ASMS 2026: Phantom Metabolites and the Hidden Chemistry of Electrospray Ionization

Among the most thought-provoking presentations discussed at the 2026 American Society for Mass Spectrometry (ASMS) Conference in San Diego was new evidence suggesting that a substantial fraction of the unknown features observed in untargeted metabolomics experiments may not be metabolites at all. Instead, many spectral peaks may be artifacts generated during the electrospray ionization (ESI) process itself. The findings have significant implications for metabolomics, biomarker discovery, and the interpretation of liquid chromatography–mass spectrometry (LC–MS) data.

For more than two decades, metabolomics researchers have been confronted by the problem of the so-called "dark metabolome." Modern high-resolution LC–MS systems routinely detect thousands of spectral features in biological samples, yet only a relatively small percentage can be confidently identified. Even with expanding databases and increasingly sophisticated computational tools, such as artificial intelligence (AI), the majority of detected ions often remain unexplained.^1,2

Historically, these unknown signals were assumed to represent undiscovered endogenous metabolites. However, a recent publication by Song and co-workers has challenged this assumption by demonstrating that electrospray microdroplets can generate extensive chemical transformations during ionization, producing ions that were not present in the original sample.³ The work has generated considerable discussion within the metabolomics community and was a topic of interest throughout ASMS 2026.

The Challenge of Dark Metabolomics

Metabolomics has emerged as one of the most powerful tools for understanding biological systems because metabolites represent the downstream products of gene expression and cellular activity. As Patti, Yanes, and Siuzdak observed more than a decade ago, metabolomics can be viewed as the "apogee of the omics trilogy" because it provides a direct chemical snapshot of biological function.¹

Despite major advances in instrumentation and data analysis, metabolite identification remains one of the field's greatest challenges. Many untargeted LC–MS studies report that more than 80% of detected features cannot be assigned to known metabolites. This persistent gap has fueled speculation that a vast reservoir of undiscovered biochemical compounds exists within living systems.²

The new findings suggest another possibility: many of the unexplained ions may be reaction products formed during the ESI analytical measurement process itself.

Electrospray Ionization as a Reactive Environment

Electrospray ionization is generally considered a soft ionization technique because it produces intact molecular ions with relatively little fragmentation. Since its introduction, ESI has been viewed primarily as a mechanism for transferring molecules from solution into the gas phase while preserving their chemical identity.

Research over the past decade has revealed a more complicated picture. Studies from the laboratories of luminaries, such as Richard Zare and R. Graham Cooks, demonstrated that charged microdroplets generated during electrospray can dramatically accelerate chemical reactions compared with bulk solution environments.^4,5 These microdroplets possess unique properties, including large surface-to-volume ratios, strong electric fields, rapid solvent evaporation, and unusual charge distributions.

As a result, reactions that might require minutes or hours in bulk solution can occur within microseconds inside electrospray droplets.^4,5

The recently published study by Song and colleagues extends these observations and suggests that microdroplet chemistry may be far more widespread than previously recognized.³

The Origin of Phantom Metabolites

In their study, Song et al. investigated the formation of previously unexplained ions during electrospray ionization and demonstrated that numerous chemical reactions occur within the electrospray plume itself.³

The authors identified evidence for oxidation reactions, reduction reactions, condensation reactions, radical-mediated transformations, reductive amination processes, methylation reactions, and solvent-assisted acetylation reactions. These reactions generated molecular species that were subsequently detected by the mass spectrometer and could easily be interpreted as authentic metabolites.

One of the most revealing examples involved serotonin. The investigators observed multiple oxidation products, radical species, and secondary reaction products derived from serotonin during electrospray ionization. Many of these ions would be difficult to distinguish from naturally occurring metabolites using conventional untargeted metabolomics workflows.³

The authors concluded that electrospray ionization should not be viewed solely as a passive ion-transfer process but rather as a dynamic chemical environment capable of generating substantial molecular complexity.

Solvents as Contributors to Artifact Formation

An especially noteworthy finding was the role of common LC–MS solvents in generating artifact peaks.

Song and colleagues demonstrated that methanol and acetonitrile can actively participate in microdroplet reactions, producing methylated and acetylated products that were not present in the original sample.³ These observations suggest that solvent composition may influence not only chromatographic behavior and ionization efficiency but also the creation of previously unrecognized analytical artifacts.

For analytical chemists, this finding raises important questions regarding method development, interlaboratory reproducibility, and metabolite annotation strategies.

How Large Is the Problem?

The magnitude of the phenomenon appears substantial. In experiments involving a mixture of only 70 known metabolites, Song and colleagues detected 1744 ions, many of which originated from reaction pathways occurring during electrospray ionization.³

Perhaps even more significant was the effect on metabolite identification. By incorporating known microdroplet reaction pathways into their computational analysis, the investigators increased annotation rates in a metabolomics dataset from approximately 9% to greater than 50%.³

These results suggest that a considerable fraction of the dark metabolome may consist of predictable products of electrospray chemistry rather than previously unknown true endogenous compounds.

Implications for Metabolomics

The implications of this work extend throughout the metabolomics field. Untargeted metabolomics is widely used for biomarker discovery, disease diagnostics, pharmaceutical development, toxicology, environmental analysis, and systems biology. All of these applications rely on the assumption that detected ions correspond to compounds present in the sample.

If significant numbers of ions arise during ionization, researchers may need to reconsider how metabolomic data are interpreted. Future annotation software may require reaction-aware algorithms capable of distinguishing authentic metabolites from electrospray-generated byproducts.

Importantly, these findings do not undermine metabolomics. Rather, they provide a pathway toward improved annotation accuracy and a better understanding of the chemical processes occurring during measurement.

A New View of Electrospray Ionization

The discovery of widespread microdroplet-induced reactions may ultimately represent one of the most significant developments in ESI-MS interpretation since the original introduction of electrospray ionization. The work builds upon earlier observations of accelerated microdroplet chemistry,^4,5 while providing a direct explanation for many previously unexplained metabolomic features.

As noted by Derek Lowe in his commentary on the study, the findings provide a compelling explanation for numerous "phantom metabolites" that have puzzled researchers for years.⁶

The central message emerging from ASMS 2026 is clear: not every ion detected in a metabolomics experiment necessarily corresponds to a metabolite that existed in the original sample. Some may instead represent products of an extraordinarily active microchemical reactor operating within the electrospray plume itself.

If these observations are confirmed broadly across laboratories and instrument platforms, they may fundamentally change how scientists interpret untargeted LC–MS data and may significantly improve confidence in metabolite identification and biomarker discovery efforts.

References

1. Patti, G. J.; Yanes, O.; Siuzdak, G. Metabolomics: The Apogee of the Omics Trilogy. Nat Rev Mol Cell Biol 2012, 13, 263–269. DOI: 10.1038/nrm3314
2. Wishart, D. S. Emerging Applications of Metabolomics in Drug Discovery and Precision Medicine. Nat Rev Drug Discov 2016, 15, 473–484. DOI: 10.1038/nrd.2016.32
3. Song, X.; Xu, J.; Sun, C.; et al. Dark Reactions in Microdroplets Explain Widespread Artifacts in Metabolomic Profiling. ACS Meas Sci Au 2026, 6 (2), 311–323. DOI: 10.1021/acsmeasuresciau.5c00146
4. Lee, J. K.; Kim, S.; Nam, H. G.; Zare, R. N. Microdroplet Fusion Mass Spectrometry for Fast Reaction Kinetics. Proc Nat. Acad Sci U.S.A. 2015, 112, 3898–3903. DOI: 10.1073/pnas.1503689112
5. Yan, X.; Bain, R. M.; Cooks, R. G. Organic Reactions in Microdroplets: Reaction Acceleration Revealed by Mass Spectrometry. Angew Chem Int Ed 2016, 55, 12960–12972. DOI: 10.1002/anie.201602270
6. Lowe, D. Phantom Metabolites. In the Pipeline (Science Magazine Blog), September 2025. Available at: https://www.science.org/content/blog-post/phantom-metabolites (accessed June 2026).

This article was prepared June 2, 2026, following an exclusive interview with Prof. Gary Siuzdak of The Scripps Research Institute, La Jolla, CA. The interview was conducted at the 74th ASMS Annual Conference.