Using Bayes' Theorem to Evaluate Time-Resolved Photoluminescence Data in n-DimensionalParameter Space
Manuel Kober-Czerny a, Akash Dasgupta a, Henry Snaith a
a Department of Physics, University of Oxford, Clarendon Laboratory, Parks Road, OX1 3PU, United Kingdom
Materials for Sustainable Development Conference (MATSUS)
Proceedings of MATSUS Fall 2023 Conference (MATSUSFall23)
#MHPN3 - Fundamental Advances in Metal Halide Perovskites and Beyond: new materials, new mechanisms, and new challenges
Torremolinos, Spain, 2023 October 16th - 20th
Organizers: Paola Vivo, Qiong Wang and Kaifeng Wu
Poster, Manuel Kober-Czerny, 065
Publication date: 18th July 2023

Time-resolved photoluminescence (TRPL) is a powerful technique to probe transientbehaviour of charge carriers in semiconductors after photoexcitation. In most reports however, a mono- orbiexponential models are used to evaluate the data obtained from TRPL. This is of course a crude model, as itwill only allow the extraction of a single lifetime (or two lifetimes) from the data, which doesn't necessarily havea physical meaning. Following a recent report, we attempted using Bayes' theorem to evaluate the TRPL datainstead.
In this approach all possible parameter combinations are used to simulate TRPL traces and each is assigned alikelihood, which depends on their overlap with the TRPL data. However, this process becomes very timeconsuming with a larger parameter space.
Here, we show an updated method, in which we include a Markov-Chain Monte-Carlo (MCMC) algorithm toexplore the n-dimensional parameter space. The computational time is reduced, as not all combinations ofparameters need to be simulated, but only the ones with a higher probability of overlapping with the data. Thisallows us to run this evaluation of the TRPL data on a normal laptop within a few hours. The power of thetechnique becomes apparent, after we evaluate a set of fluence-dependent TRPL data with up to 11parameters. For each parameter we obtain a probability distribution as well as a correlation plot with all otherparameters. This will be helpful for the investigation of novel and existing materials , as it allows a deeperinsight into underlying physics that may have been undiscovered before.
We then move on to compare the extracted parameter values with other methods, such as photoluminescencequantum yield (PLQY) or transient photoconductivity (TPC).
Overall, we show that this method of evaluating data can be extremely powerful, as it won't be limited by'overfitting' and in the case of TRPL allows the extraction of a variety of fundamental parameters from just asimple optical measurement.

© FUNDACIO DE LA COMUNITAT VALENCIANA SCITO
We use our own and third party cookies for analysing and measuring usage of our website to improve our services. If you continue browsing, we consider accepting its use. You can check our Cookies Policy in which you will also find how to configure your web browser for the use of cookies. More info