Mass Spectrometry Peels Back the Layers of Art History
MALDI imaging and machine learning decode paintings at molecular resolution
A new analytical workflow is allowing scientists to “read” paintings layer by layer – revealing not just pigments, but binders, degradation products, and even hidden conservation treatments. By combining MALDI mass spectrometry imaging (MSI) with machine learning, researchers have achieved unprecedented molecular detail from both historic and contemporary artworks.
The challenge is formidable: paintings are chemically complex, multilayered systems where organic binders, inorganic pigments, and environmental aging processes intertwine. Traditional techniques can identify materials, but often lack specificity or destroy fragile molecules. The new approach overcomes this by using MALDI-MSI, which enables label-free detection of molecules ranging from lipids to proteins with micrometer-scale spatial resolution. As the authors explain, the method delivers “an unprecedented level of molecular detail using a single technique.”
To make sense of the rich spectral data, the team developed a machine learning model – MSIpredictART – capable of automatically assigning pigment and binder compositions across layers. The model achieved high specificity (>90% for pigments, >95% for binders) and could distinguish visually identical layers that microscopy alone cannot resolve.
Applied to a 17th-century painting (The Marriage of the Virgin), the workflow reconstructed the full stratigraphy: from gesso ground layers and animal glue binders to iron-based pigments, gold leaf, and later restoration varnishes that had penetrated into the paint. Crucially, the technique identified both original materials and conservation interventions in a single analysis.
The method also proved powerful for modern art, disentangling complex pigment mixtures and revealing how artists blend materials to achieve specific colors – such as creating green hues from blue and yellow pigments rather than using a single green compound.
Mapping the Cochlea in Striking Molecular Detail
High-resolution imaging reveals distinct metabolite and lipid signatures across the inner ear’s tiny, complex structures
Researchers have developed a streamlined mass spectrometry imaging workflow that can map the molecular landscape of the mouse cochlea at 5-micrometer resolution – fine enough to resolve distinct tissues, fluid spaces, and even signatures within the organ of Corti.
The cochlea is a notoriously difficult structure to study. It is tiny, fragile, buried deep in bone, and packed with highly specialized cell types that support hearing. To tackle this, the team optimized a rapid sample preparation method using flash-frozen neonatal mouse heads and paired it with MALDI mass spectrometry imaging using a sublimated NEDC matrix. That approach outperformed conventional spraying, delivering stronger signals, less molecular delocalization, and better spatial resolution – especially for lipids and fatty acids.
The resulting images revealed a rich patchwork of metabolites and lipids across cochlear substructures, including the otic capsule, spiral ducts, nerves, and the organ of Corti. Some molecules were broadly distributed, while others showed sharply restricted localizations. Notably, the researchers detected differences between endolymph and perilymph and identified lipid and sulfatide patterns associated with the organ of Corti.
Fiber-optic MSI Tracks Apoptosis
Submicrometer mass spectrometry imaging reveals how anticancer drugs reshape lipids with striking single-cell heterogeneity
Researchers have built a mass spectrometry imaging platform that can map lipids inside single cells at roughly 800 nm resolution, offering a close-up view of how apoptosis rewires membrane biology. The system uses a commercially available single-mode fiber image relay to deliver a tightly focused laser beam from a working distance of at least 25 mm, avoiding contamination of the optics while maintaining submicrometer sampling.
Paired with plasmonic gold nanoparticles, the setup visualized lipid distributions in mouse brain, HeLa, and HepG2 cells. In cultured cancer cells, the team used the method to follow emodin-induced apoptosis over time and across dose. The data suggest that emodin does not trigger a sudden catastrophic collapse, but instead drives a progressive metabolic reprogramming: structural phospholipids such as PCs and PEs decline, lysophosphatidylcholine rises, and energy-storage lipids are mobilized in a controlled, stage-dependent way. That pattern contrasted with UV-induced apoptosis, which produced a more acute breakdown of membrane and lipid homeostasis.
The study also exposed marked variability between individual cells, between HeLa and HepG2 lines, and across drugs including luteolin, imatinib, and 5-fluorouracil – highlighting how mechanism-specific lipid fingerprints may help decode therapeutic response at single-cell resolution.
Nanoplastics May Be the Atlantic’s Biggest Plastic Burden
Ocean-basin survey finds PET, PVC, and PS nanoparticles throughout the North Atlantic water column, with estimated mixed-layer loads rivaling or exceeding previous global plastic budgets
Nanoplastics may make up a far larger share of marine plastic pollution than previously thought. In a transect spanning the North Atlantic – from the subtropical gyre to the European shelf – researchers detected polyethylene terephthalate (PET), polystyrene (PS), and polyvinyl chloride (PVC) nanoparticles at concentrations of roughly 1.5–32 mg m⁻³ throughout the water column.
The highest concentrations appeared in mixed-layer waters near Europe, pointing to strong coastal inputs, while intermediate waters in the subtropical gyre held more nanoplastic than comparable depths outside the gyre – suggesting an additional source from fragmentation of plastic already accumulating offshore. Bottom waters contained lower overall concentrations, but PET remained common even near the seafloor.
Using their mixed-layer measurements, the team estimates that nanoplastics in the temperate to subtropical North Atlantic could amount to 27 million tonnes – comparable to, or greater than, earlier estimates for macroplastics and microplastics in the entire Atlantic or even the global ocean.
Lehigh Opens Shared Mass Spec Hub for Health and Environment Research
New HEAL Service Center aims to support exposomics, metabolomics, and proteomics studies while giving students and partners access to high-resolution analytical tools
Lehigh University has launched the Health and Environmental Assessment Laboratory (HEAL) Service Center, a shared core facility designed to expand access to high-resolution mass spectrometry for faculty, students, academic collaborators, and industry. Housed in the university’s Health, Science, and Technology building, the center will support projects spanning environmental contaminants, metabolomics, and proteomics.
At the core of the new facility is a Thermo Fisher Vanquish LC coupled to a Q Exactive Orbitrap mass spectrometer, enabling both targeted and untargeted analyses across samples ranging from water, soil, and airborne particulates to plasma, urine, and tissue. According to faculty lead Gabrielle String, “This kind of access is essential if we want to accelerate discoveries that link environmental conditions to human health in data-driven ways.”
Beyond instrumentation, the center will offer method development, sample preparation, data analysis, preliminary data generation for grant applications, and end-to-end project support. It also has a training mission: students will gain hands-on experience in mass spectrometry workflows and data interpretation. “There’s no substitute for hands-on experience with processing samples, troubleshooting methods, and tackling research questions,” said Abe Moghaddam, who emphasized the center’s role in preparing students for careers with real-world impact.
Newsletters
Receive the latest analytical science news, personalities, education, and career development – weekly to your inbox.
