What The fluorescence of organic molecular emitters is suppressed by non-emissive excited states of the molecule that are thermodynamically favored over the emissive state in accordance with Hund’s rule. It has recently been found that in rare cases molecules can violate this fundamental rule of chemistry, such that the emissive state is lower energy than the non-emissive state. Unfortunately, there are currently no design strategies that enables us to find new molecules with this desirable property. In this project, I will search for new classes of molecular emitters that violate Hund’s rule and establish new design principles for optimizing the emissive properties of fluorescent molecules. Why The brightness and energy efficiency of organic light emitting diodes (OLEDs) are limited by non-emissive electronic states. Only 25% of the electricity that enters an OLED display is emitted as the bright colors we see on our smartphone screens. This limit can be exceeded only if the energy gap between the emissive and non-emissive excited states is inverted. The discovery of new classes of molecules that violate Hund’s rule may lead to new generations of more energy-efficient OLEDs with brighter colors. How To overcome the current lack of a design strategy, our research idea is to develop a high-throughput screening approach to screen molecular databases containing millions of molecules. A highly effective computational approach is required, which we can achieve using state-of-the-art machine learning methods. Upon identification of new types of molecules with inverted energy gap between the emissive and non-emissive state, we will further examine the emissive properties using high-level quantum chemical methods. Ideally, the finding of such molecules will lead us toward more systematic design principles for molecular emitters, which we can disseminate to experimental collaborators.