Electrochemical Cells Integrated in ESI Emitters
One way of combining EC with ESI is to integrate the emitter electrode in a controlled-potential electrochemical cell (Fig. 13). A simple setup for incorporating an electrochemical cell into the ESI source was presented by Van Berkel and coworkers (Zhou & Van Berkel, 1995) and Brajter-Toth and coworkers (Zhang et al., 2002; Mautjana et al., 2008a,2008b, 2009; Looi, Eyler, & Brajter-Toth, 2011). The ESI emitter was the working electrode of a two-electrode cell (Fig. 13a). This cell was constructed by connecting the ESI emitter through low volume plastic tubing with the stainless-steel counter/reference electrode. The voltage was applied with a 9 V battery across a variable resistor. Decoupling of the electrochemical cell from the ESI high voltage was accomplished by allowing the EC system to float on the potential induced by the ESI high voltage.
Figure 13. Schematic diagrams of (a) a two-electrode electrochemical ESI emitter (Zhang et al., 2002), (b) a three-electrode electrochemical ESI emitter with the high voltage applied to the auxiliary electrode (Xu, Lu, & Cole, 1996), and three-electrode electrochemical ESI emitters with the high voltage applied to (c) a planar working electrode as well as (d) a porous working electrode (Van Berkel, Asano, & Granger, 2004a).
Download figure to PowerPoint
Cole and coworkers have integrated a three-electrode cell into the ESI source (Xu, Lu, & Cole, 1996; Lu, Xu, & Cole, 1997). In the three-electrode cell, the working electrode was a platinum wire covered by insulating fused silica (Fig. 13b). The fused-silica capillary prevented electrical contact between the working electrode and the sample solution until immediately prior to the electrospray region. Furthermore, it insulated the working electrode from the stainless-steel auxiliary electrode, which also functioned as the ESI emitter electrode. The reference electrode was placed upstream outside of the sprayer in a separate compartment. The voltage was applied with a potentiostat and was floated at the ESI voltage. The most important feature of the developed setup was that it generated electrochemical intermediates (e.g., radical cations) and products in situ at the tip of the ESI emitter, thus keeping response times to a minimum.
Van Berkel and coworkers have also incorporated three-electrode cells into the electrospray emitter circuit (Van Berkel, Asano, & Granger, 2004a; Kertesz, Van Berkel, & Granger, 2005). Two different basic cell designs were used, namely, a planar flow-by working electrode (Fig. 13c) and a porous flow-through working electrode design (Fig. 13d), each operated with a potentiostat floated at the ESI voltage. In each case the working electrode also functioned as the emitter electrode of the ESI source. Reserpine was used as sample to test the cells. The authors showed that reserpine oxidation was tunable by the electrochemical potential applied. Extensive reserpine oxidation was observed at potentials more positive than the potential necessary to induce reserpine oxidation. This oxidation occurred at the working electrode. Unexpectedly, reserpine oxidation was also observed at very negative potentials. This unwanted analyte electrolysis occurred at the auxiliary electrode and was prevented either with the auxiliary electrode removed from direct contact with the analyte in the flow stream, or with mass transport of the analyte to that electrode limited by another method (e.g., small surface area).
Hyphenation of Discrete Systems
Another way of combining EC with ESI is to hyphenate discrete systems (Fig. 14). In such a setup ESI-MS is employed to specifically detect and characterize the products of electrochemical reactions produced in an independently working electrochemical cell.
Figure 14. Schematic diagrams of electrochemical cell designs used for EC/ESI-MS: (a) a three-electrode cell with a porous flow-through working electrode, (b) a three-electrode cell with a planar flow-by working electrode, and (c) a microfluidic two-electrode cell in a lab-on-chip format. Reproduced from Mengeaud et al. (2002) with permission of the Royal Society of Chemistry (Copyright 2002).
Download figure to PowerPoint
The simplest electrochemical cell hyphenated to ESI-MS consisted of two electrodes only (Bond et al., 1995). In such a flow cell, two tubular electrodes were separated by insulating tubing (e.g., PEEK, Teflon). The setup was successfully applied to induce oxidation of copper, nickel and cobalt diethyldithiocarbamates. A similar two-electrode setup was used to study the electrochemical reduction of metalloproteins (Johnson et al., 2001).
In state-of-the-art EC/ESI-MS, three-electrode cells are hyphenated to ESI-MS. In one commonly applied cell design a porous flow-through working electrode is used (Fig. 14a). This cell design has been introduced to EC/MS by Brajter-Toth and coworkers (Volk et al., 1988; Volk, Yost, & Brajter-Toth, 1989, 1992). In this pioneering work, thermospray was used as ionization technique, and the redox reactivity of uric acid, 6-thioxanthine and two purines was studied at a glassy carbon working electrode. The porous-electrode flow cell has been commercialized by ESA, Inc. (Bedford, MA) which is now part of Thermo Fisher Scientific (Waltham, MA).
Cells with porous electrodes are considered to provide good conversion rates even at high flow rates due to the large surface area provided. A further characteristic of these coulometric cells is the low maintenance effort. The electrodes are usually simply cleaned by flushing with appropriate solvents. Even though adsorption can take place on the electrode surface and residues might not be fully removed, the effects on the oxidation process often remain negligible (Baumann & Karst, 2010). Sometimes, however, life history and/or age of the electrochemical cell could have an impact on the oxidation reactions observed (Permentier & Bruins, 2004).
The first EC/ESI-MS setup using a porous electrode was presented by Van Berkel and coworkers (Zhou & Van Berkel, 1995), and was applied to study the oxidation of nickel(II) octaethylporphyrin. The cyclic voltammogram of this compound and the mass spectrometric voltammograms of the oxidation products are shown in Figure 15. These diagrams clearly indicated that nickel(II) octaethylporphyrin undergoes two reversible one-electron oxidation reactions giving rise to the corresponding monocation (M•+, m/z = 590) and dication (M2+, m/z = 295). The observed differences in the “appearance potentials” were 0.5 V.
Figure 15. a: Cyclic voltammogram and (b) mass spectrometric voltammogram of nickel(II) octaethylporphyrin. Reproduced from Zhou and Van Berkel (1995) with permission of the American Chemical Society (Copyright 1995).
Download figure to PowerPoint
Bruins and coworkers have extensively used three-electrode cells with the porous electrode to study the redox reactivity of drugs (Jurva, Wikstrom, & Bruins, 2000; Jurva et al., 2003). This work laid the foundation for the use of EC/ESI-MS techniques to mimic phase I oxidative reactions in drug metabolism. EC was found to be particularly useful in cases where the P450 enzyme catalyzed reactions are supposed to proceed via a mechanism initiated by a one-electron oxidation, such as N-dealkylation, S-oxidation, P-oxidation, alcohol oxidation, and dehydrogenation. As valuable information concerning the sensitivity of a substrate towards oxidation can be obtained from EC/ESI-MS experiments, the technique is regarded as an efficient tool in the drug development process.
One of the first examples of successful mimicking of drug metabolism is shown in Figure 16. In a proof-of-principle study, Bruins and coworkers have applied EC to oxidize lidocaine (Jurva, Wikstrom, & Bruins, 2000). Two different oxidation products were detected by ESI-MS. Of particular importance was the formation of dealkylated lidocaine, because this compound is also formed in vivo catalyzed by CYP3A4.
Figure 16. Extracted ion voltammograms of [M + H]+ ions of (a) lidocaine and (b) its oxidation products plotted against the applied cell potential. Reproduced from Jurva, Wikstrom, and Bruins (2000) with permission of John Wiley & Sons (Copyright 2000).
Download figure to PowerPoint
Another important application of EC/ESI-MS introduced by Bruins and coworkers using the porous electrode system was oxidative cleavage of peptides and proteins (Permentier et al., 2003; Permentier & Bruins, 2004; Roeser et al., 2010b). Generally, oxidation of peptides and proteins involves the amino acids tyrosine, tryptophan, cysteine, methionine, and histidine. The electrochemical oxidation of tyrosine- or tryptophan-containing species gives further rise to backbone cleavage at the C-terminal side of these specific amino acids (Fig. 17). Due to its distinct amino acid specificity, its speed of analysis, its easy coupling to MS, EC was considered a potential instrumental alternative to chemical and enzymatic cleavage with immediate applicability in proteomics.
Figure 17. Proposed reaction mechanism for electrochemical oxidation and cleavage at tyrosine (top) and tryptophan (bottom). R1 and R2 are the parts of the protein N-terminal and C-terminal, respectively, to tyrosine and tryptophan. Reproduced from Permentier and Bruins (2004) with permission of the Springer-Verlag (Copyright 2004).
Download figure to PowerPoint
Another important cell type commonly used in EC/ESI-MS is based on the planar flow-by working electrode design (Fig. 14b). This cell design has been introduced to EC/ESI-MS by Van Berkel and coworkers to study the oxidation reaction of β-carotene on a glassy carbon electrode (Zhou & Van Berkel, 1995). Thin-layer cells are commercially available from Bioanalytical Systems, Inc. (West Lafayette, IN), Antec (Zoeterwoude, The Netherlands), or Kimoto Electric Co (Osaka, Japan). In comparison to porous electrodes, the surface area of planar electrodes is rather low, and the flow rate can be used to efficiently control the conversion rate (Erb et al., 2012). For obtaining very high conversion rates the cells are usually operated at very low flow rates (<10 µL/min). Another parameter that might influence the conversion efficiency of this kind of electrochemical cell is the analyte concentration. Within the residence time the analyte needs to migrate to the electrode to become oxidized. Theoretically, the more analyte available, the more analyte should be converted. It seems, however, that a maximum conversion efficiency exists for thin-layer cells (Erb et al., 2012). The electrochemical cell can be overloaded. As half-wave potentials and peak potentials, which are often used as specific parameters to characterize redox systems, can be shifted to higher values with increasing analyte concentrations, loading effects can have particular implications for mechanistic studies. A clear advantage of thin-layer cells is the possibility to use different electrode materials. Besides glassy carbon, platinum and boron-doped diamond electrodes are the most promising alternatives. Usually, less adsorption on these electrode materials occurs (Baumann & Karst, 2010). If adsorption residues are observed, they can be manually polished from the surface. In some cases analyte adsorption can be of analytical advantage. Van Berkel and coworkers were the first that demonstrated the usefulness of electrochemically controlled sample preconcentration and purification prior to ESI-MS analysis (Pretty et al., 2000). Tamoxifen and its metabolite 4-hydroxytamoxifen were accumulated on a glassy carbon electrode via nonelectrolytic adsorption. Once on the electrode, the analytes were washed free of sample matrix. For release and subsequent mass spectrometric detection of the unaltered analytes, the potential of the working electrode was changed. With the developed method, nanomolar levels of the analytes were detected in urine samples. Nyholm and coworkers took advantage of nonelectrolytic adsorption for the detection of thiols (Bökman et al., 2004). Thiols are known to strongly adsorb on gold surfaces, and this was used to preconcentrate 1-hexenthiol on a gold working electrode. Desorption was made by applying an electrochemical potential that was sufficiently high to induce oxidation of the thiols to the corresponding sulfinates and sulfonates. These species were detected by ESI-MS. More recently, Lev and coworkers presented another approach for electrochemical preconcentration of analytes (Gutkin, Gun, & Lev, 2009). Their approach was based on the electrodeposition of an active silver layer, subsequent specific accumulation of the target analyte onto the active layer, and finally oxidative electrostripping of the conductive layer along with the supported analyte to ESI-MS. The technique was found to be useful for the analysis of homocysteine and other organothiols.
When using a thin-layer cell, electrode cross-talk may become a problem (Deng & Van Berkel, 1999a; Bökman et al., 2004; Zettersten et al., 2006). The counter and working electrodes are often positioned so close to each other that only a thin spacer, defining the flow channel, separates them. In such an arrangement, where the solution containing the analytes passes over both the working and auxiliary electrodes, there is a risk that electrochemical reactions at the auxiliary electrodes may effect the appearance of the mass spectra. One such possibility involves redox cycling, by which the product formed at the working electrode undergoes a reverse reaction on the counter electrode. Another possibility is that species directly or indirectly formed by electrochemical reactions at the auxiliary electrode may appear in the mass spectra. Such interference, however, can be avoided by using a cell with discrete compartments for working, auxiliary, and reference electrodes.
Another cell type used in EC/ESI-MS is based on microfluidic devices. These chip cells have been introduced by Girault and coworkers (Mengeaud et al., 2002), and they are commercially available from Antec and ALS (Tokyo, Japan). In one setup used the microfluidic electrochemical device consisted of arrays of interdigitated electrodes (Mengeaud et al., 2002; Liu et al., 2012). A schematic representation of this chip is provided in Figure 14c. Nyholm and coworkers integrated gold microcoil electrodes into their chip system (Liljegren et al., 2005). Odijk and coworkers designed an on-chip three-electrode electrochemical cell (Odijk et al., 2009, 2010, 2012). The microfluidic cells were successfully applied to study the redox reactivity of different drug compounds.
Microfluidic cells in a lab-on-chip format can provide specific advantages. By miniaturization, the surface-to-volume ratio is increased giving rise to improved mass transport properties necessary to achieve a high conversion efficiency of introduced analytes. Furthermore, cell volumes are small, giving the possibility to work with small volumes and amounts of sample. The miniaturized cells can easily be combined with ESI emitters. Such integrated systems offer fast response times (<1 sec), even though they are operated at low flow rates (Liljegren et al., 2005). Also, the lab-on-chip format may allow the production of low-cost disposables to circumvent the need for extensive cleaning of the electrodes after use.