When does surface polymorphism matter for the work function of organic/metal interfaces?

When molecules adsorb on metal surfaces, they might form a variety of different surface polymorphs. However, the influence of this polymorphism on interface properties is a priori unknown. In this work, we use a combination of ab-initio calculations and machine learning to investigate the influence of this surface polymorphism on a relevant interface property, namely the adsorption-induced work function change. Our investigation focuses on three different model systems: a flat-lying physisorbed system, a flat lying chemisorbed system, and a chemisorbed system that additionally undergoes a phase transition from flat-lying to upright-standing.

Employing a combination of exhaustive structure search and ab-initio thermodynamics we elucidate, for which model systems and under which circumstances the specific polymorphs that form have an influence on the interfacial work function change. We find that the influence of polymorphism is smallest for lying, chemisorbed systems and intermediate for physisorbed systems. For a substantial influence of polymorphism on the work function change, a large remaining molecular dipole normal to the surface must be present which in our model systems was only present for the upright-standing molecules.