The authors were using for dimethyl sulfoxide (DMSO) a value of the dielectric constant of 36.7, which is lower than the correct value of 46.7. We are sorry for this confusion. The purpose of this erratum is to demonstrate the consequences on the results using the correct value of the DMSO dielectric constant. With the larger value of the dielectric constant the individual electrostatic solvation energies of the protonated and deprotonated molecular species are all more negative. However, for the computation of pKA values differences of these solvation energies are used, where these changes are nearly compensated. As final result no change in the value of proton solvation in DMSO was obtained remaining at −266.4 kcal/mol.1 The individual computed pKA values exhibit minor changes (see Table 1) with a root mean square deviation (pKA-RMSD) of 0.42 and a mean signed deviation (pKA-MSDnew-old) of −0.2 between the new and old pKA values. The pKA-RMSD between computed and measured pKA values is with 0.65 (without outliers) slightly smaller than before, where it was 0.75. The value of pKA-MSDcomp-exp remains practically the same.1 Hence, pKA values seem to be robust against moderate deviations (less than 20%) of the dielectric constant used for the computation. It can be expected that this is less the case for variations of atomic radii and atomic partial charges. As an aprotic solvent DMSO does not possess polar hydrogen atoms. Due to the lack of polar hydrogens the atomic charges of the solvent molecules come less close to atoms of solute molecule than in case of a protic solvent like water. To account for this solvation effect the solute atomic radii must be enlarged compared to the atomic radii in water. For this purpose, we use radii enhancement factors, which are determined by minimizing the pKA-RMSD of measured and computed pKA values. Figure 1 shows the dependence of the pKA-RMSD and the proton solvation energy as a function of the enhancement factor α. The optimal value of the enhancement factor is α = 1.28, if the three compounds, 17, 18, 19, involving a sulfur atom are ignored. The resulting proton solvation energy of −266.4 kcal/mol is again in a plateau regime. Considering the three sulfur compounds separately, the optimized sulfur radius is 2.18 Å, which is the same value as before.1 However, when including these three sulfur compounds in the optimization procedure the smallest pKA-RMSD of 0.65 is obtained with the enhancement factor α = 1.25 and the proton solvation energy is −266.4 kcal/mol as obtained before.1 pKA-RMSD value (•, left scale) and proton solvation energy [kcal/mol] (◊, right scale) for DMSO are plotted as a function of the multiplicative atomic radius enhancement factor α. Ignoring the three sulfur compounds (No. 17–19) the optimal enhancement factor is 1.28 yielding a pKA-RMSD value of 0.65. The resulting proton solvation energy is −266.4 kcal/mol.1 Finally we like to correct Table 1 in the referred article, since the references to the literature were skewed. Again we are very sorry for the confusion.
Read full abstract