Element-Specific Donor Numbers for Anions.

Anion interactions with cations/molecules are crucial across chemistry and biology: in battery electrolytes, synthesis, atmospheric processes, and protein denaturation. Selecting the optimum anion represents an immense challenge, as anion interactions are diverse and complex. Descriptors are needed that quantitatively capture the ability of anions to interact with cations and molecules. However, only a small number of experimental anion interaction descriptors exist, all with similar limitations. Here we show that core-level anion electron binding energies, EB(core,anion), can be used to produce a new atomic-level descriptor for each of the key donor elements for anion interactions (O, N, F, Cl, Br, I, S), the element-specific donor number (DNE-XPS). DNE-XPS descriptors are intrinsic, capturing the anion interaction abilities independent of any probe, countercation, or solvent. DNE-XPS descriptors are also interpretable, as EB(core,anion) and therefore DNE-XPS are proportional to the electrostatic potential at the specific donor atom nucleus. Experimental core-level X-ray photoelectron spectroscopy (XPS) is used to measure EB(core,anion) and therefore DNE-XPS. Furthermore, DNE-XPS descriptors are produced easily and at low cost using lone-anion-SMD (solvation model density) calculations, a significant advance on the current anion descriptors, greatly reducing and potentially removing the need for experimental anion characterization. This work will greatly facilitate anion selection, especially for complex anions capable of forming interactions through multiple different atoms. We envisage our calculation method enabling the production of a very large database of DNE-XPS for each key element, ideal for use as machine learning training data sets.