What drives the MIT?: disentangling pre-transitional fluctuations in VO₂

Metal–insulator transitions (MITs) often occur concomitantly with structural changes, making it difficult to disentangle the underlying mechanism. Two competing scenarios lie at the core of a longstanding debate: (i) the transition is driven purely by electronic interactions, or (ii) it is driven by electron–phonon coupling. These two scenarios should feature very different types of pre-transitional fluctuations as the MIT is approached: they should be predominantly electronic in nature in the second case. Using VO₂ as a case study, we investigated fluctuations within the metallic state by measuring diffuse X-ray scattering. We find that pre-formed structural dimers give rise to a strong diffuse scattering signal that follows the Thomson scattering energy dependence across the K-edge. Considering both the shape of the diffuse features and experimental noise limitations, our results rule out purely electronic fluctuations with coherence lengths <2 unit cells or >4–6 unit cells, instead favouring electron–phonon coupling as the driving mechanism of the transition.