VAN BERKEL, 2005 BIEMANN MEDAL AWARDEE Expanded Use of a Battery-Powered Two-Electrode Emitter Cell for Electrospray Mass Spectrometry

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector FOCUS: VAN BERKEL, 2005 BIEMANN MEDAL AWARDEE Expanded Use of a Battery-Powered Two-Electrode Emitter Cell for Electrospray Mass Spectrometry

Vilmos Kertesz and Gary J. Van Berkel Organic and Biological Mass Spectrometry Group, Chemical Sciences Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee, USA

A battery-powered, controlled-current, two-electrode electrochemical cell containing a porous flow-through working electrode with high surface area and multiple auxiliary electrodes with small total surface area was incorporated into the electrospray emitter circuit to control the electrochemical reactions of analytes in the electrospray emitter. This cell system provided the ability to control the extent of analyte oxidation in positive ion mode in the electrospray emitter by simply setting the magnitude and polarity of the current at the working electrode. In addition, this cell provided the ability to effectively reduce analytes in positive ion mode and oxidize analytes in negative ion mode. The small size, economics, and ease of use of such a battery-powered controlled-current emitter cell was demonstrated by powering a single resistor and switch circuit with a small-size, 3 V watch battery, all of which might be incorporated on the emitter cell. (J Am Soc Mass Spectrom 2006, 17, 953–961) © 2006 American Society for Mass Spectrometry

lectrochemistry is an inherent part of the normal Basicprinciplesofelectrochemistrydictate[5]and operation of an electrospray ion source as typi- ES-MSexperimentationandcalculation[3]haveshown Ecally configured for electrospray mass spectrom- that varied degrees of control over the electrochemical etry(ES-MS)[1,2].Oxidationreactionsinpositiveion processes involving the analytes can be achieved by mode and reduction reactions in negative ion mode are managing one or more of three basic parameters, viz., the predominate reactions at the emitter electrode con- mass transport to the ES emitter electrode, the magnitude tact (i.e., the working electrode in this system) that of the current (more precisely, current density) at the ES supply the excess of one ion polarity in solution re- emitter electrode, and the ES emitter electrode potential. quired to maintain the quasi-continuous production of In our most recent research efforts, we have developed a unipolar charged droplets and subsequently gas-phase porousflow-through(PFT)electrodeemitter[6,7],replac- ions. ing the standard capillary electrode emitter, to provide Our interest in this electrochemical process is very efficient mass transport of analytes in solution to the aimed towards better understanding the process to electrode even at flow rates approaching 1 mL/min. With devise means to control it for analytical advantage. this emitter electrode design all of the analyte in solution The electrochemical reactions at the emitter electrode will contact the surface of the PFT electrode on passage alter the composition of the solution being electros- through the emitter and very efficient oxidation or reduc- prayed, and they can also directly involve the ana- tion of analytes can be achieved as long as the reactions lytes being investigated. Thus, under certain condi- are not current limited, limited by the interfacial electrode tions, with particular types of analytes, the potential, or limited by other reaction rate considerations. electrochemical reactions in the ES ion source can A PFT electrode emitter enhances the ability to have a significant influence on the identity and directly involve the analytes under study in the electro- abundance of ions observed in an ES mass spectrum chemistry of the ES process. However, one would like [1,3,4].Inparticular,wehavebeeninterestedin to use this basic electrode configuration as a general ES controlling direct heterogeneous electron-transfer emitter so that it need not be replaced for experiments chemistry of the analytes under study and their in which analyte electrochemistry is not of interest or is potential homogeneous chemical follow up reactions. to be avoided. Control over which reactions can take place at the emitter electrode, and the rate at which they Published online May 12, 2006 take place, is achieved by controlling the interfacial Address reprint requests to Dr. V. Kertesz, Organic and Biological Mass potential of the electrode. This can be accomplished to Spectrometry Group, Chemical Sciences Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831-6131, USA. E-mail: some degree with a single electrode system by limiting [email protected] the magnitude of the current at the emitter electrode

© 2006 American Society for Mass Spectrometry. Published by Elsevier Inc. Received November 23, 2005 1044-0305/06/$32.00 Revised February 10, 2006 doi:10.1016/j.jasms.2006.02.007 Accepted February 10, 2006 954 KERTESZ AND BERKEL J Am Soc Mass Spectrom 2006, 17, 953–961 through adjustable parameters like solution conductivity or ES voltage drop. Lower current magnitudes either current limit the reaction (Faraday’s law) or lower the current density at the electrode so that the potential at the electrode drops to a level lower than that required for the analyte reaction (but still sufficient for another reaction, e.g., solvent electrochemistry, to provide the required current). Another means to control the potential with a single electrode emitter is through the use of redoxbuffers[8,9]. With a small surface area auxiliary electrode and a reference electrode in addition to the PFT electrode in the emitter cell, precise control of the PFT working electrode potential can be maintained with a potentiostat. We have demonstrated such a controlled-potential electrochemistry (CPE)-ES emitter operated with a Figure 1. Diagram of electrical circuit for the CCE-ES emitter. W floatedpotentiostatpoweredby120Vmainpower[10, and A represent the working and auxiliary electrodes in the

11].Cole’sgroupdiscussedaCPE-ESemitterofa electrochemical cell, respectively. IW, IAUX, IES, and IEXT are the different design controlled with a floated battery-pow- currents in the working electrode, auxiliary electrode, ES spray eredpotentiostat[12,13].Wefoundthatuseofasmall current, and upstream external current loop, respectively. Other parameters are defined in text. area auxiliary electrode was crucial in our emitter cell design because analyte reactions that were excluded at the PFT working electrode would occur at the auxiliary mode, powered by a battery floated at the ES high electrode unless mass transport to that electrode was voltage(Figures1and2).Thetwocircuitsweused hindered. By keeping the auxiliary electrode area small, for this controlled-current electrochemistry (CCE)-ES reactions of the dilute analyte species are kept to a emitter setup provided the ability to monitor the minimum at flow rates greater than about 30 ␮L/min current conditions at each electrode and thereby [10].Solventsandsolutionadditivesreacttosupplythe correlate these conditions with the collected mass required current at this electrode. spectra. Just as in the case of the previous two- Emitter working electrode potential control interme- electrodeESemittersystemsreported[14,15]andin diate to that of the single electrode emitter and to that of the case of our three-electrode emitter cell controlled the CPE-ES system can be obtained using a two-elec- byapotentiostat,[10,11],thisCCE-ESsystempro- trode emitter system and a very simple battery-pow- vides the ability to enhance the oxidation of analytes ered voltage or current supply. In this case, one of the in positive ion mode. Also, like our potentiostat two electrodes is used as the working electrode, while controlled three-electrode emitter cell, this CCE-ES the other acts simultaneously as the quasi-reference and emitter cell provides the ability to turn off analyte the auxiliary electrode. This configuration results in oxidation in positive ion mode or reduction in nega- limited control over the working electrode potential tive ion mode. Furthermore, one also has the ability because the potential of the quasi-reference electrode is to reduce analytes in positive ion mode and oxidize not well-defined. The potential of the latter electrode is analytes in negative ion mode. In the present case, a function of many parameters including the current this control was gained by controlling the magnitude density (defined by applied voltage or current and and polarity of the current at the large surface area electrode surface area), solution conductivity, solution working electrode rather than controlling the work- composition, etc. However, one statement is always ing electrode potential directly. The practicality and true regarding such a two-electrode cell: on either economics of such a battery-powered, two-electrode electrode a cathodic current results in reduction and an CCE-ES emitter system is further demonstrated with anodic current results in oxidation. We demonstrated a a simple circuit that includes just one resistor and a floated ES emitter system utilizing two-tubular stainless toggle switch, and is powered by a small-size 3 V electrodesinthemid-1990s[14]andmorerecently watch battery, all of which are, or could be made, Brajter-Tothandcoworkers[15]discussedaverysimi- small enough to fit on the body of the emitter cell. lar two-tubular electrode emitter cell. In each case, these simple battery-powered cells were used only to either enhance the mass spectral signal level of the molecular Experimental ion of analytes that could be ionized (oxidized) by Samples and Reagents electron-transfer or to follow the reaction path of analytes that could be oxidized in the emitter cell. Reserpine (Compound 1, Aldrich, Milwaukee, WI), In this paper, we use the same emitter cell with a methylene blue (Compound 2, Aldrich), and 3,4- PFT working electrode as in our previous CPE-ES dihydroxybenzoic acid (Compound 4, Aldrich) were studies[10,11],butoperateitinatwo-electrode obtained commercially and used without further pu- J Am Soc Mass Spectrom 2006, 17, 953–961 TWO-ELECTRODE EMITTER CELL 955 rification. Solutions were prepared using acetonitrile ES-MS (Burdick and Jackson, Muskegon, MI), water (Burdick Experiments were performed on a PE Sciex API 165 and Jackson), ammonium acetate (99.999%, Aldrich), single quadrupole mass spectrometer (MDS Sciex, Con- and acetic acid (PPB/Teflon grade, Aldrich). cord, Ontario, Canada). An HP 1090 Series II LC system (Hewlett-Packard, Palo Alto, CA) was used to deliver solvent and analyte solutions to the ion source. All data were acquired using a modified TurboIonSpray source incorporating a two-electrode controlled-current emitter cell configured for pneumatic nebulization. This emitter cell incorporating a porous carbon flow-through working electrode has been described in detail else- where[10,11].Forthiswork,thefoursmallpalladium electrodes on each side of the working electrode (two normally used as auxiliary electrodes and two as quasi- reference electrodes) were connected together in the circuit as the auxiliary electrode. The solution exited the cell and was sprayed through a 3.5 cm length of 50 ␮m i.d., 360 ␮m o.d. fused silica capillary with a Taper Tip (New Objective, Inc, Woburn, MA). To protect the working electrode emitter from plugging, a precolumn filter (0.5 ␮m frit) was placed upstream of the emitter cell in all experiments. The diagram of the battery-powered circuit used withtheCCE-ESemitterisshowninFigure1.The

current (IBAT) in the battery circuit (containing the working and auxiliary electrodes and the battery) was

defined by the resistor total resistance (RBAT) and the solution resistance (RSOL) between the two electrodes. RBAT was realized as a resistor chain built by serially connecting 100 k⍀ (5 pieces), 560 k⍀,1M⍀, and 2.2 M⍀ resistors. An 6LR61 type 9 V battery (Duracell, Bethel, CT) was used to power this extra circuit. In experiments using 3,4-dihydroxybenzoic acid, the negative pole of the battery was coupled to the auxiliary electrode, while in experiments using reserpine and methylene blue, the positive pole of the battery was coupled to the auxiliary

electrode. Changing the resistance in the working (RW) and auxiliary (RAUX) leg of the circuit, i.e., distributing RBAT between these two circuit legs, was accomplished by connecting the ES high voltage lead to different points of the resistance chain. The second circuit used to power the CCE-ES emitter utilized a CR2032 type 3 V watch battery (Rayovac, Madison, WI) and a single 390 k⍀resistor(Figure2).Thepolarityofthebatteryinthe

Figure 2. (a) Diagram of electrical circuit for the CCE-ES emitter usinga3Vwatch battery. W and A represent the working and auxiliary electrodes in the electrochemical cell,

respectively. IW, IAUX, IES, and IEXT are the currents in the working electrode, auxiliary electrode, ES spray current, and upstream external current loop, respectively. The total current ϭ ϩ at the emitter (ITOT IES IEXT) is divided between IW and IAUX.(b) Details on the battery circuit part of the system showing the polarity change switch and the 390 k⍀ resistor in the working electrode leg circuit. (c) Photograph of the circuit components used to build the 3 V watch battery powered circuit showing the relative size of the parts and the emitter cell. Other parameters are defined in text. 956 KERTESZ AND BERKEL J Am Soc Mass Spectrom 2006, 17, 953–961

circuit could be reversed using a double-pole, double- Table 2. Electrospray (IES), upstream loop (IEXT), working state, type 35-018 toggle switch (GC/Waldom, Inc., electrode (IW) and auxiliary electrode (IAUX) currents measured with different resistance applied in the working electrode leg Rockford,IL),asshownbythedottedarrowsinFigure ϭ ϩ ϭ (RW) of the CCE-ES emitter circuit. RBAT RW RAUX 2.06 2b.Toreduceshockhazard,thewholebatterycircuit M⍀. Data obtained using large volume injections (500 ␮L) of 5 was placed in a plexiglass box and the switch was ␮M reserpine at 50 ␮L/min in 50/50 (v/v) acetonitrile/water operated by an isolated toggle. containing 5.0 mM ammonium acetate and 0.75% by volume acetic acid. See the corresponding mass spectra in Figure 3b The current values in the working (IW) and auxiliary R (M⍀) I (␮A) R (M⍀) I (␮A) I (␮A) I (␮A) (IAUX) electrode circuit legs were simultaneously mea- W W AUX AUX EXT ES sured by inserting GB Instruments model GDT-11 mul- 0 2.9 2.06 4.8 Ϫ7.5 Ϫ0.2 timeters (Gardner Bender, Milwaukee, WI) into the 0.5 1.4 1.56 6.3 Ϫ7.5 Ϫ0.2 respective circuits. The current in the battery circuit 1.06 0 1 7.7 Ϫ7.5 Ϫ0.2 Ϫ Ϫ Ϫ (IBAT) when no high voltage was applied to the emitter 2.06 3.3 0 11.0 7.5 0.2 was also recorded. The ES current (IES) was measured on the mass spectrometer side of the circuit by ground- ing the curtain plate (normally 1.0 kV) of the mass spectrometer through a Keithley model 610C electrom- and the upstream ground point in the solution flow ϭ ϩ eter (Cleveland, OH) and lowering the emitter voltage stream (ITOT IES IEXT). These two current loops are by 1.0 kV. The spray capillary was moved laterally also present in the classic ES ion source when the beyond the sampling orifice so all of the charged emitter electrode is held at high voltage. droplets impacted the curtain plate. The current in the Direct current measurements showed that ITOT was divided between the current realized in the working external upstream current loop (IEXT) was measured by connecting the upstream ground of the ES ion source to electrode leg (IW) and that in the auxiliary electrode leg ground through the same electrometer with no other (IAUX) without implementing a battery in the circuit conditions altered. [11].Theparticularsplitof ITOT between IW and IAUX was a reflection of the distribution of RBAT (resistance in the resistor chain) between the resistance in the working Safety electrode leg (RW) and the resistance in the auxiliary ϭ ϩ All electrical circuit components that float at the ES electrode leg (RAUX), i.e., RBAT RW RAUX.By high voltage should be handled with extreme care and implementing a battery in the circuit, IW and IAUX were preferably isolated from the user with appropriate superimposed on IBAT (i.e., the current supplied as a shields and safety interlocks. The electrospray voltage result of the battery in the circuit). As a result, adding was set to 0 V before any changes were made in a the battery allowed the current distribution through the circuit to avoid a shock hazard. two legs of the cell circuit to be redistributed between the two electrodes. Note also that I ϩ I ϩ I ϩ I ϭ0(seeTables Results and Discussion EXT ES W AUX 1and2).Bymeasuring IBAT without high voltage ϭ ␮ ϭ ␮ Figure1showsaschematicoftheentireCCE-ESsetup applied (IEXT 0 A, IES 0 A), the solution including the battery, resistor chain, the mass spectrom- resistance (RSOL) between the two electrodes can be eter, and the various current paths. As defined here, the estimated on the basis of eq 1, if the charge-transfer resistance and the battery’s inner resistance (Ͻ1k⍀) total current, ITOT, at the electrodes making high voltage contact to solution in the ES emitter is the sum of the ES [16]arenegligible:

current, IES, in the current loop formed between the ϭ ϩ emitter electrode and the curtain plate (the mass spec- IBAT EBAT ⁄ ͑RBAT RSOL͒ (1) trometer counter electrode), and the external upstream

current, IEXT, in the current loop between the emitter With the electrolyte solution used in this study flowing ␮ through the system at 50 L/min, the measured IBAT ␮ ⍀ ϭ Table 1. Electrospray (I ), upstream loop (I ), working was 1.8 A when RBAT was 4.26 M using EBAT 9V. ES EXT Ϫ ␮ electrode (I ) and auxiliary electrode (I ) currents measured The measured IBAT was 3.3 A when RBAT was 2.06 W AUX ⍀ ϭϪ with different resistance applied in the working electrode leg M using EBAT 9 V (negative pole is connected to ϭ ϩ ϭ (RW) of the CCE-ES emitter circuit. RBAT RW RAUX 4.26 the working electrode leg). Using these values, 0.74 and ⍀ ␮ M . Data obtained using large volume injections (500 L) of 5 0.66 M⍀ (average: 0.70 M⍀) were calculated for R , ␮M reserpine at 50 ␮L/min in 50/50 (v/v) acetonitrile/water SOL containing 5.0 mM ammonium acetate and 0.75% by volume respectively. Considering that the current meters had ␮ acetic acid. See the corresponding mass spectra in Figure 3a 0.1 A resolution, the two calculated RSOL values were ⍀ ␮ ⍀ ␮ ␮ ␮ equal within experimental error. Knowing RSOL allows RW (M ) IW ( A) RAUX (M ) IAUX ( A) IEXT ( A) IES ( A) calculation of the highest current possible when the 0 5.3 4.26 2.4 Ϫ7.5 Ϫ0.2 ϭ ϩ total circuit resistance, RTOT RBAT RSOL, is minimal, 0.5 4.4 3.76 3.3 Ϫ7.5 Ϫ0.2 ϭ ⍀ Ϫ Ϫ i.e., RBAT 0M (no extra resistance in the circuit, e.g., 2.06 1.8 2.2 5.9 7.5 0.2 ⍀ϭ ␮ 4.26 Ϫ1.8 0 9.5 Ϫ7.5 Ϫ0.2 9 V/0.70 M 12.9 A). This current value can in turn be used to calculate the maximum analyte concentra- J Am Soc Mass Spectrom 2006, 17, 953–961 TWO-ELECTRODE EMITTER CELL 957

(a)

N CH3O N 1 H H H OCH3 H

CH3OOC OOC OCH3

OCH3 OCH3 (b) (c) (1 + H)+ m/z 609 (1 - H)- m/z 607 reserpine reserpine - + - + -2e , +H2O, -2H -2e , +H2O, -2H - - -H O -H O 2e 2e 2 m/z 607 2 m/z 625 +0.6 V m/z 623 m/z 605 - + - + - + - + -2e , -2H -2e , -2H -2e , -2H -2e , -2H

- - 4e 4e m/z 605 m/z 623 +1.0 V m/z 621 m/z 603 - + - + -2e , -2H -2e , -2H

- - 6e 6e m/z 621 > +1.0 V m/z 619 - + - + -2e , -2H -2e , -2H

- - 8e 8e m/z 619 > +1.0 V m/z 617

Scheme 1. (a) Reserpine (1) structure, proposed oxidation pathways, oxidation potentials and ions observed in (b) positive and (c) negative ion mode. tion that can be oxidized/reduced at a given flow rate.

The electrolysis current, I100%, required for complete analyte oxidation (or reduction) in the emitter cell can be calculated by using eq 2

ϭ I100% nFvc (2) where n is the number of electrons transferred, F is the Faraday constant (96,485 C/mol), c is the concentration of the analyte, and v is the volumetric flow rate.

Analyte Oxidation in Positive Ion Mode The normal operation of the ES ion source dictates that if the analyte is involved in the ion source electrochemistry it will be oxidized in positive ion mode (reduced in negativeionmode)[1].Controllingthecurrentmagni- tude and polarity at the large surface area working electrode of the two-electrode cell can be used to either enhance the extent of oxidation or turn off the reaction completely.Table1liststhemeasuredvaluesof IW, RAUX, IAUX, IEXT, and IES when RW was set at 0, 0.5, 2.06, ⍀ ϭ ⍀ and 4.26 M , with RBAT 4.26 M . Using eq 2, complete oxidation of a 5 ␮M solution of reserpine (1, ϩ (M ϩ H) ϭ m/z 609) flowing through the emitter cell at Figure 3. Positive ion mode CCE-ES mass spectra obtained at the Ϫ Ϫ Ϫ 50 ␮L/min, for reactions involving 2e ,4e ,6e ,or plateau region of a 500 ␮L injection of a 5 ␮M solution of reserpine Ϫ ϩ ϩ ϭ 8e , was determined to require oxidative currents of [(1 H) m/z 609] in 50/50 (vol/vol) water/acetonitrile, 5.0 mM ammonium acetate, 0.75% by volume acetic acid at a flow rate 0.80, 1.6, 2.4, and 3.2 ␮A, respectively (Scheme 1a). ␮ 50 L/min using the battery circuit with (a) RW set at 0, 0.5, 2.06, ⍀ ϭ ⍀ Figure3ashowsthereserpinemassspectraobtained and 4.26 M with RBAT 4.26 M and (b) RW set at 0, 0.5, 1.06, ␮ ⍀ ϭ ⍀ when a 5 M reserpine solution was injected into the and 2.06 M with RBAT 2.06M (seeTables1and2). 958 KERTESZ AND BERKEL J Am Soc Mass Spectrom 2006, 17, 953–961

(a) tion product (m/z 623) indicating a general increase in R R EXT ES the reserpine oxidation efficiency with increasing mag- W A A nitude of the oxidative current at the working electrode. ∩∩∩∩ IES = 0.2 µA IEXT = 7.5 µA When the whole resistance chain was in the auxiliary A I µ ϭ ⍀ ϭ ⍀ AUX = 2.4 A electrode leg (RW 0M , RAUX 4.26M ,Figure4a), Iw = 5.3 µA A +_ E = - 9 V an oxidative current was measured at both electrodes BAT ϭ ␮ ϭ ␮ (IW 5.3 A, IAUX 2.4 A). Because of mass transport R Ω AUX = 4.26 M issues, all the reserpine oxidation observed in the cor- ϭ ␮ _ respondingmassspectrum(Figure3a, IW 5.3 A) + ITOT = 7.7 µA A must have taken place at the working electrode. Some HV of the current at the working electrode (not enough reserpine to supply all the current, eq 2) and essentially all of the current at the auxiliary electrode was due to (b) R EXT RES oxidation of the solvent system. A W A Table2liststhemeasuredvaluesof IW, RAUX, IAUX, ∩∩∩∩ I = 0.2 µA IEXT = 7.5 µA ES IEXT, and IES values when RW was set at 0, 0.5, 1.06, and ⍀ ϭ ⍀ A IAUX = 9.5 µA 2.06 M , with RBAT 2.06M .Figure3bshowsthe Iw = -1.8 µA A _ corresponding reserpine mass spectra. Comparison of + EBAT = - 9 V IW valuesinTables2,andthecorrespondingreserpine

RW = 4.26 MΩ oxidationefficienciesapparentinthespectrainFigure ITOT = 7.7 µA _ 3brevealsagainthatthemagnitudeandpolarityof IW + A defines the reserpine oxidation efficiency, even if RBAT is changed. RBAT has an important effect by defining HV IBAT in this extra circuit as ITOT is practically defined by Figure 4. Detaileddiagramsoftheelectricalcircuitsusedfor thesolutioncomposition,[6,10,11] IBAT defines the ϭ recordingmassspectrainFigure3acorrespondingto(a) IW 5.3 current window within which I can be changed. This ␮ ϭ ⍀ ϭ ⍀ ϭϪ ␮ ϭ W A(RW 0M , RAUX 4.26 M ) and (b) IW 1.8 A(RW Ϫ ϩ ⍀ ϭ ⍀ observation is well demonstrated by the 1.8 to 5.3 4.26 M , RAUX 0M ). ␮ Ϫ ϩ ␮ A and 3.3 to 2.9 A ranges for IW accessible using ϭ ⍀ RBAT 4.26and2.06M ,respectively(Tables1and2). The selection of RBAT thus determines the current mag- system under the corresponding circuit conditions in nitude range within which the working electrode cur- Table1.Inspectionofthecurrentdatainthetableand rent can be tuned to achieve the desired efficiency of the corresponding mass spectra indicate that the extent analyteoxidationorreduction. of compound oxidation correlated with the magnitude of the oxidative current at the working electrode. These current and mass spectral data are best under- Analyte Reduction in Positive Ion Mode/Oxidation stoodwithreferencetothecircuitdiagraminFigure1 in Negative Ion Mode ϭ ⍀ andtothespecificcircuitsfor RW 0 and 4.26 M illustratedinFigure4aandb,respectively.Whenthe The basic CCE-ES emitter system presented here pro- wholeresistancechainwasintheworkingelectrodeleg vided an architecture that also made possible the effi- ϭ ⍀ ϭ ⍀ (RW 4.26M , RAUX 0M ,Figure4b),areductive cient reduction/oxidation of analytes in positive/nega- ϭϪ ␮ current was measured at this electrode (IW 1.8 A) tiveionmode,respectively.Scheme2showsthetwo- and no reserpine oxidation was observed, even though electron, one-proton reduction of methylene blue cation ␮ ϩ IAUX was 9.5 A. Because the current at the working [(2) , m/z 284] forming Compound 3, which was ob- electrode was reductive, no oxidation of the analyte servedastheprotonatedmoleculeat m/z 286.Figure5a could take place at this electrode. The current at this is the mass spectrum of methylene blue recorded with ϭ ⍀ ϭ ⍀ ϭϪ electrode was due to reduction of some other compo- RW 0M (RAUX 2.06 M ) and EBAT 9 V. Under ϭ ␮ nent(s) in the solvent system (i.e., solvent and/or sup- these settings, oxidative currents of IW 2.7 A and ϭ ␮ porting electrolyte). Inefficient analyte mass transport IAUX 5.0 A were measured and reduction of meth- to the small surface area auxiliary electrodes inhibited ylene blue was not observed as indicated by the lack of ϩ reserpine oxidation at those electrodes at this flow rate. signalat m/z 286[(3 ϩH) ]inFigure5a.However,with ϭ ⍀ ϭ ⍀ ϭϪ ␮ The current measured at the auxiliary electrodes was RW 2.06 M (RAUX 0M ) a reductive IW 3.3 A ϭ ␮ due to oxidation of the solvent system. When RW was and oxidative IAUX 11.0 A were measured. The stepped through the resistor chain from 4.26 to 0 M⍀, corresponding mass spectrum indicated efficient reduc- Ϫ ␮ ϩ IW increased from the reductive 1.8 A to the oxida- tionof(2) to 3 (Figure5b).Also,awell-resolvedpeak tive5.3␮A(Table1),andthebasepeakofthecorre- at m/z 285 was observed in the spectrum. This can be spondingmassspectrainFigure3ashiftedfrompro- tentatively assigned as the methylene blue radical ϩ tonated reserpine (m/z 609) through the two-electron cation intermediate [(2H)· ] formed by one-electron oxidation product (m/z 625) to the four-electron oxida- reduction[17]. J Am Soc Mass Spectrom 2006, 17, 953–961 TWO-ELECTRODE EMITTER CELL 959

N N E = -0.27 V CH3 CH3 CH3 - CH3 N S N + e N S N CH3 CH3 CH3 CH3 m/z 284 (2)+ (2)

+ + H

H H

N N E = -0.37 V CH CH3 CH CH3 3 - 3 N S N + e N S N CH3 CH3 CH3 CH3 + m/z 285 (3 + H) m/z 286 + 3 (2H)

Scheme 2. Methylene blue (2) structure, oxidation pathway, and ions observed.

Similar observations were made regarding analyte ␮A were measured. The base peak in the corresponding oxidation in negative ion mode. Scheme 3 shows the mass spectrum was observed at m/z 151 indicating very two-electron, two-proton oxidation of 3,4-dihydroxy- efficient oxidation of 4 to 5 at the working electrode benzoic acid (4) forming Compound 5, which was (Figure5d)duetotheoxidative IW. observed as the deprotonated molecule at m/z 151 [(5 Ϫ Ϫ H) ].Figure5cshowsthemassspectrumof 4 thatwas ϭ ⍀ ϭ ⍀ Watch Battery Circuit recorded by setting RW 0M (RAUX 2.06 M ) and ϭϩ using EBAT 9 V. Under these conditions, reductive To demonstrate the practical implementation, econom- ϭϪ ␮ ϭϪ ␮ currents of IW 2.6 A and IAUX 4.9 A were ics, and ease of use of a battery-powered CCE-ES measured. No oxidation product was observed in the emitter cell, we built a simple circuit that included only ⍀ spectrum in agreement with the measured reductive IW. a 390 k resistor, a toggle switch, and a small-size 3 V, ⍀ ϭ ⍀ However, when RW was set at 2.06 M (RAUX 0M ), CR2032typelithiumwatchbattery(Figure2).The ϭ ␮ ϭϪ an oxidative IW 3.4 A and reductive IAUX 10.9 switch enabled simple reversal of the polarity of the battery in the circuit resulting in either an oxidative or

+ - a reductive current at the working electrode. This (2) (4-H) 100 100 . (a) R = 0 M (c) R = 0 M circuit enabled a basic turn on/turn off control for W W I I =-2.6 A 2D Graph 1 =2.7 A 2D Graph 1 W analyte electrochemistry using our two-electrode 50 W 50 emitter cell. Rel. Abun Rel. Abun. 0 0 Figure6ashowsthenegativeionmodemassspec- 282 284 286 288 290 148 150 152 154 156 ␮ ϭ m/z m/z trum of 5 M reserpine obtained while applying EBAT Ϫ ϭϪ ␮ + - 3 V, which resulted in IW 9.8 A. The mass 2D Graph(3+H) 2 (5-H) 100 (b) R = 2.06 M 100 2D Graph 2R = 2.06 M W (d) W spectrum exhibited only one peak at m/z 607 that + (2H) Ϫ I = -3.3 A I =3.4 A Ϫ W W corresponded to deprotonated reserpine, (1 H) . 50 50 l. Abun. e Rel. Abun. 0 R 0 282 284 286 288 290 148 150 152 154 156 OH O m/z m/z OH O E = +0.3 V Figure 5. Positive ion mode CCE-ES mass spectra obtained at the + - + 2 H + 2 e plateau region of a 500 ␮L injection of a 5 ␮M solution of methylene blue (2) in 50/50 (vol/vol) water/acetonitrile, 5.0 mM A ammonium acetate, 0.75% by volume acetic acid at a flow rate 50 ␮ ϭ ⍀ COOH COOH L/min using the battery circuit (RBAT 2.06 M ) with RW set to - (a)0M⍀ and (b) 2.06 M⍀. Negative ion mode CCE-ES mass (4 - H)- = m/z 153 (5 - H) = m/z 151 spectra obtained at the plateau region of a 500 ␮L injection of a 20 ␮M solution of 3,4-dihydroxybenzoic acid (4) in 50/50 (vol/vol) 4 5 water/acetonitrile, 5.0 mM ammonium acetate at a flow rate 50 ␮ ϭ ⍀ L/min using the battery circuit (RBAT 2.06 M ) with RW set to Scheme 3. 3,4-Dihydroxybenzoic acid (4) structure, oxidation (c)0M⍀ and (d) 2.06 M⍀. pathway, and ions observed. 960 KERTESZ AND BERKEL J Am Soc Mass Spectrom 2006, 17, 953–961

100 (a) m/z 607 EBAT = -3 V (b) EBAT = +3 V 6 m/z 621 an easy upgrade to the currently used circuit, where a I = -9.8 µA I = +3.4 µA W 4 W switch enables placing the battery either in the working 50 2 or in the auxiliary electrode leg. Rel. Abun. Rel. Rel. Abun. 0 0 600 605 610 615 620 625 630 600 605 610 615 620 625 630 m/z m/z Conclusions m/z 607 starting material m/z 621 4 e- ox. product (c) (d) 100 6 The implementation of a battery-powered, controlled- 4 50 current two-electrode emitter system provides a simple, +3V +3V +3V -3V -3V -3V 2

Rel. Abun. economical means for control over analyte electrochem- 0 Abun. Rel. 0 02468 02468 ical processes that take place in the emitter of an ES ion t (min) t (min) source. This modification of a regular electrospray m/z - m/z 617 8 e- ox. product 619 6 e ox. product (e) 1.2 (f) 3 emitter introduces an extra current loop into the system 0.8 2 in addition to the upstream current loop (floated emit-

1 0.4 ter to upstream ground point loop) and the electrospray Rel. Abun. Rel. 0 Rel. Abun. 0.0 02468 02468 t (min) t (min)

Figure 6. Negative ion mode CCE-ES mass spectra obtained at (a) R R the plateau region of a 500 ␮L injection of a 5 ␮M solution of EXT ES Ϫ reserpine [(1 Ϫ H) ϭ m/z 607] in 50/50 (vol/vol) water/ A W A acetonitrile, 5.0 mM ammonium acetate at a flow rate 100 ␮L/min I = 0.2 A I = 7.3 A ES using (a) E ϭϪ3 V and (b) E ϭϩ3 V. Selected ion EXT BAT BAT A I monitoring ion current profiles for the (c) original reserpine AUX = 10.9 A Ϫ Ϫ I = 3.4 A A _ compound at m/z 607; (d)4e oxidation product at m/z 621; (e)6e w E Ϫ + BAT = 3 V oxidation product at m/z 619 and (f)8e oxidation product at m/z

617. RW = 390 k

_ + Figure6bshowsthemassspectrumofthesamesolu- A I = 7.5 A ϭϩ ϭ HV TOT tion when EBAT 3Vwasappliedresultingin IW ϩ3.4 ␮A. This mass spectrum exhibited several peaks (b) that corresponded in m/z to reserpine oxidation prod- R R ucts that could be observed in negative ion mode EXT ES A W A (Scheme 1c). The major peak in the spectrum was at m/z I Ϫ I ES = 0.2 A 621 and corresponding to the 4e oxidation product. EXT = 7.3 A A IAUX = 10.9 A The peak for reserpine anion (m/z 607) was just visible I A w = 3.4 A _ in the spectrum. Using eq 2, the complete 4-electron + E oxidation of 5 ␮M reserpine at 100 ␮L/min was calcu- BAT = 3 V R lated to require ϩ3.22 ␮A, which corresponds well with W = 390 k ϩ ␮ _ the measured value of IW ( 3.4 A) in this experiment. + This simple circuit was further tested by switching A ITOT = 7.5 A back and forth the polarity of EBAT while using single HV ion monitoring to observe the abundance of deproto- Ϫ Ϫ Ϫ nated reserpine and the 4e ,6e , and 8e oxidation (c) productsofreserpine(Figure6c–f).Theseioncurrent connection to the connection to the working electrode auxiliary electrodes profiles show reproducible multiple-electron reserpine Iw A A I ϭϩ E AUX oxidation using EBAT 3 V, and complete avoidance BAT of oxidation when the battery polarity was reversed. R Also, the response time of the system to adjust to the W changing applied EBAT voltage was observed to be relatively fast (Ͻ2 s). A I While the location of the battery in the circuit should TOT not affect the observed analyte electrochemistry, in HV practical applications the current drain on the battery Figure 7. (a) Detailed diagram of the electrical circuit used for shouldbeminimized.Figure7adescribesthedetailed ϭ recordingthemassspectruminFigure6bshowingthe IAUX 10.9 circuit used while recording the mass spectrum in ␮A current drain on the battery. (b) Detailed diagram of an Figure6b.Thecurrentthroughthebatterywas10.9␮A. alternative electrical circuit that could be used to record the same massspectrumasinFigure6bwithdecreasedcurrentdrainonthe By placing the battery in the working electrode leg (i.e., ϭ ␮ battery (IW 3.4 A). (c) Diagram of another alternative electrical the high voltage connected to the auxiliary electrodes as circuit with an additional double throw switch (positions indi- showninFigure7b),thecurrentdrainonthebattery cated by the dotted arrows) to switch the battery between either wouldhavebeendecreasedto3.4␮A.Figure7cshows the working or the auxiliary electrode leg. J Am Soc Mass Spectrom 2006, 17, 953–961 TWO-ELECTRODE EMITTER CELL 961 current loop (floated emitter to mass spectrometer ORNL was supported by the Division of Chemical Sciences, Geosciences, and Biosciences, Office of Basic Energy Sciences, counter electrode loop). The total emitter current (ITOT) divided between the working and auxiliary electrodes United States Department of Energy under contract no. DE- AC05-00OR22725 with ORNL, managed and operated by UT- of the emitter cell can be manipulated by controlling the Battelle, LLC. resistance between the ES high voltage lead and the This manuscript has been authored by a contractor of the corresponding electrodes. The measured working (IW) U.S. Government under contract no. DE-AC05-00OR22725. Ac- and auxiliary (IAUX) electrode currents are the sum of cordingly, the U.S. Government retains a paid-up, nonexclu- these individual partial inherent currents superimposed sive, irrevocable, worldwide license to publish or reproduce the published form of this contribution, prepare derivative works, on IBAT, the current produced by the battery. Effective analyte electrochemistry takes place only at the high distribute copies to the public, and perform publicly and display publicly, or allow others to do so, for U.S. Government surface area porous flow-through working electrode purposes. regardless of the current magnitude and polarity at the auxiliary electrode, because of the small total surface area of the four linked auxiliary electrodes. Thus, analyte oxidation or reduction with this cell was accom- References plished by controlling the polarity and magnitude of 1. Van Berkel, G. J. The Electrolytic Nature of Electrospray, Chap. II. In the current at the working electrode. Electrospray Ionization Mass Spectrometry; Cole, R. B., Ed.; Wiley: New York, 1997; pp 65–105. The analytical utility of the CCE-ES emitter was 2. De la Mora, J. F.; Van Berkel, G. J.; Enke, C. G.; Cole, R. B.; Martinez- Sanchez, M.; Fenn, J. B. Electrochemical Processes in Electrospray illustrated by demonstrating enhanced analyte oxida- Ionization Mass Spectrometry. J. Mass Spectrom. 2000, 35, 939–952. tion in positive ion mode (enhance reduction in nega- 3. Van Berkel, G. J. Insights into Analyte Electrolysis in an Electrospray Emitter from Chronopotentiometry Experiments and Mass Transport tive ion mode should also be possible), the ability to Calculations. J. Am. Soc. Mass Spectrom. 2000, 11, 951–960. turn off analyte oxidation in positive ion mode (or 4. Diehl, G.; Karst, U. On-Line Electrochemistry-MS and Related Tech- niques. Anal. Bioanal. Chem. 2002, 373, 390–398. reduction in negative ion mode), and the ability to 5. Wang, J. Analytical Electrochemistry, 2nd ed.; Wiley-VCH: New York, reduce analytes in positive ion mode and oxidize NY, 2000. 6. Van Berkel, G. J.; Kertesz, V.; Ford, M. J.; Granger, M. C. Efficient analytes in negative ion mode. Thus, a simple and Analyte Oxidation in an Electrospray Ion Source Using a Porous inexpensive battery and resistor circuit used with a Flow-Through Electrode Emitter. J. Am. Soc. Mass Spectrom. 2004, 15, 1755–1766. two-electrode emitter cell provided the same basic 7. Van Berkel, G. J., Kertesz, V. Expanded Electrochemical Capabilities of electrochemical capabilities as a three-electrode emitter the Electrospray Ion Source using Porous Flow-Through Electrodes as the Upstream Ground and Emitter High Voltage Contact. Anal. Chem., cell controlled by a sophisticated floated potentiostat in press. [10,11].Moreover,sincethecurrentrequirementsofa 8. Van Berkel, G. J.; Zhou, F.; Aronson, J. T. Changes in Bulk Solution pH Caused by the Inherent Controlled-Current Electrolytic Process of an typical floated emitter ES system with an upstream Electrospray Ion Source. Int. J. Mass Spectrom. Ion Processes 1997, 162, ground contact are 10 ␮A or less, a small-size watch 55–67. 9. Van Berkel, G. J.; Kertesz, V. Redox Buffering in an Electrospray Ion battery with, e.g., 200 mAh capacity could power the Source Using a Copper Capillary Emitter. J. Mass Spectrom. 2001, 36, controlled-current cell in continuous use for over a year. 1125–1132. 10. Van Berkel, G. J.; Asano, K. G.; Granger, M. C. Controlling Analyte By building and using a simple controlled-current cir- Electrochemistry in an Electrospray Ion Source with a Three-Electrode cuit that included just a single resistor, polarity toggle Emitter Cell. Anal. Chem. 2004, 76, 1493–1499. 11. Kertesz, V.; Van Berkel, G. J.; Granger, M. C. Study and Application of switch, and a small-size 3 V watch battery, we were also a Controlled-Potential Electrochemistry-Electrospray Emitter for Elec- trospray Mass Spectrometry. Anal. Chem. 2005, 77, 4366–4373. able to demonstrate that the possibility exists to include 12. Xu, X.; Lu, W.; Cole, R. B. On-Line Probe for Fast Electrochemistry/ all the CCE-ES emitter electronics on the body of the Electrospray Mass Spectrometry. Investigations of Polycyclic Aromatic Hydrocarbons. Anal. Chem. 1996, 68, 4244–4253. emitter cell. 13. Lu, W.; Xu. X.; Cole, R. B. On-Line Linear Sweep Voltammetry- Electrospray Mass Spectrometry. Anal. Chem. 1997, 69, 2478–2484. 14. Zhou, F.; Van Berkel, G. J. Electrochemistry Combined On-Line with Acknowledgments Electrospray Mass Spectrometry. Anal. Chem. 1995, 67, 3643–3649. 15. Zhang, T.; Palii, S. P.; Eyler, J. R.; Brajter-Toth, A. Enhancement of ES-MS instrumentation and electrochemical cell and associated Ionization Efficiency by Electrochemical Reaction Products in On-Line Electrochemistry/Electrospray Ionization Fourier Transform Ion Cyclo- components, respectively, were provided through a Coopera- tron Resonance Mass Spectrometry. Anal. Chem. 2002, 74, 1097–1103. tive Research and Development Agreement with MDS SCIEX 16.http://www.duracell.com and ESA Biosciences, Inc. (CRADA no. ORNL02-0662). VK 17. Nicolai, S. H. D.; Rodrigues, P. R. P.; Agostinho, S. M. L.; Rubim, J. C. Electrochemical and Spectroelectrochemical (SERS) Studies of the Re- acknowledges an ORNL appointment through the ORNL Post- duction of Methylene Blue on a Silver Electrode. J. Electroanal. Chem. doctoral Research Associates Program. The work carried out at 2002, 527(1/2), 103–111.