Predicting Photoluminescence Properties of OLEDs for High Throughput Screening

Widespread adoption of organic light emitting diodes (OLEDs) displays is hindered by the short operational lifetime of blue emitters in comparison to red and green ones despite their high energy efficiency, color contrast and the possibility for flexible and transparent displays. The discovery of new photoluminescent molecules that emit in the 450-465~nm range can be accelerated by high throughput screening (HTPS) and matter modeling. However, existing first principle prediction models have only been applied to a limited number of specific systems and are not suitable for HTPS. In this article, we develop a systematic methodology to predict wavelengths and photoluminescence quantum efficiency (PLQY) from first principle for any blue-green irridium complexes using density functional theory. We validate our predictions on a set of 450 molecules for which in-solution emission wavelengths and PLQY have been reported in literature. We show that our methodology is sufficiently versatile, efficient and accurate to be utilized on large amounts of data and suitable for large scale data generation for machine learning training.