Quantum Annealing Optimization Method for the Design of Barrier Materials in Magnetic Tunnel Junctions
Kenji Nawa, Tsuyoshi Suzuki, Keisuke Masuda, Shu Tanaka, Yoshio Miura
Abstract
Materials informatics has boosted materials design, but the search for optimal atomic configurations in spintronic devices is challenging, due to many degrees of freedom and the need to design at the atomic level. Quantum annealing offers a breakthrough for such challenges in huge search spaces. The authors propose a combination of quantum annealing, machine learning, and first-principles calculations that is computationally cheaper than ordinary machine learning in designing atomically disordered spinel oxides (promising materials for magnetoresistive devices). Furthermore, the origins of physical properties of interest can be interpreted from the obtained Ising model Hamiltonian.