Why is the spin field effect transistor elusive?

The primary obstacle to the realization of the spin field effect transistor was thought to be the difficulty of spin injection from a ferromagnetic contact into a semiconductor channel. While this remains a major roadblock, there are other obstacles that are equally daunting. One overlooked fact is that the ferromagnetic contacts cause a magnetic field in the transistor´s channel, which introduces a new kind of spin relaxation mechanism. This allows even non-magnetic impurities (with symmetric or asymmetric scattering potential) to flip spin. Another overlooked fact is that if the contacts are reflective, then multiple reflections of carriers between the contacts gives rise to Ramsauer resonances at low temperatures which could affect transistor operation and make transistor characteristics particularly sensitive to impurities in the channel. Finally, the channel must be quasi one-dimensional to mitigate the harmful effect of ensemble averaging and D´yakonov-Perel´ relaxation. Taking into consideration all of this, we have designed a structure that may be promising.