Chemically Engineered Carbon Nanotubes Map Class-Selective Metabolite Enrichment from Human Plasma

Abstract

The spontaneous adsorption of biomolecules onto nanoparticle surfaces has been extensively characterized at the protein level, but the metabolite corona remains poorly defined, in part because low annotation rates inherent to untargeted metabolomics limit systematic, class-level analysis. To address this analytical challenge, we developed a multimodal liquid chromatography–mass spectrometry workflow combined with machine-learning classifiers, extending metabolite class annotations from 10 to 45% coverage to the full detected set and enabling the first class-level mapping of metabolite corona composition across 25 chemically modified carbon nanotubes in human plasma. Complementary analytical conditions detected approximately 9,000 metabolite features, of which over 5,000 yielded valid corona-versus-plasma enrichment measurements. We find that polymer wrapping dominates corona composition, and covalent quantum well defects further modulate class-level enrichment in a structure- and chemistry-dependent manner. Our findings demonstrate that chemically modified carbon nanotubes can serve as tunable, selective scaffolds for metabolite enrichment, potentially enhancing the capability to recruit and detect structurally diverse, low-abundance small molecules in complex biofluids.