Publications

The most conspicuous feature of boron carbides' electronic transport properties is their having both high carrier densities and large Seebeck coefficients. The magnitudes and temperature dependencies of the Seebeck coefficients are consistent with large contributions from softening bipolarons: singlet bipolarons whose stabilization is significantly affected by their softening of local vibrations. Boron carbides' high carrier densities, small activation energies for hopping ({approx} 0.16 eV), and anomalously large Seebeck coefficients combine with their low, glass-like thermal conductivities to make them unexpectedly efficient high-temperature thermoelectrics.

Ceramic samples of (La{sub 1-x}Gd{sub x}){sub 2/3}Ca{sub 1/3}MnO{sub 3} were prepared and used as targets to grow films onto LaAlO{sub 3} substrates by pulsed laser deposition. Electrical resistance and thermopower, measured vs temperature and applied magnetic fields indicate transport dominated by positive small polarons in the high temperature paramagnetic state. The Hall effect was measured in 0.5 {mu}m thick films of composition x=0 and x=0.25. No evidence for extraordinary hall effect was found in the paramagnetic regime. Instead, the magnitude of the Hall coefficient decreases exponentially with temperature. This behavior and its anomalous negative sign are interpreted to result from face-diagonal hopping of small polarons in the Mn sublattice.

Three issues concerning phonon-assisted hopping in molecularly doped polymers are considered. The first issue is whether Arrhenius jump rates in the vicinity of room temperature arise from single- phonon or small-polaronic hopping. It is concluded that Arrhenius hopping only occurs above low temperatures through small-polaronic hopping. Second, hopping in molecularly doped polymers is compared with small- polaronic hopping of other systems. Small-polaronic hopping typically occurs between similar chemical structures whose energies are relatively insensitive to their surroundings. Thus, disorder energies experienced by carriers are often modest, values of several hundredths of an eV are common. Nonetheless, the effects of large electric fields on carrier mobilities differ significantly among disordered systems. Data reported for molecularly doped polymers is unlike that for either transition- metal-oxide or chalcogenide glasses. In no case is high-field transport well understood. Finally, I stress that steady-state flow is driven by differences in sites` quasielectrochemical potentials (QECPs). With disorder, differences of QECPs are not simply related to the driving emf. Solution of the (nonlinear) stochastic equations for the QECPs shows that bottlenecks produced by disorder result in nonohmic conduction. Solving the linearized stochastic (disordered resistor network) equations underestimates bottleneck effects. Linearization is inappropriate when interest differences in the QECPs exceed [kappa]T.

General master equations are used to study steady-state hopping transport in a disordered solid. We express a site`s occupancy in terms of its quasi-electrochemical potential (QECP); currents flow between sites whose QECP`s differ. Coupled nonlinear circuit equations for the QECP`s result from the steady-state condition and the boundary condition that the total QECP drop is the applied emf. When the site-to-site QECP differences are much smaller than the thermal energy, K{sub B}t, the effect of current flow on site occupancies is ignorable. These equations then reduce to those of a resistance network. However, the resistor-network model fails: (a) at low temperatures, (b) with increasing disorder, and (c) with increasing emf. We therefore study hopping conduction beyond this approximation. Exact examples show the importance of current-induced charge redistribution in non-ohmic steady-state flow.

Superconductivity has long been speculated to result from charge carriers paired as mobile charged bosons. Although the pairing of carriers as small (single-site) bipolarons is known, small bipolarons readily localize. By contrast, large (multi-site) bipolarons, in analogy with large polarons, should be mobile. It is shown that large bipolarons can form in solids with very displaceable ions, e.g., many oxides. Large-polaronic (but not small-polaronic) carriers produce absorption spectra like the carrier-induced absorptions observed in cuprates. Redistribution of the self-trapped carriers of large bipolarons among sites of carriers' molecular orbitals in response to atomic motions lowers phonon frequencies. The dependence of the phonon zero-point energy on the spatial distribution of large bipolarons produces a phonon-mediated attraction between them. This dynamic quantum-mechanical attraction fosters the condensation of large bipolarons into a liquid. Superconductivity can result when the large-bipolarons' groundstate remains liquid rather than solidifying. © 1995 Plenum Publishing Corporation.

Small-polarons will only form in covalent crystals whose electronic halfbandwidths are sufficiently narrow, E{sub b} > W. The absence of small polaronic carriers in most covalent crystals presumably indicates that E{sub b} < W in these instances. However, evidence of small polarons is commonly found in disordered materials despite the estimates of E{sub b} and W not being significantly different from those of crystals. This result is ration by stating that disorder has slowed carrier motion enough to permit small-polaron formation. Recently the question of how disorder affects the stability of quasifree carriers with respect to small-polaron formation has been addressed. It is found that only modest energetic disorder is required to induce small-polaron formation. Here I first succinctly describe essential elements of this work. Second, I address the role of disorder on the adiabatic hopping motion of small polarons. Energy bands in most materials in which small-polarons are found are thought to be sufficiently wide (> a phonon energy) that the small-polaronic hopping is ``adiabatic.`` That is, the electronic carriers move between sites sufficienfly rapidly to follow the atomic motions. In this situation the small-polaron jump rates are independent of intersite separations. The magnitudes of the preexponential factors of the measured hopping mobilities typically support this view. Further support for this picture is found from experiments that determine weak dependences of the mobility on hydrostatic pressure.

Bipolaronic superconductivity requires two exceptional circumstances. First, at least some charge carriers must form bipolarons. Second, these bipolarons must be mobile (move coherently). However, beyond quasi-one-dimensional electronic systems, bipolarons have heretofore always been found to be ''small.'' Small (bi)polarons are very compact (bi)polarons that localize rather than move coherently. Thus, small bipolarons are not suitable carriers for bipolaronic superconductivity. However, in analogy with the case of large polarons, less compact bipolarons, ''large'' bipolarons, are expected to be mobile. These observations lead one to ask if and when large bipolarons can form in multi-dimensional electronic systems. Here the results of studies of the formation, motion and superconductivity of large bipolarons are summarized. 5 refs.