A new study by Zhang et al. in Langmuir (Langmuir 2025, 41, 37, 25146-25166; 10.1021/acs.langmuir.5c01994) compared their previously measured nucleation rates of ice and clathrate (gas) hydrates to the predictions of Classical Nucleation Theory (CNT) to probe the nature of the so-called "building unit" considered by CNT. The identity of this structural unit remains an outstanding question when using CNT as an interpretive framework. Their approach involved adjusting the entropy of dissociation of the building unit (Δse) so that their nucleation rate data, J, were linear in the coordinates (ln J - ΔseΔT/kT vs (1/(TΔT2)) suggested by CNT. Linearity enables the nucleation rate data to be interpreted using a single pair of the kinetic and thermodynamic parameters, A and B'. However, there is no reason to expect that a single pair of these parameters should characterize the nucleation sites contributing to the observed J over a given range of subcooling. Furthermore, Δse is strongly correlated with A and B' making it difficult to independently assess the reliability of such parameters obtained by least-squares regression. Finally, in the first stage of their data analysis, Zhang et al. fit a functional form to their experimental data which biases the resulting B' parameters obtained by subsequent regression to values near zero. We recommend alternative, independent methods of estimating Δse and treating it as a fixed quantity when analyzing measurements of the nucleation rate.
Barwood et al. (2026) studied this question.