Tc

HD

a

ARR1AA

KSSCD

1

mA[[etin

is

A

a

E

p

Eg

A

0d

Electrochimica Acta 56 (2011) 2781–2791

Contents lists available at ScienceDirect

Electrochimica Acta

journa l homepage: www.e lsev ier .com/ locate /e lec tac ta

he effect of silver chloride formation on the kinetics of silver dissolution inhloride solution

ung Ha ∗, Joe Payer1

epartment of Materials Science and Engineering, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, OH 44106, USA

r t i c l e i n f o

rticle history:eceived 11 October 2010eceived in revised form3 December 2010ccepted 14 December 2010

a b s t r a c t

The precipitation and growth of AgCl on silver in physiological NaCl solution were investigated. AgCl wasfound to form at bottom of scratches on the surface which may be the less effective sites for diffusionor the favorable sites for heterogeneous nucleation. Patches of silver chloride expanded laterally on thesubstrate until a continuous film formed. The ionic transport path through this newly formed continuous

vailable online 4 January 2011

eywords:ilverilver chloride

film was via spaces between AgCl patches. As the film grew, the spaces between AgCl patches closed andion transport was primarily via micro-channels running through AgCl patches. The decrease of AgCl layerconductivity during film growth were attributed to the clogging of micro-channels or decrease in chargecarrier concentration inside the micro-channels. Under thin AgCl layer, i.e. on the order of a micrometer,the dissolution of silver substrate was under mixed activation-Ohmic control. Under thick AgCl layer,

f mic

orrosionissolution kinetics

i.e. on the order of tens oresistance of AgCl layer.

. Introduction

Growth of silver chloride on silver substrate in chloride environ-ent has been studied by many researchers due to the widely use ofg/AgCl electrode in laboratories as secondary reference electrode

1,2], in sea water activated batteries as well as in dry batteries3–6], and the application of the system for modeling of porouslectrode [7,8]. The interesting and useful behavior of Ag/AgCl elec-rode is partly attributed to the high reversibility of Ag/Ag+ systemn chloride environments [3,9], however this behavior makes thisoble metal prone to corrosion in chloride environment.

The oxidation of silver in chloride containing solution resultingn the formation of AgCl film has been reported in a number oftudies [6,10–17]. The reaction can be expressed as:

g + Cl− = AgCl + e−

nd the half cell potential follows a Nernstian behavior:

= −0.015 − 0.059 log[Cl−] (VSCE) (1)

A precursor monolayer of AgCl was reported to be formed atotentials below the thermodynamic Ag/AgCl reversible poten-

∗ Corresponding author. Present address: Center for Electrochemical Science andngineering, Department of Materials Science and Engineering, University of Vir-inia, 395 McCormick Road, Charlottesville, VA 22904, USA.

E-mail address: hmh2n@virginia.edu (H. Ha).1 Present address: Corrosion and Reliability Engineering, The Akron University,kron, OH 44325, USA.

013-4686/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.oi:10.1016/j.electacta.2010.12.050

rometers, the dissolution of silver substrate was mediated by the Ohmic

© 2010 Elsevier Ltd. All rights reserved.

tial [10,14,15]. The formation of the monolayer AgCl followsan adsorption–desorption mechanism. At a certain overpotentialmore positive than the Ag/AgCl reversible potential, multi-layerAgCl grows laterally via nucleation and growth [6]. Gradually a thickAgCl film was observed on silver electrode at longer exposure time[16,17].

Morphology of non-continuous AgCl patches was described hav-ing rounded smooth surfaces without manifestation of any crystalorientation [18]. The thickness of the non-continuous layer is fairlyconstant, and is <0.5 �m. The top view of the continuous AgCl film iscomprised of dense, fine particles [6,18]. At longer anodic polariza-tion exposure, the morphology of AgCl top layer changes to a mosaicstructures [6]. Several attempts have been made to characterizethe inner structure of the continuous AgCl layer with various crosssectioning techniques including mechanical polishing followed byetching [17], sharp scalpel cutting [6] and breaking by liquid N2[17]. However details of the inner structure of AgCl layer were notavailable and in many cases were reported with speculations.

The growth of continuous AgCl film under galvanostatic condi-tion results in a shift of the applied potential to the more positivedirection. This behavior is attributed to the Ohmic resistance of theAgCl layer [11]. Depending on the magnitude of the applied anodiccurrent density, the growth of AgCl follows either a low field or ahigh field or a mixed low and high field conduction mechanism.

The low field conduction mechanism takes place at low currentdensity range (<10 mA/cm2) in which AgCl film is thought to beporous and AgCl exhibits extrinsic conductivity, i.e. Ohmic con-duction via pores [11]. However, no attempt has been made tocharacterize the morphology, the size, the density of these pores

2 mica Acta 56 (2011) 2781–2791

ofirtpm

talm[tepttihctAdt

lsubtwr(wtcOiptmaa

2

2

Acgibwt

wmcmstw

0

782 H. Ha, J. Payer / Electrochi

r to examine the solution concentration inside the pores. The higheld conduction mechanism takes place at high current densityange (>100 mA/cm2) in which AgCl exhibited intrinsic conduc-ivity, i.e. exponential relationship between current density andotential [11]. A mixed conduction mechanism can occur at inter-ediate current densities.Transport of ions through the corrosion product layer is essen-

ial for continuous corrosion [19]. In the case of silver oxidationnd formation of AgCl, the dominant charge carrier inside the AgClayer is still controversial. Some researchers claimed that Ag+ is the

ain charge carrier inside the micro-channels through AgCl layer6] while some others suggested both Ag+ and Cl− contribute to theransport process [17]. Beside the charge carrier, the physical prop-rties of the AgCl layer also play an important role on the transportrocess. As the AgCl layer grows thicker, diffusion and migrationhrough this layer become more difficult due to the lengthening ofhe transport distance and/or the decrease of the ion concentrationnside the micro-channels. The formation of a thick AgCl layer withigh Ohmic resistance on the silver surface as the consequence oforrosion is likely to slow the rate of the silver dissolution. However,he inter-relationship between the AgCl structure, the resistivity ofgCl layer, the transport process through this layer and the silverissolution kinetics in chloride environments was not addressed inhe literature.

The objective of this work is to examine the growth of AgClayers on silver, and its effect on silver dissolution in 9 g/L NaClolution which is more diluted than the concentration commonlysed in the previous studies. The thick AgCl layers were growny potentiodynamic and galvanostatic polarization in Cl− solu-ion. The morphology and crystal structure of thick AgCl layersere characterized with scanning electron microscope (SEM), X-

ay energy dispersive spectroscopy (XEDS) and X-ray diffractionXRD). The inner structure of the AgCl layers were characterizedith focused ion beam (FIB) and SEM. FIB provides the ability to cut

he AgCl layer at high accuracy giving a smooth cross section thatannot be obtained with conventional mechanical methods. Thehmic resistance of the AgCl layer was measured by electrochem-

cal impedance spectroscopy (EIS). The effect of AgCl corrosionroduct layer on the kinetics of silver dissolution was studied inerms of the structure, thickness and conductivity of AgCl layer. A

athematical model is presented in a separated paper. The resultsre relevant to the dissolution and release of silver in biomedicalpplications [20,21].

. Experiments

.1. Specimens and electrochemical cell

Silver wires of 99.9% purity with 0.5 mm diameter supplied bylfa Aesar (Ward Hill, MA) were used in this study. Physiologi-al saline solution of 9 g/L NaCl was prepared from experimentalraded crystalline NaCl (Fisher Chemical, Fair Lawn, NJ) and de-onized water. This solution is commonly used in corrosion tests ofiomaterials. A saturated calomel electrode (SCE) and pure Pt wireere used for reference and counter electrodes, respectively. The

est cell setup followed a 3-electrode configuration.Specimens for AgCl formation were 0.5 mm-diameter silver

ires of 2 cm long. The silver wires were ultrasonically cleaned inethanol prior to each experiment. Further cleaning was done by

athodic polarization at −1.0 VSCE for 10 min. To observe the for-ation and growth process of AgCl on Ag substrate, the Ag wire

pecimens were polarized from a cathodic potential of −0.50 VSCEo final anodic potentials varied from −0.04 VSCE up to +0.60 VSCEith a scan rate of 0.5 mV/s.

Specimens for AgCl Ohmic resistance measurement were.5 mm-diameter silver wires. The wires were wrapped in

Fig. 1. Current vs. potential curve of the silver wire specimens during potentiody-namic polarization in physiological 9 g/L NaCl solution. The terminated potentialsfor preparation of different AgCl layers are shown in the curve.

adhesive-line polytetrafluoroethylene (PTFE) heat shrink tube. Thecross section of the silver wire was polished with 600 grit sili-cone carbide paper and ultrasonically cleaned with methanol. Then,the wire was pulled inside the tube to form a cylindrical cavity of0.5 mm diameter and 1 mm depth. The purpose of this treatmentwas to help the formation of a uniform AgCl layer rather than amushroom cap on top of the Ag wire cross section as observed onflush mounted sample [22]. To obtain AgCl layers with differentthicknesses for resistance measurement, the specimens were anod-ically oxidized to pass different amounts of charges up to 10 C/cm2

at different applied current densities of 0.1, 0.2, 0.5, 1 and 2 mA/cm2.

2.2. Characterization of silver chloride layer

The morphology of AgCl formed on the specimens was observedwith a Hitachi S4500 field-emission scanning electron microscope(FE-SEM) (Hitachi, Ltd.). The chemistry and structure of the layerwere examined with a Noran X-ray energy dispersive spectrometry(XEDS) incorporated with the FE-SEM and a Scintag X-1 advancedX-ray diffractometer (XRD) (Scintag, Inc.) using Cu (K�) radiation.The inner structure of the AgCl layer was observed with SEM aftercross sectioning the layer with a dual-beam focused ion beam (FIB)Nova Nanolab 200 (FEI Company). A thin Pt layer of 1 �m wasdeposited on top surface to protect the AgCl layer during cross sec-tioning. An ion beam of 30 kV and 7 nA was used for cross sectioning,and an electron beam of 5 kV and 0.4 nA was used for SEM imaging.The illumination time was minimized for both cross sectioning andimaging activities to avoid decomposition of AgCl under energizedbeams.

2.3. Ohmic resistance measurement of AgCl layer

The resistances of AgCl layers formed after passing differ-ent amounts of coulombs were measured with electrochemicalimpedance spectroscopy (EIS) using a Gamry FAS2 femtostat(Gamry Instruments). The impedance was measured by applyinga 10 mV oscillation perturbation and sweeping frequency from

100 kHz to 1 Hz to minimize the effect of EIS measurement onAgCl thickness. The value of the AgCl layer resistance was takenas approximately the impedance value at 100 Hz when the phaseshift angle between the response current and the applied potentialwas ca. −1◦.

H. Ha, J. Payer / Electrochimica Acta 56 (2011) 2781–2791 2783

F rmatioa 0.08, (

2

egt11(vw

3

3

tb

ig. 2. SEM micrographs showing the morphology of different AgCl layers during fond terminated at different anodic potentials of (a) −0.04, (b) +0.04, (c) +0.06, (d) +

.4. Potential-pulse experiment

The effect of the AgCl layer on silver dissolution kinetics wasxamined with potential-pulse experiments. A AgCl layer wasrown on a heat shrink tube wrapped silver wire specimen as men-ioned earlier by galvanostatic polarization at a current density ofmA/cm2 for 1000 s. Potential pulses with the pulse duration ofs were applied using a Solartron 1280B Electrochemical System

Solartron Analytical). The potentials were increased from an initialalue of 100 mVSCE with a step of 100 mV and the response currentas recorded.

. Results

.1. Precipitation and growth of AgCl

A representative current vs. potential curve during the poten-iodynamic polarization experiment is shown in Fig. 1. The anodicehavior of silver was characterized by a low anodic current density

n on silver surface. The specimens were prepared by potentiodynamic polarizatione) +0.15, and (f) +0.60 VSCE.

at the potentials near the corrosion potential followed by an abruptincrease at a potential of ca. +0.05 VSCE. To observe the formationand growth of AgCl on the silver surface, the polarization was ter-minated at different anodic potentials of −0.04, +0.04, +0.06, +0.08,+0.15 and +0.60 VSCE as indicated in the figure.

SEM micrographs showing the morphologies of the speci-mens after polarization are presented in Fig. 2. At the terminatedanodic potential of −0.04 VSCE the anodic current density wasca. 8 × 10−7 A/cm2. The SEM micrograph of the specimen afterpolarization (Fig. 2a) shows a surface free of AgCl. At the anodicpotential of +0.04 VSCE, the anodic current density increased toca. 1.3 × 10−6 A/cm2. The morphology of the polarized specimenterminated at +0.04 VSCE (Fig. 2b) was featured by particles withsize of <100 nm scattered over the surface. The anodic current

density increased significantly at the potential of +0.06 VSCE to ca.1.2 × 10−5 A/cm2. Clusters of 100 nm size particles were observedon specimen at this terminated potential (Fig. 2c). At +0.08 VSCE, thecurrent density was ca. 1.8 × 10−3 A/cm2, and the specimen surfaceafter polarization was covered with patches of non-continuous cor-

2784 H. Ha, J. Payer / Electrochimica Acta 56 (2011) 2781–2791

Ftt

rtslotsa

3

tIscff

lA(tt

Ftf

ig. 3. XEDS spectrum of the non-continuous film on the specimen after polarizationo +0.08 VSCE in the potentiodynamic polarization experiment. The spectrum showshe existence of Ag and Cl in the corrosion product layer.

osion product film. During growth, the patches expanded laterallyo cover the silver substrate more than growing thicker from theurface. A preferential expanding direction was observed that fol-owed the scratch direction on the surface (Fig. 2d). At higher anodicverpotentials, the current density did not increase but stayed athe higher magnitude on the order of 10−3 A/cm2. The surface ofpecimen after polarization was covered with a continuous layers shown in Fig. 2e and f.

.2. Structure of AgCl layer

An XEDS spectrum taken from a patch on the specimen polarizedo +0.08 VSCE is shown in Fig. 3. Silver and chlorine were detected.n addition, an XRD pattern of the corrosion product formed on thepecimen after polarization to +0.60 VSCE is shown in Fig. 4. Therystalline peaks in the XRD pattern were identified as AgCl withcc structure (PCPDFWIN software Version 2.1, International Centeror Diffraction Data).

The morphologies of the non-continuous and continuous AgClayers grown by potentiodynamic polarization are shown in Fig. 5.

gCl patches in the non-continuous layer had rounded edges

Fig. 5a) and tended to expand laterally rather than thickening. Inhe continuous AgCl film formed after potentiodynamic polariza-ion to +0.15 VSCE (Fig. 5b) the top layer was comprised of rounded

ig. 4. XRD pattern of the continuous film formed on the specimen after polarizationo +0.60 VSCE in the potentiodynamic polarization experiment. The pattern confirmscc crystal structure of AgCl layer.

Fig. 5. SEM micrographs of the surface morphology of different AgCl layers afterpotentiodynamic polarization in physiological 9 g/L NaCl solution. The anodic poten-tials were terminated at different potentials of (a) +0.08, (b) +0.15 and (c) +0.60 VSCE.

particles with size <1 �m. Spaces between AgCl particles can beseen in the SEM micrograph. The top layer of the continuous AgCl

film changed to irregular-shaped grains at the terminated polariza-tion potential of +0.60 VSCE (Fig. 5c). The spaces between AgCl grainswere not observed, and instead micro-channels running throughthe grains can be seen from the top surface.

H. Ha, J. Payer / Electrochimica Acta 56 (2011) 2781–2791 2785

Fig. 6. SEM micrographs of the cross sections of the AgCl layers grown by galvano-static polarization in physiological 9 g/L NaCl solution. The layers were grown at0lfl

isgs

0.0 0.5 1.0 1.5 2.0 2.50

1

2

3

4

i (m

A/c

m2 )

Overpotential (V)

Fig. 7. Silver dissolution current density vs. overpotential in potentiostatic step

.5 mA/cm2 to pass a charge amount of 5 mC/cm2; (a) FIB cross sectioning at topayer showing a portion of a micro-channel; (b) FIB cross sectioning at Ag/AgCl inter-ace; and (c) a cross-section of AgCl layer after ultrasonic cleaning to break the topayer.

The inner structure of a thick, continuous AgCl layer was exam-ned by cross sectioning in the direction vertical to the substrateurface with FIB followed by SEM observation. The SEM micro-raphs of the cross sections are shown in Fig. 6a and b. Fig. 6chows a cross section of the AgCl layer in the direction parallel

experiments. The duration of each applied potential was 1 s to minimize the growthof AgCl layer. The overpotential was calculated with the assumption that the elec-trode potential is equal the Nernstian potential of Ag/AgCl in 9 g/L solution, i.e.Ea = +0.033 VSCE.

to the substrate surface after removal part of the layer by ultra-sonic cleaning. AgCl grains of several micrometers were observed inFig. 6a. Some micro-channels running from the top surface throughthe grains into deeper layers were also observed in Fig. 6a and c. Themicro-channel in Fig. 6a indeed penetrated deep into the AgCl layer,however due to the tortuosity of the channel, only a portion of it wasobserved in the present cross section. After ca. 20 s exposure underelectron beams, the AgCl grains were quickly decomposed creatinga porous nanostructure with the pore size of approximately 100 nm(Fig. 6b). The decomposition of AgCl to silver under the illuminationof energized beams is well known and is utilized in photography. Nosignificant delamination at the Ag/AgCl interface was observed inFig. 6b. To maintain such adherent interface during continuous dis-solution of the silver substrate, AgCl must be formed at the bottomof the AgCl layer.

3.3. The Ohmic response behavior of silver dissolutionunderneath AgCl layers

To examine the dependence of silver dissolution currents onapplied potentials with the existence of a continuous AgCl layer,potentiostatic step experiments were performed. The thickness ofthe AgCl layer after passing 1 C/cm2 was approximately 20 �m.Fig. 7 shows the relationship between the anodic current densityand the overpotential during the potentiostatic step experiment.The overpotential was calculated with the assumption that thepotential of the silver electrode underneath the AgCl layer wasequal the Ag/AgCl reversible potential of +0.033 VSCE in 9 g/L NaClsolution as obtained from Eq. (1). For a large range of applied poten-tial from +0.1 to +2.0 VSCE, the response current density followeda linear relationship with overpotential indicating the dissolutionof silver underneath a thick AgCl layer was under Ohmic control asdiscussed below.

3.4. Ohmic resistance of AgCl layer

The resistances of the AgCl layers grown by galvanostaticpolarization experiments were measured by EIS at different AgClthicknesses during the experiments. Fig. 8 shows a typical poten-tial vs. time curve during galvanostatic polarization experiments.The potential at the oxidizing electrode increased with time. Poten-

2786 H. Ha, J. Payer / Electrochimica Acta 56 (2011) 2781–2791

Fii

tetbtsavpmvootsvEm

tEttarcAditw

4

4

Nsiaat

ig. 8. Potential vs. time curve during galvanostatic preparation of AgCl layer with= 0.5 mA/cm2. The arrows indicate the time when EIS experiments took place. Thenset shows a segment of the E–t curve at a higher magnification.

ial fluctuations were observed in all experiments; however, thevents occurred at different times and lasted for different dura-ions depending on the applied current density. This behavior haseen observed by others [17]. One notable behavior is that despitehe intermittences of the galvanostatic experiments for EIS mea-urements, the overall potential vs. time curves was smooth. Therrows in Fig. 8 indicate the times of interruption during the gal-anostatic experiment. The inset in Fig. 8 shows a segment of theotential vs. time curve in which two interruption events for EISeasurement took place. The restoration of the potential to the

alue before the interruption for EIS measurement, even duringscillation period, indicates the strong dependence of the potentialn the physical properties of the AgCl layer itself rather than onime dependence processes such as diffusion, at least in the timecale of the experiments. The smoothness of the overall potentials. time curves indicated the alternation between galvanostatic andIS experiments had negligible effect on the galvanostatic experi-ents.The resistances of the AgCl layers formed during galvanos-

atic experiments at different current densities were measured byIS. Fig. 9 shows the dependence of the AgCl layer resistance onhe amount of coulombs passed per unit area on the silver elec-rode during the formation of the AgCl layers at different appliednodic current densities of 0.1, 0.2, 0.5, 1 and 2 mA/cm2. The AgClesistances increased with the coulombs passed per unit area ororresponding as the resistance increased with the thickness of thegCl layer. However, the resistance of the AgCl layer was not depen-ent on the magnitude of the applied current density under which

t grew. Compared to the resistances of thin AgCl layers, the resis-ances of thicker AgCl films grown under different current densitiesere more scattered.

. Discussion

.1. Formation and growth of AgCl

The formation and growth of AgCl on silver substrate in 9 g/LaCl solution was studied during potentiodynamic polarization. At

mall anodic overpotentials, the anodic current density was small,.e. the concentration of dissolved Ag+ in the solution was small,nd no AgCl was observed on the silver substrate (Fig. 2a). As thenodic current density increased due to higher anodic overpoten-ial, the solution near the silver surface became more concentrated

Fig. 9. Dependence of AgCl layer resistances on coulomb passed per unit area duringgalvanostatic experiments at different applied current densities of 0.1, 0.2, 0.5, 1 and2 mA/cm2. The resistances of AgCl layer were taken as the impedance value at 100 Hzwhen the phase shift angle was ca. −1◦ .

with Ag+. Precipitation of AgCl occurred when the solution satu-rated with Ag+. AgCl particles with size of <100 nm were observedat the current density of ca. 1.3 × 10−6A/cm2 (Fig. 2b). AgCl patchesexpanded laterally at higher anodic overpotentials and current den-sities with preferential directions along scratches on the surface.This may be due to a higher extent of Ag+ supersaturation at thescratch bottom or a decrease in the free energy for nucleation atthese heterogeneous sites. However, no further study was con-ducted to elucidate this speculation.

When patches impinged and AgCl formed a continuous layer onthe silver substrate, the layer started thickening. Spaces betweenAgCl particles were visible during this period (Fig. 5b). The spacesbetween AgCl particles are likely to be the primary route for theionic transport between the dissolving silver electrode and theelectrolyte during this early stage of film growth. As the AgCllayer thickened, the spaces between the AgCl particles were closed(Fig. 6a), and the ionic transport via this route was limited. Inaddition, the field strength through AgCl layer was estimated onthe order of 103 V/cm, thus intrinsic conduction via solid AgCllayer by high field mechanism which needs a field strength onthe order of 10−5 V/cm or higher, was ruled out. Therefore micro-channels running through the AgCl particles are likely to be themain conduction path at this later stage. Fig. 10 summarizes thegrowth of AgCl patches and the evolution in the structure of AgCllayer.

The change in the morphology of the AgCl top surface and theadherence of the AgCl layer to the silver substrate indicates that theformation of AgCl occurred at both the top and bottom of the AgCllayer. This observation implies that there must be transport of Ag+

from the silver substrate toward the electrolyte and Cl− from theelectrolyte toward the silver electrode to support the continuousformation of AgCl. In the thin, continuous AgCl layer, Ag+ and Cl−

ions may be transported via pores between the AgCl particles. How-ever, in thick AgCl layers, micro-channels through AgCl are likelyto be the main path for ionic transport.

The effect of the AgCl corrosion product layer on limiting thecorrosion of the silver substrate has been reported [22]. This effect

of the AgCl layer is due to the retardation of the ionic transportthrough the layer. Transport of ions in an electrolyte can be dueto a concentration gradient (diffusion), to an electrostatic field(migration), or to a bulk fluid motion (convection). During silver

H. Ha, J. Payer / Electrochimica A

FdA

disO

p(sb

i

ws(baniwuEe

o

ig. 10. Schematics of AgCl growth and the evolution in the structure of AgCl layeruring potentiodynamic polarization in 9 g/L NaCl solution, (a) lateral growth ofgCl patches, (b) impingement of AgCl patches, and (c) thickening of AgCl layer.

issolution and AgCl growth, convection inside the micro-channelss negligible, therefore it does not affect the dissolution current ofilver. Nevertheless, the process can be under diffusion control orhmic control or mixed diffusion-Ohmic control.

The appearance of a constant current density at the highotential ranges during potentiodynamic polarization experimentsFig. 1) mimics the behavior of a limiting current density in diffu-ion controlled regime. The limiting current density, iL, is calculatedy:

L = zDFCb

(1 − tr)ı(2)

here z is the valence of the charge carrier; D is the diffu-ivity of the charge carrier (cm2/s); F is the Faraday constantF = 96,500 C/mol/equiv); Cb is the concentration of Cl− ion in theulk (mol/cm3); tr is the transport number of the charge carrier;nd ı is the diffusion layer thickness (cm). During potentiody-amic polarization experiments, the thickness of the AgCl layer

ncreased and therefore the diffusion layer thickness associatedith the micro-channels also increased. Hence, if the process was

nder diffusion control, the limiting current density calculated byq. (2) would decrease during the course of the potentiodynamicxperiment rather than kept a constant value.

The argument that the dissolution of silver after the formationf a continuous AgCl layer is under Ohmic control is needed to

cta 56 (2011) 2781–2791 2787

be examined. This argument implies that (i) if the resistance ofthe AgCl layer is constant, the rate of the silver dissolution willincreases proportionally with the overpotential and (ii) if the anodicdissolution current is controlled at a constant value, the over-potential at the electrode will increase proportional to the AgClresistance.

The results of the galvanostatic step experiments are shownin Fig. 7. In these experiments, the pulse duration was chosen as1 s to minimize the change in the AgCl thickness. Thus, the resis-tance of the AgCl layer is considered constant during the test. Fig. 7shows that the dissolution current of silver indeed increased lin-early with the overpotential. Extrapolation of this data crosses theorigin in the i vs. overpotential plot. These experiments validatepoint (i).

Point (ii) is validated by the galvanostatic experiments. Duringthese experiments, the thickness of the AgCl layer increased, andso did the resistance of the layer. An increase in the overpoten-tial was observed in Fig. 8 as the result of an increasing AgCl layerresistance. Furthermore, the restoration of the overpotential to thevalue before the interruption time for EIS measurement indicatesthe strong dependence of the overpotential on the resistance ofthe AgCl layer. Quantitative examination of this point will be per-formed after the conductivity of the AgCl layer is analyzed in thesection below.

4.2. Ionic conductivity of the electrolyte inside the micro-channels

The resistance vs. coulombs passed per unit area curves in Fig. 8shows that the resistance of the AgCl layer increased approximatelylinearly with the coulomb passed to a certain thickness from whichthe resistance increased at a faster rate. Assuming that the micro-channels were the main ionic transport path through the AgCl layer,the dependence of the AgCl layer resistance on the coulombs passedcan be explained as the result of the lengthening of the micro-channels and/or the thinning or restriction of some of the channels.The resistance of the AgCl layer, R, is calculated from the film thick-ness, x, by:

R = ˇ · x

AmcKmc· 1

N(˝) (3)

where ˇ is a structural factor representing the increase in theionic transport distance due to the tortuosity of the micro-channels(ˇ > 1); Amc is the average cross-section area of the micro-channels(cm2); Kmc is the conductivity of the electrolyte inside the micro-channels (S/cm); and N is the number of the micro-channelsrunning through the AgCl layer. As the film thickened, the ionictransport distance in the micro-channels increased proportionallywith the film thickness. If the conductivity of the electrolyte insidethe micro-channels does not change, and if the structural factor andthe number of the micro-channels are constant, the resistance ofthe AgCl layer will increase linearly with the film thickness. Theseconditions are more likely to be met during a short time of AgClgrowth, i.e. when the thickness of the AgCl layer is thin. Thereforea linear relationship was observed in the R vs. q curves at low qregions (Fig. 9).

For longer times, the precipitation of AgCl inside themicro-channels might occur if the saturation condition wasmet. Precipitation of AgCl was observed at both Ag/AgCl andAgCl/solution interfaces. Therefore it is likely that AgCl also pre-cipitates inside the micro-channels. Thus, the radiuses of themicro-channels are likely to decrease with time which resulted in a

reduction of the micro-channel cross-section area, Amc. Eventuallysome of the channels might be blocked and the number of micro-channels, N, decreased. In addition, the formation of AgCl at theAg/AgCl interface and inside the channels consumed Cl− and Ag+

and decreased the concentration of Cl− and/or Ag+ ions inside the

2 mica Acta 56 (2011) 2781–2791

cttq

mmaa

m

weiwtp(c

x

S

K

watnscmbs[K

c

K

wec(dNE

2

CTt1lb

C

1o

788 H. Ha, J. Payer / Electrochi

hannels, i.e. a decrease in the conductivity of the electrolyte insidehe micro-channels, Kmc. Either the decrease in Amc, N or Kmc leadso a faster increase in the resistance of the AgCl layer, and the R vs.curve deviates from a linear relationship.

To estimate the concentration of charge carriers inside theicro-channels, the conductivity of the electrolyte inside theicro-channels is calculated from Eq. (3). The AgCl layer formed

fter passing 6 C/cm2 is chosen for this estimation. Assuming thatll dissolved Ag+ cations precipitated as AgCl, Faraday’s law gives:

AgCl = A · x · �AgCl = MAgCl

n· A · q

F(4)

here mAgCl is the mass of AgCl (g); A is the surface area of thelectrode (cm2); x is the thickness of the AgCl layer (cm); �AgCls the density of the AgCl layer (g/cm3); MAgCl is the molecular

eight of AgCl (g/mol); n is the number of electron transferred inhe silver dissolution reaction (n = 1); q is the number of coulombassed per unit area (C/cm2); and F is the Faraday’s constantF = 96,500 C/mol/equiv). Therefore, the thickness of this layer isalculated by:

= MAgCl

nF· q

�AgCl(5)

ubstitution of Eq. (5) into Eq. (3) and rearranging yields:

mc = 1nF

· MAgCl

�AgCl· q

R· ˇ

AmcN(6)

From Fig. 9, the resistance of the AgCl layer after passing 6 C/cm2

as ca. 200 k�. The number of the micro-channels and the aver-ge cross-section area of the micro-channels were estimated fromhe SEM micrographs which gave approximately 14,400 chan-els and 0.5 × 10−8 cm2 cross-section area on the entire sampleurface of 2 × 10−3 cm2. Then, the micro-channels accounted fora. 4% volume fraction of the AgCl layer. The tortuosity of theicro-channels was assumed to increase the transport distance

y a factor of two, i.e. ˇ = 2. The molecular weight and the den-ity of AgCl was taken as 143.5 g/mol and 5.56 g/cm3, respectively23]. Substitution of the numerical values into Eq. (6) yieldedmc = 2.15 × 10−4 S/cm.

The concentration of the charge carriers inside the micro-hannel was estimated using the relation:

mc =∑

j

�jzjCj (7)

here j represents the charge carrier species of interest; �j is thequivalent ionic conductivity of the species (S cm2/mol); zj is theharge of the species; and Cj is the concentration of the speciesmol/cm3). In micro-channels, it is likely that Na+ and Cl− are theominant charge carriers. The equivalent ionic conductivities ofa+ and Cl− ions are 50 and 76 S cm2/mol, respectively [24]. Thus,q. (7) can be rewritten as:

.15 × 10−4 = 50CNa+ + 76CCl− (8)

This expression indicates that the concentrations of Na+ andl− inside the micro-channels were on the order of 10−6 mol/cm3.his concentration is approximately 100 times more diluted thanhe concentration of these ions in the bulk solution which is ca.0−4 mol/cm3. The concentration of Ag+ in the micro-channels is

imited by the solubility product of AgCl, Ksp, which is representedy:

Ag+ · CCl− ≤ Ksp (9)

At room temperature, the solubility product of AgCl is.8 × 10−16 (mol2/cm6) [23]. If the concentration of Cl− was on therder of 10−6 mol/cm3 as calculated earlier, the concentration of

Fig. 11. The apparent conductivity of the AgCl layers during galvanostatic exper-iments in 9 g/L NaCl solution at different current densities of 0.1, 0.2, 0.5, 1 and2 mA/cm2. The apparent conductivities were calculated from Eq. (10).

Ag+ in the micro-channels was on the order of 10−10 mol/cm3 orsmaller.

4.3. The apparent conductivity of the AgCl layer

In addition to the conductivity of the electrolyte inside themicro-channels, another important parameter is the apparent con-ductivity of the AgCl layer. Based on the resistance of the AgCl layermeasured from EIS, the apparent conductivity of the AgCl layer, K,is calculated by:

K = x

R · A(10)

where x is the thickness of the AgCl layer (cm); R is the resistance ofthe AgCl layer (�); and A is the surface area of the electrode (cm2).The thickness of the AgCl films was calculated from Eq. (5). Theresistance of the AgCl layer at different coulombs passed per unitarea was taken from EIS measurements.

The apparent conductivity of the AgCl layer calculated at dif-ferent charges passed per unit area is shown in Fig. 11. A trend ofdecrease in the apparent conductivity as the film grew thicker wasobserved. After passing an amount of 10 C/cm2, the apparent con-ductivity of the AgCl layer was approximately 2 × 10−6 S/cm whichwas 5 times lower than the conductivity of the film after passing2 C/cm2. The apparent conductivities obtained in this study are con-sistent with the values of the AgCl conductivities reported in theliterature and as summarized in Fig. 12 [6,11,25]. The decrease inthe AgCl apparent conductivity with the film thickness supportsthe hypothesis of either decrease in the number of micro-channels,the restriction of some of the channels or the decrease of thecharge carrier concentration inside micro-channels as the filmgrows.

4.4. The effect of the continuous AgCl layer on the silverdissolution mechanism

As the apparent conductivity of the AgCl layer decreases withthe thickness, the transport via this layer becomes more diffi-

cult. The continuous increase of the applied potential to pass thesame amount of current in the galvanostatic experiment is due tothis effect. The applied potential, V, is expressed in terms of theelectrode potential, Ea, the anodic activation overpotential, �a

a, theanodic concentration overpotential, �a

c , and the Ohmic overpoten-

H. Ha, J. Payer / Electrochimica Acta 56 (2011) 2781–2791 2789

10-1 100 101 102 103 104

10-8

10-7

10-6

10-5

10-4

4N KCl, Beck & Rice 4N NaCl, Beck & Rice 6N NaCl, Beck & Rice 6N HCl, Beck & Rice

0.154 M NaCl, Ha & Payer 0.1N KCl, Kurtz et al. 0.1M HCl, Jin et al. 1M KCl, Lal et al. 0.5M KCl, Lal et al. 0.25M KCl, Lal et al. 0.1M KCl, Lal et al.

K (

Scm

-1)

i (mA/cm2 )

Fig. 12. Comparison between the conductivities of AgCl layer obtained in this studyacf[

t

V

te

�

�

aop

sicstbawcoecfdmA

ost(m

Table 1Summary of the calculated exchange current density.

i (mA/cm2) �aa (V) io (mA/cm2) io (average)

(mA/cm2)Standarddeviation

0.1 0.021 0.119 0.13 0.03

nd literature data. The AgCl layers in this study were formed at different appliedurrent densities in 0.154 M NaCl solution. The AgCl layers in the literature wereormed at different applied current densities and in different chloride solutions6,11,25].

ial, �˝, by:

= Ea + �aa + �a

c + �˝ (11)

The Ohmic overpotential at the silver electrode is mainly due tohe Ohmic resistance of the AgCl layer. Thus, during galvanostaticxperiments, the Ohmic overpotential is expressed by Ohm’s law:

˝ = i · x

K(12)

Combination of Eqs. (5) and (12) yields:

˝ = MAgCl

nF· q

�AgCl· i

K(13)

The galvanostatic polarization experiments were performed atpplied current densities of 0.1, 0.2, 0.5, 1 and 2 mA/cm2. The valuesf the apparent AgCl conductivity after passing different chargeser unit area were calculated earlier.

Fig. 13 shows the overpotential vs. time curves during galvano-tatic polarization experiments calculated from Eq. (13) plottedn the same graph with the experimental data. In all cases, thealculated curves resembled the experimental data with some off-ets. This indicated the dominance of the Ohmic overpotentialhrough the AgCl layer during the dissolution of the silver electrodeeneath thick AgCl layers. For instance, the Ohmic overpotentialfter 4 h at the applied current density of 0.5 mA/cm2 (Fig. 13c)as accounted for ca. 80% of the measured potential. At the applied

urrent densities of 1 and 2 mA/cm2, some deviation in the slopesf the calculated curves from the experimental curves toward thends of these experiments was observed and may be due to thehange in the ionic conduction mechanism through the AgCl layerrom extrinsic to an intrinsic one. At these high applied currentensities and therefore high overpotentials, the conduction viaicro-channels was complemented by high field conduction ingCl solid phase as observed by others [25].

Fig. 13 also shows that there are offsets between the calculated

verpotentials and the measured potentials. These offsets are theum of the electrode potential, the anodic activation overpoten-ial and the anodic concentration overpotential as indicated in Eq.11). The offsets between the calculated overpotentials and the

easured potentials were measured at the time when 2 C/cm2 has

0.2 0.033 0.1450.5 0.053 0.2041 0.103 0.137

been passed, i.e. at the first EIS measurements. Fig. 14 shows thedependence of the offset values on the applied current density inthe galvanostatic experiments. A linear relationship between theoffset values and the applied current density was observed. Thisfinding indicates that beside the Ohmic drop through the AgCl layer,another current dependence factor contributed to the overall over-potential at the silver electrode at the beginning of the experiments,i.e. when the AgCl layer was thin.

From Eq. (11), the offset values can be expressed as:

Offset = Ea + �aa + �c

a (14)

The concentration overpotential for the anodic reaction isexpressed by:

�ac = RT

nF· ln

(1 + |i|

iL

)(15)

In the above equations, R is the gas constant (R = 8.314 J/mol K); Tis temperature (K); io is the exchange current density of the silverdissolution reaction (A/cm2); and iL is the limiting current densityof the silver dissolution reaction (A/cm2). Eq. (15) does not indicatean overpotential limit, unlike the one for the cathodic reaction. Atthe anode, insufficient diffusion does not cause depletion of thereactant in the solution but may result in precipitation at the elec-trode when the solution is saturated with cations released from theanodic reaction. If the precipitation forms an insulation layer, thecurrent will be physically limited unless the layer is porous. In thesituations considered here, since the AgCl deposit remains porous,the anodic concentration overpotential as indicated in Eq. (15) isnegligible. Thus, the anodic activation overpotential is calculatedby:

�aa = Offset − Ea (16)

The value of the electrode potential, Ea, is taken as the Nern-stian potential of a Ag/AgCl electrode in 9 g/L NaCl solution(Ea = +0.033 VSCE). Therefore, the exchange current density, io, canbe obtained from the Butler–Volmer equation:

io = i

exp((˛aF/RT)�aa) − exp((−˛cF/RT)�a

a)(17)

where ˛a, ˛c are the charge transfer coefficients of the anodic andcathodic reactions, respectively, at the silver electrode (˛a + ˛c ≈ n).Assuming that ˛a = ˛c = 0.5, the value of the exchange current den-sity for the silver dissolution reaction at the temperature of 298 Kcalculated from Eq. (17) is summarized in Table 1.

The value of the exchange current density for the Ag/Ag+ redoxreaction varied over large ranges depending on the nature of theelectrolyte. An extremely high exchange current density on theorder of 10−1 A/cm2 for Ag/Ag+ redox reaction in perchlorate solu-tions has been reported in literature [26,27]. In other cases, theexchange current density in the range of 10−5 to 10−3 A/cm2 for the

silver electrodeposition in cyanide electrolyte has been reported[28–30]. However no value was found for the reaction in dilutedNaCl solution.

Since the concentration overpotential in the situation consid-ered here is negligible, the measured potential given in Eq. (11) can

2790 H. Ha, J. Payer / Electrochimica Acta 56 (2011) 2781–2791

F larizaw ere o

b

V

tooia

ig. 13. Calculated and measured potential vs. time curves during galvanostatic poere (a) 0.1; (b) 0.2; (c) 0.5; (d) 1 and (e) 2 mA/cm2. The calculated overpotentials w

e rewritten as:

= Ea + x

K· i + �a

a (18)

This expression indicates the contribution of the Ohmic overpo-

ential and the activation overpotential on the dissolution kineticsf silver. When AgCl is a non-continuous layer, the value of the sec-nd term in Eq. (18), which represents the Ohmic overpotential,s small in comparison with the third term, which represents thectivation overpotential, therefore the activation overpotential is

tion experiment in physiological 9 g/L NaCl solution. The applied current densitiesbtained from Eq. (13).

dominant and determines the dissolution current. When the AgClis a thin, continuous layer, e.g. after 1 h at i = 0.5 mA/cm2, the secondterm and the third term in Eq. (18) are of the same order, thereforethe dissolution of silver is under mixed Ohmic-activation control.

After a long time when the thickness of the AgCl layer becomeslarge, e.g. after 4 h at i = 0.5 mA/cm2, the second term in Eq. (18)dominates the last term. Thus, after a long enough time, the acti-vation overpotential becomes negligible, and the silver dissolutionis under Ohmic control.

H. Ha, J. Payer / Electrochimica A

0.0 0.5 1.0 1.5 2.0 2.50.00

0.05

0.10

0.15

0.20

0.25O

ffse

t (V

)

Fp

5

pnbnlttswTdwfi

omTertn9

ttaoOpoc

A

aNHs

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[28] G. Baltrunas, The mechanism of electrode process in the system silver/silver

i (mA/cm2 )

ig. 14. Dependence of the offsets between the calculated and the measured over-otentials on the applied current densities during galvanostatic experiments.

. Conclusion

The precipitation and growth of AgCl on a silver substrate inhysiological 9 g/L NaCl solution were investigated. AgCl particlesucleated at the bottom of the scratches on the surface which maye less effective sites for diffusion or favorable sites for heteroge-eous nucleation. The particles grew to patches which expanding

aterally on the substrate until forming a continuous film. Ionicransport through the newly formed continuous AgCl film was viahe spaces between the AgCl grains. As the film thickened, thepaces between the AgCl grains were closed and the ionic transportas mainly via micro-channels running through the AgCl grains.

he Ohmic resistance of the continuous AgCl layer restricted theissolution of the silver substrate. At high current densities andhen the AgCl layer is thick, the conduction mechanism transits

rom extrinsic, i.e. via electrolyte in the micro-channels, to intrinsic,.e. via AgCl solid phase.

The micro-channels were accounted for ca. 4% volume fractionf the AgCl layer. The conductivity of the electrolyte inside theicro-channels was estimated to be on the order of 10−4 S/cm.

he concentration of Cl− and Ag+ inside the micro-channels wasstimated to be on the order of 10−6 mol/cm3 and 10−10 mol/cm3,espectively. The apparent conductivity of the AgCl layer was onhe order of 10−6 to 10−5 S/cm and decreased with the film thick-ess. The exchange current density of the Ag/Ag+ redox reaction ing/L NaCl was ca. 1.3 × 10−4 A/cm2.

In the current density range from 0.1 to 2 mA/cm2, the concen-ration overpotential at the silver electrode was negligible. Whenhe AgCl layer is non-continuous, the dissolution of silver is underctivation control. When the AgCl layer is thin, i.e. on the orderf several �m, the dissolution of silver is under mixed control ofhmic overpotential through the AgCl layer and activation over-otential at the silver electrode. When the AgCl layer is thick, i.e.n the order of tens of �m or thicker, the dissolution of silver isontrolled only by the Ohmic overpotential through the AgCl layer.

cknowledgements

This research is supported by NIH-NINDS Grant No. NS-041809nd NIH-NIBIB Grant No. EB-001740, Development of Implantableetworked Neuroprosthesis System (NPPS), Principal Investigatorunter Peckham. We are grateful to Dr. Brian Smith for his discus-

ions and support along with others in the Cleveland Functional

[

[

cta 56 (2011) 2781–2791 2791

Electrical Stimulation (FES) Center at CWRU. This group providedtechnical support and assistance during this work. Useful discus-sions with Professor John Lewandowski and Professor Uziel Landauat CWRU are highly appreciated. The financial support of the Viet-nam Education Foundation to one of the authors (HH) during thestudy is acknowledged.

References

[1] E.F. Sieckmann, E. Grunwald, The standard E.m.f. of the silver–silver chlorideelectrode and the degenerate activity coefficient of chloride ion in the sys-tem ethanol–water1a, Journal of the American Chemical Society 76 (14) (1954)3855.

[2] W.R. Carmody, A study of the silver chloride electrode, Journal of the AmericanChemical Society 51 (10) (1929) 2901.

[3] P. Bro, N. Marincic, The high rate oxidation of silver electrodes in chloridesolutions, Journal of the Electrochemical Society 116 (10) (1969) 1338.

[4] T. Katan, S. Szpak, N.B. Douglas, Silver/silver chloride electrode: reaction pathson discharge, Journal of the Electrochemical Society 120 (7) (1973) 883.

[5] S. Szpak, A. Nedoluha, T. Katan, Silver/silver chloride electrode: charging of aporous structure, Journal of the Electrochemical Society 122 (8) (1975) 1054.

[6] X. Jin, et al., The electrochemical formation and reduction of a thick AgCl depo-sition layer on a silver substrate, Journal of Electroanalytical Chemistry 542(2003) 85.

[7] J. Dunning, D. Bennion, J. Newman, Analysis of porous electrodes with sparinglysoluble reactants, Journal of the Electrochemical Society 120 (7) (1973) 906.

[8] S. Szpak, T. Katan, An experimental study of reaction profiles in porous elec-trodes, Journal of the Electrochemical Society 122 (8) (1975) 1063.

[9] C. Cachet, et al., The Ag/Ag+ system: an impedance model for nucleation andgrowth, Journal of Electroanalytical Chemistry 100 (1979) 745.

10] V.I. Birss, C.K. Smith, The anodic behavior of silver in chloride solution-I. Theformation and reduction of thin silver chloride films, Electrochemica Acta 32(2) (1987) 259.

11] H. Lal, H.R. Thirsk, W.F.K.W. Jones, A study of the behavior of polarized elec-trodes. Part I – the silver/silver halide system, Transactions of the FaradaySociety 47 (1951) 70.

12] R.D. Giles, The anodic behaviour of silver single crystal electrodes in con-centrated chloride solutions, Electroanalytical Chemistry and InterfacialElectrochemistry 27 (1970) 11.

13] G.T. Burstein, R.D.K. Misra, Growth of oxidized monolayers on scratched silverelectrodes in chloride solutions, Electrochimica Acta 28 (3) (1983) 371.

14] S. Jaya, T.P. Rao, G.P. Rao, Mono- and multilayer formation studied of silver chlo-ride on silver electrodes from chloride-containing solutions, Journal of AppliedElectrochemistry 17 (1987) 635.

15] B.M. Jovic, V.D. Jovic, D.M. Drazic, Kinetics of chloride ion adsorption and themechanism of AgCl layer formation on the (1 1 1), (1 0 0) and (1 1 0) faces ofsilver, Journal of Electroanalytical Chemistry 399 (1–2) (1995) 197.

16] A.G. Brolo, S.D. Sharma, Using probe beam deflection (PBD) to investigate theelectrochemical oxidation of silver in perchlorate media in the presence andabsence of chloride ions, Electrochimica Acta 48 (10) (2003) 1375.

17] B. Bozzini, G. Giovannelli, C. Mele, Electrochemical dynamics and structure ofthe Ag/AgCl interface in chloride-containing aqueous solutions, Surface andCoatings Technology 201 (8) (2007) 4619.

18] T. Katan, S. Szpak, N.B. Douglas, Silver/silver chloride electrodes: surface mor-phology on charging and discharging, Journal of the Electrochemical Society121 (6) (1974) 757.

19] J.H. Payer, et al., Role of transport properties in corrosion product growth,Materials Science and Engineering: A 198 (1995) 91.

20] B. Smith, et al., An externally powered, multichannel, implantable stimulator-telemeter for control of paralyzed muscle, Biomedical Engineering, IEEETransactions on 45 (4) (1998) 463.

21] B. Smith, et al., Development of an implantable networked neuroprosthesis,in: Neural Engineering, 2005. Conference Proceedings. 2nd International IEEEEMBS Conference on, 2005.

22] H.M. Ha, J.H. Payer, Corrosion of Silver-cored MP35N LT Composite for Net-worked Neuroprosthetic System. Corrosion (accepted).

23] R.C. Weast, Handbook of Chemistry and Physics, The Chemical Rubber Co,Cleveland, OH, 1969.

24] J. Newman, K.E. Thomas-Alyea, Electrochemical Systems, 3rd edition, JohnWiley & Son, Inc., 2004.

25] T.R. Beck, D.E. Rice, Conductivity of anodic silver chloride during formation,Journal of the Electrochemical Society 131 (1) (1984) 89.

26] W. Mehl, J.O.M. Bockris, Mechanism of electrolytic silver deposition and disso-lution, The Journal of Chemical Physics 27 (3) (1957) 818.

27] A.R. Despic, J.O.M. Bockris, Kinetics of the deposition and dissolution of silver,The Journal of Chemical Physics 32 (2) (1960) 389.

cyanide complexes, Electrochimica Acta 48 (24) (2003) 3659.29] M. Saitou, Kinetic electrode reactions in silver electrodeposition using a mul-

tipulse current measurement, Journal of the Electrochemical Society 152 (3)(2005) C113.

30] G. Senanayake, The cyanidation of silver metal: review of kinetics and reactionmechanism, Hydrometallurgy 81 (2) (2006) 75.