Basic research: IBM scientists rotate electron spin synchronously


For the first time, they were able to directly map the appearance of a persistent spin-helix in a semiconductor where electron spins synchronously rotate, according to an IBM report. The new insights help to manipulate and control the spin in electronic components. Thus, they are of central importance for the practical application of spintronics.

In the paper published in the renowned science journal Nature Physics (DOI: 10.1038 / NPHYS2383), the researchers from IBM Research and the Laboratory for Solid State Physics at the ETH Zurich also show that the spin helix increases the lifetime of a spin by a factor of 30 to 1, 1 nanosecond extended. This corresponds to the clock rate of a 1 GHz processor, ie the time required for a processing step.

Until now, it was unclear whether electron spins have the ability to preserve the encoded information long enough to undergo a change in spin state. The now demonstrated extension of the spin lifetime corresponds exactly to the expected value based on measured material parameters. Thus, the researchers have demonstrated the limiting mechanism and can make further optimizations in a next step.

Spins instead of charge shifts

Today, information in computer chips is encoded and processed with the electric charge of the electron. With increasing reduction of the switching elements, however, this technology approaches physical limits beyond which control of the electron flow is hardly possible anymore.

Spintronics is one possible approach to overcoming this looming dead end. It uses the magnetic moment of the electron instead of the charge. This magnetic moment comes from the electron's angular momentum and is called spin. The basic unit of digital information processing, a state 0 or 1, would be the direction of the spin in such devices. Because spin-based electronics are not based on shifting charge but on a change in spin state, spintronics could enable far more energy-efficient computers and memory.

Spinhelix first observed

In their experiment, the IBM researchers observed how electron spins propagate in a semiconductor over a distance of 20 microns, with all spins spinning synchronously. The formation of such a spin helix has never before been observed in a semiconductor.

The IBM physicist Gian Salis explains: "Normally our spins would all rotate at different speeds and lose their alignment after a short time, so we can now link the spinning speed of the spins to their direction of motion in a special way all spins have exactly the same orientation locally, and the ability to control the alignment and movement of spins in this mass is an important step in the development of spin-based transistors that are electrically programmable. "

Salis and his colleagues used ultrashort laser pulses to simultaneously align and track thousands of electron spins in a very small area. For the first time they were able to demonstrate and reproduce the appearance of a stable spin helix. Theoretical concepts about this phenomenon in semiconductors already appeared in 2003; Since then, various experiments have indicated their occurrence. However, direct evidence has only now been provided. For this, the IBM researchers used a time-resolved scanning microscope technique with which the synchronous spin motion could be tracked and imaged.

Practical application for the time being very difficult

Crucial to generating the spin helix was a targeted coordination of the spin-orbit interaction, a physical mechanism that couples the spin with the motion of the electron. The semiconductor material required for this purpose consists of gallium arsenide (GaAs) and was produced by researchers from ETH Zurich, who are known worldwide for the growth of ultra-pure and atomically accurate semiconductor structures. GaAs belongs to the group of so-called III / V semiconductors and is currently used in integrated circuits, infrared light emitting diodes and highly efficient solar cells.

However, the path from the lab to practical spintronics applications in future computers and novel storage remains challenging. So far, many of the experiments can be carried out only at very low temperatures - in the present work, for example, at -233 degrees Celsius, because it reduces the interactions of the electron spin with the environment to a minimum.