Search

Electron-nuclear double resonance (ENDOR) spectra have been recorded for samples of pristine cis- and trans-polyacetylene at temperatures ranging from 15°K to 290°K. The ENDOR spectra of cis-polyacetylene consist of resolved contributions arising from distant and local ENDOR mechanisms. The temperature dependence of the peak positions of the local ENDOR spectra is analyzed in terms of a simple two-site electron jump model resulting in modulation of a 1.8 MHz isotropic hyperfine coupling. This analysis leads to an activation energy of 34°K. The ENDOR spectra of trans-polyacetylene at temperature above 70°K consist of a single line centered at the free proton Larmor frequency and exhibiting a temperature-dependent linewidth. These spectra are assumed to be composed of distant and local ENDOR contributions with motional narrowing of the local ENDOR structure accounting for the temperature dependence of the ENDOR linewidths. This assumption is supported by observation of two-component ENDOR spectra below 40°K. Analysis of the linewidth variations of the high temperature spectra suggests that the linewidths are determined by modulation of an approximately 1.4 MHz hyper-fine interaction by a dynamic process characterized by a 400-K activation energy. These numbers are in excellent agreement with values obtained from analysis of electron spin echo (ESE) spin-spin relaxation data establishing that the same process determines both ENDOR and ESE data. No evidence is obtained from ENDOR spectra for the existence of large hyperfine interactions. Thus, the EPR, ESE, and ENDOR results for trans-polyacetylene are dominated by an intermediate motion process (τ = 10-6 to 10-9 sec) modulating approximately 2 MHz hyperfine couplings and characterized by a relatively high activation energy (400°K) and not by a fast, highly directional motion modulating large hyperfine interactions as suggested by recent soliton theory analysis of EPR spectra.

Spin and charge excitations in pristine and AsF5-doped polyacetylene have been investigated to resolve the issue of whether or not topological solitons have been observed. We present experimental results for both materials suggesting that the simple model from which the soliton-based interpretations have been derived is inapplicable. In pristine trans-polyacetylene, both protonated and deuterated, conclusions are deduced from spin-echo and continuous-wave EPR about the spin diffusion rate, its temperature dependence and anisotropy that contradict predictions of the soliton theory. In AsF5-doped cis-polyacetyelene, we conclude from magnetic and transport measurements that the doping is very inhomogeneous, the doped and undoped parts being largely distinct with the doped regions metallic. The conductivity changes previously interpreted as arising from a transition wthin the doped region are attributed to a change in the topology of these regions, i. e. the crossing of a percolation threshold to a regime in which the metallic regions are no longer isolated. Other experiments for and against the soliton interpretation are also discussed.We have given evidence and arguments for the following picture: 1) The doping of cis-(CH)x is inhomogeneous for all concentrations of dopant that we have studied (0.3-4%), the doped and undoped parts of the polymer forming separate phases. 2) The doped parts form metallic-like clusters exhibiting a Pauli spin susceptibility. 3) At low dopant concentrations, the dc conductivity, which was previously interpreted to be activated or three-dimensional variable-range hopping as in a normal semiconductor, possibly with kink solitons as carriers, or to be via tunneling between charged and neutral kink solitons,12 involves conduction between isolated metallic particles dispersed in a dielectric, as described by Sheng, Abeles and Arie. 4) At high dopant concentrations, the conductivity, which was thought to be that of a disordered metal or degenerate-band semiconductor, appears to be essentially that, but with conduction along percolation paths connecting such regions, and hopping (i. e. tunneling) between the fibrils, as analyzed by Sheng. 5) The semiconductor-to-metal transition, which was previously interpreted as arising within the doped region, results from a change in the topology of these regions, i. e. it is the crossing of a kind of percolation threshold to a regime in which the metallic regions nearly touch to form a continuum.