Development of GW-based substrate screening methods for strongly coupled molecule-metal interfaces

The understanding of the molecule-metal interfaces is of paramount importance to the development of novel electronic devices. This requires an accurate characterization of the quasiparticle electronic structure at these heterogeneous interfaces, for which the GW approach within the framework of many-body perturbation theory is generally accurate. The substrate screening approximation effectively reduces the computational cost of GW calculations for large weakly coupled interfaces, by assuming that the non-interacting polarizability of the interface is additive of those from the adsorbate and from the substrate. However, this approximation breaks down for strongly coupled interfaces. Here, we develop a new GW-based method that captures the significant orbital hybridization at the interface, extending the substrate screening approximation. We use two experimentally well-studied systems to demonstrate our approach, namely the benzene dithiol adsorbed on Au(111) and the naphthalene tetracarboxylic dianhydride adsorbed on Ag(111). Our work holds promise in studying a wide range of interfaces including defects, covalently bound molecules on metal surfaces, and many others.