The conductivity and Seebeck coefficient in the nearest-neighbour hopping regime of a Fermi glass

Colossal Seebeck Coefficient of Hopping Electrons in (TMTSF)_{2} PF_{6}

We report on a study of the Seebeck coefficient and resistivity in the quasi-one-dimensional conductor (TMTSF)_{2} PF_{6} extended deep into the spin-density-wave state. The metal-insulator transition at T_{SDW}=12  K leads to a reduction in carrier concentration by 7 orders of magnitude. Below 1 K, charge transport displays the behavior known as variable range hopping. Until now, the Seebeck response of electrons in this regime has barely been explored and is even less understood. We find that, in this system, residual carriers, hopping from one trap to another, generate a Seebeck coefficient as large as 400  k_{B}/e. The results provide the first solid evidence for a long-standing prediction according to which hopping electrons in the presence of the Coulomb interaction can generate a sizable Seebeck coefficient in the zero-temperature limit.

Thermoelectric power in alloys with phase separation (composites)

A thermopower formula is derived for composites: [Formula: see text] (κ(i) and κ are the specific thermal conductivities, S(i) and S are the Seebeck coefficients of the phase i and the composite, respectively, and υ(i) is the volume fraction of the phase i). This formula can be applied for calculating the Seebeck coefficient (thermoelectric power) of amorphous transition-metal-metalloid alloys, for which amorphous phase separation occurs for large ranges of concentration. There are two contributions to S(i), a scattering term and a contribution due to electron transfer between the phases maintaining a common electrochemical potential μ. The theory predicts discontinuities in the concentration dependence of the Seebeck coefficient of metallic composites. It is argued that in amorphous composites these discontinuities occur very precisely at υ(i) = 1/3. This phenomenon can be used to characterize the crystallization kinetics of amorphous alloys. The theory is applied to a- Cr(1-x)Si(x) alloys for calculation of S versus x. Both the calculated S(x) dependence and the discontinuities agree very well with the experimental data, as long as x<0.67; the deviations at x>0.67 are interpreted to be caused by the p-d bonds at the phase boundaries.