/low-bias-quenching

Analyzing photoluminescence quenching in OLEDs operated at low biases.

Primary LanguageJupyter NotebookMIT LicenseMIT

Low-bias exciton quenching

This repository contains code for simulating photoluminescence quenching in organic light-emitting devices (OLEDs) operated at low biases. If you find this code useful, please cite our publication:

J. S. Bangsund, J. R. Van Sambeek, N. M. Concannon, R. J. Holmes, Sub–turn-on exciton quenching due to molecular orientation and polarization in organic light-emitting devices. Sci. Adv. 6, eabb2659 (2020). DOI: 10.1126/sciadv.abb2659

Photoluminescence quenching below device turn-on

Running this code

Most of the python code for the quenching simulations uses standard scientific python packages (numpy, matplotlib, scipy)--any python 3.5+ version of these packages should work. For an exact copy of the environment used in this publication, see the oled.yml environment file in https://github.com/jsbangsund/conda-envs.

To edit the optical simulations of the optical generation profile and outcoupling efficiency, you will need oledpy (https://github.com/jsbangsund/oledpy) or your own preferred optical simulation package. Otherwise, the simulated optical profiles used in our publication are saved in the Reference Data folder.