An international team of researchers has reported an experimental demonstration of a transistor whose functionality is based on an electron’s spin. The work was published in the December 24th issue of the Science journal. The team is formed by physicists from the Academy of Sciences and Charles University in the Czech Republic, the Hitachi Cambridge Laboratory and University of Cambridge and Nottingham in the UK, and Texas A&M University in the US.
60 years after the discovery of a transistor its operation is still based on the same physical principle of electrical manipulation and detection of electron’s charge in a semiconductor. The technology has focused on down-scaling the device size which has brought transistor dimensions from the original table-top size close to the inter-atomic-distance scale in an astonishingly short time of several decades. Since we are quickly approaching the ultimate down-scaling limit, it is now an eminent task to establish new physical principles of transistor operation. One extensively studied possibility is utilizing the second basic attribute of the electron which is its elementary magnetic moment, the so-called spin.
Theoretical proposal of electrical manipulation and detection of electron’s spin in semiconductors is 20 years old. However, its experimental realization turned out to be unexpectedly difficult. The team has engaged recently discovered quantum-relativistic phenomena for both spin manipulation and detection to realize the spin transistor and to demonstrate spin-logic operation.
To observe the electrical manipulation and detection of spins, the team utilized a specially designed planar photo-diode placed next to the transistor channel. By shining light on the diode, photo-excited electrons are injected into the transistor channel. In the present work, a circularly polarized light is used to generate spin-polarized electrons. The quantum-relativistic effects are employed to control the precession of spins by input gate-electrode voltages. Quantum-relativity is also responsible for the onset of transverse electrical voltages which depend on the local orientation of precessing electron spins in the transistor channel and represent the output signal.
The new device can have a broad range of applications in spin-electronic research as an efficient tool for manipulating and detecting spins in semiconductors without destroying the spin-polarized current and without using magnetic elements.
The observed output electrical signals remain large at room temperature and are linearly dependent on the degree of circular polarization of the incident light. The device represents a realization of an electrically controllable solid-state polarimeter which directly converts polarization of light into electric voltage signals. Medium term applications may exploit the device to detect the content of chiral molecules in solutions, for example, to measure the blood sugar level of patients or the sugar content of the wine. Whether spin-transistors will become a viable alternative or complement of current electron-charge based transistors in information processing devices is yet to be discerned. Due to the present work, this question has turned from theoretical academic grounds to the real world of prototype microelectronic devices.
Contact: Tomáš Jungwirth