Simulation of Atomic Layer Deposition with a Quantum Computer

In this work, we present the study of an atomic layer deposition (ALD) of zirconium by means of a quantum computation on an emulator representing the features of an architecture based on qubits implemented on carbon nanotubes. ALD process control is key in several technological applications such as spintronics, catalysis and renewable energy storage. We first derive a large ab-initio model of the precursor molecule approaching the infinite hydroxylated silicon (100) surface. In particular, we optimize geometry in three configurations: reactants, transition state and products. Subsequently, we derive an effective small cluster model for each state. Atomic valence active space (AVAS) transformation is then performed on these small clusters, leading to an effective qubit Hamiltonian, which is solved using the Variational Quantum Eigensolver (VQE) algorithm. We study the convergence of the reaction activation barrier with respect to the active space size and benchmark quantum calculations on a noiseless emulator and on an emulator representing a carbon nanotube qubit architecture, including an appropriate noise model and post-selection error mitigation. These calculations reveal an excellent agreement between the two emulation modes. Our VQE calculations provide the multi-configurational corrections to the single determinant DFT and HF states and pave the way for the routine quantum calculations of ALD reactions.