Quantum information processing is an ideal information technology whose operation accompanies low-energy dissipation and high information security. We aim at demonstrating the ability of the solid-state information processing, and finally paving the way for the realization with innovative concepts and technology. The specific research targets are implementation of small-scale quantum processing circuits with spins in silicon, development of control methods of quantum coherence and entanglement in the circuits, and development of innovative quantum information devices, and in addition development of control methods of topological particles providing new concepts of quantum information.

### Quantum non-demolition measurement of semiconductor quantum bits

Quantum computing with single electron spin in silicon has been intensively studied, motivated by prospect for the qubit scale-up using semiconductor device processing technology. However, implementing useful measurement-based protocols, including error correction remains a challenge because the qubit measurement usually demolishes the spin state. Here we probe a neighboring electron spin Ising-coupled to the qubit spin and first succeed in the quantum non-demolition measurement of the electron spin in silicon.

In this experiment an electron spin is first initialized to an arbitral qubit state (main qubit) using a spin resonance technique. Next, an Ising-type coupling is applied to the main qubit and the partner qubit (ancilla qubit) to form entanglement between them. When the ancilla qubit is measured being spin-up or spin-down, the main qubit spin is readout without measuring it. This is a nondemolition projective measurement that causes no substantial error in the main qubit, and allows to raise the readout fidelity of the main qubit by repeating the ancilla measurement. We confirmed this feature using a Si/SiGe double quantum dot device having one electron in each dot.

Our work offers a promising route to construct an error correction circuit for implementing fault tolerant spin-qubit systems in silicon.

Figure (Left) Schematic view of the quantum double dot device used for the experiment. (Right) Concept of quantum nondemolition measurement.

###

### An ultra-high fidelity quantum dot qubit in a silicon quantum dot

The building block of quantum computers, or the smallest unit of quantum information, is called a qubit. A very large number of high-quality qubits will be needed to build a quantum computer. Single electron spins in quantum dots are a strong candidate as qubits, as they likely benefit from modern electronics integration technology, once sufficient quality is reached.

Enhancing the qubit quality means improving both coherence time and control time, challenging the trade-off commonly observed between these. In this work, we fabricated a quantum dot on an isotopically-clean silicon wafer to increase the coherence time, and deposited a micromagnet nearby to decrease the control time by its slanting magnetic field. With the coherence time ten times longer and the control time two orders of magnitude shorter than conventional within a single device, we implement highly coherent qubit operations (Fig. B). We further demonstrate >99.9% control fidelity (precision), above the fault-tolerance threshold.

Our work offers a promising route to large-scale, ultra-high-fidelity spin-qubit systems in silicon.

(Fig. A) Schematic view of the quantum-dot qubit device

(Fig. B) Chevron resulting from ultra-high fidelity spin rotations