Doped Oxides
Substituting a small amount of one metal cation into the lattice of another oxide (doping) or mixing two oxides is a common way to tune the surface chemistry of oxide catalysts, in particular the adsorption energies of reaction intermediates that control activity in electrocatalysis, fuel cells, photocatalysis and selective oxidation. Dopants are often classified by their valence relative to the host: low-valence and high-valence dopants change the electron count in ways that simple rules can rationalize, while same-valence and flexible-valence dopants lack such rules. Computational screening of dopant-host pairs benefits from models that link tabulated atomic properties of the two cations to adsorption energies. See Strain and ligand effects, d-band model and Oxygen vacancies in oxides.
Ligand and strain decomposition for doped rutile dioxides (2015)
Xu and Kitchin studied same-valence doping among RuO2, RhO2, PdO2, OsO2, IrO2 and PtO2 with PBE calculations of O adsorption at the (110) 5cus site, 30 dopant-host pairs and 96 adsorption energies in all. They split the change in adsorption on the dopant cation into a ligand effect (new neighboring cations) and a strain effect (host lattice constant), whose sum matched direct calculations, with the ligand effect up to ten times larger. The t2g-band center of the dopant was the best single descriptor. The ligand effect arose from charge transfer between dopant and host cations through the lattice oxygen, even though all cations are formally +4, and it correlated with the difference in effective d-orbital radius (to the 3/2 power) of the pair, a tabulated quantity that the authors propose as a basis for predictive models (Xu 2015). Summary: xu2015b.
Co-doping RuO2 with Cr and Ti for activity and stability (2024)
Abed and coworkers substituted 46 elements into rutile RuO2-type lattices in a machine learning stability screen and used metal-oxygen covalency from crystal orbital Hamilton population analysis as a bulk stability descriptor. Ti gave the largest covalency increase (+10.7%), followed by Sn and Ge. Experimentally the two dopants played separate roles in Ru0.6Cr0.2Ti0.2Ox: Cr raised mass activity about fourfold over RuOx (more than 25 at.% Cr formed inactive Cr2O3), while Ti, limited to 20 at.% by the sol-gel route, formed a strong Ti-O network that slowed Ru dissolution about 20-fold. The computed covalency of the co-doped oxide was 93.62% versus 85.20% for RuO2 (Abed et al. 2024). Summary: abed2024; see also Electrochemical stability.
Mixed Ni-Fe oxides: phase separation instead of substitution (2012)
Landon and coworkers showed that adding Fe to NiO at the low calcination temperature of 550 °C does not dope the NiO lattice: the NiO lattice constant stayed at 4.18 Å across compositions and XPS showed no Ni binding energy shift. Fe instead formed a NiFe2O4 spinel (seen by XRD and, from 5 mol % Fe, by Raman) and, above about 25 mol % Fe, Fe2O3. XANES placed the Fe as Fe3+ with some tetrahedral sites. The best OER activity, near 10 mol % Fe, therefore came from a two-phase NiO/NiFe2O4 material, a reminder that a nominally mixed or doped oxide must be checked for its actual phase content (Landon et al. 2012). Summary: landon2012; see also Oxygen evolution reaction.