RIKEN researchers have performed repeated measurements of a semiconductor quantum system without disturbing the property they are measuring1. This is promising for the development of quantum computers based on such systems.
One dictum of quantum physics is that it is impossible to perform a measurement on a quantum system without disturbing it. This is problematic for quantum computing because it means that you get only one chance to read a qubit—the quantum equivalent of bits in classical computers. This in turn implies that it is not possible to detect and correct errors in qubits, which is vital for achieving accurate calculations.
One way to overcome this problem is to use a special type of measurement known as a quantum non-demolition measurement (QND), which is designed such that the property being measured remains unchanged (other properties of the system will be altered). This permits multiple measurements to be performed on the system without affecting the property being measured.
While QND measurements have been performed in various systems, until now they had not been realized in semiconductor quantum dots—tiny semiconductor regions that accommodate only a few electrons. Semiconductor quantum dots are especially attractive for quantum computers because they have the key advantage of being able to use existing manufacturing technology.
“Semiconductor devices are particularly good for larger-scale quantum computers because we can employ existing technologies of the integrated circuit industry,” explains Takashi Nakajima at the RIKEN Center for Emergent Matter Science.
Now, Nakajima and his co-workers have succeeded in realizing QND measurements in a device containing semiconductor quantum dots.
By entangling two quantum dots to a third quantum dot, they could use the two quantum dots as probes to measure the spin of an electron in the third quantum dot. A single measurement of the spin of a quantum dot was not particularly accurate, but performing multiple measurements on the quantum dot allowed a much higher accuracy to be realized.
Their technique will allow errors to be corrected in quantum computers. “Being able to detect an error in a qubit without disturbing the qubit further, will allow us to correct the error by an appropriate control pulse,” says Nakajima. “Such an error-correcting protocol is essential for realizing a fault-tolerant quantum computer.”
In the present study the team used gallium arsenide as the semiconductor. They note that the next challenge will be to realize QND measurements in silicon, as that will enable accuracies of more than 99% to be achieved.