Observations of anomalies in the anodic behaviour of microcrystalline aluminium

Observations of anomalies in the anodic behaviour of microcrystalline aluminium

Corrosion Science 52 (2010) 2612–2615 Contents lists available at ScienceDirect Corrosion Science journal homepage: www.elsevier.com/locate/corsci ...

328KB Sizes 1 Downloads 23 Views

Corrosion Science 52 (2010) 2612–2615

Contents lists available at ScienceDirect

Corrosion Science journal homepage: www.elsevier.com/locate/corsci

Observations of anomalies in the anodic behaviour of microcrystalline aluminium Bo Zhang, Ying Li *, Fuhui Wang State Key Laboratory for Corrosion and Protection, Institute of Metal Research, Chinese Academy of Sciences, Wencui Rd. 62, Shenyang 110016, China

a r t i c l e

i n f o

Article history: Received 13 January 2010 Accepted 12 April 2010 Available online 4 May 2010 Keywords: A. Aluminium A. Sputtered films C. Passive films

a b s t r a c t The anodic behaviour of sputtered microcrystalline Al (mc-Al) was investigated in neutral Na2SO4 electrolyte under varied conditions. Our results revealed that Cl addition led to a reduction in the anodic current density, which we considered unusual. Mott–Schottky analysis showed that Cl introduction altered the semiconducting property of the passive film from n-type to p-type, implying that the p-type film can possessed a relative higher stability. Immersion of mc-Al in other electrolytes yielded films with n-type, p-type and positive p–n junction structure. The results also indicate that the p-type film was most stable and the positive p–n junction film least stable. Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction Chloride ions have been known as aggressive anions for several decades; they make passive metals suffer pitting corrosion. By comparing the potentiodynamic curves of passive metals in chloride-containing electrolytes with those in chloride-free electrolytes, two distinctive characteristics are obvious: a well-defined inflexion in anodic branch where the anodic current density increased sharply with potential, corresponding to the occurrence of pitting corrosion as well as increased anodic passive current density. Usually, in a chloride-containing solution, Cl is adsorbed on the oxide surface, penetrates the passive film and subsequently reacts with the film to form a very soluble Cl-containing complex on the film surface [1–4]. Consequently, the surface of the passive film is dissolved at a higher rate, which is reflected in a higher anodic current density. In the present paper, reports an unusual observation of a somewhat contradictory effect of Cl ions on the anodic polarization behaviour of mc-Al, wherein the addition of Cl ions to an aqueous electrolyte solution resulted in a lower anodic current density.

2. Experimental A planar magnetron sputter machine (SBH 5115D) was used to produce the mc-Al coating on a substrate of quartz glass. The target was a 380 mm  128 mm sheet cut from commercial grade pure Al (99.9%) ingot. The sputtering parameters were as follows: argon pressure 0.25 Pa; power 1.2 kW; substrate temperature 100 °C. The total sputtering time was 4 h. Prior to fabricating the coatings, * Corresponding author. Tel.: +86 24 2392 5323; fax: +86 24 2389 3624. E-mail address: [email protected] (Y. Li). 0010-938X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.corsci.2010.04.009

the target was sputtered for 10 min to remove the surface contaminants. The commercially pure Al (99.9 wt.%) ingot for the sputtering target was also employed as a reference specimen in the experiments (bulk Al). It was ground with 2000 grit SiC paper, washed in distilled water, and degreased with acetone. The mc-Al was only washed without grinding. All the specimens were kept in a desiccator to avoid further corrosion in air. The electrolytes used in this study were prepared using analytical grade reagents and distilled water. Electrochemical tests were performed in aerated solutions at 30 ± 2 °C. The electrochemical cell contained 300 ml of the solution. For all experiments, a three-electrode cell was used, with an Ag/AgCl, KCl (saturated) reference electrode and a platinum counter electrode; the reference was immersed in saturated KCl electrolyte and was connected by a salt bridge with the measurement solution; throughout this work all potentials are referred to that reference electrode. Potentiodynamic polarization curves were measured by a PAR 2273 potentiostat, with a scan rate of 1 mV/s from OCP to anodic direction. The semiconducting property of the passive films was studied using capacitance measurements according to the Mott–Schottky theory. Capacitance measurements were carried out within a suitable potential range at a frequency of 100 Hz and perturbating AC amplitude of 5 mV. The measured potential range was selected according to the passive range on potentiodynamic polarization curves. An important point to be considered in the Mott–Schottky analysis is that the measured capacitance value is influenced by frequency. To get an optimized frequency, a series of capacitance values were measured as a function of frequency. There existed a frequency range in which the capacitance did not vary with frequency. A frequency value included in that range (for example, the 100 Hz chosen in our case) was an optimized one for the Mott–Schottky test.

2613

B. Zhang et al. / Corrosion Science 52 (2010) 2612–2615

2.0

All the electrochemical measurements were repeated at least three times with good reproducibility. The surface and cross-sectional morphology and the grain size of mc-Al have been examined in our previous work [5]. The grain size was about 400 nm.

The anodic polarization curves of the mc-Al in the different test solutions are shown in Fig. 1, and those of the bulk Al are shown in Fig. 2 for comparison. On one hand, for bulk Al in Na2SO4, its anodic passive current density greatly increased in the presence of Cl ions (Fig. 2, curves a and b), which is in agreement with the generally accepted consensus. Song et al. [4] proposed that in addition to the higher corrosivity of chloride-containing solutions, a less stable passive film caused by incorporation of Cl ions also contributes to the higher current density. Such Cl ion incorporation in passive films on aluminium has been illustrated by means of XPS [6]. We have carried out a simple test to further confirm the detrimental effect of Cl ions on passive film stability. First, a bulk Al specimen was immersed in 0.1 mol/L NaCl solution for 1 h, washed with distilled water, then immediately put into 0.5 mol/L Na2SO4 to obtain the anodic polarization curve (shown in Fig. 2, curve c). It is clear that the anodic current density was greater than that for a specimen which had been immersed only in 0.5 mol/L Na2SO4, which indicates that the Cl ions renders the passive film less stable, since the polarization measurements were run in identical solutions. Interestingly, similar tests on the mc-Al in 0.5 mol/L Na2SO4 without and with Cl ions yielded contradictory results, i.e., that the anodic passive current density was lower in chloride-containing solution than that in chloride-free solution (Fig. 1, curves a and b). In other words, it seems that Cl ions made the passive film more stable, which is rather unusual. A similar phenomenon was also observed in our previous work in fluoride and chloride-containing systems [7]. The resistance of a metal to corrosive attack has been found to depend upon the solid-state characteristics of the oxide film [8,9]. Oxide or passive films formed on metals usually exhibit semiconducting properties and capacitance measurements have been successfully used for characterization of passive films formed on metals, the so-called Mott–Schottky approach [10–17]. The effect of the applied potential E on capacitance values is described by the Mott–Schottky equation as follows [18]:

C 2 ¼ ð2=ee0 qN d ÞðE  Efb  kT=qÞ

sat'd

c

1.0

E V/Ag/AgCl, KCl

3. Results and discussion

a

Bulk Al

1.5

0.5 0.0

b

-0.5 -1.0 -8

10

-7

10

-6

10

-5

-4

10

-3

10

10

-2

-1

10

0

10

10

2

i A/cm

Fig. 2. The anodic potentiodynamic polarization curves of bulk Al in (a) 0.5 mol/L Na2SO4, (b) 0.5 mol/L Na2SO4 + 0.1 mol/L NaCl and (c) 0.5 mol/L Na2SO4 after immersion in neutral 0.1 mol/L NaCl for 1 h.

where Nd is the carrier concentration (donor or acceptor), e dielectric constant of the semiconductor, e0 vacuum permittivity, q elementary charge (e for electrons and +e for holes), k Boltzmann constant, T temperature and Efb the flatband potential. This equation predicts a linear plot of C2 versus E, with slope equal to (2/ee0qNd). The semiconductor type can be distinguished according to the slope (positive slope indicates n-type, negative slope indicates p-type). Mott–Schottky plots given in Fig. 3 show the passive film on bulk Al to be an n-type semiconductor (with positive slope) both in the absence and presence of Cl in the solution. On the other hand, Fig. 4 reveals that the passive film on mc-Al is an n-type semiconductor (positive slope) in chloride-free solution and a p-type semiconductor (negative slope) in chloride-containing solution. Similar results were also obtained in our previous work [5] in acidic electrolytes system. This behaviour gives us a clue that a ptype semiconductor film could be more stable than an n-type one, at least in the Cl ion containing solution. Based on this hypothesis, it could be suggested that the addition of chloride ions provides two competing and conflicting effects on the corrosion process of the mc-Al: (1) it makes the solution more corrosive and (2) it changes the film from an n-type semiconductor to a more stable

ð1Þ

E vs. OCP V/Ag/AgCl, KCl sat'd 2.0

0.00

0.04

0.08

0.12

0.16

b 9

1.8x10

a

-2

1.0

4

cm F

0.5

9

1.6x10

-2

0.0

C

sat'd

1.5

E V/Ag/AgCl, KCl

9

2.0x10

mc-Al

-0.5

9

1.4x10

b a

-1.0 -8

10

-7

10

9

1.2x10

c -6

10

-5

10

-4

-3

10

10

-2

10

-1

10

0

10

2

i A/cm

0.2

0.3

0.4

0.5

0.6

0.7

E vs. OCP V/Ag/AgCl, KCl sat'd Fig. 1. The anodic potentiodynamic polarization curves of mc-Al in (a) 0.5 mol/L Na2SO4, (b) 0.5 mol/L Na2SO4 + 0.1 mol/L NaCl and (c) 0.5 mol/L Na2SO4 after immersion in neutral 0.1 mol/L NaCl for 1 h.

Fig. 3. The Mott–Schottky curve of bulk Al measured in (a) 0.5 mol/L Na2SO4, (b) 0.5 mol/L Na2SO4 + 0.1 mol/L NaCl.

2614

B. Zhang et al. / Corrosion Science 52 (2010) 2612–2615 9

1.3x10

7

3.0x10

a 9

1.3x10

9

7

1.2x10

2.5x10

b

-2 4 -2

4

1.1x10

cm F

cm F

7

2.0x10

9

1.1x10

-2

C

9

1.1x10

9

C

7

1.5x10

-2

c 9

1.0x10

8

7

9.5x10

1.0x10

8

9.0x10

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

E vs. OCP V/Ag/AgCl, KCl sat'd Fig. 4. The Mott–Schottky curve of mc-Al measured in (a) 0.5 mol/L Na2SO4, (b) 0.5 mol/L Na2SO4 + 0.1 mol/L NaCl, (c) 0.5 mol/L Na2SO4 after immersion in 0.1 mol/ L NaCl solution for 1 h.

p-type one. Our results point towards a more predominant effect of (2). Again, in order to provide further support for the statement that the p-type semiconductor film is more stable than the n-type film, the anodic polarization curve was recorded in 0.5 mol/L Na2SO4 after immersion in to 0.1 mol/L NaCl for 1 h. The measured passive current density should be indicative of the dissolution rate of the p-type semiconductor film (which formed in chloride-containing solution). We anticipate that the measured anodic passive current density should be lowest in 0.5 mol/L Na2SO4. However, as depicted in Fig. 1c, the passive current density was larger than that in NaCl solution, even larger than the passive current density of the relatively less stable n-type semiconductor film in the corresponding solution, which seems to contradict our conclusion on bulk Al above. A Mott–Schottky curve was also plotted for the mc-Al which was immersed in 0.1 mol/L NaCl for 1 h and then in 0.5 mol/L Na2SO4, the same conditions as that of the polarization curve measurement. The result is shown in Fig. 4, curve c. It can be seen that the passive film show a p-type character at low potential and an n-type character at high potential, which implies a bi-layer structure for the passive film. The formation mechanism of that bi-layer film is easy to understand. Since the mc-Al was immersed in NaCl solution for 1 h, a steady state was obtained and a p-type semiconductor film formed. When it was moved into the chloride-free Na2SO4 solution and an anodic potential was applied, the p-type semiconductor film formed in NaCl solution will have been further thickened. Since the former passive film is a p-type semiconductor, cation vacancies, as major carriers, are excessive and facilitate metal cations migration from the metal/film (m/f) interface to the film/solution (f/s) interface and then form a new layer of oxide at the p-type film/solution interface (Fig. 5). Because the new layer was formed in chloride-free electrolyte, it will be an n-type semiconductor in which oxygen vacancies are excessive as major carriers and facili-

tate oxygen anions migration. As a result, in the further thickening of the passive film, metal cations flowing from the metal/p-type film (m/fp) interface to the p-type/n-type film (fp/fn) interface, and oxygen anions flowing from the fn/s interface to the fp/fn interface, produced oxide at the fp/fn interface until a new steady state was reached. The bi-layer film with an inner p-type and an outer n-type semiconductor can be considered as a p–n junction. In the junction region, there exists a barrier potential to carriers. We herein attempt to suggest a possible process for formation of the barrier potential. For simplicity, the bi-layer structure is assumed to consist of an n-type semiconductor and a p-type semiconductor in close contact. Before contact, holes are the major carriers and electrons are the minor carriers in the p-region. Whereas, electrons are the major carriers and holes are the minor carriers in the n-region. When the two semiconductors come into contact, the electrons in the n-region will diffuse to the p-region, and the holes in the p-region also will diffuse to the n-region. As a result, a positive space charge consisting of ionized donors forms on the side of the n-region, and a negative space charge consisting of ionized acceptor forms on the side of the p-region. Therefore, a built-in potential comes into being in the space charge region, which points from the n-region to the p-region. The built-in potential holds back the diffusion of the electrons of the n-region and the holes of the pregion. Thus, in the junction region (the space charge region), an equilibrium state will arise when the exclusion effect of the built-in potential is equal to the diffusion effect. For electrons, the potential energy is larger in the p-region than in the n-region, that is, the junction region is a barrier potential to the electrons diffusing from the n-region to the p-region. In the same way, the junction region also is barrier potential to holes diffusing from the p-region to the n-region. Since there are few free carriers in the space charge region, it is a high-resistance region. Therefore the formation process of the barrier potential can be well understood in terms of the changes in the energy band. Fig. 6a shows the energy band when the p-region and n-region have just been brought into contact. (EF)n is the Fermi level of the n-region, which is near to the bottom of the conduction band, and (EF)p is the Fermi level of the p-region, which is near to the top of the valence band. As a result of the difference between the two Fermi levels, electrons transfer from the n-region with higher Fermi energy to the p-region. The two Fermi levels will be the same when a steady state is reached. From Fig. 6b, it can be seen that the energy band of the p-region increased by a value of (EF)n  (EF)p at steady state. Apparently, the bent energy band is a barrier potential for both electrons in the n-region and holes the in p-region. When such an electrode with a bi-layer film is polarized in the anodic direction, it is equivalent to applying a positive voltage to the p–n junction. As a result, the potential of the p-type region increases, that is, the barrier potential decreases. The exclusion effect

a (E )

F n

n p-type film

M

solution

p-type film

M

M3+

M3+

O

VM3+

VM3+

VO..

p (EF)p

n-type film

solution

b p

EF m/f

f/s

m/f

fP/fn

f/s

Fig. 5. The schemes of the formation process of the double-layer passive film on mc-Al.

n

Fig. 6. The formation of barrier potential in p–n junction.

B. Zhang et al. / Corrosion Science 52 (2010) 2612–2615

of the built-in potential is weakened and the diffusion effects of the carriers become more pronounced. Electrons diffuse from the n-region to the p-region; at the same time, holes diffuse from the p-region to the n-region, yielding a considerable positive current. The current through the junction is equal to the sum of the diffusion current of the non-balanced minor carriers (electrons and holes). So, the positive current may be larger than the current produced by the application of the same potential on a pure n-type or pure p-type semiconductor film. In another words, the conductivity of the positive p–n junction may be higher than that of a pure n-type or pure p-type semiconductor film. It is known that,

U ¼ Uf þ UH where U is the positive voltage applied to the electrode, Uf is the potential drop in the film and UH is the potential drop in the Helmholtz layer. When a positive potential is applied to the electrode, Uf of the electrode with the double-layer film is smaller compared with that of the single-layer film. Thus, the potential drop in the Helmholtz layer UH (which drives the anodic reaction) is relatively larger, resulting in a high anodic current density. This could be the reason why the measured current density (curve c) is largest contrary to our anticipation. This provides vital insight into the relationship between the stability and the semiconducting property of passive films; it seems that for pure Al, the p-type semiconductor film is most stable and the semiconductor film with a positive p– n junction structure is the least stable. Certainly, the relationship between film stability and semiconductor type is obtained qualitatively from mc-Al, which is as yet not universally acknowledged and needs further confirmation through investigation of other passive metals or alloys.

4. Summary Chloride ions change the passive film on mc-Al from an n-type semiconductor to a p-type one, thus resulting in enhanced stability. This could be responsible for the unusual observation that Cl addition lowers the anodic current density. The unusual observation and the corresponding alteration of semiconducting property of passive film give such a clue on film stability that a p-type semiconductor film is most stable, an n-type one is less stable and film with a positive p–n junction structure is least stable.

2615

Acknowledgements The project was supported by the National Natural Science Fund of China under the contract Nos. 50001013 and 50671113. The authors are also grateful to Prof. Wu Weitao, Dr. D.G. Lees and Prof. Emeka for useful suggestions and modification of English. References [1] E. McCafferty, Sequence of steps in the pitting of aluminum by chloride, Corros. Sci. 45 (2003) 1421–1438. [2] S. Ahn, H. Kwon, D.D. Macdonald, Role of chloride ion in passivity breakdown on iron and nickel, J. Electrochem. Soc. 152 (2005) B482–B490. [3] S.Y. Yu, W.E. O’Grady, D.E. Ramaker, P.M. Natishan, Chloride ingress into aluminum prior to pitting corrosion, J. Electrochem. Soc. 147 (2000) 2952– 2958. [4] Guangling Song, Chun-Nan Cao, Sheng-Hao Chen, A study on transition of iron from active into passive state, Corros. Sci. 47 (2005) 323–339. [5] B. zhang, Y. Li, F.H. Wang, Electrochemical corrosion behaviour of microcrystalline aluminium in acidic solutions, Corros. Sci. 49 (2007) 2071– 2082. [6] L.D. Atanasoska, D.M. Drazic, A.R. Despic, A. Zalar, Chloride-ion penetration into oxide-films on aluminum Auger and XPS studies, J. Electroanal. Chem. 182 (1985) 179–186. [7] B. Zhang, Y. Li, F.H. Wang, Electrochemical behaviour of microcrystalline aluminium in neutral fluoride containing solution, Corros. Sci. 51 (2009) 268– 275. [8] N. Sato, Toward a more fundamental understanding of corrosion processes, Corrosion 45 (1989) 354–368. [9] Z. Szklarska-Smialowska, Mechanism of pit nucleation by electrical breakdown of the passive film, Corros. Sci. 44 (2002) 1143–1149. [10] A.M.P. Simoes, M.G.S. Ferreira, B. Rondot, M. da Cunha Belo, Study of passive films formed on AISI 304 stainless steel by impedance measurements and photoelectrochemistry, J. Electrochem. Soc. 137 (1990) 82–87. [11] U. Stimming, Photoelectrochemical studies of passive films, Electrochim. Acta 31 (1986) 415–429. [12] P.C. Searson, R.M. Latanision, U. Stimming, Analysis of the photoelectrochemical response of the passive film on iron in neutral solutions, J. Electrochem. Soc. 135 (1988) 1358–1363. [13] A. Fattah-alhosseini, M.A. Golozar, A. Saatchi, K. Raeissi, Effect of solution concentration on semiconducting properties of passive films formed on austenitic stainless steels, Corros. Sci. 52 (2010) 205–209. [14] J. Amri, T. Souier, B. Malki, B. Baroux, Effect of the final annealing of cold rolled stainless steels sheets on the electronic properties and pit nucleation resistance of passive films, Corros. Sci. 50 (2008) 431–438. [15] S. Fujimoto, H. Tsuchiya, Semiconductor properties and protective role of passive films of iron base alloys, Corros. Sci. 49 (2007) 195–202. [16] A.M. Schmidt, D.S. Azambuja, E.M.A. Martini, Semiconductive properties of titanium anodic oxide films in Mcllvaine buffer solution, Corros. Sci. 48 (2006) 2901–2912. [17] F.J. Martin, G.T. Cheek, W.E.O. Grady, P.M. Natishan, Impedance studies of the passive film on aluminium, Corros. Sci. 47 (2005) 3187–3201. [18] S. Menezes, R. Haak, G. Hagen, M. Kendig, Photoelectrochemical characterization of corrosion inhibiting oxide-films on aluminum and its alloys, J. Electrochem. Soc. 136 (1989) 1884–1886.