The study of secondary isotope substitutions on adjacent carbon atoms affecting the tunnelling rate in acetylacetone

Heavy-atom isotope effects on hydrogen-tunnelling splittings have been reported for malonaldehyde and tropolone but not for acetylacetone, where the splitting has not yet been measured. We predict the effects of 13C substitution at positions not directly attached to the transferring proton on its tunnelling splitting. The model is a relaxed one-dimensional reaction path on a semiempirical GFN2-xTB surface that is scaled to the coupled-cluster barrier, with the parent splitting fixed by a diffusion Monte Carlo calculation. It is tested against instanton calculations and the experimental 13C effects in malonaldehyde, and against a fulldimensional ring-polymer instanton calculation for acetylacetone on the same GFN2-xTB potential energy surface, free of the model’s fitted parameter. Substitution of 13C at both carbons bonded to the oxygens decreases the splitting by 2.1% (model range 1.5–2.5%; instanton 2.3%), and a single oxygen-bonded 13C by 1.1%. The central-and methyl-carbon isotope effects are not established. These two oxygen-bonded carbons move with the proton during tunnelling, but the instanton fluctuation prefactor cancels about a third of this kinematic effect. In malonaldehyde the model reproduces the instanton results for substitution at the oxygen-bonded carbons and the measured total effect for all three carbons, although the only positionresolved measurement gives about half as much for the oxygen-bonded pair. Predicted shifts of 0.69 cm−1 (20.5 GHz) for 2,4-13C2-acetylacetone and about half that for the naturally abundant singly labelled species present an experimental test.