Is the gap of $\mathrm{Ta_2NiSe_5}$ excitonic or lattice driven?

$\mathrm{Ta_2NiSe_5}$ has an orthorhombic-to-monoclinic transition at $T_c \sim 326\,\mathrm{K}$ with valence-band flattening and a gap of order 0.1 eV seen in photoemission; the excitonic and structural order parameters have the same symmetry and so coexist. Determine which instability drives the transition, i.e. whether a purely electronic (excitonic) susceptibility diverges at $T_c$ with the lattice frozen, or whether the cooperative zone-center phonon distortion is primary, and what fraction of the low-temperature gap each produces. An answer is an identified primary order parameter with a quantitative partition of the gap.

Observation or exclusion of a soft electronic collective mode at $T_{c}$ by Raman or resonant inelastic x-ray scattering, combined with a frozen-lattice many-body computation of the electronic susceptibility.