Noncovalent interactions determine both the structure and properties of molecular aggregates ranging from clusters and supramolecular nanostructures, to biosystems and condensed phases. Consequently, theoretical methods for noncovalent interactions in excited-state complexes are essential for understanding of photoactive materials and their in-silico-driven design for technological applications, such as organic light-emitting diodes or fluorescent probes. Interactions in complexes involving excited states pose a serious challenge for ab initio methods as compared to ground states. First, they require accounting for the multiconfigurational character of excited state wave functions and going beyond an independent-electron approximation. Second, resonance interactions, which are unique for excited states, demand treatment of degenerate states. The latter challenge has recently been addressed with energy decomposition schemes (EDAs). However, these approaches are based on single-reference linear response theories and are seriously limited by altogether missing or recovering only part of effects which are crucial for an accurate description of noncovalent interactions.

The symmetry-adapted perturbation theory (SAPT) is one of the most successful methods for interaction energy calculations in noncovalently bound complexes. In SAPT the essential components of the interaction: electrostatic, exchange (Pauli repulsion), induction and dispersion energies, are obtained directly from monomer properties. SAPT based on density functional theory (DFT) generates potential energy surfaces of accuracy on a par with high-level coupled cluster results at a much lower computational cost. This has resulted in numerous applications of the method, the most spectacular ones being structure prediction of organic crystals. Unfortunately, DFT-based SAPT is a ground state theory, and by construction, it is not applicable to systems in electronically excited states.

The aim of this project is to develop a SAPT variant for the description of noncovalent interactions in excited state complexes. The proposed method will be based on range-separated multiconfigurational density functional theory. This approach rigorously combines the accuracy of DFT with multiconfigurational wave functions. In its degenerate formulation the new SAPT method will extend to excimers and exciplexes, i.e., systems with resonance interactions that currently are beyond the capabilities of all existing SAPT formulations. Compared to EDAs, it will offer a well-defined and accurate description of the dispersion energy and the possibility to target multireference states, including excimers and exciplexes.
The novel method will provide an efficient description of large systems, in particular of organic light-emitting crystals and optoelectronic materials. The synthesis of such materials can be theory-driven, and these studies will help to establish an intrinsic connection between the structure of light-emitting molecular assemblies and their optical properties.