Abstract
Metabolomics, the study of metabolic profiles in a biological sample, has seen rapid growth due to advances in measurement technologies such as mass spectrometry (MS). While MS metabolite reference libraries have been generated for metabolomics applications, mass spectra alone are unable to unambiguously identify many metabolites in a sample; these unidentified compounds are typically annotated as "features". Surface-enhanced Raman spectroscopy (SERS) is an interesting technology for metabolite identification based on vibrational spectra. However, no reports have been published that present SERS metabolite spectra from chemical libraries. In this paper, we demonstrate that an untargeted approach utilizing citrate-capped silver nanoparticles yields SERS spectra for 20% of 80 compounds chosen randomly from a commercial metabolite library. Furthermore, prescreening of the metabolites according to chemical functionality allowed for the efficient identification of samples within the library that yield distinctive SERS spectra under our experimental conditions. Last, we present a reference database of 63 metabolite SERS spectra for use as an identification tool in metabolomics studies; this set includes 30 metabolites that have not had previously published SERS spectra.