Anion effects on the underpotential deposition of lead on Cu(111)

ANION

EFFECTS DEPOSITION

ON THE UNDERPOTENTIAL OF LEAD ON CU(lll)

J.R. VILCHE*

and K. J~~TTNER+

*Institute de Investigaciones Fisicoquimicas Teoricas y Aplicadas (INIFTA), Facuitad de Ciencias Exactas, Universidad National de La Plats, Sucursal 4-Casilla de Correo 16, (1900) La Plats, Argentina ‘Institute of Physical Chemistry and Electrochemistry, University of Karlsruhe, Kaiserstrasse 12, 7500 Karlsruhe 1, Germany (Received

29 January

1987; in revised form

23 March 1987)

Abstract-The underpotential deposition of lead on Cu(l11) single crystal surfaces was studied in aqueous solutionscontaining different anions such as perchlorate, acetate, and chlorideat various pH within the range 1.5 G pH < 3.55 using voltammetricand complex potential controlled perturbation techniques. The UPD of lead appears to be a highly irreversible process which depends on pH, the electrical history of the metal-electrolyte interface and the type of anions present in the electrolyte. A critical threshold potential for the cathodic monolayer formation is revealed by the cyclic voltammogramms as well as by the charge isotherms. The nucleation like. behaviour is interpreted in terms of a negative 2-D nucleation and growth process within a preformed surface layer consisting of specifically adsorbed anions or oxygen containing

INTRODUCTION

EXPERIMENTAL

The underpotential deposition (UPD) of metals on foreign substrates is a well known phenomenon which has been extensively studied in a great variety of systems during the last two decades. An overview of the subject can be found in several review papers[lA]. In the last few years the main point of interest was directed towards measurements on single crystal substrates in order to find specific correlations between the substrate orientation and the thermodynamic, kinetic, structural and catalytic properties of the deposited metal[3,4]. The results were mainly explained by the formation of well ordered superlattice structures with defined equilibrium adsorption properties. However, in many systems there are strong indications that UPD is a more complex process presumably involving 2-D nucleation and growth phenomena, surface reconstruction assisted by place exchange between atoms of the substrate and the deposited metal, and other interactions between the metal deposit and preexisting surface species leading to highly irreversible and time dependent UPD behaviour[5-91. The UPD of lead on Cu(ll1) substrates is one of the systems which exhibit a remarkably irregular behaviour as was already reported by Bewick et al.[lO] and Siegenthaler and Jiittner[9]. In the former paper a 2-D nucleation of lead was postulated to occur on the Cu(l11) substrate in solutions containing acetate ions at constant pH, while in the latter work a strong pH dependence was established and 2-D nucleation due to the desorption of a removable oxygen coverage was observed at pH > 1.5 in the absence of specifically adsorbed anions. The aim of the present paper is to elucidate both the influence of pH on the formation of oxygen containing surface species and the effect of anions present in the electrolyte on the kinetics and mechanism of the UPD of lead on Cu(ll1) faces.

Potentiodynamic measurements were carried out on the (111) plane of a cylindrical copper single crystal (Materials Research Ltd.) applying the “dipping technique”, originally developed by Bachmann et al.[ 111 and later on employed by Dickertmann et aI.[12] for UPD studies without use of any embedding materials. The pretreatment of the electrode surface (0.38 cm’ geometrical area) consisted of successive mechanical polishing with different grades of diamond paste (-3 pm) followed by anodic electropolishing under a controlled current density of ca 2.5 Acme2 on a polyester cloth soaked with a solution containing 130 ml H3P04 (85 %) + 20 ml HZSO, (95 %) + 60 ml HzO. A more detailed description of the polishing procedure has been given elsewhereC9, 133. The supporting (blank electrolyte, 0.5 M NaClO, +x M HCIOI (1.5 < pH < 3.55) aqueous solution, was prepared from HClO, and NaOH Suprapure (Merck) grade reagents and triply distilled water. The UPD of Pb was studied at fixed pH 3.55 with additions Of:

(i) 5 x 10e3 M Pb(C104)2, prepared from p-a. Pb(CO& and HC104 Suprapure (Merck) reagents; (ii) 10m2 M Pb(CH3COO),, p.a. grade (Merck); and (iii) lo-* M Pb(CHPCOO)L fy M NaCl (lo-’ < y < lo-“), with Suprapure NaCl (Merck). Measurements were made under purified nitrogen gas atmosphere at 298 K in a conventional three compartment glass cell. The counter electrode was a large area platinum sheet and the reference electrode was either a Hg/Hg,SO,, Na,SO, (sat.) electrode for solutions free of Pb’+ ions or a Pt sphere covered with electrodeposited lead in the case of solutions containing Pb2+ ions. Potentials in the text are referred either to the nhe scale (E”) or to the reversible potential of the

1567

J. R. VILCHEAND

1568

Pb/Pb’+ electrode in the corresponding solution (underpotential scale E-EPbIPbz +). The special polarization routines used are depicted in the figures.

RESULTS Blank

solutions

The potentiodynamic behaviour of the Cu(ll1) surface in solutions free of Pb2+ ions was investigated in the potential region between the onset of the anodic copper dissolution and the cathodic hydrogen evolution in dependence on the solution pH, the potential scan rate (II), and the anodic (I&_ .) and cathodic (I?., .) switching potentials of single and repetitive triangular potential sweeps. Figure 1 shows typical cyclic voltammograms recorded between the limits Es. L1= 200 mV and Es* E = - 470 mV at different v in the blank solution at pH 1.5. The voltammograms reveal two clearly distinguishable peaks, one anodic and one cathodic, with an overall charge in the order of 110-130 PC cm-* after base line correction. Both peak potentials and peak currents exhibit a pronounced dependence on the scan rate D. Progressively changing either E,, D or Es, c gave experimental proof for a strong correlation between the electro-oxidation and electro-reduction processes associated with these peaks. Correspondingly, the anodic (Q.) and cathodic (Q,) charges varied simultaneously while their Q,,/Qcratio remained close to one. The same type of measurements described in Fig. 1 were also performed at different pH values in the range 1.5 < pH 4 3.55. From a complete set of such experiments in blank solutions the dependence of the cathodic and anodic peaks on both v and pH was evaluated. The results are presented in Figs 2 and 3. At constant v, both the anodic E,. (1and the cathodic E,,. peak potentials fit

K. JOITNER linear E, vs pH relationships with a mean slope of about (aE,,/apH), = - 30 mV (Fig. 2). On the other hand, at constant pH (Fig. 3), EP, L1 as well as E,, Efit also linear Ep us logu relationshlps, their slopes being (dEp, ,/alog vjBH,x and (W. =/a log U&H z 120 mV - 60 mV, respectively. These results mdicate slgmficantly different kinetics for the corresponding anodic and cathodic processes. The remarkably large peak potential difference, E,,. .-E,.,, which decreases from cn

-5ool 1

2

3

L

PH Fig. 2. pH-dependence of the anodic (full symbols) and cathodic (open symbols) peak potentials obtained from cyclic voltammograms in blank solution at different scan rates U.

csthodi c

CU(llO/CtOi

ptl

1.5

-M\/ -50 -500

-LOO

20 mv5’

-3co

-100

-200

0

100

a

EH/mV

Fig. 1. Cyclic voltammograms

-I

recorded at different scan rates u = 20, 15, 10, and 6 mV s-’ in the system: CU( 111)/0.5 M NaClO, + HC104 (pH 1.5).

Anion effects on the UPD of Pb on Cu(ll1) ,200

1.5

1569

. Cu~llOl Pb2: CIO;

b_

o, mVr,

300 z ‘0

0.5 -

u? ,400

.?I00

Fig. 3. Dependence of the anodic

(E 3 and cathodic (E& peak potentials on scan rate v from cyc PICvoltammograms run in blank solution at pH 1.5 and pH 3.55.

-1.0.

-1.5

0

N)O

2ca

300

4

I

E -Epb,pb2+hV

350mVatv=30mVs-‘to2OOmVatu=3mVs-’, reveals a strong irreversibility of the overall process. In addition to these facts the observed pH dependence indicates that the processes associated with the observed peak couple involve the electroformation and electroreduction of an O-containing surface film, although the experimentally found pH dependence deviates from the thermodynamically expected value (aE/apH) = - 59 mV at 298 K. This is probably due to the observed high irreversibility of the electrooxidatiorwzlectro-reduction processes occurring at the metal surface. UPD

of Pb on Cu(ll1)

Fig. 4. Pseudocapacitance iv-’ vs E- Epb,pbz+ plots obtained at different scan rates u = 10, 1, 0.1 mV SK’ in the system: Cu(l11)/0.5 M NaCIO, + 5 x IO-’ M Pb(ClO& + HC104 (pH 3.55)

1.0,

. Cu (111)IPbh.C14 pH 3.55

in solutions containing Cl02

The potentiodynamic response of the Cu(1 ll)/ 0.5 M NaClO, +0.05 M Pb(ClO& + HC104 (pH 3.55) interface has been analyzed at potentials within the underpotential range Epb,pba+ 4 E < Epb,pbz+ OS nhe. + 370 mV, with EPb/Pb’+ = -200 mV iv-’ OS _&Epb{pb’+ plots obtained at Characteristic different values of v are shown in Fig. 4. The cathodic branch exhibits a single sharp peak with a relatively sharp increase of the current at a distinct underpotential, &Epb,pbl+ = 200 mV, and a small shoulder which becomes visible at lower underpotential values as u decreases. With increasing v two main contributions can be also distinguished in the anodic branch. The overall anodic iv-’ vs &Epb,pb’+ profile shifts to more positive underpotentials and the highly symmetric peak contour changes to a composite structure as v increases. The separation of the two contributions is observed at higher scan rates v and implies two energetically different surface states. From the results obtained in blank solutions at the same pH it is evident that the UPD process of Pb on Cu(ll1) takes place in the potential range where the detected oxygencontaining surface film is prominent. For comparison, linear potential sweeps including a constant polarization time of 7 = 60 min at a fixed underpotential, E-EpbIpbt + = 220 mV, just preceding the cathodic peak, have been performed. The dependence of the voltammetric peak profile on the potential scan rate v obtained under these conditions is clearly seen in Fig. 5. After the long-time polarization, the cathodic peak exhibits a higher degree of symmetry

Fig. 5. Pseudocapacitance iw-’ DS E-EPVPbz+ plots obtained in the system: Cu(ll1)/0.5 M NaClO, + 5 x 10e3 M Pb(CIO,), + HClO, (pH 3.55) according to the polarization routine indicated in the figure. The stabilized cyclic voltammogram (o = 1 mV s-l) was stopped during the cathodic sweep at (E-EpblPbz+) = 220 mV and after a fixed holding time of T = 60 min the potential sweep was started again at different scan rates u’= 0.1, 1, and 10 mVs_‘.

and its peak potential shifts to lower underpotentials as the scan rate u increases. The overall charge remains practically constant. These results differ clearly from the response obtained under cyclic voltammetric conditions shown in Fig. 4. However, in both types of experiments the strong influence of the polarization routine on the UPD process is evident. This can be attributed to significant variation of the surface properties of the oxygen-containing surface film, as is clearly revealed in the voltammograms recorded under extremely severe conditions of long potential holding times preceding the cathodic potential sweep recorded at a very low sweep rafq (Fig. 6).

1570

J. R. VILCHE AND K. JWTNER

i

i clo;*cn~oa*c~~ II

Cu (lll)/Pb’

ii

Ii

N

s

o

Ii

E

5_i

-0.5 -1.0 I I

. 0

100

200

300

100

500

E-Epblpbz./mV

I

Fig. 6. Dependence of the holding time at E-E,,@+ = 200 mV on the cathodic voltammogram run at a low sweep rate v’= 0.05 mVs_‘. The stabilized voltammogram at v = 1 mV s-’ was previously attained after 360 min potential

cyclic over the whole underpotential range.

Obviously, the formation of the Pb-coverage under UPD conditions is accomplished by a partial desorption of oxygen-containing surface species. Siegenthaler et aI.[9] came to the same conclusion from the results of twin-electrode thin layer (TTL) experiments. They found significant deviations of the electrosorption valency (y > z) from the ideal charge coverage stoichiometry (y = z) for the UPD of Pb in the same system.

Effect of CH,COOCu(ll1)

and Cl-

on the UPD

of Pb on

Three cyclic voltammograms obtained at constant u = 1 mVs_l in perchlorate solutions in the absence and presence of CH,COO - and Cl - at constant pH 3.55 are shown in Fig. 7. On addition of CH,COOand Cl- the main UPD peaks shift towards lower underpotentials as compared with the perchlorate solution. In the case of CH,COOa splitting of the anodic peak is seen. The addition of Cl- even in small amounts (5 x lo-’ M NaCI) provokes strong enhancement of the peak heights and a simultaneous decrease in the half width of the peaks. The observed specific dependence of the UPD process on the nature of the anions present in the solution can be interpreted in terms of strong specific interactions between these anions, the underpotential lead deposit, and the oxygen containing species at the Cu(l11) surface. It should be noted that the formation of lead complexes in the solution can be neglected under the present experimental conditions. Competitive adsorption and co-desorption of surface active anions should also be considered as a possible explanation for the observed anion effects on the UPD of Pb on the Cu(ll1) substrate. Charge

isotherms

Two polarization routines PRl and PR2 have been employed to determine the charge density q,associated with the UPD of Pb on Cu(ll1) as a function of the underpotential E-Epb,pp+. The polarization prog-

ii ii ii j! ; i cIo~*cH.po-

‘CL!

I !

0

Km

. 200

300

E-Epbp+/mV

Fig. 7. Effect of anions on the cyclic voltammograms = 1 mVs_ ‘) of UPD of Pb in the system: Cu(l1 l)/O.S NaClO,+HClO, (pH 3.55) with addition of either. x lo-” M (----); or Pb(ClO& x IO-+ M Pb(CH,COO), 1 x lo-’ (-); Pb(CH,COO)2 + 5 x 10-s M NaC?(-.-.-.-).

(o M 5 1 M

rams PRl and PR2 are shown schematically in Fig. 8. In both routines the equilibrium charge q,, at fixed underpotentials is determined by anodic stripping subsequent to 15 min holding time at the corresponding underpotential. In the case of PRl the equilibrium state is attained from a surface free of lead whereas in the case of PR2 the surface is at first fully covered with lead under UPD conditions. This means that the equilibrium state is approached from two opposite q. us E-Epb,P,,z + curves directions. The corresponding are denoted in the following as PRland PR2isotherms. Figure 8 shows the 4. isotherms obtained in the perchlorate solution in the absence and presence of CH,COOions. In the presence of CH,COOthe isotherms are shifted by about 20 mV towards lower underpotentials relative to the isotherms in the perchlorate system in accordance with the observed shift of the peak potentials of the cyclic voltammograms in Fig. 6. Moreover the steepness of the isotherm steps increases when CHaCOOis present in the solution. In both systems a remarkable hysteresis between PRland PRZ-isotherms of at least 10 mV is observed. This hysteresis suggests a first order phase transformation associated with the UPD process within a defined

Anion effects on the UPD of Pb on Cu( 111)

100

I

Fig. 9. Dependence of the stripping charge difference 6q, = q,‘-q, on the amplitude 6E (curve a) and duration 6t (curve b) of a cathodic “overpotential” pulse SE(&) applied at a fixed underpotential AE, = E,-EPbIPbl+ in the system Cu(l1 l)/Pb’+, CIO;. The ql’ and q, values are the stripping charges in the presence and absence of the pulse excitation, respectively.

Fig. 8. Charge density isotherms q. us E-EPbIPbz + obtained with the polarization routines PRl and PR2 in the system: Cu( 111)/0.5M NaCIO, + HCIO, (pH 3.55) with addition of either 5 x lo-” M Pb(CIO,)Z (0, o) or I x 10e2 M Pb(CH,COO), (A, A). Open and full symbols correspond to PRl and PR2 polarization routines, respectively.

narrow underpotential range of only few millivolts. However, a clear discontinuity in the isotherm as sufficient criterion for a first order phase transition process[3, 14-163 is neither observed in the absence nor in the presence of CHSCOO ions in the solution. The saturation charge of qs = 320 PC cm-’ is found to be close to the theoretical charge value of 310 ~Ccm-* for a close packed Pb monolayer in both systems[9, lo]. Potential

pulse

1571

experiments

obtain further evidence for a first order phase transition envolved in the UPD process the following potential pulse experiments were carried out in the perchlorate system. After attaining a stable cyclic UPD voltammogram at a sweep rate u = 1 mVs_’ the cathodic potential sweep was stopped for 5 min at a fixed underpotential AE, = Sr-SRb/Pbl+ = 200 mV slightly positive with respect to the critical threshold potential at which the UPD of Pb starts to proceed at an appreciable rate. After 5 min the charge q, associated with this initial state was determined by anodic stripping. Due to the strongly retarded UPD process at this stage q1 amounts to only few @cm’. On the other hand, when the polarization routine includes an additional cathodic “overpotential” pulse of defined amplitude aE and duration at just on reaching the defined amplitude aE and duration at just on reaching the underpotential AEl, the stripping charge q’, increases remarkably. Obviously the cathodic pulse initiates a nucleation and growth process at

lower underpotentials. The charge difference 8q1 = (q’,-q,) which corresponds to the amount of lead deposited due to this nucleation and growth initiation is shown in Fig. 9 as a function of aE and at. For short dt values (at = 5 ms) the charge 84, increases practically linearly with aE approaching a limiting value at pulse amplitudes L?E > 50 mV (curve (I in Fig. 9). For constant aE = 50 mV a fast increase of dq, with at is observed and a saturation value is obtained after about 15 ms (curve b in Fig. 9). The saturation charge correlates satisfactorily well with the value indicated by the PR2 isotherm in Fig. 8 at the corresponding underpotential AS,. These results support the idea of a 2-D nucleation like process participating in the UPD of Pb on Cu(ll1). However, it should be mentioned that the growth of a complete lead monolayer could not be initiated by such pulse experiments.

To

DISCUSSION The experimental results clearly show that the UPD of Pb on a Cu(ll1) surface is a rather complex process which depends on pH, on the electrical history of the metalelectrolyte interface and also strongly on the type of anions present in the solution. A highly irreversible UPD process is indicated by its dynamic behaviour under different polarization conditions, ie the dependence of the cyclic voltammograms on both sweep rate and prepolarization and the influence of the polarization routines on the charge isotherms. These irreversibilities can not simply be attributed to pure 2D nucleation and growth processes of Pb. Strong interactions with oxygen-containing surface species (Figs l-3) and adsorbed anions should be taken into account for the kinetic interpretation of these effects. On the other hand the appearance of a critical threshold potential in the cathodic branch of the

1572

J. R. VILCHEAND K. JOTTNER

voltammograms as well as in the charge isotherms (Figs 4-8) point to a first order phase transition presumably involving 2-D nucleation and growth in the UPD process. This critical threshold potential slightly depends on the solution pH and on the type of anions present in the solution. Similar results have been reported by Bewick et al. [lo] in the case of acetate ions in the solution at constant pH 3. These authors found for the first time a discontinuous isotherm from cathodic step polarization experiments. It should be mentioned at this point that a real discontinuity should be found to be independent of either anodic or cathodic polarization. This means that anodic stripping should also reveal a discontinuity. However, despite numerous * efforts Bewick’s result could not be reproduced in this work or by other authors under comparable experimental conditions[9,13]. Other criteria to distinguish between simple adsorption and nucleation-growth models are commonly based on the characteristics of sweep voltammograms, ie the dependence of potential, current height, and symmetry of the voltammetric peaks on the sweep rate[ 15, 171. These models however can not be directly applied in the present system, because they neglect the existence of preformed species on the substrate surface. Moreover, a comparison between the shape of amperograms and potentiograms as recently suggested by Sadkowski[16] appears also of limited applicability. According to Pangarov[ 181 the shift of the voltammetric peaks in a cathodic direction demonstrates the influence of specific adsorption of foreign species. The dependence of the peak potential in voltammograms run at a fixed very low potential sweep rate on prolonged holding times at potentials preceding the UPD of lead on Cu(ll1) in perchlorate containing solutions (Fig. 6) reveals also a pronounced effect due to foreign species on the substrate interfering the kinetics of the UPD process even in the absence of specifically adsorbed anions. Therefore we prefer to interpret the observed nucleation like behaviour in terms of a negative 2-D nucleation within a preformed surface layer consisting of oxygen containing species rather than by a spontaneous agglomeration of lead adatoms on the bare Cu(l11) substrate surface. Further, contributions due to specifically adsorbed anions like CH,COOand Cl- assist in obtaining sharper voltammetric peaks (Fig. 7) as well as abrupter changes in the charge isotherms (Fig. 8). This is also supported by the observed dependence of the UPD process on the type of specifically adsorbed anions and on the different polarization conditions. Severe retardation of the UPD process correlates with specific adsorption of anions in the following sequence Cl- > CH,COO> ClO,. In solution containing ClO,. a weakly or practically non-adsorbing anion, the peak potential shift in the course of the cathodic scan after prolonged polarization at potentials just preceeding the UPD

range of lead (Fig. 5) suggests the stabilization of a preformed oxygen containing layer which exhibits long-time range effects. According to this interpretation the observed critical threshold potential has to be related to an activation barrier for the disruption of the pre-existing surface film. The underpotential decomposition of water yielding CuOH species as a precursor layer which exhibits ageing effects has been recently proposed for the initial reaction of Cu in alkaline solutions[ 193. Finally, it should be pointed out that up to now there is no clear experimental evidence for a pure positive nucleation process in the wide field of UPD studies on metal deposition at foreign substrates. Acknowledgements-The authors are indebted to the “Arbeitsgemeinschaft Industrieller Forschungsvereinigungen” (AIF) and the “Fonds der Chemischen Industrie” for financial support. One of us (J. R. V) is grateful to the “Deutscher Akademischer Austauschdienst” (DAAD) for making this cooperation work possible. REFERENCES 1. W. J. Lorenz. H. D. Herrmann, H. Wiithrich and F. Hilbert, J. elecrrochem. Sot. 121, 1167 (1974). 2. D. M. Kolb, in Adoonces in Electrochemistry and Electrochemical Engineering, (Edited by H. Gerischer and C. W. Tobias) John Wiley, New York, Vol. 11, pp. 125-271 (1978). 3. K. Jiittner and W. J. Lorenz, 2. Phys. Chem. NF, 122,163 (1980). in Electrochemistry and 4. R. R. Adzic, in Advances Electrochemical Engineering, (Edited by H. Gerischer), John Wiley, New York, Vol. 13, 159-260 (1984). 5. K. Jiittner and H. Siegenthaler, Electrochim. Acta 23,971 (1978). 6. H. Siegenthaler, K. Jiittner, E. Schmidt and W. J. Lorenz, Electrochim. Acta 23, 1009 (1978). 7. H. Siegenthaler and K. Jiittner, Electrochim. Actn 24,109 (1979). 8. E. Schmidt and H. Siegenthaler, J. &ctroanal. Chem. 150, 59 (1983). 9. H. Siegenthaler and K. Jiittner, J. electroanal. Chem. 163, 327 (1984). 10. A. Bewick, J. Jovicevic and B. Thomas, in Faraday Symposia of rheChemical Sociery, London, Vol. 12,24-35 and 165167 (1977). 11. K. J. Bachmann and J. K. Dohrmann, J. efecaoanol Chem. 21, 311 (1969). 12. D. Dickertmann, F. D. Koppitz and J. W. Schultze, Electrochim. Acta 21, 916 (1976). 13. Ch. Lierse, Diplomarbeit, University of Karlsruhe (1979). 14. K. Jiittner, G. Staikov, W. J. Lorenz and E. Schmidt, J. electroanal. Chem. 80, 67 (1977). 15. H. Angerstein-Kozlowska, B. E. Conway and J. Klinger, J. elecrroanal. Chem. 75,45,61 (1977); 87 301,321 (1978). 16. A. Sadkowski, J. electroonal. Chem. 208, 69 (1986). 17. L. Bosco and S. K. Rangarajan, J. electroanal. Chem. 129, 25 (1981). 18. N. Pangarov, Electrochim. Acta 28, 763 (1983). 19. M. R. Gennero de Chialvo, J. 0. Zerbino, S. L. Marchiano and A. J. Arvia, J. appl. Electrochem. 16, 517 (1986).