In conjugated polymers the concept of spectroscopic units belonging to different spatial segments of the chain, which are responsible for the spectroscopic properties of the polymer, has been introduced. This concept might enable one to explain the spectral heterogeneity of conjugated polymers. Furthermore, it may be used to describe the migration of excitations on the conjugated polymer chain by (Förster type) hopping transfer between the spectroscopic units.Gra01
The aim of our study is to get a deeper molecular level insight to the concept addressing the molecular features resulting in a segmentation of a polymer chain into spectroscopic units are still uncertain. Are these only the breaks of the π-conjugation by interjectional single bonds or are kinks and twists of the polymer chain also playing part?
Absorption as well as luminescence spectra of polythiophenes show a significantly inhomogeneous broadening (see example for PTOPT films in figure below from Gra03)
To explain the heterogeneity of the transition energies of conjugated
polymers as well as the associated spectral diffusion, it has been assumed
that the absorption energies (ES1) and oscillator strengths
(fosc) of spectral units depend on the number of conjugated
thiophene rings (N) like those of oligothiophenes. The results for
several straight oligothiophenes as obtained by ZINDO calculation are
shown below.
These results has to be compared with that obtained for
polythiophenes containing kinks, twists or interjectional single bonds.
To characterize the spectral units spatially we have used the transition density (e.g. above for the S1 state of dodecithiophene) which is given as
where the Cμν represent the eigen-vectors of the single-CI Hamiltonian for transitions between occupied and unoccupied MOs φμ(r) and φν(r) , respectively. The CI eigen-vectors as well as the MOs are calculated by the semi-empirical quantum chemical method ZINDO method as implemented in Gaussian 98 Rev. A9.
The molecular features resulting in a segmentation of a polymer chain into spectroscopic units are supposed to be breaks of the π-conjugation by:
Though the transition energies and the oscillator strengths fit more or less to the length of the branches, only the interjectional single bonds result in a clear segmentation of the π-conjugated system. However, to my knowledge there exist no chemical evidence for interjectional single bonds in polythiophenes, yet. For other polymers, like MeH-PPV, there exists a clear evidence for this kind of segmentation Gra03. In the case of horizontally kinked or twisted chains, the segmentation is more or less diffuse and not always unequivocal (e.g. S2-transitions). For azimuthal kinks segmentation is completely absent.
It has to be noted that at the present level of our study only static segmentation has been considered. Nevertheless, we can qualitatively describe how the localization process acts. From the difference of the electron density Δρii(r) between the excited state Si and the ground state S0 results forces, which will drive the nuclei will be driven out of the ground state potential minimum by forces. For the horizontally kinked chain the driving forces in the excited state are concentrated either on the right side of the kink (S2-state; see figure below) or the left side (S1-state)
Notably, this situation is very similar to what has been found for excitonically coupled dimers. Hence, we expect that the feedback from the changed positions of the nuclei to the electronic system will result in a localization for the relaxed excited state (as represented by its transition and different electron density) on one of the respective sides of the kink. Thus, dynamic localization sharpens the already seeded diffuse segmentation (see above) by horizontally kinked and twisted chains. The localization dynamics are important for understanding fluorescence and excitation transfer by Förster hopping. Since it happens on the timescale of nuclear dynamics it can be related to the experimentally observed ultra-fast transient anisotropy decay Gra03.