Superconducting devices, from particle accelerator cavities (SRF) to quantum circuits, are incredibly sensitive to their surfaces. Even a few nanometers of oxide or absorbed hydrogen on niobium (Nb) or tantalum (Ta) can introduce defects that act as “two-level systems” (TLS), which degrade qubit coherence times and increase losses in SRF cavities. This makes surface chemistry one of the central bottlenecks for advancing both quantum technologies and superconducting accelerators. Protecting these surfaces requires ultrathin metallic caps—thick enough to prevent oxygen and hydrogen uptake, but thin enough to avoid reducing superconductivity through the “proximity effect.” Balancing these constraints has been difficult; despite major experimental progress, the field has lacked a rational, first-principles design framework.

This paper proposes that framework. Using ab initio calculations that link interface energetics to superconducting proximity effects, it identifies gold (Au) and Au-rich alloys (AuPd, AuPt) as excellent outer “passivation” layers that resist O/N/H, but notes that pure Au can dewet on realistic Nb/Ta surfaces unless made too thick. The key innovation is a wetting/adhesion layer (WAL): a 1–2-monolayer underlayer—especially copper (Cu)—that bonds strongly to both the substrate and the Au cap, preventing pinholes and letting the Au stay in the 2–3-monolayer “sweet spot” where passivation saturates while superconductivity is preserved. Zirconium (Zr) is also highlighted as a sacrificial oxygen “getter” when interfacial O cannot be avoided. The resulting, ready-to-test design rules—Au/Cu/(Nb or Ta) and AuPt/(Nb or Ta)—offer a clear path to oxide-free, air-stable surfaces without sacrificing superconducting performance, enabling longer-lived qubits and higher-Q SRF cavities.