THE JOURNAL OF CHEMICAL PHYSICS 149,044309 (2018) J. Chem. Phys. 149, 044309 (2018)044309-2 Oncak et al. conc Photochemistry and spectroscopy of smallhydrated magnesium clusters Mg(H2O)n,n= 7-5 Cite as: J. Chem.Phys. 149,044309 (2018); https://doi.org/10.1063/1.5037401 Submitted: 24 April 2018. Accepted: 11 July 2018 . Published Online: 27 July 2018 Milan OncakDThomas Taxer, Erik Barwa, Christian van der Linde, and Martin K. Beyer V iew Online Export Citation CrossMark ARTICLES YOU MAY BE INTERESTED IN NH, in a cold ion trap with He buffer gas: Ab initio quantum modeling of the interactionpotential and of state-changing multichannel dynamics The Journal of Chemical Physics 148, 184305 (2018); https://doi.org/10.1063/1.5022633 Photodetachment spectroscopy of cold trapped NH near threshold The Journal of Chemical Physics 149, 104302 (2018); https://doi.org/10.1063/1.5042621 Communication: Charge transfer dominates over proton transfer in the reaction of nitricacid with gas-phase hydrated electrons The Journal of Chemical Physics 147, 101101 (2017); https://doi.org/10.1063/1.4999392 Lock-in Amplifiers up to 600 MHz starting at Zurich Watch the Video J. Chem. Phys. 149, 044309 (2018); https://doi.org/10.1063/1.5037401 @ 2018 Author(s). Photochemistry and spectroscopy of small hydrated magnesiumclusters Mg+(H2O)n,n=1-5 Milan Oncak,a) Thomas Taxer, Erik Barwa, Christian van der Linde, and Martin K. Beyera)Institut fiir Ionenphysik und Angewandte Physik, Universitat Innsbruck, TechnikerstraBe 25, 6020 Innsbruck, Austria (Received 24 April 2018; accepted 11 July 2018;published online 27 July 2018) Hydrated singly charged magnesium ions Mg*(H2O)n,n ≤5, in the gas phase are ideal model systemsto study photochemical hydrogen evolution since atomic hydrogen is formed over a wide range ofwavelengths, with a strong cluster size dependence. Mass selected clusters are stored in the cell of anFourier transform ion cyclotron resonance mass spectrometer at a temperature of 130 K for severalseconds, which allows thermal equilibration via blackbody radiation. Tunable laser light is used forphotodissociation. Strong transitions to D1-3 states (correlating with the 3s-3px,y,z transitions of Mg*)are observed for all cluster sizes, as well as a second absorption band at 4-5 eV for n=3-5. Due tothe lifted degeneracy of the 3pxyz energy levels of Mgt, the absorptions are broad and red shiftedwith increasing coordination number of the Mg* center, from 4.5 eV for n = 1 to 1.8 eV for n=5.In all cases, H atom formation is the dominant photochemical reaction channel. Quantum chemicalcalculations using the full range of methods for excited state calculations reproduce the experimentalspectra and explain all observed features. In particular, they show that H atom formation occurs inexcited states, where the potential energy surface becomes repulsive along the O….H coordinateat relatively small distances. The loss of H2O, although thermochemically favorable, is a minorchannel because, at least for the clusters n= 1-3, the conical intersection through which the systemcould relax to the electronic ground state is too high in energy. In some absorption bands, sequentialabsorption of multiple photons is required for photodissociation. For n=1, these multiphoton spectracan be modeled on the basis of quantum chemical calculations. C2018 Author(s). All article content,except where otherwise noted, is licensed under a Creative Commons Attribution (CC BY) license(http://creativecommons.org/licenses/by/4.0/). https://doi.org/10.1063/1.5037401 l. INTRODUCTION Hydrated magnesium ions represent an interesting sys-tem to understand the mechanisms of hydrogen production viacatalysis on metal centersl-7 as well as corrosion effects. Atthe same time, they play a role in processes in the Earth’s andother planets’upper atmosphere where magnesium is presentdue to the influx of interplanetary particles.9.10 Hydrated metalions M(H2O)n are well-defined moieties to study the transi-tion of various properties from a metal atom solvated by asingle water molecule to bulk behavior.11-14 There has beena long history of studies on microhydrated metal ions inthe gas phase over the last decades. Especially, the hydratedmagnesium ion has attracted considerable attention over theyears15and is certainly among the best studied of thesesystems.16 First ab initio calculations to optimize structures anddetermine binding energies, electronic transition energies,and vibrational frequencies of Mg*(H20) were done byBauschlicher, Jr. in the early 1990s. Shortly afterwards, Dun-can and co-workers measured vibrationally resolved electronicand partially resolved rotational structures in the photodis-sociation spectrum of Mg*(H20).18,19 The results confirmed ( a)Authors to whom correspondence s h ould be addressed: milan.oncak@uibk. ac.at and martin.beyer@uibk.ac.at ) the C2y structure of the complex.19 Fuke and co-workers gen-erated hydrated singly charged magnesium ions with up to20 water molecules and found a dominance of Mg*(H2O)nfor n ≤ 5 and n ≥ 15, whereas MgOH*(H2O)n-1 speciesbeing almost exclusively present for n=6-14.20 Photodisso-ciation spectra for Mg*(H2O)n, n = 1-5, showed transitionscorrelating with 3s-3p excitations in Mg* and provided rel-ative cross sections and fragment branching ratios.20,21 Twodifferent dissociation processes were observed: water evapo-ration and an intra-cluster reaction forming MgOH(H2O)n-m(m 3:2tthis was however questioned in subsequent studies we discuss below. The bands in the experi-mental spectra were assigned to an s-p transition, and, for n> 2,the presence of different isomers in the spectra was sug-gested.26 For n>6, they proclaimed a negative energy for thehydrogen elimination process, explaining the product switch-ing to MgOH(H2O)n-1 observed in the experiments and forn > 14, Mg(H2O)nn was presented as a candidate toexplain the observed re-switching from MgOH+(H2O)n-ito Mg*(H2O)n in the ion formation above this thresh-old.25 Plowright et al. revisited Mg*X and Mg*XY sys-tems and presented structures and vibrational frequencies forMg*(H2O)1,2. Dunbar and Petrie simulated the formationof Mg*(H2O) by radiative association and found it to beinefficient even at T=10 K.2/ Berg et al.28.29produced Mg*(H2O)n up to n=80 andmeasured black body infrared radiative dissociation (BIRD)rates up to n = 41, as well as intracluster charge transfer(CT) processes and chemical reactions. The existence of asolvent separated ion pair of Mg2+ and a hydrated electron forn > 17 was proclaimed by Berg et a1.2 In the early 2000s,Reinhard and Niedner-Schatteburg investigated the electronicstructure of Mg(H2O)n with up to 20 water molecules.30.31For n ≤ 5, a quasi-valence state exists; for 6 ≤ n<17, acontact ion pair state exists; and for n ≥ 17, a solvent sepa-rated ion pair is formed.30 The existence of a hydrated electronand a Mg di-cation was proclaimed for n ≥8.30 Siu and Liuexplained the minimum cluster size for the hydrogen loss reac-tion based on ab initio molecular dynamics calculations andinvestigated the influence of the coordination number on theprocess. They also explained the switch off for the hydro-gen loss process for larger clusters due to the barrier increasewhen the solvated electron moves beyond the third solvationshell.15 Our group has a long history in studying the reactivityof hydrated metal ions,13,16,33-47 using Fourier transform ioncyclotron resonance mass spectrometry (FT-ICR-MS) whichis well suited for this task as it allows highest mass resolu-tion coupled with long storage times to allow multiple colli-sions with reactant gases. In the present study, we couple themass spectrometer with a tunable optical parametric oscillator(OPO)/amplifier system to conduct photodissociation experi-ments. Experimental results for Mg*(H2O)n, n=1-5, includephotodissociation cross sections and product branching ratios.The results are compared with earlier experiments. Theoreticalcalculations are used to model the spectra and to explain theobserved reactions on excited state potential energy surfaces. I. EXPERIMENTAL AND THEORETICAL METHODS The experimental setup has been described in detailearlier.48-51 The experiments were performed on a FourierTransform Ion Cyclotron Resonance Mass Spectrometer (FT-ICR-MS), equipped with a 4.7 T superconducting magnet.The setup contains a laser vaporization ion source, usingthe 2nd harmonic of a Nd: YAG laser to generate Mg* ionsvia vaporization of a rotating target disk, consisting of iso-topically enriched 24Mg (99.9%). Hydrated magnesium ionsMg*(H2O)n are formed via supersonic expansion into a highvacuum in a helium gas pulse seeded with water vapor, at a backing pressure of 20 bars. For storage and detection of theions, a liquid nitrogen cooled ICR cell (T~130±20K) wasused to minimize the influence of blackbody infrared radiativedissociation (BIRD).38,52-61 Ions of a specific mass to chargeratio were isolated via resonant ejection ofunwanted ions. TheMg*(H2O)n clusters were irradiated for typically 1 s by thebeam of a tunable wavelength, pulsed ultra violet/visible/near-infrared (UV/VIS/NIR) laser system (Nd:YAG pumped OPOsystem EKSPLA NT342 B-20-SH-SFG). Typical pulse ener-gies are shown in the supplementary material. Details on thelaser setup are available elsewhere.2 The photon flux insidethe cell was on the order of 0.1-1 mJ cm-2, with the strongwavelength dependence typical for OPO systems. Mass spec-tra of fragment and parent ions were recorded immediatelyafter irradiation. Relative photodissociation cross sectionswere calculated from the parent and fragment intensities andthe laser power using Lambert-Beer’s law, taking into accountthe contribution of BIRD to product formation (for details seethe supplementary material). The electronic ground state structures of the investi-gated ions were optimized with Moller-Plesset (MP2) Per-turbation theory and recalculated at the Coupled Clusterwith Single and Double and perturbative Triple excita-tions [CCSD(T)] level. The excited states were calculatedusing Time-Dependent Density Functional Theory (TDDFT),Equation of Motion Coupled Clusters Singles and Dou-bles (EOM-CCSD), Second-Order Approximate CoupledCluster (CC2),63 and Multireference Configuration Interac-tion (MRCI). The width of the spectra was modeled usingthe linearized reflection principle (LRP) within the harmonicapproximation.64-67 For smaller clusters (n=1-3), the stan-dard reflection principle was used along with sampling ofthe ground state density by path integral molecular dynam-ics (PIMD).68 The latter was performed with a step size of30 a.u., 17 000 steps per simulation out of which 3000 stepswere taken as an equilibration period, and 10 random walk-ers (see the supplementary material for benchmark calcula-tions).More than 700 points were used to construct eachspectrum. To account for vibrational levels disregarded withinthe reflection principle, the spectrum for n=1 was also treatedwithin the Franck-Condon approximation,accounting for theDuschinsky rotation69-71 with a full-widthathalf-maximum of270cm-. A basis set benchmark for modeling excited states isprovided in the supplementary material, showing that aug-cc-pVDZ gives reasonable accuracy for the lowest excited states.The triple-zeta aug-cc-pVTZ basis set or another larger basisset (see Table S1 of the supplementary material) is needed forhigher lying states, e.g., for the ninth excited state in doubletspin multiplicity (D9) and further states in Mg*(H2O). For ground-state optimizations as well as TDDFT andEOM-CCSD calculations, Gaussian72 was used; MRCI cal-culations were performed in Molpro,73 CC2 calculations inTurbomole,74 PIMD calculations in the Abin program. |ll. RESULTS AND DISCUSSION A. Experimental photodissociation spectra Measured photodissociation spectra for Mg(H2O)n,n = 1-5, at 130 ±20 K are shown in Fig. 1(a); experiments energy/eV FIG. 1. Photodissociationanddpho-toabsorption spectra of Mg*(H2O)nclusters. (a) Photodissociation spectrameasured during the present exper-iment.Note the relative intensitydifference of about 100 for the firstand second band of the Mg*(H2O)spectrum (see text for details). Forn=4, the spectral onset is also shown.(b) Photodissociation spectra recordedr 20by Misaizu eta1.20((c) Modeledphotoabsorptionnsspectra of isomersshown in Fig. 2. Linearized reflectionprinciple (LRP): Excitation energiesand oscillatorstrengthscalculatedatthee CC2/aug-cc-pVTZ level.ooftheory, frequencies and forces inthe excited states calculated at theCAM-B3LYP/aug-cc-pVTZ level.ReflectionprincipleWwith PIMD(RP/PIMD): Sampled on the B3LYP/6-31+g* potential energy surface, withexcitation energies calculated at theCC2/aug-cc-pVDZlevel. Franck-Condon principle (FCP): Calculated atthe EOM-CCSD/aug-cc-pVDZ level. by Misaizu et al.20 are shown in Fig. 1(b) for comparison.The detected fragment ions can be grouped into two differentchannels. The Mg*(H2O)m (m 1) can be explainedin two ways. First, water dissociation coordinate may beactivated within two-photon processes after reaching D4 orhigher states where enough energy is available due to struc-tural relaxation. Alternatively, it might be reached throughfluorescence back to the ground state or due to opening non-radiative channels to Do. Due to significant structural changesin the Di state for n> 1, the ion has enough energy to dis-sociate a water molecule after switching back into the Dostate. D. Simulation of the mixed one- and two-photonphotodissociation spectrum for Mg*(H2O) The analysis of potential energy surface scans has shownthat hydrogen dissociation in Mg*(H2O) might take placeafter absorption of two photons (initial excitation in Di, D2states) or one photon (initial excitation in D3), see schemein Fig. 4. Using this scheme, we modeled the photodissoci-ation spectrum by direct simulations of the ion interactionwith a laser pulse. For this purpose,the absorption spectrumfor the first photon was modeled using the Franck-Condonapproximation [Fig. 4(a)]. The absorption of the second pho-ton in the D and D2 states was calculated using the linearizedreflection principle in the respective excited state minima[Fig.4(b)]. To model the two-photon photodissociation spectra shownin Figs. 4(c) and 4(d), we simulated the direct interactionof the laser pulse with the ion based on the calculated pho-toabsorption spectra. We decomposed the laser pulse intofinite time steps of 10-12 s and approximated it with a Gaus-sian curve. We picked two different values of laser flux, 100and 0.4 J/m’/pulse (further denoted as high and low flux,respectively) that turned out to reproduce the experimentalspectra, see below. We modeled the total probability of pro-moting the system into the given state by integrating alongthe interaction of the laser pulse with the system in time. For 6e-16 FIG. 4. Modeled photodissociation spectra. Left-hand side: Scheme used tomodel the photodissociation spectra. Right-hand side: (a) Absorption spec-tra for the first photon calculated using the Franck-Condon approximation,decomposed into D1-D3 states. Calculated at the EOM-CCSD/aug-cc-pVDZlevel of theory.(b) Absorption spectra for the second photon calculated usingthe linearized reflection principle approximation for excitation starting fromDi and D2 minima. Calculated at the EOM-CCSD/aug-cc-pVDZ level oftheory, with the EOM-CCSD/aug-cc-pVTZ used for excitation energies andMRCI(1,14)/aug-cc-pVTZ for transition dipole moments between excitedstates. (c) Modeled photodissociation spectra for low laser flux in comparisonwith the present experimental results. (d) Modeled photodissociation spectrafor high laser flux in comparison with the data of Misaizu et al.20 the states that do not dissociate directly (D1, D2), the sec-ond photon can be absorbed within the remaining durationof the laser pulse or fluorescence might take place, with therate given by the respective Einstein coefficient. The moleculewas assumed to dissociate once the D3 or higher states werereached. Within the low laser flux limit, we simulate the situationwhen the Di/D2 states are rather little populated. With thislaser flux choice, we can reproduce the relative intensity of theDi/D2 and D3 absorption bands measured in our experiment[Fig. 4(c)]. While the position and width of the one-photonD3 band can be well reproduced by our calculation, the Di/D2band is off by about 0.2 eV and its structure is smeared. Underthe approximations used, we consider both spectra to be ingood agreement. For the high laser flux, we obtain considerably populatedDi/D2 states, with further excitation into higher states occur-ring within a pseudo one-photon process. Here, we are ableto reproduce the relative intensity of both bands as recordedin the experiment of Misaizu et al. [Fig.4(d)].20 With respectto the peak width and structure, the Di/D2 band seems to bebetter reproduced by our calculations, again with a shift ofabout 0.2 eV with respect to the maximum. Note that thisband is also broader compared to the one with the scaledphoton flux in Fig. 4(c), in agreement with the experimen-tal trend. The position and structure of the D3 band couldbe reproduced only semi-quantitatively. Here, the neglectof temperature influence represents an important source oferror. By comparing our experimental data with the theoreti-cal model and using the measurements of the Duncan groupthat recorded the onset of the D2 transition in Mg*(H20) at~3.75 eV,18 we can conclude that the first band sampled inour measurement is probably exclusively the D2 state, withDi absorption lying lower in energy and expected to be ofonly very limited intensity as shown in our photodissoci-ation spectra modeling [Fig. 4(c)]. In the previous exper-iment of the Fuke group,20 on the other hand, we expectboth Di and D2 states to be smeared into the first absorptionband. We have shown that different relative peak intensitiesin both experiments can be reproduced by considering lowand high photon flux interacting with the ions. This providesevidence for two-photon absorption in Di/D2 states and one-photon for D3. Based on this modeling and the potential energyscans presented in Fig. 3, we expect two-photon processes toplay an important role also for n=2,3. Here, the photodissocia-tion modeling is complicated by significant structural changesafter excitation. IV. CONCLUSIONS Relative photodissociation cross sections were measuredfor Mg*(H2O)1-5 clusters in the range of 0.6-5.0 eV. Theresults are overall in good agreement with the theoreticalpredictions, as well as earlier experiments, although the dis-sociation bands all seem to be shifted to the red, especiallyin the case of the Mg*(H20)3 cluster. A second dissocia-tion band was observed in this work for the Mg*(H2O)3-5 clusters. The BIRD influence was documented at cell temper-atures ofT~130±20 K,at the rate of2%-3% and 9%-17%for Mg*(H2O)4 and Mg*(H2O)5, respectively. It was foundthat for the Mg*(H2O)1-3 clusters, only a single isomer waspresent, whereas several isomers contribute to the dissocia-tion spectrum of Mg*(H2O)4,5. For all investigated clustersizes, hydrogen dissociation producing MgOH*(H2O)m wasthe main observed dissociation channel. By analysis of potential energy curves and photodisso-ciation spectra modeling, we have shown that two photonsare needed for hydrogen dissociation in Di and D2 states ofMg*(H2O). We argue that, for a high photon flux, absorp-tion of two photons proceeds in a pseudo one-photon regime.This behavior might also be expected for n =2,3 (as alreadyindicated for Mg*(H2O)2 in a previous study20). However,simulations for these systems will be more challenging due tothe large amount of energy released during structural changesafter excitation. SUPPLEMENTARY MATERIAL See supplementary material for method benchmarks,details on analysis of the experimental data, and Cartesiancoordinates of calculated ions and molecules. ACKNOWLEDGMENTS T.T. and M.K.B. acknowledge financial support from theAustrian Science Fund (FWF) through Project No. P29174.M.O. acknowledges the support through the Lise MeitnerProgramme of the FWF Project No. M2001-NBL. The com-putational results presented have been achieved using the HPCinfrastructure LEO of the University of Innsbruck. The tunableOPO systems are part of the Innsbruck Laser Core Facility,financed by the Austrian Federal Ministry of Science,Researchand Economy. V.Artero, M. Chavarot-Kerlidou, and M. Fontecave, Angew. Chem., Int.Ed. 50,7238(2011). C. Tard and C. J. Pickett, Chem. Rev. 109,2245(2009). S. Canaguier, V. Artero, and M. Fontecave, Dalton Trans. 25, 315(2008). “V. Fourmond, S. Canaguier, B. Golly, M. J. Field, M. Fontecave, andV. Artero, Energy Environ. Sci. 4, 2417 (2011). B. E. Barton, C. M. Whaley,T. B. Rauchfuss, and D. L. Gray,J. Am. Chem.Soc. 131, 6942 (2009). R. Brimblecombe, G. F. Swiegers, G. C. Dismukes, and L. Spiccia, Angew.Chem., Int. Ed. 47,7335(2008). Y. Xu, T. Akermark,V. Gyollai, D. Zou, L. Eriksson, L. Duan, R. Zhang,B. Akermark, and L. Sun, Inorg. Chem. 48,2717 (2009). R. Burgert, H. Schnockel, A. Grubisic, X. Li, S. T. Stokes, K. H. Bowen,G. F. Gantefor, B. Kiran, and P. Jena, Science 319, 438 (2008). R. J. Plowright, T. J. McDonnell, T. G. Wright, and J. M. C. Plane, J. Phys.Chem. A 113, 9354 (2009). ( 18K. F . Willey, C . S. Y eh, D. L . R o bbins, J. S. Pilgrim, and M. A. Dun c an, J . Chem . P h ys . 97, 8 886 (1992). ) ( 19c. S. Yeh, K. F. Willey, D. L. Robbins, and M. A. D uncan, I nt. J. Mass Spectrom. Ion Processes 131, 307 (1994). ) ( 20F. Misaizu, M. Sanekata, K. Fuke, and S. Iwata , J . Che m . Ph ys. 100, 1 161 ( 1 994). ) ( 21F. Misaizu, M. Sanekata, K. Tsukamoto, K . F u ke, a n d S. I w ata, J . P hys. Chem. 9 6, 8259 (1992). ) ( 22M. Sanekata,F. Misaizu, K. Fuke,S. Iwata, and K. Hashimoto,J . Am. Chem. Soc. 117,747 (1995). ) ( 23Y. Inokuchi, K . Ohshimo, F. Misaizu, and N. Ni s hi, J. P h ys. C h em. A 1 0 8, 5034 (2004). ) ( 24Y. Inokuchi, K. Ohshimo, F. M isaizu, and N . N ishi, Chem. Phys. L ett. 390, 1 40 (2004). ) ( 25H. W atanabe, S. Iwata, K . H ashimoto, F . Misaizu, and K. F u ke, J . A m. C h em. S oc. 117,755 (1995). ) ( 26H. Watanabe and S. Iwata, J . C h em. Phys . 1 08,10078(1998). ) ( 2 7R. C. D unbar and S. Petrie, J . Phy s. C h em . A 1 09 , 141 1 (2 00 5). ) ( 28C. B erg, U . A chatz, M. Be y er, S. Joos, G. Albert, T. S c h indler, G. N iedner-Schatteburg, and V. E. Bondybey, Int. J. Mass Spectrom . 167, 723 (1997). ) ( 29C. B erg, M . B e yer, U . Ac h atz, S. Joos, G. Nie d ner-Schatteburg, and V. E. Bondybey, Chem . Phy s. 239,379 (1998). ) ( 3B. M. Reinhard and G. Niedner-Schatteburg, Phys . C hem. Chem. Ph y s. 4 , 1471 (2002). ) ( B. M. Re i nhard an d G. Ni e dner-Schatteburg, J. C he m . P h ys . 1 1 8, 3 5 71 (2003). ) ( 3 2C. K . Siu and Z. F. Liu, Che m . - E u r. J . 8 , 3177 (2002). ) ( 3 3B. S. F ox-Beyer, Z. Sun, I. B a lteanu, O. P. Balaj, and M. K. Bey e r, Phys . C h e m. C hem. P h ys . 7 , 981 (2005). ) ( 3 4C. v an der Linde, A . A k hgarnusch, C.-K. Siu , and M. K . B e ye r , J. P h ys . Chem. A 115,10174(2011). ) ( 35C. van der Linde a nd M. K . B e yer, P h ys . C h em. C h e m. Phy s . 1 3, 6776 (201 1 ). ) ( 36M. B eyer, C. B erg, H . W. Gorlitzer, T. Schindler, U. Achatz, G. Albert, G. N iedner-Schatteburg, and V . E. Bondybey,J. A m. C h em. S o c. 11 8 ,7386 (1996). ) ( 37M. B eyer, U. A c hatz, C. Be r g, S. Joos, G. Nie d ner-Schatteburg, and V . E. Bondybey, J . P hy s . C hem. A 103,671 ( 1999). ) ( 38O. P. B alaj, C. B. B erg, S . J . Reitmeier, V. E. Bondybey, and M. K. Beyer, I n t. J . Mass Spectrom. 279, 5( 2 009). ) ( 39C. van der Linde and M. K. Beyer, J . P hy s . Che m. A 116,10676 (2012). ) ( 4 0B. Scharfschwerdt, C. va n der Linde, O. P . Balaj, I. Herber,D. Sch u itze, and M . K. B eyer, Low T emp. Phys. 38,717(2012). ) ( 4 I. Herber, W.-K. Tang,H.-Y.Wong, T.- W . Lam, C.-K. Siu, and M. K. Beyer, J. Phys . Chem. A 1 19, 5566 (2015). ) ( 42I. Gernert and M. K. Beyer, J . Phy s . Che m . A 121, 9557(2017). ) ( 43C. van d er L inde, S . H emmann, R. F. H o c kendorf, O. P . B al a j, and M .K. Beyer, J. Phys. C he m. A 1 17,10 1 1 (2013). ) ( 44C. van d er L inde, R . F. H ockendorf, O . P . B a laj, and M. K. Be y er, C hem. - Eur. J . 19,3741 (2013). ) ( 45B. S. Fox, O. P . Balaj, I. Balteanu,M. K . B eyer, and V. E. Bondybey, Chem . - E u r . J. 8 , 5534 (2002). ) ( 4B. S. Fox, O . P. Balaj, I . Balteanu,M. K . Beyer, a n d V. E . B o ndybey,J. Am . Chem. S oc. 1 24, 1 72 (2002). ) ( 47B. S. Fox, M. K. Beyer, and V. E. Bondybey,J . Am. C hem. S o c. 1 2 4, 13613 (2002). ) ( 48R. F. Hockendorf, O. P. Balaj, C. van der Linde , an d M. K. Beyer, P hy s. Chem. Chem. P hy s. 1 2,3772 (2010). ) ( 49P. Caravatti and M. Allemann, Org. M ass Spectrom. 2 6 , 5 1 4 (1991). ) ( 50P. Kofel, M. Allemann, H. Ke l lerhals, and K. P. Wanczek, Int. J . M a s s S pect r om. I o n P r ocesses 72 , 5 3 ( 1 986). ) ( C. Berg,T. Schindler,G.N i edner-Schatteburg, and V. E. Bondybey,J. Chem. P hys. 102, 4870 (1995). ) ( 52R. C. Dunbar, Mass S pectrom. R ev. 23, 127 (2004). ) ( 53B. S. Fox, M. K. Beyer, and V. E. Bondybey, J . Phys. C hem. A 1 05,6386 (2001). ) ( 54R.C. Dunbar and T . B. McMahon, Sci en c e 27 9 , 194 (1998). ) ( 5 50. H ampe, T. Karpuschkin, M . V o nderach, P. Weis, Y . M. Y u , L. B . G an, W. K lopper, and M. M. K appes, P hy s. Ch e m. C hem. P h ys. 1 3 , 9 818 (2011). ) ( 56s. W. L ee, P. Freivogel, T. Schindler , and J. L. Beauchamp, J . Am. Chem. S oc. 120,11758(1998). ) ( 5T.Schindler, C. Berg, G. Niedner-Schatteburg, and V. E. Bondybey, C hem. P hys . L ett. 250, 301 (1996). ) ( 58P. D. Schnier, W. D . Price, R . A . Jockusch, a n d E. R. Williams, J. A m . Chem. S oc. 118,7178 (1996). ) ( 59M. S ena a nd J . M. Riveros, Rapid Comm u n. M as s S p e ctrom. 8 , 1 031 (1994). ) ( 60D. Tholmann, D. S. Tonner, and T. B. McMahon, J. P hys . Chem. 9 8 , 2002 (1994). ) ( R . L. Wo n g, K. P aech, and E. R . W i lliams, Int. J. M a s s Spectrom. 23 2 , 59 (2004). ) ( 62A. Herburger, C. van der Linde, and M. K. Beyer, Phy s. Che m . Ch e m. Phys. 19, 1 0786(2017). ) ( 63C. Hattig and A. Kohn, J . Chem. Phys. 117, 6939 (2002). ) ( 64R. S chinke, P hotodissociation D ynamics (C a mbridge University Press, Cambridge, 1993). ) ( 65S. Y. Lee, R. C . Brown, a n d E. J. H e ller, J . P hys. C h e m. 8 7 , 2045 (1983). ) 66S.-Y. Lee, J. Chem. Phys.82,4588(1985). ( 67M. K. Prakash, J. D. Weibel, and R. A. Marcus, J . Geop h ys. Res. 110, 380, https://doi.org/10.1029/2005jd006127 (2005). 58 ) ( 68M. O ncak, L. S i stik, and P . S lavicek, J . Chem. P hys . 1 33, 17 4303 (2010). ) ( 69J. Franck and E . G. D ymond, Tr a ns. F a r a day Soc. 21 , 53 6 (192 6 ). ) 70E. U.Condon, Phys. Rev. 32, 858(1928). ( 7V. Barone, J. Bloino,M. Biczysko, and F. Santoro , J . Che m . T heory Compu t. 5,540(2009). ) ( 7M.J. Frisch, G. W . Trucks, H . B . Schlegel, G . E . Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. P etersson, H . N akatsuji, M.Caricato,X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino,G. Z heng, J.L. Sonnenberg, M . Hada, M. Ehara, K. Toyota, R. Fu k uda, J. Hasegawa, M . I shida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. V reven, J. J . A . M o ntgomery, J. E. Peralta, F. Ogliaro, M. Be a rpark,J. J. Heyd, E. Brothers, K. N. Kudin, V.N. S taroverov, R. Kobayashi,J. Normand, K . Ra g havachari, A. Ren d ell, J. C . B ur a nt, S. S. Iye n gar,J. Tomasi, M. C ossi, N. Rega, J . M. M illam, M. K l ene, J . E. Knox,J. B. Cross, V. Bakken, C. A damo, J . Jaramillo, R. Gomperts, R . E. S tratmann,O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli,J. W.Ochterski, R . L. Martin, K. Morokuma, V. G . Z a krzewski, G. A. Voth, P. Salvador, J .J. Dannenberg, S. D a pprich, A. D. Daniels, O. Far k as, J. B. Foresman,J. V. Ortiz, J. Cioslowski, and D. J . Fox, GAUSSIAN 0 9, R evision D . 01,Gaussian, Inc., Wallingford , CT, 2013. ) ( 73H.-J. Werner,P . J. K nowles, R . Lindh, F. R. M a nby, M. Schutz, P. Ce lani, T. Korona, A . M itrushenkov, G . Rauhut, T. B. A d ler, R. D . A m os,A. B ernhardsson, A. B erning,D. L. Cooper, M. J. O. D eegan, A. J. D obbyn,F. Eckert, E. Goll, C. Hampel, G. Hetzer, T. Hrenar, G. Knizia, C. Koppl, Y . Liu, A. W. Lloyd, R. A. M ata, A . J. May, S . J. McNicholas, W. Meyer, M . E. Mura, A . N i cklaB, P. P a lmieri, K. P f luger, R . P itzer, M. R eiher,U. Schumann, H. Stoll, A. J. Stone, R. Ta r roni, T. Thorsteinsson, M. Wang,and A. Wolf, MOLPRO, version 2012.1, a package of ab initio programs, 2012, see h t t p://www. mo lpro.net . ) ( 74TURBOMOLE V6.2 2010, a development of University of Karlsruhe and Forschungszentrum Karlsruhe GmbH, 1989-2007,TURBOMOLE GmbH, s ince 2007; available from h ttp://www.turbomole . com. ) ( 75D. H ollas, O. Svoboda, M. O ncak, and P. S lavicek, ABIN, S ource code available at h t t ps :/ /gi t hub.com/PHOTOX / ABIN. ) ( 76C. S. Yeh, K . F . Willey, D. L. Robbins,J. S. P ilgrim, and M. A. Duncan, C hem . P h y s. L e t t. 196, 2 33 ( 1 992). ) ( 7R. M. F o rck, I. Dauster, Y. Schieweck, T . Zeuch, U. Buck, M. Oncak, an d P . Slavicek, J . Chem. P hy s. 132, 221102 (2010). ) ( 78M. S anekata, F. M isaizu, a nd K. Fuke, J . Chem. P h y s . 1 0 4 , 97 6 8 (1996). ) O Author(s) / Hydrated singly charged magnesium ions Mg+(H2O)n, n 5, in the gas phase are ideal model systems to study photochemical hydrogen evolution since atomic hydrogen is formed over a wide range of wavelengths, with a strong cluster size dependence. Mass selected clusters are stored in the cell of an Fourier transform ion cyclotron resonance mass spectrometer at a temperature of 130 K for several seconds, which allows thermal equilibration via blackbody radiation. Tunable laser light is used for photodissociation. Strong transitions to D1–3 states (correlating with the 3s-3px,y,z transitions of Mg+) are observed for all cluster sizes, as well as a second absorption band at 4–5 eV for n = 3-5. Due to the lifted degeneracy of the 3px,y,z energy levels of Mg+, the absorptions are broad and red shifted with increasing coordination number of the Mg+ center, from 4.5 eV for n = 1 to 1.8 eV for n = 5.In all cases, H atom formation is the dominant photochemical reaction channel. Quantum chemical calculations using the full range of methods for excited state calculations reproduce the experimental spectra and explain all observed features. In particular, they show that H atom formation occurs in excited states, where the potential energy surface becomes repulsive along the O H coordinate at relatively small distances. The loss of H2O, although thermochemically favorable, is a minor channel because, at least for the clusters n = 1-3, the conical intersection through which the system could relax to the electronic ground state is too high in energy. In some absorption bands, sequential absorption of multiple photons is required for photodissociation. For n = 1, these multiphoton spectra can be modeled on the basis of quantum chemical calculations.