Light absorption sits at the heart of many of our hopes for a more sustainable future. Whenever we want to turn sunlight into electricity, into fuel, or into a chemical reaction that cleans water or captures CO₂, the first step is always the same: a material has to absorb light. Getting that first step right is what a whole field of research is built on.

In many ways, measuring how much light a material absorbs is a routine procedure. In the ultraviolet and visible range - roughly the colors our own eyes respond to - we shine light through a sample and record how much comes out the other side. The result is a simple curve, and typically a chemistry lab can produce one in an afternoon.

But behind even a "simple" measurement can lie a surprising number of microscopic phenomena. When a material absorbs light, an electron is excited to a higher energy level and leaves behind an empty space, a "hole". The electron and the hole attract one another and can travel together as a bound pair, which we call an exciton. At the same time, the atoms of the material are never perfectly still: they vibrate in collective patterns we call phonons. Excitons and phonons are not independent. They can interact, and this interaction affects the absorption curve we end up measuring.

Theory can describe each of these phenomena on its own. What is still an open research direction is combining them well enough to reproduce what an experiment actually sees. That is difficult for any material, and it is especially difficult for metal-organic frameworks, or MOFs.

MOFs are materials built from metal centers connected by organic molecules. Their tunable and porous nature makes them appealing candidates for light harvesting and photocatalysis, but it also makes their optical behavior complicated. When we started this project, we knew that light absorption in MOFs is a complex phenomenon, but we simply did not know to what extent excitons and phonons mattered. So we chose a system that could help us answer these open questions.

That system is SU-101, a bismuth-based MOF that is photocatalytically active. This MOF is synthesized with water molecules bound to its metal sites, and that water can be removed in a process that is known as activation. This raised yet another question. Does the bound solvent affect light absorption at all? From the experiments alone, the answer was not so clear. The absorption peaks did change in intensity when we removed the water, but we could not say why, or how.

This is where our computational approach could help. By computing excitons and phonons, we could look underneath the measured spectra. We found that removing the bound water switches on a different exciton-phonon coupling mechanism. In the hydrated and the dehydrated material, different atomic vibrations end up coupling to the exciton, and this shapes how the exciton looks like inside the framework. That change in how excitons and phonons interact is what shows up, indirectly, as the change in absorption we measure, and our calculations followed the same trend as the experiment.

In the end, this work taught us two things. First, excitons and phonons can be of great importance for materials like MOFs. Second, their effect can shift with a routine protocol in MOF synthesis, i.e., removal of bound solvent, even when the difference in the experiment is not loud or clear. For us, that is the real message: a modest change in a spectrum can carry a rich microscopic story, on which theory and experiment together can help explain.