Corrosion and corrosion inhibition of Cu–20%Fe alloy in sodium chloride solution

Corrosion and corrosion inhibition of Cu–20%Fe alloy in sodium chloride solution

Available online at www.sciencedirect.com Corrosion Science 50 (2008) 928–937 www.elsevier.com/locate/corsci Corrosion and corrosion inhibition of C...

279KB Sizes 3 Downloads 55 Views

Available online at www.sciencedirect.com

Corrosion Science 50 (2008) 928–937 www.elsevier.com/locate/corsci

Corrosion and corrosion inhibition of Cu–20%Fe alloy in sodium chloride solution S.S. El-Egamy * Department of Chemistry, Faculty of Science, University of Cairo, Giza, Egypt Received 20 February 2006; accepted 20 November 2007 Available online 22 January 2008

Abstract The electrochemical behavior of copper (Cu), iron (Fe) and Cu–20%Fe alloy was investigated in 1.0 M sodium chloride solution of pH 2. The effect of thiourea (TU) addition on the corrosion rate of the Cu–20%Fe electrode was also studied. Open-circuit potential measurements (OCP), polarization and electrochemical impedance spectroscopy (EIS) were used. The results showed that the corrosion rates of the three electrodes follow the sequence: Cu < Cu–20%Fe < Fe. Potentiostatic polarization of the Cu–20%Fe electrode in the range 0.70 V to 0.45 V (SCE), showed that iron dissolves selectively from the Cu–20%Fe electrode surface and the rate of the selective dissolution reaction depends on the applied potential. At anodic potential of 0.45 V, thiourea molecules adsorb at the alloy surface according to the Langmuir adsorption isotherm. Increasing thiourea concentration (up to 5 mM), decreases the selective dissolution reaction and the inhibition efficiency g reach 91%. At [TU] > 5 mM, the dissolution rate of the Cu–20%Fe electrode increases due to formation of soluble thiourea complexes. At cathodic (0.6 V), the inhibition efficiency of thiourea decreases markedly owing to a decrease of the rate of the selective dissolution reaction and/or desorption of thiourea molecules. The results indicated that thiourea acts mainly as inhibitor of the selective dissolution reaction of the Cu–20%Fe electrode in chloride solution. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: A. Cu–Fe alloy; B. EIS; B. Polarization; C. Corrosion; C. Inhibition

1. Introduction Cu–Fe alloys are materials that are more and more extensively applied in industry. Firstly, they are utilized as master alloys to produce new copper alloys for very special purposes. They are used as materials for electrical device components, for example, semiconductor lead frames, electrical connectors, and electrical fuses [1]. The phase diagram of the Cu–Fe system shows that copper solubility in iron is practically high at high temperatures. However, in the range of low temperatures, the copper solubility significantly drops to 1.88 at.% at the eutectoid temperature of 850 °C. A new melting process has been used to prepare Cu–Fe alloys with 10, 20, and 30 wt% Fe, but these alloys show a tendency to segregate [1]. On *

Present address: Chemistry Department, Taif Teacher’s College, Taif University, Taif, P.O. Box 1070, Saudi Arabia. E-mail address: [email protected] 0010-938X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.corsci.2007.11.018

determining and applying the proper parameters of the melting process, it is possible to produce Cu–Fe alloys with such high iron content [1]. Studies of the corrosion behavior of Cu–Fe alloys in aqueous media seem to be rare. It has been reported that iron-alloying addition to copper alloys may be used to improve their corrosion resistance [2]. The mechanisms of the dissolution and passivation of bulk polycrystalline icosahedral Al63Cu25Fe12 specimens during electrolytic corrosion in sodium hydroxide and sulfuric acid solutions were studied [3]. Selective dissolution of Al and Fe from the alloy surface was found to occur at the open-circuit potential, which leads to precipitation of porous layer of re-crystallized copper. After anodic polarization, the dissolution of the alloy is followed by re-deposition of Cu and formation of Cu2O [3]. The adsorption of thiourea onto metal surfaces in aqueous solutions of electrolytes is of great interest since it is known to be effective as a corrosion inhibitor in acid

S.S. El-Egamy / Corrosion Science 50 (2008) 928–937

media [4]. It has been found that thiourea is strongly chemisorbed and the adsorption characteristics are similar to those of halide ions, that is involving sharing or donation of electron pairs [5,6]. As with other organic molecules, the extent of thiourea adsorption depends on several factors including temperature, applied potential, type of electrolyte and thiourea concentration. Previous work showed contradictory results. Most of the authors agreed that the quantity of thiourea adsorbed increases as the potential becomes more positive and the extent of adsorption decreases at negative potentials [7]. The results of the SERS studies by Brown et al. [5], on the other hand, showed the opposite trend. These authors found strong thiourea adsorption at the copper surface polarized at negative potentials and this behavior was observed for potentials up to 0.70 V. The reactions involved in the electro-oxidation of copper in aqueous thiourea-containing solutions were investigated by Bolzan et al. [8]. In the range 0.30 6 E 6 +0.07 V (vs. SCE), the main electrochemical reactions are the electro-oxidation of thiourea to formamidine disulphide FDS, and of copper to several Cu(I)–TU complex ions [8]. At E P +0.07 V, electro-decomposition of FDS and Cu(I)–TU complex ions takes place [8]. At potentials less than 0.30 V, oxidation of thiourea to FDS is not expected and thiourea is assumed to be chemisorbed at the electrode surface as molecular (TU) or protonated (TUH+) species [9]. Thiourea interacts strongly with the surfaces of d-metals (e.g. iron). This interaction has the character of chemisorption. The structure of the adsorbed over-layer, the orientation of the adsorbed particles and the dissociative or associative character of the adsorption were determined [10,11]. Adsorption of thiourea molecules on the surfaces of group IB metals (e.g. copper), is on the other hand, much weaker than with d-elements, and the adsorption process has a reversible character [12]. The effect of adsorption of thiourea on the corrosion kinetics of pure copper and iron samples has been investigated in detail [12]. Investigation of the adsorption and inhibition effects of thiourea on the Cu–20%Fe alloy seems, therefore to be very interesting. The objective of the present work is to investigate the corrosion rate of the Cu–20%Fe electrode in aqueous 1.0 M sodium chloride solution of pH 2. The effects of the polarization potential and thiourea concentration on the corrosion rate of the Cu–20%Fe electrode were studied. The data were compared with those of the constituent elements, namely copper and iron. In these investigations, conventional electrochemical techniques were used including open-circuit potential measurements (OCP), polarization and the electrochemical impedance spectroscopy (EIS). 2. Experimental details The copper and iron test samples were cut from spectroscopically pure rods (Johnson & Matthey). The copper– iron alloy with chemical composition (Cu–20%Fe), was one of the products of the company of the ferrous and

929

non-ferrous metals and alloys, Helwan, Egypt. All electrodes were prepared with a constant surface area of 0.196 cm2 as described elsewhere [13–15]. Before each experiment, the working electrode was polished with emery papers of different grades, wrapped against smooth cloth, degreased with ethanol, washed with tri-distilled water, and then immediately immersed in the test solution. The purity of thiourea and sodium chloride chemicals used in this study was 99% (Fluka). An aqueous solution of 1.0 M sodium chloride was adjusted at pH 2 and used as a stock solution. The different thiourea concentrations were prepared by appropriate dilutions using the sodium chloride stock solution. Electrochemical measurements were made in a conventional three-electrode cell with a platinum spiral wire counter electrode and saturated calomel (SCE) reference electrode. All reported potentials in this study were measured and referred to the SCE. Open-circuit potential, impedance (EIS) measurements, data analysis and potentiostatic polarization were performed with an IM5d impedance and electrochemical measurement system (Zahner, Germany). The sinusoidal potential perturbation was 10 mV in amplitude with frequencies ranging from 100 kHz to 0.1 Hz. The polarization curves were recorded for the copper electrode from 0.40 V to 0.10 V vs. the SCE. For the other two electrodes, polarization was started from 0.70 V and terminated at 0.40 V. linear polarization (LP) experiments were carried out in the vicinity of the corrosion potential (±10 mV). The scan rate in all polarization experiments was 1 mV/s. Most of the experiments were repeated twice to ensure reproducibility. The Polarization resistances of the different electrodes, which are criteria of the corresponding corrosion rates, were obtained from impedance and linear polarization experiments. The results were compared and interpreted. 3. Results and discussion 3.1. Open-circuit potential measurements Fig. 1 shows the variation in open-circuit potentials of the mechanically polished copper, iron and the Cu– 20%Fe electrodes with time in 1.0 M sodium chloride solution of pH 2. For all electrodes, the open-circuit potentials drift with time in the active direction and tend to stabilize within 60 min. This negative shift of potentials is related to active dissolution of the different electrodes in the sodium chloride solution. The steady state open-circuit potential of the copper electrode in the chloride solution amounted to 0.256 V, which is in the stability region of  CuCl 2 predicted by the E–pH diagram for the Cu–Cl –  H2O system [16]. In acidic chloride media (0.1 Ml 6 [HCl] 6 3 Ml), the diffusion of chloride ion does not exist because the amount of chloride ions consumed is negligible compared to the large quantities present in the solution [17]. Also, Cu+ and Cu++ do not exist in the diffusion layer on account of the high chloride concentration. The diffu-

930

S.S. El-Egamy / Corrosion Science 50 (2008) 928–937

3.2. Corrosion rate measurements

-0.20

It is well known that the polarization resistance is inversely proportional to the corrosion rate. The polarization resistances of the different electrodes were determined from impedance measurements and linear polarization plots. The results obtained from the two methods were compared and interpreted.

Cu

-0.22

-0.26

Cu-20%Fe -0.48 -0.52

Fe -0.56 -0.60

0

20

40

60

80

100

Time / min Fig. 1. Potential vs. time curves for copper, iron and Cu–20%Fe electrodes in 1.0 M sodium chloride solution of pH 2.

sion of CuCl 2 complex is therefore considered to be the rate determining step. The results from ring-disk studies also showed that the slow cuprous chloride complexes transferred towards the bulk is the rate determining step in the copper dissolution reaction [18,19]. A mechanism was suggested that describes the adsorption of CuCl on the corroding copper surface and the diffusion of CuCl 2 complex away from the surface [20,21] Cu þ Cl ! CuClads þ e 

CuClads þ Cl !

CuCl 2

ð1Þ ð2Þ

A mechanism for the iron dissolution reaction in acidic solutions containing halide ions was proposed [22]. This mechanism suggests that iron dissolves through two dissolution paths in the presence of chloride ions, i.e. reaction paths through a FeOH adsorbed intermediate and through a FeCl adsorbed intermediate. In addition to the probable formation of FeOH(ads), iron reacts in the presence of chloride ions according to [22] Fe þ Cl ! FeClðadsÞ þ e

ð3Þ

FeClðadsÞ ! FeClþ ðadsÞ þ e

ð4Þ

the quantity of FeOH(ads) on the iron surface decreases as the concentration of the chloride ion increases. The close values of the open-circuit potentials of the pure iron and the Cu–20%Fe electrodes (cf. Fig. 1), suggests that iron dissolves from both electrodes as Fe++. A possible mechanism for the dissolution of the Cu–20%Fe electrode in acidic chloride solution is formulated as follows: Anodic: ðCu–20%FeÞ ! xFe2þ þ 2xe

3.2.1. Results from EIS measurements The impedance spectra of the copper, iron and the Cu– 20%Fe electrodes were recorded at the corresponding rest potentials at various immersion times in 1.0 M sodium chloride solution of pH 2. Fig. 2 shows comparative impedance spectra recorded for the three electrodes after 60 min immersion. It is obvious that the investigated electrodes display different frequency responses under the same conditions. The impedance spectrum of the copper electrode consisted of one capacitive loop in the high frequency region corresponding to a charge-transfer reaction. At low frequencies, the spectrum shows the beginning of a diffusional contribution. This indicates that the corrosion reaction at the copper electrode is influenced by both charge-transfer and mass transport processes. The mass transport limitation is always due to diffusion of CuCl 2 away from the copper electrode surface [16]. The impedance spectrum recorded for the iron electrode consisted of two capacitive semicircles (cf. Fig. 2). The high frequency capacitive loop is small and is a result of a fast charge-transfer process. The low frequency capacitive loop arises from specific adsorption of the different reaction products at the iron electrode surface. The presence of two capacitive semicircles indicates the presence of two time constants corresponding to iron dissolution through two dissolution paths in the presence of chloride ions [22]. In these paths, FeOH and FeCl adsorbed intermediates are 200 Cu 150

Z'' / Ω.cm2

E (SCE) / V

-0.24

Cu-20%Fe 100

Fe

50

0

ð5Þ 0

Cathodic: 1 2ðxÞHþ þ ðxÞO2 þ ð2xÞe ! ðxÞH2 O 2

100

200

300

400

Z' / Ω.cm2

ð6Þ

Fig. 2. Nyquist plots for copper, iron and Cu–20%Fe electrodes after 60 min immersion in 1.0 M sodium chloride solution of pH 2.

S.S. El-Egamy / Corrosion Science 50 (2008) 928–937

formed. The diameter of each of the two semicircles increases with the increase of the immersion time (not seen), and at a given time, the diameter of the second semicircle is greater than that of the first semicircle. This indicates that the corrosion process is controlled mainly by the second time constant in the low frequency region. For the Cu– 20%Fe electrode, the impedance spectrum consisted of only one depressed semicircle indicating the presence of only one time constant. Detailed analysis of the open-circuit impedance data for the different electrodes presented in Fig. 2 showed appreciable deviation from the ideal behavior. This is characteristic of what is termed as frequency dispersion. In this type of deviation, the Nyquist presentations of the impedance plots consist of semicircular arcs with their centers lying below the real axis. This indicates the generation of micro roughness surface heterogeneities at the surface during the corrosion process [23,24]. Similar results were reported in several studies [13,14]. According to the AC circuit theory, an impedance plot obtained for a given electrochemical system can be correlated to one or more equivalent circuit models. At the same time, a given model can be applied for different systems. When comparing the corrosion rates of the different electrodes and for simplicity, no fitting was done. Only the difference between the values of the absolute impedance |Z|, measured at low (0.1 Hz) and high frequency (100 kHz) were obtained for the different electrodes. These values

931

were taken qualitatively as measures of the corresponding polarization resistances Rimp p , as suggested by Macdonald’s [25]. The calculated Rimp values of the different electrodes p are listed in Table 1. It is obvious that Rimp of the different p electrodes follow the sequence: Cu > Cu–20%Fe > Fe. This indicates that the Cu–20%Fe alloy is more corrosion resistant than the pure iron electrode. 3.2.2. Results from polarization data Fig. 3a shows Tafel plots for the three electrodes recorded in 1.0 M sodium chloride solution of pH 2. These curves were taken after equilibration of the electrodes in the test solution under open-circuit condition for 60 min. The corrosion current densities icorr of the different electrodes were calculated from the intercepts of the straight lines (cathodic and/or anodic branches) of the Tafel plots with the corresponding corrosion potentials Ecorr. The different corrosion parameters are listed in Table 1. The anodic polarization curve of the copper electrode exhibits a linear region with slope amounting to +67 mV, which is close to the 60 mV per decade of the current. This is typical of mass-transfer controlled copper dissolution process occurring in chloride media [16]. Similar linear Tafel regions with slopes close to 60 mV were identified previously for all anodic curves recorded for copper in sodium chloride solutions [20,26,27], which become less sensitive to hydrodynamic conditions at longer exposure times. The cathodic branch of the copper polarization curve

Table 1 Corrosion parameters of the different electrodes in 1.0 M sodium chloride solution of pH 2 after 60 min immersion Sample

Ecorr (V)

icorr (lA/cm2)

ba (mV)

bc (mV)

2 Rpol p ðX cm Þ

Rimp ðX cm2 Þ p

Cu Fe Cu–20%Fe

0.247 0.578 0.572

43.6 120.2 71.5

67 59 61

– 138 122

524 160.7 265

503.0 153.0 255.5

a

b

-1

10

Cu

-2

Fe

5

Cu-20%Fe

η / mV

log (i / A.cm-2)

-3

-4

0

-5

Cu Fe Cu-20%Fe

-5 -6

-7

-10 -0.7

-0.6

-0.5

-0.4

-0.3

E (SCE) / V

-0.2

-0.1

0.0

-0.04

-0.02

0.00

0.02

0.04

i / (mA/cm2)

Fig. 3. (a) Tafel plots for copper, iron and Cu–20%Fe electrodes in 1.0 M sodium chloride solution of pH 2. (b) Polarization resistance determination for copper, iron and Cu–20%Fe electrodes in 1.0 M NaCl solution of pH 2.

932

S.S. El-Egamy / Corrosion Science 50 (2008) 928–937

shows a cathodic peak (a), at E = 0.328 V, which perturbate the linear region. This peak is probably attributed to reduction of CuCl film. A cathodic plateau is then observed between 0.37 and 0.50 V due to O2 reduction. Similar cathodic peak was identified previously and its position was found to depend strongly on the rotation rate as well as the chloride ion concentration [21]. The anodic Tafel slopes for the iron and the Cu–20%Fe electrodes are very close (cf. Table 1), indicating dissolution of these electrodes by the same mechanism. In theory, for processes occurring under activation control, it is possible to determine the corrosion rate (corrosion current density icorr), from Tafel lines by back extrapolation of the linear sections of the anodic or cathodic polarization curves to the corrosion potentials. This gives the corrosion current which can be introduced in the Stern– Geary equation to calculate the corresponding polarization resistance. However, this method cannot be so easily applied in unstirred solutions, which correspond to corrosion conditions commonly encountered in service. This is because under these circumstances, the transport of reactants (e.g. H+ ions or O2) or products (e.g. M+z ions) away from the electrode surface proceed more sluggishly than the charge-transfer reaction and so can contribute in determining the overall rate of the corrosion reaction. Under such conditions, the electrode process is controlled by mass-transfer, and the partial cathodic line may no longer be a Tafel line. This behavior is clearly observed in the present study for the copper electrode in sodium chloride solution (cf. Fig. 3a). Under these circumstances, back extrapolation procedures lead to very dubious results and corrosion rate cannot be obtained from the Tafel plots. The polarization resistances of the different electrodes, which are inversely related to the corresponding corrosion rates, can therefore be calculated using the linear polarization (LP) method. In this method, the electrode is polarized in the vicinity of the corrosion potential (±10 mV) at a scan rate 1 mV/s. The polarization resistances of the  different dg electrodes Rpol , were calculated as the slopes of the linp di ear plots, where g is the over-potential in mV and i is the current density in mA/cm2. Fig. 3b shows linear polarization plots for the copper, iron, and the Cu–20% Fe electrodes in 1.0 M sodium chloride solution of pH 2. These plots have been constructed from the average of two independent experiments. The values of the polarization resistances Rpol p , of the different samples were obtained as the slopes of these plots and the data are listed in Table 1. It is clear that there is a good agreement between the data obtained from the impedance and the linear polarization techniques. 3.3. Potentiostatic data In this series of experiments, the Cu–20%Fe electrode was polarized at different constant potentials in the range 0.70 to 0.45 V. At each potential, the current was monitored for a period of 60 min, whereby steady state condi-

tion was achieved. At the end of each polarization, the impedance spectrum was recorded while the electrode is still polarized. Fig. 4a shows current transients for the Cu–20%Fe electrode polarized at different potentials in 1.0 M sodium chloride solution of pH 2. It is evident that, at E 6 0.60 V, negative currents were recorded. At these potentials, oxygen reduction seems to be the predominant reaction. At E > 0.60 V, positive currents were observed due to alloy dissolution. The current increases initially with time reaching a maximum, and then falls down at longer immersion times. The initial current maximum increases as the polarization potential increases (cf. Fig. 4a). This behavior may be connected to selective leaching of the more active component, iron, from the Cu–20%Fe electrode surface. The Nyquist presentations of the impedance data for the Cu–20%Fe electrode polarized at different potentials are shown in Fig. 4b. The existence of a single depressed semicircle at all potentials depicts the presence of a single charge-transfer process during alloy dissolution. It is observed that the diameter of the semicircular arc decreases with the increase of the polarization potential, indicating increase of the alloy dissolution rate. The experimental impedance data in Fig. 4b were fitted to the proposed equivalent circuit model shown in Fig. 4c. This model is made of a solution resistance, Rs, in series with a parallel combination of a constant phase element, Q, and polarization resistance Rimp p . The constant phase element was introduced as a substitute of the double layer capacitance to account for frequency dispersion [12,23]. Fig. 4d presents fitting of the impedance data for the Cu–20%Fe electrode polarized at 0.45 V for 60 min as an example. The percentage error in the impedance value amounted to 2.1% and the deviation in the phase angle was equal to 1.8%. These results indicate good fitting between the experimental impedance data and the proposed model. The different electrochemical impedance parameters obtained from fitting of the impedance data at different potentials are listed in Table 2. It is obvious that Rimp decreases markedly with p the increase of the polarization potential indicating increase of the alloy dissolution rate. In duplex and multiphase alloys, the phases have different electrochemical potentials and there is consequently always a tendency for the most anodic phase to corrode preferentially. The extent, to which this occurs, depends on how great the potential difference is between the anodic phase and the surrounding phases. The dissolution of iron as (Fe2+) into acid solutions occurs above 0.617 V, whereas copper dissolution occurs above +0.165 V [28], assuming the concentration of each of Fe2+ and Cu2+ are 106 Ml1. Between these potentials, and for thermodynamic reasons, only iron dissolution can occur. Enrichment of copper on the alloy surface is expected [28]. It is therefore, evident that the Cu–20%Fe electrode, under the present experimental conditions, corrodes by selective dissolution of iron from the alloy surface. This is supported by the appearance of a reddish color on the alloy surface at

S.S. El-Egamy / Corrosion Science 50 (2008) 928–937

-0.45 V

1.0

i / (mA.cm-2)

Table 2 Impedance parameters of the Cu–20% Fe electrode polarized at different potentials in 1.0 M NaCl of pH 2

a

1.5

-0.475 V

0.5

933

E (V)

Rs (X cm2)

Rp (X cm2)

Q (X1 cm2 sn)

n

0.600 0.550 0.525 0.500 0.450

0.90 1.00 1.30 1.84 1.70

242.0 199.0 151.0 58.1 24.5

4.80  103 2.78  103 1.50  103 1.40  103 1.50  103

0.62 0.61 0.69 0.81 0.93

-0.50 V -0.55 V

0.0

the end of the polarization experiments. Although preferential dissolution is expected in this study, the kinetics characteristics of the Cu–20%Fe electrode don’t shift towards the copper characteristics (cf. Fig. 1). This may be attributed to the short immersion time (60 min) considered in the present study, where the deposited copper film is weak (not coherent) and not completely cover the alloy surface. Similar preferential dissolution was reported for a Ni– Fe alloy in 1.0 M sulfuric acid solution at room temperature [29].

-0.6V -0.7V

-0.5

0

20

40

60

Time / min

-0.6V

Z'' / Ω.cm2

120

b

80

3.4. Effect of thiourea

-0.55V 40

-0.525V

-0.5V

0 -0.45 V 0

40

80

120

160

Z' / Ω.cm2 150

d

simulated experimental

Z" / Ω.cm2

100

Q Rs

50

R imp p

c 0

0

100

200

300

Z' / Ω.cm2 Fig. 4. (a) Current transient plots for the Cu–20%Fe electrode polarized at different potentials in 1.0 M sodium chloride solution of pH 2. (b) Nyquist plots for the Cu–20%Fe electrode polarized at different potentials in 1.0 M sodium chloride solution of pH 2. (c) Equivalent circuit model for the impedance data under polarization. (d) Comparison of the experimental data (d) with the simulated values (—) for the Cu–20%Fe electrode polarized at 0.45 V.

The effect of thiourea addition on the anodic and the cathodic reactions of the Cu–20%Fe electrode was investigated in 1.0 M sodium chloride solution of pH 2 using impedance and polarization techniques. Fig. 5a and b shows potentiostatic current transients recorded for the Cu–20%Fe electrode anodically polarized at 0.45 V in sodium chloride solution containing different thiourea concentrations. Increasing thiourea concentration, (up to 5 mM), decreases the current maximum from 1.81 to 0.16 mA/cm2 (cf. Fig. 5a). This gives an inhibition efficiency of 91% for the selective dissolution process. Above 5 mM, the current maximum disappeared and the current decreases abruptly reaching steady state value, (cf. Fig. 5b). This decrease of the current density may be attributed to accumulation of dissolution products at the alloy surface. The value of the steady state current increases with the increase of thiourea concentration indicating acceleration of the alloy dissolution rate and formation of soluble thiourea complexes. Fig. 6a and b presents Nyquist plots recorded for the Cu–20%Fe electrode polarized at 0.45 V for 60 min in 1.0 M sodium chloride solution of pH 2 at two ranges of thiourea concentrations. All the spectra consisted of single depressed semicircles. It is obvious that the effect of thiourea on the corrosion rate of the alloy is inverted at a critical thiourea concentration. At low thiourea concentrations, the diameter of the semicircular arc increases with the increase of thiourea concentration (cf. Fig. 6a). This continues until a critical thiourea concentration of 5 mM. Above this limit and with further increase of thiourea concentration, the opposite trend was observed (cf. Fig. 6b). The initial increase of the diameter of the Nyquist semicircle with increasing thiourea concentration

934

S.S. El-Egamy / Corrosion Science 50 (2008) 928–937 2.0

a

b

0.4 1.5

i / (mA cm-2)

i / (mA.cm-2)

50 mM TU

0.3

nil

1.0

1 μM 10 μM

0.5

0.2

100 μM

25

1 mM 10

5 mM TU

0.0 0.1

0

20

40

60

80

Time / min

0

20

40

60

80

Time / min

Fig. 5. (a) Current transient plots for the Cu–20%Fe electrode polarized at 0.45 V in 1.0 M sodium chloride solution of pH 2 containing low thiourea concentrations. (b) Same as in Fig. 5a but with higher thiourea concentrations.

Z''/ Ωcm

2

60

5 mM TU 1 mM

40 100 μM

20 1μ μM

0 0

40

a

10 μM

80

Z'/ Ωcm

120

2

60

Z"/ Ωcm

160

2

10 mM TU

40 25 mM 20 50 mM

b

0 0

40

80

120

160

Z'/ Ωcm2 Fig. 6. (a) Nyquist plots for the Cu–20%Fe electrode polarized at 0.45 V in 1.0 M sodium chloride solution of pH 2 at low thiourea concentrations. (b) Same as in Fig. 6a but with higher TU concentrations.

indicates increase of the inhibition by thiourea addition. The opposite trend observed at higher thiourea concentrations may be attributed to the possible formation of soluble thiourea complexes leading to increase of the dissolution rate [5]. Under the present experimental conditions, oxidation of thiourea to FDS is not expected since the applied

potential is too negative to allow for such oxidation to occur [8]. Thiourea, is therefore, assumed to be chemisorbed at the alloy surface. Both the molecular (TU) and the protonated (TUH+) species would undergo specific adsorption. In the present study, where the acidity of the medium is weak (pH 2), the inhibition effect of thiourea seems to be more likely due to specific adsorption of the molecular species (i.e. blocking effect), than to specific adsorption of the cationic species [9]. The impedance data in Fig. 6a and b were fitted to the equivalent circuit shown in Fig. 4c. The different impedance parameters are listed in Table 3. The inhibition efficiency g of thiourea can be calculated using the impedance data in Table 3 and the following formula:   tcorr  tcorr g ð%Þ ¼  100 tcorr 0 1  1 1 imp Rimp  R p B p inh C ð7Þ ¼@ A  100  1 imp Rp where t and t0 are the corrosion rates in the uninhibited and the inhibited solutions, respectively. The corresponding values of the polarization resistances are ðRimp p Þ and ðRimp Þ . Values of the percentage inhibition g obtained inh p on the basis of impedance data are listed in Table 3. It is obvious that there is a critical concentration of thiourea at 5 mV, below which thiourea acts as a corrosion inhibitor, whereas at higher thiourea concentrations, it accelerates the dissolution rate of the alloy. The relatively higher values of Rimp observed at [TU] > 5 mM, compared to the p value in the pure solution, may be attributed to the existence of corrosion products still adsorbed at the alloy surface. With further increase of thiourea concentration, detachment of some of these adsorbed species takes place leading to a decrease of the value of Rimp p .

S.S. El-Egamy / Corrosion Science 50 (2008) 928–937

935

Table 3 Impedance parameters of Cu–20% Fe electrode in 1.0 M sodium chloride solution containing different thiourea concentrations at 0.45 V [TU] (mMl1) Nil 0.001 0.01 0.1 1.0 2.0 5.0 10.0 25.0 50.0

Q (X1 cm2 sn)

Rs (X cm2)

3

2.15  10 2.85  103 3.33  103 2.72  103 2.39  103 3.20  103 3.70  103 3.86  103 4.13  103 6.30  103

1.8 1.6 1.7 1.2 1.5 1.5 1.3 1.4 1.3 1.3

The inhibition efficiency of thiourea was also determined by polarization technique. The Tafel plots for the Cu– 20%Fe electrode were traced after equilibration for 60 min in 1.0 M sodium chloride solution of pH 2 containing different thiourea concentrations. For clarity, Tafel plots for the Cu–20%Fe electrode in the pure solution and in the presence of 5 mM thiourea, as a typical example, are presented in Fig. 7a. It is obvious that addition of thiourea shifts the polarization curve to the right. Further, both the cathodic and the anodic curves show lower current densities in the presence of thiourea. This indicates that thiourea inhibits both the anodic and the cathodic processes suggesting that it is a mixed-type inhibitor. The polarization resistances Rpol p , of the Cu–20%Fe electrode in the pure and in the inhibited sodium chloride solutions, were calculated from the slopes of the linear polarization plots shown in Fig. 7b. The estimated values are 265 and 937 X cm2, for the pure and the inhibited solutions, respectively. This gives an inhibition efficiency of 71.7% for the corrosion process. The corresponding Rimp values obtained from impedance p spectra recorded at the corrosion potentials after 60 min immersion (not shown), are 275 and 818 X cm2 for the pure

a

g (%)

n

24.6 28.7 53.2 111.4 185.5 202.0 296.0 253.0 135.0 47.30

– 14.28 53.78 77.91 85.73 87.80 91.68 90.27 83.92 47.99

0.87 0.84 0.72 0.65 0.65 0.63 0.59 0.61 0.59 0.55

and the inhibited solutions, respectively. This produces an inhibition efficiency of 66.4%. It is clear that the inhibition efficiency estimated from the linear polarization method is much agreeing with that obtained from impedance measurements. When considering adsorption isotherm, it is conventional to adopt a definition of surface coverage (h) which defines the ratio of adsorbate species to surface substrate atoms. When the surface is completely covered with the adsorbate, h reaches its maximum value (hmax = 1). The ratio of the surface coverage in the inhibited solution to the value in the uninhibited one gives the inhibition efficiency of the inhibitor used and is given the symbol g. Assuming a direct relation between the inhibition efficiency g and the surface coverage h, of the inhibitor, the impedance data recorded for the Cu–20% Fe electrode polarized at 0.45 V, were used to analyse the adsorption mechanism. The Langmuir adsorption isotherm [30] may be expressed as h¼

b

-3

Rp (X cm2)

KC 1 þ KC

ð8Þ

10

nil 5

5 mM TU -5

η / mV

log (i / A.cm-2)

-4

0

-5

-6

nil 5 mM TU

-7 -0.72

-10 -0.64

-0.56

E (SCE) / V

-0.48

-0.02

0.00

0.02

i / (mA.cm-2)

Fig. 7. (a) Tafel plots for the Cu–20%Fe electrode in 1 M sodium chloride solution without and with 5 mM thiourea. (b) Polarization resistance determination for the Cu–20%Fe electrode in 1.0 M sodium chloride solution without and with 5 mM thiourea.

936

S.S. El-Egamy / Corrosion Science 50 (2008) 928–937

where K is the equilibrium constant for the adsorption process, C is the inhibitor concentration in mM and h is the surface coverage. Rearranging Eq. (8): C 1 ¼ þC h K

ð9Þ

Fig. 8 shows plot of Ch vs. C for the thiourea inhibitor. The data fit straight line indicating that thiourea molecules adsorb at the surface of the Cu–20%Fe electrode according to the Langmuir adsorption isotherm. The effect of thiourea addition on the cathodic reaction occurring at the Cu–20%Fe electrode surface was studied by impedance spectroscopy. Fig. 9 presents Nyquist plots recorded for the Cu–20%Fe electrode cathodically polar-

ized at 0.60 V in pure sodium chloride solution and in the presence of 5 mM thiourea. It is observed that addition of thiourea increases the impedance of the alloy indicating that it also inhibits the cathodic reaction. The values of the polarization resistances Rimp p , were calculated as the difference in the values of the absolute impedances at the highest and the lowest frequencies [25]. The estimated values are 186 and 318 X cm2 for the pure and the inhibited solutions respectively. This gives an inhibition efficiency of 41.5%. The inhibition efficiency of thiourea at 0.6 V is obviously much lower than that obtained at 0.45 V. A probable reason for this is the decrease of the rate of the selective dissolution reaction and desorption of thiourea molecules at negative potentials [5]. Similar behavior was observed previously [31,32]. The results showed that thiourea acts as a mixed-type inhibitor with a predominately anodic action.

6

4. Summary

(C/θ) / mM

4

2

0 0

1

2

3

4

5

6

C / mM Fig. 8. Fitting Langmuir’s adsorption isotherm for the Cu–20%Fe electrode in 1.0 M sodium chloride solution of pH 2 with different thiourea concentrations.

1. The corrosion rates of copper, iron and the Cu–20%Fe electrodes in 1.0 M sodium chloride solution of pH 2 follow the sequence Cu < Cu–20% Fe < Fe. 2. The Cu–20%Fe electrode corrodes in sodium chloride solution by selective dissolution of the more active, Fe, component and the rate of the selective dissolution reaction increases with increasing applied potential. 3. Under the present experimental conditions, thiourea molecules act both as inhibitor and accelerator. At [TU] < 5 mM, it inhibits the selective dissolution reaction while above 5 mM, it acts as accelerator of the corrosion reaction. 4. The inhibition efficiency of thiourea at 0.45 V is higher than that at 0.60 V due to desorption of thiourea molecules and decrease of the rate of the selective dissolution reaction at negative potentials. Thiourea, therefore acts as a mixed-type inhibitor with a predominant anodic action.

200

References 5 mM TU

Z"/ Ωcm

2

150

nil

100

50

0 0

50

100

Z'/ Ωcm

150

200

2

Fig. 9. Nyquist plots for the Cu–20%Fe electrode polarized at 0.60 V in 1.0 M sodium chloride solution without and with 5 mM thiourea.

[1] N. Nomoto, T. Chingping, M. Ohta, K. Yamakawa, 62-Hitachi Cable Review, No. 18, October 1999. [2] J.H. Cairs, Ferrous sulfate treatment to prevent corrosion of condenser and exchange tubes, Technical Bulletin TB/7, Yorkshire Imperial Metals Ltd., Leeds, 1990. [3] Ru¨diger, U. Ko¨ster, in: MRS Symposium: Quasicrystals, MRS Fall Meeting, Boston, November 27–December 1, 2000. [4] N. Tantavichet, M.D. Pritzker, Electrochim. Acta 50 (2005) 1849. [5] G.M. Brown, G.A. Hope, D.P. Schweinsberg, P.M. Fredericks, J. Electroanal. Chem. 380 (1995) 161. [6] M. Tian, W.G. Pell, B.E. Conway, J. Electroanal. Chem. 552 (2003) 279. [7] J.O’M. Bockris, M.A. Habib, J.L. Carbajal, J. Electrochem. Soc. 131 (1984) 3032. [8] A.E. Bolzan, A.S.M.A. Haseeb, P.L. Schilardi, R.C.V. Piatti, R.C. Salvarezza, A.J. Arvia, J. Electroanal. Chem. 500 (2001) 533. [9] B.G. Ateya, B.E. El-Anadouli, F.M. El-Nizamy, Corros. Sci. 24 (1984) 497. [10] A.T. Hubbard, Chem. Rev. 88 (1988) 633.

S.S. El-Egamy / Corrosion Science 50 (2008) 928–937 [11] J. Lipkowski, L. Stolberg, D.F. Yang, B. Pettinger, S. Mirwald, F. Henglein, D.M. Kolb, Electrochim. Acta 39 (1994) 1045. [12] Z.D. Stankovic, M. Vukovic, Electrochim. Acta 41 (1996) 2529. [13] S.S. El-Egamy, W.A. Badawy, J. Appl. Electrochem. 34 (2004) 1153. [14] S.S. El-Egamy, W.A. Badawy, H. Shehata, Corros. Prevent. Control 51 (2004) 89. [15] S.S. El-Egamy, Corrosion 62 (2006) 739. [16] D. Tromans, J.C. Silva, Corrosion 53 (1997) 171. [17] A. Moreau, J.P. Frayret, F. Del Ray, R. Pointeau, Electrochim. Acta 27 (1982) 1281. [18] C.H. Bonfiglio, H.C. Albaya, O.A. Cobo, Corros. Sci. 13 (1973) 717. [19] F.K. Crundwell, Electrochim. Acta 37 (1992) 2707. [20] H.P. Lee, K. Nobe, J. Electrochem. Soc. 133 (1986) 2035. [21] C. Deslousis, B. Tribollet, G. Mengoli, M.M. Musiani, J. Appl. Electrochem. 18 (1988) 384.

937

[22] M. Itagaki, M. Tagaki, K. Watanbe, Electrochim. Acta 41 (1996) 1201. [23] K. Juttner, Electrochim. Acta 35 (1990) 1501. [24] T. Pajkossy, J. Electroanal. Chem. 364 (1994) 111. [25] J.R. Macdownald’s, Impedance Spectroscopy, John Wiley & Sons, New York, 1987. [26] A.L. Bacarella, J.C. Griess, J. Electrochem. Soc. 120 (1973) 459. [27] L. Brossard, J. Electrochem. Soc. 130 (1983) 403. [28] M. Pourbaix, Atlas of Electrochemical Equilibria in Aqueous Solution, NACE, Houston, TX, USA, 1974. [29] B.N. Popov, K.M. Yin, R.E. White, J. Electrochem. Soc. 140 (1993) 1321. [30] I. Langmuir, J. Chem. Soc. 39 (1947) 1848. [31] M. Fleischmann, I.R. Hill, G. Sundholm, J. Electroanal. Chem. 157 (1983) 359. [32] Z.Q. Tian, Y.Z. Lian, M. Fleischmann, Electrochim. Acta 35 (1990) 879.