Fresenius J Anal Chem (1996) 356: 476—479 ( Springer-Verlag 1996

ORIGINAL PAPER

Pedro Navarro · Claude Jambon · Olivier Vittori

Electrochemical response of a carbon paste electrodecontaining an adsorbed species on incorporated alumina

Received: 20 November 1995/Accepted: 6 January 1996

Abstract The electrochemical response of carbon pasteelectrode was investigated containing an inorganic anda non-electroactive compound on which an electro-active organic molecule was adsorbed. Alizarin S waschosen as the test molecule adsorbed on 150 m2/g ofalumina, because this molecule is either easily reducedor oxidized on carbon. The electrode response wasfound to be linear for low amounts of alumina contain-ing a known amount of Alizarin S. The electrical yieldwas never higher than 50%, but was reproducible. Anincreasing alumina amount in the paste increased theresponse, but lowered the electrical yield.

Introduction

Carbon paste electrodes were first introduced byAdams [1,2] for the electrochemical study of organiccompounds insoluble in water. Because of its efficiencythe electrode was further widely used to investigateeither organic or inorganic compounds [3—19]. A re-cent review by Kalcher [20] with 173 references hasbeen devoted to all aspects of the response of thecarbon paste electrode when modified with a largevariety of reagents. In most cases, the carbon paste ismade of pure graphite powder and an organic bindersuch as nujol, paraffin, silicone oil etc. These electrodesare used after a preconcentration step and the electro-chemical analysis is performed by cyclic voltammetry.Less work has been devoted to aqueous conductingbinders except those from several French and Spanishlaboratories [3—15]. In the latter case the solid powder

P. Navarro · C. Jambon · O. Vittori ( )Laboratoire d’Electrochimie Analytique, LICAS Universite Lyon I,CPE Lyon, 43 Bd du 11 novembre 1918, F-69622 VilleurbanneCedex, France

under study is mixed with an ultra-pure graphite pow-der and an aqueous electrolyte in order to obtaina paste. The main advantage of this procedure is thecomplete electrochemical transformation of the solidduring the potential sweep and a considerable increasein sensitivity.

The aim of this work is the evaluation of the responseefficiency of a carbon paste electrode containing aninorganic solid on which an organic molecule has pre-viously been adsorbed. More precisely, this work isfocused on the electrochemical yield and the linearity ofthe current response. The model molecule was AlizarinS because this presents a reversible two-electron trans-fer reaction both in oxidation and in reduction. Thefuture objective is to evaluate the electroactive naturalorganic molecules adsorbed on soils in a short-timeanalysis and if possible with low-cost electrodes.

Experimental

Cyclic voltammetry was performed by a PJT 24-1 potentiostat anda GSTP 4 signal generator (Tacussel, France) and a LY 1600 XYtrecorder (Linseis, Germany). Pulse polarography using a PRG5 pulse polarograph (Tacussel, France) was used to evaluate theAlizarin concentration in the samples.

The carbon paste electrode has been described elsewhere [3]. Itconsists of a glassy carbon rod (3.5 mm diameter) embedded ina 10 mm diameter Teflon rod. The cavity where the carbon paste ispoured is about 8 mm3; it contains about 8 mg of paste. The carbonpaste was usually made of 300 mg of carbon powder, a knownamount of alumina on which was previously adsorbed AlizarinS according to the procedure described below, and 300 ll of 1 mol/lHClO

4as binder. The amount of paste poured in the cavity of the

electrode was always evaluated by weighing. Reference electrodewas a saturated calomel one, separated from the bulk by a saltbridge filled with the appropriate electrolyte. Auxiliary electrode wasa large size platinum wire. Oxygen was removed by nitrogen bubb-ling. Water was deionized and then distilled. All the reagents were ofanalytical grade. Alumina was purchased from Aldrich and hada specific area of 150 m2.g~1. Alizarin S (or 3,4-dihydroxy-9,10-dioxo-2-anthracene sulfonic acid, sodium salt, MW"342.28) waspurchased from Fluka and used without further purification.

Fig. 1 Typical cyclic voltammetry for a carbon paste electrodecontaining Alizarin S adsorbed on alumina (7.75 lg of sample 1mixed with 300 mg of carbon and 300 ll of 1 mol/l HClO

4)

Adsorption procedure. The adsorption of Alizarin S was performedas follows: a known weight of alumina was added to a knownvolume of an aqueous solution of Alizarin S, and then stirred.(Usually 1 g of alumina in 100 ml of 0.01 mol/l Alizarin S). Afterequilibrium was reached, (approximately 10 min), the mixture wasagain stirred for 2 h in order to break possible aggregates. Then thealumina was filtered by paper filter (high-purity grade with less than0.01% ashes) and carefully dried at 40 °C. Using this procedure,alumina still remained as a powder without aggregates and was usedwithout further treatment. The same procedure was applied tocarbon powder alone in order to determine the weak adsorptionwhich occurred.

Results and discussion

Alizarin S was chosen among various organic com-pounds because it has two independent electrochemicalmechanisms: the reduction of the quinonic groups andthe oxidation of the ortho diphenol groups (Fig. 1).First was determined the maximum adsorbed amountof Alizarin S per gram of graphite powder according tothe above procedure. Only 3 mg of Alizarin S wereadsorbed on 1 g of carbon. Then this powder was usedto evaluate the response of the carbon paste electrode.Small known amounts of this powder were added tofresh carbon and mixed with the binder to obtainseveral pastes containing known amounts of Alizarin S.Then these pastes were poured into the electrode andstudied electrochemically.

In Fig. 2 are given the expected theoretical response.Q

5)%03and the experimental one Q

%91expressed as the

areas under the peaks. It was observed that the yieldwas only about 50%. Figure 3 shows the maximumpeak intensities which varied linearly in the low levelrange chosen here.

The adsorption of Alizarin S on carbon powder wasweak, and it was decided to use an inorganic powder toadsorb more. The choice of alumina was made afterseveral preliminary tests with some other inorganicsolids such as silica or titanium oxide. Adsorption of

Fig. 2 Experimental (d) and theoretical (m) response for a carbonpaste electrode containing increasing Alizarin S amounts (AlizarinS is adsorbed on carbon in this case). The response is evaluated fromthe peak area expressed in millicoulomb

Fig. 3 Linear response of a carbon paste electrode containingincreasing amounts of Alizarin S adsorbed on carbon

Alizarin S on 150 m2.g~1 Al2O

3powder was per-

formed at room temperature. The standard deviation ofthe adsorption procedure was tested on five indepen-dent experiments and was found to be better than 1%.The reproducibility of these experiments may be con-sidered as satisfying. Several samples of alumina con-taining known amounts of Alizarin S were obtainedaccording to the same procedure. One of them (notedsample 1) only contained 74.3 mg of Alizarin S (for 1 gof alumina) (Table 1). As previously described, thecarbon paste was made of a known amount of AlizarinS impregnated alumina (sample 1) mixed with carbonpowder and 1 mol/l HClO

4as binder.

The alumina content of the carbon paste waschanged and was expressed in %w/w of solids. Be-cause the quantity of paste in the electrode was deter-mined by careful weighing, the theoretical amount ofAlizarin S to be oxidized was known for any experi-ment and consequently the necessary electrical charge.Alizarin S needs two Faradays per mole for com-plete oxidation. The electrical yield was measured by

477

Table 1 Electrical response of a carbon paste electrode containingincreasing percentages of alumina with adsorbed Alizarin S.(Sample 1: 74.3 mg/g alumina)

Al2O

3Alizarin S Q

5)%03Q

%91Yield

percentage amount in the mC mC %in the paste electrode% mg

1.61 0.0050 2.8190 1.3600 48.22.70 0.00694 3.9050 1.6338 41.83.85 0.01083 6.0938 2.3046 37.87.46 0.02032 11.4570 3.8770 33.8

14.43 0.03887 21.8715 6.6394 30.4

Fig. 4 Experimental (d) and theoretical (m) response of a carbonpaste electrode containing adsorbed Alizarin S with large aluminaamounts (sample 1)

comparing the electrical charge calculated from theoxidation peak area with the theoretical one. It wasobserved that Q

%91/Q

5)%03decreased when the alumina

amount in the paste was increased (Table 1). Linearityof the electrode response was obtained only for lowalumina percentages, but as shown in Fig. 4, when thepercentage was above 50%, the peak intensity was lessthan expected (the electrical charge was changing in thesame way).

It was also noticed that the peak potential was pro-gressively shifted towards more positive values. Severalexplanations may be advocated. The first one is thechange of the internal resistance of the electrode, sincethe carbon percentage is decreasing. The number ofelectrical contacts was progressively lowered. The sec-ond one may arise from the rate of diffusion of protonsand oxidized Alizarin from the paste towards the bulksince the internal structure of the paste was changing asthe alumina percentage increased, and the adsorptionof the oxidized form was perhaps stronger than that ofAlizarin S. It was interesting here to compare thisbehaviour with that of the electro-dissolution of a thinmetallic film from an inert but conducting electrodesurface, or the electro-dissolution of an inorganic solidincorporated in the paste. The first case was previously

Fig. 5 Variation of the peak height for a constant amount of adsor-bed Alizarin S in the electrode (5 lg) as a function of the sweep rate

Table 2 Alizarin S content for five synthetized samples accordingto the adsorption procedure (see text)

Sample Amount of Number of mole ofAlizarin S Alizarin S in the pastemg mole

1 74.3 2.17]10~42 131.0 3.83]10~43 185.7 5.42]10~44 325.2 9.50]10~45 491.1 1.43]10~3

studied in detail by Brainina [21] who clearly demon-strated that an increase of the thickness of the metallicfilm led to a shift of the oxidation peak potentialtowards more positive values. The second case wasstudied by Vittori et al. [3] for incorporated inorganiccompounds such as Fe

2O

3, FeOOH, and PbCrO.

They proposed an equation well fitting the experimentsand showing again a similar potential shift every timethe amount of incorporated solid was increased.

Probably these two explanations are partly true inthe present case since the voltammetric investigationswere made at 10 mV/s and at such a speed equilibriumwas never reached in the paste. In addition, to outlinethe importance of the surface phenomena, the potentialsweep rate was changed from 0.2 to 20 mV/s. As shownin Fig. 5, the peak intensity varied linearly with thesweep rate, indicating that diffusion was not an influentparameter. (For a pure diffusion process, the intensitymay vary with the square root of the sweep rate).

When a small amount of alumina was added to thepaste, the electrical yield appeared to be higher than fora larger amount. So it was interesting to prepare severalsamples containing known amounts of Alizarin Sadsorbed on alumina. (Table 2).

As shown in Table 3, the electrical charge decreasedas the Alizarin amount increased, despite a lowalumina percentage (see also Fig. 6). Consequently theyield decreased in the same way.

These observations have to be carefully analyzed,since it was expected that for these low alumina

478

Table 3 Comparison of experimental and theoretical charge for theoxidation of Alizarin S adsorbed on alumina, for low aluminapercentages in the carbon paste

Sample % Alumina Q Q Yieldin the carbon (Theoretical) (Experimental) %paste mC mC

1 1.61 2.819 1.360 48.22 1.78 4.062 2.286 48.23 1.69 8.568 3.464 40.44 1.43 7.810 3.183 40.7

4.87 28.85 10.25 35.55 1.75 12.38 3.753 30.3

Fig. 6 Variation of the peak height with increasing amounts ofadsorbed Alizarin S in the electrode (same conditions as in Fig. 4)

percentages the yield will be constant and close to 50%The probable explanation which may be advocatedhere is the possible adsorption of Alizarin in two ways:one part on the external surface of the particles and onepart in the pores of the aggregates. This last part had noelectrical contacts with the carbon particles and did notgive any electrical response.

Conclusion

A carbon paste electrode with an incorporated inor-ganic solid on which an organic molecule was adsorbed

gave a linear electrical response, provided two ruleswere followed: low alumina amount in the electrodeand low amount of organic molecule adsorbed on thealumina. Despite these two restrictions, the carbonpaste electrode is an interesting tool for the determina-tion of organic compounds in natural waters, since it isonly necessary to strip the electroactive molecules ona known amount of alumina under vigorous stirringand then to incorporate the alumina, after filtration, ina carbon paste. Further developments for the evalu-ation of pollutants are under investigation.

References

1. Adams RN (1958) Anal Chem 30:15762. Adams RN (1969) Electrochemistry at solid electrode. Dekker,

New York3. Mouhandess MT, Chassagneux F, Vittori O, Accary A, Reeves

R (1984) J Electroanal Chem 181:93—1054. Bennouna A, Durand B, Vittori O (1986) Electrochim Acta 31:

831—8355. Magri El, Ouadseri M, Vittori O, Durand B (1986) Electrochim

Acta 31:1335—13396. Chassagneux F, Durand B, Vittori O (1987) High Tech Ceram

1083—10907. Bennouna B, Durand B, Vittori O (1991) Electrochim Acta 36:

1505—15148. Tascon ML, Vazquez MD, Pardo R, Sanchez-Batanero P, Iza

J (1989) Electroanalysis 1:363—3669. San Jose MT, Espinosa AM, Tascon ML, Vazquez MD,

Sanchez-Batanero P (1991) Electrochim Acta 36:1209—121810. Espinosa AM, Tascon ML, Vazquez MD, Sanchez-Batanero

P (1991) Electrochim Acta 36:1561—157111. Espinosa AM, Tascon ML, Vazquez MD, Sanchez-Batanero

P (1992) Electrochim Acta 37:1165—117212. Encinas P, Lorenzo L, Tascon ML, Vazquez MD, Sanchez-

Batanero P (1994) J Electroanal Chem 367:99—10813. Gaillochet P, Bauer D, Henion MC (1975) Analusis 3:513—51614. Lamache M, Bauer D (1977) J Electroanal Chem 79:359—36415. Adekola FD, Colin C, Bauer D (1992) Electrochim Acta 37:

2009—201416. Adekola FD, Diaw M, Colin C, Bauer D (1992) Electrochim

Acta 37:2491—249517. Lecuire JM (1975) J Electroanal, Chem 66:195—20518. Lecuire JM, Henry B, Bessieres JM (1977) J Electroanal Chem

83:111—11519. Vallot R, N’Diaye A, Bermont A, Jakubowicz C, Yu LT (1980)

Electrochim Acta 25:1501—151220. Kalcher K (1990) Electroanalysis 2:419—43321. Brainina KZ (1974) Stripping voltammetry in chemical analysis.

Wiley, New York

.

479