Recent spin transport measurements on quasi-one-dimensional wire samples employ a nonlocal geometry. When current is injected at the center of a wire which is grounded at the left end, the portion of the wire to the right of the injector is approximated as having a uniform potential. A voltage measurement between two points on this portion gives a null result. This geometry is useful for studies of spin injection because a spin-sensitive electrode can detect a spin dependent electrochemical potential variation associated with nonequilibrium spin accumulation while contribution from any Ohmic resistance is eliminated. In any real experiment, however, the sample may deviate from the ideal geometry and a small baseline resistance may result. Accurate knowledge of the true baseline is important for the correct interpretation of spin transport experiments. An analytic solution for this nonlocal baseline resistance is found for nonideal injection or detection. The general solution is reduced to a specific case of interest, point injection and detection at the edges of the wire, a result which is compared with experiment.

• Received 21 August 2006

DOI:https://doi.org/10.1103/PhysRevB.76.153107