Copper determination in ethanol fuel by differential pulse anodic stripping voltammetry at a solid paraffin-based carbon paste electrode modified with 2-aminothiazole organofunctionalized

Page 1

A

dpapeaafkeFr©

K

1

cwoiBate

A

0d

Talanta 71 (2007) 771–777

Copper determination in ethanol fuel by differential pulse anodic strippingvoltammetry at a solid paraffin-based carbon paste electrode modified

with 2-aminothiazole organofunctionalized silica

Regina M. Takeuchi a,∗, Andre L. Santos a, Pedro M. Padilha b, Nelson R. Stradiotto a

a Departamento de Quımica Analıtica, Instituto de Quımica, UNESP, Araraquara, Brazilb Departamento de Quımica e Bioquımica, IB, UNESP, Botucatu, Brazil

Received 24 March 2006; received in revised form 11 May 2006; accepted 12 May 2006Available online 27 June 2006

bstract

Solid paraffin-based carbon paste electrodes modified with 2-aminothiazole organofunctionalized silica have been applied to the anodic strippingetermination of copper ions in ethanol fuel samples without any sample treatment. The proposed method comprised four steps: (1) copper ionsreconcentration at open circuit potential directly in the ethanol fuel sample; (2) exchange of the solution and immediate cathodic reduction of thebsorbate at controlled potential; (3) differential pulse anodic stripping voltammetry; (4) electrochemical surface regeneration by applying a positiveotential in acid media. Factors affecting the preconcentration, reduction and stripping steps were investigated and the optimum conditions weremployed to develop the analytical procedure. Using a preconcentration time of 20 min and reduction time of 120 s at −0.3 V versus Ag/AgClsat

linear range from 7.5 × 10−8 to 2.5 × 10−6 mol L−1 with detection limit of 3.1 × 10−8 mol L−1 was obtained. Interference studies have showndecrease in the interference effect according to the sequence: Ni > Zn > Cd > Pb > Fe. However, the interference effects of these ions have not

orbidden the application of the proposed method. Recovery values between 98.8 and 102.3% were obtained for synthetic samples spiked withnown amounts of Cu2+ and interfering metallic ions. The developed electrode was successfully applied to the determination of Cu2+ in commercial

thanol fuel samples. The results were compared to those obtained by flame atomic absorption spectroscopy by using the F-test and t-test. Neither-value nor t-value have exceeded the critical values at 95% confidence level, confirming that there are no significant differences between the

esults obtained by both methods.2006 Elsevier B.V. All rights reserved.

unctio

blfw

ico

eywords: Ethanol fuel; Solid paraffin-based carbon paste electrodes; Organof

. Introduction

The use of ethanol fuel as an alternative energy source hasontributed to reduce CO2 emission levels in Brazil. This fuelas initially used as an additive in gasoline which enhances itsctane number, replacing the organolead compounds and signif-cantly reducing the CO2 emission [1,2]. Nowadays, its use inrazil is as a new car fuel option with economic and ecological

dvantages [3]. The Brazilian’s ethanol fuel is obtained fromhe fermentation of sugar cane and this process is now totallystablished [4]. Thus, ethanol is an attractive alternative fuel

∗ Corresponding author at: Av. Prof. Francisco Degni s/n, CEP14800-900raraquara, SP, Brazil. Tel.: +55 16 3301 6621; fax: +55 16 3322 7932.

E-mail address: [email protected] (R.M. Takeuchi).

fmeti

aa

039-9140/$ – see front matter © 2006 Elsevier B.V. All rights reserved.oi:10.1016/j.talanta.2006.05.035

nalized silica; Copper ions; Differential pulse stripping voltammetry

ecause it is a renewable bio-based resource and it is less pol-utant than petroleum derivatives. The characteristics of ethanoluel indicate that this fuel will play an important role in world-ide energetic polities in the next years.The presence of trace metals ions in ethanol fuel plays an

mportant role in the engine maintenance, since metallic speciesan accelerate the corrosion of engines or promote the formationf gums and sediments [5]. In addition, these metals contributeor the environmental contamination, mainly in big cities. Theain sources of metal ions are associated to the sugar cane

thanol production process, storage and transportation [6]. Thus,he development of analytical methods to determine trace metal

ons in ethanol fuel is a relevant research field.

The determination of metals by electrothermal atomicbsorption spectroscopy in ethanol fuel is well established [7–9]nd flame atomic absorption spectroscopy (FAAS) can be used

Page 2

7 Talan

adoc

msSbfwmfelcmbmeiuai

atmiptbisiDevdomsawbcmci[c[c

sac

iv

2

2

taMS5tsStopcecl

2

ftoiiauAta

2

wppotpschc(

72 R.M. Takeuchi et al. /

fter a preconcentration step [10–12]. Recently, a FAAS proce-ure for iron determination in ethanol fuel samples was devel-ped by introducing the sample in spectrophotometer in flowonditions using nickel as internal standard [13].

Electroanalytical techniques such as anodic stripping voltam-etry can be used for metals determination providing high

ensitivity and precision with relatively low instrumental costs.ome works have shown that anodic stripping voltammetry cane successfully applied to metals ions determination in ethanoluel [14,15], being obtained detection limit values compatibleith those obtained by AAS. The introduction of a chemicalodifier able to preconcentrating metallic ions at electrode sur-

ace either by complexation or electrostatic attraction can lead tolectroanalytical procedures more sensitive with lower detectionimit values. Silica or organofunctionalized silica with groupsontaining S, N or O atoms are very efficient materials to pro-ote the metal ions preconcentration [16]. These materials have

een widely used to preconcentrate metals for subsequent deter-ination by AAS in several kinds of matrices [17–19]. Roldan

t al. [12] have performed the determination of Cu, Ni and Znn ethanol fuel after preconcentrarion of these metals in a col-mn packed with 2-aminothiazole functionalized silica gel. Theuthors found that 2-aminothiazole groups presented high capac-ty of metal ions preconcentration by complexation.

The extremely advantageous silica properties like its highdsorption capacity, chemical stability and possibility of func-ionalization with a large variety of functional groups make these

aterials excellent modifier agents to construct chemically mod-fied electrodes [20,21]. Several works have demonstrated theotentialities of using silica modified electrodes for metals elec-roanalytical determination in different matrices [22–25]. Car-on paste electrodes are easily modified, since that no specificnteractions between modifier and electrode surface are neces-ary, thus these type of electrode have been widely employedn electroanalysis of organic and inorganic compounds [26].espite the great success of using carbon paste electrodes in

lectroanalysis, the chemical and mechanical proprieties of con-entional carbon paste electrodes limit their use in flow con-itions or in totally non-aqueous media. The main problemf using conventional carbon paste electrodes in non-aqueousedia is the dissolution of binder agent promoted by organic

olvents; the binder agent is also responsible by the inappropri-te mechanical strength of conventional carbon paste electrodeshich limit its use in flow conditions. Several strategies haveeen used to improve the chemical/mechanical proprieties ofarbon paste electrodes, all of them replace mineral oil, theost used binder agent, by other able to produce more rigid

omposites. Thus, the literature have presented several promis-ng alternative binder agents such as epoxy [27,28], silicone29], Teflon® [30] polyurethane [31]. The solid paraffin-basedarbon paste electrode was introduced by Petit and Kauffmann32], these authors have successfully used this electrode in flowonditions [32] and in the development of biosensors [33].

In this context, this work describes the construction of aolid paraffin-based carbon paste electrode modified with 2-minothiazole organofunctionalized silica (SiAt-SPCPE) whichan be successfully employed to Cu2+ determination directly

psnp

ta 71 (2007) 771–777

n ethanol fuel samples by differential pulse anodic strippingoltammetry.

. Experimental

.1. Reagents

Stock solutions of all metallic species were prepared by dilu-ion of corresponding analytical purity grade chloride salts innalytical purity grade ethanol from Merck. Nitric acid fromerck was used to prepare the electrolyte supporting solution.

ilica gel (Merck) with specific surface area between 486 and20 m2 g−1 and average pore diameter of 0.6 nm was utilizedo construct both the SiAt-SPCPE and the non-functionalizedilica modified solid paraffin-based carbon paste electrode (Si-PCPE). 2-aminothiazole analytical purity grade was employed

o achieve the silica functionalization. The 2-aminothiazolerganofunctionalized silica was synthesized according to therocedure previously described in literature [12]. Spectroscopicarbon powder from Merck was used to obtain the carbon pastelectrode, as binder agent it was used paraffin from Synth. Theommercial ethanol fuel samples were acquired from differentocal gas stations.

.2. Apparatus

Differential pulse anodic stripping voltammetry were per-ormed with an AUTOLAB PGSTAT 30 potentiostat coupledo a microcomputer and controlled by GPES 4.9 software. Ane-compartment electrochemical cell with a platinum wire aux-liary electrode and an Ag/AgClsat reference electrode was usedn all voltammetric measurements. The SiAt-SPCPE was useds working electrode. FAAS experiments were carried out bysing a spectrometer from Perkin-Elmer model Analyst 100.copper hollow cathode lamp from Perkin-Elmer was used in

hese experiments. The analytical procedure adopted in FAASnalysis was the same used by Oliveira et al. [14].

.3. Electrode preparation

Unmodified carbon pastes, used for comparison purpose,ere prepared by sluggish addition of carbon powder in meltedaraffin in the proportion of 60:40% (w/w). Modified carbonastes were prepared by substituting corresponding amountsf the carbon powder by silica or 2-aminothiazole orgafunc-ionalized silica to obtain the desired composition. The carbonowder–modifier was homogenized in a mortar and pestle, sub-equently the mixture was added to melted paraffin whose per-entage was always kept at 40%. The new mixture was againomogenized in a thermostatized bath (65–75 ◦C). The obtainedarbon pastes were then packed in a cylindrical plastic tubeinternal diameter of 4.8 mm) equipped with a copper piston

roviding an inner electrical contact. Smoothing of the electrodeurface was made by hand-polishing on a weighing paper. Whenecessary, a new surface was obtained by pushing an excess ofaste out of the plastic tube.

Page 3

Talant

2

pCb1etcf

iea(ci

icitdiippmsv

twcm

3

3S

eotbtwhCsicphs

fs

SS(Sbwic0

aHpBptaiTisostpeak is observed after preconcentration and cathodic reductionsteps, indicating that this electrode is unable to preconcentrateCu2+.

Fig. 1. Linear sweep voltammograms registered in 0.2 mol L−1 HNO3 solu-tion, scan rate: 100 mV s−1. (A) Using SiAt(20%)-SPCPE after immersionby 14 min in analytical-grade ethanol followed by electrochemical reductionstep at −0.3 V vs. Ag/AgClsat by 50 s. (B) Using Si-SPCPE after preconcen-

R.M. Takeuchi et al. /

.4. Analytical procedure

Before its utilization, the SiAt-SPCPE was preconditioned byerforming the preconcentration step in a 1.0 × 10−3 mol L−1

u2+ solution in analytical-grade ethanol for 5 min, followedy electrochemical reduction at −0.4 V versus Ag/AgClsat for20 s in 0.2 mol L−1 HNO3 solution. Subsequently, the differ-ntial pulse anodic stripping voltammogram was registered inhis same solution. After that, the electrode surface was electro-hemically regenerated by applying +0.3 V versus Ag/AgClsator 120 s in 1.0 mol L−1 HNO3 solution.

The analytical procedure comprised four steps: (1) copperons preconcentration at open circuit potential directly in thethanol fuel sample, (2) exchange of the solution and immedi-te cathodic reduction of the absorbate at controlled potential,3) differential pulse anodic stripping voltammetry, (4) electro-hemical surface regeneration by applying a positive potentialn acid media.

For the preconcentration step, the SiAt-SPCPE was immersedn a stirred 10 mL Cu2+ solution in analytical-grade ethanol,ontaining a known amount of this metal for a controlled timenterval in open circuit potential conditions. The electrode washen removed from the preconcentration cell, briefly rinsed witheionized water, and placed in an electrochemical cell contain-ng 20 mL of 0.2 mol L−1 HNO3 solution as electrolyte support-ng. In this electrochemical cell the cathodic reduction of Cu2+

reconcentrated was performed by electrolyses at controlledotential. Finally, differential pulse anodic stripping voltam-ograms were registered. After the steps above, the electrode

urface was electrochemically regenerated by applying +0.3 Versus Ag/AgClsat for 60 s in 1.0 mol L−1 HNO3 solution.

For Cu2+ determination in commercial ethanol fuel samples,he preconcentration step was performed directly in the sample,ithout any pretreatment or dilution step. The Cu2+ quantifi-

ation in these samples was performed by standard additionethod.

. Results and discussion

.1. Anodic stripping voltammetric behavior of Cu2+ oniAt-SPCPE

We have initially tried to use conventional carbon pasteslectrodes with Nujol® as binder agent, however we havebserved a swelling effect when these electrodes were in con-act with ethanolic solutions, probably due to binder dissolutiony ethanol. As a consequence, we have observed that these elec-rodes provide peak current values relative to cooper strippingith very high dispersion presenting relative standard deviationigher than 25% after preconcentration in a 1.0 × 10−6 mol L−1

u2+ solution in analytical-grade ethanol. These results demon-trate that conventional carbon paste electrodes are unstablen ethanolic solutions and therefore they are inappropriate to

ooper determination in ethanol fuel samples. To overcome thisroblem we have replaced Nujol® by solid paraffin, thus weave obtained a more rigid and compact electrode, which istable in ethanolic solutions and can be successfully employed

t1UsA

a 71 (2007) 771–777 773

or Cu2+ determination directly in commercial ethanol fuelamples.

Fig. 1 presents the linear sweep voltammograms of Si-PCPE after Cu2+ preconcentration (curve B) and of SiAt(20%)-PCPE without (curve A) and with Cu2+ preconcentrationcurve C) in 0.2 mol L−1 HNO3 solution. The SiAT(20%)-PCPE corresponds to a carbon paste with composition: car-on powder:paraffin:modifier (40:40:20%, w/w/w). Curve Aas registered after the immersion of SiAt-SPCPE for 10 min

n analytical-grade ethanol without Cu2+ followed by electro-hemical reduction at −0.3 V versus Ag/AgClsat for 60 s in.2 mol L−1 HNO3 solution.

Fig. 1 shows that both Si-SPCPE and SiAt(20%)-SPCPEre able to preconcentrate Cu2+ from ethanolic solutions.owever, the anodic stripping peak observed for Si-SPCPEresents low intensity and a poor voltammetric profile (curve). When SiAt(20%)-SPCPE is used the anodic strippingeak current obtained is approximately four times higher thanhe obtained with Si-SPCPE. This result indicates that 2-minothiazole organofunctionalized silica has a higher capac-ty of Cu2+ preconcentration than non-functionalized silica.he SiAt(20%)-SPCPE immersed in analytical grade ethanol

n absence of Cu2+ followed by cathodic reduction stephowed any voltammetric peak, indicating that the anodic peakbserved on curves A and C is associated to Cu2+ anodictripping and not to other species present in ethanolic solu-ion. For unmodified carbon paste electrode any voltammetric

ration in a 1.0 × 10−4 mol L−1 Cu2+ solution in analytical-grade ethanol by4 min, reduction potential: −0.3 V vs. Ag/AgClsat, reduction time: 50 s. (C)sing SiAt(20%)-SPCPE after preconcentration in a 1.0 × 10−4 mol L−1 Cu2+

olution in analytical-grade ethanol by 14 min, reduction potential: −0.3 V vs.g/AgClsat, reduction time: 50 s.

Page 4

7 Talan

ceocSaei

3

vaiicuctfaofw

rtppsPA

Frat5

ivrm

wFctie

d−pT5tfmpesl

mifoto

74 R.M. Takeuchi et al. /

The concentration of SiAt in the carbon paste had a signifi-ant influence on the voltammetric response this influence wasvaluated by using modifier percentages from 10 to 30%. It wasbserved that a maximum peak current was obtained when SiAtoncentration in the paste was 20%. Higher concentrations ofiAt significantly decreased the peak current. This is presum-bly due the reduction of conductive area (carbon particles) at thelectrode surface. Thus, the SiAt(20%)-SPCPE was employedn all subsequent experiments.

.2. Analytical experimental conditions optimization

The dependence of the differential pulse anodic strippingoltammetric peak current with the preconcentration time for5.0 × 10−7 mol L−1 Cu2+ solution in analytical-grade ethanol

s shown in Fig. 2. The peak current increases by increas-ng preconcentration time, indicating an enhancement of Cu2+

oncentration at the electrode surface. The peak current val-es increase rapidly until 20 min, and then become almostonstant. This behavior indicates that after 20 min preconcen-ration time, a steady-state equilibrium of adsorption/complexormation–desorption/complex dissociation is reached. This istypical behavior often observed for stripping methods based

n adsorptive or complex formation accumulation [34]. Thus,or all subsequent experiments, preconcentration time of 20 minas used.The influence of the reduction potential on the peak cur-

ent values was studied by varying this parameter from −0.2o −0.6 V versus Ag/AgClsat. It was observed that reductionotential values less negative than −0.3 V versus Ag/AgClsat

roduce an anodic stripping voltammetric peak with low inten-ity and, consequently, inappropriate for Cu2+ determination.otential reduction values more negative than −0.3 V versusg/AgClsat, did not improve the peak current values. Therefore,

ig. 2. Effect of preconcentration time on the anodic stripping peak cur-ent obtained after preconcentration in a 5.0 × 10−7 mol L−1 Cu2+ solution innalytical-grade ethanol. Reduction potential: −0.4 V vs. Ag/AgClsat, reduc-ion time: 30 s. Voltammetric conditions—pulse amplitude: 50 mV, pulse width:.0 ms and scan rate: 10 mV s−1.

Rpe

FagA5

ta 71 (2007) 771–777

n all subsequent experiments the reduction potential of −0.3 Versus Ag/AgClsat was employed, since more negative potentialeduction could negatively affect the selectivity of the proposedethod.The effect of the reduction time on peak current values

as also examined and the obtained results are shown inig. 3. It was observed a significant increase in the peakurrent values until 120 s reduction time. For higher reductionime the peak current values become almost constant. Thus,n all subsequent experiments a reduction time of 120 s wasmployed.

After the optimization of all analytical experimental con-itions (preconcentration time: 20 min, potential reduction:0.3 V versus Ag/AgClsat, time reduction: 120 s) it was

erformed an optimization of the voltammetric parameters.hese studies were performed after preconcentration in a.0 × 10−7 mol L−1 Cu2+ solution in analytical grade ethanolesting pulse amplitude values from 10 to 100 mV, pulse widthrom 5.0 to 50 ms and scan rate from 5.0 to 20 mV s−1. The opti-ized voltammetric parameters were—pulse amplitude: 50 mV,

ulse width: 5.0 ms and scan rate: 10 mV s−1. All optimizationxperiments were carried out using the same electrode withouturface renewal, indicating that SiAt(20%)-SPCPE presents aong lifetime.

Under the optimized conditions, five successive measure-ents were performed in a 5.0 × 10−7 mol L−1 Cu2+ solution

n analytical grade ethanol. When these experiments were per-ormed using the same electrode without surface renewal it wasbserved a relative standard deviation (R.S.D.) of 3.7%, by usinghe same electrode with surface renewal a R.S.D. of 4.8% wasbtained, finally by using different electrodes it was obtained a

.S.D. of 5.7%. These results indicate that the proposed methodresents good precision and repeatability even when differentlectrodes are used.

ig. 3. Effect of reduction time on the anodic stripping peak current obtainedfter preconcentration in a 5.0 × 10−7 mol L−1 Cu2+ solution in analytical-rade ethanol. Preconcentration time: 20 min, reduction potential: −0.3 V vs.g/AgClsat. Voltammetric conditions—pulse amplitude: 50 mV, pulse width:.0 ms and scan rate: 10 mV s−1.

Page 5

R.M. Takeuchi et al. / Talanta 71 (2007) 771–777 775

Fig. 4. Differential pulse anodic stripping voltammograms registered afterpreconcentration in different Cu2+ concentrations in analytical-grade ethanol.Preconcentration time: 20 min, reduction potential: −0.3 V vs. Ag/AgClsat,reduction time: 120 s. Voltammetric conditions—pulse amplitude: 50 mV, pulsew(2

3

fs

i2dfee(eenCol

3

imtTmismic

Table 1Effect of interfering metal ions on anodic stripping peak current observed afterpreconcentration in a 5.0 × 10−7 mol L−1 Cu2+ solution in analytical-gradeethanol in the presence of interfering metal ions

Interfering ion Interfering ion (Cu2+ ratio) Relative sign (%)

Fe3+ 0.1 98.21.0 92.0

10 62.0

Ni2+ 0.1 95.71.0 68.7

10 43.8

Zn2+ 0.1 95.81.0 72.8

10 38.8

Pb2+ 0.1 97.51.0 86.7

10 45.4

Cd2+ 0.1 96.9

sm

ihcctfam2eaefe5iptdfta

sscsZ

idth: 5.0 ms and scan rate: 10 mV s−1. (A) 0.0; (B) 7.5 × 10−8 mol L−1;C) 2.5 × 10−7 mol L−1; (D) 7.5 × 10−7 mol L−1; (E) 1.3 × 10−6 mol L−1; (F).0 × 10−6 mol L−1; (G) 2.5 × 10−6 mol L−1. Analytical curve inserted.

.3. Analytical curve

The differential pulse anodic stripping voltammograms at dif-erent Cu2+ concentrations under the optimized conditions arehown in Fig. 4.

The peak current values lineally increase with increas-ng Cu2+ concentration in the range from 7.5 × 10−8 to.5 × 10−6 mol L−1. At higher Cu2+ concentrations negativeeviation from linearity were observed due to electrode sur-ace saturation. The analytical curve was linear according toquation: ip (�A) = 0.35 + 1.6 × 107 CCu2+ (mol L−1) with lin-ar correlation coefficient equal to 0.9997. The detection limitDL) was found to be 3.1 × 10−8 mol L−1, evaluated by thequation: DL = 3S.D./(analytical curve slope), where S.D. isquivalent to the standard deviation of the average of the sig-al of five measurements of the blank in the peak potential ofu2+ anodic stripping. The obtained DL value is close to thosebtained by other stripping voltammetric methods proposed initerature to Cu2+ determination in ethanol fuel samples [14,15].

.4. Interference studies

The interference studies were carried out by evaluating thenterference effects of Ni2+, Pb2+, Zn2+, Cd2+ and Fe3+. These

etal species were used because some works have shown thathey are present in commercial ethanol fuel samples [9,12,14].he interference studies were performed by using interferingetal ion/Cu2+ ratios equal to 0.1, 1.0 and 10. Higher interfer-

ng metal ion contents were not used because some works have

hown that Cu2+ is the main transition metal present in com-ercial ethanol fuel samples [9,14,15]. Thus, the presence of an

nterfering metal ion excess higher than ten times the Cu2+ con-entration is an extreme situation hardly found in commercial

rUoi

1.0 76.710 40.8

amples. The influence of interfering ions on the Cu2+ voltam-etric response is shown in Table 1.It was observed that the presence of interfering metal ions

n concentrations ten times smaller than Cu2+ concentrationas only little effect on the Cu2+ voltammetric response,ausing a decrease smaller than 5.0% in voltammetric peakurrents. When the others ions and Cu2+ were presented inhe same concentration, it was observed the following inter-erence order: Ni > Zn > Cd > Pb > Fe, indicating that the 2-minothiazole groups are able to form complex with othersetallic species besides the Cu2+. Table 1 also shows that the

-aminothiazole groups preferentially accumulate Cu2+ becauseven in presence of a ten-fold excess of interfering metal ions,Cu2+ relative signal close to 40% is still observed. In order to

valuate if the interference effect of the other metallic ions couldorbid the method application in commercial samples, recoveryxperiments were performed in synthetic matrices containing.0 × 10−7 mol L−1 of Cu2+ and this same concentration of eachnterfering ion. In these experiments Cu2+ determination waserformed by standard addition method. Recoveries from 98.8o 102.3% were obtained indicating that Cu2+ can be reliablyetermined by the proposed method even in the presence of inter-ering ions when the standard addition method is used. Amonghe interfering ions studied only Cd2+ and Pb2+ have presentednodic stripping voltammetric peak as shown in Fig. 5.

Fig. 5 shows that the SiAt(20%)-SPCPE could be used inimultaneous determination of Cd2+, Pb2+ and Cu2+, howeverome works have shown that Cd2+ and Pb2+ are in extremely lowoncentrations in commercial ethanol fuel samples. Althoughome works have presented stripping voltammetric methods forn2+ [35], it is well known that this metal is electrochemically

educed only in potential values more negatives than −1.0 V.

nder the experimental conditions adopted in this work it wasbserved that potential values as negative as −1.0 V promote anntense hydrogen evolution, causing damages in electrode sur-

Page 6

776 R.M. Takeuchi et al. / Talanta 71 (2007) 771–777

Fig. 5. Differential pulse voltammogram registered in 0.2 mol L−1 HNO3 solu-tion after preconcentration in a 5.0 × 10−7 mol L−1 Cu2+ solution in analytical-grade ethanol containing this same concentration of each interfering metal ion.Prw

ftuaaao

3

nlgw

aiimtivowss

mbteFb

Fig. 6. Differential pulse voltammogram registered in 0.2 mol L−1 HNO3 solu-tion after preconcentration in the commercial ethanol fuel sample 1. Precon-centration time: 20 min, reduction potential: −0.85 V vs. Ag/AgClsat, reductiont5

tmrettwisecidpisatCCtrode disintegration or modifier loss by dissolution in ethanolicmedium.

Table 2Comparison between the Cu2+ concentrations determined in commercial ethanolfuel samples by proposed method and by FAAS

Sample Proposed method(×10−7 mol L−1)a

FAAS(×10−7 mol L−1)a

tb Fc

1 4.15 ± 0.12 4.02 ± 0.08 1.58 2.25

reconcentration time: 20 min, reduction potential: −0.85 V vs. Ag/AgClsat,eduction time: 120 s. Voltammetric conditions—pulse amplitude: 50 mV, pulseidth: 5.0 ms and scan rate: 10 mV s−1.

ace and a significant increase in the background current. Thus,he Zn2+ anodic stripping voltammetric peak was not observednder the experimental conditions adopted. Fig. 5 shows annodic peak at −0.78 V versus Ag/AgClsat which is not associ-ted with any interfering specie studied, probably this peak isssociated to electrooxidation of 2-aminothiazole group previ-usly reduced at −0.85 V versus Ag/AgClsat.

.5. Commercial samples analysis

The proposed analytical method was used for Cu2+ determi-ation in two commercial ethanol fuel samples acquired fromocal gas stations. Fig. 6 shows the differential pulse voltammo-ram registered after preconcentration directly in the sample 1ithout any treatment or dilution.The anodic stripping voltammetric peaks relative to Cd2+

nd Pb2+ were not observed in any sample analyzed, indicat-ng that these ions are absent or are in concentration levelsmpossible to detect with the proposed method. The Cu2+ deter-

ination in commercial ethanol fuel samples was performed inriplicate by using the standard addition method. In these exper-ments the previously optimized reduction potential of −0.3 Versus Ag/AgClsat was used, since no voltammetric signal wasbserved for Cd2+ and Pb2+. The obtained results were comparedith those obtained by FAAS, the mean values, the respective

tandard deviations, t-values and F-values obtained are pre-ented in Table 2.

Table 2 shows that the results obtained by the proposedethod are concordant with the results obtained by FAAS for

oth samples. The obtained t-values were always smaller than

he critical value, indicating that there is not statistical differ-nce between the results obtained with both methods. Obtainedvalues have not exceeded the critical value, indicating that

oth methods presents similar precisions. Finally, the Cu2+ con-

2

ime: 120 s. Voltammetric conditions—pulse amplitude: 50 mV, pulse width:.0 ms and scan rate: 10 mV s−1.

ents determined in the commercial samples are in good agree-ent with those values reported in literature [9,14,15]. These

esults show that Cu2+ can be reliably determined in commercialthanol fuel samples by using the proposed method. In addi-ion, the proposed method presents the advantage of to allowhe Cu2+ preconcentration directly from commercial samplesithout any treatment or dilution. The use of chemically mod-

fied electrodes for metal determination directly in ethanol fuelamples is often limited by modifier dissolution in ethanol. How-ver, in this work this problem was overcome by using a modifierhemically attached on silica. Another problem of using CPEsn a totally non-aqueous medium is associated with binder agentissolution, leading to CP disintegration [26]. In this work thisroblem was overcome by using solid paraffin as binder agentnstead mineral oil, which is the most used binder agent to con-truct CPEs. The use of paraffin as binder agent has producedmore rigid CP, which is stable in ethanol by long periods of

ime. Thus, the strategies adopted to modify and to prepare thePEs have allowed the application of the developed electrode foru2+ directly in commercial ethanol fuel samples without elec-

2.56 ± 0.24 2.34 ± 0.19 1.28 1.60

a N = 3.b tcritical = 2.78 (P = 0.05 with 4 degrees of freedom).c F2/2 critical = 19.0 (P = 0.05).

Page 7

Talant

4

easmauletiaTmemlfs

A

7f

R

[

[[

[

[

[[

[

[

[[[[

[

[

[[

[[

[

[

[

[32] C. Petit, J.-M. Kauffmann, Anal. Proc. 32 (1995) 11.

R.M. Takeuchi et al. /

. Conclusions

A new chemically modified solid paraffin-based carbon pastelectrode was developed using 2-aminothizole organofunction-lized silica for Cu2+ determination in commercial ethanol fuelamples. The use of paraffin as binder agent has produced aore rigid carbon paste, which is stable in ethanol and has

llowed the Cu2+ determination in ethanolic solutions. These of a modifier chemically attached on silica has avoidedosses of modifier by dissolution by ethanol. The developedlectrode is able to preconcentrate Cu2+ from ethanolic solu-ions by complexation with 2-aminothizole groups. Despite thenterference of other ions it was observed that Cu2+ can be reli-bly determined when the standard addition method is used.his is the first work in literature using a stripping voltam-etric method for Cu2+ determination directly in commercial

thanol fuel samples. The obtained results were in good agree-ent with those obtained with FAAS and those reported in

iterature, indicating that the developed electrode can be success-ully used for Cu2+ determination in commercial ethanol fuelamples.

cknowledgements

The authors are grateful to FAPESP (contract no. 03/09334-(RMT) and no. 03/05567-7 (ALS), CTPetro/FINEP and ANP

or financial support.

eferences

[1] I.C. Macedo, Biomass Bioenergy 14 (1998) 77.[2] F. Rosillo-Calle, L.A.B. Cortez, Biomass Bioenergy 14 (1998) 115.[3] K. Sandelim, R. Backman, Environ. Sci. Technol. 33 (1999) 4508.[4] A.E. Wheals, L.C. Basso, D.M.C. Alves, H.V. Amorin, Trends Biotechnol.

17 (1999) 482.[5] D.B. Taylor, R.E. Synovec, Talanta 40 (1993) 495.[6] I.M.R.A. Bruning, E.B. Malm, Bol. Tec. Petrobras 25 (1982) 217.[7] A.P. Oliveira, M. Morais, J.A.G. Neto, E.C. Lima, Atom. Spectrosc. 23

(2000) 39.

[[[

a 71 (2007) 771–777 777

[8] A.P. Oliveira, M. Morais, J.A.G. Neto, E.C. Lima, Atom. Spectrosc. 23(2000) 190.

[9] T.D. Saint’Pierre, V.L.A. Frescura, A.J. Curtius, Talanta 68 (2006) 957.10] L.A.M. Gomes, P.M. Padilha, J.C. Moreira, N.L. Dias-Filho, Y. Gushiken,

J. Braz. Chem. Soc. 9 (1998) 494.11] P.M. Padilha, C.C.F. Padilha, J.C. Rocha, Chim. Anal. 18 (1999) 299.12] P.S. Roldan, I.L. Alcantara, G.R. Castro, J.C. Rocha, C.C.F. Padilla, P.M.

Padilla, Anal. Bional. Chem. 375 (2003) 574.13] J.E. Silva, F.A. Silva, M.F. Pimentel, R.S. Honorato, V.L. Silva,

M.C.B.S.M. Montenegro, A.N. Araujo, Talanta 70 (2006) 522–526.14] M.F. Oliveira, A.A. Saczk, L.L. Okumura, A.P. Fernandes, M. Moraes,

N.R. Stradiotto, Anal. Bioanal. Chem. 380 (2004) 135.15] R.A.A. Munoz, L. Angnes, Microchem. J. 77 (2004) 157.16] R.S.A. Machado-Junior, M.G. Fonseca, L.N.H. Arakaki, J.G.P. Espınola,

S.F. Oliveira, Talanta 63 (2004) 317.17] N.L. Dias-Filho, Y. Gushikem, W.L. Polito, J.C. Moreira, E.O. Ehirim,

Talanta 42 (1995) 1625.18] P. Lessi, N.L. Dias-Filho, J.C. Moreira, J.T.S. Campos, Anal. Chim. Acta

327 (1996) 183.19] A.G.S. Prado, C. Airoldi, Anal. Chim. Acta 432 (2001) 201.20] A. Walcarius, Electroanalysis 10 (1998) 1214.21] A. Walcarius, Electroanalysis 13 (2001) 701.22] L.M. Aleixo, M.F.B. Souza, O.E.S. Godinho, G. Oliveira-Neto, Y.

Gushikem, Anal. Chim. Acta 271 (1993) 143.23] A. Walcarius, N. Luthi, J.-L. Blin, B.-L. Su, L. Lamberts, Electrochim.

Acta 44 (1999) 4601.24] W. Yantasee, Y. Lin, G.E. Fryxell, B.J. Beusche, Anal. Chim. Acta 502

(2004) 207.25] I. Jureviciute, A. Malinauskas, Chem. Anal. (Warsaw) 49 (2004) 339.26] I. Svancara, K. Vytras, J. Barek, Z. Zima, Crit. Rev. Anal. Chem. 31 (2001)

311.27] M.F.S. Teixeira, A.Z. Pinto, O. Fatibello-Filho, Talanta 45 (1997) 249.28] L. Moreno-Baron, A. Merkoci, S. Alegret, Electrochim. Acta 48 (2003)

2599.29] Y. Ballesteros, M.J.G. Huebra, M.C. Quintana, P. Hernandez, L. Hernandez,

Microchem. J. 74 (2003) 193.30] A.G.-V. Prada, N. Pena, C. Parrado, A.J. Reviejo, J.M. Pingarron, Talanta

62 (2004) 896.31] R.K. Mendes, S. Claro-Neto, E.T.G. Cavalheiro, Talanta 57 (2002)

909.

33] C. Petit, A. Gonzalez-Cortes, J.-M. Kauffmann, Talanta 42 (1995) 1783.34] H. Alemu, B.S. Chandravanshi, Anal. Chim. Acta 368 (1998) 165.35] M.F. Oliveira, A.A. Saczk, L.L. Okumura, N.R. Stradiotto, Eclet. Quim.

27 (2002) 153.