These new ultra-low consumption transistors could revolutionize electronics

These transistors could enable the creation of a new generation of ultra-low consumption electronic devices. About the size of a few nanometersnanometersthese transistors use the properties of the quantum physicsquantum physics to overcome the limits of transistors based on siliconsilicon.

Silicon-based technology faces a physical limit called the tyranny of Boltzmann. The transistors used in processors function as switches that are switched under a certain voltage to create the 0s and 1s of binary calculations, an operation that requires minimal voltage. In an article published in the journal Nature Electronicsresearchers at MIT have managed to create transistors that use thetunnel effecttunnel effect of quantum physics to operate with much lower voltage.

Transistors based on quantum physics

With the tunnel effect, electronselectrons can cross a potential barrier. Researchers have created transistors based on antimonide of galliumgallium and arsenideindiumindium which use this effect to encourage electrons to pass through the barrier rather than around it. This makes them very easy to switch. These transistors are vertical nanowire heterostructures, with a diameter of just six nanometers, which would be the smallest three-dimensional transistors ever created.

This precision makes it possible to obtain a phenomenon of quantum confinement, where the electron does not have enough space to move, which modifies its massemasse effective and the properties of materialmaterialreinforcing the tunnel effect. The resulting transistors operate below the fundamental limit of silicon-based transistors, and are 20 times more efficient than other transistors that use tunneling.

Researchers need to further improve production methods to create transistors uniformly across an entire chip, including exploring vertical structures in the form of fins. Ultimately, these transistors could enable the creation of ultra-low consumption processors, and significantly reduce the power consumption of artificial intelligence, for example.