A Meta‐Analysis of Published Brønsted Plots Supports Brønsted's “Conventional” pKa Values: 14.00 for Water and 0.00 for H+(aq)

Abstract

In the early 1900s, Brønsted used two different pairs of pKa values for water and the aqueous proton: He denoted 14.00 and 0.00 (respectively) as the “conventional” values, and 15.74 and −1.74 (respectively) as the “rational” values. Although the conventional values have been supported by both experiment and theory, the rational values have been employed in the vast majority of studies involving Brønsted plots up to the present day. Here we present a meta‐analysis of 67 Brønsted plots published from 1924 to 2010, revealing that rational pKa’ values more frequently appear as statistical outliers (a greater than 80 % rejection rate) compared to conventional values (a roughly 50 % rejection rate) based on Q‐tests. Furthermore, in 85 % of examined plots, the conventional points align more closely with best‐fit lines. An evaluation of Hupe and Pohl's Marcus plots also shows a somewhat better fit with the conventional rather than the rational pKa’ values. Finally, in the bond length versus pKa investigations published by Brydon et al., all the rational points were statistically rejected, whereas the conventional values were accepted. From our statistical analyses, we conclude that a convincing majority of experimental evidence from published Brønsted plots leads one to question the validity of Brønsted's rational pKa’ values for water and the aqueous proton (15.74 and −1.74, respectively).