Correction to: Nature Catalysis https://www.nature.com/articles/s41929-024-01177-6, published online 14 June 2024.

We found and corrected a mistake in our computer code that estimated the electrode surface roughness based on measured current densities. In the corresponding routine, the parameters a and b in the function j = a ⋅ exp(−b ⋅ U) are determined via two consecutive fits to experimental datasets of measured current density j and applied potential U. A typographical error in the code unintentionally changed the function for the first fit to j = a + exp(−b ⋅ U), leading to wrong values for the initially determined b which then also affected the second fit in determining the target value a. The pre-factor a is proportional to the density of (catalytically) active sites and thus to a relative roughness which we directly use as the experimental estimate.

The affected code was only used to obtain reference experimental data when no roughness estimates were made available in the original studies. This was the case for roughness estimates of oxygen-derived (OD) Cu in Fig. 3a which changed from 30 to 16 and those from two studies on CuAg- and CuPd-alloys shown in Fig. 4. The corresponding figures changed as shown in Figs. 1 and 2, below.

Fig. 1
figure 1

Original and corrected Fig. 3a.

Full size image
Fig. 2
figure 2

Original and corrected Fig. 4.

Full size image

Correcting the error does not lead to any qualitative changes in our analysis or our discussion. In the case of Fig. 4, it even leads to improved agreement between the experimental reference data and our theoretical model.

We have now corrected the data and slightly adjusted some wording in the HTML and PDF versions of the manuscript as well as the Supplementary information and data sheet. We have also provided updated scripts for roughness evaluation in the GitHub repository (https://github.com/hheenen/model_transport_selectivity) that is published alongside our paper.