General technique to predict spin-phonon relaxation published in Nature Commun.
Long-lived spin states of electrons in materials are of great interest for future quantum information processing technologies. Reliable spin-qubit operations require life times on the order of milliseconds, but spins can scatter against other spins, impurities and phonons (crystal vibrations) rapidly destroying the quantum information encoded in the spin state. Of these, spin-phonon relaxation is the dominant mechanism at higher temperatures and is the intrinsic limit of a perfect material containing few impurities and a very low concentration of spins. Prior studies have employed empirical or material-specific models to estimate spin-phonon relaxation, but a general predictive method independent of experimental inputs is missing.
Here, we present an accurate first-principles method that predicts spin-phonon relaxation lifetimes very generally in all material classes. Specifically, we show that it compares well with experimental measurements on inversion-symmetric semiconductors such as silicon, ferromagnetic metals including iron and 2D materials such as graphene and transition metal dichalcogenides (TMDs). These calculations additionally provide insights about the contributions of different types of phonons, for example those that connect the same valley (intra-valley) or different valleys (inter-valley) in the TMDs, and different modes of vibration such as the flexural phonon mode compared to acoustic and optical phonons in the 2D materials. The ability of our approach to quantitatively compute spin-phonon relaxation in any material with arbitrary degrees of spin mixing provides an unprecedented opportunity to explore the design space for new materials apt for spin-based quantum computing.