Maki-Thompson correction to magnetoresistance in three-dimensional disordered alloys

Abstract

Measurements of the magnetoresistance of three-dimensional (3D) amorphous MgCuZn alloys were carried out with high precision $(\ensuremath{\Delta}\ensuremath{\rho}/\ensuremath{\rho}<~{10}^{\ensuremath{-}7})$ between $1.8$ and $12 \mathrm{K}$ in magnetic fields up to 2 T. Depending on concentration these alloys are normal or superconducting ${(T}_{C(\mathrm{max})}=123 \mathrm{mK}).$ The Maki-Thompson contribution (MT) of superconducting fluctuations was separated by subtracting weak localization (WL) and electron-electron interaction (EEI) contributions. It was shown that for the latter (WL and EEI) there is a 1:1 agreement between experimental data and the theoretical expressions within the above data accuracy. For the MT contribution in 3D significant and systematic deviations from the corresponding conductivity correction expression of Altshuler, Aronov, Larkin, and Khmelnitskii (AALK) occur for fields in excess of limiting values already given by these authors, i.e., ${B}_{limop}{=(k}_{B}T)/(4eD)\mathrm{ln}{(T/T}_{C}).$ In our case this amounts to ${B}_{limop}\ensuremath{\approx}0.5 \mathrm{T}$ at $T=4.2 \mathrm{K}.$ In fact, our data can be described with this theory and within experimental resolution only for $B<{1/2B}_{limop}.$ Based on a theory of M. Reizer a numerical calculation of the 3D MT contribution has been made which reasonably describes the experimental findings up to $B\ensuremath{\approx}{2B}_{limop}.$ The main feature of the experimental MT contribution as compared with our theoretical estimate is that the experimental data, especially at the lower temperatures, exhibit a stronger curvature near $B\ensuremath{\approx}{2B}_{limop}$ and stay below the theoretical curves. In other words, the high-field saturation is even more pronounced than in our estimate. It is suggested that a more detailed calculation of the magnetoresistance of the MT effect is being attempted.