a The |D | and |E| values in 3Car(I) are normalized.

a The |D | and |E| values in 3Car(I) are normalized.

Finally, we tried to find the reason why such conformational changes can cause the dissipation of triplet population (energy) by examining the time-dependent changes in spin polarization (the difference in the populations on the spin sublev-els, that is, N0 - N+1 (N-1)). When the triplet energy is transferred from 3P that has been generated by the radical-pair mechanism [20], the initial population on the spin sublevels should be N0 = 1 and N+1 = N-1 = 0. Then, N0 must decay to zero, keeping N+1 = N-1 to be zero. Therefore, the spin polarization of N0 - N+1 (N-1) can never become negative in principle.

Fig. 3.21 The chemical structure and the principal axes of spheroidene (top), and the conformation of 3Car(I) assumed and those of3Car(R) and 3Car(II) determined by comparison between the observed and the

calculated zero-field splitting parameters, |D| and |E| (bottom 3); see Table 3.3 for the comparison of the zero-field splitting parameters [19].

The spectrum of each triplet species originates from an assembly of molecules in all the different spherical orientations. We decomposed the spectra of 3Car(I), 3Car(R), and 3Car(II) into the x-x', y-y', and z-z' components, taking a certain opening angle of a cone around the X, Y, and Z axes, and determined the scaling factors with the + or - sign to fit the observed spectra in reference to a set of simulated unit spectra along each direction. We used the scaling factors thus determined as a measure of spin polarization, and determined its time profile by the use of time-dependent changes in population of each triplet species (the lower panels of Fig. 3.19B). The results for each component and a sum of them are presented in Fig. 3.22.


Delay time / jjs

Fig. 3.22 Time-dependent changes in spin polarization, i.e. N0- N+1 (N^), for the x-x', y-y', and z-z' components and a sum of them [19].

We found something totally unexpected, but it turned out that the observation was along the line of our hypothetical mechanism of triplet-energy dissipation that accompanies the conformational changes of the triplet carotenoid. Figure 3.22 shows the time-dependent changes in spin polarization, which can be characterized as follows: (1) The time profile of a sum of the three components shows that the inversion of spin polarization does take place contrary to our expectation. The only mechanism we can think of to explain this observation is spin-orbit coupling. (2) Concerning the three components, the strongest inversion of the spin polarization takes place along the Z axis, which is approximately in the same direction as that of the C11=C12, C13=C14, and C15=C15' double bonds (see Fig. 3.21), around which the rotational motions take place during the transformation of 3Car(I) fi 3Car(R) fi 3Car(II). The results strongly suggest that the orbital angular momentum generated by the conformational changes is the origin of the unexpected change in spin polarization (or, in other words, in spin angular momentum). (3) The timing in the inversion of spin polarization along the Z axis (3.8 is) approximately agrees with that in the generation of 3Car(R) (3.4 is as shown in Fig. 3.20B).

The results of this investigation, including (a) the conformational changes around the central double bonds, (b) the leak channel of triplet population, and (c) the inversion of the spin polarization during the conformational changes, have provided us with evidence for the hypothetical mechanism of triplet-energy dissipation we proposed previously [17,18,21,22]: "The rotational motions around the central double bonds cause a change in the orbital angular momentum, and through the spin-orbit coupling, a change in the spin angular momentum which facilitates the T1 fi S0 intersystem crossing accompanying the triplet-energy dissipation."

Was this article helpful?

0 0

Post a comment