Proton Transfer 200 Years after von Grotthuss: Insights from Ab Initio Simulations

Figure 8 in ref. 1 is reproduced from Figure 1 in ref. 2. Panels (a)–(c) thereof, reprinted as Figure 1 1in this Addendum, illustrate schematically the essential steps of the proton diffusion mechanism in liquid water 3–12 as illustrated in Figure 2 in ref. 6 and Figure 1 in the review ref. 13. It should be noted that Figure 1 in ref. 2 and thus Figure 8 in ref. 1 is only a very schematic sketch of the structural diffusion process that has been previously sampled from actual ab initio molecular dynamics trajectories based on selected snapshot configurations such as those depicted in Figures 1 and 3 in ref. 3, Figure 3 in ref. 2, and, in particular, in Figure 1 in ref. 10. Reproduction of Figure 8 in ref. 1: 'Schematic representaion of the mechanism of structural diffusion of a)–c) H+ and d–f) OH− charge defects in liquid water.'; for panels a)–c) see also Figure 2 in ref. 6 and Figure 1 in the review ref. 13. The quotation marks in the sentence 'Shortly after, and independently from the first ab initio molecular dynamics simulation,[62] a molecular mechanism was "suggested"[76] that is "qualitatively consistent with the above mechanism".' were used in ref. 1 in order to highlight explicit quotes from the following two sentences of ref. 6 (which is ref. [76] in ref. 1): 'It is suggested that the molecular mechanism behind prototropic mobility involves a periodic series of isomerizations between H9O4+ and H5O2+, the first triggered by hydrogen-bond cleavage of a second-shell water molecule and the second by the reverse, hydrogen-bond formation process.' and 'The animated trajectories are qualitatively consistent with the above mechanism.' whereby the specific mechanism referred to in the latter sentence is the one described in the former sentence and the animated trajectories are those published in ref. 3 (which is ref. [62] in ref. 1). The publication chronology of refs. 3, 6 collected in [76] in ref. 1 is provided in order to substantiate the conclusion drawn by the reviewer that both publications report results from independent research (see also the two publications refs. 14 and 5 that precede refs. 3 and 6, respectively), which is explicitly stated in the review 1. In particular, the quotation marks in the above-mentioned sentence from ref. 1 are not intended to be connected to or indicative of this chronology. Finally, it is added that proton transfer between two specified water molecules embedded in liquid water was investigated in ref. 15 using the empirical valence bond (EVB) approach in terms of a two-state model (see also ref. 16 in relation to the parameterization of this particular model) and, by construction, the simulation refers necessarily to a localized protonic defect that is confined to a particular hydrogen bond. Shortly later, so-called extended 17 or multi-state 9 EVB models were successful in simulating actual proton transport within the three-dimensional hydrogen-bonded water network and thus charge migration and Grotthuss structural diffusion in liquid water 7–9, 11, 12 as reviewed in ref. 18. I am grateful to Noam Agmon for bringing ref. [5] to my attention and to Greg Voth for commenting on the two-state EVB model 15.