On-Line Investigation of the Generation of Nonaqueous Intermediate

Department of Chemistry, University of Florida, Gainesville, Florida 32611-7200 ... on-line with mass spectrometry (MS) through a particle beam (PB) i...
32 downloads 0 Views 106KB Size
Anal. Chem. 2000, 72, 2533-2540

On-Line Investigation of the Generation of Nonaqueous Intermediate Radical Cations by Electrochemistry/Mass Spectrometry Tianyi Zhang and Anna Brajter-Toth*

Department of Chemistry, University of Florida, Gainesville, Florida 32611-7200

Anodic oxidation of triphenylamine (TPA) in acetonitrile was investigated by electrochemistry (EC) combined online with mass spectrometry (MS) through a particle beam (PB) interface, EC/PB/MS. Electrooxidation of TPA generates a TPA•+ radical cation (m/z 245) which dimerizes to tetraphenylbenzidine (TPB, MW 488). TPB is readily oxidized to TPB•+ (m/z 488) and TPB2+ (m/z 244) at the oxidation potential of TPA. In EC/PB/MS, direct monitoring of the oxidation of TPB to TPB•+ radical cation as a function of the electrode potential was achieved via selective ion monitoring of the ion peak at m/z 488. By using the relative intensity ratio of ions at m/z 244 (TPB2+) to 245 (TPA•+), the formation of TPB2+ as a function of the electrode potential was also monitored. EC/PB/MS showed a maximum rate of formation of TPB•+ at +1.2 V vs Pd, while TPB2+ is generated at a maximum rate at +1.6 V vs Pd. The effect of spectral interference from the electron impact ionization of TPA, on EC/PB/ MS results, is also discussed. Finally, a significant signal enhancement is observed in the presence of tetrabutylammonium perchlorate (TBAP) and is reported for the first time. Compatibility of coupling of EC with MS via PB interface for EC/MS studies in nonaqueous solvents is demonstrated. The observation of significant signal enhancement in the presence of TBAP may facilitate other applications of LC/MS. Electrochemistry coupled on-line with mass spectrometry (EC/ MS) has attracted much attention from chemists.1-21 By using MS as a detector, structural information about reactants, short* Corresponding author: (phone) 352-392-7972; (fax) 352-392-4651; (e-mail) [email protected]. (1) Hambitzer, G.; Heitbaum, J. Anal. Chem. 1986, 58, 1067-1070. (2) Hambitzer, G.; Heitbaum, J.; Stassen, I. Anal. Chem. 1998, 70, 838-842. (3) Hambitzer, G.; Heitbaum, J.; Stassen, I. J. Electroanal. Chem. 1998, 447, 117-124. (4) Volk, K. J.; Lee, M. S.; Yost, R. A.; Brajter-Toth, A. Anal. Chem. 1988, 60, 720-724. (5) Volk, K. J.; Yost, R. A.; Brajter-Toth, A. Anal. Chem. 1989, 61, 1709-1717. (6) Volk, K. J.; Yost, R. A.; Brajter-Toth, A. J. Chromatogr. 1989, 474, 231243. (7) Volk, K. J.; Yost, R. A.; Brajter-Toth, A. J. Electrochem. Soc. 1990, 137, 1764. (8) Volk, K. J.; Yost, R. A.; Brajter-Toth, A. Anal. Chem. 1992, 64, 21A-33A. (9) Regino, M. C. S.; Brajter-Toth, A. Anal. Chem. 1997, 69, 5067-5072. (10) Regino, M.; Weston, C.; Brajter-Toth, A. Anal. Chim. Acta 1998, 369, 253262. (11) Regino, M. C. S.; Brajter-Toth, A. Electroanalysis 1999, 11, 374-379 10.1021/ac9912348 CCC: $19.00 Published on Web 04/26/2000

© 2000 American Chemical Society

lived intermediates, and products can be derived during the electrochemical reaction.8 This offers the possibility to monitor and identify the intermediates and products as a function of the electrode potential, in close to real time, which can help clarify pathways of redox reactions.7 For example, on-line combination of electrochemistry with thermospray mass spectrometry (EC/ TSP/MS) has been used to gain insights into biological redox reactions of purines.5 As a detector, MS is universal and sensitive because all compounds, upon ionization, will show a signal at the electron multiplier.22 Selectivity is achieved by monitoring characteristic ions of a species or by using tandem MS/MS methods.4-8 Thus, it is possible to perform an EC/MS experiment without a prior separation of analytes, intermediates, and products. In EC/MS experiments, electrochemical reactions of analytes usually take place in a flow cell. To generate a sufficient abundance of ions for MS detection, high cell conversion efficiency is critical. Much attention has been paid to cell design, aimed at improving cell conversion efficiency and reducing the delay time between the electrochemical reaction, in an EC flow cell, and the subsequent MS detection by on-line LC/MS.2,3,9 Since electrochemistry is usually performed in liquid solutions while MS is operated under vacuum, an appropriate LC/MS interface is needed for on-line coupling of EC with MS. EC/MS studies have been carried out with thermospray1-8 and electrospray (ES) interfaces.13-21 The use of a particle beam (PB) interface in EC/MS has also been reported.9-11 Despite many recent studies, most EC/MS work has been focused on aqueous solution-based reactions. In the studies with the TSP interface,1-8 ammonium acetate has been used as both an electrolyte for the electrochemical reaction and an ionizing (12) Bittins-Cattaneo, B.; Cattaneo, E.; Konigshoven, P.; Vielstich, W. In Electroanalytical Chemistry; Bard, A. J., Ed.; Marcel Dekker: New York, 1991; Vol. 17, pp 181-220. (13) Xu, X.; Nolan; S. P.; Cole, R. B. Anal. Chem. 1994, 66, 119-125. (14) Xu, X.; Lu, W.; Cole, R. B. Anal. Chem. 1996, 68, 4244-4253. (15) Lu, W.; Xu, X.; Cole, R. B. Anal. Chem. 1997, 69, 2478-2484. (16) Van Berkel, G. J.; McLuckey, S. A.; Glish, G. L. Anal. Chem. 1992, 64, 1586-1593. (17) Van Berkel, G. J.; Zhou, F. Anal. Chem. 1995, 67, 2916-2923. (18) Van Berkel, G. J.; Zhou, F. Anal. Chem. 1995, 67, 3958-3964. (19) Zhou, F.; Van Berkel, G. J. Anal. Chem. 1995, 67, 3643-3649. (20) Bond, A. M.; Colton, R.; Da´gnostino, A.; Downnard, A. J.; Traeger, J. C. Anal. Chem. 1995, 67, 1691-1695. (21) Tang, L.; Kebarle, P. Anal. Chem. 1993, 65, 3654-3668. (22) Niessen, W. M. A. Liquid Chromatography-Mass Spectrometry, 2nd ed.; Marcel Dekker: New York, 1999.

Analytical Chemistry, Vol. 72, No. 11, June 1, 2000 2533

reagent in mass spectrometry to facilitate proton transfer and adduct ion formation during the thermospray ionization. EC/MS studies in nonaqueous solvents have not received much attention. Some work with organic solvents such as acetonitrile and methylene chloride has been reported with an electrospray interface, in the study of electrochemistry of metallocenes,13 alkyl-substituted metalloporphyrins, and polycyclic aromatic hydrocarbons.16-18 In these studies, ES served as an electrolytic cell, where the redox reactions took place at the spray needle of the ES probe. To improve the performance of the ES interface in detection of neutral compounds, coupling of an additional electrochemical cell on-line with the ES/MS has been examined.14,15,19,20 However, high concentrations of nonvolatile supporting electrolytes, such as tetrabutylammonium hexafluorophosphate (Bu4NPF6) and tetrabutylammonium perchlorate (TBAP), which were used to increase solution conductivity and to reduce error in cell voltage measurements, were shown to cause problems which resulted in a loss of sensitivity. The loss of sensitivity was attributed to signal suppression by nonvolatile electrolyte, and to plugging and contamination of the ES interface.14,15,17-21 It has been reported, for example, that addition of 5 × 10-4 M TBAP to the analyte solution consisting of 10-4 M 9,10-dimethylanthracene (DMA) in methylene chloride resulted in an almost complete suppression of the analyte signal of the radical cation, M•+.14 To minimize the magnitude of the signal suppression, an uncommon electrolyte, i.e., one incorporating a small metal ion and a relatively volatile anion (e.g., lithium trifluoromethane sulfonate or lithium triflate), was used.14,15,17-19 The studies and applications of EC/MS in nonaqueous solvents were hence retarded by problems arising from the use of electrolytes. The PB interface is known to support efficient volatilization of nonaqueous solvents.22 Significant analyte enrichment can be achieved in the desolvation chamber via rapid evaporation of volatile solvent.22 PB also has the advantage of being coupled to an electron impact (EI) mass spectrometric ion source, which allows acquisition of EI mass spectra that may be library searched and readily interpreted.22-31 In addition, using an EI source in an EC/PB/MS experiment permits ionization and identification of neutral intermediates and products generated from the electrochemical reactions and follow-up chemical reactions in solution. A disadvantage of PB is its poor sensitivity, which results from the relatively low mass transport efficiency through the PB momentum separator.22-31 Since many organic compounds are only soluble in pure organic solvents, where many important redox reactions occur, the performance of EC/PB/MS in nonaqueous solvents is of interest. The intermediate polarity, volatility, and low viscosity (23) Incorvia Mattina, M. J. J. Chromatogr. 1991, 549, 237-245. (24) Incorvia Mattina, M. J. J. Chromatogr. 1991, 542, 385-395. (25) Bellar, T. A.; Behymer, T. D.; Budde, W. L. J. Am. Soc. Mass Spectrom. 1990, 1, 92-98. (26) Ho, J. S.; Behymer, T. D.; Budde, W. L.; Bellar, T. A. J. Am. Soc. Mass Spectrom. 1992, 3, 662-671. (27) Apffel, A.; Perry, M. L. J. Chromatogr. 1991, 554, 103-118. (28) Bellar, T. A.; Budde, W. L.; Kryak, D. D. J. Am. Soc. Mass Spectrom. 1994, 5, 908-912. (29) Wilkes, J. G.; Zarrin, F.; Lay, J. O., Jr.; Vestal, M. L. Rapid Commun. Mass Spectrom. 1995, 9, 133-137. (30) Wilkes, J. G.; Freeman, J. P.; Heinze, T. M.; Lay, Jr., J. O.; Vestal, M. L. Rapid Commun. Mass Spectrom. 1995, 9, 138-142. (31) Li, Y.; Koropchak, J. A. Instrum. Sci. Technol. 1998, 26, 389-407.

2534

Analytical Chemistry, Vol. 72, No. 11, June 1, 2000

(0.345 cP at 25 °C) of acetonitrile makes it a good solvent for EC/PB/MS. Acetonitrile is widely used as a solvent in organic electrochemistry owing to its high dielectric constant (38 at 25 °C), wide accessible potential range (-3.5 to +2.4 V vs SCE), and convenient liquid range (-42 to +82 °C).14,32 EC/MS requires the presence of an electrolyte to support current flow in the EC flow cell. Quaternary ammonium salts are frequently used as electrolytes in organic electrochemistry due to their high solubility and electrochemical inertness in organic solvents. An alkyl group with at least two carbon atoms can be used, but higher homologues (e.g., butyl and propyl) are often preferred for their wider potential range.32 It has also been shown that Bu4N+ can be converted more efficiently from solution to gas phase than Et4N+.20,21 The corresponding anions, from the point of view of solubility, are usually limited to halides, nitrates, perchlorates, tetrafluoroborates, and hexafluoroantimonates.32 The size of the negative counterions does not have a significant effect on the interionic separation.33 Therefore, TBAP is a reasonable electrolyte for EC/PB/MS in nonaqueous solvents. In this paper, we demonstrate the use of EC/MS with a PB interface (EC/PB/MS) in the investigation of the anodic oxidation of triphenylamine (TPA) in acetonitrile, with TBAP as the supporting electrolyte. The electrochemical cell used in the experiment has been described before.9 TPA was selected as a test analyte because the on-line EC/MS detection of radical cations generated in the oxidation of TPA was of interest. In addition, TPA was selected as a test analyte because of its well-known electrochemical behavior and the generation of relatively stable radical cations during electrooxidation in acetonitrile.34-40 The electrochemical oxidation of TPA generates a cation radical TPA•+ in a one-electron-transfer step. The radicals undergo a follow-up dimerization reaction to generate tetraphenylbenzidine (TPB), which is more readily oxidized than TPA at the oxidation potential of TPA. Oxidation of TPB occurs in two discrete one-electron steps.34-40 The oxidation pathway of TPA can be shown schematically (Scheme 1). Direct monitoring of the ion abundance of the radical cation TPB•+ at m/z 488 as a function of the electrode potential was achieved by on-line EC/PB/MS. A maximum ion abundance of TPB•+ was observed at +1.2 V vs Pd. The effect of the electrode potential on the rate of generation of the dication TPB2+ was additionally investigated by EC/MS. A maximum rate of TPB2+ formation was detected at +1.6 V vs Pd. The significance of the MS results vis a` vis the electrooxidation pathway is discussed. In addition, an important new observation of a significant signal enhancement in PB/MS in the presence of electrolyte, TBAP, in solution, which can play a role in facilitating MS detection of intermediate radical cations by EC/PB/MS, is described. (32) Janz, G. J.; Tomkins, R. P. T. Nonaqueous Electrolytes Handbook; Academic Press: New York, 1972; Vol. 1. (33) Bennett, J.; Gillen, G. J. Am. Soc. Mass Spectrom. 1993, 4, 930-937. (34) Seo, E. T.; Nelson, R. F.; Fritsch, J. M.; Marcoux, L. S.; Leedy, D. W.; Adams, R. N. J. Am. Chem. Soc. 1966, 88, 3498-3503. (35) Nelson, R. F.; Adams, R. N. J. Am. Chem. Soc. 1968, 90, 3925-3930. (36) Marcoux, L. S.; Adams, R. N.; Feldberg, S. W. J. Phys. Chem. 1969, 73, 2611-2614. (37) Nelson, R. F.; Feldberg, S. W. J. Phys. Chem. 1969, 73, 2623-2626. (38) Nelson, R. F.; Philp, Jr., R. H. J. Phys. Chem. 1979, 83, 713-716. (39) Oyama, M.; Nozaki, K.; Okazaki, S. Anal. Chem. 1991, 63, 1387-1392. (40) Sumiyoshi, T. Chem. Lett. 1995, 645-646.

Scheme 1. Anodic Oxidation Pathway of Triphenylamine (TPA)34

EXPERIMENTAL SECTION Chemicals. TBAP was purchased from Alfa Aesar (Ward Hill, MA). TPA (98%) was from Aldrich (St. Louis, MO), and the HPLCgrade acetonitrile was from Fisher (Pittsburgh, PA). Acetonitrile was degassed in an ultrasonic bath under reduced pressure for use as a mobile phase in flow injection analysis (FIA) and EC/ MS. All chemicals were used as received. Acetonitrile with 0.1 M TBAP was used as a solvent and electrolyte, respectively, in cyclic voltammetry and, additionally, as a mobile phase in off-line FIA with electrochemical detection. In EC/PB/MS, the composition of the mobile phase was modified and the concentration of TBAP was reduced to 0.01 M. This reduced the contamination of the PB desolvation chamber and the EI ion source. An aqueous solution of 1 mM K3Fe(CN)6 in 0.1 M KCl was used in the determination of the working electrode area by chronocoulometry.41 All solutions were freshly prepared prior to analysis. Apparatus. A Bioanalytical Systems electrochemical analyzer (BAS-100, Bioanalytical Systems, Inc., West Lafayette, IN) was used for cyclic voltammetry and electrochemical detection in offline FIA. A model 173 PAR potentiostat (EG&G Princeton Applied Research, Princeton, NJ) was used to apply a constant potential to the flow-through electrochemical cell in EC/PB/MS experiments. A Hewlett-Packard 5989A quadrupole mass spectrometer with a Hewlett-Packard 59980B particle beam interface (Palo Alto, CA) was used in EC/PB/MS. A Hewlett-Packard 1050 Series HPLC pump with a 20-µL injector loop was used to deliver the mobile phase in off-line FIA and EC/PB/MS experiments. Electrochemical Cell. The description and characterization of the cell has been detailed before.9 The home-built cell contains three electrodes, i.e., a rough pyrolytic graphite (RPG) working electrode (3 × 3 mm), a Pd wire (1-mm diameter) counter and reference electrodes. Before use, the RPG working electrode was polished on a 600-grit silicon carbide paper (Mark V Laboratory, East Granby, CT) using a metallographic polishing wheel (Ecomet I; Buehler, Evanston, IL), followed by a 15-min ultrasonication (41) Bard, A. J.; Faulkner, L. R. Electrochemical Methods; Wiley: New York, 1980.

in a water bath. The area of the RPG working electrode was determined to be ∼0.093 cm2 by chronocoulometry41 with a potential step from +0.40 to 0.00 V vs SCE in 1 mM K3Fe(CN)6/ 0.1 M KCl aqueous solution. In on-line EC/PB/MS experiments, the electrochemical flow cell was positioned between the HPLC pump and the ion source of the mass spectrometer. The volume of the electrode compartment of the cell is ∼4.4 µL. With the flow channel (∼51.9 µL) and the capillary connective tubing (∼10.1 µL), the total dead volume introduced by the cell on-line with PB/ MS is ∼66.4 µL. Cyclic Voltammetry (CV). CV of TPA was performed with a BAS-100 electrochemical analyzer. The solution was 0.001 M TPA/ 0.1 M TBAP in acetonitrile. The electrodes and the pretreatment of the RPG working electrode were the same as in EC/MS. The Pd wire reference and counter electrodes were polished on a felt cloth (Mark V Laboratory, East Granby, CT) with a γ-alumina suspension of 1-µm particle size (Gamal, Fisher Scientific, Pittsburgh, PA). The potential was scanned from +0.4 to +1.2 V vs Pd at a scan rate of 0.1 V/s. In cyclic voltammetry, two scan cycles were recorded. Off-Line Flow Injection Analysis. In off-line FIA, the mobile phase of 0.1 M TBAP in acetonitrile was pumped continuously through the electrochemical cell.10 The analyte solution of 100 µM TPA and 0.1 or 0.01 M TBAP in acetonitrile, was injected into the flow stream through a 20-µL sample injection loop. The thin-layer bulk electrolysis (TLBE) method available directly from the BAS-100 electrochemical analyzer was used to apply voltage to the cell and monitor charge with time. To reduce the effect of the background charging current on the measured charge of TPA oxidation, the TPA sample injection was made 30 s after cell voltage was applied.10 An injection of a blank of 0.1 M TBAP in acetonitrile was made before each analyte injection, at each applied voltage.10 The blank-subtracted charge was used to determine cell conversion efficiency.9,10 With the cell voltage controlled at +1.2 V vs Pd, the total number of moles of oxidized TPA was determined from the net charge increase following the injection of TPA.9 The cell conversion efficiency was simply defined as the Analytical Chemistry, Vol. 72, No. 11, June 1, 2000

2535

ratio of the total number moles of analyte consumed in the oxidation to the total number of moles of analyte injected.9 The current-time (i-t) response curve of the cell in FIA was obtained by taking a derivative of the charge-time curve obtained by the BAS 100 during the oxidation. In the calculation, i ) δQ/ δt, where δt ) 1 s and δQ is simply the difference between two successive Q values at 1-s time intervals. The average of two successive time values was used to construct the time axis of the i-t curve. Hydrodynamic voltammograms were obtained by measuring blank-subtracted current values, obtained at different voltages from the i-t curves, and plotting the current as a function of the electrode potential applied to the electrochemical cell. PB and MS Conditions. The Hewlett-Packard 59980B PB interface contains a concentric, pneumatic nebulizer. The nebulizing gas was high-purity helium, and its pressure was controlled at 50 psi. The PB desolvation chamber temperature was kept at 60 °C. The two-stage momentum separator rough pumps of the PB interface reduced the pressure to ∼300 Torr at the first stage and to ∼0.2 Torr at the second stage. The mobile phase, pure acetonitrile, was pumped continuously through the PB interface at a flow rate of 0.4 mL/min. The differentially pumped mass spectrometer diffusion pumps maintained a pressure of about (2-3) × 10-5 Torr in the manifold surrounding the conventional EI ion source. The ion source temperature was kept at 250 °C, and the electron energy was 70 eV. The ion source and lens potentials were tuned automatically to optimize the ion intensity of perfluorotributylamine (PFTBA) in the range of 50-550 amu. Both scan and selective ion monitoring (SIM) modes were used. In the scan mode, the mass spectrometer was repetitively scanned from 50 to 550 amu at 0.8 scan/s. In the SIM mode, multiple ions were monitored at m/z 488, 245, and 244. A fixed integration time of 235 µs is used in the SIM data acquisition mode, and the dwell time for sampling of a specific ion was 100 ms. All mass spectra were acquired by EI in the positive ion mode. On-Line EC/PB/MS. The electrochemical flow cell was positioned between the HPLC pump and the PB interface. Mass spectra of TPA, TBAP, and the mixture of TPA/TBAP were first acquired in a scan mode, without applying voltage to the electrochemical flow cell. To initiate the electrooxidation of TPA, a constant voltage was applied to the flow cell with a model 173 PAR potentiostat. When a solution consisting of 100 µM TPA with 1 mM TBAP in acetonitrile was tested, no ion peak at m/z 488 (of TPB•+) was detected in either scan or SIM mode. By injecting a solution (20 µL) consisting of 1 mM TPA with 0.01 M TBAP in acetonitrile, the ion peak at m/z 488 was detected. Background subtraction of each ion chromatogram was performed with a blank mass spectrum, obtained just prior to the foot of the first peak in the ion chromatogram. The integrated ion abundance of each ion was measured from the background-subtracted extracted-ion chromatogram as an integrated peak area. RESULTS AND DISCUSSION Cyclic Voltammetry of TPA. The cyclic voltammogram of TPA obtained in bulk solution is shown in Figure 1. During the first anodic scan, only one oxidation peak at +0.883 V vs Pd is observed, which is attributed to the oxidation of TPA. Two reduction peaks, at +0.750 and +0.634 V, respectively, are observed in the following cathodic scan. During the second scan cycle, two oxidation peaks and two reduction peaks are observed. 2536 Analytical Chemistry, Vol. 72, No. 11, June 1, 2000

Figure 1. Cyclic voltammetry of TPA in acetonitrile. Conditions: 1 mM TPA/0.1 M TBAP, RPG (0.093 cm2) working electrode, Pd reference and counter electrodes, and scan rate 100 mV/s.

Figure 2. Electrochemical detection in off-line FIA: (a) 0.1 M TBAP; (b) 100 µM TPA/0.1 M TBAP in acetonitrile. Conditions: injection volume 20 µL, mobile phase 0.1 M TBAP/acetonitrile, flow rate 0.4 mL/min, voltage applied +1.2 V vs Pd, and injections 30 s after cell voltage was applied.

The cyclic voltammetry results are consistent with the reported ECE electrochemical oxidation pathway of TPA.34-40 According to this pathway (Scheme 1), the initial oxidation of TPA produces an unstable radical cation TPA•+, and a chemical follow-up reaction product, a dimer TPB. TPB is more easily oxidized than TPA. The oxidation of TPB occurs in two discrete one-electron steps to a radical cation TPB•+ and a dication TPB2+. The electrochemical formation of TPB•+ (m/z 488) and its conversion to TPB2+ (m/z 244) was further investigated by EC/PB/MS, as described below. Off-Line FIA. The electrochemical behavior of TPA under flow conditions was examined by off-line FIA. Figure 2 shows the FIA current response curves obtained with the injection of a blank solution of 0.1 M TBAP in acetonitrile (a), followed by a mixture of 100 µM TPA and 0.1 M TBAP in acetonitrile (b), with the electrochemical flow-through cell voltage controlled at +1.2 V vs Pd. As can be seen from Figure 2, the injection of the solution of 100 µM TPA /0.1 M TBAP gives rise to a current peak, while a smooth baseline is observed after the injection of 0.1 M TBAP. The results verify the electrochemical oxidation of TPA in the flow cell. The electrochemical cell conversion efficiencies in the oxidation of TPA, at several different flow rates, are summarized in Table 1. The contribution of the oxidation of TPB product was

Table 1. Cell Conversion Efficiency for the Oxidation of TPAa flow rate (mL/min)

TPA injected (mol)

charge detected (C)

conversn effic (%)b

0.4 1.0 2.0

2.0 × 10-8 2.0 × 10-8 2.0 × 10-8

3.38 × 10-4 1.76 × 10-4 8.34 × 10-5

17.5 9.1 4.3

a Conditions: 1 mM TPA/0.1 M TBAP in acetonitrile, injection volume 20 µL, mobile phase 0.1 M TBAP in acetonitrile, and voltage applied +1.2 V vs Pd. b Oxidation of chemical follow-up reaction product was not considered. Fluctuations of the cell conversion efficiency were estimated to be 30-40%.

Figure 4. EI mass spectra of TPA, TBAP, and TPA/TBAP obtained by PB/MS: (a) 1 mM TPA; (b) 0.01 M TBAP; (c) 1 mM TPA/0.01 M TBAP in acetonitrile. Conditions: injection volume 20 µL, mobile phase acetonitrile, flow rate 0.4 mL/min, EI electron energy 70 eV, and source temperature 250 °C.

Figure 3. Hydrodynamic voltammogram of TPA obtained by FIA: 100 µM TPA with 0. 01 M TBAP. Potential window from +0.5 to +1.9 V. Other experimental conditions as in Figure 2. Voltammogram of TPA was constructed by plotting the blank-subtracted peak current response.

not considered in the calculation of the cell conversion efficiency. Table 1 shows that a higher conversion efficiency is achieved at lower flow rates, which is as expected.9,10 This observation is related to a longer residence time of the analyte in the flow cell at lower flow rates, which increases the amount of TPA that can be converted in the electrochemical reaction during its passage through the cell.9 The high conversion efficiencies in Table 1 may be related to the effect of the organic solvent on the conversion efficiency.9-11 The presence of an organic solvent in the mobile phase has been shown to improve the conversion efficiency in a mixed solvent system.10 Figure 3 shows a hydrodynamic voltammogram that was constructed from the blank-subtracted FIA data, as described in the Experimental Section. The voltage applied to the cell was in the range of +0.5 to +1.9 V vs Pd. Compared to the voltammograms in Figure 1, the onset of the electrode reaction of TPA in the flow cell, which is indicated by the increase in the cell current, has been shifted to somewhat more positive potentials. The current of TPA starts to increase at +0.8 V vs Pd, and a maximum current is observed at ∼+1.1 V. The positive shift of the reaction potential in the flow cell may be due in part to the ohmic drop in the flow-through electrochemical cell9-11 and may also be a result of different surface conditions at the working RPG electrode in

the electrochemical cell during the FIA experiment, under flow conditions. In Figure 3, the current drops to a low value at ∼+1.2 V, then increases to a relatively stable level, and fluctuates around this level until the potential applied to the flow cell reaches ∼+1.6 V. The observed current fluctuations can be attributed to the oxidation in the flow cell of TPA oxidation products such as the dimer product. In Figure 3, the initial increase of the current at ∼+0.8 V reflects the onset of the oxidation of TPA. As the voltage that is applied to the electrochemical cell is increased, the rate of formation of the dimer product as a result of the oxidation of TPA, and thereby the contribution of the oxidation of TPB to the current response, also increases. The oxidation of TPB•+ to TPB2+ is a slow kinetic step.38-40 The current observed at ∼+1.1 V includes mainly the contributions from the oxidation of TPA and from oxidation of TPB to TPB•+. TPB•+ can also be produced via solution chemistry, i.e., oxidation of TPB by TPA•+.39 When the voltage is greater than ∼+1.1 V, TPB and TPB•+ predominate at the electrode surface, and the current is mainly controlled by the oxidation of TPB to TPB•+, while the contribution from the oxidation of TPA to TPA•+ and of TPB by TPA•+ becomes less significant. Between about +1.3 and + 1.6 V, the oxidation of TPB•+ to TPB2+ starts to be the major process contributing to the current. Above ∼+1.6 V, a rapid increase in current is observed, which can be interpreted as due to further oxidation of TPB2+ at more positive potentials. The off-line FIA results are in agreement with the EC/MS results described below. On-Line FIA in EC/PB/MS. Shown in Figure 4a-c are the mass spectra obtained by PB/MS following the injections of acetonitrile solutions of 1 mM TPA, 0.01 M TBAP, and 1 mM Analytical Chemistry, Vol. 72, No. 11, June 1, 2000

2537

Figure 5. Mass spectrometric ion chromatogram at m/z 488 with change of cell voltage in EC/PB/MS. Conditions: 1 mM TPA/0.01 M TBAP in acetonitrile, injection volume 20 µL, mobile phase acetonitrile, flow rate 0.4 mL/min, and injection at 30 s after cell voltage was applied. Selective ion monitoring at m/z 488.

TPA/0.01 M TBAP, respectively, with the electrochemical flow cell on-line, but without applying voltage to the flow-through electrochemical cell. Figure 4a illustrates the mass spectrum of TPA. The ions identified following EI ionization of TPA show significant intensities at m/z 245 (TPA•+), 167 (loss of C6H6 from TPA•+), 77 (C6H5+), and 51 (loss of C2H2 from C6H5+). Figure 4b shows the EI mass spectrum of TBAP, which shows ions at m/z 242 ((C4H9)4N+), 142 ((C4H9)2N+dCH2), 100 ((C4H9)(CH3)N+d CH2), and 57 ((C4H9)+). The mass spectra in Figure 4 are in agreement with the on-line library spectra and the literature.42,43 When TPA and TBAP are present together in acetonitrile solution, ions from the ionization of both TPA and TBAP appear in the mass spectrum, as shown in Figure 4c. When voltage is applied to the electrochemical flow cell in the EC/PB/MS experiment, no new ions are detected by MS in the scan mode. However, selective ion monitoring at m/z 488 (TPB•+) produces the ion chromatograms shown in Figure 5. The voltage that was applied to the cell is shown in Figure 5. When the applied voltage is changed from +0.6 to +2.0 V, as shown in Figure 5, the change in the intensity of the selected ion at m/z 488, with the electrode potential, is clearly observed. As can be seen in Figure 5, when the applied voltage is below +0.8 V, only a small response is observed at m/z 488. When the voltage is increased to +1.0 V, the ion intensity of m/z 488 ion increases. A maximum response was observed at ∼+1.2 V, and above +1.2 V, the ion peak due to the m/z 488 ion starts to decrease with the increase in the applied cell voltage. The MS response at m/z 488 decreases to nearly background levels when the voltage is increased to ∼+2.0 V. With the voltage set back to a smaller value, e.g., +1.1 V, the ion peak increases again, which indicates that the small response at +2.0 V is caused by the cell voltage, and the related electrochemical reactions at this cell voltage, and is not a result of the MS response decreasing with time, after multiple injections. The ion chromatogram obtained at 1.1 V (Figure 5) also shows that the inactivation of the working electrode in the flow-through cell as a result of the multiple injections during a continuous operation in EC/PB/MS was not significant. The results confirm the formation of TPB•+ monocation (42) Gierlich, H. H.; Rollgen, F. W.; Borchers, F.; Levsen, K. Org. Mass Spectrom. 1977, 12, 387-390. (43) Veith, H. J. Org. Mass Spectrom. 1983, 18, 154-158.

2538 Analytical Chemistry, Vol. 72, No. 11, June 1, 2000

Figure 6. Effect of cell voltage on TPB•+ ion abundance at m/z 488 and on ion intensity ratio of ions at m/z 244 (TPB2+) to 245 (TPA•+) in EC/PB/MS. (b) Ion abundance at m/z 488, solid line, left Y-axis; ([) ratio of ion abundance ratio at m/z 244 to that at m/z 245, dashed line, right Y-axis. The ion abundance was measured as the integrated peak area of the corresponding ion peaks.

in the electrochemical oxidation of TPA, which can be monitored by EC/PB/MS at m/z 488. Direct EC/PB/MS Monitoring of Radical Intermediate and Product. Figure 6 shows a plot of the integrated ion abundance at m/z 488 (TPB•+), as well as the ratio of ion abundance ratio at m/z 244 (TPB2+) to that at m/z 245 (TPA•+), obtained by EC/PB/MS as a function of the electrode potential. As shown in Figure 6, an increase in the abundance of the radical cation intermediate TPB•+ is observed with the increase in the electrochemical cell voltage between +0.8 and +1.2 V. The rapid increase of ion abundance of TPB•+ between +0.8 and +1.2 V can be attributed to a rapid formation of the radical cation as a result of the electrochemical oxidation of TPA in the flow cell. The abundance of the TPB•+ intermediate decreases between +1.2 and +1.6 V vs Pd, with a rapid decrease above +1.6 V. The high rate of formation of TPB•+ between +1.2 and +1.6 V is accompanied by a high rate of formation of the dication product TPB2+ (m/z 244) as shown in Figure 6. Figure 6 shows that, at potentials positive of 1.2 V, the latter reaction rate increases while the rate of formation of the intermediate TPB•+ decreases. Above +1.3 V, the rate of conversion of TPB•+ to TPB2+ exceeds the rate of generation of TPB•+. Above +1.6 V, the ion abundances at m/z 488 and at m/z 244 decrease rapidly with an increase in cell voltage, probably as a result of further oxidation of TPB2+ at very positive potentials. The EC/PB/MS results are in agreement with the results of off-line FIA analysis illustrated in Figure 3. The current in Figure 3 at less positive potentials of ∼1.2 V is lower than at more positive potentials of 1.6 V, while the ion abundance in EC/PB/MS is similar and high throughout the same potential window, and this likely reflects high sensitivity of PB/MS to the species being detected.11 Ion Formation in EC/PB/MS. According to the oxidation pathway of TPA, in EC/PB/MS, the electrochemical oxidation of TPA in a flow-through electrochemical cell can be expected to produce ions at m/z 245, 488, and 244. If the ions have a long enough lifetime to enter the ion source of the mass spectrometer, and reach the detector, they will give rise to ion peaks in the ion chromatogram. However, in EC/PB/MS, the reverse is not true; i.e., not all ions that are observed are due to the electrochemical

oxidation of TPA. Since the electrochemical cell conversion efficiency is rather low in the oxidation of TPA, EI ionization of TPA may generate ions in large enough quantities for them to show significant spectral overlap with the products of the electrochemical oxidation of TPA in the electrochemical cell during EC/PB/MS. It is therefore important to consider the possible origin of all ions, to verify the electrochemical formation of the intermediates and products. Ion Formation by EI. Electrooxidation of TPA in the electrochemical cell generates TPA•+ (m/z 245), an intermediate radical monocation TPB•+ (m/z 488), and a dication TPB2+ (m/z 244). At the same time, as shown in Figue 4a and c, EI ionization of TPA produces ions at m/z 245 and 244. The ion at m/z 245 is formed by EI after a loss of one electron by TPA, a reaction that is equivalent to the formation of the cation radical TPA•+, with the same m/z (245), in the electrochemical oxidation of TPA in the flow cell. However, the ion at m/z 244 formed in the EI ion source is not the same as the dimer TPB2+ (m/z 244), formed by the electrochemical oxidation of TPA in the electrochemical flow cell. Because the EI ion source is operated under high vacuum, the residence time of ions that are formed in the ion source by EI is very short (∼10-6 s). Combined with a relatively low rate constant for the formation of TPB2+ dication from TPB•+, which is ∼1.1 × 103-3 ×103 dm3 mol-1 s-1,38-40 the probability of formation of TPB2+ dication in the EI ion source is small. The ion intensity of the ion at m/z 244 obtained in the EI ion source is ∼42% that at m/z 245 (of TPA•+), as can be seen from Figure 4c. Consequently, the ion abundance observed at m/z 244 as a result of EI ionization is much too high to be due to the formation of TPB2+ in the EI ion source. Furthermore, the ion expected at m/z 488 of TPB•+, a dimer radical cation precursor ion of TPB2+ dication, is not detected after EI ionization. It can thus be concluded that the ion generated by EI at m/z 244 is not due to TPB2+ but more likely results from a loss of a hydrogen atom from TPA•+, as a consequence of high excess internal energy during the ionization of TPA to TPA•+ in the EI source. Ion Intensity Ratio for Ion Monitoring. Since the ionization of TPA by EI does not produce an ion at m/z 488, the electrochemical formation of TPB•+ (m/z 488) radical monocation can be monitored directly at m/z 488 via SIM. The ion abundance at m/z 488 is in fact controlled by cell voltage, as can be seen from the change of the ion abundance with the change in cell voltage in Figure 6. The electrochemical formation of TPB2+ dication (m/z 244) cannot be monitored directly at m/z 244 because of spectral interference from TPA-H•+ formed via a loss of hydrogen atom from TPA•+ during EI ionization, which generates an interfering ion at the same m/z 244. Instead, as shown in Figure 6, the intensity ratio of ions at m/z 244 vs 245 can be monitored with the change in the applied cell voltage to reflect the status of TPB2+ at m/z 244, which is only generated in the electrochemical reaction. Concentrations of TPA and TBAP. Both TPA and TBAP are nonvolatile, and when their concentrations are high, PB/MS analysis of TPA/TBAP mixtures causes plugging of the PB interface, degradation of the interface performance, and contamination of the EI ion source. To alleviate these problems, low concentrations of TPA and TBAP are preferred. But at low concentrations, e.g., 100 µM TPA and 1 mM TBAP, the TPB•+

Figure 7. Mass spectrometric extracted-ion chromatogram at m/z 245 obtained by PB/MS with EI ionization. Conditions: 1 mM TPA, followed by 1 mM TPA/0.01 M TBAP in acetonitrile. Same experimental conditions as in Figure 4.

intermediate is not detected via either scan or SIM mode. When the concentrations of TPA and TBAP are higher, e.g., 1 mM and 0.01 M, respectively, TPB•+ can be detected via SIM. High concentrations of TBAP also improve conductivity in the flow cell, which improves the control of cell voltage and reduces ohmic drop. When the concentrations of TPA and TBAP are high, a decrease in the intensity of TPA ions is observed at m/z 244 and 245 after each injection. Nevertheless, monitoring of the intensity ratio at m/z 244/245 allows detection of the TPB2+ dication (m/z 244) formed in the electrochemical reaction. The ion peak at m/z 488 of TPB•+, of the radical intermediate formed in the electrochemical oxidation reaction of TPB, is less sensitive to the deteriorating performance of the PB interface and the ion source, which is observed at high concentrations of TPA and TBAP. This is consistent with a higher mass and higher momentum of TPB•+ which favors transfer of the ion into the ion source.11 An apparent carrier effect in the presence of TBAP, which is discussed in the following section, additionally contributes to the increase in the transmission efficiency of TPB•+. When concentrations of TPA and TBAP are lower, e.g.,100 µM TPA and 1 mM TBAP, contamination of the PB interface and the ion source is reduced, and good sensitivity is observed after repeated sample injections. Signal Enhancement with TBAP. A comparison of the abundance scales of the mass spectra in Figure 4a and c shows that, in the presence of TBAP, a significant increase in ion intensity is obtained in EI ionization of TPA. To better illustrate the observed enhancement, Figure 7 compares the extracted-ion chromatogram of TPA•+ at m/z 245 in the absence and presence of TBAP. The apparent enhancement in the presence of TBAP, quantified as peak area, is estimated at ∼35-fold. As a result, MS detection sensitivity is improved, and the detectable concentration of TPA is reduced from about 100 to 1 µM, or the mass detection limit is reduced from about 490 to 4.9 ng. The observed enhancement of ion signal in the presence of TBAP is similar to the signal enhancement observed previously in LC/PB/MS analysis of different classes of compounds in the presence of ammonium acetate, neutral analogues of analytes, and isotope-labeled compounds.23-31 The effect was found previously to influence sensitivity in water-containing organic mobile phases. Unlike ammonium acetate, whose volatility and decomposition were believed to play a role in previously observed signal enhanceAnalytical Chemistry, Vol. 72, No. 11, June 1, 2000

2539

ment,28,31 TBAP is nonvolatile and is relatively stable under mild temperature operating conditions of the PB interface. TBAP is also different from neutral analyte analogues and/or isotopelabeled compounds that were observed previously to contribute to signal enhancement in that it contains ionic charge. Here, we report for the first time that in a nonaqueous solvent, acetonitrile, ion signal enhancement can be achieved in PB/MS via addition of a nonvolatile electrolyte, TBAP. The effect of TBAP on sensitivity enhancement in EC/PB/MS may play an important role in enabling the direct monitoring of the formation of the intermediate radical cation TPB•+ by EC/PB/MS. CONCLUSIONS Anodic oxidation of TPA was investigated by on-line EC/PB/ MS, with EI. Direct detection of an intermediate radical ion, TPB•+, in the electrochemical oxidation of TPA was achieved. Electrochemical formation of TPB•+ and its subsequent electrochemical oxidation to TPB2+ was verified by EC/PB/MS, and the rate of formation of each ion was shown to be a function of the electrode potential. In conventional cyclic voltammetry, TPA showed an oxidation peak at +0.883 V vs Pd, while hydrodynamic voltammograms obtained by FIA indicated a small positive shift of the reaction potential in the electrochemical flow cell. EC/PB/MS allowed identification of cell potentials at which the rate of generation of the radical intermediate was the highest. The results confirmed a slower rate of electrochemical formation of the final dication product. The EC/PB/MS results showed that the oxidation of

2540

Analytical Chemistry, Vol. 72, No. 11, June 1, 2000

the dimer TPB to a monocation TPB•+ and then the dication TPB2+, following the initial electrooxidation of TPA to TPA•+ in the flow cell, occurs at highest rates at different potentials that are more positive than the potentials of the voltammetric oxidation of TPA. We also report, for the first time, a significant signal enhancement in PB/MS in the presence of TBAP. The observed signal enhancement may facilitate other EC/MS studies including in nonaqueous solvents. The EC/PB/MS results show the advantages of the PB interface in EC/MS studies, such as ease of coupling of EC with MS via the PB interface, flow rate compatibility, and the advantage of EI ionization, which generates mass spectra that can be library searched. Mainly, the compatibility of PB/MS with EC/MS stems from the ability to use an electrolyte at high concentrations. The resulting signal enhancement observed in the presence of a nonvolatile electrolyte TBAP may facilitate other applications of LC/MS. ACKNOWLEDGMENT We thank Dr. J. P. Toth for many helpful discussions. Access to the IFAS mass spectrometer facilities is very much appreciated. T.Z. also acknowledges Ms. Susan Hillier for her assistance with the use of the mass spectrometer. Received for review October 29, 1999. Accepted March 4, 2000. AC9912348