ELECTROCHEMICAL STUDIES OF REDOX SYSTEMS

FOR ENERGY STORAGE

Report No. II

by

Cherng-Dean Wu

Daniel Scherson

Ernest Yeager

Annual Report : December 1981 to November 1982

NASA Project

Contract : NAG 3-219

NASA-Lewis

Cleveland, Ohio 44106

Case Center for Electrochemical Sciences

and The Chemistry Department

Case Western Reserve University

Cleveland, Ohio 44106

1 December 1982

https://ntrs.nasa.gov/search.jsp?R=19830006396 2018-06-01T13:37:43+00:00Z

TABLE OF CONTENTS

Page

I. ABSTRACT OF REPORT 2

II. INTRODUCTION AND OBJECTIVES OF RESEARCH 3

III. SUMMARY OF REPORT 4

IV. DATA AND PRELIMINARY DISCUSSION 6

1. Cr(H00)+/Cr(H00)

+ Couple on Rotating£ D 2. D

Nickel-Mercury Disk Electrode 6

A. Nickel-Mercury Electrode 6

B. The NaCIO, Supporting Electrolyte 7

C. Effect of Halides 12

D. NH.X as Supporting Electrolyte 14

2. Chromic Chloride Reduction in Acid Media on Au,Pb/Au and Graphite Electrode • 17

A. Gold Electrode 17

B. Lead/GoId System 24

i) Deposition and Stripping of Lead onGold Electrode 24

ii) Potential Step Experiment 28

C. Ordinary Pyrolytic Graphite 40

3. Fe(H.O) +/Fe(H.O)f+ Couple, Rotating Gold Ring-Disk2. D t. D

Electrode Studies in Acid Media 42

A. Effect of Halides 46

B. Effect of UPD of Bismuth on Reduction Kinetics 56

4. Spectroscopic Studies 65

A. Time Effects on Solution Composition(Visible Spectra) 65

B. XPS Measurement on Pre-electrolysis: Gold Cathode 65

I. ABSTRACT OF REPORT

Both chromium (II)/ (III) and iron (II)/ (III) couples have been studied

on various electrode surfaces in acidic perchlorate solution by using

rotating ring-disk techniques. It was found that chloride which forms

inner sphere coordination complexes with the redox species enhances the

electrode kinetics dramatically. Those are consistent with the results of

Weaver for chromium (II) / (III) couple and those of Weber for iron (II) / (III)

couple.

The effects of lead underpotential deposition and surface alloy

formation on the kinetics of the chromium (II)/ (III) couple on gold have

been studied using both linear sweep voltammetry and potential step tech-

niques . The lead IFD was found to slow down the kinetics of the reduction

of the Cr (III) species on gold surface although increase the hydrogen

overvoltage. The effect on the chromium(II)/ (III) kinetics can be explained

in terms of principally a double layer effect. The IFD lead species with

its positive charge results in a decrease in the concentration of the

Cr (III) species at the electrode surface. Similar phenomena were also

observed with bismuth underpotential deposition on gold for the iron(II)/

(III) couple.

II. INTRODUCTION AND OBJECTIVES OF RESEARCH

NASA-Lewis has developed a promising redox electrochemical system

for energy storage. This system is based on an electrochemical cell with

two compartments separated by an ion selective membrane. The electrode

reaction at the negative electrode involves the chromium (II)/ (III) couple

while that at positive electrode involves the iron (II)/ (III) couple.

Solutions containing these species are. circulated through carbon felt

electrodes where the reduction of Cr (III) and oxidation of Fe (II) occur

during charging and the reverse reaction occur during discharge.

NASA-Lewis has greatly improved the performance of the carbon electrodes

used for the chromium (II)/ (III) reactions by incorporating gold in the carbon

felt matrix and by adding lead chloride to the solution. The resulting gold-

lead catalyst greatly reduces the overvoltage normally associated with this

redox couple while maintaining the hydrogen overvoltage sufficiently high

to depress the competing H_ generation reaction.

While the gold-lead catalyzed carbon felt electrodes afford high perfor-

mance, a full understanding of the electrochemical and other processes involved

at these electrodes has not yet been achieved. The proposed research is directed

to the study of the factors controlling the kinetics of oxidation-reduction

couples on various electrode surfaces.

During the past 12 months, NASA-Lewis has sponsored research at Case

Center for Electrochemical Sciences on the behavior of the chromium (II)/ (III)

couple on electrode surfaces including gold, lead and graphite. This investi-

gation is aimed at understanding the role of the electrode and electrolyte

components in controlling the electrode reaction kinetics. Such fundamental

understanding should provide the insight, necessary for the optimization of

the present system and for the identification of alternative and perhaps

more effective electrode-electrolyte systems for energy storage applications.

III. SUMMARY OF RESEARCH

The research during the first 12 months of this project is summarized.

The experimental studies have included the following measurements :

1. Linear sweep voltammetry studies

The underpotential electrodeposition of lead has been examinedon gold surface using linear sweep voltammetry. At more cathodicpotential, bulk deposition and alloy formation occur. The under-potential deposition of bismuth on gold surface also has beenstudied.

2. Rotating ring-disk electrode studies of the chromium (II)/ (III)couple

Rotating ring-disk electrode techniques have been used to examinethe effect of the underpotential deposition of lead on the kineticsof the chromium (II)/ (III) couple on gold. This technique permits theseparation of the current components for the redox couple and forthe underpotential deposition of lead. In some instances it shouldalso be possible to examine the reduction kinetics for the chromium(II)/(III) couple in the presence of a substantial 1L generationcurrent using a ring with high hydrogen overvoltage to monitor eitherthe consumption of Cr (III) or the production of Cr(II). The rotatingdisk electrode provides quantitative control of the transport of react-ant and products to and from the rotating disk electrode surface.

3. Potential step studies of the effects of UPD layers on the kinetics ofchromium (III) reduction

These experiments involve polarizing a rotating gold disk electrodeat a potential where UPD does not occur and then stepping the poten-tial to a value where UPD occurs.Ihder these conditions with a lowbulk concentration of the UPD species, the deposition of the UPD layeroccurs very slowly under pure diffusion control. The coverage of theUPD layer can then be calculated as a function of time from the diffu-sion flux. With chromium (III) also present, the dependence of the rateof the chromium (III) reduction on the UPD coverage can be determinedfrom the time dependence of the chromium (III) reduction current.

4. Raman studies of adsorbed chromium complexes on electrode surfaces

Surface enhancement of the Raman signals (SERS) by factors of up to10 are obtained with adsorbed species on metals such as silver andgold. Advantage is being taken of this enhancement to check on thepossible specific adsorption of various inner sphere chromium (III)complexes on these electrode surfaces. So far, preliminary measure-ments have been carried out with the mono- and di-thiocyanate chromium

complexes adsorbed on silver. Strong signals have been obtained,presumably from monolayers, but detailed analyses of the Ramandata have not yet been made. Further work is planned.

5. Ultraviolet-visible absorption spectroscopy

The [Cr(lLO)6]3+, [ Cr (H Cl]2* and [ Cr (I O)4C12]

+ species havebeen monitored using W-visible absorption spectroscopy. Furthermeasurements using UV-visible absorption are planned to establishthe concentration changes for these species as a consequence ofthe electrode processes as well as homogeneous processes.

The voltammetry and polarization measurements have led to the followingconclusions :

1. The reversible potentials differ substantially for the [Cr(ILO),][ Cr (10) Cir and [ Cr ( I O C l couples. i b

3+

2. The UPD lead on the gold slows down the kinetics of the reductionof the various chromium (III) species on the gold surface. This effect isprobably principally associated with a double layer effect. The UPDlead species with its positive charge results in a decrease in theconcentration of the chromium (III) species at the electrode surface(in the outer telmholtz plane).

3. The principal beneficial role of the lead on the reduction of chromium(III) species on gold is to increase the hydrogen overvoltage and thussuppress the competing IL generation reaction.

4. Anion bridging does not play an important role in the reduction of[Cr(H_0) Cl] on gold, probably because the negative charge on thecathode causes this complex to approach the electrode surface in theouter Helmholtz plane with the chloride ligand directed away from theelectrode surface.

5. Anion"~bridging does appear to be important in the oxidation of chromium(II) to chromium (III). The model proposed to explain the kinetics isthat the chloride ion is specifically adsorbed on the electrode surfaceand then transfered into the inner coordination sphere of the chromium(III) ion to yield [ Cr_(lLO)5Cl] . UPD can increase the surface concen-tration of adsorbed Cl and thus promote the formation of the chloridecomplex. In addition to the IPD of lead, the UPD of bismuth has alsobeen examined on gold. Some kinetic measurements have also been carriedout for the iron (II) / (III) couple on the UPD Bi/Au surfaces.

IV. DATA AND PRELIMINARY DISCUSSION

1. Cr(H00)^+/Cr(H00)i:+ Couple on Nickel-Mercury Disk Electrode/ D / D

A. Nickel-Mercury Electrode

The chromium(II)/(III) couple in an acidic aquo solution is a simple

one-electron transfer reaction having very irreversible kinetics in the

absence of bridging ligands. It is necessary to have a high hydrogen

overvoltage electrode to study the reduction kinetics of this couple.

Mercury affords such a high hydrogen overvoltage and has been used for

previous studies of this couple at Case Center for Electrochemical

Sciences. To check out this couple under steady state and well defined

transport conditions, a rotating mercury coated nickel disk electrode

has been used in the early phases of the present study. For practical

purposes the electrochemical surface properities are close to those

of a mercury electrode. The solubilities of nickel in mercury and vice

verse are very low. The surface layer is almost pure mercury.

The mercury coated electrode is easily made and is easy to use. The

following procedure will give a nickel-mercury electrode with high

hydrogen overvoltage. A polished nickel disk is subjected to a few

minutes of ultrasonic cleaning immediately before the mercury-deposi-

tion procedure to remove any organic film and trapped particles on the

electrode. The electrode is then introduced into a solution of ca. 0.5

mM % (NO ) and 0.1 M K10,. A potential of -1.10 V vs SCE is then

applied to the nickel disk electrode and maintained until hydrogen

evolution is apparent from bubble formation. A nickel grid is used as

(1). S. Donovan and E. Yeager, " The Electrode Kinetics of the Chroraous-

Chromic Couple " OXR. Technical Report No. 24, Case Western Reserve

University, Cleveland, Ohio, Oct. 1969.

counter electrode. Deposition is repeated for several times.

The mercury coated surface of the electrode has a reasonable high

hydrogen overvoltage, although not as high as very pure mercury. Actu-

ally the hydrogen overvoltage on a stationary film of mercury was

intermediate between that on smooth solid electrode and on dropping

mercury.

B. The NaCIO, Supporting Electrolyte

The specific adsorption of the ions of the supporting electrolyte

has long been known to influence profoundly the rate of electron-trans-

fer reactions at metal-solution interfaces. A difficulty faced with

-2+ 3+redox couples of the type M /M with high charged ions is that the

double-layer effects in the kinetics are quite significant even in con-

centrated electrolytes. The double-layer corrections are smallest and

can be applied with greatest confidence using supporting electrolytes,

at high ionic strengths in the absence of specific ionic adsorption

with the concentration of the redox species small in comparison with

that of the supporting electrolyte. Perchlorate was chosen because

this anion combines weak specific adsorption with only a very small

tendency toward complex formation in the bulk of the solution. Per-

chlorate ions are adsorbed to a very low degree on conventional elec-

trodes such as Ft, Au and %. Their adsorbability is far lower than

for sulfate ions and, except at higher positive potential, their

adsorption can be neglected. Inner coordination complexes between

chromium ions and the CIO, form only to a very small extent except in

very high concentration of perchlorate. The concentration of chromium

(III) employed was always kept sufficiently below those of the supporting

electrolytes so that these ions made a negligible contribution to the

diffuse-layer structure.

Among the solid electrodes, it seems that a mercury coated electrode

is the only one which can be seen both cathodic and anodic waves of

chromium (II)/ (III) couple in perchlorate electrolyte. In perchlorate

supporting electrolyte, the chromium(II)/(III) couple gives well defined

3+cathodic and anodic waves, as is shown later. The reduction of Cr

proceeds at negative potential where the perchlorate is not signifi-

cantly specifically adsorbed. The reduction and oxidation waves are

well separated and have different shapes that reflect, in part, the

different influence of the diffuse double layer on the reduction and

oxidation reactions, respectively.

Rotating nickel-mercury disk electrode data are shown in fig.2.

(2). K. M. Jones and J. Bjerrum, Acta. Chem. Scand. 19 (1965) 974.

10

oLT>

oo

ooCNJ

current

omCM

oo

oi

oi

O1

00

oI

o•

1—II

wu

w a61 oO

J-lOo

rH CM

o

in s-iT-l 4J•O O

0)rH iH(1) W

C O T3i-l Q)

T3 CJ 4J(U ffl n)w a oCO -HO S T)o e c

U O(-1 <-< ••0) U 0)e* to1C S n0 o cc • oO O -H•H 4J4J C tOU -r\ W3 O

T) rn«SU x-sV4 «* •

U — i a)vO JJ

M • tOO O H

c a)O <U

wQJ>M 4-> CO3 3O rH •

OTMC w go o•rl .AJ (1) CNtO T3 •N O O

CO O COH Q) Q)0 «H l-i

PL, (U CO

tN

0)^J

W•rl '

11

1.5

01

o

H 1.0

0.5

0

-1 -1/2Figure 3: I vs. (jj plots of data in Fig. 2 for reduction of

Cr (ILO) on nickel-mercury electrode.

12

C. Effect of Halides

In perchlorate supporting electrolyte with a pH of 2 or less (to avoid

hydrolysis), the hexaaquochromium (II)/ (III) couple exhibits an irreversible

wave. With halides in solution, the anodic wave of couple is shifted toward

more negative potential. When the supporting electrolyte contains bromide

ions, the couple produces two reduction waves: one corresponding to the

uncomplexed chromium (III) reduction, the other to the bromide complex

chromium (III) reduction. Only a single anodic wave is apparent corres-

ponding to the formation of the complexed chromium (III) from chromium

(II). The bromide is adsorbed on the electrode surface and transfered

into the inner coordination sphere of the Cr. The kinetics are very slow

for the reduction without the bridging bromide ligand. A representative

voltammetry curve with bromide is given in"Fig. 4.

13

E vs, SCE (V)

3+Figure 4: Voltammetry curves for 2mM Cr(tLO)6 on stationary nickel-mercury disk

electrp.de. Electrolyte: 0.01M NaCIO, and 0.01M HC10,. Electrode area:0.2 cm . Sweep rate: 50 mV/s. Temperature: 22°C. A. without NaBr added.B. with 1.67 mM NaBr added. Quasi-steady-state measurement (curves recordedafter many cycles).

14

D. NH,X as Supporting Electrolyte

In 0.1 M ammonium chloride or ammonium bromide, chromium(II)/(III) couple

gives well defined anodic and cathodic waves. The voltammetry curve of chromium

(II)/(III) couple shows a much more reversible behavior in NH.X supporting

electrolyte than that in pure perchlorate solution. While the cathodic wave

potentials are not changed very much and remain almost the same, the anodic

potentials vary with the supporting electrolyte used.

In presence of ammonium halide the anodic wave potential is found to be

more negative than that with NaX in acidic solution. It appears that both

ammonium and halide ions have an influence on the behavior of the chromium

(II)/(III) couple. It is likely that ammonium ions (NH.) adsorb on mercury

surface to modify the kinetics of electrode reaction.

When NH.Br was used as supporting electrolyte, the cyclic voltammetry

curve was changed from what it was in a NH.C1 supporting electrolyte. (Fig.6).

The original cathodic peak was slightly changed from what it was in the pure

perchlorate or NH.C1 media, the anodic peak was shifted toward negative poten-

tials, and a new cathodic peak at potentials positive to the original cathodic

peak appeared. The new cathodic peak which was not seen in NH.C1 electrolyte

2+corresponds to the reduction of [Cr(H?0) Br] . It is believed some'bromide

2+ions have already coordinated to the Cr to form [Cr(H 0) Br] during the

back scan, but not chloride ions.

15

E (SCE)/V,

20

10

n

20

o

Figure 5: Voltammetry curve for 2 mM Cr (1LO), on nickel-mercury disk electrode.Quasi steady-state measurement. Xcurve recorded after many cycles).Electrolyte: 0.01M HC10,, 0.1M NH.C1 (pH: 2.4). Electrode area: 0.2 cm''Sweep rate: 100 mV/s.

-1,1 -0.9 -0,7E(SCE) /V,

16

-0.5—i—

-0.3

A

o _

n>

cathode

25

15

5

0

5

15

25

35

45

55

Figure 6: Voltammetry curves for 2 mM Cr(ILO).. on nickel-mercurydisk electrode with NH.X as supporting electrolyte. _Quasi steady-state measurements. Electrode area: 0.2 cm .A. in 0.01 M ffilO , 0.1 M NH.C1 <ptt 2 .4)B. in 0.01 M HClo£, 0.1 M NRVBr (pH: 2.3)

17

2. Chromic Chloride Reduction in Acid Media on Au, Pb/Au and GraphiteElectrodes

A. Gbld Electrode

3+ 2+As indicated in section 1, the Cr (ILO) /Cr (ILO) couple has

sluggish electrode kinetics on nickel-mercury electrode in acid media

at pHof 2. Such low pH solution was used to prevent the hydrolysis of

3+chromium couple. Under these conditions, the reduction wave of Cr(ILO),

occurs at potentials close to that for hydrogen evolution. It is difficult

3+to study the reduction of Cr(ILO), without interference of hydrogen

evolution.

The high overvoltage for Cr (III) reduction is due to the high free

energy of activation just as in the homogeneous case. In contrast to

the nickel-mercury electrode, the reduction of Cr(ILO), does not occur

on crystalline gold disk electrode without hydrogen evolution. Unlike

Cr(C10.)_, the reduction in solutions prepared from CrCl_ occurs on gold

disk electrode before hydrogen evolution with a well defined reduction

wave. Voltammetry and polarization curves are shown in Fig. 7-10. In

3+ 2+view of the results for the Cr (ILO) /Cr (ILO) couple even with the

introduction chloride into solution, the inner coordination sphere

corapiexes--of chromium with chloride play the major role in the reduction

on gold electrode in chromic chloride solution. From a coordination-com-

plex standpoint the binding chloride perturbs the transition-state of

the electron-transfer process as well as the initial-state of the

chromic complex. From a electrochemical standpoint the bound chloride

may adsorb on the electrode surface and participate as a bridging ligand,

18

thus facilitating the electron transfer.

It was found that the results are not reproducible; currents increase

with extended cycles. Probable reason for this is the changing of solu-

tion composition with time. The commercially available chromic chloride

is assigned the molecular formula [ Cr (ILO) ,C1 ]C1* 2ILO. On dissolving

this complex in water, however, the metal-bound chlorides are successively

replaced by water molecules, giving first blue-green [ Cr (ILO) C1]C1»1LO

and finally violet [ Cr (1 0) ]C1 .

With bridging chloride adsorbs on electrode surface to facilitate

2+ +electron transfer, both Cr (1LO) Cl and Cr (ILO) Cl are electrode

active species. Unfortunately their electrode kinetics are not the same

2+making the analysis work difficult. CrClLCO.-Cl with its dipositive

charge will have a higher concentration in the outer Ifelmholtz plane

than Cr(lLO),Cl7 at potentials negative to the potential of zero charge.

Beside that, the steric interactions between ligands are unfavorable to

+ 2+Cr(lLO),Cl within the double layer. The closest distance of Cr (H,0) Cl

to electrode surface is therefore shorter than that of Cr(H_0).Cl2.

2+Among these components in chromic chloride solution, the Cr (H-,0) Cl

has the fastest kinetics on gold electrode. As time passing after freshly

prepared, the Cr (ILO) Cl loses its bridging chloride producing more

2+Cr(H_0) Cl species in solution and therefore the current increases

with extended cycles. (Fig. 8).

19

Figure 7: Voltammetry curves for 3mM CrCl_ on gold disk electrode.Electriolyte: 0.1M NaCIO, and 0.01M HC10,. Electrode area:0.2 cm. Sweep rate: A. lOmV/s B. 50mV/s C. lOOmV/s.

20

cycl

900 rpm

50 oc

re40 3

30

20

10

-0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1£(SCE)/ V

0.1 0.2

-0.6 -0.5 -0.4 -0.3 -0.2 -0.1

cycle900 rpm

0 0.1 0.2

20

40

60

80 1

n>

100

120

Figure 8: Polarization curves for reduction of freshly prepared CrCl- ongold ring-disk electrode. Ring potential: 0.6V. Rotation rate:900 rpm. Sweep rate: lOmV/s. Solution: 3mM CrCl in 0.1M KaCIO,and 0.1M HC10..

4

21

Large hyteresis of CrCl_ reduction on gold surface under convection

was observed. (Fig. 8-10). The I-E curves do not level off at cathodic

potentials, -0.57V vs. SCE, where the reduction processes are under pure

diffusion control. Before the potential of hydrogen evolution, the currents

decrease quickly even during the back scan of potential. A re-enhanced

current occurs during back scan at the same potential where current chan-

ges dramatically during forward scan. The turnover of current occurs at

-0.57V vs. SCE., which is independent of the rotation rate. (Fig. 10).

This phenomenon is probably associated with the inner-sphere mechanism

in which the bridging chloride adsorbs on the gold surface. At potentials

negative to -0.57V vs. SCE. the bridging chloride de-adsorbs from the

electrode surface. Therefore the orientation of the oxidized species on

gold surface changes with the water ligand directing toward the electrode.

Because of the slowness of the electron transfer with the water ligand com-

pared with the chloride ligand, the electrode kinetics slow down at these

potentials. The re-enhancement of currents during back scan is due to the

re-adsorption of bridging chloride, as is .shown below:

CATHDDE CATHDDE

Chloro- o Aquo-

22

900 rpm

E(SCE) (V)

Ring (0.6v)

0.2

E(SCE) (V)-n.fi

900 rpm

bridged Clre-adsorpti

evolution

=JDL4L JLZ

30

60

150

Disc

bridged Cl desorption

Figure 9: Polarization curve for reduction of CrCl, on goldring-disk electrode. Ring potential: 0.6V. Rotationrate: 900 rpm. Sweep rate: lOmV/s. Solution:3mM CrClin 0.1M NaClO. and 0.01M HC10..A 4

23

Ring (0.6V)

2500

1600

900

rpm

E(SCE)/ V.

-0.2

80

60

40

20

0

oc-sn>

0.2

Figure 10: Polarization curves for chromic chloride reduction on rotating

gold ring-disk electrode in acid media. Solution: 3mM CrCl,O

in 0.1 M N a C l O . , 0.01 M HC10.. Sweep rate: 10 mv/s. Electrode2

area: 0.20 on , Rotation rates as specified.

24

B. Lead/Gold System

i). Deposition and Stripping of Lead on Gold Electrode

The role of lead on gold surface is important in the understanding

of the kinetics of chromium(II)/(III) couple. Underpotential deposition

as well as the bulk deposition and stripping of lead on crystalline gold

electrode have been studied in perchloric acid.

The effect of lead on the kinetics of the chromic reduction can be

studied by rotating ring-disk electrode technique. Cr(II) which is reduced

at the disk can be reoxidized to Cr(III) at the ring. The ring potential is

fixed at such a positive value (0.6V vs. SCE) that the reoxidation of Cr(II)

occurs under pure diffusion control. Because no lead deposition occurs at

the ring, the ring currents are totally due to the oxidation of Cr(II) to

Cr(III).

It was found that the underpotential deposition of lead on gold

disk results in a net decrease of the reaction kinetics as compared

with that on the bare metal. This effect is probably principally asso-

ciated with a double layer effect. The UPD lead species with its posi-

tive charge results in a decrease in the concentration of the Cr (III)

species at the electrode surface (in the outer Helmholtz plane).

-25

E(SCE) (V)

E(SCE) (V)

FIGURE 11: UNDERPOTENTIAL DEPOSITION OF LEAD ON GOLD

SOLUTION: 0,3 MM PstNO-^)R9

10 iW/s, ELECTRODE AREA: -- n ° ""*-STAGNANT SOLUTION,

ELECTRODE, SOLUTION: U,3 MH PBtNO-^ IN

0,01 M HCLO^, 0,1 M NACLO/,, SCAN RATE:0,2 CM

26

IESL9001600250036004900

-100

FIGURE 12": SAME AS FIG, n WITH ELECTRODE UNDER ROTATION, ROTATIONRATES AS SPECIFIED,

27

900 rpm

Ring

-0.5 -0.4 -0.3V-0.2 -0.1

12

10 £-S

ft)

E(SCE)/ V.

900 rpm

Figure 13: Polarization curves for chromic chloride reduction on rotatinggold_ring-disk electrode. Solution: 3 mM CrCl in o.l M 1C10,6-10 M PbO. Sweep rate: 10 mV/s. Electrode area: 0.2 cm.Ring potential: 0.6V vs. SCE.

28

ii). Potential Step Experiments

In order to study the transient behavior of UPD and the transient

effect of UPD on redox couple, controlled potential step experiments

have been arranged. The working electrode potential is initially set

to a known constant value where no Faradaic processes occur, and then

stepped to a potential at which the kinetics are of interest The

current which is the response of the system to the step perturbation

is measured as a function of time: The waveform applied in a series of

potential step experiments are depicted in Fig. 14.

These experiments of UPD phenomenon studies involve polarizing a

rotating gold disk electrode at a potential where UPD does not occur

and then stepping the potential to a value where UPD occurs. Under

these conditions with a low bulk concentration of the UPD species, the

deposition of the UPD layer occurs very slowly under pure diffusion

control. (Fig. 15). The coverage of the UPD layer can then be calculated

as a function of time from the diffusion flux. (Fig. 17-18). With the redox

couple also present, the dependence of the rate of the couple reduction

on the UPD coverage can be determined from the time dependence of the

couple reduction current. (Fig. 19-21). Figure 19 shows the current vs.

time curves for simultaneous underpotential deposition of lead and

chromic chloride reduction on gold rotating ring-disk electrode upon

potential step from 0.6V vs. SCE. into specificed potential. The

curve of time-depended ring current which totally results from the

oxidation of Cr(II) has the same shapes as those of disk current at

constant potential. Steady state currents are observed after 40 seconds

with electrodes under 900 rpm rotation.

The time dependence of the chromic chloride reduction current

2+ -with both Pb and Cl presence in solution is shown in Figure 24.

29

time

Figure 14: Step waveforms applied in a series of experiments.

30

0 w ?n ant/ sec.

fie d e f

1

2

< 3 .fc

4OJs_S-3u

0

Co>s_S-3o

10 30

( f t )

Figure 15: Disk current vs. Time curves for gold rotating ring-diskelectrode in 0.1 M HC10, in absence of Cr (III) and Pb(ll).Potential step from 0.6V into indicated potential. Rotationrate: 900 rpm. A. No NaCl B. with addition of 3mM NaCl.a, b, c, d, e, f, g, h, i are same as in figure 16.

31

a). 0.0V

b). -0.05V

c). -0.10V

d). -0.15V

e). -0.20V

f). -0.25V

g). -0.30V

h). -0.35V

i). -0.40V

Figure 16:"Current vs. Time curves for rotating gold ring-disk electrode

in a solution containing 0.1 M NaC104> 0.01 M HC104 and 6'10~ f MPbO upon potential step from 0.6V into indicated potential.Rotation rate: 900 rpm, ID= 0.K

32

E = -0.40V

Figure 17: Charge vs. Time curve for underpotential depositionof Lead on gold at fixed applied potential.Data taken from Fig. 16.

33

'20

Ring (0.6V.)

CJ

o

10

5

-0.40V

-0.35v

-0.30v-0.25v

15 25 35 45 55 65 75 85 t(Sec)

10

2O

30

40oi-

50

t(Sec)

5 .15 25 35 • 45 55 65 75 85

-0.25v

-0.30v

-0.35v

-0.40v

Disc

Figure 18: Current vs. Time curves for simultaneous underpotential depositionof lead and chromic chloride reduction on gold rotating ring-diskelectrode upon potential step from 0.6V into indicated potential.Solution: 3 mM CrCU, 6. 10- M PbO in 0.1 M NaCIO, and 0.01 M HC10

3'Rotation rate: 900 rpm. Ring potential: 0.6V vs.0.2 on.

i *

>CE. Electrode area:

34

A for Cr(III) reduction

Q/^coul.

lead depositionin charge units

Figure 19: Three dimensional representation of current vs. potentialand charge of UPD. Data from figure 16, 18.

35

I//I A

co•H4-1U3

E(SCE)/V.

20 QAUCOU,?0

lead UPD coverage in charge units

Figure 20: Same as figure 19, at Q = 74.3>Ucoul., E = -0.40V

36

3, 15

10

Rina (0.6v)

-f-

-0.42v-0.40v

-0.35v

-0.30v-0.25V

0 10 20 30 40 50 60 70 80 90t(Sec)

0 L 1.0 2P 30 40 50 60 70 80 9JD

5

10

15

20

25

30

35

40

t(Sec)

40v

-0.40v Disc

Figure 21: Current vs. Time curves for the simultaneous underpotentialdeposition of lead and chromic chloride reduction on a-rotating ring-disk Au electrode upon potential step from0.6V into indicated potential. Solution: 3mM CrCl_ in0.1 M HC10 andf 6-10 M PbO + 3mM NaCl,--6.10 M PbO, ••'• 3mM NaCl.Rotation rate: 900 rpm.

37

The rotation dependences of the transient behavior for simultaneous

underpotential deposition of lead and chromic chloride reduction on the

gold disk electrode at constant potential are shown in figure 22-23. The

time required to reach steady state decreases with rotation rate.

38

o>3

10 20 3D 40 60t/ sec.

rpm

E = -0.25V

10oc.-5

n>3

20

_2P_ 30t/ sec.60

900

1600250036004900

E = -0.30 V.

Figure 22: Disk current vs. Time curves for simultaneous underpotential depositionof lead and chromic chloride reduction on gold electrode with rotationrates change as indicated. Potential gtep from 0.6V into indicatedpotential. Solution: 3mM CrCl , 6'10~HC10,.

M PbO in 0.1 M NaC104 and 0.01M

39

30 40 50t/ sec.

60

900

1600

2500

3600

4900

20 30 40 50 60t/ sec.

E = -0.40V.

rpm

900

16002500

Figure 23: Same as figure 22.

40

C. Ordinary Pyrolytic Graphite

With lead deposited on graphite surface to prevent hydrogen evolution,

both reduction and oxidation of chromic chloride couple occur on graphite

electrode. Cyclic voltammogram shows two new waves for the reduction and

oxidation of chromium couple. (Fig. 24)

The plating of lead on ordinary pyrolytic graphite rotating disk

electrode was also investigated in the past period. Further studies are

still going on.

E(SCE) (V)

Figure 24: Cyclic voltammogram of ordinary pyrclytic graphite disk electrode ina solution containing 0.1 M NaCIO , 0.01 M IC10, and 3 mM Pb (NO ) .Sweep rate: 10 mV/s. A). No CrCl B). with 6mM CrCl .

41

E(SCE) (V)

rpm

900

1600

2500

3600

4900

2000

1000

25

50

75

100

125

150

175

200

r>-s-jn>

Figure 25: Polarization curve for lead deposition and stripping on ordinary ^pyrolytic graphite rotating disk electrode. Electrode area: 0.2cm .Solution: 0.1 M NaCIO,, 0.01 M HC10 and 3 mM PbO. Sweep rate: lOmV/s.Note: change in ordinate for anodic vs. cathodic curves.

3. Fe(H00)^+/Fe(H-0)f*~ Couple, Gold Ring-Disk Studies in Acid Media2 o 2 o

3+ 2+The kinetics of the Fe(H00),, /Fe(H00), couple were studied on a/ D / O

gold rotating disk electrode in 0.5 M HC10, solution. Without anions

other than the CIO, of the supporting electrolyte, moderate electrode

kinetics have been observed. It was found that the process is very

sensitive to the surface condition; trace concentrations of impurities

made the results unreproducible. The cyclic voltammogram and polariza-

3+ 2+tion curve of Fe(H.O), /Fe(H»0), couple are shown in Figures 26 - 27.

2. O f- O

The resulting Levich plots in which the curves are parallel confirm

that the electron transfer reaction is first order in kinetics. For

first order reaction: ,. .-1 ... v-1 , . JL/2N-1CO = diJ + (B tw )

i,: current for pure kinetic control

1/2BCO ?= i, : diffusion limiting current

43

anode

E (SCE)/ V.

300"

Figure -26 : VoItammetry curves for 6mM Fe(H»0) on gold disk electrode

in 0.5 M H310,. Electrode area: 0.458 cm . Sweep rate:

A: IChnV/s, B: 50mV/s, C: lOOmV/s.

oa

0)•ot-i4Jo0)

CL0)0)

13r-lO00

00

•H4JCO4JOJ-4coco•H4JCJ3

"D0)

CO0)

3O

O•H4-1CON

CO

OIX,

CNJ

oo

CM

O

GOm

ncu

cu13OM4J

U0)

rHw

OiH

a

3i-lO

T30)

•HC O

•H OJO.CO

COcfl

x-x CO<• 0)

O 4J»-l Cfl

co•H4JCO

CUe

CU4Jcfl

45

1 1.5

1.0

0.5

01.0 2.0 3.0

(0»-1/2 102 rpnf1/2

-1 -1/2:igure 28: I vs. GJ plot of data in Fig. 27

46

A. Effect of Hal ides

3+ 2+The electrode reaction of the Fe(H00),. /Fe(H00), couple is stronglyC O C . V

affected by halide anions adsorbed on the electrode surface. In the absence

of halides the shape of the voltammogram with separated cathodic and anodic

waves indicates a relatively slow electrode reaction. However, it was found

that the addition of even a small amount of halides leads to a change in

polarization curves to a shape corresponding to a much more rapid electrode

reaction. The marked increase in the rate of the electrode reaction of3+ 2+Fe(H20)g /Fe(H20)g couple has long been believed due to changing the

electron-transfer process from outer-sphere mechanism to inner-sphere

mechanism with halide as a bridge between electrode and iron complex.

Corresponding cyclic voltammogram and polarization curves of iron(II)/(III)

couple with halides presence in solution are shown in the following figures.

47

E(SCE)(V)

-300

Figure 29: Effect of chloride on the kinetics of iron(II)/(III) couple ongold electrode in acidic media. Solution: 6 mM FeCClO.)., in,,0.5 M HC10,. Electrode: gold disk (Pine) Area: 0.458 J cm.

B). 3 mM NaCl added.•> A • L. i CU Ul l/UC • QO I 0

Sweep rate: 50 mV/s. A). No chloride

48

anode

cathode

E(SCE)/ V.

Figure 30: Voltammetry curves for ferric-ferrous couple on gold disk electrode

in acid media containing chloride. Solution: 6mM Fe (CIO,) in

20.5 M HC10, and 3mM NaCl. Electrode area: 0.458 on (Pine). Sweep

rate: A: lOmV/s, B: 50mV/s, C: lOOmV/s.

49

Figure 31: Polarization curves for 6 mM Fe(C10j, in 0.5 M HC10., 3 mM NaCl onrotating gold disk electrode. Electrode area: 0.458 cm .(Pine).Sweep rate: 10 mV/s. Rotation rate as indicated.

50

-1/2 ..2 -1/2.O «10 (rpm )

Figure 32: I"1 vs. 6o"1/2 plot of data in figure 31.

51

LUQOtrHCJLLJ_lLU

CO -3""- OQ -I

OQ ^ Q_l LU LUO U_ QID Q

2: hO LU— Qf- .. _< z cel- o oo -- -io: i- DC

o o ooo 2:

zO CO -—• ^^ QH > LUO S HZ5 <Q O OLU i—I —•ce Q

-- z+ LU —•

x-s < COO Qi <£

CM

LU O LULL. 00 I-

<o: - o:OCX]u- s: z

o o111 f—I

> -, I-cx C* <

Oi—i •» ^3I- < O< LU _lM Cf. C_3

—I COo --

COCO

LU

CD

52

C_3I/O

oc01

co

ocIT)CM

CCvoCO

LHIO

oLU

LU

LUo:Q-

LU

COco

CO

LUS

CO

• *

CO

LU\-

co

>-Q<LUI-co

53

IOi—I

CsJ

U_oLU

LUCOLUo:CL

UJX

cooo

CO

UJ

CO

LT>oo

LU01

CD

LU

150

Figure36 : Voltammetry curves for ferric-ferrous couple on

gold disk electrode in acid media containing Br .

Solution: 6mM Fe (CIO,) ' in 0.5M HC10, andf

3mM NaBr and 0.32mM Pb (NO ) . Electrode area:0.458cm'

Sweep rate: A: 10 mv/s, B: 50 mv/s, C: 100 mv/s.

55

w

w

ir

O

CM^

O

mCM\o

oo

oo

oomCN

oo

O O O O O OO O O O O Ot— i CN ro .3- m vo

OO

OOco

01T3O

OOJ

acoco

u3

TJ(U

n

COatgCJ

co

N

CO

o3oo

OllJ C1J

cfl COz

0)c

CN

1

O01

r-l

01f*-i Ci

O

O 4J

OCO

c•H

OJ•O

CU(0(U

PQCO

00

•a-o

coO)

01•ao

u01

oiH

a

oc

o&XI

SCMCO

•O01i-lM-l•HU01D.toCOCO

CO(U

co•H4-1

CO4J

or-<

0)JJcfl

56

B. Effect of the Underpotential Deposition of Bismuth

The plating and stripping of bismuth on gold was studied in perchloric

acid media using a rotating gold disk electrode. The underpotential deposi-

tion of bismuth occurs at potential region between -0.05V and +0.40V vs. SCE.3+ 2+where the Fe(HpO)g /Fe(H20)6 couple undergoes both diffusion and kinetic

controlled reaction. It was found that the underpotential deposition of

bismuth on gold results in a net decrease of the reaction kinetics as compared3+with that on the bare metal for Fe(H20)6 reduction.

57

E (SCE) / V

900 rpm0.6V_ IR.= 0

Figure 38: Voltammetry curves for IPD of bismuth on gold disk electrode.

Voltage window opening experiments. Electrolyte: 0.1M 1C10,,

-5 23.2*10 M Bi203. Electrode area: 0.2 cm . Sweep rate: 10 mv/s.

Rotation rate: 900 rpm

58

A

E(SCE) (V)

SCAN RATE: 10MV/S

fl .5 . 10 . 15 .t(Sec)

B

a,b.c.d.e.f.

0.4 V0.3V0.2 V0. 1 V0.05V0.00V

g.-O.05V

(stepped from 0.8V)

FIGURE 39: UNDERPOTENTIAL DEPOSITION OF Bl ON Au

ROTATING DISK ELECTRODE IN A SOLUTION

0.1M "5, 3,2'10"M Bl203, ELECTRODE

AREA:^0.2CM . ROTATION RATE: 900 RPN.

A, CYCLIC VOLTAKMOGRAM. B, TRANS-IENT EXPERIMENT

59

60

nc

50

40

30

20

10

Bi bulk-1/ 40deposition

E (SCE) / V

Figure 40 : Deposition and Stripping of bismuth on gold disk electrode.

~4Electrolyte: 3.5*10~M Bi^ in 0.1M

10 mv/s. Electrode area: 0.2 cm2.

. Sweep rate:

60

10

15

20

25

t/sec.

a),b).c).d).e).f).g).h).

0.40V0.30V0.20V0.10V0.00V-0.05V-0.10V-0.15V

900 rpm

Figure "1: Current vs. Time curves for gold disk electrode in a solution containing

0.1M IClO and 3.2 10 M Bi 0 upon potential step from 0.8V into indicated

2potential. Rotation rate : 900 rpm. Electrode area: 0.2 cm . I = 0, E =0.8V.

R R

61

Ring(O.SV)

-0.3 -0.2 -0..1 3 0.1 0.2 0.3

f(SCE)/V

-0.3 -0.2 -0.1 0 O.-l 0.2 0.3 0.4 0.5 0.6 0.7 0..8

Disk

FIGURE42 : POLARIZATION CURVE FOR FEt^O)"1" REDUCTIONON ROTATING GOLD RING-DISK ELECTRODE IN

ACIDIC MEDIA, SOLUTION: 3ttM FE(CLO^)^

IN 0.1M HCLO, ROTATION RATE: 900 RPM.

DYNAMIC POLARIZATION . SCAN RATE: lO/W/s,

62

900 rpm

100

80

360

20

-0.4 -0.3 -0.2 -0.1

R i n g ( 0 . 9 V )

0 0.1 0.2 0.3 0.4 0.5 0,6 0.7 n.R n.Q

E(SCE)/ V

-0.4 -0.3 -0.-2 -0.1 0 0.1 0.2 0.3 0.4 0.5 LTo-. 8 ;- 0-9"

E(SCE)/ V

50 Bistripping

n>

900 rpmDisk

FIGURE 43: EFFECT OF Bi. UNDERPOTENTIAL DEPOSITIONON THE REDUCTION OF FECh^O)^" ON GOLD

RING-DISK ELECTRODE, ELECTRODE AREA:

RiNG=DisK^O,2cM2, SOLUTION: 3,2'10~5MBl203, 3.2MM FE(CLO^)3 IN 0.1MROTATION RATE: 900 RPM, RING POTENTIAL: 0, VDYNAMIC POLARIZATION, SCAN RATE: lOMV/s

63

et

3,

TOO

80+jcQJ

b 60

40

20

Ring (0.9V)-0.1,0.0V0.1V

0.4V

0.5V

t/ sec.0 -5 10 15 20 25 30 35 40 45

10 15 20 ?5 3n 35 40 45.t/ sec.

O)

50

100

150

200

250

Disk 900 rpm

0.5V

0.4V-

0.3V

0,2V0.1V

-0.1, 0.0V

Figure 44: Current vs. Time curves for ferric-ferrous couple on gold

ring-disk electrode upon potential step from 0.9V into

indicated potential. Solution: 3.2mM Fe(C10.)3 in 0.1M

HC104. Rotation rate: 900 rpm. Ring potential: 0.9V.2

Electrode area: 0.2 cm .

<3

cuS-3o

80

60

40

20

D.05to-0.3\0.1V

0.2,0.3V

Ring (0.9V), 900 rpm

10 15 20 25 30 35 40 45 t/sec.

t/ sec.

0.3, 0.2V

0.1V0.05V

9o?ifv-0.2, -0.3VJ

Disk , 900 rpm,

Figure 45: Current vs.. Time curves for simultaneous underpotential deposition

of bismuth and Fe(H20)g reduction on gold ring-disk electrode upon

potential step from 0.9V into indicated potential. Solution:3.2mM

Fe(C104)3 , 5MO"5M Bi^ in 0.1M HC104- Rotation rate: 900 rpm.

Ring potential: 0.9V. Electrode area: 0.2 cm2.

65

4. SPECTROSCOPIC STUDIES

A. Time Effects on Solution Composition. (Visible Spectra).

Ultraviolet and visible spectroscopy have been utilized in connection

with the chromium(II)/(III) couple part of NASA's Redox system as a means of

analyzing the solution compositions. These studies have been aimed at deter-

mining the concertration of individual metal complexes at different states

of charge and discharge in an effort to correlate the performance of the

overall electrode with the presence of certain species in solution.

The time dependence of absorption spectra of CrCK solution is shown in

figure 46. It was found that the absorption maximums shift toward shorter2+wavelength with time. However, the spectrum of Cr(H 0) 01 which can be

separated using an ion-exchange column eluting with HC10. does not change

even after 12 hours.(Fig. 47).

B. XPS Measurement on pre-electrolysis Gold Cathode

In order to check out the effects of impurity on the reduction

kinetics of couple, a pre-electrolyzed gold cathode out from the same

supporting electrolyte for the couple is examined by ESCA. No obvious

impurity has been detected, except oxygen and carbon.which exist else

where.(Fig. 48).

66

c c c c c c

'i "i 'i "i 'i 'io o o o o ovo en oo LT> co «—'

.—t i—I .—I CM

CO -O

CDcCD

r—O)

O OCO — •

c_> a.re LU

o:

o u-- o

o

cr --

<_> o< H

21 a:rH 01

- u.C3 LU

ccz>_ UJ

ce

CM

LU

LULL. —•O h-

s: u_ZD ODCH- Z.O OLU —Q_ Hco O

z.LU ra_) u_cq

CO

> <

VO

LU

LU

67

•4JD>CdJ

O)

COLUCC10-

LUO

XOX

O LU

> LUOQ O

Q ZLU O

< u\- -z.O Ll_

<

CO<•3-

+ sCVJ —1

LO

3 ^CSJ -

ce 2

ce Nj- _iO OLU o:o_ <_JCO

QLU LU—I OHCQ <—• O.CO LU—• o:> CL.

68

oCD.CVJ

Lt_O <

CO O—. ZCO •—i>- CO-J !Docc -.\- CO(_> a:uu x_lLU .3-I CM

LUDC 010. O

U.o:LU cI- OU_ _l

Q s:O t— IX O\- - UJ< O Qo o

2 ce

o -dU_ o

uj— IUJ

Q O_J < UJo 2: u

< .— i ceU- Oos: o

UJo:

« Q

f- r>o _Juj oQ- CO

ujQ

CO OQ "Z.UJ <

OOL <H Q:o nUJ H-I <UJ CO

CO

M

0 ujre o 2:a. o =>

o: z> Q —•< >- f-ce x <1 -J

X < Q-

oo<3-

LU