Cyclic voltammetry study of arsenic in acidic solutions

Electrochimica Acta52(2007)

6644–6649

Cyclic voltammetry study of arsenic in acidic solutions

Fabiola Brusciotti?,Paul Duby

Department of Earth and Environmental Engineering,Columbia University,NY,United States

Received19February2007;received in revised form10April2007;accepted10April2007

Available online25April2007

Abstract

An electrochemical investigation of arsenic species in acidic solutions,using cyclic voltammetry at a platinum rotating disk electrode,has been carried out.A well de?ned peak at about1V SCE,with a large overpotential(>0.5V)is attributed to the oxidation of As(III)to As(V).The reaction is under the mixed control of ionic transport and charge transfer.The corresponding reduction of As(V)is not observed in the potential range of the study(?0.6to+1.2V SCE).As(V)is not electrochemically active in this range and it has no effect on hydrogen evolution.With As(III)in solution, hydrogen evolution on Pt is shifted towards more negative potentials and it takes place simultaneously with arsine production.A small cathodic peak at?0.2V SCE and a corresponding anodic peak at+0.3V SCE are attributed to the deposition of a thin layer of As by reduction of As(III)and its dissolution.

?2007Elsevier Ltd.All rights reserved.

Keywords:Arsenic;Cyclic voltammetry;Rotating disk electrode;Kinetics

1.Introduction

Most methods currently employed for arsenic removal from drinking water are based on oxidation followed by adsorption or co-precipitation of As(V)with iron and other metal oxides. An electrochemical investigation of the behavior of arsenic in aqueous solutions will improve the understanding of these reac-tions.This is an important step towards optimizing the process and making the removal of arsenic more ef?cient.

Cyclic voltammetry(CV)is a powerful technique to study redox reactions in aqueous solutions,surface deposition and adsorption[1,2].When paired with a rotating disk electrode (RDE),it can be used to investigate the kinetics of charge trans-fer at the electrode/electrolyte interface[3–7].Solution of the convection–diffusion equations(assuming laminar?ow to the rotating electrode surface)leads to a Nernst diffusion layer(δ) given by:

δ=1.61D1/3ν1/6kω?1/2(1) where D is the diffusion coef?cient of the species(cm2/s),νk the kinetic viscosity of the solution(cm2/s)andωis the rotating speed of the electrode(rad/s).

?Corresponding author.Tel.:+3236293534.

E-mail address:fb118@https://www.360docs.net/doc/f15068047.html,(F.Brusciotti).

The current generated by an electrochemical process under a given potential gradient is determined by the transport of ions to the electrode and the kinetics of its conversion at the electrode surface.Under steady-state conditions,prevalent at an RDE,the Koutecky-Levich equation expresses the relative contribution of mass transport and kinetics(in terms of resistance to charge transfer)to the current generated at a RDE experiment[8]:

I?1=(0.62FAD2/3ν?1/6

k

ω1/2C)?1+(FAkC)?1(2) where C is the bulk concentration of the reacting species (mol/cm3),F the Faraday constant(C/mol),A the electrode area(cm2),and k is the rate transfer coef?cient(cm/s).Thus, at potentials in the range where mass transfer and kinetic effects are signi?cant,a plot of I?1versusω?1/2should be linear.In the case of the mass transport-limited region of the current,the Levich equation is veri?ed:

I l=0.62FAD2/3ν?1/6

k

ω1/2C(3) hence,a plot of I versusω1/2that is linear and passes through the origin is a good test of diffusion control.

Cabelka et al.[9]and Wei[9,10]have performed an electro-chemical study of arsenic reactions using cyclic voltammetry. The investigation reported here con?rms some of the results pre-viously obtained and provides experimental veri?cation of some reactions among arsenic species.Particular attention is given to

0013-4686/$–see front matter?2007Elsevier Ltd.All rights reserved. doi:10.1016/j.electacta.2007.04.071

F.Brusciotti,P.Duby/Electrochimica Acta52(2007)6644–66496645

the oxidation of As(III)to As(V)and the corresponding reduc-

tion.Constant potential studies and analysis of the solutions are

performed to check that the mentioned reactions take place. 2.Experimental setup

2.1.Cyclic voltammetry and rotating disk electrode

The cyclic voltammetry experiments were carried out using

solutions of As(III)and As(V)at different ratios.As(III)solu-

tions were prepared by dissolving solid arsenic trioxide(As2O3,

Fisher,Certi?ed)in the minimum amount of4M NaOH and then

adjusting pH with H2SO4.As(V)solutions were prepared dis-

solving solid sodium hydrogen arsenate(Na2HAsSO4·H2,Alfa Aesar,ACS)into distilled water and then adjusting pH with

H2SO4.Solutions referred to as“blank”contained no arsenic

and were prepared using sulfuric acid or sodium hydroxide,

according to pH requirements.High purity nitrogen gas was

purged into the solutions before applying the potential.All the

experiments were performed at room temperature(24±2?C).

A three-electrode system was employed.The working elec-

trode was a Pt RDE(Pine)with an exposed area of0.5cm2

mounted to a Pine rotator.A platinum electrode was the counter

electrode and a saturated calomel electrode(SCE)was the

reference one.Cyclic voltammograms were obtained using a

potentiostat(Pine AFRDE5)together with a signal genera-

tor(PAR M175),which controlled the initial,switching,and

lower potential parameters for each experiment.A linear poten-

tial sweep was imposed on the working electrode through a

potentiostat and the current at the electrode/electrolyte interface

recorded by a computer,using the LabView software[10].

2.2.Constant potential study

Solutions of As(III)and As(V)equal to1000ppm each were

used in the second part of the study,prepared as described above.

The pH of all starting solutions was between1and2.

Potentiostatic experiments were performed on a Pt electrode

immersed in As(III)or As(V)solutions.In one case,the anode

and the cathode were both Pt electrodes immersed in the As(III)

solution in an undivided cell,as the one described above.In the

other experiments,the anode was a Pt electrode with an area

of～52cm2(considering both sides)and the cathode was a Cu

foil or Pt wire.The cell had two compartments separated by a

glass frit.In both cases,the potential was applied by a power

supply and the resulting current measured and recorded through

a computer software(LabView).

In these experiments,the reference was a saturated calomel

electrode,and N2was sparged into the solution.The initial

and?nal solutions were analyzed for As(III)by titration with

iodine[11]when its concentration was quite high,in the range

of1000ppm;at the conditions of the experiments,the sensitivity

of this method was estimated at7.5ppm.When As concen-

tration was lower(up to0.5ppm)the total arsenic has been

analyzed by graphite furnace atomic absorption spectropho-

tometer(GFAAS),with a sensitivity of5ppb.Table1

Reversible potential of reactions among arsenic species,calculated at As(III) and As(V)concentrations equal to1000ppm

Reaction E(V SCE)

at pH1.20

E(V SCE)

at pH4.70 H3AsO4+2H++2e?→HAsO2+2H2O+0.262–

H2AsO4?+3H++2e?→HAsO2+2H2O–?0.008 HAsO2+3H++3e?→As+2H2O?0.103?0.309

H3AsO4+5H++5e?→As+4H2O+0.043?0.163 As+3H++3e?→AsH3?0.551?0.757 HAsO2+6H++6e?→AsH3+2H2O?0.326?0.533

H3AsO4+8H++8e?→AsH3+4H2O?0.422?0.628 3.Results and discussion

3.1.Cyclic voltammetry study of arsenic

In the potential range(?1.2to+1.5V SCE)and at the solution pH(1.20and4.70)of these experiments,the stable forms of arsenic are elemental As,As(III)as HAsO2,As(V)as H3AsO4 (pH<3.60)or H2AsO4?(3.60

Fig.1shows the voltammogram of a solution containing As(III)=1000ppm+As(V)=1000ppm and that of a blank,both with sulfuric acid at pH1.20.All peaks related to the blank solution are referred to with numbers,while all peaks on the voltammograms with arsenic solutions are referred to with let-ters.Six main peaks(A to F)can be detected on the curve for the arsenic solution.Peaks E and D seem to correspond to peaks 2and3,respectively,for oxygen evolution and the reduction of the oxide layer,even though they are slightly shifted to a more negative potential than the blank.The baseline current preceding peak C is positive with about the same intensity of peak1,which represents the formation of the PtO layer.Peak C appears at a high anodic potential consistent with the oxidation of As(III)to As(V).On the cathodic side of the voltammogram, the large negative current5,which represents hydrogen evolu-tion on the blank scan,disappears or is strongly reduced

and

Fig.1.CV of a Pt RDE in1000ppm As(III)+1000ppm As(V)at pH1.20,in comparison with a blank solution(ω=1000rpm,ν=500mV/s).

6646F.Brusciotti,P.Duby/Electrochimica Acta52(2007)6644–6649

Table2

Electric charge from area of respective peaks

Peak(letter)Tot.charge(×10?3C/cm2)

3 1.02

D 1.15

A0.88

A F 1.63

B0.80

C 3.83

shifted towards more negative potentials,as peak F,in presence of dissolved arsenic.Peaks B and A correspond most likely to the reduction of As(III)to As(0)and the oxidation of As(0)to As(III),respectively.

The features of the voltammogram are in agreement with the results of previous investigations[9,10].The position of the oxidation potential of Pt relative to peak C is consistent with the electrocatalytic mechanism proposed by Cabelka et al.for the oxidation of As(III)[9].Furthermore,the cathodic peak3 (reduction of PtO)of the blank has become peak D,as con?rmed by the data in Table2,which show that the electric charge esti-mated from the areas of the two peaks is the same for both of them.Peak D is at a potential of about0.2V more negative,after oxidation of As(III)to As(V).This could be explained by a slow reduction of the same intermediate,PtOAs(OH)3,suggested in the electrocatalytic mechanism.

The amount of electric charge was calculated from the esti-mated areas of the various peaks and reported in Table2with a precision estimated at0.08×10?3C.Assuming a?lm of PtO, its reduction requires two electrons per platinum atom with a surface concentration of1.31×1015Pt atoms/cm2[13],or an electric charge of0.42×10?3C/cm2.The estimated elec-tric charge of peak3,divided by the above number,yields a roughness factor of2.4comparable to that observed by others [14].

Within the precision of the measurements,peaks A and B have the same area and they may be interpreted as the deposition (B)and dissolution(A)of As(0)on the platinum electrode.The concentration of arsenic atoms on the electrode,based on a two-dimensional compact hexagonal lattice with a=3.76×10?8cm [15],is1.6×1015As atoms/cm2.For a reaction involving three electrons,the charge is0.77×10?4or1.84×10?3C/cm2after accounting for the roughness factor of the electrode.The mea-sured value would indicate a coverage of about50%.

When the voltammetric scan extends to?1.2V(Fig.2),peak A is enlarged.The amount of charge estimated for this peak(A F) is approximately twice than that for peak A.This could be due to the presence of more arsenic on the platinum electrode,resulting from reduction to As(?III)at the lower potential(peak F)and the chemical deposition of more arsenic:As(?III)+As(III)→2 As(0)[16].

Fig.3shows that the voltammograms for a solutions with As(III)only is almost a perfect overlap of that with As(III)and As(V).The voltammogram of the solutions with As(V)only completely overlaps with the one for the blank.This indicates that As(V)is not electrochemically active in the studied

potential Fig.2.CVs of a Pt RDE in1000ppm As(III)+1000ppm As(V)solutions at pH 1.20,at different potential ranges(ω=1000rpm,ν=500

mV/s).

Fig. https://www.360docs.net/doc/f15068047.html,parison of CVs of a Pt RDE in As(III)=1000ppm,in As(V)=1000ppm,and in a As(III)=1000ppm+As(V)=1000ppm solution, at pH1.20(ω=1000rpm,ν=500mV/s).

range(?0.6to1.5V SCE).It is worth noticing also that As(V) does not modify the hydrogen evolution.

Figs.4and5show the effect of varying the concentration of As(III)as well as the potential scan range.Peak C depends on As concentration both in intensity and potential,con?rming the oxidation of As(III)to As(V).Peaks A,B and D are not affected by the amount of As(III)in the solutions at500and1000ppm levels,but all three peaks are smaller when the concentration goes down to100ppm(see Fig.4).Fig.5indicates that peak A depends on the magnitude of peak F,as discussed above,rather than on arsenic concentration.Peak B is associated with the reduction of As(III)to elemental arsenic,while peak F to

the Fig. 4.V oltammogram of As(III)solutions at different concentrations (ω=1000rpm,ν=500mV/s).

F .Brusciotti,P .Duby /Electrochimica Acta 52(2007)6644–6649

6647

Fig.5.V oltammogram of As(III)at different concentrations and potential ranges (ω=1000rpm,ν=500mV/s,pH

1.20).

Fig.6.CV of As(III)at pH 4.70and different concentrations (ω=1000rpm,ν=500mV/s).

simultaneous evolution of hydrogen and arsine,as con?rmed by further experiments performed by the author [17].Peak A,on the anodic side,could represent the dissolution of the arsenic layer,and of eventual arsine formed at peak F.

Fig.6shows that the behavior of As(III)is not signi?cantly affected by increasing the pH to 4.70.The anodic peak C occurs at about the same potential (see Table 1)and its amplitude increases with the arsenic concentration.Between ?0.6and +0.6,the voltammetric scan differs little from the blank at the same pH,except for a small cathodic peak B,attributed to reduc-tion to elemental arsenic.As expected from the data in Table 1,hydrogen and/or arsine evolution do not occur in the potential range of the scan.The voltammogram (Fig.7)in the presence of As(V)almost overlaps the one of the blank.3.2.Constant potential study of oxidation of As(III)to As(V)

Fig.8shows a plot of current density versus potential for a slow anodic scan (0.33mV/s)performed on Pt in

an

Fig.7.CV of As(V)=500ppm solutions at different pH values (ω=1000rpm,ν=500

mV/s).

Fig.8.Anodic scan of As(III)=1000ppm on Pt electrode (ν=0.33mV/s).

As(III)solution.The graph is consistent with the cyclic voltam-mograms.The peak current density is 6.44mA/cm 2and the reaction potential is 0.75V SCE ,which is much higher than the reversible potential calculated for the corresponding reaction as ?0.008V SCE (see Table 1),thus indicating a large overpotential (～0.8V).

Table 3summarizes the main parameters and results for the experiments at constant potential:pH,working electrode mate-rial and area,potential applied,length of experiment,average current,concentration of As(III)in the initial and ?nal solu-tions;the theoretical amount of As(V)that should form at the anode given the duration of the experiment and the measured average current is calculated using Faraday’s law.

The titration of the ?nal solution after electrolysis in the undi-vided cell,at a potential of 1V SCE for 46h,shows that the initial and ?nal concentrations of As(III)were the same,within the range of titration sensitivity.Analysis of the nitric acid solution used to wash the Pt counter electrode proves that some arsenic (36?g)deposited on its surface.While the only possible process taking place at the anode potential is the oxidation of As(III)to As(V),an amount of the latter is reduced at the cathode,either directly to As(0)or with simultaneous arsine formation.

Table 3

Parameters and results of As(III)→As(V)investigation pH Work.el (cm 2)E applied (V SCE )Time (h)Avg I (mA)As(III)in (ppm)As(III)?n (ppm)As(V)theoretical (ppm)Notes

1.40Pt (0.483)+1460.21004±71004±72636?g As deposited on counter el.

1.70

Pt (～52)

+1

2.22

14

1009±7

794±7

217

Divided cell.CE:Cu foil.Manual device

6648

F .Brusciotti,P .Duby /Electrochimica Acta 52(2007)6644–6649

Table 4

Veri?cation of Koutecky-Levich equation

Koutecky-Levich equation:i ?1=(0.62nFD 2/3ν?1/6C ω1/2)?1+(nFKC )?1nF (mC/mol)

ν(cm 2/s)C (mol/cm 3)Slope Intercept (mA/cm 2)?1D (cm 2/s)k (cm/s)Peak C

1.9E+08

1.E ?02

1.33E ?05

0.5372

0.0191

1.3E ?05

0.0204

Fig.9.CV of a Pt RDE in 1000ppm As(III)+1000ppm As(V)in acid,at different rotating speeds (ν=500mV/s).

In the experiment performed with a two-compartment cell and a larger Pt anode,titration of the initial and ?nal solutions shows that 215ppm of As(III)have been removed.This con-?rms the hypothesis that peak C represents oxidation of As(III)to As(V).The amount of As(III)removed from solution is also in agreement with Faraday’s calculations (Table 3),for the oxi-dation of As(III)to As(V)under an approximate average current of 14mA.

The last experiment has been carried out with a potential equal to ?0.3V SCE applied to a Pt electrode immersed in an As(V)solution.The titration of the ?nal solution shows that no As(III)forms in the process (within the detection limit of 7.5ppm).This is in agreement with the results obtained with the CV analysis in the previous section,where As(V)solutions do not show any activity in the potential range under investigation (?0.6to 1.5V SCE ).

Fig.9shows the voltammograms of a solution containing 1000ppm As(III)+1000ppm As(V)at pH 1.20with a scan rate of 500mV/s and different rotating speeds,varying from 0to 3600rpm.Peak C is the one where the effect of rotating speed is measurable.The other peaks are small and only peak B shows an effect while A and D do not seem

affected.

Fig.10.Plot of i ?1vs.ω?1/2for peak C,to verify Koutecky-Levich equation.

Fig.10shows plots of i ?1versus ω?1/2for peak C to verify the Koutecky-Levich (Eq.(2)above).The curve indicates that there is direct proportionality between the two parameters,and the intercept on the i axes is different from zero.Following the discussion above,this means that there is a mixed control (ion transport and kinetics)in the reaction taking place at this poten-tial [8,9].A line has been ?tted through the points obtained and an equation for the resulting line is shown in the ?gure.From the value of the slope,using Eq.(2)above,an estimate of the diffu-sion coef?cient has been calculated.Table 4shows the results of all calculations and the values of the parameters used in the equa-tion:all the parameters are experimental,except for νk ,which has been estimated (×10?2cm 2/s)[18].It has been assumed that the reaction taking place at peak C is As(III)→As(V)(hence n =2).The value obtained for the diffusion coef?cient D (1.3×10?5cm 2/s)is in the same range and comparable to typical values in the literature (for other reactions)[18].The rate transfer coef?cient k has also been calculated,from the value of the intercept of the ?tted line,and it is equal to 0.0204cm/s,in the same range as the one calculated by Cabelka et al.[9].4.Conclusions

Cyclic voltammetry and rotating disk electrode techniques are used to investigate the behavior of arsenic in aqueous solu-tions.The study covers a potential range between ?0.6and +1.5V SCE with particular attention given to the As(III)/As(V)redox process.A fully developed anodic peak starting at about 0.7V SCE and peaking at 1.0V SCE is observed,which represents the oxidation of As(III)to As(V).The peak current is propor-tional to the concentration of As(III)and not affected by the presence of As(V).The large overvoltage and the relationship of the current to the rotation speed indicate that the oxidation is under mixed control of mass transport and charge transfer.The process has been also con?rmed by a potentiostatic exper-iment on a platinum anode at 1.0V SCE in a divided cell,which yielded a decrease in As(III)concentration of 215ppm compared to 217ppm predicted by Faraday’s law.

There is no cathodic peak that can be related to the anodic one and the reaction is irreversible in the investigated potential range.Solutions containing As(V)also con?rm that it is electro-chemically inactive in the investigated potential range and that it does not affect hydrogen evolution.

The presence of As(III)modi?ed the evolution of hydrogen;it occurs at a lower potential and is accompanied by arsine evo-lution.A cathodic peak at about ?0.2V SCE and a corresponding anodic one at +0.3V SCE have the same area within the precision of the measurement and they are attributed to the deposition of a thin layer of arsenic by reduction of As(III)and its dissolution.

F.Brusciotti,P.Duby/Electrochimica Acta52(2007)6644–66496649

The anodic peak is larger when the scan is extended towards more negative potential and As(?III)is formed concurrently with H2evolution.

References

[1]K.Polat,M.L.Aksu,A.T.Pekel,J.Appl.Electrochem.32(2002)217.

[2]R.Wartena,J.Winnick,P.H.Pfromm,J.Appl.Electrochem.32(2002)725.

[3]J.Prentice,Electrochemical Engineering Principles,Prentice Hall,1991.

[4]J.M.Maciel,S.M.L.Agostinho,J.Appl.Electrochem.30(2000)981.

[5]A.Neville,T.Hodgkiess,A.P.Morizot,J.Appl.Electrochem.29(1999)

455.

[6]J.O.M.Bockris,A.K.N.Reddy,Modern Electrochemistry,vols.1and2,A

Plenum/Rosetta Edition,New York,1977.

[7]A.J.Bard,L.R.Faulkner,Electrochemical Methods.Fundamentals and

Applications,second ed.,John Wiley&Sons,Inc.,2001.

[8]P.A.Christenses,A.Hamnett,Techniques and Mechanisms in Electro-

chemistry,B.A.Professional,Chapman&Hall,1994.

[9]T.D.Cabelka,D.S.Austin,D.C.Johnson,J.Electrochem.Soc.131(7)

(1984)1595.

[10]Z.Wei,P.Somasundaran,J.Appl.Electrochem.34(2004)241.

[11]D.A.Skoog,Fund.Anal.Chem.,second ed.,Holt,Rinehart and Winston,

Inc.,1969.

[12]D.D.Wagman,et al.,J.Phys.Chem.Ref.Data11(Supplement No.2)

(1982).

[13]J.P.Hoare,The Electrochemistry of Oxygen,Interscience

Publishers—John Wiley and Sons,1968.

[14]M.W.Breiter,Electrochemical Processes in Fuel Cells,Springer Verlag,

1969,p.58.

[15]American Society of Metals,Metals Handbook,D.Edition,1985,p.2.

[16]A.J.Bard(Ed.),Encyclopedia of Electrochemistry of the Elements,vol.II,

Marcel Dekker,Inc.,New York,1974,p.21.

[17]F.Brusciotti,Arsenic removal from water.An Electrochemical Investiga-

tion,in Earth and Environmental Engineering,Columbia University,New York,2006.

[18]R.C.Weast(Ed.),Handbook of Chemistry and Physics,58th ed.,CRC

Press,West Palm Beach,1977–1978.