Ten‑electron rule helps predict when dopant atoms stay isolated or pair on copper and silver surfaces
This paper maps which combinations of metal dopants prefer to sit alone or pair up when added in tiny amounts to copper and silver surfaces. The authors used atom‑level computer simulations to build an “atlas” that predicts whether a dopant will form a single‑atom alloy (an isolated active atom in an inert host) or a dual‑atom alloy (a paired active site). Their main finding is a simple guiding rule: two dopants tend to pair most strongly when their combined d‑electron count is close to ten.
To make the atlas the team ran density functional theory (DFT) calculations, a standard quantum simulation method, on nearly 3,000 configurations of transition‑metal dopants embedded in Cu(100), Cu(111), and Ag(111) surfaces. They used the optB86b‑vdW exchange‑correlation functional and checked that results were robust with other functionals and with a machine‑learned potential. The study reports two useful thermodynamic metrics: an aggregation energy that measures whether two isolated dopants would rather form a dimer, and a segregation energy that measures whether a dopant prefers to sit at the surface or sink into the bulk.
The ten‑electron trend they found links to the d electrons of transition metals, which occupy orbitals important for bonding at metal surfaces. When two dopants together approach ten d electrons, pairing usually lowers the energy and so dimers (dual‑atom alloys) are favored. The host metal (copper versus silver) and the crystal face (the surface facet) shift the balance and change the detailed geometry of the active site. The authors also identify departures from the rule that come from differences in atomic size and from magnetic spin effects.
They tested some of their predictions experimentally with scanning tunneling microscopy (STM), an imaging method that can see single atoms on a surface. For example, CrAg(111) and PdAg(111) showed isolated single‑atom sites, while a trimetallic CrPdAg(111) sample displayed hetero‑dimers as predicted. The paper also introduces a “reactor‑anchor” concept: one dopant element can act as an anchor that holds a second dopant at the surface when that second element would otherwise prefer to move into the bulk.