Photofragmentation Dynamics in Solution Probed by Transient IR

Letter pubs.acs.org/JPCL

Photofragmentation Dynamics in Solution Probed by Transient IR Absorption Spectroscopy: πσ*-Mediated Bond Cleavage in p‑Methylthiophenol and p‑Methylthioanisole Daniel Murdock,† Stephanie J. Harris,† Tolga N. V. Karsili,† Gregory M. Greetham,‡ Ian P. Clark,‡ Michael Towrie,‡ Andrew J. Orr−Ewing,† and Michael N. R. Ashfold*,† †

School of Chemistry, University of Bristol, Cantock’s Close, Bristol, BS8 1TS, United Kingdom Central Laser Facility, Research Complex at Harwell, Science and Technology Facilities Council, Rutherford Appleton Laboratory, Didcot, Oxfordshire, OX11 0QX, United Kingdom

‡

S Supporting Information *

ABSTRACT: The 267 nm photodissociation dynamics of p-methylthiophenol (pMePhSH) and p-methylthioanisole (p-MePhSMe) dissolved in CD3CN have been probed by subpicosecond time-resolved broadband infrared spectroscopy. Prompt (τ < 1 ps) S−H bond fission in p-MePhSH is confirmed by monitoring the time-evolution of the parent (S0) bleach and the transient absorption of the p-MePhS products. Vibrational relaxation of the latter occurs on a ∼8.5 ps time scale, and ∼40% of the total radical population undergoes geminate recombination over a ∼150 ps time scale, yielding (mainly) the p-MePhSH(S0) parent. S−Me bond fission following photoexcitation to the S1 state of p-MePhSMe occurs over a much longer timescale, with a rate that is very dependent on the degree of vibrational excitation within S1. The various findings are compared and contrasted with results from complementary gas-phase photofragmentation studies of both molecules, which are shown to provide a valuable starting point for describing the solution-phase dynamics. SECTION: Spectroscopy, Photochemistry, and Excited States

T

resolved UV/vis transient absorption (TA) studies of pMePhSH (and its photolysis products) in ethanol and cyclohexane showed that the dissociation mechanism is largely immune to the presence of solvent molecules, with S−H bond fission following excitation at 267 nm deduced to occur within the instrument response time (∼50 fs). Solution-phase specific effects (e.g., vibrational cooling, geminate recombination) were evident, but no alternative relaxation pathways were observed (as gauged by the absence of any signal attributable to solvated electrons). Here we illustrate the utility of time-resolved infrared (TRIR) spectroscopy for exploring the solution-phase photofragmentation dynamics of molecules whose gas-phase photochemistry and photophysics are well-established using the 267 nm photolysis of p-MePhSH and its alkylated analogue pmethylthioanisole (p-MePhSMe; Figure 1b) as exemplars. Figure 2a depicts a series of TRIR spectra obtained following 267 nm photolysis of a 45 mM solution of p-MePhSH dissolved in CD3CN. Two spectral features dominate at all delays: a bleach centered around 1495 cm−1 (with a shoulder extending to ∼1455 cm−1) reflecting the depletion of ground-state population induced by the pump pulse, and an absorbance at ∼1570 cm−1 attributable to the ν7(a′) mode (ν8a in Wilson

he nonradiative relaxation of electronically excited heteroaromatic molecules in the gas phase has attracted much recent attention, with the result that the important role of optically “dark” n/πσ* electronic states is becoming increasingly recognized.1−4 The potential energy surfaces (PESs) of states formed by σ* ← n/π excitation are typically repulsive with respect to X−Y (X = O, N, S, etc.; Y = H, CH3, etc.) bond extension and thus offer a route to X−Y bond fission or to internal conversion (IC) via conical intersections (CIs) located at extended X−Y bond lengths, RX−Y. The applicability of this paradigm to dissociation processes in the condensed phase is a question of considerable interest. To what extent does the presence of a solvent bath promote alternative, solution-phase specific, relaxation mechanisms (e.g., autoionization, proton coupled electron transfer, etc.)5 or influence the CIs at extended RX−Y (i.e., extending into the solvent shell) and thus affect the relative radiative/nonradiative quantum yields or quantum-state distribution of the radical products? Two recent studies have sought to compare and contrast the gas- and solution-phase photodissociation dynamics of pmethylthiophenol (p-MePhSH; Figure 1a) for a range of UV photolysis wavelengths.6,7 In the gas-phase, prompt S−H bond fission can be inferred from the measured vibrational energy disposal in the ground (X̃ ) and electronically excited (Ã ) state p-MePhS radical products, and from the evident recoil anisotropy of the H atom fragments. Complementary time© 2012 American Chemical Society

Received: October 30, 2012 Accepted: November 28, 2012 Published: November 28, 2012 3715

dx.doi.org/10.1021/jz301755p | J. Phys. Chem. Lett. 2012, 3, 3715−3720

The Journal of Physical Chemistry Letters

Letter

Figure 1. Cuts along the RS−Y stretch coordinate (with all other nuclear degrees of freedom held at their respective ground-state equilibrium values) for the ground and first two singlet electronically excited states for (a) p-MePhSH and (b) p-MePhSMe calculated at the CASPT2(10/10)/aug(S)cc-pVTZ level of theory. The electronic-transition characters and locations of the S2/S1 and S2/S0 CIs are denoted. The highest occupied molecular orbital in these two molecules involves (different) admixtures of sulfur 3px (n) and ring-centered πx character, but, for simplicity, we label the S2 state 1 πσ* in each case.

notation) of the p-MePhS(X̃ ) radical.8 The former is noteworthy because it allows direct probing of the loss and recovery of the ground-state population (unlike the UV/vis TA studies of this molecule). The prompt appearance of the radical product, and the observed vibrational cooling (on a picosecond time scale, as revealed by the simultaneous narrowing and blueshifting of the ∼1570 cm−1 feature with increasing pump− probe time delay, δt) agrees well with the previous TA results.6,7 The bleach feature recovers partially at longer times, indicating repopulation of the parent S0 state. A small absorption signal at ∼1475 cm−1 is evident at early times but decays rapidly and has disappeared by δt ≈ 40 ps. This shortlived transient is likely to be attributable to electronically excited p-MePhSH or to vibrationally excited p-MePhSH(S0) molecules (i.e., photoprepared p-MePhSH molecules that internally convert to “hot” p-MePhSH(S0) molecules and subsequently vibrationally relax). Three weak features centered around 1425, 1520, and 1625 cm−1 develop at later times. Given their delayed appearance, we assign these as signatures of p-MePh(H)S (i.e., an adduct formed by the H atom recombining on the aromatic ring rather than at the S atom), as invoked in the TA studies of this molecule.6,7 Figures 2b−e present kinetic traces along with accompanying fits obtained from analysis of the TRIR spectra of p-MePhSH. The time evolution of the red wing of the 1570 cm−1 feature provides information about the relaxation of the vibrationally excited p-MePhS(X̃ ) radical products (Figure 2b, obtained by numerical integration of the IR signal in the range 1535−1550 cm−1). The resultant data fit satisfactorily to a singleexponential function with a rate constant kvib = 0.118(7) ps−1 (where the number in parentheses represents one standard deviation uncertainty in the last significant digit(s)), corresponding to a mean vibrational lifetime of 8.5(5) ps. Numerical integration over the entire width of the 1570 cm−1 feature provides an estimate of the total population of p-MePhS(X̃ ) products (assuming that radicals with v = 0 and with v > 0 have similar IR absorption cross sections when exciting the

v7 = 1 ← 0 transition). Figure 2c shows a ∼40% reduction in the p-MePhS(X̃ ) radical population by δt = 150 ps, consistent with geminate recombination of this fraction of the primary photofragments to reform the ground-state parent molecule or form a p-MePh(H)S adduct. Further evidence of geminate recombination is provided by the kinetic traces obtained by integrating the time-delayed features at ∼1425 and ∼1625 cm−1 (Figure 2d), which both grow at rates mirroring the loss of pMePhS(X̃ ) products. Simultaneous fitting of these three kinetic traces to single-exponential functions yields a best-fit geminate recombination rate constant kgem = 0.037(2) ps−1 (τgem = 27(2) ps), consistent with that obtained in the TA studies of this same photolysis (in ethanol).6 Interrogation of the p-MePhSH(S0) population (peak at 1495 cm−1) is hampered by the overlapping absorption feature at small δt (vide supra), which has the effect of decreasing the bleach intensity. However, numerical integration over the bleach signal (Figure 2e) shows that the longer time kinetics (δt > 10 ps) are well-described by a single exponential function with the same rate constant as that for p-MePh(H)S adduct formation. Because geminate recombination is a diffusion-controlled process (with the bond-forming step assumed to be fast once the radical pair approaches to within a certain encounter radius), the equivalence of these rate constants thus suggests that recombination (rather than IC) is the major route to repopulating p-MePhSH(S0). Furthermore, the observation that the reduction in bleach intensity (∼40%) by δt ≈ 150 ps matches the concurrent decline in radical absorption (Figure 2c) implies that most recombination leads to p-MePhSH, with p-MePh(H)S adduct formation being a relatively minor pathway. The present results and those from the recent TA studies6,7 imply prompt S−H bond fission following 267 nm excitation of p-MePhSH in solution, as in the gas phase. This can be rationalized by inspecting cuts through the relevant PESs (Figure 1a), which reveal that the S2(1πσ*) PES intersects that of the S1(1ππ*) state near its minimum. The potential barrier 3716

dx.doi.org/10.1021/jz301755p | J. Phys. Chem. Lett. 2012, 3, 3715−3720

The Journal of Physical Chemistry Letters

Letter

Figure 2. (a) TRIR spectra measured at different δt following 267 nm photoexcitation of a 45 mM solution of p-MePhSH in CD3CN. Numerical integration over portions of these spectra allows extraction of kinetic traces (open symbols) monitoring the (b) vibrational cooling of p-MePhS(X̃ ; 1535−1550 cm−1), (c) overall population of p-MePhS(X̃ ; 1535−1590 cm−1), (d) buildup of p-MePh(H)S adduct peaks at 1425 (red circles) and 1625 cm−1 (green squares), and (e) p-MePhSH(S0; 1495 cm−1) population. The solid blue lines are least-squares fits of the data to singleexponential functions.

the spectral signature of the S1 state population. The evident blue shift with increasing δt implies that the initial photoexcitation yields vibrationally excited S1 molecules, which then relax (through collision with solvent molecules) to the S1; v = 0 level. The time-evolving p-MePhS(X̃ ) population out to δt = 1600 ps (Figure 3b) shows a rapid initial buildup, followed by a slower, secondary production. The kinetics of the p-MePhSMe(S0) and p-MePhSMe(S1) features are harder to gauge due to the substantial overlap of their respective spectra, which necessitates a deconvolution using model functions. The kinetic traces shown in Figure 3c−e were obtained as follows: the solution-phase Fourier transform IR spectrum of pMePhSMe was used as a guide for the p-MePhSMe(S0) bleach signal, whereas the p-MePhSMe(S1) peak was modeled as a sum of two Gaussians (one representing molecules with v > 0, the other describing p-MePhSMe(S1; v = 0) absorption, with the respective peak centers and widths held fixed through simultaneous fitting of all available data). Figure 3c depicts the amplitude of the Gaussian function modeling the p-MePhSMe(S1; v > 0) population, which affords a measure of the rate of vibrational cooling within the S1 manifold. The p-MePhSMe(S1; v = 0) population (Figure 3d) increases at small δt due to vibrational cooling of S1; v > 0 molecules but thereafter declines slowly as a result of IC to the S0 state, dissociation, or both. The

under the S1/S2 CI thus presents minimal impediment to dissociation at all excitation wavelengths (i.e., at all levels of vibrational excitation within the S1 state), and bond fission is fast. In concluding this section, we note one difference between the gas- and solution-phase data: no transients attributable to pMePhS(Ã ) products are identified in the present TRIR data, yet these constitute ∼40% of the total radical yield in the 267 nm photolysis of gas-phase p-MePhSH. We return to consider this difference after reviewing the corresponding data for pMePhSMe. Cuts through the S0, S1, and S2 PESs for p-MePhSMe (i.e., along RS−Me, Figure 1b) reveal a much larger barrier under the S1/S2 CI than is the case for p-MePhSH. TRIR spectra obtained following 267 nm photolysis of a 11.25 mM solution of pMePhSMe in CD3CN are shown in Figure 3a. By analogy to the corresponding p-MePhSH data (Figure 2a), the bleach centered at ∼1495 cm−1 is attributable to UV-pump-induced depletion of the ground-state population, and the small absorption signal at ∼1570 cm−1 is attributable to p-MePhS(X̃ ) products. In contrast with p-MePhSH, this radical signal is absent at the earliest time delays, and we see no evidence of vibrational cooling. The early time spectra are dominated by a broad absorption signal peaking at ∼1485 cm−1. Given the long lifetime of this signal, we eliminate vibrationally excited pMePhSMe(S0) molecules as the carrier and conclude that this is 3717

dx.doi.org/10.1021/jz301755p | J. Phys. Chem. Lett. 2012, 3, 3715−3720

The Journal of Physical Chemistry Letters

Letter

Figure 3. (a) TRIR spectra measured at different δt following 267 nm photoexcitation of a 11.25 mM solution of p-MePhSMe in CD3CN. (b) Timedependent population of the p-MePhS(X̃ ) radical photoproduct obtained through numerical integration. Deconvolution of the region between 1450 and 1510 cm−1 in terms of model functions enables extraction of kinetic traces for the (c) vibrational cooling of p-MePhSMe(S1; v > 0) [note the different time scale], (d) population of p-MePhSMe(S1; v = 0), and (e) p-MePhSMe(S0) population. The solid blue lines are least-squares fits of the data to analytical functions obtained from the kinetic scheme described in the text.

kinetic trace obtained from analysis of the p-MePhSMe(S0) bleach signal (Figure 3e) shows the expected recovery as a result of IC from the S1 state. The apparent sharp decrease in S0 population at short delay times is likely to be an artifact of the deconvolution process and probably reflects an underestimation of S1; v = 0 population at the very earliest delays (δt < 5 ps). The biexponential nature of the kinetic trace shown in Figure 3b suggests that p-MePhSMe(S1) molecules with v > 0 yield pMePhS(X̃ ) radicals with a different (faster) rate than those with v = 0. Such a conclusion concurs with that reported in recent femtosecond time-resolved velocity map ion imaging studies of gas-phase thioanisole (PhSMe),9 which found time constants of 1.4 ns following excitation centered at 289.8 nm (i.e., the S1 − S0 origin), cf. 74 ps when exciting at 275.0 nm. This encourages fitting of the present TRIR data according to the following kinetic scheme:

dissociation rate constants for vibrationally excited and vibrationally cold p-MePhSMe(S1) molecules, respectively. This kinetic scheme neglects possible repopulation of pMePhSMe(S0; v = 0) or loss of p-MePhS(X̃ ) through geminate recombination. Under the assumption that photoexcitation populates S1; v > 0 levels only, analytical solutions for the timedependent populations of the four species of interest were obtained,10 and the solid lines depicted in Figure 3b−e show fits to the experimental data using the best-fit rate constants: kVib = 0.054(2) ps−1, kDiss(v>0) = 0.042(2) ps−1, kDiss(v=0) = 0.00179(7) ps−1; and kIC = 0.00066(6) ps−1. The dissociation rate constants returned by the kinetic fits are orders of magnitude smaller than in the case of p-MePhSH, as expected given the potential barrier impeding access to the dissociative S2 state (Figure 1b). Recent studies of the gasphase photodissociation dynamics of p-MePhSMe and related thioanisoles (unpublished results) have highlighted the importance of out-of-plane S−Me torsional motion in promoting nonadiabatic coupling between the S1 and S2 states and subsequent S−Me bond fission. 267 nm excitation provides p-MePhSMe with enough energy and a sufficient density of vibrational states to access overtones and combination levels involving such torsional distortion − directly, or after intramolecular vibrational redistribution, leading to τDiss(v>0) =24(1) ps. S1; v = 0 molecules, in contrast, sample a more limited range of configuration space, and most transfer to the S2

where kIC denotes the rate of S1; v = 0 → S0; v ≫ 0 IC (vibrational cooling within the S0 manifold is expected to be much faster than the IC rate and will thus have minimal impact on the observed rate constant), kVib describes the vibrational cooling within the S1 manifold, and kDiss(v>0) and kDiss(v=0) are 3718

dx.doi.org/10.1021/jz301755p | J. Phys. Chem. Lett. 2012, 3, 3715−3720

The Journal of Physical Chemistry Letters

■

Letter

EXPERIMENTAL AND COMPUTATIONAL METHODOLOGY The TRIR spectra were recorded using the ULTRA laser facility at the Rutherford Appleton Laboratory.12 An amplified titanium sapphire laser system generated 800 nm (band center) pulses with 50 fs pulse duration and 10 kHz repetition rate. A portion of this light was used to pump an optical parametric amplifier to produce mid-IR radiation of ∼500 cm−1 bandwidth, whereas some of the light was frequency tripled to generate the requisite 267 nm photolysis radiation. The two optical pulses were overlapped in the sample with their linear polarization vectors aligned at the magic angle, before the transmitted IR radiation was dispersed by a grating onto a 128 element mercury cadmium telluride array detector. The ∼1 ps experimental response function is limited by background noise induced by the cell windows and solvent, thus masking the true instrumental response time of ∼100 fs. p-MePhSH (98%), p-MePhSMe (99%), and CD3CN (99.96 atom % D) were obtained from Sigma-Aldrich and used without further purification. The samples were flowed continuously through a Harrick cell with a 100 μm PTFE spacer between CaF2 windows at solution concentrations chosen to ensure an absorbance, A ≈ 0.5, at 267 nm. The ab initio potential energy curves presented in Figure 1 were calculated at the CASPT2(10/10)/aug(S)-cc-pVTZ level of theory using Molpro2010.1.13

PES is assumed to occur in regions of phase space sampled by the outer extremities of the S1; v = 0 wave function. Such an interpretation also accords with the finding that the dissociation rate of solvated p-MePhSMe(S1) molecules (τDiss(v=0) = 560(20) ps) is faster than that reported for jet-cooled gasphase PhSMe(S1) molecules (τDiss(v=0) ≈ 1.4 ns),9 given that the ∼300 K solvent bath will provide thermal energy sufficient to ensure some steady-state population of the low-frequency S− Me torsional mode at all times (excitation that is not available in the gas-phase sample). The photodissociation dynamics of p-MePhSH and pMePhSMe in solution show many similarities with those found in the gas phase, but the present data leave open one major question: the relative yield of à state p-MePhS products. These constitute ∼40% of the total radical yield in the 267 nm photolysis of gas-phase p-MePhSH6 and are the dominant radical product in the case of gas phase PhSMe11 and pMePhSMe (unpublished results). The electronic branching in the radical is determined at extended RS−Y (∼3 Å): molecules with near-planar geometries can follow the diabatic route through the S2/S0 CI to X̃ state products, whereas those with the S−Y bond directed out of the ring plane follow the adiabatic path to à state products. Non-observation of the latter in the present solution-phase experiments might reflect some solvent-induced changes to the S2 and S0 PESs, or solventinduced funneling of population toward the local minimum of the S2/S0 CI. Although appealing, the present data are not sufficient to demonstrate solvent-induced modification of the primary dissociation dynamics. The recent UV/vis TA studies of p-MePhSH photolysis at 267 nm assign very short-lived features to à -state radical formation,6 and the non-observation of any such transient signals in the present TRIR studies may simply reflect the efficiency of à state quenching and the background limited experimental response time. In summary, TRIR spectroscopy has been used to probe the 267 nm photodissociation dynamics of p-MePhSH and pMePhSMe in solution in CD3CN. In both cases, the early time dynamics closely mimics the corresponding gas-phase data. pMePhSH dissociates promptly, little or no S1 population is detected, and the p-MePhS(X̃ ) product yield peaks within the experimental response time. The time dependence of the radical absorption signal yields time constants for vibrational relaxation within the radical and for geminate recombination; the latter time constant is confirmed by following the time dependence of the S0 bleach signal. Geminate recombination also yields an alternative p-MePh(H)S adduct (in which the returning H atom is attached to the ring rather than the S atom), but the available evidence suggests that this is a minor pathway. p-MePhSMe displays richer dynamics consistent with a hindered dissociation process. The TRIR spectra show a clear absorption signal attributable to S1 population, the time dependence of which shows that the rate of S−Me bond fission scales with vibrational energy content. This, too, mirrors behavior identified in recent gas-phase photolysis studies of PhSMe9 and can be understood by recognizing the importance of torsional motion in promoting nonadiabatic coupling from S1 to the dissociative S2 PES. No signals attributable to à -state radical products were observed with either molecule, a somewhat surprising result given their prevalence in the corresponding gas-phase photodissociations, but this may simply reflect the efficiency of à -state quenching in collision with solvent molecules.

■

ASSOCIATED CONTENT

S Supporting Information *

More detailed description of the deconvolution procedure used to disentangle the kinetics of p-MePhSMe(S1) and pMePhSMe(S0). This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Funding from EPSRC (Programme Grant EP/G00224X) is gratefully acknowledged, as are contributions from A. M. Wenge, S. E. Bradforth, and T. A. A. Oliver. M.N.R.A. and S.J.H. thank, respectively, the Royal Society for the award of a Leverhulme Trust Senior Research Fellowship and the University of Bristol for a Postgraduate Research Scholarship. We acknowledge STFC for programme access to the ULTRA laser facility (STFC Facility Grant ST/501784).

■

REFERENCES

(1) Ashfold, M. N. R.; King, G. A.; Murdock, D.; Nix, M. G. D.; Oliver, T. A. A.; Sage, A. G. πσ* Excited States in Molecular Photochemistry. Phys. Chem. Chem. Phys. 2010, 12, 1218−1238. (2) Sobolewski, A. L.; Domcke, W. Conical Intersections Induced by Repulsive 1πσ* States in Planar Organic Molecules: Malonaldehyde, Pyrrole and Chlorobenzene as Photochemical Model Systems. Chem. Phys. 2000, 259, 181−191. (3) Sobolewski, A. L.; Domcke, W.; Dedonder-Lardeux, C.; Jouvet, C. Excited-State Hydrogen Detachment and Hydrogen Transfer Driven by Repulsive 1πσ* States: a New Paradigm for Nonradiative Decay in Aromatic Biomolecules. Phys. Chem. Chem. Phys. 2002, 4, 1093−1100.

3719

dx.doi.org/10.1021/jz301755p | J. Phys. Chem. Lett. 2012, 3, 3715−3720

The Journal of Physical Chemistry Letters

Letter

(4) Venkatesan, T. S.; Ramesh, S. G.; Lan, Z.; Domcke, W. Theoretical Analysis of Photoinduced H-Atom Elimination in Thiophenol. J. Chem. Phys. 2012, 136, 174312. (5) Elles, C. G.; Crim, F. F. Connecting Chemical Dynamics in Gases and Liquids. Annu. Rev. Phys. Chem. 2006, 57, 273−302. (6) Oliver, T. A. A.; Zhang, Y.; Ashfold, M. N. R.; Bradforth, S. E. Linking Photochemistry in the Gas and Solution Phase: S-H Bond Fission in p-Methylthiophenol Following UV Photoexcitation. Faraday Discuss. 2011, 150, 439−458. (7) Zhang, Y.; Oliver, T. A. A.; Ashfold, M. N. R.; Bradforth, S. Contrasting the Excited State Reaction Pathways of Phenol and paraMethylthiophenol in the Gas and Liquid Phases. Faraday Discuss. 2012, 157, 141−163. (8) Oliver, T. A. A.; King, G. A.; Tew, D. P.; Dixon, R. N.; Ashfold, M. N. R. Controlling Electronic Product Branching at Conical Intersections in the UV Photolysis of para-Substituted Thiophenols. J. Phys. Chem. A 2012, DOI: 10.1021/jp308804d. (9) Roberts, G. M.; Hadden, D. J.; Bergendahl, L. T.; Wenge, A. M.; Harris, S. J.; Karsili, T. N. V.; Ashfold, M. N. R.; Paterson, M. J.; Stavros, V. G. Exploring quantum phenomena and vibrational control in σ* mediated photochemistry. Chem. Sci. 2012, DOI: 10.1039/ C2SC21865H. (10) Berberan-Santos, M. N.; Martinho, J. M. G. The Integration of Kinetic Rate Equations by Matrix Methods. J. Chem. Educ. 1990, 67, 375−379. (11) Lim, J. S.; Kim, S. K. Experimental Probing of Conical Intersection Dynamics in the Photodissociation of Thioanisole. Nat. Chem. 2010, 2, 627−632. (12) Greetham, G. M.; Burgos, P.; Cao, Q.; Clark, I. P.; Codd, P. S.; Farrow, R. C.; George, M. W.; Kogimtzis, M.; Matousek, P.; Parker, A. W.; et al. ULTRA: a Unique Instrument for Time-Resolved Spectroscopy. Appl. Spectrosc. 2010, 64, 1311−1319. (13) Werner, H.-J.; Knowles, P. J.; Lindh, R.; Manby, F. R.; Schütz, M.; Celani, P.; Korona, T.; Mitrushenkov, A.; Rauhut, G.; Adler, T. B.; et al.. MOLPRO, version 2010.1, A Package of Ab Initio Programs, 2010. http://www.molpro.net.

3720

dx.doi.org/10.1021/jz301755p | J. Phys. Chem. Lett. 2012, 3, 3715−3720