Solitons or inhomogeneous doping in AsF5-doped polyacetylene—EPR and dc conductivity evidence

The question of whether several properties of As${\mathrm{F}}_{5}$-doped ${(\mathrm{CH})}_{x}$ are due to solitons or doping inhomogeneities is investigated. For dopant concentrations ranging from 0.3 to 4.0 mol% in cis-${(\mathrm{CH})}_{x}$, the EPR spin susceptibility is found to be essentially independent of temperature from 300 K down to the lowest temperatures measured (100 K) but is proportional to the concentration. In trans-${(\mathrm{CH})}_{x}$ results are found for two dopant concentrations that are completely consistent with those in the cis isomer. From these results we propose that over this concentration range the doping is very inhomogeneous, there being distinct doped and undoped regions with the doped regions being metallic. For concentrations of 0.95 and 16 mol% in cis-${(\mathrm{CH})}_{x}$ the respective temperature dependences of the dc conductivity, $\mathrm{exp}(\ensuremath{-}A{T}^{\ensuremath{-}\frac{1}{2}})$ and $\mathrm{exp}(\ensuremath{-}{A}^{\ensuremath{'}}{T}^{\ensuremath{-}\frac{1}{4}})$, are taken to support the idea of very inhomogeneous doping. The transition to metallic conductivity, previously interpreted as a semiconductor-to-metal transition throughout the doped sample, is attributed to the crossing of a percolation threshold at which the metallic regions form a metallic continuum. The effect of doping on the degree of isomerization, which is important for interpreting some experiments on cis-${(\mathrm{CH})}_{x}$, is critically examined. A variety of other experimental evidence, particularly evidence that others have advanced in support of the soliton-doping mechanism, is reviewed and shown to be compatible with inhomogeneous doping, in most cases, while incompatible with soliton doping in several cases.