Streltsov, N. F. Borrelli, Opt. Lett. 2001,26, 42.

[27] K. Minoshima, A. M. Kowalevicz, I. Hartl, E. P.Ippen, J. G. Fujimoto,Opt. Lett. 2001, 26, 1516.

[28] R. Osellame, S. Taccheo, G. Cerullo, M.Marangoni, D. Polli, R. Ramponi, P. Laporta, S.De Silvestri, Electron. Lett. 2002, 38, 964.

[29] R. R. Thomson, S. Campbell, I. J. Blewett, A.K. Kar, D. T. Reid, S. Shen, A. Jha, Appl. Phys.Lett. 2005, 87, 121102.

has been hypothesized[56] and recently ex-perimentally confirmed.[57] The exact for-mation mechanism of these nanostructuresis still actively debated and various modelshave been proposed (see among others refs[12,58–60]).

Interestingly, the pulse duration triggersor suppresses the occurrence of Regime I.Above 200 fs, this regime seems not to beobserved,[54] at least using classical focus-ing techniques. The evolution of these laser-modified zones and their behavior stronglydepends on the energy deposited in the ma-terial (also called net fluence or dose). Forinstance, one can observe a change from avolume reduction to an expansion of thelaser-affected zone while continuously in-creasing the pulse energy of the laser, fromregime I to regime II.[57] Noteworthy, thistransition from shrinkage to expansion isalso observed while continuously increas-ing the exposure dose in regime I.[57]

If the repetition rate of the laser isdriven to a high level (typically above~2–3 MHz for silica), one observes thebulk heating of the material[61,62] and theformation in particular of gas bubbles thatmay themselves self-organize by feed-back mechanisms.[63]

Here, we have briefly surveyed some ofthemorewell-known effects.Much remainsto be done to fully understand this peculiarlaser–matter interaction and to explore allits subtleties, but also to optimize it.

Conclusion and Future Prospects

Femtosecond lasers can be used to ef-ficiently tailor the material properties offused silica in order to create not only opti-cal functions (such as waveguides), but al-so mechanical (flexures) or fluidic (micro-channels) elements. These functions can becombined to form complex optofluidics oroptomechanical devices or a combinationof both in a single substrate of material.This dramatically reduces the fabricationcomplexity and opens new horizons forfurther integration at the micro-/nanoscale,in particular by offering the possibility tofabricate three-dimensional devices, whichis difficult using conventional fabricationmethods based on lithographic techniques.

The next step is to expand the taxonomyof functions that can be directly written inthe material. To do so, one could imagineusing femtosecond lasers to induce high-pressure polymorphic phases of the samematerial. For instance, in the case of SiO

2,

there are numerous existing polymorphicphases, stable at ambient temperature, thathave physical properties different fromtheir amorphous counterpart. Well-knownexamples are the quartz phases that exhibitpiezoelectricity while amorphous silicadoes not.

http://dx.doi.org/10.1117/12.2086709

http://dx.doi.org/10.1117/12.2086709

http://dx.doi.org/10.1117/12.2086709

http://dx.doi.org/10.1117/12.2086709


*Correspondence: Dr. C. J. MilneSwissFELPaul Scherrer InstituteCH-5232 Villigen-PSIE-mail: [email protected]://www.psi.ch

Opportunities for Chemistry at theSwissFEL X-ray Free Electron Laser

Christopher J. Milne*, Paul Beaud, Yunpei Deng, Christian Erny, Rolf Follath, Uwe Flechsig, ChristophP. Hauri, Gerhard Ingold, Pavle Juranic, Gregor Knopp, Henrik Lemke, Bill Pedrini, Peter Radi, andLuc Patthey

Abstract: X-ray techniques have long been applied to chemical research, ranging from powder diffraction toolsto analyse material structure to X-ray fluorescence measurements for sample composition.[1] The developmentof high-brightness, accelerator-based X-ray sources has allowed chemists to use similar techniques but on moredemanding samples and using more photon-hungry methods. X-ray Free Electron Lasers (XFELs)[2] are the latestin the development of these large-scale user facilities, opening up new avenues of research and the possibility ofmore advanced applications for a range of research. The SwissFEL XFEL project at the Paul Scherrer Institute willbegin user operation in the hard X-ray (2.1–12.4 keV) photon energy range in 2018 with soft X-ray (240–1930 eV)user operation to follow and here we will present the details of this project, it’s operating capabilities, and someaspects of the experimental stations that will be particularly attractive for chemistry research. SwissFEL is arevolutionary new machine that will complement and extend the time-resolved chemistry efforts in the Swissresearch community.

Keywords: SwissFEL · X-ray free electron laser

Introduction

Chemistry has long been a forward-looking field, with its researchers con-stantly developing new tools and pushingthe limits of current techniques to an-swer a broad range of research questions.Chemistry has rapidly adopted X-rayFree Electron Lasers (XFELs) as yet an-other research tool in their panoply oftechniques, taking full advantage of thesehighly-specialized sources of high energyphotons. XFEL techniques are often simi-lar to those developed at third-generationsynchrotron light sources, which includeX-ray spectroscopy, X-ray scattering, andX-ray imaging methods.[3] The fundamen-tal difference between the properties ofthe two facilities is that whereas a storagering produces a quasi-continuous beam ofX-rays for experiments, XFELs produceshort bursts of very intense X-rays at peri-odic intervals.A facility such as SwissFELoperates at a repetition rate of 100 Hz, pro-

ducing 50 fs duration X-ray pulses withover 1 mJ of energy per pulse (1012 pho-tons/pulse). These properties make XFELsideal for certain types of experiments, andin some cases have driven the developmentof entirely new techniques, such as non-linear X-ray techniques.[4] In the followingsections, we will introduce the details ofthe SwissFEL machine, beamlines and ex-perimental stations, concluding with someexamples of how the facility will be usedfor chemistry research.

Facility

The SwissFEL is located at the PaulScherrer Institute (Villigen, Switzerland),which is the home to other large-scale userfacilities including the Swiss Light Source,a third-generation synchrotron light source,SµS, the Swiss Muon Source, SINQ, theSwiss Spallation Neutron Source, andHIPA, the high-intensity proton accelera-tor. SwissFEL is located in the forest onthe East side of the Aare river, with thebuilding located partially underground.The building itself is just under 1 km long,starting from the injector and ending at theexperimental hall. The facility includeslarge areas for the experiments, includ-ing 692 m2 for the three soft X-ray Athosbeamlines and instruments, and three largehard X-ray hutches for three experimentalstations, which are located on the threeAramis beamlines (see Fig. 1 for a sche-matic layout of the experimental areas). Inaddition to the experimental hutches therewill be laboratory space for chemical andbiological sample preparation, and techni-

cal work areas for instrument assembly andtesting. Optical lasers will be available forboth Athos and Aramis experiments, withthe laser infrastructure installed directly inthe soft X-ray experimental hall and on thefirst floor above the hard X-ray experimen-tal hutches (see Fig. 1).

The SwissFEL accelerator scheme isshown in Fig. 2. The initial installation ofthe Aramis hard X-ray branch was com-pleted in 2017, and is scheduled to beginuser operation in 2018. It consists of theelectron gun and injector, followed bythree linear accelerator (linac) stages toaccelerate the electrons to their highest en-ergy of 5.8 GeV. In order to achieve the up-per target X-ray photon energy of 12.4 keV(1 Å) this requires a low-emittance elec-tron beam. After the electron accelerationstages the photons are generated by 12in-vacuum undulator modules with vari-able gap.[5,6] This variable gap allows thephoton energy to be easily changed, whichis important for X-ray spectroscopy exper-iments.[7] The soft X-ray Athos branch ofthe accelerator uses the first two Aramislinac stages to achieve an electron ener-gy of 3 GeV, and a fast electron kicker todirect the beam into the Athos undulators.Linac 3 can be used to accelerate or decel-erate the electrons, allowing it to cover thefull electron energy range (2.1–5.8 GeV)while maintaining a stable electron energyfor the Athos branch. The Athos undula-tor section allows full polarization controlover the X-rays while still allowing foreasy scanning of the photon energy, whilethe Aramis undulators produce horizontal-ly polarized photons, also tunable over awide energy range.

http://www.psi.ch


cover the full X-ray photon energy rangewith varying energy resolution. Aramis 1hasvertically deflectingharmonic rejectionmirrors following the monochromator toremove the higher-order X-ray harmonicswhen operating in the lower range of X-rayphoton energies (2–5 keV) where signifi-cant harmonic contamination can be espe-cially problematic for spectroscopic meas-urements.[12] The beamlines are equippedwith flexible Kirkpatrick-Baez focussingmirrors, allowing the X-ray focus to bechanged depending on the experimental re-quirements (1.5–100 µm). Because of theachromatic nature of these optics they willnot need to be adjusted when changing theX-ray photon energy. Future possible up-grades to the beamlines include compoundrefractive lenses for simpler focussing re-quirements and phase-retarders to be ableto control the hard X-ray polarization onthe sample.[13]

In addition to the X-ray optics thebeamlines also include several differentdiagnostics elements to report on the prop-erties of the X-ray beam. Depending ontheir application they can be destructive,precluding any other measurement fromtaking place with the XFEL beam, and on-line, providing information concurrent toan experiment for every X-ray pulse. Theinformation recorded for every pulse in-

Through the use of vertical mirrors bothof these modes share the same beam path,allowing all the downstream optics, diag-nostics and experimental positions to re-main unchanged when switching betweenmodes. When using monochromatic modethe X-ray beam direction will remain un-changed when scanning the X-ray energythrough the use of fixed-exit monochroma-tor designs. Eachmonochromator has threecrystal pairs mounted, which will includeSi(111), Si(311) and InSb(111), which will

SwissFEL will operateAramis initiallyusing twoelectronchargemodeswhichwillallow users to choose between high per-pulse X-ray flux (200 pC, 50 fs FWHM) orshort pulse duration (10 pC, 1 fs FWHM).In the future several more advanced opera-tion modes are anticipated including broadX-ray bandwidths (5–7%),[8] attosecondpulse durations (100 as), and two-pulseoperation with tuneable pulse energies andtime delays.[9] The Athos branch has sim-ilar advanced capabilities, but with someadditional possibilities introduced by theinstallation of magnetic chicanes betweenthe undulator modules, the so-called CHICmode of operation.[10,11] In final operationcapacity SwissFEL will be able to operateboth Athos and Aramis simultaneously byproducing and accelerating two electronbunches separated by 28 ns, then usinga fast electron kicker to direct one of thebunches into the Athos accelerator branch.The SwissFEL design parameters are giv-en in Table 1.

Beamlines

The hard X-ray branch of SwissFELconsists of three planned beamlines, eachdelivering the FEL radiation into three ex-perimental hutches.[12] The initial phase ofoperation will cover the first two experi-mental hutches, with the third planned forinstallation in the coming years. The op-tics scheme for the beamlines is shown inFig. 3.

The beamlines are physically separatedby the use of horizontal offset mirrors thatdirect the XFEL beam either to the left orthe right, allowing for three different beamtrajectories. Each beamline will be capableof operating in either so-called pink beammode, with the full XFEL X-ray spectrumdelivered to the experiment (0.05–7%), ormonochromaticmode where only a narrowX-ray bandwidth is used (e.g. 0.015%).

Table 1. SwissFEL design parameters covering both the soft and hard X-raybranches. Values in red are advanced operation modes.

Description Range

Wavelength 1 Å to 70 Å

Photon energy 0.24 to 12.4 keV

Photons/pulse @ 1Å 1011

Pulse duration (FWHM) 1 to 50 fs (100 as)

Spectral bandwidth 0.05–0.16% (5–7%)

Maximum electron energy 5.8 GeV

Electron bunch charge 10–200 pC

Repetition rate 100 Hz

Pump laser infrastructureabove ESA X-ray hutch

ESA ESBESC

(Phase II)

Pump laserinfrastructure

Aramis Experimental Area:• 3 hutches (522.6 m2)• Hard X-rays 2.1Û12.4 keV (0.1Û 0.7nm)• 1st Pilot Experiment Q4 2017• Full operation (5.8 GeV) Q2 2018

Athos Experimental Area:• 1 single hutch 692 m2

• Soft X-rays 0.24Û 1.93 keV (0.65Û 5nm)• Phase II: 2017-2020

Fig. 1. SwissFEL experimental hutch layout. The hard X-ray experimental stations ESA and ESBwill be ready for user operation in 2018, with ESC and Athos coming online in the followingperiod. The X-rays come from the accelerator tunnel, which is to the left on the above layout,and move through the above building layout from left to right.

2nd construction phase2017-20

Linac 3Linac 1Injector Linac 2

Athos 240-1930 eV

Aramis 2.1-12.4 keV0.35 GeV 2.1 GeV 3.0 GeV 2.1-5.8 GeV

userstations2.7-3.3 GeVBC1 BC2

1st construction phase2013-2016

Fig. 2. SwissFEL accelerator design. The hard X-ray Aramis branch will begin user operation in2018 with the soft X-ray Athos branch installation to begin in 2017. Legend: Linac = linearaccelerator, BC = bunch compressor.


imental stations have similar beamlinecapabilities, as illustrated in the previoussection, but the instruments located at theexperimental stations are focussed on em-phasizing different X-ray techniques. Theconceptual design reports for all instru-ments presented in this section are availa-ble on the SwissFEL website.[23]

Experimental Station Alvra[24]

The layout of the ESA experimentalhutch is shown in Fig. 4. ESA is focussedprimarily on two techniques: X-ray spec-troscopy[7,25] and Serial FemtosecondCrystallography (SFX).[26,27] X-ray spec-troscopy involves measuring the X-raytransmission or X-ray fluorescence of asample as a function of monochromaticX-ray energy.[3] These measurements canprovide information on the local electronicand geometric structure around the absorb-ing species.[25] SwissFEL will be particu-larly suited for these types of experimentsdue to its variable-gap undulators[28,29]which can easily scan the X-ray energy ofthe XFEL over a very wide range, allowingtechniques such as X-ray absorption near-edge structure (XANES) and ExtendedX-ray absorption fine structure (EXAFS)to be used.[25] SFX is a technique that hasbeen developed for protein crystallographywhere a stream of small crystals is deliv-ered into the focussed X-ray beam using arange of different injector techniques, andthe diffraction pattern from each crystal isrecorded on a large two-dimensional de-tector.[30] Though the intense XFEL pulsedestroys the crystal, because an ultrashortX-ray pulse is used the diffraction occursbefore the atoms have time to move fromtheir lattice positions.[31,32] SFX can re-solve room-temperature protein structuresto better than 2 Å resolution on very smallcrystals,[26,30] which expands the techniqueto include samples that are difficult to crys-tallize, such asmembrane proteins.[33]ESAhas the additional capability of performingX-ray emission spectroscopy (XES) whichuses X-ray diffraction from an analysercrystal to measure the scattered or fluo-rescence X-ray photons with high energyresolution.[34] ESA uses short focal lengthcrystals (25 cm) in a dispersive von Hamosgeometry[35] to measure a range of XESenergies in a single measurement. Thisspectrometer can also be used for a varietyof other scattering measurements includeoff-resonant techniques[36,37] and inelasticX-ray scattering (IXS).

These techniques will be applied at twoinstruments, which are located in line withthe X-ray beam: ESA Prime and ESA Flex(see Fig. 4). Due to the flexible KBmirrorsthe X-rays can be focussed at either instru-ment, with the minimum focus of 1.5 µmachieved at ESA Prime. Both instruments

delivered with uncompressed pulses downto the experimental hutch where it is splitinto two branches with separate pulse dura-tion compressors: one for the X-ray timingtool diagnostic[18,19] and the other for theexperiment. For photochemistry and pho-tobiology the samples will often requireexcitation in the UV to IR range, which iscovered by a high-power Topas optical par-ametric amplifier (OPA). The anticipatedpulse energies delivered to the experimentare shown in Table 2. In addition to the UVto IR wavelengths the laser is capable ofgenerating a range of additional frequen-cies for sample excitation, including THzgeneration.[22]The laser can provide pulsesat frequencies ranging from the maximumrepetition rate of SwissFEL (100 Hz) downto single pulses.

Experimental Stations

SwissFEL will operate initially withtwo experimental stations. Experimentalstation Alvra (ESA) is focussed on pho-tochemistry and photobiology whileExperimental station Bernina (ESB) oncondensed matter physics. Both exper-

cludes the X-ray spectrum,[14,15] the X-raypulse energy and the X-ray beam posi-tion.[16,17] When pump-probe experimentsare performed a specialized timing diag-nostic is used to provide information onthe relative time-delay between the opticallaser excitation pulse and the probe X-raypulse.[18,19] This information allows the ex-periment to achieve the best time resolu-tion possible and corrects for the varioustiming instabilities that can affect thesetypes of experiments.[20] The goal of thetiming diagnostic is to provide 10 fs timeresolution for the experiments.

Laser

In order to fully take advantage of thefemtosecond time resolution the XFEL iscapable of providing, each experimentalstation has access to a highly reliable com-mercial laser system. These laser systemsare based on Ti:Sa technology and provide20 mJ, 35 fs pulses at 100 Hz.[21] Theselasers are located in a climate controlledroom on the first floor above the ESAX-ray hutch (see Fig. 1). The laser is then

Table 2. Laser pulse energies delivered to the experiment assuming 8 mJ and 30 fs pump laservalues delivered to the OPA. Numbers are extracted from ref. [21].

Wavelength Pulse energy Pulse duration (FWHM)

240–295 nm >26 µJ at peak <90 fs

290–480 nm >40 µJ at peak 36–60 fs

475–533 nm >466 µJ at peak 30–45 fs

533–600 nm >306 µJ at peak 30–45 fs

600–1160 nm >320 µJ at peak 30–45 fs

1160–2600 nm >2000 µJ at peak 36–45 fs

ESC

ESA

ESB

CRL

154.7142.34125.69123.29 152.364.5 84.5 9866.5 76.5 95.8 105.0 107.9 138.9

Distance from end of undulator (m)

124.19109.292.5 139.8

OM

M11 M31 M12 M32 M21 M22M23 M24

M14M13

M33 M34

4 mrad

6 mrad 6 mrad

4 mrad

6 mrad

X-ray delay line

153.2121.2 136.8

M15

M16

2 – 32 mrad

91.5 104 120 135.744.5

AU8AU4

AU4

AU4

AU4

CRL

8-12 mradDCM-1

DCM-2

ARAMIS-1

ARAMIS-2

ARAMIS-3

KB-1

KB-28-12 mrad

HRM

pink

C

monochromatic

D

pink

A Bmonochromatic

M13

M14

M13

M14

M22

M21

M22

M21

Fig. 3. Layout of the Aramis hard X-ray beamlines. Aramis-1 (ESA) and Aramis-2 (ESB) will beready for users in 2018, while Aramis 3 (ESC) will be installed in the following period. Oval insetsindicate the two possible beam paths for Aramis-1 (A, B) and Aramis-2 (C, D). Legend: M = mirror,DCM = double-crystal monochromator, KB = Kirkpatrick-Baez focussing mirrors, OM = offsetmirror, CRL = compound refractive lenses, pink = broad X-ray spectrum. The X axis coordinatesrepresent the distance of the optical elements from the end of the undulators.


Bragg angles (>85°) and segmented X-raycrystals[35] this spectrometer is capable of100 meV energy resolution. Both instru-ments are shown in Fig. 6.

Experimental Station BerninaThe Aramis 2 beamline is very similar

in scope to Aramis 1, making the X-rayproperties available at ESB very similar tothose available at ESA. The primary dif-ference is longer focal length KB mirrors,providing a larger minimum X-ray focusof 2.5 µm.[12] ESB is focussed primarily onsolid state physics experiments,[47] build-ing upon the expertise acquired[48] at theFEMTO laser-electron slicing source in-stalled at the Swiss Light Source.[49,50] Theprimary tool for this will be a six-circlehard X-ray Kappa-diffractometer (XRD)with a 1.5M Jungfrau 2D detector mount-ed on the detector arm, which is capable ofboth resonant and non-resonant diffractionexperiments. In addition to this instrumentthe experimental station will have a gener-al-purpose instrument (GPS) upon whichusers can mount or build their experimen-tal setup. These instruments are mountedon a rail system transverse to the beam toallow them to be moved in and out of po-sition, while maintaining the same X-rayfocus position. Both instruments will haveaccess to the optical laser for pump-probeexperiments and a large-area 2D Jungfraudetector, which is mounted on a robot armon theX-ray hutch ceiling to allow for flex-ible positioning. A specific focus for theseinstruments is the ability to use strong THzfields for sample excitation,[21,22] which al-low for direct excitation of lattice dynam-ics in the condensed phase.[51] The layoutof the ESB hutch is shown in Fig. 7.

In addition to the two instruments de-scribed above a third instrument will be inuse at ESB. This instrument, named ESB-MX, is designed to perform protein crys-tallography measurements using sampleson a solid support.[52] This has the advan-tage of increasing the hit rate of the pro-tein crystals, which is a restriction of theliquid jet delivery methods. ESB-MX willconsist of a chamber filled with He to re-duce X-ray background scatter and high-speed translation stages that are capableof positioning the samples to within 1 µmaccuracy at 100 Hz. The X-ray crystallog-raphy measurements will be performedwith the Jungfrau 16M detector on theceiling-mounted robot arm. The instrumentwill include a robot for automated sampleexchange, allowing for very high through-put measurements. This allows for twomethods of operation: 1) pre-location ofthe protein crystals on the solid support,[53]allowing pre-programming of the supporttrajectory upon insertion at ESB-MX, or 2)raster scanning of the support to efficiently

is below 1.5 Å and the X-ray spectrome-ter will be capable of measuring the fullphoton energy range of SwissFEL (2.1–12.4 keV) resonantly, with non-resonantmeasurements below 1.5 keV. This energyrange is shown in Fig. 5 with the elementslabelled and the specialized analyser crys-tals shown.

ESA Flex is a flexible instrument thatallows users to build up the experimentas required. It is mounted on a motorizedtable, allowing user-supplied chambersto be installed for the measurement. ESAFlex also includes a configurable X-rayspectrometer that can be mounted in avariety of positions to measure a range ofscattering angles, in both vertical and hori-zontal geometries. When used with large

can be usedwith the optical laser for pump-probe experiments, with an anticipatedtime resolution of better than 50 fs.[19] ESAPrime has a chamber that can be operatedunder vacuum, He, or neutral atmosphereand combines a large 2D Jungfrau scatter-ing detector[38–40] and a dual-crystal from aHamos X-ray emission spectrometer. Thisallows experiments to be performed usingboth scattering and emission techniquessimultaneously, which has proven to be apowerful combination for molecular[41,42]and protein[43,44] samples. The chamberhas the possibility of using different typesof sample injectors, including several spe-cifically for SFX sample delivery.[45,46]The expected achievable resolution of thecrystallography measurements at 12.4 keV

ESAPrime

ESAFlex

PhotonDiagnostics

KBmirrors

Fig. 4. Hutch layout for Experimental station A showing the location of the various elementswithin the hutch. The X-rays go through the hutch from left to right. Figure courtesy of DominiqueHauenstein.

1000 2000 3000 4000 500030

40

50

60

70

80

ADP(101)

TlAP(002)

PET(002)

InSb(111)

SiO2(10310)

Si(111)SiO2(11320)

SiO2(10312)

Si (220)

Energy (eV)

Braggangle(deg)

Energy

coverage

Na Mg Al Si P S Cl K Ca Sc TiK3edges

ZnGaGe

AsSe Br Kr Rb Sr Y

L3edgesInCdAgPdRhRuTcMoNbZr CsXeITeSbSn

Ar

TlAP(002)ADP(101)

PET(002)InSb(111)

SiO2(10310)Si(111)

SiO2(11320)SiO2(10312) Si (220)

Fig. 5. The X-ray energy range covered by the ESA Prime X-ray spectrometer. The dashed linesare the various crystals available for use with the spectrometer, covering the full range of energiesat the range of Bragg angles available in the spectrometer. The energies of various X-ray emissiontransitions are also marked to provide a sense as to what elements can be measured in this pho-ton energy range. Figure courtesy of Jakub Szlachetko.


measure randomly located crystals. In addi-tion to techniques specific to solid supportmethods, ESB-MX will also be capable ofmore conventional protein crystallographymethods, including goniometer rotationscans with large protein crystals.[54,55] Fig.8 shows how the temporary installation ofESB-MXwill look in the ESBX-ray hutch.

Chemistry at SwissFEL

Chemistry is by necessity a multi-tech-nique field of research since the samplesthemselves cover all forms of matter. Inthis section we will highlight some ex-amples of how XFELs have been used toinvestigate chemical problems to illustratehow the SwissFEL experimental stationswill be applied in the future.

AramisThe field of photochemistry has long

been investigated using time-resolvedtechniques. The ability to monitor energyrelaxation pathways after photoexcitationusing optical laser techniques has rep-resented the core of femtochemistry re-search.[56]This has recently been expandedto X-ray based structural techniques, in-cluding X-ray absorption spectroscopy[25]and X-ray scattering,[57,58] and X-ray tech-niques to probe electronic structural dy-namics, such as X-ray emission spectros-copy.[25] These techniques have often beencombined since they provide complemen-tary information on disordered systems,such as molecules or proteins in solution.

An example of how X-ray techniqueshave contributed to a better understandingof molecular photochemistry is in the fieldof spin-crossover molecular systems.[59,60]

The metal centres of these molecules (e.g.Fe2+, Fe3+, Co2+) can undergo a spin transi-tion through changes of temperature, pres-sure, or by absorption of light. Early opticalexperiments had reported on the relaxationpathways of a prototypical Fe2+-based spincrossover complex (iron trisbipyridine,FeII(bpy)

3) with a large spin-state gap that

could only be excited with visible light,[61]but though the timescales of the dynam-ics were well established, the characterof the electronic states involved were un-der debate. X-ray absorption experimentswere able to identify the structure of thequintet high-spin state,[62] and to establishthat this structural change took place morequickly than previously assumed possible(<150 fs).[63] Developments in UV tran-sient-absorption spectroscopy were able toconfirm this time-scale of population of thehigh-spin state and identify a vibrationalmode coupled with this singlet-to-quintetspin state transition.[64] The high spin statecharacter was also confirmed using X-rayemission spectroscopy with 100 ps timeresolution,[65] demonstrating this tech-nique’s sensitivity to the Fe spin state[66]but without the ability to achieve bettertime resolution to resolve the relaxationcascade. The open question was how theinitial excitation into a singlet metal-to-li-gand charge transfer state (1MLCT) couldthen populate a metal-centred quintet high-spin state (5T) so quickly. Several inter-mediate states were suggested, includingtriplet MLCT and metal-centred states,[67]but no conclusive explanation was forth-coming since no spin-sensitive experi-mental techniques were available with therequired time resolution. In an attempt tounderstand the solvent interaction withthese types of excited states,[68] time-re-solved experiments were performed com-bining both XES and wide-angle X-rayscattering (WAXS, also called X-ray dif-fuse scattering or XDS).[69] These resultsillustrated the complementary nature ofXES, which provides information on theelectronic state of the spin-crossover mole-cule, andWAXS, which provides informa-tion on both the structural changes in themolecule, but also on the solvent cage sur-rounding the molecule and the bulk solventdynamics in response to the photoexcita-tion. With the introduction of XFELs boththe XAS[70] and combined XES/WAXS[41]

experiments were repeated with these newsources of high-brightness femtosecondX-rays, though for technical reasons thetime resolution was limited to around 500fs, which restricted their ability to providenew information on the spin state relaxa-tion cascade. The primary new informationwas thatWAXS reported on ultrafast solva-tion changes around the excited moleculewhich corresponded to an increase in bulksolvent density on a timescale of 1 ps. The

Fig. 6. Models of the two instruments available at ESA. Left: ESA Prime, a chamber capable ofoperating under vacuum, He, or neutral atmosphere, which combines an X-ray emissionspectrometer and a large 2D Jungfrau scattering detector for photochemistry or photobiologyexperiments. Right: ESA Flex, a flexible station for user-mounted experiments with a configurablehigh-energy resolution X-ray spectrometer.

Fig. 7. Hutch layout for Experimental station B showing the location of the various elements withinthe hutch. The X-rays go through the hutch from left to right.


advantage of the development of viscoussample injectors,[45] which allow crystalsto be mixed into the delivery medium[84,85]

and then extruded from the injector intothe XFEL beam. In addition to room-tem-perature structural measurements, thistechnique has also been demonstrated fortime-resolved protein crystallography,[86,87]which can undoubtedly be extended to mo-lecular crystals as well. Preliminary exper-iments have already been demonstrated ona powder of a perovskite manganite in acryo-cooled liquid ethanol jet,[88] illustrat-ing that SFX need not be restricted solelyto protein crystals.

AthosWhile the focus has initially been on

the installation and commissioning ofthe hard X-ray branch of SwissFEL, theextension to the soft X-rays in the Athosbranch has also been in development.A so-phisticated undulator approach (CHIC[11])ensures a flexible photon source for exper-iments,[9,10] and a novel undulator designallows for easy control over polarizationand photon energy.[8] The definition of thethree beamlines and their experimental sta-tions is currently in progress in consulta-tion with the SwissFEL user community.Experiments performed at other soft X-rayFEL facilities around the world have estab-lished certainmethods as being specificallyrelevant to chemistry research, and will beconsidered for development at SwissFEL.Surface chemistry has proven a rich areafor soft X-ray FEL research, due to thehigh absorption cross-section for low ener-gy X-rays which allow them to be surfacesensitive. This has allowed experiments toinvestigate monolayers of molecules, suchas CO and atomic oxygen, on a metal sur-face and their response to photoexcitationof the metal.[89,90] These measurementsrepresent the first structurally-sensitiveexperiments able to probe transition statesin chemical reactions during catalytical-ly-active processes.[91] A second class ofexperiments has investigated photochem-ical processes of species in solution usingresonant inelastic X-ray scattering (RIXS)in the soft X-ray regime. This has allowedresearchers to follow themost fundamentalof chemical processes: bond-breaking andbond-making. By photo-dissociating a COligand off a parent complex (iron pentacar-bonyl, Fe(CO)

5) researchers were able to

follow the dissociation and immediate for-mation of a solvent-molecule complex, andby using simulations were able to suggestan electronic relaxation pathway consist-ent with their experimental results.[92,93] Athird class of experiments takes advantageof a combination of different probe tech-niques in a general-purpose experimentalchamber,[94] which is capable of electron/ion coincidencemeasurements,[95,96]Auger

structural changes.[76–78] ESA is designedexpressly for these kinds of experimentaltechniques: combining X-ray spectrosco-py with X-ray scattering with excellenttime resolution.We anticipate that as thesetechniques become more mature and theiranalysis more routine, combined with theintroduction of additional facilities capa-ble of these experiments,[79] photochemis-try on species in solution will undergo anexplosion of interest in the chemistry com-munity. The solution-phase experimentswill be complemented by the recent devel-opment of gas-phase molecular scatteringtechniques.[80,81]

Though a majority of hard X-raychemistry experiments performed to dateat XFELs have been as described above,there are several other opportunities atSwissFEL which can be envisioned. Oneaspect which has not been well-exploredhas been the ability to measure the photo-chemistry of molecular crystals. The mainrestriction to these experiments involvesthe sample damage inherent to XFELmeasurements and the inability to refreshthe solid sample easily. These problemshave been overcome at storage rings[82] andexperiments have demonstrated their po-tential at XFELs.[83] The ability to measurecrystal structures using large protein crys-tals andmoving to a new spot between FELpulses has been demonstrated so it is clearthat similar measurements on molecularcrystals at SwissFEL using an instrumentsuch as ESB-MX will be possible in thefuture.A second approach would be to take

conclusion drawn was that upon spin stateexcitation the structural change caused theexpulsion of two water molecules fromthe solvation shell around the molecule, inagreement with theoretical predictions.[71]The most recent X-ray measurement tocontribute to this discussion used XESwith 150 fs time resolution, courtesy ofthe development of a timing tool for X-ray-laser jitter correction at LCLS.[72] Thesemeasurements were compared with XESmeasurements on reference compoundsin various Fe spin- and oxidation states.The conclusions drawn were that at earlytimes (50 fs) the spin-state was not purelya quintet, but more consistent with a tripletligand-field excited state (3T). This resultis not the final conclusion since subsequentoptical results[73] with even better timeresolution indicate that at 50 fs the mol-ecule is in the high-spin state, essentiallyprecluding the ability of the molecule torelax through a ligand-field 3T unless thisprocess was occurring on a timescale of20 fs. This overview illustrates how thecombination of ultrafast optical and X-raytechniques have been used to address thequestion of ultrafast spin-state excitation,a question which has still not been fullyresolved and will require X-ray measure-ments with 20 fs time resolution or better.

Though here we have presented buta single research example these tech-niques have been applied to many dif-ferent fields of investigation, includingdonor-bridge-acceptor complexes,[42] in-tramolecular dynamics,[74,75] and protein

ESB-GPS

ESB-XRD

Fig. 8. ESB-MX installation in the ESB hutch positioned between the diffractometer (XRD) andgeneral-purpose (GPS) instruments on the hutch rail system. X-rays propagate from right to left.Inset: Details of the ESB-MX installation including the sample chamber and robot sample changer.Figure courtesy of Jan Hora and Pirmin Boehler.


M.Yabashi, R. Abela, Opt. Express 2014, 22,30004.

[19] I. Gorgisyan, R. Ischebeck, C. Erny, A. Dax,L. Patthey, C. Pradervand, L. Sala, C. J.Milne, H. T. Lemke, C. P. Hauri, T. Katayama,S. Owada, M. Yabashi, T. Togashi, R. Abela,L. Rivkin, P. Juranic, Opt Express 2017, 25,2080.

[20] M. Harmand, R. Coffee, M. R. Bionta, M.Chollet, D. French, D. Zhu, D. M. Fritz, H.T. Lemke, N. Medvedev, B. Ziaja, S. Toleikis,M. Cammarata, Nature Photon. 2013, 7, 215.

[21] C. Erny, C. P. Hauri, IUCr, J. SynchrotronRad. 2016, 23, 1143.

[22] C. Vicario, C. Ruchert, C. P. Hauri, J. ModernOptics 2015, 62, 1480.

[23] https://www.psi.ch/swissfel/publications[24] https://www.psi.ch/swissfel/aramis-experi-

mental-station-a[25] C. J. Milne, T. J. Penfold, M. Chergui, Coord.

Chem. Rev. 2014, 277, 44.[26] S. Boutet, L. Lomb, G. J. Williams, T. R. M.

Barends,A.Aquila, R. B. Doak, U.Weierstall,D. P. DePonte, J. Steinbrener, R. L. Shoeman,M. Messerschmidt, A. Barty, T. A. White, S.Kassemeyer, R. A. Kirian, M. M. Seibert, P.A. Montanez, C. Kenney, R. Herbst, P. Hart, J.Pines, G. Haller, S. M. Gruner, H. T. Philipp,M. W. Tate, M. Hromalik, L. J. Koerner,N. van Bakel, J. Morse, W. Ghonsalves,D. Arnlund, M. J. Bogan, C. Caleman, R.Fromme, C. Y. Hampton, M. S. Hunter, L. C.Johansson, G. Katona, C. Kupitz, M. Liang,A. V. Martin, K. Nass, L. Redecke, F. Stellato,N. Timneanu, D. Wang, N. A. Zatsepin, D.Schafer, J. Defever, R. Neutze, P. Fromme, J.C. H. Spence, H. N. Chapman, I. Schlichting,Science 2012, 337, 362.

[27] T. R. M. Barends, L. Foucar, A. Ardevol, K.Nass, A. Aquila, S. Botha, R. B. Doak, K.Falahati, E. Hartmann, M. Hilpert, M. Heinz,M. C. Hoffmann, J. Kofinger, J. E. Koglin,G. Kovacsova, M. Liang, D. Milathianaki,H. T. Lemke, J. Reinstein, C. M. Roome, R.L. Shoeman, G. J. Williams, I. Burghardt, G.Hummer, S. Boutet, I. Schlichting, Science2015, 350, 445.

[28] M. Calvi, M. Aiba, M. Brügger, S. Danner, R.Ganter, C. Ozkan, T. Schmidt, Proc. FEL20142014, 111.

[29] M. Calvi, M. Aiba, M. Brügger, S. Danner,T. Schmidt, R. Ganter, T. Schietinger, R.Ischebeck, Proc. FEL2014 2014, 107.

[30] J. C. H. Spence, U.Weierstall, H. N. Chapman,Rep. Prog. Phys. 2012, 75, 102601.

[31] A. Barty, C. Caleman, A. Aquila, N.Timneanu, L. Lomb, T. A. White, J.Andreasson, D. Arnlund, S. Bajt, T. R. M.Barends, M. Barthelmess, M. J. Bogan, C.Bostedt, J. D. Bozek, R. Coffee, N. Coppola,J. Davidsson, D. P. Deponte, R. B. Doak,T. Ekeberg, V. Elser, S. W. Epp, B. Erk,H. Fleckenstein, L. Foucar, P. Fromme, H.Graafsma, L. Gumprecht, J. Hajdu, C. Y.Hampton, R. Hartmann, A. Hartmann, G.Hauser, H. Hirsemann, P. Holl, M. S. Hunter,L. Johansson, S. Kassemeyer, N. Kimmel,R. A. Kirian, M. Liang, F. R. N. C. Maia, E.Malmerberg, S. Marchesini, A. V. Martin,K. Nass, R. Neutze, C. Reich, D. Rolles, B.Rudek, A. Rudenko, H. Scott, I. Schlichting,J. Schulz, M. M. Seibert, R. L. Shoeman,R. G. Sierra, H. Soltau, J. C. H. Spence, F.Stellato, S. Stern, L. Strüder, J. Ullrich, X.Wang, G. Weidenspointner, U. Weierstall, C.B.Wunderer, H. N. Chapman, Nature Photon.2011, 6, 35.

[32] H. N. Chapman, C. Caleman, N. Timneanu,Phil. Trans. Royal Soc. B: Biol. Sci. 2014,369, 20130313.

[33] R. Neutze, G. Brändén, G. F. X. Schertler,Curr. Opin. Struc. Biol. 2015, 33, 115.

from everyone working on the SwissFELproject at PSI. CJM acknowledges supportby the Swiss NSF through the NCCR-MUST and the contributions of JakubSzlachetko and Julien Réhault to the de-sign and concept behind the ESA instru-ments.


[1] R. Jenkins, in ‘Encyclopedia of AnalyticalChemistry’, John Wiley & Sons, Ltd, 2000.

[2] C. Pellegrini, A. Marinelli, S. Reiche, Rev.Mod. Phys. 2016, 88, 015006.

[3] P. Wilmott, ‘An Introduction to SynchrotronRadiation:Techniques and Applications’,Wiley, 2011.

[4] S. Tanaka, V. Chernyak, S. Mukamel, Phys.Rev. A 2001, 63, 63405.

[5] B. D. Patterson, R. Abela, H.-H. Braun, U.Flechsig, R. Ganter, Y. Kim, E. Kirk, A.Oppelt, M. Pedrozzi, S. Reiche, L. Rivkin,T. Schmidt, B. Schmitt, V. N. Strocov, S.Tsujino, A. F.Wrulich, New J. Phys. 2010, 12,035012.

[6] R. Ganter, M. Aiba, H.-H. Braun, M. Calvi, P.Heimgartner, H. Joehri, G. Janzi, R. Kobler,F. Loehl, M. Negrazus, E. Prat, L. Patthey, S.Reiche, S. Sanfilippo, U. Schaer, T. Schmidt,L. Schulz, V. Vrankovic, J. Wickstroem, Proc.FEL2013 2013, 263.

[7] W. Gawelda, J. Szlachetko, C. J. Milne, in‘X-Ray Absorption and X-Ray EmissionSpectroscopy’, John Wiley & Sons, Ltd,Chichester, UK, 2016, p 637.

[8] E. Prat, M. Calvi, S. Reiche, J. SynchrotronRad. 2016, 23, 874.

[9] S. Reiche, E. Prat, J. Synchrotron Rad. 2016,23, 869.

[10] E. Prat, S. Reiche, Phys. Rev. Lett. 2015, 114,244801.

[11] E. Prat, M. Calvi, R. Ganter, S. Reiche, T.Schietinger, T. Schmidt, J. Synchrotron Rad.2016, 23, 861.

[12] R. Follath, U. Flechsig, C. J. Milne, J.Szlachetko, G. Ingold, B. Patterson, L.Patthey, R. Abela, AIP Conf. Proc. 2016,1741, 020009.

[13] V. Scagnoli, C. Mazzoli, C. Detlefs, P.Bernard, A. Fondacaro, L. Paolasini, F.Fabrizi, F. De Bergevin, J. Synchrotron Rad.2009, 16, 778.

[14] P. Karvinen, S. Rutishauser, A. Mozzanica,D. Greiffenberg, P. N. Juranic, A. Menzel,A. Lutman, J. Krzywinski, D. M. Fritz, H. T.Lemke, M. Cammarata, C. David, Opt. Lett.2012, 37, 5073.

[15] M. Makita, P. Karvinen, D. Zhu, P. N. Juranic,J. Grünert, S. Cartier, J. H. Jungmann-Smith,H. T. Lemke, A. Mozzanica, S. Nelson, L.Patthey, M. Sikorski, S. Song, Y. Feng, C.David, Optica 2015, 2, 912.

[16] K. Tono, T. Kudo, M. Yabashi, T. Tachibana,Y. Feng, D. Fritz, J. Hastings, T. Ishikawa,Rev. Sci. Instrum. 2011, 82, 023108.

[17] K. Tiedtke, A. A. Sorokin, U. Jastrow, P.Juranic, S. Kreis, N. Gerken, M. Richter,U. Arp, Y. Feng, D. Nordlund, R. Soufli, M.Fernandez-Perea, L. Juha, P. A. Heimann, B.Nagler, H. J. Lee, S. Mack, M. Cammarata, O.Krupin, M. Messerschmidt, M. Holmes, M.Rowen, W. Schlotter, S. Moeller, J. J. Turner,Opt. Express 2014, 22, 21214.

[18] P. N. Juranic, A. Stepanov, R. Ischebeck,V. Schlott, C. Pradervand, L. Patthey, M.Radovic, I. Gorgisyan, L. Rivkin, C. P. Hauri,B. Monoszlai, R. Ivanov, P. Peier, J. Liu, T.Togashi, S. Owada, K. Ogawa, T. Katayama,

electron measurements,[97] and X-ray scat-tering measurements.[98–100] When used incombination with a variety of sample in-jectors, including molecular beams, clus-ters, liquid jets and single-particle injec-tors,[101] this type of instrument allows formany different kinds of experiments. Thedevelopment and design of soft X-ray in-struments for use at SwissFEL is currentlyongoing.

Conclusions

SwissFEL will begin user operationwith two hard X-ray experimental stationsin 2018, providing researchers with accessto high-intensity femtosecond X-ray puls-es that can be used for a broad range ofresearch activities. Experimental stationA provides users with instruments that aredesigned for X-ray spectroscopy and scat-tering experiments, including serial femto-second crystallography. In addition to thesegeneral features the ESA Prime instrumentwill also be able to investigate a spectros-copy regime currently not accessible at anyother XFEL: the tender X-ray energy rangefrom 2–5 keV. This spectral range coverselements such as S, P, Cl, and Ca, all ofwhich are biologically relevant, and the4d metals, which include Ag, Pd, Rh, andRu, many of which are important for catal-ysis chemistry. The ability to probe theseelements and others using resonant X-rayemission spectroscopy (RXES, also some-times called RIXS) allows researchers toobtain information on both the occupiedand unoccupied states of the absorbingspecies,[102,103]whether that be the HOMO/LUMO of a molecule or the valence/con-duction band of a semiconductor. This in-formation, when combined with the localstructure of an excited state, provides anenormous amount of information relevantto chemistry research. Experimental sta-tion B is focussed on condensed matter in-vestigations, with the ability to investigatethin films or bulk crystals under a rangeof different environments using resonantor non-resonant X-ray diffraction. Thesetechniques can be applied to systems suchas photoactive molecular crystals[82,83,104]or possibly photochemically-active nanos-tructured thin films.[105] With these experi-mental stations, and the development of fu-ture instruments at SwissFEL, we hope toenable chemists to perform ground-break-ing research at our facility. We are lookingforward to working closely with the Swisschemistry community in the coming years.

AcknowledgmentsThe authors would like to acknowl-

edge the ESB-MX project team, ThomasSchmidt, Romain Ganter, Rafael Abela,Gabriel Aeppli, and all the contributions

http://www.psi.ch/swissfel/publications

http://www.psi.ch/swissfel/aramis-experi-


[68] M.-E. Moret, I. Tavernelli, M. Chergui, U.Rothlisberger, Chem-Eur. J. 2010, 16, 5889.

[69] K. Haldrup, G. Vankó, W. Gawelda, A. Galler,G. Doumy, A. M. March, E. P. Kanter, A.Bordage, A. Dohn, T. B. van Driel, K. S.Kjær, H. T. Lemke, S. E. Canton, J. Uhlig,V. Sundstrom, L. Young, S. H. Southworth,M. M. Nielsen, C. Bressler, J. Phys. Chem. A2012, 116, 9878.

[70] H. T. Lemke, C. Bressler, L. X. Chen, D. M.Fritz, K. J. Gaffney, A. Galler, W. Gawelda,K. Haldrup, R. W. Hartsock, H. Ihee, J. Kim,K. H. Kim, J. H. Lee, M. M. Nielsen, A. B.Stickrath, W. Zhang, D. Zhu, M. Cammarata,J. Phys. Chem. A 2013, 117, 735.

[71] L. M. Lawson Daku, A. Hauser, J. Phys.Chem. Lett. 2010, 1, 1830.

[72] M. R. Bionta, H. T. Lemke, J. P. Cryan, J. M.Glownia, C. Bostedt, M. Cammarata, J. C.Castagna, Y. Ding, D. M. Fritz, A. R. Fry, J.Krzywinski, M. Messerschmidt, S. Schorb,M. Swiggers, R. Coffee, Opt. Express 2011,19, 21855.

[73] G. Aübock, M. Chergui, Nature Chem. 2015,7, 629.

[74] E. Biasin, T. B. van Driel, K. S. Kjaer, A. O.Dohn, M. Christensen, T. Harlang, P. Chabera,Y. Liu, J. Uhlig, M. Pápai, Z. Németh, R.Hartsock, W. Liang, J. Zhang, R. Alonso-Mori, M. Chollet, J. M. Glownia, S. Nelson,D. Sokaras, T. A. Assefa, A. Britz, A. Galler,W. Gawelda, C. Bressler, K. J. Gaffney, H.T. Lemke, K. B. Møller, M. M. Nielsen, V.Sundstrom, G. Vankó, K. Wärnmark, S. E.Canton, K. Haldrup, Phys. Rev. Lett. 2016,117, 013002.

[75] R. W. Hartsock, A. O. Dohn, T. Harlang, M.Chollet, M. Christensen, W. Gawelda, N. E.Henriksen, J. G. Kim, K. Haldrup, K. H. Kim,H. Ihee, J. Kim, H. T. Lemke, Z. Sun, V. S. O.m, W. Zhang, D. Zhu, T. B. van Driel, K. S.K. A. r, K. B. M. O. ller, M. M. Nielsen, K. J.Gaffney, Nature Commun. 2016, 7, 1.

[76] M. Levantino, G. Schiró, H. T. Lemke, G.Cottone, J. M. Glownia, D. Zhu, M. Chollet,H. Ihee, A. Cupane, M. Cammarata, NatureCommun. 2015, 6, page.

[77] M. Levantino, H. T. Lemke, G. Schiró, M.Glownia, A. Cupane, M. Cammarata, Struct.Dyn. 2015, 2, 041713.

[78] D. Arnlund, L. C. Johansson, C. Wickstrand,A. Barty, G. J. Williams, E. Malmerberg, J.Davidsson, D. Milathianaki, D. P. Deponte, R.L. Shoeman, D. Wang, D. James, G. Katona,S.Westenhoff, T. A.White, A.Aquila, S. Bari,P. Berntsen, M. Bogan, T. B. van Driel, R. B.Doak, K. S. Kjær, M. Frank, R. Fromme, I.Grotjohann, R. Henning, M. S. Hunter, R. A.Kirian, I. Kosheleva, C. Kupitz, M. Liang, A.V. Martin, M. M. Nielsen, M. Messerschmidt,M. M. Seibert, J. Sjohamn, F. Stellato, U.Weierstall, N. A. Zatsepin, J. C. H. Spence,P. Fromme, I. Schlichting, S. Boutet, G.Groenhof, H. N. Chapman, R. Neutze, Nat.Meth. 2014, 11, 923.

[79] In 2018 five facilities worldwide will haveinstruments capable of performing thesekinds of experiments: LCLS (Stanford, CA,USA), SACLA (Spring-8, Japan), PAL-FEL(Pohang, Korea), European XFEL (Hamburg,German), and SwissFEL (Villigen-PSI,Switzerland).

[80] M. P. Minitti, J. M. Budarz, A. Kirrander, J.S. Robinson, D. Ratner, T. J. Lane, D. Zhu,J. M. Glownia, M. Kozina, H. T. Lemke, M.Sikorski, Y. Feng, S. Nelson, K. Saita, B.Stankus, T. Northey, J. B. Hastings, P. M.Weber, Phys. Rev. Lett. 2015, 114, 255501.

[81] J. M. Budarz, M. P. Minitti, D. V. Cofer-Shabica, B. Stankus, A. Kirrander, J. B.Hastings, P. M. Weber, J. Phys. B: At. Mol.Opt. Phys. 2016, 49, 034001.

R. L. Shoeman, A. Barty, H. N. Chapman, R. A.Kirian, K. R. Beyerlein, R. C. Stevens, D. Li, S.T. A. Shah, N. Howe, M. Caffrey, V. Cherezov,Nature Commun. 2014, 5, 1.

[46] U. Weierstall, J. C. H. Spence, R. B. Doak,Rev. Sci. Instrum. 2012, 83, 035108.

[47] G. Ingold, J. Rittmann, P. Beaud, M. Divall,C. Erny, U. Flechsig, R. Follath, C. P. Hauri,S. Hunziker, P. Juranic, A. Mozzanica, B.Pedrini, L. Sala, L. Patthey, B. D. Patterson, R.Abela, AIP Conf. Proc. 2016, 1741, 030039.

[48] P. Beaud, S. L. Johnson, E. Vorobeva, C. J.Milne, A. Caviezel, S. O. Mariager, R. A. DeSouza, U. Staub, G. Ingold, Chimia 2011, 65,308.

[49] P. Beaud, S. L. Johnson, A. Streun, R. Abela,D. Abramsohn, D. Grolimund, F. Krasniqi,T. Schmidt, V. Schlott, G. Ingold, Phys. Rev.Lett. 2007, 99, 174801.

[50] G. Ingold, R. Abela, P. Beaud, S. L. Johnson,U. Staub, Z. Kristallogr. 2008, 223, 292.

[51] T. Kubacka, J. A. Johnson, M. C. Hoffmann,C. Vicario, S. de Jong, P. Beaud, S. Grübel,S. W. Huang, L. Huber, L. Patthey, Y. D.Chuang, J. J. Turner, G. L. Dakovski, W. S.Lee, M. P. Minitti, W. Schlotter, R. G. Moore,C. P. Hauri, S. M. Koohpayeh, V. Scagnoli, G.Ingold, S. L. Johnson, U. Staub, Science 2014,343, 1333.

[52] M. S. Hunter, B. Segelke, M. Messerschmidt,G. J.Williams, N.A. Zatsepin,A. Barty,W. H.Benner, D. B. Carlson, M. Coleman, A. Graf,S. P. Hau-Riege, T. Pardini, M. M. Seibert, J.Evans, S. Boutet, M. Frank, Sci. Rep. 2014, 4,6026.

[53] J. A. Wojdyla, E. Panepucci, I. Martiel, S.Ebner, C.-Y. Huang, M. Caffrey, O. Bunk, M.Wang, J. Appl. Crystallogr. 2016, 49, 944.

[54] K. Hirata, K. Shinzawa-Itoh, N. Yano,S. Takemura, K. Kato, M. Hatanaka, K.Muramoto, T. Kawahara, T. Tsukihara, E.Yamashita, K. Tono, G. Ueno, T. Hikima,H. Murakami, Y. Inubushi, M. Yabashi,T. Ishikawa, M. Yamamoto, T. Ogura, H.Sugimoto, J.-R. Shen, S. Yoshikawa, H. Ago,Nat. Meth. 2014, 1, 734.

[55] M. Suga, F. Akita, K. Hirata, G. Ueno, H.Murakami, Y. Nakajima, T. Shimizu, K.Yamashita, M. Yamamoto, H. Ago, J.-R.Shen, Nature 2015, 517, 99.

[56] A. H. Zewail, Science 1988, 242, 1645.[57] H. Ihee, Acc. Chem. Res. 2009, 42, 356,.[58] K. Haldrup, M. Christensen, M. M. Nielsen,

Acta Crystallogr. A 2010, 66, 261.[59] A. Hauser, Top. Curr. Chem. 2004, 234, 155.[60] A. Hauser, Top. Curr. Chem. 2004, 233, 49.[61] W. Gawelda, A. Cannizzo, V.-T. Pham, F.

van Mourik, C. Bressler, M. Chergui, J. Am.Chem. Soc. 2007, 129, 8199.

[62] W. Gawelda, V.-T. Pham, M. Benfatto, Y.Zaushitsyn, M. Kaiser, D. Grolimund, S. L.Johnson, R. Abela, A. Hauser, C. Bressler, M.Chergui, Phys. Rev. Lett. 2007, 98, 057401.

[63] C. Bressler, C. J. Milne, V. T. Pham, A. ElNahhas, R. M. van der Veen, W. Gawelda,S. L. Johnson, P. Beaud, D. Grolimund, M.Kaiser, C. N. Borca, G. Ingold, R. Abela, M.Chergui, Science 2009, 323, 489.

[64] C. Consani, M. Prémont-Schwarz, A. ElNahhas, C. Bressler, F. van Mourik, A.Cannizzo, M. Chergui, Angew. Chem. Int. Ed.2009, 48, 7184.

[65] G. Vankó, P. Glatzel, V.-T. Pham, R. Abela,D. Grolimund, C. N. Borca, S. L. Johnson, C.J. Milne, C. Bressler, Angew. Chem. Int. Ed.2010, 49, 5910.

[66] G. Vankó, T. Neisius, G. Molnar, F. Renz, S.Karpati, A. Shukla, F. M. F. de Groot, J. Phys.Chem. B 2006, 110, 11647.

[67] A. Cannizzo, C. J. Milne, C. Consani, W.Gawelda, C. Bressler, F. van Mourik, M.Chergui, Coord. Chem. Rev. 2010, 254, 2677.

[34] U. Bergmann, P. Glatzel, Photosynth. Res.2009, 102, 255.

[35] J. Szlachetko, M. Nachtegaal, E. de Boni, M.Willimann, O. Safonova, J. Sa, G. Smolentsev,M. Szlachetko, J. A. van Bokhoven, J.C. Dousse, J. Hoszowska, Y. Kayser, P.Jagodzinski, A. Bergamaschi, B. Schmitt, C.David, A. Lücke, Rev. Sci. Instrum. 2012, 83,103105.

[36] J. Szlachetko, C. J. Milne, J. Hoszowska, J.C. Dousse, W. Błachucki, J. Sa, Y. Kayser,M. Messerschmidt, R. Abela, S. Boutet,C. David, G. Williams, M. Pajek, B. D.Patterson, G. Smolentsev, J.A. van Bokhoven,M. Nachtegaal, Struct. Dyn. 2014, 1, 021101.

[37] J. Szlachetko, M. Nachtegaal, J. Sa, J.-C.Dousse, J. Hoszowska, E. Kleymenov, M.Janousch, O. V. Safonova, C. König, J. A. vanBokhoven, Chem. Commun. 2012, 48, 10898.

[38] J. H. Jungmann-Smith, A. Bergamaschi,M. Brückner, S. Cartier, R. Dinapoli, D.Greiffenberg, A. Jaggi, D. Maliakal, D.Mayilyan, K. Medjoubi, D. Mezza, A.Mozzanica, M. Ramilli, C. Ruder, L. Schädler,B. Schmitt, X. Shi, G. Tinti, Rev. Sci. Instrum.2015, 86, 123110.

[39] J. H. Jungmann-Smith, A. Bergamaschi,S. Cartier, R. Dinapoli, D. Greiffenberg,I. Johnson, D. Maliakal, D. Mezza, A.Mozzanica, C. Ruder, L. Schaedler, B.Schmitt, X. Shi, G. Tinti, J. Inst. 2014, 9,P12013.

[40] A. Mozzanica, A. Bergamaschi, S. Cartier,R. Dinapoli, D. Greiffenberg, I. Johnson, J.Jungmann, D. Maliakal, D. Mezza, C. Ruder,L. Schaedler, B. Schmitt, X. Shi, G. Tinti, J.Inst. 2014, 9, C05010.

[41] K. Haldrup, W. Gawelda, R. Abela, R.Alonso-Mori, U. Bergmann, A. Bordage, M.Cammarata, S. E. Canton, A. O. Dohn, T. B.van Driel, D. M. Fritz, A. Galler, P. Glatzel,T. Harlang, K. S. Kjaer, H. T. Lemke, K. B.Møller, Z. Németh, M. Pápai, N. Sas, J. Uhlig,D. Zhu, G. Vankó, V. Sundstrom, M. M.Nielsen, C. Bressler, J. Phys. Chem. B 2016,120, 1158.

[42] S. E. Canton, K. S. Kjaer, G. Vankó, T. B. vanDriel, S.-I. Adachi, A. Bordage, C. Bressler,P. Chabera, M. Christensen, A. O. Dohn, A.Galler, W. Gawelda, D. Gosztola, K. Haldrup,T. Harlang, Y. Liu, K. B. Møller, Z. Németh,S. Nozawa, M. Pápai, T. Sato, T. Sato, K.Suarez-Alcantara, T. Togashi, K. Tono, J.Uhlig, D. A. Vithanage, K. Wärnmark, M.Yabashi, J. Zhang, V. Sundstrom, M. M.Nielsen, Nature Commun. 2015, 6, 6359.

[43] J. Kern, R.Alonso-Mori, R. Tran, J. Hattne, R.J. Gildea, N. Echols, C. Glockner, J. Hellmich,H. Laksmono, R. G. Sierra, B. Lassalle-Kaiser, S. Koroidov, A. Lampe, G. Han, S.Gul, D. DiFiore, D. Milathianaki,A. R. Fry,A.Miahnahri, D.W. Schafer, M. Messerschmidt,M. M. Seibert, J. E. Koglin, D. Sokaras, T.C. Weng, J. Sellberg, M. J. Latimer, R. W.Grosse-Kunstleve, P. H. Zwart, W. E. White,P. Glatzel, P. D. Adams, M. J. Bogan, G. J.Williams, S. Boutet, J. Messinger, A. Zouni,N. K. Sauter, V. K. Yachandra, U. Bergmann,J. Yano, Science 2013, 340, 491.

[44] J. Kern, J. Hattne, R. Tran, R.Alonso-Mori, H.Laksmono, S. Gul, R. G. Sierra, J. Rehanek,A. Erko, R. Mitzner, P. Wernet, U. Bergmann,N. K. Sauter, V. Yachandra, J. Yano, Phil.Trans. Royal Soc. B: Biol. Sci. 2014, 369,20130590.

[45] U. Weierstall, D. James, C. Wang, T. A. White,D. Wang, W. Liu, J. C. H. Spence, R. B.Doak, G. Nelson, P. Fromme, R. Fromme, I.Grotjohann, C. Kupitz, N. A. Zatsepin, H. Liu,S. Basu, D. Wacker, G. W. Han, V. Katritch, S.E. B. Boutet, M.Messerschmidt, G. J.Williams,J. E. Koglin, M. M. Seibert, M. Klinker, C. Gati,


[98] T. Ekeberg, M. Svenda, C. Abergel, F. R. N. C.Maia, V. Seltzer, J.-M. Claverie, M. Hantke, O.Jönsson, C. Nettelblad, G. van der Schot, M.Liang, D. P. Deponte, A. Barty, M. M. Seibert,B. Iwan, I. Andersson, N. D. Loh, A. V. Martin,H. Chapman, C. Bostedt, J. D. Bozek, K. R.Ferguson, J. Krzywinski, S. W. Epp, D. Rolles,A. Rudenko, R. Hartmann, N. Kimmel, J.Hajdu, Phys. Rev. Lett. 2015, 114, 098102.

[99] G. van der Schot, M. Svenda, F. R. N. C.Maia, M. Hantke, D. P. Deponte, M. M.Seibert, A. Aquila, J. Schulz, R. Kirian, M.Liang, F. Stellato, B. Iwan, J. Andreasson,N. Timneanu, D. Westphal, F. N. Almeida,D. Odic, D. Hasse, G. H. Carlsson, D. S. D.Larsson, A. Barty, A. V. Martin, S. Schorb, C.Bostedt, J. D. Bozek, D. Rolles, A. Rudenko,S. Epp, L. Foucar, B. Rudek, R. Hartmann,N. Kimmel, P. Holl, L. Englert, N. T. DuaneLoh, H. N. Chapman, I. Andersson, J. Hajdu,T. Ekeberg, Nature Commun. 2015, 6, 5704.

[100] M. M. Seibert, T. Ekeberg, F. R. N. C. Maia,M. Svenda, J. Andreasson, O. Jönsson, 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. Messerschmidt, J. Krzywinski,G. Williams, K. O. Hodgson, M. J. Bogan,C. Y. Hampton, R. G. Sierra, D. Starodub, I.Andersson, 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. Foucar, N. Kimmel,P. Holl, B. Rudek, B. Erk, A. Hömke, C.Reich, D. Pietschner, G. Weidenspointner,L. Strüder, G. Hauser, H. Gorke, J. Ullrich,I. Schlichting, S. Herrmann, G. Schaller,F. Schopper, H. Soltau, K.-U. Kühnel, R.Andritschke, C.-D. Schröter, F. Krasniqi,M. Bott, S. Schorb, D. Rupp, M. Adolph, T.Gorkhover, H. Hirsemann, G. Potdevin, H.Graafsma, B. Nilsson, H. N. Chapman, J.Hajdu, Nature 2011, 469, 78.

[101] R. A. Kirian, S. Awel, N. Eckerskorn, H.Fleckenstein, M.Wiedorn, L.Adriano, S. Bajt,M. Barthelmess, R. Bean, K. R. Beyerlein, L.M. G. Chavas, M. Domaracky, M. Heymann,D. A. Horke, J. Knoska, M. Metz, A. Morgan,D. Oberthuer, N. Roth, T. Sato, P. L. Xavier,O. Yefanov, A. V. Rode, J. Küpper, H. N.Chapman, Struct. Dyn. 2015, 2, 041717.

[102] J. Szlachetko, M. Pichler, D. Pergolesi, J. Sa,T. Lippert, RSC Adv. 2014, 4, 11420.

[103] J. Szlachetko, J. Sa, CrystEngComm 2013, 15,2583.

[104] R. Bertoni, M. Lorenc, H. Cailleau, A.Tissot, J. Laisney, M.-L. Boillot, L. Stoleriu,A. Stancu, C. Enachescu, E. Collet, NatureMater. 2016, 15, 606.

[105] X. Chen, C. Li, M. Grätzel, R. Kostecki, S. S.Mao, Chem. Soc. Rev. 2012, 41, 7909.

Ogasawara, H. Ostrom, L. G. M. Pettersson,W. F. Schlotter, J. A. Sellberg, F. Sorgenfrei,J. J. Turner, M. Wolf, W. Wurth, A. Nilsson,Science 2013, 339, 1302.

[91] A. Nilsson, J. LaRue, H. Öberg, H.Ogasawara, M. Dell’Angela, M. Beye, H.Ostrom, J. Gladh, J. K. Nørskov, W. Wurth, F.Abild-Pedersen, L. G. M. Pettersson, Chem.Phys. Lett. 2017, 675, 145.

[92] P. Wernet, K. Kunnus, I. Josefsson, I.Rajkovic, W. Quevedo, M. Beye, S. Schreck,S. Grübel, M. Scholz, D. Nordlund,W. Zhang,R.W. Hartsock,W. F. Schlotter, J. J. Turner, B.Kennedy, F. Hennies, F. M. F. de Groot, K. J.Gaffney, S. Techert, M. Odelius, A. Föhlisch,Nature 2015, 520, 78.

[93] K. Kunnus, I. Josefsson, I. Rajkovic, S.Schreck, W. Quevedo, M. Beye, C. Weniger,S. Grübel, M. Scholz, D. Nordlund,W. Zhang,R.W. Hartsock, K. J. Gaffney, W. F. Schlotter,J. J. Turner, B. Kennedy, F. Hennies, F. M. F.de Groot, S. Techert, M. Odelius, P. Wernet,A. Föhlisch, Struct. Dyn. 2016, 3, 043204.

[94] L. Strüder, S. Epp, D. Rolles, R. Hartmann, P.Holl, G. Lutz, H. Soltau, R. Eckart, C. Reich,K. Heinzinger, C. Thamm, A. Rudenko, F.Krasniqi, K.-U. Kühnel, C. Bauer, C.-D.Schröter, R. Moshammer, S. Techert, D.Miessner, M. Porro, O. Hälker, N. Meidinger,N. Kimmel, R. Andritschke, F. Schopper, G.Weidenspointner, A. Ziegler, D. Pietschner,S. Herrmann, U. Pietsch, A. Walenta, W.Leitenberger, C. Bostedt, T. Möller, D. Rupp,M. Adolph, H. Graafsma, H. Hirsemann, K.Gärtner, R. Richter, L. Foucar, R. L. Shoeman,I. Schlichting, J. Ullrich, Nuclear Inst. Meth.Phys. Res. A 2010, 614, 483.

[95] C. E. Liekhus-Schmaltz, I. Tenney, T. Osipov,A. Sanchez-Gonzalez, N. Berrah, R. Boll, C.Bomme, C. Bostedt, J. D. Bozek, S. Carron,R. Coffee, J. Devin, B. Erk, K. R. Ferguson,R. W. Field, L. Foucar, L. J. Frasinski,J. M. Glownia, M. Gühr, A. Kamalov, J.Krzywinski, H. Li, J. P. Marangos, T. J.Martinez, B. K. McFarland, S. Miyabe, B.Murphy, A. Natan, D. Rolles, A. Rudenko,M. Siano, E. R. Simpson, L. Spector, M.Swiggers, D. Walke, S. Wang, T. Weber, P. H.Bucksbaum, V. S. Petrovic, Nature Commun.2015, 6, 8199.

[96] R. Boll, B. Erk, R. Coffee, S. Trippel, T.Kierspel, C. Bomme, J. D. Bozek, M. Burkett,S. Carron, K. R. Ferguson, L. Foucar, J.Küpper, T. Marchenko, C. Miron, M. Patanen,T. Osipov, S. Schorb, M. Simon, M. Swiggers,S. Techert, K. Ueda, C. Bostedt, D. Rolles, A.Rudenko, Struct. Dyn. 2016, 3, 043207.

[97] B. K. McFarland, J. P. Farrell, S. Miyabe, F.Tarantelli, A. Aguilar, N. Berrah, C. Bostedt,J. D. Bozek, P. H. Bucksbaum, J. C. Castagna,R. N. Coffee, J. P. Cryan, L. Fang, R. Feifel, K.J. Gaffney, J. M. Glownia, T. J. Martinez, M.Mucke, B. Murphy, A. Natan, T. Osipov, V. S.Petrovic, S. Schorb, T. Schultz, L. S. Spector,M. Swiggers, I. Tenney, S. Wang, J. L. White,W.White, M. Gühr, Nature Commun. 2014, 5,4235.

[82] P. Coppens, Struct. Dyn. 2017, 4, 032102.[83] M. Cammarata, R. Bertoni, M. Lorenc, H.

Cailleau, S. Di Matteo, C. Mauriac, S. F.Matar, H. T. Lemke, M. Chollet, S. Ravy, C.Laulhé, J.-F. Létard, E. Collet, Phys. Rev. Lett.2014, 113, 227402.

[84] M. Sugahara, E. Mizohata, E. Nango, M.Suzuki, T. Tanaka, T. Masuda, R. Tanaka, T.Shimamura, Y. Tanaka, C. Suno, K. Ihara,D. Pan, K. Kakinouchi, S. Sugiyama, M.Murata, T. Inoue, K. Tono, C. Song, J. Park, T.Kameshima, T. Hatsui, Y. Joti, M. Yabashi, S.Iwata, Nat. Meth. 2014, 12, 61.

[85] C. E. Conrad, S. Basu, D. James, D. Wang,A. Schaffer, S. Roy-Chowdhury, N. A.Zatsepin, A. Aquila, J. Coe, C. Gati, M. S.Hunter, J. E. Koglin, C. Kupitz, G. Nelson, G.Subramanian, T. A. White, Y. Zhao, J. Zook,S. Boutet, V. Cherezov, J. C. H. Spence, R.Fromme, U. Weierstall, P. Fromme, IUCrJ2015, 2, 421.

[86] P. Nogly, V. Panneels, G. Nelson, C. Gati,T. Kimura, C. J. Milne, D. Milathianaki, M.Kubo, W. Wu, C. Conrad, J. Coe, R. Bean, Y.Zhao, P. Båth, R. Dods, R. Harimoorthy, K.R. Beyerlein, J. Rheinberger, D. James, D.Deponte, C. Li, L. Sala, G. J. Williams, M. S.Hunter, J. E. Koglin, P. Berntsen, E. Nango, S.Iwata, H. N. Chapman, P. Fromme, M. Frank,R. Abela, S. Boutet, A. Barty, T. A. White, U.Weierstall, J. Spence, R. Neutze, G. Schertler,J. Standfuss,Nature Commun. 2016, 7, 12314.

[87] E. Nango, A. Royant, M. Kubo, T. Nakane, C.Wickstrand, T. Kimura, T. Tanaka, K. Tono,C. Song, R. Tanaka, T. Arima, A. Yamashita,J. Kobayashi, T. Hosaka, E. Mizohata, P.Nogly, M. Sugahara, D. Nam, T. Nomura, T.Shimamura, D. Im, T. Fujiwara,Y.Yamanaka,B. Jeon, T. Nishizawa, K. Oda, M. Fukuda, R.Andersson, P. Båth, R. Dods, J. Davidsson,S. Matsuoka, S. Kawatake, M. Murata, O.Nureki, S. Owada, T. Kameshima, T. Hatsui,Y. Joti, G. Schertler, M. Yabashi, A.-N.Bondar, J. Standfuss, R. Neutze, S. Iwata,Science 2016, 354, 1552.

[88] K. R. Beyerlein, C. Jooss, A. Barty, R. Bean,S. Boutet, S. S. Dhesi, R. B. Doak, M. Först,L. Galli, R. A. Kirian, J. Kozak, M. Lang,R. Mankowsky, M. Messerschmidt, J. C. H.Spence, D. Wang, U. Weierstall, T. A. White,G. J. Williams, O.Yefanov, N. A. Zatsepin, A.Cavalleri, H. N. Chapman, Powder Diff. 2015,30, S25.

[89] H. Ostrom, H. Öberg, H. Xin, J. LaRue,M. Beye, M. Dell’Angela, J. Gladh, M. L.Ng, J. A. Sellberg, S. Kaya, G. Mercurio,D. Nordlund, M. Hantschmann, F. Hieke,D. Kuehn, W. F. Schlotter, G. L. Dakovski,J. J. Turner, M. P. Minitti, A. Mitra, S. P.Moeller, A. Foehlisch, M.Wolf, W.Wurth, M.Persson, J. K. Nørskov, F. Abild-Pedersen, H.Ogasawara, L. G. M. Pettersson, A. Nilsson,Science 2015, 347, 978.

[90] M. Dell’Angela, T. Anniyev, M. Beye, R.Coffee, A. Föhlisch, J. Gladh, T. Katayama,S. Kaya, O. Krupin, J. LaRue, A. Møgelhøj,D. Nordlund, J. K. Nørskov, H. Öberg, H.

thelausannecentrefor ultrafastscience(lacus)the lausanne centre for ultrafast science (lacus) chimia...

Documents