Cyclic Voltammetry Experiment

James J. Van Benschoten. J a n e Y. Lewis, a n d William R. H e i n e m a n

University of Cincinnati. Cincinnati, OH 45221 Daryl A. Roston' and P e t e r T. Klsslnger Purdue University. West Lafayene. IN 47907 Cyclic voltammetry (CV) is a very useful modern electroanalytical technique; its principles and theories have been discussed by Kissinger and Heineman (I). Despite t h e advantages of the technique a n d its widespread use, there is a lack of CV experiments in t h e literature for students. Thus, this experiment has been designed specifically for introducing CV to students who are a t t h e advanced undergraduate or earlv -- " maduate level. There are three oarts to this exueriment, each of which may be performed independently. P a r t I uses t h e Fel1I(CN)c3-/Fe11(CN)s4-couule a s a n example of a well-behaveb e ~ e i t r o c h e m i c areversible ~i~ system in aoueous solution. This exoeriment demonstrates determinition of t h e following: t h e formal reduction potential (E"'): the number of electrons transferred in the redox process in);the diffusion coefficient (D); electrochemical reveriibility; and t h e effects of varying concentration and scan rate. P a r t I1 investigates t h e electrochemical behavior of acetaminophen, a n ingredient in many over-the-counter pharmaceuticals (e.g., Tylenol*, Excedrin*, Nyquil*, etc.). T h i s organic system is more complex, involving chemical a s well a s electrochemical reactions. Principles t h a t are illustrated include: effect of a coupled chemical reaction on t h e appearance of t h e CV; effect of p H on the appearance of the CV and o n t h e mechanism of reaction; use of scan rate for elucidating mechanistic information; and quantitative determinationof acetaminophen in a l'ylenolm tnhlet. P a r t 111-is designedto introduce t h e student t o practical considerations in a cyclic voltammetry experiment. Instead of followine recommended conditions t h e student now is persuaded t o investigate factors ~ u c h as: t h e effect of 0 2 on the C'Y: choosinr the aoorooriate .. . electrode: ~srablishinaPOtential $can lim& and studying Pt electrode surface waves. It is honed that. uoon comoletion of this section, t h e student can effktively design a n eiectrochemical experiment for a n y particular chemical system of interest. recorder, and an electrochemical cell (I). The potentiostat used is a Bioanalyticsl Systems, Inc. CV-1R. Alternatively, any number of commercially available instruments, or a homebuilt model constructed from conventional operational amplifiers can be used ( 3 . 4 ) . The current versus potential curves are recorded on a HewlettPackard model 7015B XY recorder, As mentioned in reference ( 1 ), the ootential scan varies linearlv with time. thus it is feasible to use a rrrip chart recorder, even tho&h 11 is not ns wnvenienr as an X ) ' . I f an <,rcill~mcope ia nvailahlr, i t can be used in place of the rerorder. The p~trntial output i s fi~lhwed independently wwlth n Keithley model 178 digital voltmeter. A three-electrode system is used for the CV experiment (I). The electrodes reauired are a Pt warkinc electrode (area = 2.5 mm2),a Pt auxiliary electrode, and either a A ~ I A ~ C orI a saturated calomel electrode (SCEI reference electrode. All ootentials in this exoeriment are referenced veraur. %'E The cell used la a H~mnalytical Systems. In,. \'('-2 ylaw wal with a t i t m l top. The u,p h a . , iour hdes machmed to nerommudate the three rlectn~des and a t u h for detrxygenating by bubbling with a stream of Ns. Procedure. Pretreatment of the platinum working electrode surface may be required. Simply polishing the surface with powdered alumina and rinsing thoroughly with distilled water should suffice. The electrode can be sonicated in an ultrasonic bath if available. The cell, as shown in reference ( I ) , is assembled and filled with 1 M KNOJ just until the ends of the electrodes are immersed (Note: avoid tilting the SCE; this may disturb the HgIHgnCln and result in incorrect potentialsl. The system is deoxygenated by purging with Nz for approximately 10 mi". Following this, N2 is allowed to flow over the solution to prevent 0 2 from re-entering the eell for the remainder of the experiment. While the system is beingdeoxygenated, the scan parameters can be set. The working electrode should be disconnected or switched off during this procedure. The initial potential is set at 0.80 V, and the scan limits at 0.80 V and -0.12 V using the recorder as a monitor. All scans are initiated in the negative direction with a scan rate of 20 mV/s. These settings will be used unless otherwise specified. When deoxygenation is complete, the working electrade is switched on. After allowing the current to attainaeonstant value (in about 10 s), the potential sean is initiated and a background CV of the supporting electrolyte solution is obtained. After turning off the working electrode, the eell is cleaned and refilled with 4 mM KaFe(CNI6 in 1M KNOs. Following the same procedure as above, a CV of the Fe'1L(CN)i"/Fe1'(CN)8- couple is obtained. The effect of the sweep rate ( u ) on the voltammoprams is observed by using this same soiutionand recording CV's at the following rates: 20,50,75,100,125,150,115,and 200mV/s. Retweeneach scan, initial conditions at the electrodesurface are restored by moving the working electrode gently up and down without actually removing it from solution. Care should be taken that no buhbles remain on the electrodes. ('wwenrration likewiseaffrcts the magnrtudr of thr penkcurrent. This i seen by uhtainmg +cansof 2.6.8. and I d m A f K ,l.'t.tC'Fi~fi uang nswwo rate uf 20 mV,s A \r,ltarnmoeram of tne unkn<,wnK8nCN)a solution should be ru'n as well. The affect of the supporting electrolyte on the appearance of the CV isdemonstrated by recordingvoltammogramsof (1) 4 mM ferricyanide in 1 M KNOs and (2) 4 mM ferricyanide in 1M Na2S01.
~

.~

~~~~

~~

Part I-Fundamentals of Cyclic Voltammetry T h e Fe111(CN)6"-/Fe"(CN)$L- couple is known t o be wellbehaved both chemically and electrochemically (2). As such, i t is often used a s a model system in electrochemical experiments. I t can be used t o determine electrode areas a n d to diagnose problems associated with new electrochemical cell designs. In this experiment, t h e couple is used t o clearly demonstrate s o m e ~ i m p o r t a n tprinciples of cyclic voltammetry. Experimental Reagents. All chemicals are reagent grade and are used assuch. A 10O-mlstock solution of apprnximateiy 10 mM K3Fe(CNk is prepared in 1 M KN03. From this, 25 ml each of approximately 2,4,6, and 8 mM solutions in 1 M KN03 are made. In addition, a 25-ml K3Fe(CN)6 solution of approximately 4 mM is prepared in 1 M NazS04.All solution concentrations must be accurately known. An unknown KsFe(CN)fisolution should be provided by the instructor. Apparatus. The cyclic voltammetry apparatus consists of a potentiastat with potential sean capability, a voltmeter (optional), a Current address: Department of Chemistry. Northern Illinois University. DeKalb. I L 60115.

~~~

~~

.~

~~~

Results and Discussion Using the data obtained, a number of simple calculations can he done that will yield considerable information about the ferri-ferrocyanide system. Anodic peak current ( i , ) cathodic peak current (iw), E"', and n can be determined as described

772

Journal of Chemical Education

"

0.8 04 00 WTENTIAL V vr S C E

;-loL

2.0 4.0 6.0 6.0 CONCENTRATION mM

'

100

POTMTIAL. V rr 5 C E in 1 MKNOJ.Pi elecFigure 1. (A) Cyclic voltammograms of 4 mMKsFe(CN)s bade. Scan rate ( 4 = 20. 50. 75. 100. 125. 150. 175. end 200 mVls. (8)Plot 01 iw YBTSUI v ' ' ~and ,1 versus uU2 horn voltammogams in (A).

F i a r e 2. IAl . . Cvcllc . voltammmam of KIF~CN). . .. In 1 MKNO*.Scan rate = 20 mVls. PL elechade. Concentration = 2. 4, 6. 8, and 10 m M . (8)Plot o f i, versus concentration and i, vsrsus concentration horn voltammograms in (A).

in reference ( I ) using the voltammogram of 4 m M K3Fe"'(CN)6 in 1M KN03. The effect of the scan rate, v, on the appearance of the CV can be seen in Figure 1A. As described by the Randles-Sevcik equation (I), ,i and ip, i n c r e m as u"~. A plot of this equation (Fig. 1B) yields a straight line, the slope of which can be used to determine the diffusion coefficient (D in cm2/s) of the system. T h e area of the working electrode will either have to be provided by the instructor or experimentally determined. Evidence indicating that the system is electrochemically reversible (discussed in reference ( I ) ) can he seen in Figure 1A as well. T h e separation of the peak potentials (AEJ is i n d e ~ e n d e nof t v. This can he discussed bv the student. , and The Kandlrs-Sevcik equation ( 1 ) also indicates that i ...i are d i r e d v oro~urtional to concentration. The effect of .. . increasing concentration on the appearance of the CV can be seen in Figure 2A. A calibration curve (Fig. 2B) can be constructed and used to determine the concentration of the unknown ferricvanide solution. A formal potential as reported in the literature is a working number that is dependent on the conditions of the experiment (5).T h e effect ofthe particular supporting electrolyte used is demonstrated by comparing CV's of K3Fen1(CN)6in KN03 and Na2SOa. Part ll-Effect of Coupled Chemlcal Reactions Acetaminophen (N-acetyl-p-aminophenol, APAP), the active ineredient in Tvlenol". is commonlv used as an a s ~ i r i n substitute. However, unlike aspirin, it is known to cause liver and kidnev damaee when administered in laree amounts. I t is suspectld that :metabolite of APAP is the actual hepatotoxic agent, thus APAP and its metabolites have been extensively investigated (6). Voltammetric studies in aqueous solution have revealed chemical as well as electrochemical steps (7). The APAP system therefore is useful in demonstrating the mechanistic information that can he obtained from CV's.

tions should span the range of 0.10 to 5.0 m M ; 10-25 ml of each are needed. An unknown solution is prepared by dissolving a Tylenolm tablet in 250 ml of pH 2.2 buffer. A workable concentration is obtained by diluting a 5-ml aliquot of this solution to 50 ml. All APAP solution concentrations need to be accurately known. Apparatus. The apparatus used in this experiment is identical to that in Part I with the exception of the carbon paste workii electrode (area = 9.1 mm2). Procedure. The carbon oaste electrode is ~renared . . as described in the Hiuanalytirnl Sysrems. Inc. manual. Thw elrrtrc,dt must he carefully parked tu avoid erroneous resulu and rare ;lhould c , e mken not to gauge it once it is ready for use. Using the 3 m M APAP solution in pH 2.2 buffer, scan limits are established at 1.0 V and -0.2 V. Scans are initiated in the positive direction at 0.0 V. Cvclic voltammomams of the 3 mM APAP solution in each of the 3 buffers are then obcined at a scan rate of 40 mV/s and 250 mV1s. The solution should be stirred between each run. As was demonstrated in Part I, CV's can be used for quantitative analysis. A standard curve can be constructed using the series of 5 APAP solutions in pH 2.2 buffer. From this, the concentration of APAP in the Tylenol solution is determined.
~ ~~~ ~ ~

Results and Discussion The oxidation mechanism of APAP is as follows

OH
I APAP

1 1

111

NAPQI

Experimental Reagents. All chemicals are reagent grade and are used as such. The fallowing supporting electrolyte solutions are needed: 500 ml of pH 2.2 and 200 mlof DH6 Mcllvaine bufferwith an ionicstrenethof 0.6 M (8); 200 ml of i.8 M HSSO~. A stock solution of approximately 0.070MAPAP is prepared in0.05 M perehlaric acid. This should be kept in the refrigerator when not in use. Using the stock solution, 10-25 ml of an APAP solution is prepared in each of the three supporting electrolyte solutions. The concentration of these APAP solutions should be approximately 3
~ ~

mM.
For the purpose of establishing a calibration curve, four additional APAP solutions in pH 2.2 buffer are needed. These four concentra-

APAP is electrochemically oxidized in a pH-dependent, 2electron. 2-proton Drocess to N-acetvl-D-~uinoneimine . . . (NAPQI') (step 1). ~ i occurrence e of follow-up chemical reactions involving NAPQl pH dependent. Bv the pH . is . . varving . . of the media and the scan rate ofthe cyclic voltammetry experiment, chemical reactions involving NAPQI can be "mapped-out." At p H values 26, NAPQI exists in the stable unprotonated form (11). Cyclic voltammograms recorded for N A P a t pH 6 are shown in Figure 3. Reasonably well-defined anodic and cathodic waves are evident. The anodic current represents step 1in the mechanism detailed above while the Volume 60 Number 9 September 1983

773

E,VOLTS vr W e p c l

roe

//

70.4

r0.z

Figure 3. Cyclic voltammogramr of 3.6 mM APAP in pH 6 Mcllvsine buffer. Carbon paste electrode.

cathodic current represents the reverse of this step. The similarity in appearance of the p H 6 cyclic voltamm~grams observed with 40 and 250 mV/s scan rates indicates that the involved species are stable in the time domain of the cyclic voltammetrv experiment. The large separation between the anodic andcathodic peak currents &the pH 6 cyclic voltammograms is a manifestation of sluggish electron transfer kinetics. Under more acidic conditions, NAPQI is immediately protonated (step 21, yieldmg a less itablebut elrctn,chernicalls artivr species (1111ahirh rapidlv yields (step ;I) a hyIYr inactiw at the exclrnted firm - I, , that is elertroch~~micallv amined potentials. Cyclic voltammograms shown in Figure 4 are consistent with this mechanism. The DH of the media was 2.2. A small cathodic wave due to the reduction of protonated NAPQI " (111) . . is evident when the scan rate of 250 mV/s is employed. This wave is even more pronounced when faster scan rates are emnloved: however. faster scan rates reauire the . " . use of an oscilloscope to record thk current-potential rksponse. With a slower scan rate of 40 mV/s, a cathodic wave for the reduction of protonated NAPQI is not observed. All of the protonated NAPQI (111) is converted to the inactive hydrated form (IV) before negative enough potentials are reached durinz the reverse scan of the cvclic voltammetrv exoeri. . . -. . . . Hvdrated NAPQI (IV) converts (step 4) to benzoquinone; however, the media has to he extreme& acidic for thk rate of the process to be sienificant enough that reduction of benzoqui&ne is observed during the cyclic voltammetry experiment. The media for the cyclic voltammograms detailed in Figure 5 was 1.8 M HzS04. A poorly defined cathodic wave for the reduction of benzoquinone (V) was observed when a scan rate of 250 mV/s was employed. The reduction wave is broad because the formation of benzoquinone (V) from hydrated NAPQI (IV) occurs during the reverse scan. When the scan rate is 40 mV/s, the increased length of time required to reach negative enough potentials during the reverse scan allows for the accumulation of benzoquinone (V). Consequently, a well-defined reduction wave was observed for henzoouinone ( V ) when the slower scan raw wnsernployrd.The sermd scan anodic wave in addition to in thr nositive direction virld3 i u ~ that o E APAP which coriesponds to the oxidation of hydroquinone, the reduction product of benzoquinone.
~~~~~~ ~ ~ ~ ~ ~ ~

Figure 4. Cyclic voltammcqramsof 3.6 mMAPAP in pH 2.2 Mcllvaine buffer.

Carbon paste electrode.

Figwe 5. Cyclic w l i a m w m m s of 3.6 mM APAP in 1.8 M W . . C m b m psste electrode.

determining whether 0 2 is an interference. Generally, this is accomplished by some amount of trial and error. This effort is justified, however, because as shown in parts I and 11, considerable information can be obtained from a relatively simple experiment. This section demonstrates how these selections are made, thus giving the student insight on how todesign an effirirnt C V experimrnt for almost anv chemical systtm of interest. Experimental Reagents. All chemicals are reagent grade and require no further and 25 purification. Solutions of approximately 25 ml1.0 M H2SO& ml 1.0 M KNOuare required. A~oarotus. The electrochemical aooaratus is described in Dart I

Part Ill-Dissolved Oxygen, Potential Limlts, Surface Elfects When investigating a chemical system for the first time with cyclic voltammetry, there are several experimental conditions that need to be established. Among these are choosing the appropriate electrode, solvent, and supporting electrolyte; and 774 Journal of Chemical Education

O1 REMOVED

0.4

0.0

-0.4

-0.8

-1.2

POTENTIAL, V vs SCE Figure 6. Cyclic volternmqlrams ol 1.0 MKWsshowing effect of OZ. Oxygen rernovd by deaxygenating wilh N2 f n 10 min. Au electrode. Scan rate = 50
rnV/s.

v '

, 1.2

0.9

0.6

0.3

00

-0.3

POTENTIAL, V vs SCE

Figure 8. Cyclic valtarnmqlram showing R eisctrDMt surface waves in 1 M H2S0,. Scan rate = 50 rnV/s. Plntinum makw a very interwing elcctrde i n that several surface wnves npprar in a 1,ackgmund rean. Thmr wavra ran he investigated in a dwrygmated solution of 1.0 M HJOI wing a pdirhed electnde. A voltammogram is obtained by initiating the scan in the negative direction at 0.30 V with potential limits of 1.425V and -0.225 V. The positive potential limit is progressively stepped in to 0.825 V (in 0.15 V increments) obtainine a CV at eaeh new limit. Leavine this limit at 0.825 V, the negativepotential limit is then likewisestepped in to 0.075 V (in 0.075 V increments) obtaining a CV at each new setting.
~ ~

~~

~~~~~~~

.
1.6

00

0.0 -08 POTPITIAL, Y vr SCE

-1.6

Frg~e 7

bmes
01

Cyclic voltarnmDgrams o f 1 0 M <NOs bs.ng d tlerenl wommg slsc(Alglassycarbon. IBicaman pase. ICIAU.ID1 Fi Cycilc vohmmogsm 1 0 MH,SO. mng IE) Au electrode Scan rate = 50 mVls

9.1 mm2),and an Au electrode (area = 2.5 mm2)are used. Any or all of these electrodes may he investigated depending on availability. Procedure. The effect andlor interference of dissolved 01 " in the supporting electrolyte is readily determined. Using the Au working electrode (polished as was the Pt electrode in part I) acyclicvoltammogram of 1.0 M KN03 with scan limits of 0.7 V and -1.0 V is run. This is initiated at 0.40 V in the negative direction with a scan rate of 50 mV1s. The solution is then deaxygenated for 15 min as described in part I, and a CV of the oxygen-free solution is recorded. The potential window for variouselectrodes can be just as readily determined. At the low current settines .. tvoicallv . , used in CV.. this is generally l~m~red m either end hg large offrcnlr wnwz ['ring n 1.0 M K N h sdutim, n scan rntr oi 50 my s,and a current range of approxmately hll pA in.. the worknhlr pute~,tialrange can be determined for eaeh electrode. Not only is the potential range dependent on the particular electrode used, but alsoon the suppurtingelectrolyteused. This isdemonstrated by comparing the potential limits available on an Au working electrode in 1.0 M KN03 and 1.0 M H2S04.
~~

Results and Discussion Though a compound is not chemically altered by 0 2 (i.e., oxidized, degraded, etc.), the voltammogram of that compound may nonetheless be subject to 0 2 interference. This results from the fact that 0. is electrochemicallv active (i.e.. These two process"es should h i reducible to Hz02 and HZO~. amarent from the CV's of oxveenated 1.0 M KNO* (see Fie. f3j.'1f the compound of i n t e r e z e d u c e s in the samepktential region. . deoxveenation is a necessarv -~rocedure. . Choosing the appropriate solvent, supporting electrolyte, and working electrode for a CV e x ~ e r i m e nreauires t several rtmsiderati(;n.s.The major requirements of the xivent are that the iveries of interest he soluble and stahle in it. The SUI>porting electrolyte must provide a relatively high ionic strength (>0.01 M ) and be compatible with the svstem. A further consideration is the window that is available. In aqueous solution, this is often limited on the negative end by the reduction of HzO to H2 and on the positive end by the oxidation of Hz0 to 0 2 . It is influenced by the choice of working electrode, solvent, and supporting electrolyte. Electrolysis of the species of interest must occur within this range for a CV experiment. Figure 7 (A-D) shows the potential windows provided by several working electrodes in 1M KNOs. Figure 7C and 7E demonstrate the influence that two different supporting electrolytes, 1 M KN03 and 1.0 M H2S04 respectively, can have. These illustrate the pH dependence of the Hz0 electrolysis waves. Metals such as Au and Pt are not totallv inert as once thought ( 9 , l O ) .When used as electrodes, ~ t a n d exhibit ~ u surface waves which limit the usable votential ranee. At an Au electnde (as shown in Fig. 7 E ) these waves are atrributed to adiorhed oxide furmation and redurtion at low pH. Figure H illustrates the numerous l't surface waves. At negative potentials, fmnation and oxidation ol adsurhed Hv UCCUW: in the positive region, formation and reduction of adsorbed oxide
Volume 60

Number 9

September 1983

775

is observed. These waves have been studied and can be used to determine accurately electrode area (9, 10). As demonstrated in Figure 8, if the potential range is sufficientlylimited, a very clean workable background can be obtained. Literature Cited
11) Kissinwr. P. T.. Heineman, W. R.. 3. CHEM.EDUC..EO, 702 119831. (21 DeAn~e1is.T. P., Heineman, W. R., J. CHBM. BDVC..53.94 11976l. (3) Schwan. W. M.,Shain, I., A n d Chem..35.177011963).

14) Kimingcr, P. T., in "Labcrstory T ~ h n i q u e a in E l e c t r 4 y f i c a l ChemhV/."KYsbger. P. T, and Heineman. W. K. IEditomI Marrel Dakkar, New York, in pma. Chapt.
E.

(51 ~ d t h o f fI. . ~ . . ~ n m a i e cW k ,. J . J ~ h y chem. s 39.945 (1935). (fil Miner. I). .I.. Rice. J. H..Rimin. R. M.. Kiasin~or.P. T.. Aml. Chsm.. 53. 2258 119811. 171 Preddy, C. A . Miner. D. .I., Meinsma. D. A . Kiasinpr, P. T., Anal. Chem., sub(8) Riving. P. J., Markoulitz. J. M.. Rorenthal. I . . A m i . Chm~.28.1179(1956). (91 T.. ~ n l x r t a JX..I. . ~ . . - ~ ~ ~~ ~ l ~i t ~ m ~h ~ t ~forchemista." ri e s tl ~ widey. Now York. 1914. pg. -9. I101 Bsuer. H. H.. "Electrdie: Modern Ideas Concerninr E l r d r d o Reactions." Gwrg Thismr l'uhl~sherr,StutOa~.Gsrmany. 1972. ~ p l W . 103.

.....".
-;*.-A

sayer, n.

776

Journal of Chemical Education