A Digital Twin of Silicon Spin Qubits

The rapid advancement of silicon-based quantum computing platforms has driven the need for sophisticated computer-aided design and fabrication of quantum devices. A key requirement is the development of advanced 3D modelling techniques to accurately simulate the dense gate structures that form quantum dot arrays, as well as the intrinsic electronic properties of each dot. Current software struggles to address the combined challenges posed by silicon's induced valley splitting, strong electron-electron correlations in multielectron quantum dots, and realistic disorder from atomic-scale roughness or charged states. We have developed a novel simulation pipeline that integrates COMSOL Multiphysics with path-integral Monte Carlo techniques to model this complex system [1-2]. Initial simulations of nearest-neighbour spin entanglement and electron tunnelling in arrays of up to 2x2 quantum dots show excellent agreement with experimental results and demonstrate high scalability with both the number of electrons and quantum dots. The recent incorporation of valley physics into the simulations paves a promising path toward developing a full digital twin of silicon spin qubits



[1] J.D. Cifuentes, Nature Communications, 15, 4299 (2024)

[2] J.D. Cifuentes, Phys. Rev. B 108, 155413 (2023)