Triplet–triplet annihilation (TTA) is one of the primary contributors to efficiency roll-off and permanent material degradation in phosphorescent organic light-emitting diodes. The two limiting case models typically used to quantify this quenching mechanism are multi-step Dexter and single-step Förster, which, respectively, assume ideal Fickian diffusion or perfect trapping of triplet excitons. For device-relevant guest doping levels (typically 5–12 vol. %), both significant diffusion of excitons and trapping due to spatial and energetic disorder exist, so neither conventional model fits experimental data well. We develop and validate an intermediate TTA model, which is a weighted average of the limiting cases of pure radiative decay (no TTA) and multi-step Dexter based TTA that returns an effective TTA rate constant and a parameter quantifying the portion of well-isolated excitons. Kinetic Monte-Carlo simulations and time-resolved photoluminescence measurements of an archetype host–guest system demonstrate that our intermediate model provides significantly improved fits with more realistic physical values, is more robust to variations in experimental conditions, and provides an analysis framework for the effects of trapping on TTA.

In this work, we report results of clustering studies across a range of length scales in Ir(ppy)3: mCBP organic light emitting diodes (OLEDs) deposited at a variety of substrate temperatures. We identify signatures of the light‐emitting guest molecules clustering at the local scale using atom probe tomography and large‐scale clusters using high‐angle annular dark‐field scanning transmission microscopy (HAADF‐STEM). Changes of to the small‐ and large‐scale clustering behavior are consistent with changes to conductivity and exciton annihilation processes observed at different substrate deposition temperatures.

No abstract