Home/ Research/ Functional Imaging
Research Line 02 · Imaging & Materials Science

Functional Imaging of
Optoelectronic Devices

How do nanoscale heterogeneities in solar cell materials — invisible to conventional characterisation — determine where and why devices lose efficiency?

Project PEROVSKOPE
Funding FWO Senior Research Project
My Role Co-Lead — CLIM development, imaging & analysis
Lab Hofkens Lab, KU Leuven
Status Active
01 Overview
Key method developed in this line: CLIM — Correlation Clustering Imaging

Halide perovskites are among the most promising materials for next-generation solar cells and LEDs — yet their efficiency remains limited by spatial heterogeneity that conventional characterisation methods cannot resolve. Different regions of the same thin film behave differently under illumination, and those differences propagate directly into device-level performance losses.

This research line develops and applies advanced optical imaging methods to make those differences visible — quantitatively, non-invasively, and at the length scales that matter. The central method is CLIM (Correlation Clustering Imaging), which segments a standard widefield photoluminescence movie into functional domains based on spatiotemporal intensity correlations. The resulting cluster maps reveal the grain microstructure of the photoactive layer — without electron microscopy, without labels, and without destroying the device.

The clusters revealed by CLIM correspond directly to the grain boundaries visible in SEM — but extracted purely from optical fluctuations, non-invasively, and even inside a working solar cell under one-sun illumination.

Beyond CLIM, the line is expanding into super-resolution approaches, deep learning for on-the-fly analysis, and in-operando characterisation of complete devices under electrical bias — bridging the gap between fundamental photophysics and applied device engineering.

02 The Concept
CLIM concept: widefield photoluminescence movie of a perovskite film segmented into correlation clusters matching the grain microstructure observed in SEM
CLIM concept. A standard widefield PL movie of a halide perovskite film (left) is segmented into correlation clusters (centre) — regions sharing correlated intensity fluctuations in time. The resulting cluster map matches the grain microstructure revealed by SEM (right), extracted entirely non-invasively. Image placeholder — replace with Content/clim-concept.jpg.

The core idea is straightforward: pixels within the same perovskite grain share correlated photoluminescence fluctuations (blinking), while pixels in adjacent grains do not. By clustering pixels according to these temporal correlations rather than by intensity alone, CLIM segments the film into its functional building blocks — directly reporting carrier dynamics, recombination, and defect distribution at the grain level.

The approach works on bare films, on complete devices under electrical bias, and across different perovskite compositions and morphologies. It is fully non-invasive and operates with a standard widefield fluorescence microscope.

03 Active Projects
01

CLIM — Algorithm Development

Improving and extending the correlation clustering algorithm: automated threshold optimisation, super-resolution extension via SOFI-inspired approaches, and deep learning for on-the-fly analysis at the microscope.

02

In-Operando Device Imaging

Mapping functional domains in complete perovskite solar cells and LEDs under operational conditions — as a function of applied voltage, current density, and illumination intensity.

03

Correlative Microscopy

Sequential and simultaneous correlative measurements combining optical (PL, lifetime, Raman) and electron microscopy on the same sample region, to directly connect structure and function.

04

Self-Healing & Degradation

Imaging the microscopic dynamics of defect formation and healing at perovskite–charge transport layer interfaces under illumination, electrical bias, and thermal cycling.

CLIM Halide Perovskites Widefield Fluorescence In-Operando Imaging Super-resolution Deep Learning Solar Cells · LEDs Correlative Microscopy
04 Collaborators
Host Lab
KU Leuven
Prof. Johan Hofkens
Host · microscopy infrastructure · device expertise
Key Collaborator
Lund University, Sweden
Prof. Ivan Scheblykin
Perovskite photophysics · PL fluctuation analysis · CLIM co-development
Key Collaborator
KU Leuven
Dr. Sudipta Seth
CLIM application · halide perovskite thin films · device characterisation
Collaborator
KU Leuven
Dr. Elke Debroye
Perovskite materials · correlative microscopy
Collaborator
KU Leuven
Toon Van Roy
CLIM development · data analysis · perovskite films
Collaborator
KU Leuven
Dr. Tejmani Behera
Self-healing in perovskite heterostructures · device imaging
Collaborator
TU Dresden, Germany
Prof. Yana Vaynzof
High-efficiency perovskite device fabrication · in-operando measurements
Collaborator
Tokyo Institute of Technology
Prof. Martin Vacha
Single-molecule spectroscopy · perovskite photophysics
05 Selected Publications
Chemical & Biomedical Imaging · 2025
S. Seth, B. Louis, K. Asano, T. Van Roy, M. B. J. Roeffaers, E. Debroye, I. G. Scheblykin, M. Vacha, J. Hofkens · Chemical & Biomedical Imaging · 2025
arXiv · 2025 · Preprint
Microscopic Intricacies of Self-Healing in Halide Perovskite–Charge Transport Layer Heterostructures
T. Behera, B. Louis, L. Paesen, R. Vanden Brande, K. Asano, M. Vacha, M. Roeffaers, E. Debroye, J. Hofkens, S. Seth · arXiv:2510.11948 · 2025
ACS Energy Letters · 2019
A. Merdasa, A. Kiligaridis, C. Rehermann, M. Abdi-Jalebi, J. Stöber, B. Louis, M. Gerhard, S. D. Stranks, E. L. Unger, I. G. Scheblykin · ACS Energy Letters, 4, 1370–1378 · 2019

For the complete record, see my Google Scholar profile.

Back to borislouis.com