KINETIC STUDY OF THE ELECTROCHEMICAL PROCESSES OF THE BROMINE/BROMIDE AQUEOUS SYSTEM ON

VITREOUS CARBON ELECTRODES

MARINA MASTRAGOSTINO and CARLA GRAMELLIM

Istituto chimico CXiamician UniversitB degli Studi di Bologna, Via S.&xi 2, 40126 Bologna, Italy

(Recked I May 1984; in reuisedform 14 Atlgust 1984)

Abstract-The electrochemical processes of the aqueous Br,/Br- system have been studied, by tbe method of the rotating disc electrode, on two different vitreous carbon eleetrodas: reticulated vitreous carbon RVC E.R.G. Inc., Lowell, Oakland, Calif.) and smooth vitreous carbon CVJ (Tacussel Lyon France). On both electrodes the cathodic and the anodic processes involve two consecutive electrochemical steps:

Br,+e*Br.adr+Br- (I) Brads+e+Br-. (2)

The step (1) is ratecontrolling in the cathodic process and the step (2) in the anodic process. The Br; reduction occurs via formation of Br,. with which Br; is in rapid equilibrium, whereupon Br, is

reduced &cording to mechanisms (1) and (2). -

INTRODUCTION

In a preceding paper[ l] we have proposed a model for the Zn-Br, battery designed to overcome the self discharge phenomenon associated with the high sol- ubility of Br, in aqueous electrolytes. We have pro- posed a polymeric salt as bromine complexing agent and reticulated vitreous carbon RVC? (trademark) as electrode material for the redox couple Brs/Br-.

RVC, an inexpensive material of recent development and now manufactured by E.R.G. Inc. (Lowell, Oakland, California) for non-electrochemical appli- cations, is available in several porosity grades. This material has received increasing attention as electrode material during the past few yearsC2-41. However much still has to be done to characterize this material and to date no kinetic study of electron-transfer reactions has been carried out with a RVC electrode.

The vitreous (glassy) carbon is an artificially pro- duced material and it is not possible to define an absolute vitreous carbon state. Fairly rmntly it has been shown[5,6] that the kinetic parameters of redox reactions on vitreous carbon electrodes from several manufacturers are different and this is related to the fact that the material is prepared at different tem- peratures from a range of polymeric resins[6,7]. Nevertheless the present state of understanding of this subject is far from satisfactory and the situation is still one of compiling empirical data.

In the present paper we report the results obtained in a kinetic study of the electrochemical processes of the bromine/bromide aqueous system on two different vitreous carbon electrodes: RVC and CVJ Tacussel (France). The latter is a smooth vitreous carbon, normally used in electrochemistry.

This study has been carried out using the rotating disc electrode method. The reduction of Brz and the oxidation of Br-, in the presence respectively of Br- and Br,, in a concentration range where Br; is

negligible, have been investigated. Also the reduction process of Br; has been investigated, because undet typical operating conditions of the Zn-Br, battery, where concentrated bromide solutions are used, Br; is the predominant Br, containing species. The kinetic parameters: reaction orders, exchange current density, Tafel slopes, transfer coefficients, polarization re- sistence and apparent rate constants were determined. Possible mechanisms for the electrode reactions are proposed and discussed in the light of the results obtained.

The objectives of this study are to calculate the rate constants of the electrochemical processes on the actual material utilized in a model of Zn-Br, battery and to determine what is the electrochemical response of RVC with respect to that of vitreous carbon, normally used electrochemically.

EXPERIMENTAL

An EDI Tacussel (Lyon, France) rotating electrode has been employed with rotation speed between 25+3000 rev min- ‘. The rotation speed was con- trolled to better than 1 y0 by an independent servocon- trol electronic amplifier Controvit, Tacussel. In the experiments on smooth vitreous carbon a disc of 2 mm diameter type CVJ Tacussel has been used. The electrode itself is enclosed in a small interchangeable ferrule made of Teflon; the working surface (the actual disc) is flush with the end of the ferrule. In the experiments on RVC the rotating RVC electrode design is similar to that reported in[8]. A cylinder of the type RVC-S100 (100 pores per inch) 2 mm dia- meter and 2.75 mm long, supported on a graphite disc by means of epoxy glue has been used. The RVC electrode once prepared was kept at 450 K in air for 2-3 h. Only in this way experimental values of active area corresponding to those calculated from the

373

314 MARINA MASTRAGOSTINO AND CARLA GRAMELLN

geometrical dimension of the electrode and using[Z] a surface area of 66cm2/cm3 for RVC-SlOO, were ob- tained. The same RVC electrode was used in all experiments over a period of more than six months. A saturated calomel electrode (see) was used as reference and all the reported potentials are referred to see. The current-voltage curves were obtained with an Electrochemolab (Milan, Italy) apparatus using a sweep rate of 5 mVs_‘.

For the sake of reproducibility of the polarization measurements, the cathodic curves had to be made scanning the potential from -0.5 to + 1.0 V, while a pre-electrolysis of the working electrode at -0.5 V for 1 min was required for anodic curves before scanning from +0..5 to more positive potentials. The tension was never set above + 1.5, not to produce irreversible transformation of electrode surface[7].

In the study of reduction of Br, and of the oxidation of Br-, 0.5 M NaCIO, was used as supporting elec- trolyte. Test solutions were made up with reagent- grade chemicals (NaBr, Br,, NaClO,, C.Erba) and twice distilled water. All experiments were carried out at 283 f 0.2 K.

The experimental data have been elaborated with help of a Commodore model 3008.

RESULTS AND DISCUSSION

Reduction of Br, and oxidation of Br-

On the system Br,/Br-, at a concentration range where Br; is negligible, the cathodic and anodic polarization curves were carried out on RVC and CVJ electrodes, at different rotational speed. Figures 1 and 2 show the cathodic and anodic currents, respectively on RVC and CVJ, as a function of oLf2 for selected fixed overvoltages. In the anodic process the potential of the limiting current region was not attained on RVC, to avoid the possibility of damage to the electrode at positive potentials (in consideration of the impossibility to mechanically polish a porous electrode).

The limiting currents, illm. are mass-transfer-limited, indeed through the origin straight lines are obtained, according to the Levich convective diffusion equation. When the potentials are fixed below the limiting current region the cathodic and anodic processes are partially controlled by the mass transfer and by the kinetics of the electrodic reactions and i us 0’1~ plots are not linear. In order to determine the reaction order with respect to Br, (I~~,) for the cathodic process and with respect to Br- (zBr-) for the anodic process, the cathodic and anodic current values at constant q, higher than 100 mV, have been analysed on the basis of the following equation:

1 I 1 -=_++ i i, Bw”Z

which holds for a reaction order equal to 1, where i,, is the kinetic current unaffected by mass transfer (i for w--r CO) and 8, for a given concentration of the diffusing species, is a constant, its value being ili,/o’ ‘*. The values of i-’ plotted, at constant q, against o-~‘~, give parallel straight lines. As shown in Table 1 the reciprocal of the slope of these straight lines (B) agrees well with the value of the slope of i,, vs LU”’ plot (a,)

a

d

b

m

Fig. I. Cathodic and anodic currents on RVC electrode trs the square root of the rotational speed, at different over- vo!tages. Cathodic process: (a) 180, (b) 218, (c) 255, (d) 305, (e) 355 mV and (f) limiting current. Anodic process (a’) 320, (b’) 358, (c’) 395, (d’) 433, (e’) 470 mV. [Br,] = 7.1 x 10s4 M,

[Br-] = 4.2 x lo- M.

Fig. 2. Cathodic and anodic currents on CVI electrode 1)s the square root of the rotational speed, at different overvoltages. Cathodic process: (a) 155, (bj 180, (c) 205, (d) 230, (e) 255 mV and (f) limiting current. Anodic process: (a’) 275, (b’) 300, (c’) 325, (d’) 350, (e’) 375 mV and (f’) limiting current. [Brz]

= 7.0 x lo-’ M, [Br-] = 8.2 x 10m5 Ad.

Bromine/bromide aqueous system of vitreous carbon electrodes

Table 1. Test for the validity of the 1st order analysis

37s

Br, reduction Br- oxidation System

LB;] LB;-1 B BL B BL

Electrode /LA s”2 pA sl’* /LA sl’* @A s”2

7.1 x 10-4 4.2 x lo-” RVC 52.4 + 1 50.2 + 0.5 250+3 242 7.0x 10-4 8.2 x 10-S CVJ 2.52f0.01 2.49 + 0.02 23.6 + 0.05 23.9 + 0.4

(a) The equilibrium concentration of Br; under these conditions is negligible. (b) B is the average. of 15 values, obtained by least squares analysis, at different 7. (c) Thecomparison of the E, values on RVC and CVJ, for Br, reduction, indicates that the active area of the RVC

electrode corresponds to that evaluated from the geometrical dimensions of the electrode on the basis of a surface area of 66 cm2/cm3.

(d) For the Br- oxidation on RVC, on assuming that a limiting current: ib = BLO”’ could be observed at E B 1.5 V, the corresponding BL value has been cdculated from the EL of Br, reduction taking into account the ditTusion coefficients ratio (deduced by experimental results on CVJ) and the difYerent concentration.

for the same system. This fit demonstrates that z~,, = 1 for the cathodic process and zBr- = 1 for the anodic process.

Moreover the good agreement of the results with the kinetic analysis of the first order in case where [Br-] $= [Brz] suggest that in the anodic process the reaction order with respect to Br, is zero. The reaction order with respect to the products, for the cathodic and anodic process are clearly shown by Figs 3 and 4, where the values of log i,, (obtained by plot i-’ us o-1/z) of two systems with the same concentration of the electroactive species and different concentrations of the products are plotted in either figure against the corresponding electrode potentials. The values of each system are placed on the same straight-line, which shows that the reaction order w.r.t. the products are zero.

The i, values (obtained by plot i-’ US cK1”) cathodic and anodic at different q (higher than lOOmV), have been analysed with least squares fitting

I I

05- \

i 2.8 3.0 3.2 3.4 3.8 3.8

log lm.*

Fig. 3. RVC electrode. Cathodic and anodic E us log I, (I = pAcmm2). (a) Cathodic: [Br2] = 7.1 x IO-’ M; l , [Br-] = 4.1 x lo-” M; o, [Br-] = 5.0 x lo-‘M. (b) Anodic: [Br-] = 4.2 x IO-’ M; o, [Br*] = 7.1 x 10m4 M; l ,

[Br2] = 1.4 x IO-’ M.

on the basis of the Tafel equation in the following form:

2.3RT 2.3 RT tl’ aFlogio -7logi_+Ri_

to consider the possible uncompensated resistance, where G and & are the transfer coefficients of the reduction and oxidation overall reaction. The systems of the least squares analysis were usually of 15 equations and the differences between experimental and calculated PJ were less than 2 mV. In Tables 2 and 3 the values of exchange current density (I,), Tafel slope (2.3 RT/aF), & & obtained under different experimental conditions on RVC and CVJ electrodes are reported. It must be noted that for the same system the analysis of thecathodic and anodic i mX gives different values of I,, as shown also in Figs 5 and 6, where plots of log i, against q, corrected for ohmic potential drop, are reported.

The current density us overvoltage curves in the range of very low overvoltages (1 f 1 15 mV) are exactly straight lines, as shown in Figs 7 and 8, hence the two directions of the electrode process are symmetrical. A few values of the polarization resistance are reported in Tables 2 and 3.

A comparison between the experimental kinetic parameters of Tables 2 and 3 and the values expected on the basis of the most likely mechanisms[4, 51 for the overall reaction: Br, + 2e $2Br -, listed in Table 4 can be carried out.

The fact that the cathodic and anodic transfer coefficients do not add up to unity (o! + & # 1), since a = p and ‘a = 1 -/3 for the mechanisms that appear acetable for the theoretical-experimental agreement of orders and for the high values of the Tafel slopes, suggest that all the mechanisms where the same /? appear in the reduction and in the oxidation process can be rejected. Therefore the following mechanisms, that also justify the different cathodic and anodic current densities, must be considered as possible on RVC and CVJ electrodes:

(b)

reduction process: Br, + e -P Brads + Br-, as rate determining step, followed by electrochemical (21 or chemical (2”) steps, with 8 + 0. oxidation process: Br =Br,, + e, as r.d.s., fol- lowed by electrochemical (- 1) or chemical ( - I’) steps, with 8 +O.

MARINA MASTRAGOSTINO AND CARLA GRAMELLINI

I 1 1 1

2.8 3.0 32 3.4 3.6 3.6 4.0 4.2 44 46 4e

log Imnx

Fig. 4. CVJ electrode. Cathodic and anodic E vs log1 (I = rAcn_‘). x lOme M; l , [Br-] = 8.9 x 10m3 M, o, [Br-] = 1.0 x 10% M. (b) Anodic:

[Br,] = 7.3 x lo-* M; o, [Brl] = 7.1 x 10e5 M.

(a) Cathodic: [Br,d = 7.1 [Br-] = 8.2~ IO- M; .,

Owing to the identical polarization resistence in the cathodic and anodic direction, with different values of the exchange cathodic and anodic current densities, it is reasonable to reject, among the possibilities ex- amined, those in which a chemical step follows an electrochemical step, that is different in the reduction and in the oxidation process.

In addition, the exchange current densities give a good fit of Vetter’s equation which is valid for two consecutive electrochemical steps:

/arr\ RT / 1 1\

where 1: and 19 are the half of the values obtained by extrapolation of Tafel’s plots to q = 0. Therefore we propose the following mechanism:

Br, +e z+ Brads + Br - (1)

Br&+e+Br- (2)

where, for B -, 0, step (I) is rate determining in the cathodic and step (2)in the anodic process respectively, both on RVC and CVJ.

En Table 5 are reported all the kinetics parameters obtained in the Br; reduction on RVCand CVJ, while in Figs 9 and 10 a few typical experimental results obtained on RVC are shown. Figure 9 shows that the limiting current is mass-transfer-limited and thar. the current flowing during the development of the catho- dic curve is partly mass-transfer controlled and partly controlled by the kinetics. Figure 10 shows that the reaction order with respect to Br- is - 1 (the exger-

imental value is - 1.15, considering an average Tafel slope).

The limitingcurrent controlled by mass transfer, the reaction order with respect to Br; , zBr = 1 and with respect to Br-, I ar- = - 1, and the high values of the Tafel slope, analogous to those obtained in the Br, reduction, suggest a mechanism where the Br; reduc- tion occurs via formation of Br,, with which Br; is in rapid equilibrium, whereupon Br, is reduced accord- ing to mechanism (1) and (2).

CONCLUSION

On the basis of the proposed mechanism the apparent rate constants, at the standard potential of the couple Br,/Br-, ks.h, for the process investigated have been calculated. The average of the k,,, values obtained in the different experiments for each process examinated on RVC and CVJ are listed in Table 6.

As shown from Table 6, the reduction of Br, and the oxidation of Br- is slightly slower on RVC than on CVJ, but the ratio: k r,htptbodic Br,)/k,h(anodic e-1 is the same on both materials. In the Br; reduction while k s,h(cathodii Br,) - shlcathdic Br,)

s,h(cathodic B.1;2i RVC.

on CVJ, k >k

s,k(calhcdw Br;)

The difference between the rates of the processes on RVC and CVJ are to be ascribed[6,7] to the different origin (temperature of treatment, starting material, preparative procedure) of the two materials, but this difference does not affect the reaction mechanism.

On the basis of the results discussed above we can therefore conclude that the electrochemical response of RVC is not very different from that of CVJ.

Tabl

e 2. K

inet

ic pa

ram

eter

s ob

tain

ed un

der d

iffer

ent e

xper

imen

tal co

nditi

ons o

n R

VC

elec

trode

Br,

redu

ctio

n B

r- o

xida

tion

7.1 x

10-Q

2.

0 x 1

0-3

9809

5 1 0 -2

03f8

0.

28 f

0.01

70

22

7.1 x

10-

4 4.

1 x IO

-’

965&

5

7.1 x

10-

4 5.

0 x lo

-’

985$

5 1 0 -2

34f9

0.

24f0

.01

65&

l

7.1 x

10-

4 5.

0x 1

0-4

985i

5 1 0 -2

43&

4 0.

23 +

0.04

55

&l

7.1 x

10-

4 4.

2 x lo

- 3

9752

5 0 1 24

0*24

0.

23yl

.02

64f4

1.4 x

10-

4 4.

1 X 10

-3

960+

5 0 1 24

9 f 3

5 0.

22 f

0.02

50

f5

1.6x

lo-”

4.

2 x lo

-’

955f

5 0

7.1 x

10-

q 20

x 10

-3

980f

5 0 1 20

2f12

0.

28 it

0.02

24

&4

4 ;, -2

51 f

4 0.

22 +

0.02

80

&l

l

239;

34

0.23

+0.0

2 35

f3

(8.5

fl) X

102

(8.5

fl)

x 10

1 -

- -

- -

-

l Th

e val

ues o

f th

e eq

uilib

rium

pote

ntia

l hav

e bee

n de

term

ined

from

the

polu

rizat

ion c

urve

s for

i =

0.

Tabl

e 3.

Kin

etic

pam

eter

s ob

tain

ed un

der d

iffer

ent c

ondi

tions

on C

VJ e

lect

rode

Br2

redu

ctio

n B

r - o

xida

tion

7.1 x

10-

4 LO

X lo

-’

995&

5 1 0 -1

23f5

0.

46fO

.M

52k7

7.1 x

10-

4 8.

9 x 1

O-3

92

535 I 0

-125

f5

0.45

f 0.

01

137A

f

6.8 x

lo-*

8.

8X 10

-3

9c0&

5 1 0 -1

39s

0.40

f0.0

1 16

1 f6

6.1 x

IO

-’

8.1 x

IO

-”

9155

5 1 0 -1

03f2

4 0.

54f0

.10

120f

4

6.8

x 1W

4 8.

8 x

IO- 3

90

035 0 1

2252

25

0.25

f 0.

03

3982

22

7.0X

IO”

8.3 x

1O

-3

9205

5 0 1 18

3&3

0.31

*0*0

1 25

6&2

7.3 x

10-

4 8.

2 x lo

- 3

915f

5 0

7.1 x

10-

5 8.

2 x lo

-’

880f

5 0 1 18

5k5

0.30

f0.0

1 12

422

t76;

9 0.

32 f

0.02

21

9&4

115

128

132

115

- -

* Th

e va

les

of th

e equ

ilibr

ium

pote

ntia

l hav

e bee

n de

term

ined

from

the p

olar

izat

ion c

urve

s for

i =

0.

378 MARINA MASTRAGCXTINO AND CARLA GRAMELLINI

-400 -300 -200 100 lb0 2bo 3’oa 4bo q/ mY

Fig. 5. RVC electrode. (a) Cathodic and (b) anod~Iic~c~&uization. [BrJ = 7.1 x 1Oe4 M. [Br-] = 2.0

-4bo _3w - 2’00 _rba

/

\

0

2

100 200 300 400 tl /mV

Fig. 6. CVJ electrode. (a) Cathodic and (b) anodic Tare1 polarization. [Brz] = 6.8 x IO-’ M, [Br-] = 8.8 x lo-‘M.

Bromine/bromide aqueous system of vitreous carbon electrodes 379

Fig. 7. RVC electrode. Current density curve ws overvoltage Fig. 8. CVJ electrode. Current density curve US overvoltages at low overvoltages. [31,~~7d x lo-* M, [Br-] = 2.0 at low overvoltages. [Br2,iirS6G x 10e4 M, [Br-] = 8.8

Table 4. Parameters for the cathodic and anodic processes (theoretical)

ZBr, =Ilr - Tafel slope Rate-determinng

Mechanism step 8 Br-+’ 0,-l e,,-0 8,,-1 0 Br -to B Br+’

Cathodic (1) Br,+e-tBrd+Brm 1 - 0 - - 2.3RT/BIF (2) Brlds+e-rBr- I 0 -1 0 -2.3RT/(l +&)F - 2.3RT/& F (1’) Br, + 2Br, I - 0

(2’) Brti+e-rBr- l/2 0 ::3RT,&F - 2.3RT/pz F (I”) Br,+e+Br,+Br- 1 - ; 0 -2.3RT/j?,F (2Y 2BrtiS -. Br, 2 0 -2 0 - 2.3RT/2F -CO

Anodic 1:;; Br- +Brds+e 0 - : 2.3RT/( 1 - f?,)F - Br- + Br& + Br, + e 0 0 1 2.3RT/(2 - &)F 2.3RT/(l -B, )F

(-2’) Br--+Brads+e 0 - I - 2.3RT/(l - &)F - (- 1’) 2Brds + Br, 0 0 2 0 2.3RT/2F cm (- 2”) Br2 -+ 2Br,, 1 0 - (-1”) Br-+Br,,+Br*+e l/2 0 1 1 2q;RT,(l -&)F 2.3RT/(l -b,)F

Table 5. Kinetic parameters obtained under diNerent experimental conditions

Br; reduction RVC CVJ

t3E [B:” ,,M

4.6 1.0 x x 10-S lo-” 4.6 1.0 x x IO-’ 10-5 9.9 4.5 x x lo-+ 10-S 5.4 5.9 x x 10-5 10-a 5.0 1.1 X x 10-5 10-a

1.0 1.0 1.0 0.5 1.0

L&mV 760+5 755+5 750*5 77x55 730& 5 2 Br; 1 1 1 1 1 L.* 1.20 0.99 1.03 0.96 1.10 %- -1 -1 -1 -1 -1

PTAI /mV -215%8 -243+X -225*5 -208*9 -131*13 t 0.26 f 0.1 0.23 f 0.01 0.25 kO.01 0.27f0.01 0.42 * 0.4

I’/pA cm-’ 143*3 183+4 225+3 144*2 111*2

* L = B/B, being assumed as a test for the validity of the 1st order analysis.

MARINA MASTRAGOSTINO AND CARLA GRAMELUNI

0.8

0.4

4 8 12 16 0% /

rad “2 ‘-“7 0.50 I

L-_z 28 3.0 %2 3.4

Fig. 9. Cathodic current on RVC electrode us the square root of the rotational speed, at different overvoltages: (a) 100, (b) 150, (c) 200, (d) 250, (e) 3OOmV (f)limitingcurrent. [Br; ]

= 9.9 x lo-+ M. [Br-] = 1.0 M.

Acknowledgements-This work was carried out with the log Im.x

financial support of the CNR Progetto finalizzato Energetica Fig. IO. Cathodic E vs log I_ (I = PA cm-‘). (a) [Br;] 2, No. 830211759. The technical assistance of Mr. Lino Cludi = 1.0 x lo-’ M, [Br-] = 1.0 M, (b) [Br;] = 5.9 x lo-* M, is acknowledged and appreciated. [Br-] = 0.5 M.

Table 6. Apparent rate constant at the standard potential of the couple Br,/Br-

ks.h[cathodic Br,) k s.h(ancdic Br-) k (cm s-11

s,h(cathodic BI;) (ems-‘) (cm s-l)

RVC CVJ RVC CVJ RVC CVJ

(1.1 +0.1)x lO-3 (2.0+0.1)x IO-” (4.5+0.5)x lo-’ (1.2+0.2)x lo+ (5.6+0.7)x IO-’ 1.4 x 10-S

1.

2. 3.

4.

5.

REFERENCES 6.

M. Mastragostino and S. Valcher. Electrochim. Acta 28. 501 (1983).

7.

J. Wang, Electrochim. Acta 26, 1721 (1981). 8. D. Calasantio, C. Sorana, S. Zampieri and R. Marassi, 9. Ann. Chim. (Rome) 73, 161 (1983). D. J. Curran and T. P. Tougas, Analyt. Chem. 56, 672 lo. (1984). L. Bjelica, R. Parsons and It. M. Reeves, Proc. elec-

trochem. Sot. 1979, 80, 190 (1980). l-l. Gunasingham and B. Fleet, Analyst 107, 896 (1982). L. Bjelica, R. Parsons and R. M. Reeves, Crwtica chemicu Acta 53, 211 (1980). W. J. Blaedel and J. Wang, An&t. Gem. 52, 76 (1980). G. Faita, G. Fiori and T. Mussini, Eleclrochim. Acta 13, 1765 (1968). D. M. Nowak, B. V. Tilak and B. E. Conway, Modern Aspects of Electrochemisrry, No. 14, p. 283. Plenum, New York (1982).