The nature of charge carriers in conjugated polymers was elucidated through optical spectroscopy following single- and multielectron reduction of 2,7-(9,9-dihexylfluorene) oligomers, F(n), n = 1-10, yielding spectra with the two bands typical of polarons upon single reduction. For short oligomers addition of a second electron gave a single band demonstrating the classic polaron-bipolaron transition. However, for long oligomers double reductions yielded spectra with two bands, better described as two polarons, possibly residing side-by-side in the F(n) chains. The singly reduced anions do not appear to delocalize over the entire length of the longer conjugated systems; instead they are polarons occupying approximately four fluorene repeat units. The polarons of F(3) and F(4) display sharp absorption bands, but for longer oligomers the bands broaden, possibly due to fluctuations of the lengths of these unconfined polarons. DFT calculations with long-range-corrected functionals were fully consistent with the experiments describing polarons in anions, bipolarons in dianions of short oligomers, and side-by-side polarons in dianions of long oligomers, while results from standard functionals were not compatible with the experimental results. The computations found F(10)(2-), for example, to be an open-shell singlet ( ≈ 1), with electrons in two side-by-side orbitals, while dianions of shorter oligomers experienced a gradual transition to bipolarons with states of intermediate character at intermediate lengths. The energies and extinction coefficients of each anionic species were measured by ultraviolet-visible-near-infrared absorption spectroscopy with chemical reduction and pulse radiolysis. Reduction potentials determined from equilibria mirrored oxidation potentials reported by Chi and Wegner. Anions of oligomers four or more units in length contained vestigial neutral (VN) absorption bands that arise from neutral parts of the chain. Energies of the VN bands correspond to those of oligomers shorter by four units.

Electrons and holes were injected selectively into poly-2,7-(9,9-dihexylfluorene) (pF) dissolved in a tetrahydrofuran (THF) and a 1,2-dichloroethane (DCE) solution, respectively, using pulse radiolysis. Transient absorption spectra of monoions of both signs revealed two bands attributable to formation of polarons, one in the visible region (pF+* at 580 nm, pF-* at 600 nm) and another in the near-IR region. Additional confirmation for the identification of pF+* and pF-* comes from bimolecular charge-transfer reactions, such as bithiophene-* + pF --> pF-* or pF+* + TTA --> +TTA+* (TTA = tri-p-tolylamine), in which known radical ions transfer charge to pF or from pF. Difference absorption spectra of pF chemically reduced by sodium in THF provided a ratio of absorbance of anions formed to bleaching of the neutral band at 380 nm. In conjunction with pulse-radiolysis results, the data show that each polaron occupies 4.5 +/- 0.5 fluorene units, most probably contiguous units. Extensive reduction of pF by sodium also revealed resistance to formation of bipolarons: excess electrons reside as separate polarons when two or more electrons were injected. Redox equilibria with pyrene and terthiophene by pulse radiolysis established reversible one-electron redox potentials of E0(pF+/0) = +0.66 V and E0(pF0/-) = -2.65 V vs Fc+/0. Together with the excited-state energy, these results predict a singlet exciton binding energy of 0.2 eV for pF in the presence of 0.1 M tetrabutylammonium tetrafluoroborate. This binding energy would increase substantially without an electrolyte.

No abstract

Polyfluorenes (pF) were synthesized having anthraquinone (AQ) or naphtylimide (NI) end caps that trap electrons or di- p-tolylaminophenyl (APT2) caps that trap holes. The average lengths of the pF chains in these molecules varied from 7 to 30 nm. End capping was found not to be complete in these molecules so that some were without caps. Electrons or holes were injected into these polymers in solution by pulse radiolysis. Following attachment, the charges migrated to the end cap traps in times near 2 ns in pF12AQ or 5 ns in pF35NI. From these observations, electron mobilities for transport along single chains to the end caps in THF solution were determined to be smaller by a factor of 100 than those observed by microwave conductivity. Despite this, the mobilities were sufficiently large to provide encouragement to the use of such single chains in solar photovoltaics. Most charges were observed to transport over substantial distances in these polymers, but 23, 18, and 37% of the charges attached to pFNI, pFAQ, and pFAPT2, respectively, were trapped in the pF chains and decayed by slower bimolecular reactions. For pFAQ and pFAPT2, all of the trapped charges were accounted for by estimates of the fraction of molecules having no end cap traps. For pF35NI, 23% of the attached electrons were found to be trapped in the chains, but only 4% of chains were expected to have no end caps. This could indicate some trapping by kinks or other defects but may just reflect uncertainties in the capping of this long polymer. When the charges reach the trap groups, their spectra have no features of pF(*-) or pF(*+), nor do the principal bands of the trapped ions resemble spectra of the radical ions of isolated trap molecules. The optical absorption spectra are rather dominated by new bands identified as charge-transfer transitions, which probably reinject electrons or holes into the pF chains. The energies of those bands correlate well with measured redox potentials.

Carbon-halogen bond dissociation rates for a series of aryl halide radical anions (ArX-: X = Cl, Br) in NMP were measured at room temperature by pulse radiolysis with 10-11 s time resolution. To obtain accurate dissociation rates, care was taken to measure and correct for competing decay channels. The observed rates correlated well with activation energies computed in the gas phase by density functional (DFT) calculations. The rates did not correlate well with electron affinities or dissociation energies obtained by the same computational methods, although such correlations are reported in the literature and are expected on the basis of simple models. The calculations also found that the transition state structures had bent carbon-halogen bonds. Bending enables large reductions of the activation energies by an electronic effect involving mixing of phi* and sigma* states. This bending-induced mixing is computed to increase the dissociation rates by a few orders of magnitude and is thus essential to understanding these reactions.