Paintings, and their chemistry, can change long after the last brushstrokes. For conservators, it can be tricky to discern the pigments and binders used and any chemical changes that have happened since. Now researchers have paired matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI-MSI) with machine learning to map the molecules in paint layers from a historical artwork from around 1690 (Sci. Adv. 2026, DOI: 10.1126/sciadv.adz4427).Many methods that researchers use to pin down the organic and inorganic compounds in paintings are limited in their identification power or can destroy fragile molecules before they can be mined for clues. MALDI-MSI, often used for biological samples, seemed like “the perfect tool” to get better intel on a painting’s layers and how they interact, says study coauthor Julie Arslanoglu, an organic chemist at the Metropolitan Museum of Art. The understudied chemistry of interactions is key to figuring out the conditions, such as lighting and temperature, needed to preserve art, she says.The researchers say this is the first time MALDI-MSI has been used on a historical painting. To do this, the team had to adapt the method for samples from oil paintings, including figuring out how to slice a cross section of brittle paint and extending the databases needed to interpret mass spectra to include paint degradation products. The team also paired the chemical imaging with machine learning to map layers in a sample, training the model on painted layers that had been prepared some 15 years ago.Using their approach, the researchers identified six layers
S. Wilke (Mon,) studied this question.