Ion Heating in Advanced Dielectric Barrier Discharge Ion Sources for Ambient Mass Spectrometry

Dielectric barrier discharges (DBD) are highly versatile plasma sources for forming ions at atmospheric pressure and near ambient temperatures for the rapid, direct, and sensitive analysis of molecules by mass spectrometry (MS). Ambient ion sources should ideally form intact ions, as in-source fragmentation can limit sensitivity, increase spectral complexity, and hinder interpretation. Here, we report the measurement of ion internal energy distributions for the four primary classes of DBD-based ion sources, specifically DBD ionization (DBDI), low-temperature plasma (LTP), flexible microtube plasma (FμTP), and active capillary plasma ionization (ACaPI), in addition to atmospheric pressure chemical ionization (APCI) using para-substituted benzylammonium thermometer ions. Surprisingly, the average extent of energy deposited by the use of ACaPI (90.6 kJ mol–1) was ∼40 kJ mol–1 lower than the other ion sources (DBDI, LTP, FμTP, and APCI; 130.2 to 134.1 kJ mol–1) in their conventional configurations, and slightly higher than electrospray ionization (80.8 kJ mol–1). The internal energy distributions did not depend strongly on the sample introduction conditions (i.e., the use of different solvents and sample vaporization temperatures) or the DBD plasma conditions (i.e., maximum applied voltage). By positioning the DBDI, LTP, and FμTP plasma jets on axis with the capillary entrance to the mass spectrometer, the extent of internal energy deposition could be reduced by up to 20 kJ mol–1, although at the expense of sensitivity. Overall, the use of an active capillary-based DBD can result in substantially less fragmentation of ions with labile bonds than alternate DBD sources and APCI with comparably high sensitivity.

S-2 Figure S1. Diagram of ion source to introduce the benzylamines through the DBD plasma of the ACaP ion source rather than in the center of the halo plasma. The inner electrode has an outer diameter of 0.907 mm and inner diameter of 0.603 mm.
This configuration permits the analytes to go through the plasma and directly interact with the more energetic portion of the plasma. The vacuum from the capillary entrance to the MS ensures the supply of ambient air towards the inner electrode to sustain the plasma.
S-3 Figure S2. Effects of the solvent composition in the internal energy distribution for two representative plasma ion sources. Methanol and water mixtures of 10:90, 50:50 and 90:10 were evaluated for DBDI Orthog (red traces) and ACaPI (green traces). Figure S3. Effects of the vaporizer temperature (175º, 225 and 275ºC) in the internal energy distributions observed for the evaluated plasma-based ion sources: a) ACaPI, b) APCI, c) DBDI Orthog, d) LTP Orthog and e) FμTP Orthog. f) Thermospray vaporized TM ions.
The ion source and the in-source collision induced dissociation in the mass spectrometer first vacuum stages can favor different degrees of dissociation. The fragment ion of the most labile TM ion (m/z = 121.1) was observed. However, as observed in Figure S2f, the abundances was, at least, three order of magnitude lower compared to the plasma-based ion sources, having negligible contributions in the internal energy distribution.
S-5 Figure S4. Average of the internal energy deposition (E max ) for a range of working voltages of ACaPI (black), DBDI (red) and FμTP (blue).
Internal energy distributions for different discharge gases. a) Scheme of the system used to introduce the evaporated TM ions in the ACaPI ion source. b) Internal energy distributions of different discharge gases (He, synthetic air (SA) and ambient air (AA)) used to help the transport of headspace-analyzed TM ions compared to the vaporized TM ions internal energy distribution (dark wine trace). c) FμTP for He, Ar and SA as discharge gases.
The built setup for the ACaPI analysis consists of a 20 mL vial with a two-inlet cap placed on a heater maintained at 60ºC. Five milliliters of the 100 µM solution was introduced in the vial. When a dragging gas was used (He and SA) a constant flow of the gas was supplied to drag the evaporated TM ions and maintain the discharge. On the other hand, the discharge gas inlet was open when the ACaPI was operated with ambient air. The atmospheric air, as well as the TM ions, were introduced in the discharge region helped by the mass spectrometer.
The miniaturized dimensions of FμTP enabled the operation with different discharge gases within the maximum operation range of the square-wave voltage supply used, 3.5 kV. The operation voltages were 1.5 kV for He The on-axis distances were controlled using an x, y, z stage. Both, DBDI and LTP, did not show survival yield changes when the distances were shorter than 15 mm.
The small plasma volume of FμTP required from shorter distances (<12 mm) to promote noteworthy abundances (>1000 a. u.). However, the increased intensities also promoted a higher degree of fragmentation. In addition, when the probe was positioned at 8 mm with respect to the mass spectrometer inlet, we observed the decrease of the TM ions signals; the probe was partially blocking the vaporized molecules path towards the plasma-jet. S-9 Figure S8. Effects of the distance of DBDI orthogonal plasma-jet with respect to the mass spectrometer inlet. S-11 Table S2. Parameters obtained for the sigmoidal fitting of the five evaluated TM ions survival yields in Figure 3. Slope value in the table corresponds to the slope of the sigmoid curve at survival yield 0.5, calculated as .

4
The sigmoidal fit was done using the following equation: being x 0 is the fitted modal average internal energy of the sigmoidal fit and dx the time constant.