Volcano Plot
A volcano plot shows catalytic activity as a function of a descriptor, typically the binding energy of a key intermediate. Activity rises as binding strengthens from very weak, peaks at an intermediate value, and falls when binding becomes too strong, reflecting the Sabatier principle. See also Hydrogen evolution reaction and Hydrogen adsorption free energy.
HER volcano from a computed descriptor (2005)
Plotting measured HER exchange currents against DFT hydrogen chemisorption energies, Nørskov, Kitchin and coworkers obtained a volcano with reactive metals on the strong-binding leg and Cu, Ag and Au on the weak-binding leg. A one-parameter kinetic model based only on thermochemistry reproduces the trend and places the maximum at zero hydrogen adsorption free energy: strong binding starves the surface of free sites, while weak binding makes proton transfer uphill (Nørskov et al. 2005). Summary: norskov2005.
Origins of the HER volcano and the role of DFT descriptors (2006)
In their reply to Schmickler and Trasatti, Nørskov, Kitchin and coworkers acknowledge that volcano plots of HER rate against hydrogen binding energy were introduced decades earlier, crediting in particular Parsons (1958). They describe their own contribution as supplying the descriptor from density functional theory, and their simple kinetic model as a way to show how a volcano peaking near zero hydrogen adsorption free energy arises from thermochemistry alone (No̸rskov et al. 2006). Summary: norskov2006.
A single OER volcano for many oxide families (2011)
Plotting the negative theoretical OER overpotential against the free energy difference between O* and HO* gives one volcano shared by perovskite, rutile, anatase, spinel and manganese oxides. The two legs correspond to HOO* formation limiting on strong binders and HO* oxidation limiting on weak binders, and the apex is capped by the HO*/HOO* scaling relation rather than reaching zero overpotential (Man et al. 2011). Summary: man2011.
Moving along the OER volcano with DFT + U (2015)
For rutile dioxides, Xu, Rossmeisl and Kitchin found that adding a linear response Hubbard U preserves the adsorption scaling relations, so the OER volcano itself is unchanged and each oxide simply slides along one of its legs toward weaker binding. IrO2 and RuO2 moved toward the peak from the strong-binding side while RhO2 and PtO2 moved away on the weak-binding side, which reordered the predicted activities but moved no oxide across the peak (Xu et al. 2015). Summary: xu2015; see also Oxygen evolution reaction.
ORR volcano versus oxygen binding energy (2004)
Using DFT O and OH binding energies on 13 close-packed metal surfaces and a Sabatier-type measure of maximal activity (the slowest elementary step relative to a nonactivated proton transfer), Nørskov, Rossmeisl, Kitchin and coworkers built an oxygen reduction volcano with Pt and Pd at the top. Strong binders such as Ni are limited by proton transfer to adsorbed O or OH, while weak binders such as Au are limited by O2 dissociation or, in the associative mechanism, by proton transfer to adsorbed O2. Including the associative path and the potential dependence leaves the volcano maximum nearly unchanged (Nørskov et al. 2004). Summary: norskov2004; see also Oxygen reduction reaction.
A volcano for photocatalytic Pd/Sn colloids (2020)
Lopato and coworkers asked whether the electrocatalytic HER volcano carries over to photocatalytic hydrogen evolution on in situ formed Pd/Sn nanoparticles (Lopato et al. 2020). Measured maximum rates peaked near a Pd mole fraction of 0.5, and when plotted against a DFT-derived, composition-weighted hydrogen adsorption free energy they followed a parabola (R2 = 0.98) with its maximum at -0.036 eV, close to the 0 eV electrocatalytic optimum; the authors note other explanations are possible. Summary: lopato2020.
Volcano screening at nanoparticle scale (2023)
This study built an HER volcano from high-throughput nanoparticle experiments by representing each metal pair with the Pourbaix-stable surface whose strongest hydrogen binding site lay closest to -0.24 eV, and plotting it against the highest observed H2 evolution rate for that pair (Broderick et al. 2023). Only the upper envelope of activity formed a volcano: favorable binding (about -0.41 to -0.23 eV) proved necessary but not sufficient for activity. Used as a filter, the volcano would exclude about half of all metal combinations from an experimental screen, illustrating both the value and the limits of Sabatier-based screening for complex in situ catalysts. Summary: broderick2023.
Beyond the Sabatier peak by oscillating binding energies (2025)
Dynamic catalysis aims to exceed the static Sabatier limit by oscillating binding energies so different steps are promoted at different times. Colombo Tedesco and co-workers note that small forcing amplitudes, which lie near the Sabatier peak of an extended volcano, are both easier to realize experimentally and the regime where their analytical perturbation method is accurate (Colombo Tedesco et al. 2025). Summary: colombotedesco2025; see also Dynamic catalysis.