Corrosion behavior of Cu55Zr35Ti10 metallic glass in the chloride media

Materials Chemistry and Physics 134 (2012) 938e944

Contents lists available at SciVerse ScienceDirect

Materials Chemistry and Physics journal homepage: www.elsevier.com/locate/matchemphys

Corrosion behavior of Cu55Zr35Ti10 metallic glass in the chloride media A.H. Cai a, b, *, X. Xiong b, Y. Liu b, W.K. An a, G.J. Zhou a, Y. Luo a, T.L. Li a a b

College of Mechanical Engineering, Hunan Institute of Science and Technology, Yueyang 414000, PR China State Key Laboratory of Powder Metallurgy, Central South University, Changsha 410083, PR China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 5 March 2011 Received in revised form 3 February 2012 Accepted 10 March 2012

Cu55Zr35Ti10 (at. %) ribbon was prepared by melt spinning. Its glassy structure was confirmed by X-ray diffraction (XRD) and differential scanning calorimetry (DSC), respectively. Its corrosion behavior in HCl and NaCl solutions was investigated by electrochemical polarization and immersion measurements. The surfaces before and after corrosion were observed with scanning electron microscope (SEM). The corrosion potential and corrosion resistance of the Cu55Zr35Ti10 metallic glass both decrease with increasing chloride concentration, and are higher in NaCl than in HCl. The current density in anodic curve sharply decreases when the potential reaches up to a value and the chloride concentration is more than 0.5 M in both HCl and NaCl solutions. The different corrosion behavior in HCl and NaCl is also carefully discussed. Ó 2012 Elsevier B.V. All rights reserved.

Keywords: Amorphous materials Microstructure Corrosion test Corrosion

1. Introduction Metallic glasses have acquired significant attention from the scientific and technological viewpoints. They usually show high strength, large elastic strain limit, and excellent wear and corrosion resistances, along with other remarkable engineering properties [1]. Many metallic glasses have been used in practical applications [2,3]. However, the applications of the metallic glasses require high chemical stability in various environments in order to ensure its lifetime. Without high corrosion resistance in the service environments, their favorable mechanical properties cannot be fully exploited. Up to now, many corrosion studies have already been reported for metallic glasses in different corrosive media [4e11]. In order to expand the fields of applications of the metallic glasses, the development of new metallic glasses with better mechanical properties and higher corrosion resistance for lower cost is desirable. Compared with Zr- and Pd-based metallic glasses, Cubased metallic glasses exhibit even higher mechanical properties and lower cost [12,13]. In addition, Cu-based metallic glasses are expected to have applications due to their potential as important catalysts. Molnár et al. [14] investigated activation of amorphous CueM (M ¼ Ti, Zr or Hf) alloy powders by being applied as catalyst in the transformation of various alcohols at elevated temperature

* Corresponding author. College of Mechanical Engineering, Hunan Institute of Science and Technology, Yueyang 414000, PR China. Tel.: þ86 730 8648516; fax: þ86 730 8648806. E-mail address: [email protected] (A.H. Cai). 0254-0584/$ e see front matter Ó 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.matchemphys.2012.03.094

(523e573 K). Active, stable and selective catalysts were generated from CueZr and CueHf during the dehydrogenation of 2-propanol to acetone and the transformations of allyl alcohol to form propanal and 1-propanol. CueZreTi ternary system is a typical glass forming system firstly explored by Inoue’s group [15]. The maximum size for glass formation in this system can be up to 5 mm [13]. In order to apply this type of metallic glass as engineering material, the corrosion behavior of Cu-based metallic glasses, such as CueZr [10,11], CueZreTie(Mo, Ta and Nb) [16], CueZreTi-Nb [17], CueHfeTie(Mo, Ta and Nb) [18], CueZreTieNie(Nb, Cr, Mo and W) [19e22], and CueZreAleY [23] has been carried out. The chloride ion is one of the common corrosive media. It is important for theory and application by the investigation of the corrosion behavior of the metallic glasses in the chloride medium. Zender et al [24] investigated the corrosion performances of Cu46Zr42Al7Y5 and Zr58.5Cu15.6Ni12.8Al10.3Nb2.8 bulk metallic glasses in 0.001e0.1 M HCl aqueous solutions. They found that in both cases the corrosion potential changed to more positive potentials and the corrosion current increased with increasing chloride concentration. However, the pitting potential decreased and the usable passive region became smaller with increasing chloride concentration. Recently, Gostin et al [5] have investigated the corrosion behavior of (Fe44.3Cr5Co5Mo12.8Mn11.2C15.8B5.9)98.5Y1.5 bulk metallic glass in 0.01e0.6 M NaCl aqueous solutions. They found that the corrosion potential was larger in 0.01 M NaCl aqueous solution than in 0.1 and 0.6 M NaCl aqueous solutions. However, few researchers have carefully investigated the influence of the chloride concentration on the corrosion behavior of

A.H. Cai et al. / Materials Chemistry and Physics 134 (2012) 938e944

939

CueZreTi ternary metallic glasses. Thus, it is important for investigating the influence of the chloride concentration on the corrosion property of Cu-based metallic glasses. In the present work, the corrosion behavior of Cu55Zr35Ti10 metallic glass is investigated in HCl and NaCl aqueous solutions with different chloride concentration. 2. Experimental procedures Cu55Zr35Ti10 (at. %) ternary alloy ingots were prepared from the mixture of pure metals by arc melting in an argon atmosphere. Ribbon samples with a thickness of 50 mm and a width of 2.5 mm were prepared by melt spinning at the wheel speed of 30 ms1. The glassy structure was confirmed by X-ray diffraction (XRD) using Cu Ka radiation and differential scanning calorimetry (DSC) at a heating rate of 30 K min1, respectively. Corrosion behavior of the glassy alloy was investigated by electrochemical polarization and weight loss measurements, respectively. Prior to electrochemical polarization and weight loss measurements, the specimens were degreased in acetone, washed in distilled water and dried in air. Electrolytes were NaCl and HCl aqueous solutions whose concentrations were 0.005, 0.01, 0.5, and 1 M, respectively, which were prepared from reagent grade chemicals and deionized water. Electrochemical measurements were conducted using a three-electrode cell with a platinum foil as a counter electrode. The reference electrode was a standard saturated calomel electrode (SCE). Potentiodynamic polarization curves were measured with an IM6ex instrument at a potential sweep rate of 0.05 mV s1 from 0.6 V to 1.0 V. The cell was open to air at room temperature and measurement started after the immersion of the samples for 20 min so that the open-circuit potentials of the samples became almost stable. The working electrode was exposed only to an area of 0.05e0.08 cm2 while the rest of the specimen was embedded in a thermoplastic resin to provide electrical isolation. All potentials given in this article are referred to the SCE electrode. The corrosion rates were estimated from the weight loss after immersion at room temperature for 30 days. The surfaces of the samples before and after corrosion were observed with a SIRION scanning electron microscope (SEM) and compositional analysis was performed using electron dispersive spectroscopy (EDS).

Fig. 1. X-ray diffraction pattern for Cu55Zr35Ti10 glassy alloy: the inset displays DSC curve of Cu55Zr35Ti10 glassy alloy at heating rate of 30 K min1.

0.01 M chloride-containing solutions. When the chloride concentration exceeds 0.5 M, a sharp decrease of the current density similar to the passivation appears at the potential of 579 mV vs. SCE in 0.5 M HCl, 484 mV vs. SCE in 1 M HCl, 748 mV vs. SCE in 0.5 M

a

3. Results The glassy structure of Cu55Zr35Ti10 alloy ribbon was confirmed by XRD, the result of which is shown in Fig. 1. The diffraction pattern exhibits the characteristic broad peak for a glassy structure without any distinct crystalline peaks within the sensitivity limit of XRD measurement. In order to further examine its glassy structure, the DSC was conducted. The inset in Fig. 1 displays the DSC curve in the mode of isochronal heating with a rate of 30 K min1 for the melt-spun ribbon. An endothermic signal associated with the glass transition and exothermic signals due to crystallization reactions are observed. The glass transition temperature Tg and onset crystallization temperature Tx are 675.3 and 729.3 K, respectively. Thus the undercooled liquid region (defined as the interval between Tg and Tx) is 54.0 K. It also manifests four crystallization peaks, indicating its complex crystallization procedure. Therefore, the XRD and DSC results indicate the glassy structure of the melt-spun Cu55Zr35Ti10 ribbon. The corrosion behavior of the Cu55Zr35Ti10 metallic glass ribbon was examined by potentiodynamic polarization measurement. Fig. 2 shows the anodic and cathodic polarization curves in NaCl and HCl aqueous solutions open to air at room temperature, respectively. In both cases, the Cu55Zr35Ti10 metallic glass exhibits the active dissolution state in the whole anodic region in 0.005 and

b

Fig. 2. The polarization curves in aqueous solutions containing different chloride concentration open to air at room temperature: (a) for HCl, and (b) for NaCl.

940

A.H. Cai et al. / Materials Chemistry and Physics 134 (2012) 938e944

NaCl, and 612 mV vs. SCE in 1 M NaCl, respectively. However, the large current density indicates that the sharp current decrease is different from the passivation. In addition, the potential for the sharp current decrease is larger in 0.5 M chloride-ion solution than in 1 M chloride-ion solution, and larger in NaCl than in HCl. On the other hand, the values of the corrosion current density icorr were determined by graphical extrapolation from the polarization curves. The intersection point of the vertical line corresponding to the corrosion potential Ecorr, with the tangents on the anodic and cathodic branches was determined. The Ecorr and icorr values in all solutions are listed in Table 1. The relationships between the chloride concentration, and Ecorr and icorr are also plotted in Fig. 3. As shown in Fig. 3(a), the corrosion current density slightly decreases when the chloride concentration is less than 0.01 M and then increases with increasing chloride concentration in HCl solution, while decreases when the chloride concentration exceeds 0. M in NaCl solution. In addition, the corrosion current densities are less in NaCl solutions than in corresponding HCl solutions, indicating that the corrosion resistance of this metallic glass is better for the former than for the latter. As shown in Fig. 3(b), the corrosion potential decreases with increasing chloride concentration in HCl and NaCl solutions. In addition, the corrosion potentials are larger in NaCl solutions than in corresponding HCl solutions except for in 0.01 M chloride-containing solution. It indicates that the stability of this metallic glass is better for the former than for the latter except for in 0.01 M chloride-containing solution. In addition, the corrosion rate Rcorr of the glassy alloy was determined by the mass loss after the immersion test using following expression [25]:

Rcorr ¼

8:76  103  W At r

a

b

(1)

where W is the weight loss (g), A is the area of exposure (cm2), t is the exposure time (h), and r is the density (gcm3). The corrosion rate values in all solutions are listed in Table 1.The relationship between the corrosion rate and chloride concentration is plotted in Fig. 4. As shown in Fig. 4, the corrosion rate increases with increasing chloride concentration in both HCl and NaCl solutions. It indicates that the corrosion resistance decreases with increasing chloride concentration. In addition, the corrosion rates are less in NaCl solutions than in corresponding HCl solutions, indicating that the corrosion resistance of this metallic glass is better for the former than for the latter. Furthermore, SEM was employed to investigate the morphologies of the samples after polarization tests. Figs. 5 and 6 show the SEM micrographs of the surfaces corroded in HCl and NaCl solutions, respectively. The pits and corrosion products can be clearly observed on the surfaces corroded in the chloride-containing solutions, indicating the occurrence of the pitting and general corrosions in all chloride-containing solutions. As shown in

Fig. 3. The relationships between the chloride concentration and corrosion current density icorr (a) and corrosion potential Ecorr (b), respectively.

Fig. 5(a)e(c), a loose structure appears on the surfaces corroded in 0.005e0.5 M HCl solutions. In addition, a rough dimple-like structure displays on the surface corroded in 0.005 M HCl, it would be due to the heavy attack of the corrosion on the corrosion

Table 1 The corrosion potential Ecorr, corrosion current density icorr, and corrosion rate of Cu55Zr35Ti10 glassy alloy. Type of solution

Concentration M

Ecorr, mV vs. SCE

icorr, A m2

RCorr, cm y1

HCl

0.005 0.01 0.5 1

18.87  0.10 10.22  0.10 119.60  0.10 322.90  0.10

2.40  0.10 1.12  0.10 1.97  0.03 11.69  0.50

3.32  0.51  103 8.25  0.51  103 3.42  0.51  102 7.02  0.51  102

NaCl

0.005 0.01 0.5 1

164.86  0.10 21.08  0.10 58.03  0.10 87.62  0.10

0.27  0.30 0.23  0.30 1.76  0.10 0.64  0.10

2.58  0.45  104 7.65  0.45  104 3.76  0.45  103 7.92  0.45  103

Fig. 4. The relationships between the corrosion rate and the chloride concentration in HCl and NaCl, respectively.

A.H. Cai et al. / Materials Chemistry and Physics 134 (2012) 938e944

941

Fig. 5. SEM micrographs of the corroded surfaces in different HCl concentrations, (a) 0.005 M, (b) 0.01 M, (c) 0.5 M, (d) 1 M, respectively: the insets are enlarged views.

products resulting in the separation of the products from the pits, as shown in Fig. 5(a). However, many grooves and corrosion products appear on the surface corroded in 1 M HCl solution, as shown in Fig. 5(d), indicating that the corrosion resistance of this metallic glass is worse in 1 M HCl than in other HCl solutions. As shown in Fig. 6(a) and (b) the size of the pits is larger in 0.01 M NaCl than in 0.005 M NaCl, but the number of the pits is more in the former than in the latter. In addition, the pits in 0.005 M NaCl do not fully develop, as shown in Fig. 6(a). It indicates that the corrosion resistance of this metallic glass is almost the same in two solutions. However, a loose structure appears on the surface corroded in 0.5 M NaCl, as shown in Fig. 6(c). As shown in Fig. 6(d), the loose structure on the surface corroded in 1 M NaCl cannot be observed, but many large pits appear. On the other hand, comparing Fig. 5(a) and Fig. 6(a), large-area exfoliation corrosion occurs, the corrosion products are broken off the pits, and some small pits can be also observed in Fig. 5(a). However, the depth of the pits is largely lower in 0.005 M NaCl than in 0.005 M HCl, although the magnitude of the pits is more in the former than in the latter. As shown in Fig. 5(b), one can observe large-area loose structure and the deep pits due to the separation of the corrosion products from the surface corroded in 0.01 M HCl. However, the corrosion products remain in some pits although many smooth pits can be observed in Fig. 6(b). As shown in Fig. 5(c) and Fig. 6(c), although many pits and loose structures can be observed in both cases, the region of smooth region is larger in 0.5 M NaCl than in 0.5 M HCl, indicating that the corrosion resistance of the studied metallic glass is better in the former than in the latter. Compared Fig. 5(d) and Fig. 6(d), although the surfaces are composed of the deep pits in both cases, the surface corroded in 1 M HCl is covered by a filiform corrosion product (Fig. 5(d)) and some corrosion products remain in the pits (Fig. 6(d)). These results

indicate that the corrosion resistance of the studied metallic glass is better in 1 M NaCl than in 1 M HCl. In order to further substantiate the difference of the corrosion resistance in HCl and NaCl solutions, the immersion tests are also conducted. The corroded surface morphologies are investigated by SEM. Fig. 7 presents the SEM surface morphologies corroded in 1 M HCl and 1 M NaCl. Many pits and corrosion products can be obviously observed on the corroded surface (not shown in here) in both cases, but the size and the depth of the pits in 1 M HCl is larger than those in 1 M NaCl (see Fig. 7). It indicates that the corrosion resistance of this metallic glass is worse in the former than in the latter. 4. Discussion As shown in Figs. 5e7, the pits display on surfaces corroded in all solutions, indicating that the studied metallic glass is susceptible to the pitting corrosion in the chloride-containing solution. The reasons would be as follows. Ideally, metallic glasses are regarded as being physically and chemically homogeneous, free from secondary phases or inclusions which should diminish or prevent the occurrence of galvanic or localized corrosion [26]. However, in practice the presence of defects in cast samples cannot be completely avoided, at least in commercial production [27]. Not surprisingly, several studies show that some metallic glasses have high pitting susceptibility and pits are initiated at the interface between such defects and the surrounding matrix [5,10,28]. The surface morphology of the melt-spun glassy ribbon is illustrated in Fig. 8. There are three regions including the relative flat region, the troughs, and the pores. Thus these defects could be energetically favorable sites where ions would be absorbed when corrosion samples were immersed into the electrolyte and the galvanic couples would be formed among these defects and between these

942

A.H. Cai et al. / Materials Chemistry and Physics 134 (2012) 938e944

Fig. 6. SEM micrographs of the corroded surfaces in different HCl concentrations, (a) 0.005 M, (b) 0.01 M, (c) 0.5 M, (d) 1 M, respectively: the insets are enlarged views.

defects and the surrounding matrix, resulting in the pitting corrosion. The other reason would be the selective dissolution of Ti and Zr, and/or simultaneous dissolution of Cu, Ti and Zr. It is wellknown that Cu is nobler than Zr and Ti, i.e. the standard equilibrium electrode potentials for the Zr/Zr4þ, Ti/Ti2þ and Cu/Cu2þcouples are 1.529 VSHE, 1.628 VSHE and 0.337 VSHE, respectively. This large electrochemical potential difference between Cu and Zr (Ti) can provide a sufficient potential window for a selective dissolution of Zr (Ti) in CueZreTi metallic glass. At surface defects where chloride ions were possibly absorbed, a preferential dissolution of less noble elements, such as Zr and Ti, resulted in the initiation of the pits. The nobler element Cu was left on the surface forming a metal cover at the beginning of the polarization. With the increase of the potential, the dissolution of Cu became possible, but it would immediately redeposit into the pit, accelerating the development of the pits [10,29e31]. It is clearly seen from Fig. 2 that the potentials for the decrease of the current density are all higher than the standard equilibrium electrode potential of Cu/Cu2þ (0.337 VSHE) and the redox potential of Cu/Cl/CuCls (46 mVSCE [10]), resulting in the active dissolution of Cu and the reaction between Cu and chloride ion. The reaction is as follows. The initial reaction is between chloride ions and copper metal, resulting in the formation of porous CuCl layer, which is the least soluble of the copper chlorides, on the anodic as the reaction [32]:

Cu þ Cl 4CuCl þ e

(2)

In chloride media, Cu electro-oxidizes to cuprous chloride complexes and the rate is dependent on the chloride concentration but independent on pH [33]. The electrodissolution rate has also shown to be strongly affected by the rate of mass transfer [34,35]. The depletion of chloride ions will yield a rather well defined electrode potential corresponding a copper electrode in a solution

saturated with CuCl in the absence of excess electrolyte, especially HCl. When the surface is covered, this reaction is supplanted by another, which may be Eq. (3) [10]:

Cu4Cuþ þ e

(3)

Cuprous ions can diffuse through the porous CuCl layer until they meet with chloride ions, and then may have the reaction [10]:

Cuþ þ Cl 4CuCl

(4)

The equilibrium potential of Eq. (4) is [33]:

i h Ee ¼ 0:105  0:059 log Cl ðVSCE Þ

(5)

We investigate the composition of the loose structure in Figs. 5 and 6 by EDS (not shown in here). The EDS results indicate that the composition of the loose structure in all solutions is mainly composed of Cu and Cl, as well as minor Zr, Ti, and O. Thus the loose structure is mainly composed of the cuprous chloride complexes as well as minor oxides. Approximately estimated by Eq. (5), the equilibrium potential of Eq. (4) decreases with increasing chloride ion. Thus the anodic current density sharply decreases when the chloride concentration reaches up to 0.5 M and the potential arrives at a value. The potential for the decrease of the current density decreases with increasing chloride concentration. The corrosion potential decreases with increasing chloride concentration. Also, CuCl is known to react with excess chloride ions to form the 2 [36]. In addition, Zr and Ti soluble complexes CuCl 2 and CuCl3 elements are easy to be passivated, resulting in the formation of an oxides film on the surface. The content of oxygen in the chloridecontaining solution decreases with increasing chloride concentra

A.H. Cai et al. / Materials Chemistry and Physics 134 (2012) 938e944

943

Fig. 8. SEM surface morphology of the melt-spun ribbon.

Fig. 7. SEM micrographs of the corroded surfaces immersed in 1 M HCl (a) and 1 M NaCl (b), respectively.

tion [37], resulting in the decrease of the protective Zr and Ti oxides with increasing chloride concentration. These results would result in the decrease of the corrosion resistance (as shown in Table 1 and Fig. 4) and corrosion potential (as shown in Table 1 and Fig. 3(b)) with increasing chloride concentration. At the same time, the above-mentioned reactions are accompanied with the hydrogen evolution when the potential is higher than the hydrogen evolution potential. In fact, we observe large number of the bubbles on the surface during electrochemical polarization tests. The hydrogen bubble erodes the surface, resulting in the exfoliation corrosion in Figs. 5 and 6. In addition, the exfoliation corrosion would be more and more serious with increasing chloride concentration due to the formation of the 2 soluble complexes CuCl 2 and CuCl3 (see Figs. 5 and 6). We also observe that the colorful products fall off from the surface when the potential arrives at the above-mentioned potential (in Fig. 2) in 0.5 and 1 M chloride-containing solutions (not shown here). As shown in Table 1 and Figs. 3 and 4, the corrosion current density and the corrosion rate are smaller in NaCl than in HCl, while inversely for the corrosion potential except for in 0.01 M chloride-

containing solutions. The different corrosion behavior between HCl and NaCl would be related with the factors as follows. Firstly, it is related to the different distribution of the alloying elements on the surface in HCl and NaCl solutions because the corrosion of metallic glass is strongly influenced by the alloying elements [24]. Asami et al [16] investigated the concentration of the alloying elements of Cu60Zr30Ti10 metallic glass on the surface immersed in 1 N HCl and 3% NaCl, respectively. They found that the cationic fraction of the surface film after immersion in 3% NaCl was almost the same as the alloy composition, but in the surface film formed in 1 N HCl solution, its cationic fraction of Cu was more than the alloy composition. Qin et al [38] also found similar phenomenon for a Cu-based metallic glass after immersion in 1 N HCl and 3% NaCl, respectively. When the specimen was immersed in 1 N HCl, it was found that the Cu content in the surface film significantly decreased, and further reduced after immersion in the 3% NaCl. At the same time, Ti and Hf were largely concentrated on the alloy surface immersed in the 1 N HCl. The content of Ti increased more after immersion in the 3% NaCl, while that of Hf slightly decreased. Secondly, the different elements play different roles in corrosive media. It is known that Ti and Zr are strong passive elements in aggressive environments [39]. Cu is not corrosion resistant in acid, especially chloride-ions-containing solution [23]. The relatively high corrosion resistance of the glassy Ti41.5Zr2.5Hf5Cu42.5Ni7.5Si1 alloy is considered to be due to the synergistic effect of Ti, Zr and Hf, in spite of the high content of copper [39]. Thirdly, the stability of the oxide of the alloying element is different from each other. For example, Zr- and Ti-oxides are more stable chemically and denser structurally than Cu-oxides [21]. Good corrosion resistances of Zr69.5Cu12Ni11Al7.5 and Cu50Zr40Al5Nb5 are resulted from the formation of an amorphous ZrO2 layer [40] and a Zr (Nb)-rich protective film [41], respectively. Lastly, the role of the chloride ion is changed by the existence of the hydrogen ion. Jiang et al [42] have investigated the corrosion behavior of CueZreAgeAl-Ti metallic glass system in 1 N H2SO4 and 1 N H2SO4 þ 0.1 N NaCl. They found that the metallic glasses exhibited excellent corrosion behavior in pure hydrogen-ioncontaining solution. In contrast, in both hydrogen- and chlorideion-containing solutions, the alloys revealed poor corrosion resistance even though the concentration of chloride ion was very low. Many researchers have proved that there would exist following

944

A.H. Cai et al. / Materials Chemistry and Physics 134 (2012) 938e944

Table 2 Content of alloying elements of surface film immersed in 1 M NaCl and 1 M HCl, respectively

1 M HCl 1 M NaCl

O (at %)

Zr (at %)

Cl (at. %)

Ti (at %)

Cu (at %)

25.72 40.63

23.48 28.67

3.85 1.58

6.85 8.56

40.10 20.56

reactions for Cu in the pure chloride-ion solution or both hydrogenand chloride-ion solution [20,21,33,43,44].

CuCl þ Cl 4CuCl 2

(6)

2CuCl þ H2 O4Cu2 O þ 2Cl þ 2Hþ

(7)

Cu2 O þ 2H2 O þ O2 42CuO þ 2H2 O2

(8)

Cu2 O þ H2 O2 42CuO þ 2H2 O

(9)

The corrosion potential decreases with increasing chloride concentration in HCl and NaCl solutions. The corrosion potential is larger in NaCl than in HCl, except for in 0.01 M chloride-containing solution. The sharp decrease of the current density appears when the chloride concentration exceeds 0.5 M in HCl and NaCl solutions. The potential for the decrease of the current density decreases with increasing chloride concentration. The corrosion rate increases with increasing chloride concentration in HCl and NaCl solutions. The corrosion rate is higher in the former than in the latter. Acknowledgements

2Cu þ O2 þ 4Hþ 42Cu2þ þ 2H2 O

(10)

Cu2þ þ 2Cl 4CuCl2

(11)

The composition of the surface film immersed in 1 M HCl and NaCl solutions was investigated by EDS, respectively. The results are listed in Table 2. It is clearly seen from Table 2 that the content of Zr and Ti on the surface immersed in HCl is lower than that in NaCl, while inversely for Cu. In addition, Zr- and Ti-oxides are more stable chemically and denser structurally than Cu-oxides [21]. It indicates that the stability and corrosion resistance of this metallic glass are better in NaCl than in HCl, resulting in higher corrosion potential in the former than in the latter (as shown in Fig. 3(b) and Table 1). It also indicates that the corrosion resistance of this metallic glass in NaCl is better than in HCl. In addition, the content of CuCl in the film formed in HCl is more than that in NaCl. The chloride ion in the solution can react with Cu to form the slightly soluble porous CuCl film and soluble CuCl 2 complex ion according to Eq. (6). It indicates that the corrosion resistance of the metallic glass is worse in HCl than in NaCl. Although the chloride ion simultaneously exists in HCl and NaCl solutions, the HCl solution contains the hydrogen ion. In both hydrogen- and chloride-ion-containing solution (HCl solution), the porous cupric chloride preferentially forms according to Eq. (7). In addition, even the existed Cu2O in the surface film is transformed into CuCl, resulting in the prohibition of the formation of CuO according to Eqs. (8) and (9). In addition, Cu is known to form a protective passive layer of CuO in chloride solution [45]. Moreover, in both hydrogen ion and oxygen, the soluble CuCl2 is formed according to Eqs. (10) and (11). These results would result in worse corrosion resistance in HCl than in NaCl. 5. Conclusions The influence of the chloride concentration on corrosion behavior of Cu55Zr35Ti10 metallic glass in HCl and NaCl solutions was investigated by electrochemical polarization and immersion measurements, scanning electron microscope, and electron dispersive spectroscopy. The results are as follows. The corrosion current density slightly decreases in HCl and NaCl solutions when the chloride concentration is less than 0.01 M, and then increases with increasing chloride concentration in HCl. However, the corrosion current density decreases when the chloride concentration exceeds 0.5 M in NaCl. In addition, the corrosion current density is larger in HCl solution than in corresponding NaCl solution.

This work was financially supported by the National Natural Science Foundation (Grant No. 50874045), Postdoctoral Science Foundation of China (Grant Nos. 20080431021 and 200902472), and Scientific Research Fund of the Hunan Provincial Education Department (Grant No. 10A044). References [1] W.H. Wang, C. Dong, C.H. Shek, Mater. Sci. Eng. R 44 (2004) 45. [2] M.L. Morrison, R.A. Buchanan, R.V. Leon, C.T. Liu, B.A. Green, P.K. Liaw, J.A. Horton, J. Biomed. Mater. Res. A 74A (2005) 430. [3] S. Hiromoto, A.P. Tsai, M. Sumita, T. Hanawa, Corros. Sci. 42 (2000) 1651. [4] A.P. Wang, X.C. Chang, W.L. Hou, J.Q. Wang, Corros. Sci. 49 (2007) 2628. [5] P.F. Gostin, A. Gebert, L. Schultz, Corros. Sci. 52 (2010) 273. [6] Z.M. Zhang, J. Zhang, X.C. Chang, W.L. Hou, J.Q. Wang, Corros. Sci. 52 (2010) 1342. [7] R.V. Subba Rao, U. Wolff, S. Baunack, J. Eckert, A. Gebert, Corros. Sci. 45 (2003) 817. [8] S.J. Pang, T. Zhang, K. Asami, A. Inoue, Corros. Sci. 44 (2002) 1847. [9] A. Gebert, P.F. Gostin, L. Schultz, Corros. Sci. 52 (2010) 1711. [10] H.B. Lu, Y. Li, F.H. Wang, Corros. Sci. 48 (2006) 2106. [11] H.B. Lu, L.C. Zhang, A. Gebert, L. Schultz, J. Alloys Compd. 462 (2008) 60. [12] A. Inoue, Acta Mater 48 (2000) 279. [13] A. Inoue, W. Zhang, T. Zhang, K. Kurosaka, Acta Mater 49 (2001) 2645. [14] Á, L. Domokos, T. Katona, T. Martinek, G. Mulas, G. Cocco, I. Bertóti, J. Szépvölgyi, Mater. Sci. Eng. A226e228 (1997) 1074 Molnár. [15] A. Inoue, W. Zhang, T. Zhang, K. Kurosaka, Mater. Trans. JIM 42 (2001) 1149. [16] K. Asami, C. Qin, T. Zhang, A. Inoue, Mater. Sci. Eng. A 375e377 (2004) 235. [17] C. Qin, K. Asami, T. Zhang, W. Zhang, A. Inoue, Mater. Trans. JIM 44 (2003) 749. [18] F.R. Niessen, Cohesion in Metals, Elsevier Science, Amsterdam, 1988. [19] T. Yamamoto, C. Qin, T. Zhang, K. Asami, A. Inoue, Mater. Trans. JIM 43 (2002) 1147. [20] B. Liu, L. Liu, Intermetallics 15 (2007) 679. [21] B. Liu, L. Liu, Mater. Sci. Eng. A 415 (2006) 286. [22] C.L. Qiu, L. Liu, M. Sun, J. Biomed Mater. Res. A75 (2005) 950. [23] M.K. Tam, S.J. Pang, C.H. Shek, J. Non-Cryst. Solids 353 (2007) 3596. [24] D. Zander, B. Heisterkamp, I. Gallino, J. Alloys Compd. 434e435 (2007) 234. [25] N.D. Nam, M.J. Kim, Y.W. Jang, J.G. Kim, Corros. Sci. 52 (2010) 14. [26] K. Hashimoto, Chemical properties, in: F.E. Luborsky (Ed.), Amorphous Metallic Alloys, Butterworth, London, 1983. [27] Y.H. Liu, G. Wang, R.J. Wang, D.Q. Zhao, M.X. Pan, W.H. Wang, Science 315 (2007) 1385. [28] A. Gebert, U. Kuehn, S. Baunack, N. Mattern, L. Schultz, Mater. Sci. Eng. A415 (2006) 242. [29] G. Kear, B.D. Barker, F.C. Walsh, Corros. Sci. 46 (2004) 109. [30] M. Chmielová, J. Seidlerová, Z. Weiss, Corros. Sci. 45 (2003) 883. [31] A.V. Vvedenskii, S.N. Grushevskaya, Corros. Sci. 45 (2003) 2391. [32] L. Stephenson, J.H. Bartlett, J. Electrochem. Soc. 101 (1954) 571. [33] H.P. Lee, K. Nobe, J. Electrochem. Soc. 132 (1985) 1031. [34] R.S. Copper, J.H. Bartlett, J. Electrochem. Soc. 105 (1958) 109. [35] A.L. Bacarella, J.C. Griess Jr., J. Electrochem. Soc. 120 (1973) 459. [36] H.P. Lee, K. Nobe, J. Electrochem. Soc. 133 (1986) 2035. [37] http://chem188.cn/Article/ShowArticle.asp?ArticleID¼2317; http://wenku. baidu.com/view/af09274ac850ad02de8041ad.html. [38] C.L. Qin, W. Zhang, K. Asami, H. Kimura, X.M. Wang, A. Inoue, Acta Mater. 54 (2006) 3713. [39] S.J. Pang, C.H. Shek, C. Ma, A. Inoue, T. Zhang, Mater. Sci. Eng. A449e451 (2007) 557. [40] D. Zander, U. KÖster, Mater. Sci. Eng. A375-377 (2004) 53. [41] C. Qin, W. Zhang, H. Kimura, K. Asami, A. Inoue, Mater. Trans. JIM 45 (2004) 1958. [42] X.P. Nie, X.H. Yang, J.Z. Jiang, J. Alloys Compd. 481 (2009) 498. [43] L. Huang, D.C. Qiao, B.A. Green, P.K. Liaw, J.F. Wang, S.J. Pang, T. Zhang, Intermetallics 17 (2009) 195. [44] L. Huang, G.Y. Wang, D.C. Qiao, P.K. Liaw, S.J. Pang, J.F. Wang, T. Zhang, J. Alloys Compd. 504S (2010) 159. [45] E. Kunze (Ed.), Koeeosion und Korrosionsschutz, Wiley-VCH, Weinheim, Germany, 2001.