Researchers at CEA-Irig have demonstrated that there is an optimal configuration that allows hole spin qubits to hide from noise and thus have long coherence times.
The promise of quantum computing lies in its potential ability to solve problems much faster than conventional computers. In the current technological race, different physical media are competing to encode quantum information. What do they have in common? They must form quantum two level system, the famous quantum bits or qubits.Unlike the classical bit, which can only take two values (0 or 1), the qubit can be placed in a superposition of the two states 0 and 1, such as 0+1 or 0-1. It is this possibility of superposition that is exploited in the algorithms to be run by future quantum computers.

To create a qubit we can use a single electron (or its positively charged alter-ego, the hole) in a semiconductor such as silicon or germanium. The qubit is then encoded using the electron's spin state. Spin has no equivalent in classical physics, but it can be imagined as a small quantum compass. When the needle of the compass points downwards we speak of low spin (↓) defining the qubit as being in the 0 state.
Conversely, the needle pointing upwards will define state 1, known as high spin (↑). Electron or hole spins are promising candidates because they can be isolated in devices similar to silicon transistors, as demonstrated in 2016 by researchers at CEA-IRIG and CEA-LETI [1]. The manufacture of these qubits is therefore compatible with industrial microelectronics processes. Small, in the order of ten nanometers, the millions, or even billions, spin qubits could be integrated on a single chip in the same way as the millions of transistors at the heart of today’s computer processors.

Like bits, qubits must be able to be initialized, read and manipulated, but above all they must be able to be kept in a state of superposition. However, this superposition is fragile. The slightest noise in the qubit's environment can destroy its superposed state. This is known as decoherence, the nightmare of quantum information! At present, no qubit possesses sufficient coherence to exploit the promise of quantum algorithms (most of which are based on perfect qubits). This is where fundamental research comes in. By trying to understand the physics of quantum systems, researchers are trying to model the effects of decoherence and find strategies for making better qubits.

Recently, researchers at Delft University of Technology (Netherlands) developed a 4-hole germanium qubit processor [2], a feat hailed by the entire spin qubit community. However, in their experiment (as in the CEA's first experiments), the coherence killer is electrical noise.By finely controlling a single spin of a hole in silicon, researchers at IRIG have just demonstrated that there is an optimal configuration for which electrical noise is no longer capable of inducing decoherence [3]. "It was incredible: when we applied the magnetic field in a very precise direction, we had a tremendous increase of the coherence of our qubit," enthuses Romain Maurand from the LaTEQS laboratory. But we had to understand this unexpected effect. So, combining theory and experiment, the physicists refined their model and "for the first time, the experimental spin responses measured at LaTEQS were in perfect agreement with our theoretical predictions" comments Yann-Michel Niquet from the IRIG's LSIM laboratory. The researchers are continuing to explore the physics of this ideal configuration and are trying to determine whether it could be used on several qubits simultaneously or in other materials such as germanium.