Scrubbing ions with molecules: kinetic studies of chemical noise reduction in mass spectrometry using ion-molecule reactions with dimethyl disulfide

The kinetics and product distributions of the reactions of dimethyl disulfide (DMDS) have been investigated with a group of chemical background ions commonly observed in atmospheric pressure ionization (API) mass spectrometry (MS) in order to assess the value of this molecule in filtering (or "scrubbing") these ions by changing their mass/charge (m/z) ratio. The measurements were taken with a novel electrospray ionization/selected ion flow tube/QqQ tandem mass spectrometer. The background ions studied include those with m/z 42 (protonated acetonitrile, ACN), 83 (protonated ACN dimer), 99 (protonated phosphoric acid), 117 (water cluster of m/z 99), 131 (methanol cluster of m/z 99), 149 (protonated phthalic anhydride, formed from the phthalates), and 327 (protonated triphenyl phosphate). In addition, reactions of DMDS have been studied with two model analytes--protonated caffeine and doubly protonated bradykinin--in order to assess the selectivity of DMDS reactivity. All the measurements were taken at 295 +/- 2 K in helium buffer gas at a pressure of 0.35 +/- 0.01 Torr. DMDS was observed to react efficiently with m/z 42 (ACNH+), 149 (from phthalates), and 99 (protonated phosphoric acid), with k/kc=0.91, 0.47, and 0.38, respectively. Only proton transfer was observed with ACNH+, followed by the secondary reaction of DMDSH]+ with DMDS to yield CH3S-S(CH3)-SCH3]+. Ligation of DMDS was the dominant primary channel observed for the reaction of the m/z 149 background ion; however, some proton transfer also was observed. Both of these primary product ions react further with DMDS to yield CH3S-S(CH3)-SCH3]+, the structure of which we have determined computationally using DFT calculations. Only the sequential ligation with two DMDS molecules was observed for the reaction of the m/z 99 ion. Reactions of DMDS with m/z 117 H3PO4 + H + H2O]+ and m/z 131 H3PO4 + H + MeOH]+ were observed to proceed with k/kc=0.71 and 0.058, respectively. Ligand substitution of DMDS for H2O predominated ( approximately 94%) over DMDS ligation ( approximately 6%) in the reaction with m/z 117, while only DMDS ligation was observed for the reaction of m/z 131 with DMDS. In contrast, the reactions of DMDS with ions of m/z 83 (protonated dimer of ACN) and 327 (protonated triphenyl phosphate) were extremely inefficient, with k/kc=0.0042 and 0.0079, respectively. The higher reactivity of DMDS toward ACNH+ (m/z 42) compared to (ACN)2H+ (m/z 83) is attributed to the lower proton affinity of the unsolvated ACN. The reactivity of DMDS toward the two model analyte ions studied-protonated caffeine and doubly protonated bradykinin-was negligible, with k/kc=0.0073 and 0.010, for the respective reactions. These results suggest that, under appropriate reagent pressure conditions, DMDS can be an appropriate reagent for chemically filtering out many common API-MS background ions, without significantly affecting the observed intensity of analyte peaks.