Theory of the effects of destruction of localization by inelastic scattering in the resistivity of pure thin potassium wires

Mary Eileen Farrell, Marilyn F. Bishop, N. Kumar, and W. E. Lawrence
Phys. Rev. B 42, 3260 – Published 15 August 1990
PDFExport Citation

Abstract

Measurements of the electrical resistivity of thin potassium wires at temperatures near 1 K have revealed a minimum in the resistivity as a function of temperature. By proposing that the electrons in these wires have undergone localization, albeit with large localization length, and that inelastic-scattering events destroy the coherence of that state, we can explain both the magnitude and shape of the temperature-dependent resistivity data. Localization of electrons in these wires is to be expected because, due to the high purity of the potassium, the elastic mean free path is comparable to the diameters of the thinnest samples, making the Thouless length lT (or inelastic diffusion length) much larger than the diameter, so that the wire is effectively one dimensional. The inelastic events effectively break the wire into a series of localized segments, whose resistances can be added to obtain the total resistance of the wire.

The ensemble-averaged resistance for all possible segmented wires, weighted with a Poisson distribution of inelastic-scattering lengths along the wire, yields a length dependence for the resistance that is proportional to [L3/lin(T)], provided that lin(T)≥L, where L is the sample length and lin(T) is some effective temperature-dependent one-dimensional inelastic-scattering length. A more sophisticated approach using a Poisson distribution in inelastic-scattering times, which takes into account the diffusive motion of the electrons along the wire through the Thouless length, yields a length- and temperature-dependent resistivity proportional to (L/lT)4 under appropriate conditions. Inelastic-scattering lifetimes are inferred from the temperature-dependent bulk resistivities (i.e., those of thicker, effectively three-dimensional samples), assuming that a minimum amount of energy must be exchanged for a collision to be effective in destroying the phase coherence of the localized state. If the dominant inelastic mechanism is electron-electron scattering, then our result, given the appropriate choice of the channel number parameter, is consistent with the data. If electron-phason scattering were of comparable importance, then our results would remain consistent. However, the inelastic-scattering lifetime inferred from bulk resistivity data is too short. This is because the electron-phason mechanism dominates in the inelastic-scattering rate, although the two mechanisms may be of comparable importance for the bulk resistivity. Possible reasons why the electron-phason mechanism might be less effective in thin wires than in bulk are discussed.

  • Received 16 April 1990

DOI:https://doi.org/10.1103/PhysRevB.42.3260

©1990 American Physical Society

Authors & Affiliations

Mary Eileen Farrell

  • Department of Physics and Atmospheric Science, Drexel University, Philadelphia, Pennsylvania 19104

Marilyn F. Bishop

  • Department of Physics, Virginia Commonwealth University, 1020 West Main Street, P.O. Box 2000, Richmond, Virginia 23284-2000

N. Kumar

  • Department of Physics, Indian Institute of Science, 560 012 Bangalore, Mysore, India

W. E. Lawrence

  • Department of Physics and Astronomy, Dartmouth College, Hanover, New Hampshire 03755

References (Subscription Required)

Click to Expand
Issue

Vol. 42, Iss. 6 — 15 August 1990

Reuse & Permissions
Access Options
Author publication services for translation and copyediting assistance advertisement

Authorization Required


×
×

Images

×

Sign up to receive regular email alerts from Physical Review B

Log In

Cancel
×

Search


Article Lookup

Paste a citation or DOI

Enter a citation
×