Reexamination of some spintronic field-effect device concepts

Current interest in spintronics is largely motivated by a belief that spin-based devices (e.g., spin field-effect transistors) will be faster and consume less power than their electronic counterparts. Here we show that this is generally untrue. Unless materials with extremely strong spin-orbit interaction can be developed, the spintronic devices will not measure up to their electronic cousins. We also show that some recently proposed modifications of the original spin field-effect transistor concept of Datta and Das [Appl. Phys. Lett. 56, 665 (1990)] actually lead to worse performance than the original construct. (C) 2004 American Institute of Physics.