Benchmarking MRSF-TDDFT and ΔSCF-DFT Against Wavefunction Methods: the S1 Decay Pathways of the DFHBDI Chromophore

<jats:p> The electronic structure of organic chromophores governs their optical properties and excited-state decay pathways, yet its accurate computational characterization remains challenging. Here, we benchmark the performance of the single-reference ΔSCF-DFT and multireference MRSF-TDDFT methods against established wavefunction-based approaches for the DFHBDI anion (3',5'-difluoro-4’-hydroxybenzylidene-2,3-dimethyl-imidazolinone), a fluorinated GFP chromophore analogue used in fluorescent RNAs. We performed a detailed analysis of the S₁ potential energy surface of DFHBDI, including the minima, saddle-points and minimum-energy S <jats:sub>0</jats:sub> -S <jats:sub>1</jats:sub> conical intersections (MECIs), associated with both P- and I-twisting decay pathways. We employed MRSF-TDDFT and ΔSCF-DFT with a range of exchange-correlation functionals, the multireference XMCQDPT2 method, and in addition the single-reference TD-DFT and CC2 approaches. All methods consistently predict that I-twisting constitutes a more favorable decay path than P-twisting. At the same time, the methods diverge significantly in their estimates of the energies of the P-twisting pathway: functionals predicting a less stabilized P-twisted minimum also predict a higher-energy MECI. Overall, the energies of the P-twisting pathway suggest that initial P-twisting does not necessarily commit the system to decay through this channel and that I-twisting is in fact the dominant quenching mechanism. MRSF-TDDFT, particularly with the DTCAM-VEE functional, and ΔSCF-DFT with the M06-2X and CAM-B3LYP functionals yield S <jats:sub>0</jats:sub> -S <jats:sub>1</jats:sub> energies in close agreement with the CC2 benchmark, for the planar and twisted S <jats:sub>1</jats:sub> minima, and with the XMCQDPT2 method for the P- and I-MECI. Our results demonstrate that when choosing the best-performing functional MRSF-TDDFT and ΔSCF-DFT are computationally efficient and accurate methods for modeling the complex, multi-pathway photophysics of biologically relevant chromophores with charge-transfer character. </jats:p>