Linear sweep voltammetry: a cheap and powerful technique for the identification of the silver...

ORIGINAL PAPER

Linear sweep voltammetry: a cheap and powerful techniquefor the identification of the silver tarnish layer constituents

S. Capelo & P. M. Homem & J. Cavalheiro &

I. T. E. Fonseca

Received: 6 July 2012 /Revised: 10 September 2012 /Accepted: 14 September 2012 /Published online: 28 September 2012# Springer-Verlag Berlin Heidelberg 2012

Abstract The potentialities of linear sweep voltammetry(LSV) for the characterization of the tarnish layer formed onpure and sterling silver samples exposed in the chapel and inthe museum, at the Cathedral of Porto, in Portugal, are welldemonstrated in this study. The technique allows the identifi-cation of the constituents of the thin tarnishing layers and alsoits relative abundance. A much more complex compositionthan the silver sulphide commonly associated with silvertarnish has been found, namely, silver chlorides, silver oxidesand minor amounts of silver sulphide on the pure silver, pluscopper oxides and a mixed copper–silver sulphide on thesterling silver samples. The tarnishing films were very thin

mainly composed by silver chloride and silver oxides layerswith estimated thicknesses ranging between 0.22 and 9.63 nm.

Keywords Silver . Chapel . Museum . Tarnishing layer .

LSV . Atmospheric pollutants

Introduction

Silver and silver alloys are applied in various fields, such asCultural Heritage. Objects made of silver or silver alloystarnish due to the interaction of the atmospheric pollutantswith the metallic surfaces. Atmospheric corrosion happens,leading to the tarnishing and/or to the unpleasant appear-ance, or even to the loss of important details of the objectsmade of silver or silver alloys. The intensity of the phenom-enon depends on the type of pollutants, on its levels and onthe exposure time. Pollutants such as, sulphur compounds,namely, sulphur dioxide (SO2), hydrogen sulphide (H2S)and carbonyl sulphide (OCS), other gases, such as carbondioxide (CO2), oxygen (O2), ozone (O3) and nitrogen dioxide(NO2), as well as other species, such as organic acids andaerosols particles (NaCl, (NH4)2SO4, Mg2+, Na+, Cl−, etc.)may interact with metals and alloys leading to the formation ofsuperficial films which can be homogeneous or heteroge-neous, thin or thick, porous or non-porous, conductors,semi-conductors or even insulators. The films can, in princi-ple, be composed by sulphides, sulphates, oxides/hydroxides,chlorides, nitrates, carbonates and organic salts [1–17]. Thedegree of humidity and the oxidant conditions of the atmo-sphere play also an important role in the corrosion process andon the nature of the corrosion products [10].

In order to avoid the tarnishing of silver and silver alloysobjects, even when exposed in a museum, inside showcases,gases that may accelerate the corrosion process should be ex-cluded and climatic conditions, such as temperature, humidity,

S. CapeloDPAO, Escola de Ciências e Tecnologia, Universidade de Èvora,Colégio Luís António Verney,Rua Romão Ramalho, 59,7000-671 Évora, Portugale-mail: [email protected]

S. Capelo : P. M. Homem : I. T. E. Fonseca (*)CCMM, DQB, Faculdade de Ciências, Universidade de Lisboa,Campo Grande, C8,1749-016 Lisbon, Portugale-mail: [email protected]

P. M. Homeme-mail: [email protected]

P. M. HomemCITCEM, DCTP, Faculdade de Letras, Universidade do Porto,Via Panorâmica s/n,4150-564 Porto, Portugal

J. CavalheiroDEMM, Faculdade de Engenharia, Universidade do Porto,R. Dr. Roberto Frias, s/n,4200-465 Porto, Portugale-mail: [email protected]

J Solid State Electrochem (2013) 17:223–234DOI 10.1007/s10008-012-1884-3

etc., must be well-established and maintained constant. Allshowcases construction and dressing materials shall be rou-tinely tested before being used. Other objects displayed withthe silver objects, in the same showcase, can also causetarnishing to the silver. Despite all the precautions theremaining gases can cause tarnishing on silver even whenpresent at very low levels, and a certain tarnish rate is ingeneral unavoidable. Slowing down the rate of silver tarnish-ing in museum context and at open spaces is highly desir-able, but for that purpose the use of surface sensitivetechniques is very important.

The problem of the influence of the atmospheric pollutantson the tarnishing rate and its mechanism are important tasksnot yet completely understood, in spite of the amount ofpublished data [1–17]. According to Graedel et al. [7], hydro-gen sulphide (H2S) and carbonyl sulphide (OCS) are theprincipal atmospheric aggressive gases for silver and copperleading to silver and copper sulphides. However, according toHallet et al. [12], the XPS technique has identified the pres-ence of several species other than silver sulphide. AlsoHomem et al. [13] have identified by SEM-EDS and XPS,on real samples taken from the Holy Altar of the Chapel ofPorto Cathedral (made of pure silver and of Cu-Ag alloys,with copper percentages varying between 2 % and 40 %)compounds as silver chloride and silver oxides, copper oxidesand copper sulphates, as well as particles of carbon, SiO2,Na2SO4.CaSO4, CaSO4.2H2O, Na2CO3, iron, lead and titani-um oxides. In that study [13], the authors have not foundevidence of silver and/or copper sulphides. In fact, in thescientific community still remains some controversy about thecomposition of the silver tarnishing layers.

Electrochemical techniques, such as linear sweep voltam-metry (LSV), cyclic voltammetry (CV) and square wavevoltammetry (SWV) have been recognized as important toolsfor the corrosion monitoring on the metallic objects exposedto polluted atmospheres [14, 18–33]. By reducing the prod-ucts formed on the metallic surface, it is possible to resolve thetotal layer into its major individual chemical constituents. Onthe other hand, the high sensitivity of technique allows meas-urements of extremely thin surface films. The potentialities ofthe electrochemical techniques can be greatly extended whenapplied to particles and droplets immobilized on carbon orother suitable electrode materials prepared and applied asdescribed by Scholz et al. [28–33].

Doménech et al. [14, 22–25] report the potentialities ofthe technique developed by Scholz and collaborators,named as voltammetry of microparticles (VMP), applied inthe fields of archaeometry, conservation and restoration.They have recently applied the technique on the identifica-tion of the corrosion products on metallic and other objectsexposed over the years in different media (air, soil, seawater,etc.). The identification being performed by comparingnumber, position and shape of the voltammetric peaks with

those from reference compounds. This is particularly impor-tant when the corrosion products are powder mixtures whichcan be easily removed from the real samples. Eletrochemicalmethods applied to the impregnated microparticulate mate-rial are, as reported and demonstrated by Doménech et al.[14, 22–25], a powerful technique for the identification ofminor amounts of corrosion products formed on artworks.

In connection with our work, as the films are so thin (orderof few nanometers), uniform and highly adherent, the applica-tion of the VPM technique could be important when applied toimmobilizied micro or nano particles of synthetized referencecompounds incorporated on suitable electrode surfaces or onfurther studies dealing with very small amounts of corrosionproducts removed from the real object, the Holy Altar of theChapel of Porto Cathedral, dated from the seventeenth toeighteenth centuries and made of silver and Ag-Cu alloys.

Previous studies on the electrochemical behaviour of puresilver and copper in aqueous solutions contain also useful datafor the present study [34–41]. Also the work by Langh et al.[42] on the delamination of silver sulphide layers can be animportant piece of work in the context of our study, as well asthe voltammetric analysis of artificially tarnished silver cou-pons described by Ankersmith et al. [43] and even the recentwork byDoménech et al. [14] performed on the analysis of thecorrosion products in highly corroded silver coins.

Concerning the pollution levels, de la Fuente et al. [44]have recently published a study on the effects of the airpollution on the cultural heritage. They report levels of SO2

and NO2 at Madrid, ranging between 6 and 60 and 36 and86 μgcm−3, respectively. Also Camuffo et al. [45] during acampaign in 1996 have obtained in winter concentrations ofNO2 and SO2 varying from 8 to 24 and 3 to 40 ppb, respec-tively. Values of H2S concentrations in museums in the pptrange have been reported byCamuffo et al. at 95–110 ppt [45],by Ankersmit et al. at 50–100 ppt [46] and by Graedel at300 ppt [9], among others. Considering that the lowest ob-served adverse effect dose (LOAED) for the production ofAg2S is 2.5 ppb/year [4], the probability of having Ag2S onthe silver exposed in museums is quite low or even null. Evenoutside a museum in different cities around the world, theaverage concentrations of H2S are fortunately too small.

This paper reports a contribution for a project which aimsto understand the tarnishing process of the Holy Altar of thechapel of Porto Cathedral, a treasure of the PortugueseCultural Heritage, made of silver and silver alloys, as de-scribed in previous studies [8, 13]. The authors have ac-quired in the market silver and sterling silver pieces, of highpurity, from which the coupons used in the present studywere prepared and then submitted to the chapel environ-ment, being placed in selected points near the Holy Altarand in the museum inside showcases (not sealed).

The Porto Cathedral is situated on the top of a hill,looking down to the Atlantic Ocean, and is one of the most

224 J Solid State Electrochem (2013) 17:223–234

visited monuments in the north of Portugal, still open tocatholic worship and frequent religious ceremonies attendedby many people. Moreover doors are open the whole year andthe number of visitants is quite high (approximately 5,000/yearin the museum and 40,000/year in the chapel), particularlyduring autumn, spring and winter (Christmas time). The out-door atmosphere of Porto is a typical urban, industrial andmaritime atmosphere, where aerosols transporting particlesof matter, such as chlorides, may play its role on the attackto the metallic samples. The relative humidity (RH) andthe ozone and UV radiation levels are also importantparameters enabling the corrosion process. In a dry atmo-sphere, corrosion will not take place; corrosion requireshumidity and its rate varies with the degree of humidity.

SO2 has not been measured in the chapel; measurementshave been carried out during a 1-year period at an atmo-spheric station, Sobreiras-Lordelo do Ouro, located in thesurroundings of the cathedral. The average concentration ofSO2 has shown a value of 3.3 ppb; this value is in the rangeof those reported by Camuffo et al. [44] for Correr museumin Venice (5.3 ppb for indoor). The interaction of SO2 withsilver in an atmosphere with a certain degree of humiditymay lead to silver sulphates or silver sulphides.

Measurements of Cl− ions have not been performed atPorto air quality stations. However, at Viana do Castelo, atown near Porto, the Instituto of Meteorologia has performedmeasurements from 1996 to 2009 and the average values ofCl− were between 4 and 11 ppm.

This paper presents a contribution of the electrochemicaltechniques for the identification of the main compounds of thetarnishing layers formed on pure and sterling silver samplesexposed in the chapel and in the museum of Porto Cathedral,during the 1-year period, from March 2010 till March 2011.Control as well as blank samples have also been producedand used as references. The technique has also allowedthe evaluation of the relative composition of the variouscompounds, as a function of exposure site and time.The higher reactivity induced by the relative small per-centage of copper on the sterling silver in comparisonwith the pure silver samples, all submitted to the sameatmospheric conditions, have also been well evidencedon the corresponding electrochemical recordings(voltammograms). The charges under the cathodic voltamm-teric peaks have been used to estimate the film thicknesses.

Experimental

Two working electrodes were made from polished coupons(1×1×0.1 cm) of pure silver (999 ‰ Ag) and of sterlingsilver (925 ‰ Ag+ 75 ‰ Cu). Both materials were suppliedby the national market (Jewelery Paçó, Torre de Moncorvo,Portugal).

Each coupon was soldered to a copper wire and theninserted into a glass tube and sealed with epoxy resin.Before experiments, the electrodes were polished with0.3–0.05 μm alumina paste to obtain a mirror like surface.The electrodes were then rinsed by doubly distilled waterand inserted into the electrochemical cell. The electrochem-ical cell was a two-compartment cell filled with about50 mL of the supporting electrolyte solution. The supportingelectrolyte solution (0.1 M of sodium citrate dehydrated)was prepared with C6H5Na3O7.2H2O, pro analysis grade99.5 %, and Milli-Q water from Millipore. The pH of thesolution was 8.2.

Coupons of pure silver and of sterling silver covered withfilms of sulphides produced under controlled laboratorialconditions (closed atmosphere of H2S 1,000 ppm, RH0

53 %, T021 °C, over periods of 24, 48 and 72 h) weremounted as described above. The silver electrodes coveredwith the sulphide films are designated by PS24, PS48 andPS72 and the corresponding electrodes made with the silveralloy as LS24, LS48 and LS72. Two electrodes made fromcoupons of pure silver and silver alloy covered with chloridefilms grown during 5 h, in a controlled and closed atmo-sphere (3.5 % of NaCl, RH041 %, T043 °C), have alsobeen mounted, as previously described, and then used asworking electrodes (PCl5 and LCl5).

Finally a set of working electrodes was made from thesilver 999 ‰ (P) and silver alloy 925 ‰ Ag (L) couponsexposed inside showcases, in the Treasure Room of themuseum and near the Altar of the Holy Sacrament of thechapel, both of Porto’ Cathedral. These two sets of couponshave been exposed during periods of 4 and 52 weeks withexposures of 4 weeks starting at the beginning of eachseason. The coupons were exposed on acrylic supports withan inclination of 45 °C.

The identification of the samples exposed in the museumand in the chapel, which will be the subject of the presentstudy, is given in Table 1.

Electrochemical measurements were made using anEcochimie Autolab PGSTAT 12 potentiostat, controlledby PC using the GPES software version 4.9. The refer-ence electrode was a commercial mercury sulphate elec-trode, saturated with K2SO4 (SME) with an E00+0.64 Vvs. SHE. The counter electrode was a platinum wire andthe working electrodes were the samples previously identified(see Table 1).

Experiments were performed at room temperature (22 °C)in the electrolyte solution de-oxygenated by bubbling nitrogenduring 15 min before measurements. The open circuit poten-tial between the working electrode and the reference electrodewas measured before each experiment. In order to convertcurrent intensity into current density, the area of all the work-ing electrodes was accurately determined and charges weremeasured by integration between the curve and the baseline

J Solid State Electrochem (2013) 17:223–234 225

with the Origin software. Cyclic voltammetry (CV) and linearsweep voltammetry (LSV)were applied to blank (bare) electro-des of pure and sterling silver (PB and LB electrodes); refer-ence electrodes covered with sulphide films (PS24, PS48,PS72, LS24, LS48 and LS72); reference electrodes coveredwith chloride films (PCl5 and LCl5) and finally to the electro-des from natural exposures, in the museum (MP4Sp, ML4Sp,MP4S, ML4S, MP4A, ML4A,MP4W,ML4W,MP52, ML52)and in the chapel (CP4Sp, CL4Sp,CP4S, CL4S, CP4A, CL4A,CP4W, CL4W, CP52, CL52) of Porto Cathedral.

The quantification of the corrosion products was obtainedfrom the charge involved in the corresponding electrochem-ical reduction process, which is determined from the areaunder the cathodic voltammetric peaks. This methodologyhas been often reported, among others, by Scholz et al.[47–50]. The mass of the corrosion products is related withthe amount of charge through the Faraday’s law:

m ¼ M

zFQ

where, m (g cm−2) represents the amount of mass, M is themolar weight (g mol−1), z is the number of electrons involvedin the redox process, F is the Faraday constant (Cmol−1) andQ(C cm−2) is the charge under the corresponding cathodic peak.

The thickness of the film was estimated through theformulae:

l ¼ m

ρ

With ρ being the density and l the thickness (ρ(AgCl)05.56 g cm−3; ρ(Ag2O)07.14 g cm−3 and ρ(Ag2S)07.23g cm−3). Non-porous and non-hydrated layers covering the

whole geometric area of the working electrodes have beenassumed. The well-defined cathodic peaks give us evidenceabout the layer structure. Also in the literature, there arestudies, such as those by Doménech et al. [14, 23, 24],reporting data, even from shiny and lustrous surfaces, show-ing a bilayer structure composed by an outer and an innerlayer.

Results and discussion

Visual observations

The colour exhibited by samples exposed during 1 month inthe chapel and the museum varied from white to slightlybrownish. The samples exposed during 1 year present auniform film, still very thin, quite adherent to the surface.The morphology of the tarnishing layers on the varioussamples is being observed by surface analysis techniques,namely by scanning electron microscopy coupled with X-ray energy dispersive analysis (SEM-EDX) and X-ray pho-toelectron spectroscopy (XPS). These and other spectro-scopic studies are in progress and will be the subject of afuture publication.

Voltammetric studies

Cyclic voltammetric responses from the blank samples

The cyclic voltammograms (CVs) of the two working elec-trodes of pure (999‰ Ag) and sterling silver (925 ‰ Ag)immersed in the 0.1 M sodium citrate solution are presentedin Fig. 1a, while Fig. 1b shows the LSVs of the sameelectrodes with the potential being scanned, at 10 mVs−1,in the cathodic direction, from 0 to −1.8 V.

The cyclic voltammogram of the pure silver electrodeagrees well with the one reported by Zaky et al. [34, 35]who have studied the electrochemical behaviour of silver inborate buffer solution. Two anodic peaks, A2 and A4 at −1.1and −0.88 V, respectively, and a broad peak, A4a, precedingthe passive region can be observed. During the cathodicscan, two well-defined peaks, C4 and C2, at −1.04 and−1.49 V vs. SME are detected. However, it should be notedthat the most relevant and well-defined peaks is the pairA4/C4, for which ΔEp00.16 V. According to the literature [34,35], peak C4 is related with the reduction of the AgO to Ag2Oand peak C2 is due to the electro reduction of Ag2O to Ag. Avery small peak, C5, between −0.6 and −0.8 V is also observedin Fig. 1a. This peak may be related with soluble silver oxidesspecies.

In the case of the alloy, the corresponding CV is almostcoincident of that of pure silver with a well-defined pair ofpeaks, A4/C4, located at −0.88 and −1.1 V. However, peak

Table 1 Codes for the pure silver (P) and silver alloy (L) samples vs.the place (chapel, C, or museum, M), time (number of weeks); seasonin which the exposure was started (spring, Sp, summer, S, autumn, A,winter, W)

Code Exposition period Exposure time(weeks)

Cathedral Museum

CP4Sp MP4Sp 23.03.10–19.04.10 4CL4Sp ML4Sp (First month in spring)

CP4S MP4S 21.06.10–19.07.10 4CL4S ML4S (First month in summer)

CP4A MP4A 22.09.10–18.10.10 4CL4A ML4A (First month in autumn)

CP4W MP4W 21.12.10–18.01.11 4CL4W ML4W (First month in winter)

CP52 MP52 23.03.10–21.03.2011 52CL52 ML52 (spring + summer + autumn

+ winter)

Figures represent the number of weeks

C chapel, M museum, L alloy (925 ‰ of Ag), P pure silver (999 ‰)


C2 is not so well defined, being more likely a broad regionbetween −1.30 and −1.50 V, probably as a result of thereduction of copper and silver oxides (CuO2 and Ag2O).The observed small differences are certainly related with therelative small percentage of copper in the silver alloy. How-ever, it is worth to mention that the reduction charges on theLSV of the silver alloy are about twice higher than thecorresponding ones of pure silver (4.88 vs. 2.01 mC cm−2).

Concerning the silver–copper binary alloys there are notmany studies published in the literature. Assaf et al. [36] havestudied the electrochemical behaviour of two binary alloys inNaOH solutions. However, the two alloys have the followingcomposition: 20% Cu+80%Ag and 80% Cu+20%Ag. Forthose alloys, they have concluded that below a critical poten-tial copper dissolves preferentially and the three pairs of peakscorresponding to the formation and reduction of Cu2O, Cu(OH)2 and CuO could be observed, while beyond a criticalpotential Ag dissolves and the CV shows the three peakscorresponding to AgO−, Ag2O and AgO. They have observedthat in the second region both components may dissolvesimultaneously and the corresponding CVs may show sixanodic and six cathodic peaks. However, the silver alloy usedin our study contains only a relative small percentage ofcopper (less than 8 %), then one shall not expect a behaviourmuch different from that of pure silver, as shown in Fig. 1a.The cathodic voltammogram (Fig. 1b) shows for the silverelectrode two peaks: C4 and C2 at −1.06 and −1.48 V, respec-tively and a shoulder at ca. −0.90 V. In the case of the alloy the

pattern is quite similar presenting peaks C4 and C2 at slightdifferent potential values: −1.05 V, attributed to the reductionof AgO to Ag2O and −1.46 V, ascribed to the reduction ofCu2O to Cu. The peak at −0.90 V may be related with thereduction of soluble silver (II) oxide species.

Hazzazi et al. [38] have studied the passivation/depassi-vation of two Cu-Ag alloys (containing 20 % and 80 % ofcopper) in borate buffer (pH08.4). The referred studypresents also CVs of pure Ag electrodes showing threeanodic peaks that were attributed to the formation of Ag2Omultilayers, a peak related to the oxidation of Ag2O to AgOand prior to the O2 evolution reaction they have foundanother anodic peak which they have ascribed to the oxida-tion of AgO to Ag2O3. During the cathodic scan they havefound three cathodic peaks related to the reduction of Ag2O3

to AgO, AgO to Ag2O and Ag2O to Ag, respectively.

LSVs of silver and silver alloy electrodes covered with filmsof sulphides

Figure 2 presents the first scans of the cathodic reduction ofthe sulphide films, grown under controlled conditions (H2S1,000 ppm, RH053 % and T021 °C), during periods of 24,48 and 72 h on the pure silver and on the silver alloysamples. The LSVs of the blank electrodes are also given.It should be mentioned that the atmosphere contains a huge

Fig. 1 a CVs. b LSVs of the following electrodes: silver, PB, andsilver alloy, LB, electrodes, both immersed in a 0.1 M sodium citratesolution. ν010 mVs−1

Fig. 2 LSVs of the reduction of the layers formed on the followingelectrodes: a pure silver: PB, PS24, PS48 and PS72; b silver alloy: LB,LS24, LS48 and LS72. All the electrodes immersed in a 0.1 M sodiumcitrate solution. ν010 mVs−1


amount of H2S, thus the films formed under those specialconditions are certainly quite thick films and the blackcolour characteristic of the silver sulphides is well visible.

The cathodic voltammograms of the silver electrodes cov-ered with the silver sulphide film, grown at about roomtemperature under low relative humidity atmosphere (53 %),show only one well-defined stripping peak, C3, located at−1.26V, while the voltammogram of the blank silver electrodeshows two small cathodic peaks C4 and C2, located at −1.06and −1.48 V, respectively. The observed peak at −1.26 V, C3,agrees well with that reported by Bernard et al. [27] and alsoby Costa and Dubus [20], which they attributed to the silversulphide. The linear sweep voltammograms of the three silverelectrodes covered with the silver sulphide films (PS24, PS48and PS72) in the case of pure silver show a very well-definedstripping peak, located at −1.26 V. The height of this peakincreases linearly with the growth time of the sulphide film, ata rate of 40 μAcm−2h−1. However, it should be emphasizedthat 1,000 ppm of H2S is a huge concentration, therefore theintensities of the sulphide peaks are expected to be quite high( jp

C3 ranging from 0.7 to 1.7 mAcm−2 and the correspondingcharges varying from 14.3 to 27.2 mCcm−2) leading to a filmwith thickness varying from 25 to 48 nm.

In the case of the sterling silver (Fig. 2b), the pattern ofthe LSV shows a broad cathodic peak, C3, at ≈−1.26 V(Ag2S), and a smaller peak, C1, at ca −1.62 V, whichaccording to Costa and Dubus [20] can be ascribed to thereduction of a mixed copper-silver sulphide. The amount ofcopper in the alloy is relative low, therefore the lowerintensity of peak C1 is to be expected. The cathodic chargeassociated to peak C3 on the voltammogram of the LS24electrode is about 232 mC cm−2, which corresponds to afilm thickness of 412 nm. When compared with the chargecorresponding to the pure silver electrode (PS24) is aboutten times higher. Since both films have grown under thesame experimental conditions then we may conclude thatthe high increase in the reactivity is induced by the relativesmall percentage of copper.

According to Tran et al. [10] and also to Graedel [9] theproducts formed on silver and on copper due to the sulphi-dation by H2S are Ag2S and Cu2S following the reactions:

2Ag þ S2� ! Ag2S þ 2e� ð1Þ

2Cuþ S2� ! Cu2S þ 2e� ð2ÞHowever, when the amounts of H2S are small then in the

presence of a certain degree of humidity and levels ofoxidants the formation of silver oxides (Ag2O, AgO or evenAg2O3) may occur according to the following reactions:

2Ag þ 2OH� ! Ag2Oþ H2Oþ 2e� ð3Þ

Ag2Oþ 2OH� ! 2AgOþ H2Oþ 2e� ð4Þ

2AgOþ 2OH� ! Ag2O3 þ H2Oþ 2e� ð5ÞThe formation of silver sulphate and silver chloride may

also happen in the presence of SO2 and of Cl− ions, respec-tively, in humid and oxidant atmospheres, following thereaction:

SO42� þ 2Ag ! Ag2SO4 þ 2e� ð6Þ

Ag þ Cl� ! AgCl þ e� ð7ÞTran et al. [10] have obtained CVs of copper showing

four peaks with two prominent peaks, located at −0.2/−0.3and −1.3/−1.4 V vs. SCE (−0.6/−0.7 and −1.7/−1.8 V vs.SME). They have attributed those peaks to the reduction ofCuO and Cu2S, respectively. They have also concluded thatunder a certain degree of humidity, Cu2S is the main com-ponent of the corrosion layer on copper. Tacconi et al. [41]have studied the early stages of copper sulphide film growthat quite negative potentials. Cyclic voltammetry combinedwith electrochemical quartz crystal microgravimetry(EQCM) showed the formation of initial Cu2S (chalcocite)phase and further oxidation leading to a non-stoichiometricover layers of CuS (covelite), for potentials ranging between−1.0 and −1.3 V vs. Ag/AgCl (−1.4 and −1.7 V vs. SME).According to Tacconi et al. [41], the Cu2S film growth canbe studied without the interference from oxides formationonly by carefully restricting the range of potential valuesnegative to −0.5 V vs. Ag/AgCl (−0.9 V vs. SME). Forprolonged atmospheric exposures of copper, Tran et al. [10]report the formation of copper sulphides and also copperoxides. From the thermodynamic point of view, Cu2O ismore stable than Cu2S, since ΔG°(25 °C) (Cu2O)0−126 andΔG°(25 °C) (Cu2S)0−80 kJmol−1.

LSVs of silver and sterling silver electrodes coveredwith films of silver chloride

As it is well-known sodium chloride is a major componentof human perspiration. When perspiration or other residueare left on silver or silver alloys, after handling, the sodiumchloride reacts with the silver to form silver chloride, whichis highly sensitive and can darken over time [39]. Figure 3shows LSVs of the pure and sterling silver electrodes cov-ered with films of chlorides, grown during 5 h, in a closedatmosphere (3.5 % of NaCl, RH041 % and T043 °C). TheLSVs of the blanks are also given.

Birss and Smith [39], in a study on the formation andreduction of the AgCl films on silver in aqueous chloridesolutions, have found a cathodic pattern with two peaks quite


similar to those observed in our study: a small peak precedingthe peak of high intensity. They have attributed the small peak(shoulder) at more positive potentials to the first stage of thedeposition of a three-dimensional AgCl nuclei and the moreintense peak to the growth of AgCl multi-layers. Silver chlor-ides are light coloured when formed, but soon became dark byphoto-oxidation; its solubility in water is 5.2 mg/L.

In Fig. 3a for the PCl5 electrode, only the cathodic peak, C6,is observed at −0.21 V, necessarily due to the reduction of thesilver chloride. On the other hand, for the alloy, Fig. 3b, twopeaks are observed on the LSV of the LCl5 electrode: C6 at−0.20 and a broad peak C4 between −0.9 and −1.5 V. Peak C6

is due to the reduction of the silver chloride, with charges of theorder of 0.89 and 0.14 mC cm−2 for the pure and sterling silver,respectively. The broad peak has to be due to the reduction ofthe silver and/or the copper oxides. Again the higher reactivityof the alloy is well denoted by the high charge of the broadpeak, between −0.9 and −1.5 V (5.1 mC cm−2), while in thecase of pure silver those peaks are not shown.

LSVs of silver and sterling silver electrodes from exposuresin the chapel and in the museum

LSVs of silver and sterling silver samples exposed in chapeland in the museum

Figure 4 gives the LSVof the pure silver samples exposed inthe chapel for a period of 4 weeks (1 month) with the

exposure starting in spring time. For the assignment of thepeaks on the LSV of the unknown films, the LSVs of theblank and of the reference electrodes covered with layers ofsilver sulphide and silver chloride are superimposed.

The LSVof the pure silver samples exposed in the chapel(CP4Sp) show two peaks, C6 at −0.23 and another one at−1.10 V, located between peaks C4 and C3. Both peakspresent about the same intensity, of the order of 25 μAcm−2 (Qp

C600.56 and QpC3-C400.74 mC cm−2). It could

be argued that the more cathodic peak of the CP4Sp samplemay correspond to Ag2S, but it can also be due to the silveroxide since this peak can be displaced in the negativedirection as peak C6 is. In fact, Graedel et al. [7] hasidentified oxides as the constituents that form on silverduring the tarnishing process, under atmospheric exposures,as well as acanthine (Ag2S) and chorargyrite (AgCl). Silveroxides have also been identified by Graedel et al. [7], byHomem et al. [13] and by Chen et al. [16], among manyothers. Nevertheless, it has been extensively reported in theliterature [1, 3–7] that the primary atmospheric agents re-sponsible for the degradation of the metallic materials areH2S, COS, particulate chloride and possibly, HCl, all actingin the presence of moderate to high humidity. Those whohave identified Ag2S on the silver tarnish layer were forexposures under high aggressive conditions (higher concen-trations of H2S, long exposure times, high degree of humid-ity, etc.). The amounts of H2S present in the atmosphere ofthe chapel of Porto cathedral during the spring time werecertainly not enough to produce Ag2S during a short periodof time (1 month).

Figure 5 gives the LSVs of the pure and sterling silversamples exposed for a period of 4 weeks (1 month) in thechapel and in the museum, with the exposures started in thespring time (Sp).

Pure silver samples exposed in the chapel showed thepeaks ascribed to silver chloride and to silver oxide exhibit-ing about the same intensity, around 35 μAcm−2, while thevoltammogram of the silver alloy presents three well-defined peaks: C6, C4 and C3 at −0.23, −0.97 and −1.26 V,

Fig. 3 LSVs of the reduction of the films on the following electrodes:a PB and PCl5; b LB and LCl5 electrodes immersed in a 0.1 M sodiumcitrate solution. ν010 mVs−1

Fig. 4 LSV of the following electrodes: PB, PS24, PCl5 and CP4Sp.All the electrodes immersed in a 0.1 M sodium citrate solution.ν010 mVs−1


respectively. The high reactivity of the silver alloy is againwell evident, as well as the appearance of the peak assignedto the sulphides. The total reduction charge is considerablyhigher: 10.18 mC cm−2 (silver alloy) vs. 1.70 mC cm−2

(pure silver). The peak assigned to the reduction of silversulphide (C3) is well defined and presents quite high chargein the case of the alloy.

For the 1-month exposures in the museum, the patternsof the LSVs of both samples are quite similar to the onespresented in the chapel (the same number of peaks) butwith much lower intensities particularly in the case of thepure silver. For pure silver the charge of peaks C6 variesfrom 0.30 (museum) to 0.56 mC cm−2 (chapel), while thatof peak C4 varies from 0.57 to 0.74 mC cm−2. In the caseof the alloy peak C6 varies from 0.030 (museum) to0.47 mC cm−2 (chapel), peak C4 from 3.83 to 4.46 mCcm−2 and peak C3 from 2.00 to 4.75 mC cm−2. The dataindicate that the amounts of chlorides, during that expo-sure period, have been much higher in the chapel than inthe museum.

LSV of the silver samples exposed in the chapeland in the museum as a function of time

Figure 6 presents the LSVs showing the effect of the expo-sure time on the silver coupons exposed in the chapel and inthe museum (4 and 52 weeks). The LSV of the CP4Spsample (Fig. 6a) corresponding to 1 month exposure ofsilver in the chapel shows two small but well-defined peaks,C6 and C4 at −0.27 and −1.12 V, respectively. After 1 year,those peaks are seen to increase in intensity and one addi-tional small peak C6a at around −0.24 preceding the bigpeak C6 is also depicted. The LSVs corresponding to thesilver samples exposed in the museum shows two smallpeaks C6 and C4 and at −0.22 and −1.1 V, while the onecorresponding to 12 months exposure shows an enhance-ment of the both peaks. Peak C4 being succeeded by a smallpeak C3 at −1.20 V, probably related with the formation ofthe silver sulphide.

The total amount of chlorides deposited from the aerosolsof the atmosphere during a 1-year period is certainly higherwhen compared with the period of 4 weeks. For thesamples exposed in the museum, during 4 weeks, whencompared with those exposed in the chapel the chargesinvolved in the AgCl formation are lower in the museum( jp

C6≈10 μAcm−2 vs. ≈30 μAcm−2). The same behaviouris observed for the peak assigned to silver oxide (jp

C4≈20 μAcm−2 vs. ≈ 30 μA cm−2).

The charges associated with the reduction of silverchloride from exposures of pure silver in the chapel variedfrom 0.56 to 3.61 mC cm−2, during the 1 year exposure,while for peak C4 related with the oxides the valuesvaried from 0.74 to 1.04 mC cm−2. Concerning the sam-ples in the museum exposed during 1 year, the charges ofthe chloride peak varied from 0.30 to 0.55 mC cm−2 andthe ones of the peaks related with the oxides from 0.57 to1.38 mC cm−2. It should be mentioned that the peakattributed to the silver sulphide starts to be seen on theLSV of the MP52 sample (Qp

C300.59 mC cm−2).

Seasonal effects on the silver samples exposed in the chapeland in the museum

The voltammograms of the reduction of the films, due to the1-month exposures of the silver in the chapel and in themuseum, with the exposures that started at the beginning ofthe four seasons are given in Fig. 7.

Fig. 5 LSVs of the reduction of the films on the following electrodes:CP4Sp, MP4Sp, CL4Sp and ML4Sp. ν010 mVs−1

Fig. 6 LSVs of the following electrodes: a CP4Sp and CP52; bMP4Sp and MP52. All the electrodes immersed in a 0.1 M sodiumcitrate solution. ν010 mVs−1


The LSVs corresponding to the pure silver samples exposedin the chapel (Fig. 7a) show peaks attributed to the silveroxides (C4 and C5) and silver chloride (C6); its definition andintensities are influenced by the season in which the exposurewas initiated. For the winter exposures, the peak of silverchloride is better defined, presenting the shape of a strippingpeak, while peak C4 almost disappears and a broad peak C5

probably also related with the reduction of silver oxides speciesis observed between −0.5 and −0.9 V. For the summer expo-sures the pattern of the LSV is quite complex, peaks C6a andC6, followed by a shoulder at ≈−0.4 V, and by another onebetween −0.5 and −0.9 V and a well-defined peak C4. Thespring exposure shows a behaviour quite similar to that ofsummer, but the peak C6 is not split. The LSVof the autumnexposure shows a quite round, small and complex peak C6,followed by a doublet of peaks in the region of peak C4.

The total charge associated to the reduction of the filmformed on pure silver in the chapel is lower for the autumnexposures. The charges associated with peaks C6 from thepure silver samples exposed in the chapel show valuesbetween 0.54 and 0.89 mC cm−2 for the autumn and winterexposures, respectively. It is clearly shown that the higheramounts of silver chloride were produced on the sampleswith exposures started in winter. The total reduction chargevaries from 1.33 to 2.62 mC cm−2.

The voltammograms of the silver samples exposed in themuseum (Fig. 7b) show the silver oxides peaks slight higherthan that of the silver chloride, but both notable big in thewinter exposures. The intensity of the silver chloride andsilver oxides peaks follows the order: jp

C6(W)> jpC6(Sp)>

jpC6(A)> jp

C6(S)≈0; jpC4 (W) >> jpC4 (S)>jp

C4(A)> jpC4(Sp),

respectively. On the other hand, the peak attributed to silversulphide is not observed. Our data are in agreement withthose who argue based on thermodynamic values that in thepresence of chlorides the formation of silver chloride ismore probable than the silver sulphide. The amount ofcharges related with the silver chloride film on pure silverin the museum is 0.08 in summer and 0.56 mC cm−2 inwinter and the corresponding to silver oxide reduction var-ied from 0.57 (spring) to 1.64 mC cm−2 in winter exposures.

Removal of films by continuous linear sweeps

Figure 8 shows the effects of the successive scans, in thecathodic direction, revealing the process of the removal ofthe films produced on the silver coupon exposed in thechapel. From data of Fig. 8, it can be concluded that duringthe first scan the peak attributed to silver chloride disap-pears, while the one related to silver oxide remains beingdisplaced to more positive potentials. This effect may berelated to the high oxidation currents observed at the anodicpotential limit when the second and third scans are initiated.Data indicate also that AgCl is quite easily reduced, whilethe silver oxides are more stable, not being reduced underthe established experimental conditions (solubility in water:(Ag2O)025 μg/L and (AgCl)05.2 mg/L). A small peak, inthe region of the silver sulphide, is observed on the LSVscorresponding to the 2nd and 3rd scans. It seems that thedisappearance of the AgCl peak allows the visibility of thesmall reduction peak which might be associated with thereduction of the silver sulphide (≈−1.3 V). Similar behav-iour has been reported by Domenéch et al. [14] on a studydealing with the identification of corrosion products on acollection of archaeological copper–silver coins.

Fig. 7 LSVs of the cathodic reduction of the films due to 1 monthexposures of pure silver in the chapel and in the museum with expo-sures started at the beginning of the four seasons: a silver in the chapel:CP4Sp, CP4S, CP4A and CP4W; b silver in the museum: MP4Sp,MP4S, MP4A and MP4W. All the electrodes immersed in a 0.1 Msodium citrate solution. ν010 mVs−1

Fig. 8 Consecutive LSVs corresponding to the reduction of the filmdue to 1 month exposure of silver in the chapel: the CP4Sp electrodeimmersed in a 0.1 M sodium citrate solution. ν010 mVs−1. First (solidline), second (broken line) and third (dotted line) scans


Analysis of the reduction charges and film thicknessas a function of exposure conditions

The charges of the various cathodic peaks observed on theLSVs presented in the preceding paragraphs allow the esti-mation of the film thickness on the silver samples exposedunder the various experimental conditions. Figure 9 givesthe graphic of the reduction charges of the peaks associatedto the reduction of the various constituents of the tarnishingfilms, as well as the total reduction charges measured be-tween −0.1 and −1.2 V.

The estimated thickness of the chloride films formed onpure silver samples exposed during 1 month in the chapelvaried from 1.43 nm, on exposures started in autumn, to2.38 nm, on those started in winter, while the corresponding

ones in the museum varied from 0.22 nm (summer) to1.48 nm (winter), while the thickness of the oxides filmson the samples exposed during 1 month in the chapel variedfrom 0.96 nm on the exposures started in the autumn to1.34 nm on those corresponding to summer and winter,while the corresponding values on the samples exposed inthe museum were 0.95 (spring) to 2.76 nm in the winterexposures.

Concerning to the influence of the exposure time, thethickness of chloride films formed on the silver samplesexposed in the chapel vary from 1.49 (4 weeks start inspring) to 9.63 nm (52 weeks start in spring) and in themuseum from 0.81 to 1.47 nm. On the silver in the chapelthe silver chloride predominates being particularly high forthe 52 weeks exposure, while silver oxides are slightlylower and particularly lower on the CP52 sample. The silversulphide was formed in a measurable quantity, but stillsmall, only on the MP52 sample. This is probably accept-able since the levels of the H2S concentration during theother exposures were too small that the LOAED of the silversulphide was not reached.

Accordingly to Franey and Graedel [17], the corrosionfilms that form on metals and alloys during field exposuresare generally chemical mixtures. On the other hand, Chen etal. [16] have concluded about the formation of silver oxideforms, which are stable during exposure, and admitted thepossibility of the formation of silver sulphate in field expo-sures. They have also concluded that in the presence of UVradiation and ozone in the atmosphere silver oxides may bethe main corrosion products formed on silver samples.Graedel et al. [7] also state that in atmospheres contaminatedwith chloride ions, AgCl may be produced instead of Ag2S.Also Vieira and Piedade [51] have concluded, based onthermodynamic data, about the higher probability of havingsilver chloride, since ΔG°(AgCl)0−109.8 kJmol−1 and ΔG°(Ag2S)0−40.67 kJmol−1. Also according to Thickett et al.[52] the percentage of silver sulphide can be minimum incomparison with the other compounds, therefore of verydifficult identification.

Conclusions

The influence of the atmosphere (chapel and museum),exposure time (1 and 12 months) and season in which theexposures were started (spring, summer, autumn and winter)has been observed and analysed on the LSVs of the electro-des made from the coupons of pure and sterling silversamples exposed to the various experimental conditions.

For the 1-month exposures that started in spring, thepeaks corresponding to the reduction of the silver oxidesand silver chloride were, as expected, higher on the samplesexposed in the chapel, while for the 12-month exposures

Fig. 9 a Charges under the peaks associated to the reduction of silverchloride (peak C6), silver oxides (peak C4), silver sulphide (peak C3)and total reduction charges (−0.1–1.2 V). b Estimated thickness of thefilms. All for the pure silver samples exposed under the programmedexperimental conditions


that started in spring, peaks have increased in size with thepeak assigned to chloride being much higher.

The seasonal influence, even for the 1-month exposures,was particularly notable on the exposures started in winter,revealing pronounced effects on the chloride peak.

Concerning the corrosion products on the pure silver,it was possible to postulate the existence of silver com-pounds, as silver chlorides, silver oxides and very smallamounts of silver sulphide. On the sterling silver sam-ples, silver chlorides, silver and copper oxides and amixed copper–silver sulphide were admitted as possibleconstituents of the tarnish layer. Whenever identified,the amounts of the silver sulphide were quite small;however, it shall be emphasized that none of the analysedsamples presented the black colour characteristic of thesulphides.

The films were very thin (0.22 and 9.63 nm), but thiswas not surprising since the samples have been exposedon relative non-polluted atmosphere during a short pe-riod (maximum 1 year). A layer-by-layer film morphol-ogy has been assumed, since the cathodic peaks werewell-defined stripping peaks. Besides the literature onsimilar studies reports surface analysis data confirming alayer-by-layer morphology. However, it should be men-tioned that such studies deal, in general, with thickerfilms.

Acknowledgments The authors would like to acknowledge the fi-nancial support from FCT (Fundação para a Ciência e Tecnologia) toCCMM (Centro de Ciências Moleculares e Materiais). S Capeloacknowledges the University of Évora for her sabbatical leave. Allthe authors also thank Doctor M. R. Nunes for the laborious anddelicate work in the preparation of all the working electrodes for theelectrochemical studies.

References

1. Leygraf C, Graedel T (2000) Atmospheric Corrosion. Wiley-Interscience, John Wiley and Sons Inc, New York

2. Watanabe M, Hokazono A, Handa T, Ichino T, Kuwaki N (2006)Corros Sci 48:3759–3766

3. de Mele MFL, Duffó G (2002) J Appl Electrochem 32:157–1644. Ankersmith HD, Tennent NH, Watts SF (2005) Atmos Environ

39:695–7075. Baer NS, Banks PN (1985) Int J Mus Manag Curatorsh 4:9–206. Arroyave C, Morcillo M (1995) Corros Sci 37:293–3057. Graedel TE, Franey JP, Gualtieri JG, Kammlott GW, Malm DL

(1985) Corros Sci 25:1163–11808. Homem PCMD (2006) Contributo para a Preservação do Património

Cultural em Ligas de Prata: O Caso do Retábulo da Sé do Porto.Master Thesis, Lisboa, Portugal

9. Graedel TE (1992) J Electrochem Soc 139:1963–197010. Tran TTM, Fiaud C, Sutter MM (2005) Corros Sci 47:1724–173711. Payer HA (2011) Electrochim Acta 56:2781–279112. Hallet K, Tickett D, McPhail DS, Chater RJ (2003) Apll Surf Sci

203–204:789–792

13. Homem P, Fonseca I, Cavalheiro J (2008) Corros Prot Mater27:82–86

14. Doménech A, Doménech-Carbó MT, Pasies T, Bouzas MC (2011)Electroanalysis 23:2803–2812

15. Rice DW, Peterson PBP, Rogby EB, Phlips PBP, Cappell RJ,Tremoureux R (1981) J Electrochem Soc 128:275–284

16. Chen Z, Liang D, Ma G, Frankel GS, Allen HC, Kelly RG (2010)Corros Eng Sci Technol 45:169–180

17. Franey JP, Graedel TE (1985) J Air Pollut Control Assoc 35:644–648

18. Doménech-Carbó A (2010) J Solid State Electochem 14:363–37919. Costa V, Leyssens K, Richard N, Scholz F (2010) J Solid State

Electrochem 14:449–45120. Costa V, Dubus M (2007) In: Padfield T & Borchersen K (eds)

Museum Microclimates. National Museum of Denmark, Denmark(ISBN 9788-77-602-080-4)

21. Degrigny C (2010) J Solid State Electrochem 14:353–36122. Doménech-Carbó A, Doménech-Carbó MT, Costa V (2009) In:

Scholz F (ed) Electrochemical methods applied to archaeometry,conservation and restoration, Monographs in Electrochemistry Se-ries, Spring, Berlin

23. Doménech-Carbó A (2010) J Solid State Electochem 14:1767–1780

24. Doménech A, Doménech-Carbó MT, Peiró MA (2011) Anal Chem83:5639–5644

25. Doménech A, Doménech-Carbó MT, Martínez-Lázaro I (2010)Anal Chim Acta 680:1–9

26. Skogvold SM, Mikkelsen O, Billon G, Garnier C, Lesven L,Barthe J (2006) Anal Bioanal Chem 384:1567–1577

27. Bernard MC, Dauvergne E, Evesque M, Keddam M, Takenouti H(2005) Corros Sci 47:663–679

28. Scholz F, Meyer B (1998) In: Bard AJ and Rubinstein I (eds)Electroanalytical Chemistry, A Series of Advances 20:1–86. MarcelDekker, New York

29. Scholz F, Schröder U, Gulaboski R (2005) Electrochemistry ofImmobilized Particles and Droplets. Springer, Berlin

30. Scholz F, Lange B (1992) Trends Anal Chem 11:359–36731. Scholz F, Meyer B (1994) Chem Soc Rev 323:341–34732. Scholz F, Nitschke L, Kemnitz E, Olesch T, Henrion G, Hass R,

Bagchi RN, Hermann R, Pruss N, Wilde W (1989) Fresenius ZAnal Chem 335:571–572

33. Scholz F (2011) J Solid State Electrochem 15:1699–170234. Zaky AM, Assaf FH, Abd El Rehim SS, Mohamed BM (2004)

Appl Surf Sci 221:349–35735. Zaky AM, Abd El-Rehim SS, Mohamed BM (2006) Int J Electro-

chem Sci 1:17–3136. Assaf FH, Zaky AM, Abd El-Rehim SS (2002) Appl Surf Sci

187:182–18737. Teijelo ML, Vilche JR, Arvia AJ (1983) J Electroanal Chem

162:207–22338. Hazzazi OA, Zaky AM, Amin MA, Abd El Rehim SS (2008) Int J

Electrochem Sci 3:489–50839. Birss VI, Smith K (1987) Electrochim Acta 32:259–26840. Birss VI, Wright GI (1981) Electrochim Acta 26:1809–181741. Tacconi NR, Rajeshwar K, Lezna RO (1996) J Phys Chem

100:18234–1823942. Langh R, Ankersmit HA, Joosten I (2004) Proceedings of Metal

2004. National Museum of Australia Camberra, Australia43. Ankersmith HA, Doménech A, Tennent NH (2004) Tarnishing of

silver; evaluation by color measurements. In: Proceedings.of theInternational Conference on Metals Conservation. Santiago deChile, Western Australian Museum, Welhpool

44. de la Fuente D, Vega JM, Viejo F, Díaz I, Morcillo M (2011)Atmos Environ 45:1242–1250

45. Camuffo D, Grieken RV, Busse H-J, Sturaro G, Valentino A,Bernardi A, Blades N, Shooter D, Gysels K, Deutsch F, Wieser


M, Kim O, Ulrych U (2001) Atmos Environ 35(Supplement 1):S127–S140

46. Ankersmit HA, Noble G, Ridge R, Stirling D, Kennedy N (2000)Scott Soc Conserv Restor J 11:12–13

47. Zhang S, Meyer B, Moh G, Scholz F (2005) Electroanalysis7:319–328

48. Scholz F, Nitschke L, Kemnitz E, Olesch T, Henrion G, Hass D,Bagchi RN, Herrmann R, Pruss N, Wilde W (1989) Fresenius ZAnal Chem 335:571–572

49. Scholz F, Nitschke L, Henrion G (1990) Electroanalysis 2:85–87

50. Scholz F, Lange B, Jaworski A, Pelzer J (1991) Fresenius J AnalChem 340:140–144

51. Vieira MT, Piedade AP (2006) In: Sudarsham TS et al. (eds) ASTInternational Materials Park, Ohio and IOM Communications Ltd,United Kingdom

52. Thickett D (2004) In: Proceedings of the 6th Indoor Air Quality2004 Meeting (IAQ2004), Pádova, Italy


Linear sweep voltammetry: a cheap and powerful technique for the identification of the silver...

Documents

Transcript of Linear sweep voltammetry: a cheap and powerful technique for the identification of the silver...