Reducing Tuning Complexity in Top Gated Semiconductor Quantum Dot Qubits with single-nm-resolution Gate Fabrication

Top gated semiconductor quantum dot qubits represent an attractive path to quantum computing. However, variations in the physical dimensions of the top gates create significant variations in the size of the electrostatic confinement and therefore the energy levels in the qubit. The variation in gate dimensions complicates the design of multi qubit systems and the required tuning of the biases on the gates for multiple qubits is so complex that machine learning is employed.

Multiple modeling runs of a generic top gated multi-qubit system carried out with the spin-qubit computer-aided design tool QTCAD will be reported on. Using finite elements, QTCAD enables the solution of non-linear Poisson and Schrödinger equations for quantum-dot systems with arbitrary 3D geometry. The models will include variations in the gate dimensions that are representative of those found in typical e-beam lithography patterns, structures with significantly less variations, and a system with no variations. Predictions on the impact of the magnitude of variations on the complexity of tuning to low error rate conditions will be made.

Finally, a path will be described which uses Atomic Precision lithography, selective Atomic layer Deposition, and reactive ion etching to make nanoimprint templates. The accuracy of templates thus produced, and the precision of Jet and Flash Nanoimprint lithography will produce far more uniform top gates with a scalable manufacturing technique.