J. Phys. Chem. 1988, 92, 3801-3805 TABLE II: Geometrical Parameters (kcal/mol) at Each Point on IRC 0.0(TS) 1.0 Be-B2 Be-HS Be-H, B~-H; B 4 B2-HS B2-H6
Be-B7 Be-Hlo Be-H, Be-H9 B7-H10 B7-H8 B7-H9 B7-H
II
B7-Be-B, BeBT-HIo BeB7-Hg Be-B,-H,
AE
1.895 1.500 1.500 1.189 1.189 1.281 1.281 1.753 1.541 1.740 1.740 1.279 1.237 1.237 1.173 178.7 58.6 68.7 68.7 0.0
(A and deg) and Relative Energies
distance 1.5
1.887 1.487 1.488 1.188 1.189 1.283 1.285 1.776 1.523 2.016 1.570 1.283 1.212 1.264 1.175 174.7 57.0 82.5 59.3 -1.6
1.881 1.480 1.483 1.188 1.189 1.287 1.287 1.798 1.508 2.153 1.527 1.286 1.205 1.275 1.177 172.6 55.6 89.4 56.6 -3.2
2.0
3.0
1.878 1.474 1.479 1.188 1.188 1.293 1.287 1.824 1.493 2.293 1.500 1.289 1.198 1.283 1.180 170.6 54.1 96.5 54.4 -4.8
1.880 1.468 1.477 1.188 1.188 1.302 1.287 1.869 1.476 2.542 1.477 1.295 1.191 1.293 1.184 170.0 51.9 109.2 51.9 -6.7
D2da
1.882(1.875) 1.473 (1.469) 1.473 (1.469) 1.187 ii.is8j 1.187 (1.188) 1.295 (1.291) 1.295 (1.291) 1.882 (1.875) 1.473 (1.469) 2.680 (2.674) 1.473 (1.469) 1.295 (1.291) 1.188 (1.189) 1.295 (1.291) 1.188 (1.189) 180.0(180.0) 519.2 (51.3) 119.8 (119.8) 51.2 (51.3) -7.4
OValues in parentheses are optimized parameters obtained by using the 3-21G** basis set.
exists at s = 1.O amu1/2.bohr,as shown in Figure 6. These changes in geometry accompany a very small decrease in energy, only 1.6 kcal/mol. The next step from s = 1.0 to 2.0 amu1i2-bohrproduces a large change in both geometry and energy. The formation of the B e H 9 bond is almost complete at s = 2.0 amu’/*.bohr (Be-H9: 1.500 A). The breaking of the weak Be-B7 bond is accomplished within the next step. The trace of the H8 movement is easily understood from Figure 6. All geometrical parameters of s = 3.00 (23)s in amu’/2.bohr is the length from the TS to some point on IRC.
3801
amu1/2.bohr are similar to those of the D2d molecule except for the B7-Be-Bz angle. Therefore, the C, molecule is the TS for the exchange of H’s in the BH4- ligand. The bending of the angle and other small changes to the D2dstructure require a very small energy decrease, Le., 0.7 kcal/mol. The barrier for the reaction is calculated to be 7.4 kcal/mol at the 3-21G//3-21G level and 6.1 kcal/mol from a MP2/6-31G//3-21G calculation. 4. Concluding Remarks In the DZdmolecule, both the ES interaction and the orbital mixing with a, symmetry play an important role in the binding of the two fragments. The bridging H’s occupy convenient positions to obtain the large stabilization from AE,,. In the C, molecule, AE, is the largest. Only one H affects the orbital mixing between metal and boron fragments. In both cases, one-way electron donation occurs from the BH4- to the Be(BH4)+, as expected. The IRC calculation indicates that the C, geometry is the transition state of the tautomeric exchange of H’s. Experimental results3 for transition-metal complexes showed that there was rapid exchange between terminal and bridging H’s of BH4-. This reaction is too fast to allow the detection of the individual N M R signals. In these experiments, the central metals were different from that used in the present calculations. However, a mechanism similar to one found here is likely in transition-metal complexes when only sp orbitals are used as valence orbitals.
Acknowledgment. K.H. thanks Prof. H. Weinstein at Mount Sinai School of Medicine for useful discussions. Permission to use FACOM M-382 and VP-200 computers at Data Processing Center of Kyoto University, HITAC M-200H computer at the Computer Center, Institute for Molecular Science, and IBM3090 at the University Computer Center of the City University of New York is gratefully acknowledged. Registry No. Be(BH&, 30374-53-9.
Pathways of SI Decay in Phenol, Indoles, and Water Complexes of Phenol and Indole in a Free Jet Expansion Robert J. Lipert, German Bermudez, and Steven D. Colson* Sterling Chemistry Laboratory, Yale University, New Haven, Connecticut 0651 1 (Received: December 4, 1987)
Pumpprobe photoionization in a supersonic free jet has been used to determine quantum yields and rates of intersystem crossing from the SIstate of phenol, phenol-H20, indole, indole-H20, and a variety of substituted indoles. Refined data analysis plus a more complete accounting of photofragmentation has resulted in lower intersystem-crossing quantum yields in phenol and phenol-H20 than previously reported. Intersystem-crossing quantum yields of all the indole compounds studied were found to be similar except for 2,3-dimethylindole for which evidence is presented for the existence of predissociation of the ‘Lb origin level.
Introduction The pumpprobe photoionization technique’ with mass spectrometric detection has made it possible to monitor photophysical processes that would be difficult or impossible to measure by any other method. In particular, under favorable circumstances, it allows the determination of singlet-state lifetimes, intersystemcrossing quantum yields and the lifetimes of the resulting triplet states of cold, isolated molecules which need not be luminescent. Our interest is in using this technique to measure Q,,
in protein chromophores and in determining the effects of solvation on Qisc,through the study of van der Waals complexes. A better understanding of the effects of complexation on the partitioning of electronic energy into the various decay channels open to a chromophore excited to its S1 state will be of value in the use of ultraviolet spectroscopy as a probe of protein structure and dynamim2 In order to determine Q,,, by the pump-probe method, the photophysical properties of the molecule must be such that singlet
(1) Dietz, T.G.;Duncan, M.A.; Smalley, R. E. J . Chem. Phys. 1982, 76, 1227.
(2) Demchenko, A. P. Ultraviolet Spectroscopy of Proteins; SpringerVerlag: Berlin, 1986. Creed, D. Photochem. Photobiol. 1984, 39, 537.
(eis,),
0022-3654/88/2092-3801$01.50/0
0 1988 American Chemical Society
3802 The Journal of Physical Chemistry, Vol. 92, No. 13, 1988 and triplet decays are clearly distinguishable in a plot of the sum of the singlet- and triplet-state populations, the observable in this experiment, as a function of time. This requires QIscto be not too close to 1 and the singlet lifetime to be sufficiently different from the triplet lifetime so that a biexponential fit of the excited-state decay can be performed. In addition, when the lifetime of the singlet state is comparable to or shorter than the width of the instrument response function, it is important to deconvolute the instrument response function from the observed time evolution of the system when extracting Q,,, from a biexponential fit of the data. This has not been done by previous workers. Recently, Sur and Johnson3 have used the pump-probe technique to study the effects of hydrogen bonding on radiationless transitions in phenol, the chromophore of the amino acid tyrosine. The singlet-state lifetime of phenol was estimated to be approximately 2 ns, which indicates the need for a proper deconvolution of the data when typical 6-10-11s lasers are used. We have repeated the phenol measurement and have tested two methods for determining the instrument response function so that a deconvolution of the data could be performed. Both methods assume that the response function can be approximated by a Gaussian. One approach was to allow the width of the Gaussian to be an additional adjustable parameter in a least-squares fit of the data. The other approach was to determine the response function width independently by using compounds with known singlet lifetimes. In this case excited-state decays were measured, and then the singlet lifetimes were held constant at their published values while the response function width was varied to produce the best fit to the decay curves. The average of 13 such determinations was then used in a fit of the phenol decay. The compounds used in the response function determination were indole, 1-methylindole, 3-methylindole, and 2,3-dimethylindole. These compounds were used because we are also involved in the study of intersystem crossing in indoles. Quantum yields for intersystem crossing in these compounds were obtained from these fits. It is also important for the determination of Q, that one makes a proper accounting of any photofragmentation that might be occurring, both in the neutral species and in the ion. This consideration is particularly relevant to the study of the effects of weakly bound complexes on photophysical properties. In the case of the phenol-water complex, intersystem crossing results in a triplet state with enough vibrational energy to dissociate the complex r a ~ i d l y .Hence ~ the decay observed in the complex mass channel is a single exponential corresponding to the decay of the singlet state of the complex. In addition, the ion formed by probe laser ionization of the singlet state undergoes partial dissociation into ionic phenol and neutral water. Since the monomer formed through dissociation in the triplet manifold is also ionized by the probe laser, the time evolution of the monomer signal will be biexponential. The Q,,, of the complex can be determined by summing the signals in the complex and monomer mass channels. As noted by Dietz et al,,' accurate measurements of Qlscalso require equally efficient ionization processes for both the singlet and triplet species. This is ensured by using sufficiently high ionizing laser powers so that saturation is reached for ionization of both states. However, ion fragmentation due to ion absorption of additional ionizing laser photons can occur before the saturation condition is reached. In this case, ions originating from singlet or triplet neutral species may have different cross sections for the absorption of the second photon. Moreover, their fragmentation channels may not be the same or may have different branching ratios, thereby modifying the observed ratio of ions derived from triplet molecules to those derived from singlets. To account for these effects properly, the signal from the parent and all of the pertinent ionic fragments must be summed in order to get an accurate estimate of Indole derivatives have commonly been used as prototypes of the chromophore in the tryptophan residue of proteins. Due to its large absorption coefficient and fluorescence quantum yield, this chromophore makes a substantial contribution to the near-
e,,,.
(3) Sur, A.; Johnson, P. M . J . Chem. Phys. 1986, 84, 1206.
Lipert et al. ultraviolet spectroscopic properties of protein molecules containing tryptophan. Furthermore, its photophysics is very sensitive to environmental effects, thus making it a convenient probe of protein structure.* Indole studies under supersonic expansion conditions have included fluorescence and photoionization measurement^.^ In a recent study of methyl-substituted indoles, Hager et aLs have found evidence for the 'La dissociative state in 2,3-dimethylindole at vibrational energies within lo00 cm-'of the SI origin. No evidence for this state was seen in indole or the monomethyl-substituted compounds at vibrational energies up to 2000 cm-' above the singlet origin. The appearance of the dissociative state in the dimethyl-substituted molecule was inferred from the observed broadening of the vibronic bands, the qualitative decrease in the fluorescence quantum yields, and the decrease in the fluorescence lifetimes as higher vibronic bands were examined in sequence. Further evidence was provided by the two-color soft ionization spectrum of complexes with solvent molecules known to decrease the energy gap between the dissociative 'La and the bound 'Lb states. The 2,3-dimethylindole-methanol and 2,3-dimethylindole-trimethylamine spectra showed characteristic broadening at the origin region which was ascribed to cluster predissociation due to the nearly isoenergetic 'La and 'Lb states. In this work we report quantum yields of intersystem crossing that are consistent with this state-coupling interpretation. These quantum yields make it possible to quantify the contribution of intersystem crossing to the observed singlet decay rates.
Experimental Section The room temperature vapor of the compounds studied was seeded into helium and continuously expanded through a 12.5-pm pinhole at a backing pressure of 10 atm into a vacuum chamber maintained at Torr. The water complexes were formed in the supersonic expansion by bubbling the carrier gas through room temperature water before the gas reached the heated (
