Neural network classification of mobility spectra for volatile organic compounds using tandem differential mobility spectrometry with field induced fragmentation.

A spectral library of field induced fragmentation (FIF) spectra for 45 oxygen-containing volatile organic compounds from 5 chemical classes was obtained using tandem differential mobility spectrometry (DMS). Protonated monomers were mobility isolated in a first DMS stage, fragmented with electric fields >10,000 V/cm in a middle (or reactive) stage, and mobility characterized in a second DMS stage. Other spectral libraries were obtained for protonated monomers and for complete mobility spectra from a single DMS stage. Neural networks from Python/Tensorflow software, prepared in-house, and from commercial NeuralWorks Professional II/PLUS were trained to assign spectra into a chemical class. The success at classification was determined for familiar and unfamiliar spectra from these three libraries. Classification test scores were best with FIF spectra with >0.99 for familiar compounds and 0.52 for unfamiliar compounds and were consistent with neural network learning of structural information from fragment ions when compared to other spectral libraries. Radar charts are introduced as measures of classification and as a tool to explore mis-classification. This work shows that ion fragmentation with multi-stage tandem DMS portends molecular identification with the portability and robustness of ambient pressure ion mobility analyzers.

Field induced fragmentation spectra from reactive stage-tandem differential mobility spectrometry.

A planar tandem differential mobility spectrometer was integrated with a middle reactive stage to fragment ions which were mobility selected in a first analyzer stage using characteristic compensation and separation fields. Fragmentation occurred in air at ambient pressure of 660 Torr (8.8 kPa) with electric fields of 10 to 35 kV cm-1 (E/N of 52 to 180 Td) between two 1 mm wide metal strips, located on each analyzer plate between the first and second mobility stages. Field induced fragmentation (FIF) spectra were produced by characterizing, in a last stage, the mobilities of fragment ions from protonated monomers of 43 oxygen-containing volatile organic compounds from five chemical classes. The extent of fragmentation was proportional to E/N with alcohols, aldehydes, and ethers undergoing multiples steps of fragmentation; acetates fragmented only to a single ion, protonated acetic acid. In contrast, fragmentation of ketones occurred only for methyl i-butyl ketone and 2-hexanone. Fragment ion identities were supported by mass-analysis and known fragmentation routes and suggested that field induced fragmentation at ambient pressure can introduce structural information into FIF spectra, establishing a foundation for chemical identification using mobility methods.