Interpreting the excited states and decay processes of bichromophoric 1-phenyl-1,3-butadiene using CASSCF calculations

Abstract

We report here a CASSCF computational investigation of the photochemistry of the S ₁ state of 1-phenyl-1,3-butadiene (PB). Our main objective is to determine whether the photochemistry of PB can be rationalized on the basis of the photochemistry of the isolated butadiene and benzene chromophores, or whether there exists an intrinsic coupling between the two chromophores. Based on our findings, we revisit the experimental observations of Song and co-workers [J. W. Buma, B. E. Kohler, J. M. Nuss, T. A. Shaler, and F. Song, J. Chem. Phys. 96, 4860 (1992)] and show that PB deactivates by a mixture of fluorescence and radiationless S ₁ → S ₀ (via conical intersection) decay processes.

Notes

† Although Song and co-workers describe PB molecules as ‘isolated’, they conduct their experiments in 3 atm of He, and some cyclohexane solvent molecules were present in the reservoir chamber of the jet apparatus during the experiment. Also there is no indication of the concentration of PB itself. In the reactor, PB molecules will engage in permanent dipole--permanent dipole interactions either with other PB molecules or with solvent molecules. PB molecules will also participate in instantaneous dipole--permament dipole interactions with He atoms. Naturally, dipole--dipole interactions preferentially stabilize the species with the larger dipole moment. 1t-S ₁ (0.357 D) being more polar than 1t-S ₀ (0.120 D), the computed S ₁→S ₀ 0–0 transition is certainly overestimated in our calculations in vacuum. The extent to which these intermolecular forces impact the computed transition is still unknown.

† As was shown for the tZt-C₆H₈ system, the kink initiated in 1c-S ₁/S ₀ and 1t-S ₁/S ₀ can propagate along the polyenic chain. It follows that isomerization would become possible at any position including here around the C2--C3 bond for the 1c↔1t conversion.

† Computationally, we encountered great difficulties in characterizing 1c -S ₁/S ₀ and 1c -S ₁/S ₀ since the geometry optimizations tended to collapse to the lower lying 1t-S ₁/S ₀ and 1c -S ₁/S ₀ crossings, respectively. This suggests that these points may belong to the same hyperline (seam) of S ₁/S ₀ crossings. Interestingly, it is clear by comparing scheme 2 and scheme 3 that the difference between 1c - S ₁/S ₀ and 1c -S ₁/S ₀, and 1t-S ₁/S ₀ and 1c -S ₁/S ₀, respectively, is the localization of the second unpaired electron on the phenyl ligand. The second radical centre is localized either on a carbon atom adjacent to the phenyl-butadiene linkage or on a carbon site in 1--4 of that linkage. The coordinate of the seam might therefore be mostly due to the relocalization of the π system on the phenyl ligand and apply to both families of crossings.

‡ In a computational study of hexatriene, Garavelli et al. 58 have shown that a high-energy symmetric conical intersection is connected to a low-energy asymmetric crossing by an intersection space segment along which the S ₁ and S ₀ states remain degenerate. A second example of such a topological feature is provided in a recent Migani et al. 59 study of the retinal rhodopsin chromophore. The authors also demonstrated that two kinked S ₁/S ₀ conical intersection minima of deca-1,3,5,6,7-pentane are topologically connected by means of an intersection space segment.

† A zwiterionic/covalent state crossing has been located by Amatatsu 60 in the styrene system. If such a CI exists in PB, it might allow a ‘direct’ radiationless decay of the trans isomer, without the necessity to invoke a cis--trans isomerization on the S ₁ surface prior to decay.