Metal-organic frameworks (MOFs) are promising materials for light-driven applications. While their optical properties have been widely investigated experimentally, the relationship between optical absorption, charged excitations, and the underlying electronic structure remains difficult to assess computationally. In particular, most computational studies of MOF photophysics have compared ground-state DFT band gaps, often obtained using standard hybrid functionals such as PBE0 or HSE06, with experimental UV-vis absorption onsets. Although these methods can sometimes appear to reproduce experiment, such agreement is fortuitous since the quantities being compared are not equivalent. Experimental absorption onsets reflect neutral excitations and therefore include excitonic and vibrational effects, which are expected to be significant in MOFs. In contrast, the fundamental gap describes charged excitations and determines the thermodynamic driving force for charge separation. Distinguishing these quantities is essential for understanding whether photoexcitation in MOFs produces bound excitons or charge carriers that can participate in light-driven processes. The goal of this work is to clarify the relationship between ground-state electronic structure, optical absorption, and charge separation in MOFs by explicitly computing both charged and neutral excitations, along with the excitonic and vibrational effects that connect them to experiment.

Here, we calculate charged and neutral excitation energies in a representative set of ten widely studied MOFs. Fundamental gaps are computed using many-body perturbation theory within the \({G_0W_0}\) approximation and compared with non-empirically tuned range-separated hybrid functionals, including Wannier-localized optimally tuned screened range-separated hybrid (WOT-SRSH) and doubly screened hybrid (DSH) functionals. Optical gaps and absorption spectra were then obtained using the Bethe–Salpeter equation and time-dependent DFT using the DSH functional (TD-DSH), with zero-point and finite-temperature phonon renormalization incorporated through finite-difference electron–phonon coupling calculations.

WOT-SRSH and DSH reproduce \(G_0W_0\)@PBE fundamental gaps well, with mean absolute errors of 0.31 and 0.42 eV, respectively, and strong linear correlations across the MOF set. For neutral excitations, TD-DSH and \(G_0W_0\)@PBE-BSE optical gaps show improved agreement with experiment after phonon corrections, whereas conventional TD-PBE0 underestimates exciton binding energies. Across the MOF set, excitons are strongly bound, often exceeding 1 eV, demonstrating that optical and fundamental gaps must be treated as distinct quantities when evaluating MOFs for light-driven applications. These results establish practical computational routes for predicting charged and neutral excitations in MOFs and show that apparent agreement between hybrid DFT band gaps and absorption measurements can obscure the underlying excited-state physics.