In organic light-emitting diodes, electrically generated excitations predominantly form triplet states. Recovering that energy as light through thermally activated delayed fluorescence is fundamentally an electronic-structure problem: how can molecular design control the energies, characters and interactions of excited states to enable efficient spin conversion without sacrificing strong emission?

I showed that the conventional two-state picture of TADF is incomplete for multiconfigurational excited states. The singlet–triplet energy gap depends not only on HOMO–LUMO exchange but also on the energetic separation between low-lying virtual orbitals. That led to a predictive expression for the gap based on ground-state parameters, turning molecular design into a more direct screening problem.

I then developed a complementary strategy for accelerating reverse intersystem crossing by controlling the orbital character of excited states. Restricting π-conjugation in carbonyl-containing frameworks promotes mixing between states of different orbital character and enhances spin–orbit coupling. Together these approaches allow excited-state energetics, spin conversion and emission properties to be controlled independently. Their predictive power has been demonstrated with experimental collaborators, giving emitters spanning deep-blue to red and OLEDs with external quantum efficiencies approaching 40%.

Singlet fission addresses the complementary problem in organic photovoltaics. Instead of converting multiple excitations into one photon, can one absorbed photon generate two triplet excitons, opening a pathway to more efficient solar energy conversion?

Using high-level many-body electronic-structure methods, I showed that singlet fission cannot be adequately described by a simple two-state or charge-transfer-mediated picture. Efficient fission emerges from the cooperative interplay of locally excited, charge-transfer and correlated triplet-pair states, with strong electron correlation and electron–vibration coupling jointly controlling their population and dynamics.

That picture also points toward design. Combining many-body calculations with systematic screening, I identified promising BN-doped π-conjugated chromophores and established how molecular packing controls the electronic couplings fission requires, with multidimensional exploration of molecular arrangements revealing that slip-stacked geometries strongly enhance vibronic population transfer. These results connect molecular electronic structure to supramolecular organization.

Generating and controlling excited states is only part of the challenge. For practical optoelectronic and photovoltaic technologies, those states must also remain chemically stable under prolonged operation. Excited-state reactions can open competing pathways involving bond dissociation, charge transfer, radical formation and other reactive intermediates, ultimately limiting the operational lifetime of organic devices.

My future research will extend the electronic-structure framework developed for exciton dynamics toward the mechanisms of excited-state reactivity and photochemical degradation: identifying how molecular structure controls the formation and evolution of reactive excited states, determining the microscopic pathways responsible for degradation and developing predictive descriptors for photochemical stability.

Rather than treating degradation as an empirical materials problem, I want to establish a molecular-level understanding of why organic optoelectronic materials fail and how their structures can be redesigned to resist it. That connects the two sides of the programme: controlling how excitations are generated, transferred and converted and controlling what happens to them afterwards.