One of the first points made when teaching advanced undergraduates about isotope effects is that isotopic labeling does not alter the reaction because the potential energy surface is not changed, only the vibrational frequencies of the molecule. Today, Robert Nixon and Guillame de Bo proved that chemistry is nothing but exceptions with their short communication in J.A.C.S.
The authors start by describing an imidazoline mechanophore that they previously showed to have three different pathways of mechanical scission depending on the amount of polarization (number of F atoms) on the arene that is cleaved from the imidazoline scaffold.
In this work, they examine how replacing the hydrogen with a deuterium atom at the site of cleavage can also change the pathway of mechanical scission. And they prove that isotopes can indeed alter the reaction pathways a molecule takes (at least in force-modified potential energy surfaces).
Here's why. The computed potential energy surface shows a bifurcation after the bond-elongated transition state in which the C–H/D bond also elongates. It is at this point that the molecule can take either the heterolytic or concerted pathway. When molecule bond reaches this point with two pathways available, bond vibrations determine the road that is taken.
The concerted path is computed to require greater atom motion than the heterolytic path. Thus, the isotope that has a greater atom motion is also more likely to follow the concerted path. Thinking about vibrational frequencies and bond lengths, the heavier D atom has a lower vibrational frequency than the H atom, causing it to have less atomic motion and a lower probability of taking the concerted pathway. This small change is large enough to reverse the preferred mode of reaction for the labeled and unlabeled mono-fluorinated substrate.
Nixon and de Bo elegantly demonstrate that half of what is taught is true. The surface does not change with isotopic labeling. But this does not mean that the pathway that is followed remains the same.