Integration of Cryogenic Ion Vibrational Predissociation Spectroscopy

*E-mail: [email protected]. ... Vibrational predissociation spectroscopy was carried out by photoevaporation of weakly bound N2 adducts over the ...
0 downloads 0 Views 897KB Size
Article pubs.acs.org/ac

Integration of Cryogenic Ion Vibrational Predissociation Spectroscopy with a Mass Spectrometric Interface to an Electrochemical Cell Joseph A. Fournier, Arron B. Wolk, and Mark A. Johnson* Sterling Chemistry Laboratory, Yale University, P.O. Box 208107, New Haven, Connecticut 06520, United States S Supporting Information *

ABSTRACT: Cryogenic ion vibrational predissociation (CIVP) spectroscopy is used to structurally characterize electrochemically (EC)-generated oxidation products of the benchmark compound reserpine. Ionic products were isolated using EC-electrospray ionization (ESI) coupled to a 25 K ion trap prior to injection into a double-focusing, tandem time-of-flight photofragmentation mass spectrometer. Vibrational predissociation spectroscopy was carried out by photoevaporation of weakly bound N2 adducts over the range 800−3800 cm−1 in a linear (i.e., single photon) action regime, thus enabling direct comparison of the experimental vibrational pattern with harmonic calculations. The locations of the NH and OH stretching fundamentals are most consistent with formation of 9-hydroxyreserpine, which is a different isomer than considered previously. This approach thus provides a powerful structural dimension for the analysis of electrochemical processes detected with the sensitivity of mass spectrometry.

M

respectively, to yield 3,4-dehydrohydroxyreserpine and 3,4,5,6tetradehydrohydroxyreserpine (structures VI and VII). Interestingly, chemical oxidation has also been performed using Ce(IV),14 peroxodisulphate,15 nitrous acid,16 and sodium nitrite,17 but these reagents do not yield hydroxylated species. Instead, 3,4-dehydroreserpine and 3,4,5,6-tetradehydroreserpine (lumireserpine) are generated and have been identified using UV−vis fluorescence spectroscopy (structures IV and V). Herein, we demonstrate how vibrational predissociation spectroscopy of the N2 adducts of species formed during EC oxidation can be used to aid in their identification. Prior structural assignments for the reserpine oxidation products relied on chemical manipulations. For example, Allen and Powell18 originally assigned the first oxidation product to compound II in Figure 1 based on the observation that oxidation was not observed when the methoxy group at the 11position was either removed or moved to the 10-position. This behavior was rationalized in the context of methoxy group activation, directing oxidation at the 10-position. Subsequently, Van Berkel and co-workers concluded that compound III was instead the correct product through mass spectrometric analysis of reserpine ether derivatives.5,19 These two alternative structures should be trivial to differentiate by their vibrational spectra in the regions of the NH and OH stretching bands, and so acquiring this information is the central point of this

ass spectrometric sampling of electrochemical cells provides a sensitive means with which to identify species prepared in redox reactions during the course of the cyclic voltammetry wave.1,2 Indeed, commercial interfaces are available that incorporate cells directly into electrospray ionization (ESI) sources.3−7 Structural identification of product masses can then be deduced using the usual array of secondary analysis tools (e.g., msn, reactive modifications, etc.) that rely on collisional activation and transformation of primary ions.8 Very recently, however, generally applicable, high-performance vibrational spectroscopy has been demonstrated based on cryogenic ion processing,9 followed by single-photon vibrational predissociation of cold, weakly bound adducts.10−12 In this report, we demonstrate how cryogenic vibrational predissociation (CIVP) spectroscopy10 can be used to structurally characterize products generated by electrochemical oxidation of the organic compound reserpine, a prototypical system for the EC-ESI interface.2,4−7,13 The structure of protonated reserpine is presented at the top of Figure 1 with a rotatable structure available in the Supporting Information. Also shown are several other compounds invoked in the oxidation scheme discussed earlier by Van Berkel and coworkers.4−6,10 These products are formed through modifications of the tricyclic domain on the left, as indicated. Although there is consensus that the primary product corresponds to hydroxylation of the indole functionality, it has not been possible to establish which of the candidate sites (I, II, or III) are formed. Further oxidation involves the sequential loss of two protons and two electrons at the 3,4 and 5,6 positions, © 2013 American Chemical Society

Received: April 24, 2013 Accepted: June 16, 2013 Published: June 16, 2013 7339

dx.doi.org/10.1021/ac401228y | Anal. Chem. 2013, 85, 7339−7344

Analytical Chemistry

Article

Figure 1. Reserpine oxidation pathways and products. The first oxidation product corresponds to hydroxylation of the indole ring system, with three candidate structures indicated in the dotted blue box.

generated through difference frequency mixing of the signal and idler output of the OPA in a AgGaSe2 crystal. Due to the additional stage of frequency mixing, the bandwidth was estimated to increase to approximately 6 cm−1 in the lower frequency region. Photoexcitation of the tagged-ion packet resulted in the evaporation of the weakly bound N2 adducts. To ensure that photoevaporation occurred in a single photonstage was lowered until the linear photofragment yield to pulse regime, the energy of the pump laser through the amplification stage was lowered until linear photofragment yield to pulse energy was observed. Vibrational action spectra were recorded by measuring the photofragment yield as a function of laser frequency. Spectra were normalized to the laser pulse energy to account for fluctuations in laser performance over the course of the scan. EC-ESI. The electrochemical cell (Figure 2) was a commercial (Bioanalytical Systems, Inc., model MW-5052) thin-film cell with a glassy carbon working electrode and Ag/ AgCl reference electrode (MF-1044, 0.0005 in. spacer). In this

demonstration study. Comparison of the observed band positions with harmonic predictions and the spectral behavior of model compounds indicates that the dominant first oxidation product is actually due to hydroxylation of the indole ring at the 9-position (compound I), which was not previously considered.



EXPERIMENTAL SECTION Cryogenic Photofragmentation Mass Spectrometer. The ESI source and photofragmentation mass spectrometer at Yale have been described in detail elsewhere.10,20 Briefly, millimolar solutions of reserpine were prepared by dissolving reserpine (Sigma-Aldrich) in 50:50 acetonitrile:water with 0.75% acetic acid by volume and 5 mM ammonium acetate buffer (solution pH 4.3). The solutions were electrosprayed (using the electrochemical cell described below) at solution flow rates between 0.02 and 0.15 mL/h. The generated ions were guided through four differentially pumped stages, using two RF-only quadrupole guides and an octopole guide. Following a 90° quadrupole bender and a second octopole guide, the ions were trapped in a quadrupole ion trap (Jordan) attached to the second stage of a closed-cycle helium cryostat (Sumitomo, 1.5 W at 4.2 K). Temperature control down to 10 K was achieved using a 100 W resistive heater and was measured using a Si diode (Lake Shore). He buffer gas was pulsed (1 ms) into the trap (Parker Hannifin series 99 pulsed valve), collisionally cooling the ions and allowing for the condensation of N2 adducts with a trap temperature of ∼25 K. The ions were held in the trap for ∼90 ms and then ejected into the extraction region of a Wiley-McClaren TOF mass spectrometer by applying ±90 V push/pull voltages to the entrance and exit lenses of the trap, respectively. Photofragments created by laser excitation at the first TOF focus were separated by a reflectron-based secondary TOF stage and detected using microchannel plates. Infrared light was generated using a pulsed, tunable OPO/ OPA infrared laser (LaserVision) pumped by a Nd:YAG source (Continuum Surelite EX, 10 Hz, 7 ns). The idler output of the OPA was tunable from 2300 to 4500 cm−1 with a bandwidth of about 3 cm−1. The lower energy 600−2500 cm−1 region was

Figure 2. Schematic of the commercial glassy carbon electrode used to oxidize reserpine. The outer housing acts as the counterelectrode and was held at the electrospray voltage (∼3000 V). An additional 0−2 V was added to the working electrode to generate the electrochemical potential difference. Oxidation products were then electrosprayed into the instrument. 7340

dx.doi.org/10.1021/ac401228y | Anal. Chem. 2013, 85, 7339−7344

Analytical Chemistry

Article

method, the stainless steel housing acts as the counterelectrode and is held at the electrospray voltage (∼3 kV). A thin film of solution flows over the glassy carbon working electrode with a residence time of 1−10 s, depending on the solution flow rate. An additional 0−2 V was added to the working electrode using a 5 V power supply (floated at the electrospray voltage) and potentiometer. Cryogenic Ion Processing and N2 Tagging. Most CIVP spectra taken using ion traps to date have been conducted using either H2 or D2 as the mass “messenger.”10,11,21−24 Note that here we are concerned with protonated reserpine, where the excess proton undoubtedly attaches to the most basic site corresponding to the ternary nitrogen atom in the 4-position indicated in Figure 1. Because the reaction products anticipated in this study often differ by two Daltons (e.g., structures IV and V at 607 and 605, respectively), H2 or D2 tagging is not optimal, as the distributions arising from the tagged clusters lead to overlapping isobars at our resolution (m/Δm = 1000). Consequently, N2 was chosen as the mass messenger, which has also been frequently used in CIVP and yields spectra similar to those obtained with Ar.25−34 Formation of the N2 adducts was carried out by pulsing a buffer gas mixture containing N2 (1%) in He into the trap, which is allowed to pump out before the ions are extracted into the photofragmentation mass spectrometer. The adducts form at trap temperatures between 23 and 30 K and appear on all species extracted from the trap, as shown in the mass spectra displayed in Figure 3. An upper bound on the N2 binding energy can be deduced from the lowest energy CIVP resonance resulting in linear photoproducts, which in this case is the 967 cm−1 transition observed in reserpineH+ (Figure 4b). Computational Methods. Geometry optimizations and harmonic frequency calculations were performed using Gaussian 09.35 ReserpineH+ calculations were initialized from its neutral crystal structure and optimized using the B3LYP/631+G(d,p) level of theory and basis set. Significant rearrangement of the cyclohexane backbone was observed upon protonation of the neutral precursor due to formation of an ionic hydrogen bond (IHB) between the protonation site and one of the carbonyls (discussed further below). However, the chemically active indole moiety was largely unaffected by protonation. The calculated frequencies for reserperineH+ were scaled by 0.95 in the 2600−3800 cm−1 region to bring the calculated indole N1−H stretch into agreement with the observed band at 3503 cm−1 and by 0.99 in the 800−1800 cm−1 to bring the free carbonyl stretch into agreement with that observed at 1753 cm−1 (vide infra). Calculations on the oxidation products were also performed at the B3LYP/631+G(d,p) level and scaled by the same factors. The calculated intensities were divided by the transition frequency so they can be directly compared to the observed spectra,36 which are normalized to the laser energy per pulse and not the number of photons per pulse.

Figure 3. (Top) The cyclic voltammogram of 1 mM reserpine in an acidic solution displays two irreversible oxidative features at 0.85 and 1.4 V. (Bottom) Mass spectra showing the formation of various oxidation products at several working electrode potentials. The red arrows along the voltammetry wave indicate where the mass spectra were collected. At (a) 0.12 V, only protonated reserpine (m/z 609) is observed along with several N2 adducts. At (b) 0.90 V, m/z 625 forms from the oxygen addition to reserpine and corresponds to the first voltammetry wave. At (c) 1.5 V, the second oxidative wave is shown to arise from dehydroreserpine, m/z 607 and 605. Finally, at (d) 1.7 V, products at m/z 621 and 623, corresponding to dehydrohydroxyreserpine, dominate.

ESI study. There are two irreversible oxidative features within the 0−1.8 V scan range. The first wave is clear and appears around 0.85 V, while the second is less well-defined and is centered around 1.4 V. The evolution of the EC-ESI mass spectra at various locations throughout the cyclic voltammetry wave are presented in the bottom panels of Figure 3. Red arrows along the voltammetry wave indicate the potential values where the mass spectra in traces a−d of Figure 3 were taken. The mass spectrum in Figure 3b was obtained at the first voltammetry feature (labeled b) at 0.90 V and exhibits a new peak at m/z 625, formally due to the oxygen addition (i.e., corresponding to one or more of the compounds (I/II/III) in the blue box in Figure 1). The species at m/z 607 and 605 (structures IV and V) occur with an onset at around 1.5 V (Figure 3c) and are accompanied by a weaker peak at 641, formally due to the addition of two oxygen atoms. The second oxidative wave is then likely oxidation of reserpine to structures IV and V (and



RESULTS AND DISCUSSION Mass Spectrometric Observation of Reserpine Oxidation. To survey the oxidation products of reserpine, we first present the cyclic voltammogram of reserpine taken using a glassy carbon disk working electrode, Pt wire counterelectrode, and Ag/AgCl reference electrode at the top of Figure 3. The solution consisted of 1 mM reserpine in 50:50 acetonitrile:water with 5 mM ammonium acetate and 0.75% volume acetic acid, conditions identical to those used in the subsequent EC7341

dx.doi.org/10.1021/ac401228y | Anal. Chem. 2013, 85, 7339−7344

Analytical Chemistry

Article

reveal peaks corresponding to condensation of N2 adducts onto reserpineH+ as well as onto each of the oxidation products indicated above. Vibrational Predissociation Spectra of Protonated and Lithiated Reserpine. Figure 4 presents the N2vibrational predissociation spectrum of reserpineH+ (Figure 4b) along with the calculated harmonic spectrum (Figure 4a) for the structure indicated at the top of the figure, which accurately accounts for the well-resolved features in the experimental spectrum. The sharp, intense indole N1−H stretch, shown in red, appears highest in energy at 3503 cm−1, essentially identical to where this motif was observed in gas-phase isolated indole (3500 cm−1).37 The numerous CH stretch modes contribute to the congested feature spanning the 2700−3050 cm−1 range. Like many protonated amine molecules in this size range, the nominally positively charged NH moiety is H-bonded to a basic functionality to form an intramolecular H-bonded cyclic motif. In reserpineH+, this Hbond occurs to the ester carbonyl colored in gold in Figure 4, which should act to red shift both the N4−H+ donor and CO acceptor constituents of the linkage.11 This linked pair of transitions is highlighted in gold as well as bracketed in Figure 4a. The higher energy CO stretch feature (green) then arises from the nonhydrogen-bonded carbonyl (green circle in structure) which bridges the trimethoxyphenyl moiety and appears in the characteristic location (∼1770 cm−1) of a freeester carbonyl.38 The two distinct bands observed in the CO stretching region at 1707 and 1753 cm−1 (Figure 4b) are clearly in excellent agreement with this prediction, but the corresponding N4−H+ stretch is less convincing as the observed feature in the critical region near 3100 cm−1 (gold) is broad and merged with the envelope arising from the CH stretches. One way to empirically support the H-bonded N4−H+ stretching assignment is to follow the spectral consequences of replacing the excess proton with a lithium ion, and the inverted trace below Figure 4b presents the comparison of the reserpineLi+ spectrum with that of reserpineH+. Although the CH stretches remain relatively unperturbed around 2900 cm−1, the putative N4−H+ stretching feature at 3100 cm−1 is conspicuously absent in the Li+ adduct, thus confirming its assignment to the internal H bond. We note that there is an additional band at 3464 cm−1 in the reserpineLi+ spectrum, labeled * in Figure 4b, just below the indole N1−H stretch (red in Figure 4). This is likely due to a second conformation involving intramolecular attachment to the N1−H group, but these complications are beyond the scope of this work. The fingerprint region below 1700 cm−1 is quite congested, with features at 1642 and 1591 cm−1 (highlighted in orange in Figure 4b) clearly traced to the indole ring and phenyl ring stretching modes, respectively. As a rough guide, the bands in pink are expected for the CH and methoxy −CH3 bending and wagging modes on the two aromatic ring systems, while the methoxy C−O stretches are highlighted in purple. Although detailed spectroscopic analysis is not warranted for our goal of product identification, we remark that the linewidths throughout the spectrum are quite narrow and as such will provide a very high degree of confidence in the assignment of an unknown species at this m/z, by comparison of its CIVP spectrum with those from a library of known compounds. Thus, by offering a structural identification component, this technique appears poised to complement bulk electrolysis experiments,39 which typically yield spectoelectrochemical or mass identi-

Figure 4. N2-vibrational predissociation spectrum of reserpineH+. The calculated harmonic spectrum [B3LYP/6-31+G(d,p)] is shown in (a) and was scaled by 0.95 in the 2600−3700 cm−1 range to bring the N1− H stretch into agreement with the observed band, shown in red, and by 0.99 in the 800−1800 cm−1 range to match the observed free-ester carbonyl stretch (green). The calculation correctly predicts the redshift in the ester carbonyl that is hydrogen bonded to the N4−H+ protonation site (indicated in gold and displayed in structures at top). The N4−H+ stretch appears as a broad shoulder to the blue of the CH stretching region and was confirmed through comparison to the corresponding reserpineLi+ spectrum, shown in the inverted trace below (b). The extra band, labeled * in the Li+ spectrum, is likely from an alternative arrangement where the Li+ attaches to the indole NH.

further oxidation of m/z 625 to 641). Finally, Figure 3d displays the mass spectrum at 1.7 V, where water oxidation and oxidation of the carbon electrode begin to occur. At this voltage, peaks at m/z 623 and 621 (structures VI and VII) gain intensity, presumably through oxygen addition to IV and V, respectively. Each of these species has been previously observed by Van Berkel and co-workers using various EC-ESI methods.4−7,13,19 Note that the ions reported here are analyzed after being processed in the cold trap, and the mass spectra 7342

dx.doi.org/10.1021/ac401228y | Anal. Chem. 2013, 85, 7339−7344

Analytical Chemistry

Article

The higher energy feature at 3640 cm−1 appears unaffected by N2 solvation, as there is no change in intensity relative to the nearby NH transition at 3505 cm−1. On the other hand, the lowest energy member of the triplet at 3612 cm−1 grows at the expense of the central 3630 cm−1 peak when a second N2 is attached. This low-energy feature is thus most likely due to a structural isomer where N2 attaches to the OH moiety, red shifting the nominally free stretch at 3630 cm−1 by about 20 cm−1. The presence of both peaks in the single N2 spectrum would thus arise from at least two distinct binding sites. Regardless of the fine structure, and most importantly, the observed vibrational spectrum of the N2 tagged m/z 625 peak is at odds with that expected for both compound II18 and compound III.5,19 Specifically, the harmonic calculation for compound III (Figure 5d) predicts only an OH stretch in the NH/OH region, as formation of the N-hydroxyindole eliminates the telltale N1−H stretch. Moreover, the prediction places this OH stretch for III well below (135 cm−1) the observed transition at 3630 cm−1. Assignment of the oxidation product to compound II is more favorable, as that arrangement would yield both the initial N1−H fundamental and a new OH stretching transition, accounting for the two features in the experimental spectrum. The exact location of the observed OH stretch, however, lies 30 cm−1 above the predicted value for compound II (Figure 5c). This lower energy transition in II occurs because of the H-bond interaction between the OH and the neighboring methoxy group. Indeed, previous gas-phase studies on o-methoxyphenol40 and para-substituted o-methoxyphenol,41 which are models for compound II, show the OH stretch to occur at 3599 cm−1, essentially identical to that predicted (3597 cm−1) for compound II. This discrepancy led us to search for other possible isomers that would not exhibit such a red shift. Indeed, compound I, in which the OH group is too remote to undergo intramolecular interaction, is predicted to display the two transitions (Figure 5b) in very good agreement with the observed bands. We note that rotation of the methoxy group at the 11-position by 180° gives a minimum structure 260 cm−1 higher in energy with a small (5 cm−1) blue shift of the OH stretch, which would naturally account for the high-energy shoulder at 3640 cm−1. For example, the gas-phase spectrum of m-methoxyphenol, a model for compound I, reveals an OH doublet at 3654 and 3557 cm−1 attributed to such rotational conformers.40 Moreover, compound I is actually also calculated to be the lowest energy of the three conformers, lying about 500 cm−1 below compound II and 15000 cm−1 below compound III. We therefore propose that compound I, 9-hydroxyreserpine, is actually the most consistent with the spectroscopic and theoretical evidence.

fication of isolated electrochemical products through UV-vis, GC-MS, LC-MS, MS-MS, etc. Vibrational Predissociation Spectra of the Protonated Oxidation Product at m/z 625. The CIVP spectrum of the first oxidation product (compounds I/II/III, m/z 625) is displayed in Figure 5a in the 2600−3700 cm−1 region critical

Figure 5. N2-vibrational predissociation spectrum of (a) m/z 625, nominally reserpineH+ + O, along with calculated harmonic spectra [B3LYP/6-31+G(d,p)] for O insertion into three candidate oxidation sites: at the (b) 9-position (I), (c) 10-position (II) and (d) 1 position (III). A new triplet feature, highlighted in blue, around 3630 cm−1 appears in the oxidation product spectrum, indicative of free OH stretches. The inverted trace in (a) displays the predissociation spectrum using 2N2 tags. Changes in intensity of the OH stretches indicate that the lower energy feature arises from N2 attachment to the OH group. The N1−H stretch observed in reserpineH+ is shown in the inset above (a) and is again observed in the product spectrum. The harmonic calculation for compound I is in excellent agreement with the observed spectrum.

for structure determination (the fingerprint region is available in Figure S1 of the Supporting Information, and rotatable structures of compounds I, II, III are available in the Supporting Information). Calculations were performed for hydroxylation at the 1, 4, 9, 10, and 12-positions (Figure S2 of the Supporting Information), with harmonic spectra for candidate compounds I, II, and III compared with the observed spectrum in Figure 5 (panels b, c, and d, respectively). The transition at 3505 cm−1 is once again identified as the indole N1−H stretch and occurs at the same frequency as that found in reserpineH+ (reproduced at the top right of Figure 5a). New to the spectrum is a close triplet of peaks, colored blue, centered at 3630 cm−1, and located in the characteristic region for free, nonhydrogen-bonded OH stretches. The predissociation spectrum taken with 2N2 tags is presented as the inverted trace at the right of Figure 5a, indicating that the peak positions are the same but the intensity distribution among them changes upon addition of the second N2 molecule.



CONCLUSIONS The N2-vibrational predissociation spectra of protonated reserpine and its +O oxidation product have been recorded using a glassy carbon electrode EC-ESI source. Comparison of the resulting band pattern with harmonic calculations indicates that this product is most likely the 9-hydroxyreserpine isomer, a different form than previously suggested based on the behavior of derivatives. This study demonstrates our ability to spectroscopically interrogate species obtained by routine ESIMS analysis. Most importantly, the highly resolved spectra from the cold ions yield a high information content fingerprint with which to identify specific structures as a library of such spectra become available for important species. On the basis of this successful application to reserpine, identification of metabolites 7343

dx.doi.org/10.1021/ac401228y | Anal. Chem. 2013, 85, 7339−7344

Analytical Chemistry

Article

(20) Robertson, W. H.; Kelley, J. A.; Johnson, M. A. Rev. Sci. Instrum. 2000, 71, 4431. (21) Kamrath, M. Z.; Garand, E.; Jordan, P. A.; Leavitt, C. M.; Wolk, A. B.; Van Stipdonk, M. J.; Miller, S. J.; Johnson, M. A. J. Am. Chem. Soc. 2011, 133, 6440. (22) Leavitt, C. M.; Wolk, A. B.; Kamrath, M. Z.; Garand, E.; van Stipdonk, M. J.; Johnson, M. A. J. Am. Soc. Mass Spectrom. 2011, 22, 1941. (23) Garand, E.; Fournier, J. A.; Kamrath, M. Z.; Schley, N. D.; Crabtree, R. H.; Johnson, M. A. Phys. Chem. Chem. Phys. 2012, 14, 10109. (24) Goebbert, D. J.; Wende, T.; Bergmann, R.; Meijer, G.; Asmis, K. R. J. Phys. Chem. A 2009, 113, 5874. (25) Bailey, C. G.; Kim, J.; Dessent, C. E. H.; Johnson, M. A. Chem. Phys. Lett. 1997, 269, 122. (26) Bandyopadhyay, B.; Cheng, T. C.; Duncan, M. A. Int. J. Mass Spectrom. 2010, 297, 124. (27) Pillai, E. D.; Jaeger, T. D.; Duncan, M. A. J. Phys. Chem. A 2005, 109, 3521. (28) Pillai, E. D.; Jaeger, T. D.; Duncan, M. A. J. Am. Chem. Soc. 2007, 129, 2297. (29) Dopfer, O.; Olkhov, R. V.; Maier, J. P. J. Chem. Phys. 1999, 111, 10754. (30) Solca, N.; Dopfer, O. J. Phys. Chem. A 2001, 105, 5637. (31) Dopfer, O.; Roth, D.; Maier, J. P. J. Am. Chem. Soc. 2002, 124, 494. (32) Solcà, N.; Dopfer, O. J. Chem. Phys. 2004, 120, 10470. (33) Solcà, N.; Dopfer, O. J. Am. Chem. Soc. 2004, 126, 1716. (34) Solcà, N.; Dopfer, O. J. Phys. Chem. A 2005, 109, 6174. (35) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö .; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09; Gaussian, Inc.: Wallingford, CT, 2009. (36) Scerba, M. T.; DeBlase, A. F.; Bloom, S.; Dudding, T.; Johnson, M. A.; Lectka, T. J. Phys. Chem. A 2012, 116, 3556. (37) Miller, D. J.; Lisy, J. M. J. Chem. Phys. 2006, 124, 184301. (38) Shimanouchi, T. Tables of Molecular Vibrational Frequencies Consolidated Volume I; NSRDS-NBS 39; U.S. Department of Commerce: Washington, D.C., June 1972; 1. (39) Zhou, A. L.; Kikandi, S.; Sadik, O. A. Electrochem. Commun. 2007, 9, 2246. (40) Fujimaki, E.; Fujii, A.; Ebata, T.; Mikami, N. J. Chem. Phys. 1999, 110, 4238. (41) Rodrigo, C. P.; James, W. H.; Zwier, T. S. J. Am. Chem. Soc. 2011, 133, 2632.

associated with similar small drug molecules appear to be ideal targets for this capability.



ASSOCIATED CONTENT

S Supporting Information *

Figures S1 and S2 and rotatable structures of reserpineH+ and compounds I, II, and III. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors would like to thank Dr. Michael Z. Kamrath for his early work on the ESI-EC interface and Dr. Karin Young for her assistance and advice in the cyclic voltammetry experiments. M.A.J. thanks the Air Force Office of Scientific Research (AFOSR) under Grant FA-9550-09-1-0139, which provided support for the instrumentation used in the electrochemical interface to cryogenic ESI, and the National Science Foundation under Grant CHE-1213634, which has supported work on catalytic oxidation of aromatic systems. J.A.F. thanks the Department of Defense for support through a National Defense Science & Engineering Graduate Fellowship (NDSEG). This work was supported in part by the Yale University Faculty of Arts and Sciences High Performance Computing facility (and staff).



REFERENCES

(1) Baltruschat, H. J. Am. Soc. Mass Spectrom. 2004, 15, 1693. (2) Van Berkel, G. J.; Kertesz, V. Anal. Chem. 2007, 79, 5510. (3) Xu, X. M.; Lu, W. Z.; Cole, R. B. Anal. Chem. 1996, 68, 4244. (4) Van Berkel, G. J.; Asano, K. G.; Granger, M. C. Anal. Chem. 2004, 76, 1493. (5) Van Berkel, G. J.; Kertesz, V.; Ford, M. J.; Granger, M. C. J. Am. Soc. Mass Spectrom. 2004, 15, 1755. (6) Kertesz, V.; Van Berkel, G. J.; Granger, M. C. Anal. Chem. 2005, 77, 4366. (7) Van Berkel, G. J.; Kertesz, V. Anal. Chem. 2005, 77, 8041. (8) Armentrout, P. B. The Encyclopedia of Mass Spectrometry: Theory and Ion Chemistry; Gross, M. L., Caprioli, R. M., Eds.; Elsevier: San Franscisco, 2003; 1. (9) Wang, X. B.; Wang, L. S. Rev. Sci. Instrum. 2008, 79, 073108. (10) Kamrath, M. Z.; Relph, R. A.; Guasco, T. L.; Leavitt, C. M.; Johnson, M. A. Int. J. Mass Spectrom. 2011, 300, 91. (11) Garand, E.; Kamrath, M. Z.; Jordan, P. A.; Wolk, A. B.; Leavitt, C. M.; McCoy, A. B.; Miller, S. J.; Johnson, M. A. Science 2012, 335, 694. (12) Yacovitch, T. I.; Heine, N.; Brieger, C.; Wende, T.; Hock, C.; Neumark, D. M.; Asmis, K. R. J. Chem. Phys. 2012, 136, 241102. (13) Kertesz, V.; Van Berkel, G. J. Int. J. Mass Spectrom. 2011, 303, 206. (14) Sanchez, M.; SanchezAibar, J. J. Analyst 1996, 121, 1581. (15) Carmonaguzman, M. C.; Balon, M.; Gonzalezarjona, D.; Maraver, J.; Sanchez, M. J. Chem. Soc., Perkin Trans. 2 1986, 409. (16) Munoz, M. A.; Gonzalezarjona, D.; Balon, M. J. Chem. Soc., Perkin Trans. 2 1991, 453. (17) Ebel, S.; Muck, W.; Placke, F. J. J. Pharm. Biomed. Anal. 1989, 7, 709. (18) Allen, M. J.; Powell, V. J. J. Electrochem. Soc. 1958, 105, 541. (19) Pasilis, S. P.; Kertesz, V.; Van Berkel, G. J. Anal. Chem. 2008, 80, 1208. 7344

dx.doi.org/10.1021/ac401228y | Anal. Chem. 2013, 85, 7339−7344