Pure Molecular Beam of Water Dimer
Helen Bieker, Jolijn Onvlee, Melby Johny, Lanhai He, Thomas Kierspel, Sebastian Trippel, Daniel A. Horke, Jochen Küpper
PPure Molecular Beam of Water Dimer
Helen Bieker,
1, 2, 3
Jolijn Onvlee, Melby Johny,
1, 2
Lanhai He, k Thomas Kierspel,
1, 2, 3, § Sebastian Trippel,
1, 3
Daniel A. Horke,
1, 3, ‡ and Jochen Küpper
1, 2, 3, ∗ Center for Free-Electron Laser Science, Deutsches Elektronen-Synchrotron DESY, Notkestraße 85, 22607 Hamburg, Germany Department of Physics, Universität Hamburg, Luruper Chaussee 149, 22761 Hamburg, Germany Center for Ultrafast Imaging, Universität Hamburg, Luruper Chaussee 149, 22761 Hamburg, Germany
Spatial separation of water dimers from water monomers and larger water-clusters through theelectric deflector is presented. A beam of water dimers with
93 % purity and a rotational temperatureof . K was obtained. Following strong-field ionization using a 35 fs laser pulse with a wavelengthcentered around 800 nm and a peak intensity of W / cm we observed proton transfer and
46 % of ionized water dimers broke apart into hydronium ions H O + and neutral OH. INTRODUCTION
Hydrogen bonding between water molecules plays animportant role in aqueous systems, e. g., for biomoleculesthat are surrounded by solvents. It is responsible forthe unique properties of water, such as its high boilingpoint [1]. While hydrogen bonds have been studied ex-tensively in many different molecular systems [2–8], oneof the most important models remains the water dimer,somehow the smallest drop of water. Numerous studieshave been conducted on this benchmark system and itsstructure with a single hydrogen bond is well known [9–12].Water molecules and water-clusters have been studiedusing various techniques to describe dynamics such asproton motion [13] or chemical processes, e. g., reactivecollisions [14]. For investigations of ultrafast moleculardynamics, such as energy and charge transfer across hy-drogen bonds in molecular systems, photoion-photoioncoincidence measurements at free-electron lasers are devel-oping as a powerful tool [8, 15, 16] and this approach wasalso used to study hydrogen bonding in the water dimerat a synchrotron [17]. Other spectroscopic techniquesutilizing synchrotron facilities [18, 19] or table-top laser-systems [5, 6, 20, 21] further improved the knowledgeabout hydrogen bonding in water and water-clusters.Most of these experiments investigating the dynamicsof hydrogen-bonded systems would benefit from sam-ples of identical molecules in a well-defined initial state.The widely used supersonic expansion technique providescold molecular beams down to rotational temperaturesof < K [22–24]. However, cluster expansions do notproduce single-species beams, but a mixture of variouscluster stoichiometries. Hence, only low concentrationsof specific species can be achieved. In the case of watermolecules, supersonic expansion produces a cold beam ofvarious water clusters [2] with a water dimer concentrationof only a few percent [17, 25]. This leads to small experi-mental event rates and requires long measurement times,e. g., in coincidence detection schemes [16, 17]. Theseexperiments with a mixture of molecules in a molecularbeam are only feasible if it can be disentangled which molecule was actually measured. Therefore, these mix-tures severely limit the applicable techniques. A purebeam of water dimers would significantly speed up themeasurements, when unwanted backgrounds from carriergas and larger water-clusters are avoided, or simply enablesuch experiments.The electrostatic deflector is an established methodto spatially separate the molecules of interest from thecarrier gas and to separate different species within a coldmolecular beam [26]. This includes the separation ofmolecular conformers [27–30], individual quantum statesof small molecules [31, 32], as well as specific molecularclusters [24, 33, 34]. The deflector was previously utilizedin investigations of water, e. g., to determine the rotationaltemperatures of “warm” molecular beams of water [35], toseparate its para and ortho species [32], and to measurethe dipole moment of small water-clusters [36]. Alterna-tively, separation by the cluster species’ distinct collisioncross sections, i. e., by the transverse momentum changesdue to scattering with a perpendicular rare-gas beam, wasdemonstrated [37]; this method is especially amenable tolarger cluster sizes [38]. Such spatially separated single-species samples enable, for instance, advanced imagingapplications of water-clusters using non-species-specifictechniques, as well as the study of size-specific effects andthe transition from single-molecule to bulk behavior.
EXPERIMENTAL METHODS
Here, the electrostatic deflector was used to spatiallyseparate water dimers from water monomers as well aslarger water-clusters in a molecular beam formed by super-sonic expansion. The experimental setup was describedpreviously [26, 39]. Briefly, liquid water was placed inthe reservoir of an Even-Lavie valve [23], heated to ◦ C,seeded in 100 bar of helium, and expanded into vacuumwith a nominal driving-pulse duration of 19.5 µs and at arepetition rate of 250 Hz. The produced molecular beamwas doubly skimmed, 6.5 cm ( ∅ = 3 mm) and 30.2 cm( ∅ = 1 . mm) downstream from the nozzle, directedthrough the electrostatic deflector [40] of 154 mm length a r X i v : . [ phy s i c s . a t m - c l u s ] J u l -0.050.0101 0.00.0020.0040.0060.0080.010 50 100 150-0.020.000.020.040.060.080.10 1 2 3 4 5 6 7 8 90 20 401 2 mass/charge (u/e)n × N o r m a li ze d I o nS i g n a l N o r m a li ze d I o nS i g n a l FIG. 1. TOF-MS in the center of the molecular beam asdepicted in Fig. 2 with a deflector voltage of 0 kV (black, red)and at a position of +3 mm with a deflector voltage of 8 kV(blue). For mass/charge ( m/q ) ratios of . . . u/e, to theleft of the dashed red line, the TOF-MS has been scaled by . . The inset shows the region of m/q = 0 . . . u/e enlarged;see text for details. and with a nominal field strength of 50 kV / cm with anapplied voltage of 8 kV across the deflector, before passingthrough a third skimmer ( ∅ = 1 . mm). The deflectorwas placed 4.4 cm behind the tip of the second skimmer.In the center of a time-of-flight (TOF) mass spectrometer,134.5 cm downstream from the nozzle, molecules werestrong-field ionized by a 35 fs short laser pulse with acentral wavelength around 800 nm and a pulse energyof 170 µJ. Focusing to 65 µm yielded a peak intensityof ∼ W / cm . The generated ions were acceleratedtoward a microchannel-plate detector combined with aphosphor screen and the generated signal was recordedwith a digitizer. The valve, skimmers, and deflector wereplaced on motorized translation stages, which allowedmovement of the molecular beam through the ionizationlaser focus and the recording of vertical molecular-beam-density profiles without moving the laser focus, resultingin fixed imaging conditions [41–43].While the employed strong-field ionization is a general,non-species specific ionization technique, it can also leadto fragmentation of molecules, such that recorded massspectra (MS) do not directly reflect the composition of themolecular beam. In combination with the species-specificdeflection process, however, this can be disentangled and,thus, even allows for the investigation of strong-field-induced fragmentation processes of a single species. RESULTS AND DISCUSSION
TOF-MS of the direct and the deflected beams areshown in Fig. 1. The spectrum of the undeflected beam × Position (mm) H O + (H O) +2 (H O) H + (H O) H + (H O) H + O + (H O)H + -3 -2 -1 0 1 2 3 × simulation N o r m a li ze d I o nS i g n a l FIG. 2. Normalized measured vertical molecular-beam-densityprofiles (dashed lines) of water cation ( H O ) + (blue), water-dimer cation ( H O ) +2 (red), and protonated water-cluster ions ( H O ) n H + up to n = 4 (yellow, green, cyan, orange) withdeflector voltages of 0 kV (black circles) and 8 kV (squares).Simulated vertical molecular-beam-density profiles of the un-deflected water monomer (grey) as well as of the deflectedwater monomer (light blue) and water dimer (dark red) with adeflector voltage of 8 kV are shown as solid lines. The shaded(dark red) area depicts the error estimate of the water dimersimulation due to the temperature uncertainty, T rot = 1 . K;see text for details. The black arrow indicates the positionin the deflected beam where the TOF-MS shown in Fig. 1was measured. The inset shows the deflection region enlargedwith a magnification factor of applied to the ( H O ) +2 and ( H O ) n H + signals. shows water-cluster ions ( H O ) + n up to n = 2 and proto-nated water-cluster ions ( H O ) n H + up to n = 10 . Evenlarger clusters were likely formed in the supersonic expan-sion, but were not observed in the recorded TOF interval.We point out that all clusters that reach the interactionregion are neutral clusters of the type ( H O ) n , and pro-tonated clusters must result from the interactions withthe femtosecond laser, i. e., due to fragmentation duringor after the strong-field-ionization process.Vertical molecular-beam-density profiles for water ions ( H O ) + , water dimer ions ( H O ) +2 , and protonated water-cluster ions ( H O ) n H + up to n = 4 , with a potentialdifference of 8 kV applied across the deflector, are shownin Fig. 2. For comparison, a field-free vertical profile forthe water ion with 0 kV across the deflector is also shown.The vertical molecular-beam-density profiles have beennormalized to the area of the field-free spatial profile ofthe water ion. For visibility the water dimer profile hasbeen scaled by a factor of 100 after normalization. Whilethe field-free molecular beam profile is centered around0 mm, application of a voltage of 8 kV to the deflectorshifted the peak of water ions, water dimer ions, andprotonated water-cluster ions by +0.5 mm, as indicatedby the red arrow in Fig. 2. In addition, water dimer ionsshowed a broadening and an increase of signal at around+3 mm, indicated by a black arrow in Fig. 2.In the inset of Fig. 2 the region around +3 mm isshown enlarged with a magnification factor of 5 appliedto ( H O ) +2 and ( H O ) n H + with n = 1 . . . . The corre-sponding TOF-MS in the deflected part of the beam ata position of +3 mm is highlighted in Fig. 1 by the blueline. Not just water dimer ions, but also hydronium ions,H O + , and water ions, H O + , showed an increased signalin the deflected beam. The shape of the vertical beamprofiles for these ions matched the water dimer profile inthe region of 2.8–3.5 mm, indicating that they originatedfrom the same parent molecule.The water dimer ion was the largest non-protonatedcluster measured in this setup. To verify that the wa-ter dimer ion was originating from the water dimer, thedeflection behaviour of water-clusters inside the electro-static deflector was simulated. Therefore, the Stark en-ergies and effective dipole moments of water monomersand water-clusters were calculated with the freely avail-able CMIstark software package [44] using rotationalconstants, dipole moments, and centrifugal distortionconstants from the literature [11, 45–48], see Suppl. Inf.Table I; contributions of the polarizability to the Starkeffect could safely be ignored [40, 49]. The rotational con-stants of the water dimer are significantly smaller thanfor the water monomer, leading to a larger effective dipolemoment for the water dimer than for water and a largeracceleration in the electric field in the deflector, see Fig. 3and Table I of the Suppl. Inf. for further information.The simulated vertical molecular-beam-density profilesof the water monomer and the water dimer are shown inFig. 2. The deviations between the measured and simu-lated undeflected vertical beam profiles are ascribed toimperfect alignment of experimental setup, which wasnot taken into account in the simulations. Due to therotational-state dependence of the Stark effect, the deflec-tion of a molecular beam in an electrostatic field dependson the rotational temperature of the molecular ensem-ble [26] and the best fit for the profiles of the watermonomer and the water dimer at a deflector voltage of8 kV was obtained assuming a Boltzmann population dis-tribution of rotational states corresponding to . K.Not only deflection of water-clusters measured as amass of 36 amu, but also of water-clusters detected asprotonated-clusters have been measured, for instance,for ( H O ) +2 , as indicated by the red arrow and sym-bols in Fig. 2. Trajectory simulations for water-clusters ( H O ) + n with n = 3 . . . using a rotational temperature of T rot = 1 . K were performed to understand the originof this deflection behavior. For the water hexamer threeand for the water heptamer two conformers have been sim-ulated, assuming an equal population of the conformers.These showed that, based on the different effective dipolemoments, a different deflection is expected for differentwater-clusters, see Fig. 5 of the Suppl. Inf.. Since the detected protonated water-clusters arose from the strong-field fragmentation of larger neutral clusters in the inter-action region, the measured vertical protonated-clusterdensity profiles are a superposition of several neutralwater-cluster density profiles. Thus, it is not possible tocompare the individual simulated molecular-beam-densityprofiles of neutral clusters directly with the measuredprotonated water-cluster density profiles. Instead, at eachposition of the deflection profile the signal from all water-clusters has been summed up, both for the simulated andthe measured molecular-beam-density profiles. The resultyields a comparable amount of deflection for simulatedand measured molecular-beam-density profiles, see Suppl.Inf. Fig. 6. The shift of 0.5 mm can, therefore, originatefrom the superimposed molecular-beam-density profilesfrom different larger clusters due to fragmentation intosmaller water-clusters. The same shift is visible for H O + and ( H O ) +2 , which indicates that water-clusters are alsofragmenting into H O + and ( H O ) +2 . Nevertheless, thesimulation for water-clusters n = 1 . . . shows that thewater dimer deflected the most, reaching a position of+3 mm and above, see Suppl. Inf. Fig. 4 and Fig. 5. Ofall the other clusters considered, only the water hexamerin its prism and book forms reaches to a position up to3.2 mm with the falling edge of the profile. In our ex-periments the water hexamer and higher order clustershave only been measured as fragments, such that theconcentration and size distribution of neutral clusters inthe molecular beam is unknown. However, the measuredfragment distributions strongly suggest that significantlylarger clusters are not present, since the ion signals de-cay exponentially and it is known that clusters primarilyfragment through loss of single water molecules [50–52].The TOF-MS in the deflected part of the beam, shownin Fig. 1, contains peaks corresponding to H + , O + , OH + ,H O + , and H O + , in addition to the water dimer ion. Asmentioned before the short-pulse ionization can lead tofragmentation. For the water dimer, two fragmentationchannels were reported for electron-impact ionization with70 eV electrons [50]: either an H O + ion and a neutral OHare formed or a H O + ion and a neutral water monomerH O. Using a size-selection method and infrared spec-troscopy, H O + has been reported as a fragment of thewater dimer [37]. Comparison of the vertical molecular-beam-density profiles of the deflected molecules allowedfurther investigation of the fragmentation channels ofthe water dimer. The measured vertical molecular-beam-density profiles of these molecules showed a similar deflec-tion behavior in the region of 2.8 to 3.5 mm as the waterdimer, see Fig. 2 and Suppl. Inf. Fig. 1. The observedconstant ratio of those fragments over this spatial regionindicates that all these fragments originated from thewater dimer.Comparing the intensity of the fragments of the waterdimer, H O + and H O + and (H O ) +2 , in the deflectedbeam, the fragmentation ratios of the intact water dimerwere estimated. These showed that of the waterdimer fragmented into one ionized water molecule, while of the water dimer underwent most likely protontransfer and formed a hydronium ion. Only of thewater dimer present in the molecular beam stayed intactafter ionization.The actual number of water dimer molecules per shotin the deflected molecular beam was estimated to ∼ . within the laser focus using the known fragmentationratios of H O + and H O + , while the fragmentation chan-nels of H + , O + , OH + have not been included. Taking theknown fragmentation channels into account, the fractionof the water dimer within the molecular beam was evalu-ated. Comparing the ratios between the water dimer andall other species visible in the TOF, a water dimer fractionof . in the center of the undeflected beam and of in the deflected beam, at a position of +3 mm,was achieved. Thus, using the electrostatic deflector thefraction of the water dimer within the interaction regioncould be increased by nearly a factor of 24. CONCLUSIONS
In summary, a high-purity beam of water dimers wascreated using the electrostatic deflector, which spatiallyseparated water dimers from other species present in themolecular beam. The resulting water dimer sample hada purity of . The fragmentation products andratios of the water dimer following strong-field ioniza-tion using a 35 fs laser pulse with a wavelength centeredaround 800 nm and peak intensity of ∼ W / cm werestudied, with of the water dimer found to forma hydronium ion and fragmenting into one wa-ter cation and one neutral water monomer, while of the water dimer stayed intact. The deflection profilescould be simulated using a rigid-rotor model and an initialrotational temperature of 1.5(5) K.The produced clean samples of water dimers arewell suited for non-species-specific experiments, e. g.,reactive-collisions, diffractive imaging, or ultrafast spec-troscopies [14, 41, 53]. Even for experiments that can dis-tinguish different species, for example photoion-photoioncoincidence measurements [8, 54], the produced cleanbeams will enable significantly faster measurements ofthis important hydrogen-bonded model system, e. g., be-cause unwanted backgrounds are avoided. Furthermore,the electrostatic separation technique can be used to sep-arate different conformers [26], which could be highlyinteresting in the purification and studies of larger water-clusters that exhibit multiple conformers [55]. ACKNOWLEDGMENTS
This work has been supported by the Clusters of Ex-cellence “Center for Ultrafast Imaging” (CUI, EXC 1074,ID 194651731) and “Advanced Imaging of Matter” (AIM,EXC 2056, ID 390715994) of the Deutsche Forschungs-gemeinschaft (DFG), by the European Union’s Horizon2020 research and innovation program under the MarieSkłodowska-Curie Grant Agreement 641789 “MolecularElectron Dynamics investigated by Intense Fields andAttosecond Pulses” (MEDEA), by the European ResearchCouncil under the European Union’s Seventh Frame-work Program (FP7/2007-2013) through the ConsolidatorGrant COMOTION (ERC-Küpper-614507), and by theHelmholtz Gemeinschaft through the “Impuls- und Ver-netzungsfond”. L.H. acknowledges a fellowship within theframework of the Helmholtz-OCPC postdoctoral exchangeprogram and J.O. gratefully acknowledges a fellowship bythe Alexander von Humboldt Foundation.
SUPPORTING INFORMATION DESCRIPTION
Supporting Information Available: Description of thefragmentation correction method and the trajectory sim-ulations k Permanent address: Institute of Atomic and MolecularPhysics, Jilin University, Changchun 130012, China § Present address: Department of Chemistry, University ofBasel, Klingelbergstrasse 80, 4056 Basel, Switzerland ‡ Present address: Institute for Molecules and Materials,Radboud University, Heijendaalseweg 135, 6525 AJ Ni-jmegen, The Netherlands ∗ An Introduction to Hydrogen Bonding (Ox-ford University Press, 1997).[2] K. Liu, J.D. Cruzan, and R.J. Saykally, “Water clusters,”Science , 929–933 (1996).[3] K. Nauta and R. E. Miller, “Formation of cyclic waterhexamer in liquid helium: The smallest piece of ice,”Science , 293 (2000).[4] G. T. Dunning, D. R. Glowacki, T. J. Preston, S. J.Greaves, G. M. Greetham, I. P. Clark, M. Towrie, J. N.Harvey, and A. J. Orr-Ewing, “Vibrational relaxationand microsolvation of DF after F-atom reactions in polarsolvents,” Science , 530–533 (2015).[5] G. Berden, W. L. Meerts, M. Schmitt, and K. Klein-ermanns, “High resolution UV spectroscopy of phenoland the hydrogen bonded phenol-water cluster,” J. Chem.Phys. , 972 (1996).[6] Timothy M. Korter, David W. Pratt, and Jochen Küpper,“Indole-H O in the gas phase. Structures, barriers to inter-nal motion, and S ← S transition moment orientation.Solvent reorganization in the electronically excited state,”J. Phys. Chem. A , 7211–7216 (1998). [7] A. L. Sobolewski and W. Domcke, “Photoinduced electronand proton transfer in phenol and its clusters with waterand ammonia,” J. Phys. Chem. A , 9275–9283 (2001).[8] X. Ren, E. Wang, A. D. Skitnevskaya, A. B. Trofimov,G. Kirill, and A. Dorn, “Experimental evidence for ul-trafast intermolecular relaxation processes in hydratedbiomolecules,” Nat. Phys. , 1745 (2018).[9] J. A. Odutola and T. R. Dyke, “Partially deuterated waterdimers: Microwave spectra and structure,” J. Chem. Phys. , 5062–5070 (1980).[10] H. Yu and W. F. van Gunsteren, “Charge-on-spring po-larizable water models revisited: From water clustersto liquid water to ice,” J. Chem. Phys. , 9549–9564(2004).[11] T. R. Dyke, K. M. Mack, and J. S. Muenter, “Thestructure of water dimer from molecular beam electricresonance spectroscopy,” J. Chem. Phys. , 498–510(1977).[12] T. R. Dyke and J. S. Muenter, “Molecular-beam electricdeflection studies of water polymers,” J. Chem. Phys. ,5011 (1972).[13] T. Marchenko, L. Inhester, G. Goldsztejn, O. Travnikova,L. Journel, R. Guillemin, I. Ismail, D. Koulentianos,D. Céolin, R. Püttner, M. N. Piancastelli, and M. Simon,“Ultrafast nuclear dynamics in the doubly-core-ionizedwater molecule observed via auger spectroscopy,” Phys.Rev. A , 063403 (2018).[14] Ardita Kilaj, Hong Gao, Daniel Rösch, Uxia Rivero,Jochen Küpper, and Stefan Willitsch, “Observation ofdifferent reactivities of para- and ortho-water towardstrapped diazenylium ions,” Nat. Commun. , 2096 (2018).[15] Rebecca Boll, Benjamin Erk, Ryan Coffee, Sebastian Trip-pel, Thomas Kierspel, Cédric Bomme, John D. Bozek,Mitchell Burkett, Sebastian Carron, Ken R. Ferguson,Lutz Foucar, Jochen Küpper, Tatiana Marchenko, CatalinMiron, Minna Patanen, Timur Osipov, Sebastian Schorb,Marc Simon, Michelle Swiggers, Simone Techert, KiyoshiUeda, Christoph Bostedt, Daniel Rolles, and ArtemRudenko, “Charge transfer in dissociating iodomethaneand fluoromethane molecules ionized by intense femtosec-ond x-ray pulses,” Struct. Dyn. , 043207 (2016).[16] Thomas Kierspel, Imaging structure and dynamics usingcontrolled molecules , Dissertation, Universität Hamburg,Hamburg, Germany (2016).[17] T. Jahnke, H. Sann, T. Havermeier, K. Kreidi, C. Stuck,M. Meckel, M. Schöffler, N. Neumann, R. Wallauer,S. Voss, A. Czasch, O. Jagutzki, A. Malakzadeh,F. Afaneh, Th. Weber, H. Schmidt-Böcking, andR. Dörner, “Ultrafast energy transfer between watermolecules,” Nat. Phys. , 139–142 (2010).[18] Bernd Winter, Emad F. Aziz, Uwe Hergenhahn, ManfredFaubel, and Ingolf V. Hertel, “Hydrogen bonds in liquidwater studied by photoelectron spectroscopy,” J. Chem.Phys. , 124504 (2007).[19] Jared D. Smith, Christopher D. Cappa, Kevin R. Wilson,Benjamin M. Messer, Ronald C. Cohen, and Richard J.Saykally, “Energetics of hydrogen bond network rearrange-ments in liquid water,” Science , 851–853 (2004).[20] Frank N. Keutsch and Richard J. Saykally, “Inauguralarticle: Water clusters: Untangling the mysteries of theliquid, one molecule at a time,” PNAS , 10533–10540(2001).[21] T. S. Zwier, “The spectroscopy of solvation in hydrogen-bonded aromatic clusters,” Annu. Rev. Phys. Chem. , 205–241 (1996).[22] G. Scoles, ed., Atomic and molecular beam methods , Vol. 1(Oxford University Press, New York, NY, USA, 1988).[23] U. Even, J. Jortner, D. Noy, N. Lavie, and N. Cossart-Magos, “Cooling of large molecules below 1 K and He clus-ters formation,” J. Chem. Phys. , 8068–8071 (2000).[24] Melby Johny, Jolijn Onvlee, Thomas Kierspel, HelenBieker, Sebastian Trippel, and Jochen Küpper, “Spa-tial separation of pyrrole and pyrrole-water clusters,”Chem. Phys. Lett. , 149–152 (2019), arXiv:1901.05267[physics].[25] J. B. Paul, C. P. Collier, R. J. Saykally, J. J. Scherer, andA. O’Keefe, “Direct measurement of water cluster con-centrations by infrared cavity ringdown laser absorptionspectroscopy,” J. Phys. Chem. A , 5211–5214 (1997).[26] Yuan-Pin Chang, Daniel A. Horke, Sebastian Trippel, andJochen Küpper, “Spatially-controlled complex moleculesand their applications,” Int. Rev. Phys. Chem. , 557–590 (2015), arXiv:1505.05632 [physics].[27] Frank Filsinger, Undine Erlekam, Gert von Helden,Jochen Küpper, and Gerard Meijer, “Selector for struc-tural isomers of neutral molecules,” Phys. Rev. Lett. ,133003 (2008), arXiv:0802.2795 [physics].[28] Frank Filsinger, Jochen Küpper, Gerard Meijer, Jonas L.Hansen, Jochen Maurer, Jens H. Nielsen, LotteHolmegaard, and Henrik Stapelfeldt, “Pure samples ofindividual conformers: the separation of stereo-isomersof complex molecules using electric fields,” Angew. Chem.Int. Ed. , 6900–6902 (2009).[29] Thomas Kierspel, Daniel A. Horke, Yuan-Pin Chang, andJochen Küpper, “Spatially separated polar samples of the cis and trans conformers of 3-fluorophenol,” Chem. Phys.Lett. , 130–132 (2014), arXiv:1312.4417 [physics].[30] Nicole Teschmit, Daniel A. Horke, and Jochen Küpper,“Spatially separating the conformers of the dipeptide Ac-Phe-Cys-NH ,” Angew. Chem. Int. Ed. , 13775–13779(2018), arXiv:1805.12396 [physics].[31] Jens H. Nielsen, Paw Simesen, Christer Z. Bisgaard, Hen-rik Stapelfeldt, Frank Filsinger, Bretislav Friedrich, Ger-ard Meijer, and Jochen Küpper, “Stark-selected beamof ground-state OCS molecules characterized by revivalsof impulsive alignment,” Phys. Chem. Chem. Phys. ,18971–18975 (2011), arXiv:1105.2413 [physics].[32] Daniel A. Horke, Yuan-Pin Chang, Karol Długołęcki,and Jochen Küpper, “Separating para and ortho wa-ter,” Angew. Chem. Int. Ed. , 11965–11968 (2014),arXiv:1407.2056 [physics].[33] Sebastian Trippel, Yuan-Pin Chang, Stephan Stern, TerryMullins, Lotte Holmegaard, and Jochen Küpper, “Spa-tial separation of state- and size-selected neutral clus-ters,” Phys. Rev. A , 033202 (2012), arXiv:1208.4935[physics].[34] Hyun Sik You, Junggil Kim, Songhee Han, Doo-Sik Ahn,Jean Sun Lim, and Sang Kyu Kim, “Spatial isolationof conformational isomers of hydroquinone and its watercluster using the stark deflector,” J. Phys. Chem. A ,1194 (2018).[35] Ramiro Moro, Jaap Bulthuis, Jonathon Heinrich, andVitaly V. Kresin, “Electrostatic deflection of the watermolecule: A fundamental asymmetric rotor,” Phys. Rev.A , 013415 (2007).[36] Ramiro Moro, Roman Rabinovitch, Chunlei Xia, and Vi-taly V. Kresin, “Electric dipole moments of water clustersfrom a beam deflection measurement,” Phys. Rev. Lett. , 123401 (2006).[37] Udo Buck and Friedrich Huisken, “Infrared spectroscopyof size-selected water and methanol clusters,” Chem. Rev. , 3863–3890 (2000).[38] Christoph C. Pradzynski, Richard M. Forck, ThomasZeuch, Petr Slavíček, and Udo Buck, “A fully size-resolvedperspective on the crystallization of water clusters,” Sci-ence , 1529–1532 (2012).[39] Sebastian Trippel, Melby Johny, Thomas Kierspel, JolijnOnvlee, Helen Bieker, Hong Ye, Terry Mullins, LarsGumprecht, Karol Długołęcki, and Jochen Küpper, “Knifeedge skimming for improved separation of molecularspecies by the deflector,” Rev. Sci. Instrum. , 096110(2018), arXiv:1802.04053 [physics].[40] Jens S. Kienitz, Karol Długołęcki, Sebastian Trippel,and Jochen Küpper, “Improved spatial separation ofneutral molecules,” J. Chem. Phys. , 024304 (2017),arXiv:1704.08912 [physics].[41] Jochen Küpper, Stephan Stern, Lotte Holmegaard, FrankFilsinger, Arnaud Rouzée, Artem Rudenko, Per Johnsson,Andrew V. Martin, Marcus Adolph, Andrew Aquila, SašaBajt, Anton Barty, Christoph Bostedt, John Bozek, CarlCaleman, Ryan Coffee, Nicola Coppola, Tjark Delmas,Sascha Epp, Benjamin Erk, Lutz Foucar, Tais Gorkhover,Lars Gumprecht, Andreas Hartmann, Robert Hartmann,Günter Hauser, Peter Holl, Andre Hömke, Nils Kimmel,Faton Krasniqi, Kai-Uwe Kühnel, Jochen Maurer, MarcMesserschmidt, Robert Moshammer, Christian Reich,Benedikt Rudek, Robin Santra, Ilme Schlichting, CarloSchmidt, Sebastian Schorb, Joachim Schulz, Heike Soltau,John C. H. Spence, Dmitri Starodub, Lothar Strüder,Jan Thøgersen, Marc J. J. Vrakking, Georg Weidens-pointner, Thomas A. White, Cornelia Wunderer, Ger-ard Meijer, Joachim Ullrich, Henrik Stapelfeldt, DanielRolles, and Henry N. Chapman, “X-ray diffraction fromisolated and strongly aligned gas-phase molecules with afree-electron laser,” Phys. Rev. Lett. , 083002 (2014),arXiv:1307.4577 [physics].[42] Stephan Stern, Lotte Holmegaard, Frank Filsinger, Ar-naud Rouzée, Artem Rudenko, Per Johnsson, Andrew V.Martin, Anton Barty, Christoph Bostedt, John D. Bozek,Ryan N. Coffee, Sascha Epp, Benjamin Erk, Lutz Fou-car, Robert Hartmann, Nils Kimmel, Kai-Uwe Kühnel,Jochen Maurer, Marc Messerschmidt, Benedikt Rudek,Dmitri G. Starodub, Jan Thøgersen, Georg Weidenspoint-ner, Thomas A. White, Henrik Stapelfeldt, Daniel Rolles,Henry N. Chapman, and Jochen Küpper, “Toward atomicresolution diffractive imaging of isolated molecules withx-ray free-electron lasers,” Faraday Disc. , 393 (2014),arXiv:1403.2553 [physics]. [43] Frank Filsinger, Jochen Küpper, Gerard Meijer, LotteHolmegaard, Jens H. Nielsen, Iftach Nevo, Jonas L.Hansen, and Henrik Stapelfeldt, “Quantum-state selec-tion, alignment, and orientation of large molecules us-ing static electric and laser fields,” J. Chem. Phys. ,064309 (2009), arXiv:0903.5413 [physics].[44] Y.-P. Chang, F. Filsinger, B. Sartakov, and J. Küp-per, “ CMIstark : Python package for the stark-effectcalculation and symmetry classification of linear, sym-metric and asymmetric top wavefunctions in dc elec-tric fields,” Comp. Phys. Comm. , 339–349 (2014),arXiv:1308.4076 [physics].[45] Frank C. DeLucia, Paul Helminger, and William H. Kirch-hoff, “Microwave spectra of molecules of astrophysicalinterest V. Water vapor,” J. Phys. Chem. Ref. Data ,211–219 (1974).[46] L. H. Coudert and J. T. Hougen, “Analysis of the mi-crowave and far infrared spectrum of the water dimer,”Journal Of Molecular Spectroscopy , 259–277 (1990).[47] Shelley L Shostak, William L Ebenstein, and John SMuenter, “The dipole moment of water. I. dipole momentsand hyperfine properties of H O and HDO in the groundand excited vibrational states,” J. Chem. Phys. , 5875(1991).[48] N. P. Malomuzh, V. N. Makhlaichuk, and S. V. Khrapatyi,“Water dimer dipole moment,” Russian Journal of PhysicalChemistry A , 1431–1435 (2014).[49] George Maroulis, “Static hyperpolarizability of the waterdimer and the interaction hyperpolarizability of two watermolecules,” J. Chem. Phys. , 1813–1820 (2000).[50] L. Angel and A. J. Stace, “Dissociation patterns of (H O) + n cluster ions, for n=2–6,” Chem. Phys. Lett. , 277–281(2001).[51] Leonid Belau, Kevin R Wilson, Stephen R Leone, andMusahid Ahmed, “Vacuum ultraviolet (VUV) photoion-ization of small water clusters,” J. Phys. Chem. A ,10075–10083 (2007).[52] Xiaojie Liu, Wen-Cai Lu, C.Z. Wang, and K.M. Ho,“Energetic and fragmentation stability of water clusters(h2o)n, n=2–30,” Chem. Phys. Lett. , 270 – 275 (2011).[53] Christopher J. Hensley, Jie Yang, and Martin Centu-rion, “Imaging of isolated molecules with ultrafast electronpulses,” Phys. Rev. Lett. , 133202 (2012).[54] Thomas Kierspel, Cédric Bomme, Michele Di Fraia, JossWiese, Denis Anielski, Sadia Bari, Rebecca Boll, BenjaminErk, Jens S. Kienitz, Nele L. M. Müller, Daniel Rolles,Jens Viefhaus, Sebastian Trippel, and Jochen Küpper,“Photophysics of indole upon x-ray absorption,” Phys.Chem. Chem. Phys. , 20205 (2018), arXiv:1802.02964[physics].[55] J. K. Gregory, D. C. Clary, K. Liu, M. G. Brown, and R. J.Saykally, “The water dipole moment in water clusters,”Science , 814–817 (1997). upplemental Material: Pure molecular beam of water dimer Helen Bieker,
1, 2, 3
Jolijn Onvlee, Melby Johny,
1, 2
Lanhai He, k Thomas Kierspel,
1, 2, 3, § Sebastian Trippel,
1, 2
Daniel A. Horke,
1, 2, ‡ and Jochen Küpper
1, 2, 3, ∗ Center for Free-Electron Laser Science, Deutsches Elektronen-Synchrotron DESY, Notkestraße 85, 22607 Hamburg, Germany Department of Physics, Universität Hamburg, Luruper Chaussee 149, 22761 Hamburg, Germany Center for Ultrafast Imaging, Universität Hamburg, Luruper Chaussee 149, 22761 Hamburg, Germany
FRAGMENTATION CORRECTION OFMEASUREMENTS
The strong-field-ionization technique employed in thiswork can lead to fragmentation, such that clusters fromthe molecular beam contributed to smaller masses in themass spectrum (MS). For example, the water monomerand the water dimer signals at m/q = 18 u/e and 36 u/e,respectively, contained contributions due to fragmentationof larger water-clusters in the molecular beam. Therefore,measured intensities needed to be corrected for thesefragmentation channels. In addition, background waterinside the chamber was measured at 18 u/e and neededto be corrected for.For the latter, background measurement were perma-nently performed during the experiments using the higherrepetition rate of the laser compared to the valve. Laserpulses were arriving in the interaction region at the sametime as the molecular beam and between two molecularbeam pulses, such that for each data point a backgroundmeasurement was performed. The background signal wassubtracted from the measurements.The fragmentation ratios of the water dimer into smallermasses could be estimated and used for the calculationof the fraction of the water dimer in the deflected andundeflected molecular beam [1]. In Fig. 1 the deflectionprofiles measured at masses corresponding to H + , O + andOH + are shown. In the region of 2.8 to 3.5 mm the de-flection curves look identical to those for the water dimer,indicating that at these positions those are fragmentsfrom the water dimer. The ratios of the water dimer toH + , O + and OH + at a position of 3 mm are 0.3, 0.8, and0.7, respectively. For the calculation of the fraction of thewater dimer in the molecular beam for the undeflectedbeam, these ratios were used to estimate the amount ofthe water dimer inside of the beam.For larger clusters, only fragments were measured,such that the measured signal was not solely due toa specific cluster stoichiometry and the overall shapeof the molecular beam profile arose from several largerwater-clusters. All protonated-water-cluster ions recordedshowed the same deflection behavior, see Fig. 2. Anestimate of the exponential decay of the measured proto-nated water-clusters distribution showed that protonatedwater-clusters n = 1 − contained . of the overallintensity. -3 -2 -1 0 1 2 30.00.10.20.30.4 H O + Position (mm) H + × × × OH + (H O) +2 O + N o r m a li ze d I o nS i g n a l FIG. 1. Column density profiles, measured for the water-monomer cation with deflector voltages of 0 kV (black) and forH + (blue), O + (purple), OH + (orange) and the water-dimercation with a deflector voltage of 8 kV (red). The inset showsthe region around y = 3 mm enlarged, with O + , OH + and thewater dimer ion scaled by a factor . TRAJECTORY SIMULATIONS
The Stark energies and effective dipole moments µ eff ofwater-clusters n = 1 . . . were calculated using the freelyavailable CMIstark software package [2], which werethen used to perform trajectory simulations [3] and toverify the measured deflection profiles of water-clusters.The rotational constants, dipole moments and centrifugaldistortion constants from the literature are summarizedin Table I. Three conformers for the water hexamer inprism-, book- and cage-like form [13] and two conformersof the water heptamer following the naming scheme of [15]were included.For these simulations the water-clusters were assumedto be rigid rotors. Since the water dimer is known to be afloppy molecule with large amplitude motions [7, 16], thecorresponding energy spectra and the description of theinteraction of the states would significantly complicatefurther analysis. Using a rigid rotor assumption enablesan easier and faster description and it has been shownpreviously that this model can be used to describe thedynamics of indole ( H O ) in strong-electric- and laser-field alignment and orientation experiments [17, 18] andto fit pure rotational transitions of the water dimer to a r X i v : . [ phy s i c s . a t m - c l u s ] J u l molecule dipole moment µ (D) rotational constants (MHz) Centrifugal Distortion constants (kHz) µ a µ b µ c A B C ∆ J ∆ JK ∆ K d J d K H O − .
86 0 [4] .
29 435351 .
72 278138 . [5] . × − . × . × . × . × [5] ( H O ) .
63 0 0 [6] . .
76 6133 . [7] .
044 4010 0 0 0 [8] ( H O ) [9] .
91 6646 .
91 0 [10] − − − − − ( H O ) [9] .
00 3149 .
00 1622 . [11] − − − − − ( H O ) .
93 0 0 [9] .
00 1818 .
00 940 . [12] − − − − − ( H O ) book .
17 2 .
46 0 . [13] .
47 1063 .
98 775 . [13] − − − − − ( H O ) cage .
63 0 .
32 1 . [13] .
61 1131 . . [14] − − − − − ( H O ) prism .
41 0 .
88 0 . [13] .
22 1362 .
00 1313 . [13] − − − − − ( H O ) . . . [15] .
44 937 .
88 919 . [15] . − .
342 0 .
842 0 . . [15] ( H O ) . . . [15] .
16 976 .
88 854 . [15] .
044 0 .
000 0 .
000 0 . [15] TABLE I. Dipole moments, rotational constants and centrifugal distortion constants of water-clusters used in the Stark-effectcalculations -3 -2 -1 0 1 2 30.00.20.40.6 cluster sum (H O) H + (H O) H + (H O) H + (H O) H + (H O) H + (H O) H + (H O) H + (H O) H + (H O) H + Position (mm) N o r m a li ze d I n t e n s i t y FIG. 2. Comparison of the averaged measured protonated-water-cluster signal (red) and the individual measured proto-nated water-clusters deflection profiles for n = 2 . . . . Theprofiles are normalized to the area under the curve. experimental measurements [8].For the rotational states J = 0 . . . of the watermonomer and the water dimer the Stark energies andthe corresponding µ eff as a function of the electric fieldstrength are shown in Fig. 3. For the water dimer allrelevant states are strong-field seeking and, hence, acceler-ated toward regions of stronger fields. For a nominal fieldstrength of 50 kV / cm the µ eff of the water dimer are sig-nificantly larger than for the water monomer, except fromthe | J, K a , K c , M i = | , , , i , | , , , i states, leadingto a larger acceleration in the electric field. All theshown states have a small asymmetry splitting, see Ta-ble I, resulting in a fast rise of µ eff at small electric fieldstrength. The discontinuous change of µ eff at an electricfield around 30 kV / cm is ascribed to an avoided crossingof the | , , , i and | , , , i states.The trajectories of the molecules inside the electrostaticdeflector were simulated using the calculated µ eff [3]. Forquantum states J = 0 . . . , trajectories were cal-culated for each set of J states and used to simulatethe spatial profiles using a weighting factor based on thethermal distributions of the state for a given temper- µ e ff ( D ) E n e r g y ( c m − ) Electric field (kV/cm) | , , , > | , , , > | , , , > | , , , > | , , , > | , , , > ; | , , , > | , , , > | , , , > FIG. 3. Calculated Stark energy and effective dipole momentsfor the J = 0 . . . states of the water monomer (blue) andthe water dimer (red, orange, yellow). J = 0 are shown inblue (water) and yellow (water dimer), J = 1 in light blue(water), red (water dimer), J = 2 in darkblue (water) andorange (water dimer) and the | , , , i state in darkred forthe water dimer. States where K a < K c are indicated bydashed lines, K a > K c by solid lines and K a = K c by dottedlines. ature. Those temperature-weighted simulated verticalmolecular-beam profiles were scaled to the area underthe curve of the corresponding experimental profile tocompare the deflection profiles. The simulations includethe nuclear-spin-statistical weights for the water monomerand the water dimer. For para - and ortho -water a room-temperature distribution of was used. The waterdimer in its equilibrium geometry is isomorphic with thepermutation-inversion point group D h including tunnel-ing splittings [19]. Neglecting tunneling splittings andacceptor switching, the rigid water dimer belongs to thesymmetry group C S ( M ) , yielding nuclear-spin-statistical Position (mm) (H O) +2 N o r m a li ze d I o nS i g n a l FIG. 4. Simulated deflection profiles for the water dimerat temperatures of 1.0 K, 1.5 K and 2 K compared to thecorrected pure water dimer profile at 8 kV (red, dots). weights of para : ortho of
16 : 16 [20].The simulated profiles for the water dimer at differ-ent rotational temperatures T rot including rotationalstates J = 0 . . . are shown in Fig. 4. An initial-beam temperature of T rot = 1 . K reproduced theexperiment the best. At this temperature the watermonomer in the para nuclear spin state has
100 % ofits population in its absolute rotational ground states | J = 0 , K a = 0 , K c = 0 , M = 0 i , while ortho-water popu-lates the | J = 1 , K a = 0 , K c = 1 , M = 0 , i state to equalamounts. . of the para-water dimer and . ofthe ortho-water dimer population is within J = 0 . . . .Trajectory simulations were performed for water-clusters up to n = 7 . Based on the estimated water-cluster distribution, vide supra , this covers . of thewater-clusters in the molecular beam, while ∼ . of themolecules in the beam are from water-clusters n ≥ . Thesimulations for water-clusters including J = 0 . . . andusing the same rotational temperature T rot = 1 . Kof the water dimer are shown in Fig. 5. We note that atthis temperature rotational states up to J = 10 mightbe populated in the molecular beam and the rotationaltemperature can differ from the one of the water dimer.Thus the simulations give just an estimate of the amountof deflection. Based on the simulations the water dimeris deflecting the most of all water-clusters, followed bythe water hexamer in prism- and book-like form, whichreaches to a position of +3.2 mm.Since for larger clusters only fragments have been mea-sured and, therefore, the shape of the recorded beamprofiles is the result of a superposition of several neu-tral cluster distributions in the molecular beam, it isnot possible to compare the single deflection profiles di-rectly with simulations. Therefore, at each position of thedeflection profile the signal of the measured protonatedwater-clusters for n = 2 . . . have been summed up. The Position (mm) -3 -2 -1 0 1 2 30.00.20.4