Reducing background ion burden in tributylamine ion-pairing LC-MS improves signal intensity and feature coverage in metabolomics
Background chemical ions are a pervasive but often underappreciated limitation in LC-MS metabolomics, where they can suppress analyte signal, obscure endogenous metabolites, increase spectral complexity, and consume MS/MS acquisition events. Tributylamine (TBA) ion-pairing reversed-phase LC-MS provides stable retention and broad coverage of polar anionic metabolites, including central carbon intermediates, nucleotides, cofactors, and bile acids, but the back-ground burden introduced by the ion-pairing reagent itself has not been systematically addressed. Here, we identify commercial TBA as a major source of nonbiological contaminant ions and develop a practical strategy to reduce background burden while preserving metabolite coverage. Serial solid-phase extraction of TBA using orthogonal reversed-phase, strong anion-exchange, and strong cation-exchange sorbents removed chemically diverse contaminants, including isobaric background ions that interfered with endogenous hydroxybutyrate isomers. We further optimized the workflow by reducing medronic acid concentration, restricting medronic acid to the organic mobile phase, replacing phosphoric-acid column conditioning with metal-passivated column hardware, and adding EDTA to the sample reconstitution solvent to improve citrate detection. In mouse liver extracts, the optimized method increased signal intensity for most annotated metabolites and improved the fraction of full-scan ion current attributable to target analytes. Method optimization also altered compound-specific retention behavior, resolving some co-elution-based interferences while introducing new suppression relationships for selected analytes. Across mouse liver, human B lymphocytes, and NIST SRM 1950 plasma, the optimized workflow increased total feature detection by 45%, 72%, and 42%, respectively, and improved the number of low-variance features, precursors with data-dependent MS/MS spectra, and MS/MS library matches. These findings establish background-ion mitigation as a central design principle for LC-MS method development. More broadly, this work provides a generalizable framework for identifying, reducing, and validating reagent- and additive-derived background to improve targeted and untargeted LC-MS data quality.