Transforming Adsorption-Energy Linear Correlations via Rescaling and Segmentation

Scaling relations between the adsorbed intermediates of the oxygen evolution reaction (OER) affect the efficiency of electrocatalysts. Recent efforts have been devoted to finding individual departures from scaling relations by numerous strategies. Beyond seeking particular deviations, are there any means of transforming an entire scaling relation? Here we show that the statistical nature of scaling relations is key to their harnessing. Employing a materials data set and a collection of high-throughput optimization techniques known as delta-epsilon optimization, we show the transformation of adsorption-energy correlations via rescaling and segmentation. Rescaling is a visible change in the slope and intercept of a linear relation, whereas segmentation creates two lines, one with a negative slope and another with a positive slope, the hinge point being the ideal catalyst. We illustrate these two phenomena for the scaling relations between *O, *OH, and *OOH and discuss the implications for OER electrocatalysis.