A Radical-Mediated Pathway for the Formation of [M + H]+ in Dielectric Barrier Discharge Ionization
- 783 Downloads
Active capillary plasma ionization is a highly efficient ambient ionization method. Its general principle of ion formation is closely related to atmospheric pressure chemical ionization (APCI). The method is based on dielectric barrier discharge ionization (DBDI), and can be constructed in the form of a direct flow-through interface to a mass spectrometer. Protonated species ([M + H]+) are predominantly formed, although in some cases radical cations are also observed. We investigated the underlying ionization mechanisms and reaction pathways for the formation of protonated analyte ([M + H]+). We found that ionization occurs in the presence and in the absence of water vapor. Therefore, the mechanism cannot exclusively rely on hydronium clusters, as generally accepted for APCI. Based on isotope labeling experiments, protons were shown to originate from various solvents (other than water) and, to a minor extent, from gaseous impurities and/or self-protonation. By using CO2 instead of air or N2 as plasma gas, additional species like [M + OH]+ and [M − H]+ were observed. These gas-phase reaction products of CO2 with the analyte (tertiary amines) indicate the presence of a radical-mediated ionization pathway, which proceeds by direct reaction of the ionized plasma gas with the analyte. The proposed reaction pathway is supported with density functional theory (DFT) calculations. These findings add a new ionization pathway leading to the protonated species to those currently known for APCI.
KeywordsDBDI Ionization mechanism Active capillary plasma ionization APCI Radical mediated protonation Hydronium clusters
This mechanism is well accepted for APPI. In APCI, it is believed that only nonpolar compounds, such as polycyclic aromatic hydrocarbons (PAH) or benzene, form radical cations via this mechanism [8, 9]. For the formation of protonated polar molecules, the hydronium cluster model is used exclusively in the literature for both plasma-based or APCI-like ambient ionization techniques . One reason for this is the solvent dependence of the APCI process, which was described by a couple of research groups [14, 15]. Because the ionization efficiency was found to depend on the solvent, the direct ionization pathway was disregarded in mechanistic investigations, since a solvent-independent behavior would have been expected . However, this neglects the possibility that both mechanisms contribute (proton transfer and/or direct ionization followed by a hydrogen atom transfer). After decades of experience with APCI and the development of a vast number of new APCI-like ionization methods , there is still insufficient experimental evidence regarding the actual protonation pathways and their contribution to the formation of [M + H]+. An improved understanding of the ionization mechanism would therefore allow the enhancement of APCI-like ambient ionization methods for the detection of specific compounds and for increased sensitivity.
Active capillary plasma ionization  is a novel dielectric barrier discharge-based ambient ionization method, where the source itself serves as the interface to the mass spectrometer. In addition to the extraordinary sensitivity , this design greatly facilitates mechanistic investigations, since the whole sample passes through the plasma and the reaction time of the analyte and the gas is well defined. In order to study the underlying APCI pathways, we investigated the ionization efficiency of various substances (homologous series of tertiary alkyl amines, phosphonates) in different deuterated solvents, and in a variety of atmospheres (air, N2, CO2).
Chemicals and Solvents
Diethyl ethylphosphonate (DEEP) (98%), methanol-d (99%) triethylamine (TEA) (>99.5%), tripropylamine (TPrA) (98%), and tributylamine (TBuA) (99%) were obtained from Sigma-Aldrich Chemie GmbH (Buchs, Switzerland). Tripentylamine (TPeA) (99%) and D2O (99.9%) were purchased from Acros Organics (Geel, Belgium); trihexylamine (THA) (99%) from Alfa Aesar GmbH and Co. KG (Karlsruhe, Germany); ethanol-d (99%), benzene-d6 (99.5%), and chloroform-d (99.8%) from Cambridge Isotope Laboratories Inc. (Tewksbury, MA, USA). All chemicals were used without further purification. Individual and combined stock solutions (2000 μg/mL each) were prepared in the corresponding solvents. Depending on the experiment, diluted samples were generated from these stock solutions with varying concentrations.
The mass spectra were acquired using Thermo LTQ Orbitrap and LCQ DECA XP mass spectrometers (both from Thermo Scientific, San Jose, CA, USA) fitted with an active capillary plasma ionization source. All spectra were recorded in positive polarity.
Active Capillary Plasma Ionization Source
The principle of the active capillary plasma source has already been described in detail in previous reports [18, 19]. Briefly, a quartz glass capillary (i.d. 0.7 mm, o.d. 1.0 mm) was connected to the inlet of the mass spectrometer. The constant underpressure in the instrument ensured a fixed flow rate of 1.5 L/min through the capillary. A stainless steel capillary (i.d. 0.5 mm, o.d. 0.6 mm) inserted into the glass capillary served as first electrode. The counter-electrode was a 5-mm-long copper ring (i.d. 1.0 mm) surrounding the capillary. By applying a sine modulated (5750 Hz) high voltage (1.5–2.5 kV, peak to peak) to the electrodes, the plasma was ignited inside the capillary by a dielectric barrier discharge, which then ionizes the passing air and sample molecules. All this was enclosed in a gas-tight housing, enabling a direct, leak-tight connection to the plasma gas stream.
The gas-phase sample generation system was similar to the one used in previous studies . Briefly, a pressurized sample reservoir was connected via a fused silica capillary (ID 40 μm) to a hollow heating cartridge held at 200 °C. The sample outflow was evaporated within the cartridge by temperature and a preheated 3 L/min plasma gas stream, controlled by a mass flow controller. The outlet of the heating cartridge was connected by a T-piece to the active capillary plasma ionization source. The other (excess) outlet is directed to the exhaust. The whole sample generation system is leak-tight and can be fed with different plasma gas streams and sample solutions.
Gas-Phase Sample Delivery
Dry air and dry N2 plasma gas was prepared by additionally drying the supply gases (purity 4.5) over columns filled with phosphorous pentoxide (Reagent plus 99%, Sigma-Aldrich Chemie GmbH, Buchs, Switzerland) followed by an active charcoal filter. Humidification was done by passing the plasma gas streams through a glass fritted bubbler filled with H2O or D2O, respectively. Two mass flow controllers (Bronkhorst High Tech B.V., Ruurlo, The Netherlands) ensured constant flow rates and mixing ratios. Relative humidity was measured using a FHA 646R capacitive humidity sensor and ALMEMO 2590A readout (Ahborn GmbH, Holzkirchen, Germany).
For quantum mechanical calculations, full geometry optimization with no constraints, and frequency calculation for reactants, intermediates, and products were performed with density functional theory (DFT) using the hybrid B3LYP functional as implemented in the GAUSSIAN09  code. For all atoms, the 6-311++G(3df,3pd) basis set was used. For integration, an “ultrafine” grid corresponding to 99 radial shells and 590 angular points per shell was used for increased accuracy, with “tight” optimization criteria. In addition, calculation of the intrinsic reaction coordinates (IRC) for each reaction pathway was performed to show the connection between reactants, products, and the corresponding transition states along the imaginary mode of vibration. All energy values reported in this work are free energies at a temperature of 298 K. For all reactions, the sum of the energies of the separated reactants was set to zero as common reference point.
Results and Discussion
In order to elucidate the ionization mechanism leading to [M + H]+ in plasma-based or APCI-like ambient ionization techniques, the active capillary plasma ionization source  was used. It features a fully enclosed ionization volume and precisely controllable gas-phase conditions. The plasma gas, the solvent, and the humidity are the three dominant chemical factors in the plasma that influence the ionization of the analyte in normal operation of the active capillary plasma ionization source. The impact of each of these factors on the [M + H]+ ion formation was studied in this work.
The activation energy (ΔG‡) and the reaction free energy (ΔGR) were calculated with DFT (B3LYP/6-311++G(3df,3pd)) for possible reactions in a CO2 atmosphere with triethylamine (black), tributylamine (red), and trihexylamine (blue). All values are given in kcal/mol, with the sum of the energies of the separated reactants set to zero as common reference point
Regardless whether the direct reduction or the water elimination prevails, the formation of all these ions observed and discussed above requires an initial reaction of the ionized plasma gas with the analyte. This is in contradiction to the current APCI model, where the pathway to the protonated molecules always involves reaction of the plasma gas with H2O.
Despite this and the occurrence of less favored reaction products with CO2 +•, no M+• species are observed in the spectra, leaving only one explanation: M+• ions are formed but rapidly react further to form [M + H]+, although, according to the calculations, the final reaction (Equation 16) is endothermic for water as hydrogen source. However, one has to consider the clustering processes or other hydrogen sources (e.g., organic solvent, impurities, and other analyte molecules).
The [M + H]+, [M + OH]+, and [M – H]+ ions were also observed under humid conditions, as shown by a repetition of the experiment in a D2O enriched atmosphere (80% relative humidity, D2O, Figure 1 bottom). Most, (around 90%) but not all, protons were exchanged by deuterium and therefore originate from solvent molecules, which supports reactions 9 and 16 (endothermic according to our calculations). This is confirmed by high-resolution measurements, as shown in detail for tributylamine (Figure 1, zoomed view). The signals at m/z 184.2083 must correspond to [M – H]+, m/z 185.2145 to [M – 2H + D]+, m/z 186.2240 to the protonated molecule [M + H]+, m/z 187.2302 to the deuterated molecule [M + D]+, m/z 202.2188 to the oxidized and protonated molecule [M + OH]+, and m/z 203.2253 to the oxidized and deuterated molecule [M + OD]+. The slightly lower signals for [M – H]+ and [M + OH]+ in D2O saturated CO2 atmosphere suggested that the concomitant contribution of protonation by H3O+ clusters is increasing, as expected, but still not exclusive.
The intensity ratio (%) of the deuterated molecules [M + D]+ to the sum of protonated and deuterated molecules ([M + H]+ + [M + D]+) of tributylamine vaporized from different deuterated solvents under dry (<100 ppm H2O) and humid (4% relative humidity, at room temperature) air atmosphere was calcultated. The numbers were corrected by the natural abundance of 13C and resolution of MS
Intensity ratio ([M + D]+/ ([M + H]+ + [M + D]+)) in different solvents
61.1% ± 1.6%
69.6% ± 1.7%
69.7% ± 3.6%
26.8% ± 8.6%
28.9% ± 4.5%
6.5% ± 2.6%
7.5% ± 1.7%
5.3% ± 2.3%
1.1% ± 3.8%
3.7% ± 2.5%
Our results provide the following evidence for a significant contribution of a radical-driven ionization mechanism for the formation of [M + H]+ ion in plasma-based, APCI-like ambient ionization methods: (1) [M + OH]+ and [M – H]+ ions are formed when using CO2 as plasma gas through a radical-driven reaction with CO2 +•, (2) the activation energy and the reaction free energy of different radical and non-radical reactions, as calculated by DFT, support our findings, (3) deuterium atoms incorporated in [M + D]+ can originate from aprotic solvents such as chloroform-d and benzene-d6, (4) the [M + H]+ ion is also formed under dry conditions (no H3O+ clusters present), and (5) an in-source fragmentation of diethyl ethylphosphonate was quenched by humidity. This does not contradict, but complements the currently accepted ionization mechanism by hydronium clusters in APCI.
Although our measurements were performed with the active capillary plasma ionization source, these results can be generalized for all plasma-based atmospheric pressure ionization methods (APCI, APPI, DART, FAPA, LTP, etc.), since similar reactive species (i.e., N2 +•) are formed in all these methods. Confirmed in various studies by means of spectroscopy, they cause the “glowing” purple color of the plasma, visible by eye. Given the same reactive species as starting point, a similar chemical mechanism can be assumed for all these methods, leading to a reasonable generalization of our findings. Additionally, these reactive species are a common starting point for both pathways to the protonated analyte ([M + H]+): the currently accepted APCI hydronium cluster mechanism and the radical cation pathway revealed in this study. The latter starts with the initial ionization of gas, solvent, and analyte attributable to electron ionization. Subsequently, the radical cations formed react with other neutral molecules or undergo fragmentation to get rid of their excess energy. Depending on the stability of the intermediates, this can lead to in-source fragmentation, in case of insufficient protic collision partners. However, most likely, the intermediates undergo stabilization by hydrogen atom abstraction from other reaction partners such as solvents, leaving mainly a protonated molecule. Depending on the experimental conditions, this pathway may be even more important than proton transfer by hydronium clusters, marking the difference to spray-based ambient ionization methods, which lack this “radical” pathway. Since the contribution of both mechanisms to the final ion yield may vary depending on the solvent content, plasma gas, discharge type, temperature, and chemical properties of each individual compound within the plasma plume, a prediction of the final ionization efficiency for an individual analyte remains challenging. This work adds an additional pathway to the APCI mechanism for the formation of [M + H]+, further elucidating positive ion formation and opening up further possibilities in the prediction of ionization efficiencies.
The authors thank Christoph Bärtschi and Christian Marro of the ETH mechanical shop for manufacturing the active capillary plasma ionization source and various other relevant parts. They gratefully acknowledge Dr. Juan Zhang (Novartis AG) for the donation of the LTQ Orbitrap instrument used in this study. This work was funded by the Federal Office for Civil Protection FOCP, SPIEZ LABORATORY (grant 353004332/Stm) that is gratefully acknowledged.
- 3.Horvatic, V., Vadla, C., Franzke, J.: Spectrochimica Acta Part B Discussion of fundamental processes in dielectric barrier discharges used for soft ionization. Spectrochim. Acta B 100, 52–61 (2014)Google Scholar
- 7.Klee, S., Derpmann, V., Wißdorf, W., Klopotowski, S., Kersten, H., Brockmann, K.J., Benter, T., Albrecht, S., Bruins, A.P., Dousty, F., Kauppila, T.J., Kostiainen, R., O’Brien, R., Robb, D.B., Syage, J.A.: Are clusters important in understanding the mechanisms in atmospheric pressure ionization? Part 1: reagent ion generation and chemical control of ion populations. J. Am. Soc. Mass Spectrom. 25, 1310–1321 (2014)CrossRefGoogle Scholar
- 19.Wolf, J.-C., Schaer, M., Siegenthaler, P., Zenobi, R.: Direct quantification of chemical warfare agents and related compounds at low ppt levels: comparing active capillary dielectric barrier discharge plasma ionization and secondary electrospray ionization mass spectrometry. Anal. Chem. 87, 723–729 (2015)CrossRefGoogle Scholar
- 21.Gaussian 09, Revision D.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, and D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.Google Scholar
- 22.Franklin J.L., Dillard J.G., Rosenstock H.M., Herron J.T., D.K.: Ionization potentials, appearance potentials, and heats of formation of gaseous positive ions. National Bureau of Standards, Washington, D.C. (1969)Google Scholar
- 23.Chan, G.C.Y., Shelley, J.T., Wiley, J.S., Engelhard, C., Jackson, A.U., Cooks, R.G., Hieftje, G.M.: Elucidation of reaction mechanisms responsible for afterglow and reagent-ion formation in the low-temperature plasma probe ambient ionization source. Anal. Chem. 83, 3675–3686 (2011)CrossRefGoogle Scholar