Deakin Research Online This is the published version: Huang, P., Howlett, P., MacFarland, D. and Forsyth, M. 2011, Anodising AA5083 aluminium alloy using ionic liquids, in ICC 2011 : 18th International Corrosion Congress : Corrosion Control, Contributing to a Sustainable Future for All, International Corrosion Council, [Perth, W. A.], pp. 1-8. Available from Deakin Research Online: http://hdl.handle.net/10536/DRO/DU:30042242 Reproduced with the kind permission of the copyright owner. Copyright : 2011, The Authors

18th International Corrosion Congress 2011 Paper 535 - Page 1

ANODISING AA5083 ALUMINIUM ALLOY USING IONIC LIQUIDS

P. Huang1, P. Howlett1, D. MacFarlane2, M. Forsyth1 1Deakin University, Melbourne, Australia, 2Monash University, Melbourne, Australia

SUMMARY: Aluminium, as the current collector in lithium batteries, has shown reduced corrosion susceptibility in room temperature molten salts (1, 2). Moreover, previous studies have established that corrosion mitigation is achieved on magnesium alloys using ionic liquids pretreatments (3, 4). This paper investigated the anodisation of AA5083 aluminium alloy in Trihexyl(tetradecyl)phosphonium bis(trifluoromethylsulfony) ([P6,6,6,14][NTf2]) ionic liquid by applying a constant current followed by holding at the maximum potential for a period of time. Potentiodynamic polarisation results show that the treated surfaces were more corrosion resistant in 0.1 M sodium chloride solution compared with the control specimen. The anodising treatment was effective both in shifting the free corrosion potential to more noble values and in suppressing the corrosion current. Optical microscope and optical profilometry images indicated that an anodising film was deposited onto the alloy surface, which is thought to have inhibited corrosion in chloride environment. Further characterisation of the anodising film will be carried out in future work.

Keywords: Anodising film, Passivation, Aluminium Alloy, Ionic liquid, Corrosion Protection.

1. INTRODUCTION

Ionic liquids (ILs), also known as room temperature molten salts, refer to organic salts that are liquid at ambient temperature (5). Generally, they are composed of an organic cation-anion pair, which can be designed to possess a desired profile of properties. Their attractive properties include non-volatility, high ion conductivity and high thermal stability (6, 7). There is an increasing amount of research devoted to exploring the possibilities of employing ILs as safer alternative electrolytes in numerous electrochemical applications, especially in the lithium-ion batteries (8, 9).

More recently, another novel application of ILs has emerged that is corrosion mitigation of reactive metals. Uerdingen and co-workers conducted one of the first investigations on the corrosion behaviour of a range of metals, including steel, nickel alloy, copper and aluminium, in various ionic liquids (10). It has been revealed that the degree of corrosion resistance of metals varies with the specific metallic material, chemical structure of the ionic liquid, and also with the water content. Nevertheless, a new corrosion protection scheme has been established, which utilises ionic liquids to impart better corrosion resistance of reactive metals.

In a series of studies on lithium-ion batteries, it was observed that lithium metal reacts with ionic liquid electrolytes based on the bis(trifluoromethanesulfony)imide (NTf2) anion and forms desirable surface films (11). Howlett et al have characterised the solid electrolyte interface (SEI) formed on lithium electrode cycled in N-methyl-N-alkylpyrrolidinium NTf2 electrolyte (8, 9). The results suggested that it was a layered structure composed of mainly the reduction product of the NTf2 anion. Following from this, there have been extensive investigations into various aspects of passivating other reactive metals, such as magnesium alloys, with ionic liquid coatings (3, 4, 12-17).

Aluminium current collectors, used in lithium-ion batteries, usually suffer from severe corrosion and pose safety concerns due to the high potential bias during cycling. However, recent studies have shown that aluminium current collectors passivate when immersed in IL environments instead of aqueous electrolytes (1, 2). In addition, anodising is a widely performed process, in suitable aqueous electrolytes, to form corrosion protective films on aluminium alloys (18, 19). Therefore, this paper explores improving corrosion properties of the AA5083 Al-Mg alloy by forming an anodising surface film in a commercially available IL, Trihexyl(tetradecyl)phosphonium bis(trifluoromethylsulfony) ([P6,6,6,14][NTf2]) (Figure 1).

18th International Corrosion Congress 2011 Paper 535 - Page 2

Figure 1. Trihexyl(tetradecyl)phosphonium bis(trifluoromethylsulfony) ([P6,6,6,14][NTf2])

2. EXPERIMENTAL DETAILS

2.1 Materials used AA5083 aluminium alloy sheet specimens were obtained from Capral Ltd:

Table 1. Elemental Composition of Commercial AA5083 Aluminium Alloy

Element (w.t.%) Si Fe Cu Mn Mg Cr Zn Ti

Others Al

Each Total

Specification <0.4 <0.4 <0.10 0.4-1.0

4.0-4.9

0.05-0.25 <0.25 <0.15 <0.05 <0.15 Remaining

Content 0.11 0.28 0.03 0.71 4.4 0.07 0 0.05 <0.05 <0.15 94.23 -94.33

The microstructure of this wrought alloy, shown in Figure 2, is composed of intermetallic particles that are embedded in the aluminium solid solution matrix.

Figure 2. SEM image of AA5083 aluminium alloy

The [P6,6,6,14][NTf2] IL (Figure 1) was purchased from Io-li-tech company. Any surface processes are extremely sensitive to impurities, therefore the IL is always purified before use. As received ILs were dissolved in high purity liquid chromatography (HPLC) grade acetone, which was pressed through a column containing purification agent, alumina, sand and activated charcoal by nitrogen. Finally, the product was filtered, dried on a roto-vap, and stored in a desiccator.

18th International Corrosion Congress 2011 Paper 535 - Page 3

2.2 Surface preparation AA5083 specimens were polished to a P2500 surface finish on SiC abrasive papers under running tap water, followed by polishing on P4000 SiC paper using ethanol as the lubricant. Between papers and after polishing, surfaces were rinsed with deionised water and acetone dried under a nitrogen stream. All specimens were allowed to stabilise in a desiccator for at least 1 hour before anodising experiments were carried out.

2.3 Anodising A VMP3 Potentiostat was used for the anodising, and the software used was ECLab V10.10. A purpose-made pipette cell was used (Figure 3), which is composed of a pipette filled with IL, then clamped onto the metal surface. This set-up defines an active surface area of 50 ± 6 mm2, determined by the bottom area of the pipette. Before each experiment, approximately 3 ml IL was injected into the cell through the pre-drilled hole in the pipette wall. Platinum wire was used as the pseudo-reference electrode, while the counter electrode was a spot-welded titanium mesh. Epoxy-mounted specimens, with alloy surfaces smaller than that of the pipette, were also treated for the subsequent PP tests. There were two steps in anodising, Chronopotentiometry (CP) and Chronoamperometry (CA), which are constant current density and constant potential, respectively. For each specimen, a constant current density was applied, allowing the potential to ramp up to the 18 V maximum potential, which was followed by holding at 18V for a period of time.

Figure 3. The purpose-made pipette cell

2.4 Potentiodynamic Polarisation (PP) tests All PP scans of both the as-polished and the treated specimens were performed 0.1 M NaCl solution using a traditional 3-electrode cell and the same potentiostat used in anodising. The specimen was allowed to rest at OCP in solution for 30 minutes, before the potential was scanned from - 200 mV vs OCP to + 500 mV vs OCP, with a standard ramping rate of 0.1667 mV s-1. A saturated calomel reference electrode (SCE) was used. Tafel extrapolation were performed on the raw data to obtain the free corrosion potential Ecorr, corrosion current density Icorr, and breakdown potential Ebr values.

2.5 Optical Profilometry (OP) OP images were taken with the Veeco Contour GT-K1 Optical Profilometer. A tilt removal filter was applied to the raw data.

3. RESULTS

3.1 Anodising Results Figure 4 and Figure 5 represent the CP and CA treatment with various current densities. The current densities were reduced to very low values, from 0.20 mAcm-2 to 0.23 mAcm-2 (Table 2) after CA treatment. It seems that the higher the applied current density in CP, the less time it took for the sample potential to reach the limiting potential. For instance, CP treatment time for 0.5 mAcm-2, 1 mAcm-2 and 2 mAcm-2 were 190.2 s, 75 s and, 15.7 s, respectively (Table 2). It is also

18th International Corrosion Congress 2011 Paper 535 - Page 4

worth mentioning that the ramping rate of the potential was lowered in all cases at around 7 V. Moreover, higher current density led to the final current dropping to a lower value in subsequent the CA treatment.

0 50 100 150 200

02468

101214161820

E v

s P

t/V

time/s

0.5 mAcm-2 18V 15min 1 mAcm-2 18V 15min 2 mAcm-2 18V 15min

Figure 4. The effect of current density on CP treatment

0 200 400 600 800 1000 12000.00.20.40.60.81.01.21.41.61.82.02.2

j mA

cm

-2

time/s

0.5 mAcm-2 18V 15min 1 mAcm-2 18V 15min 2 mAcm-2 18V 15min

Figure 5. The effect of current density on CA treatment

Table 2. CP treatment time and CA final current density values

Pretreatment CP treatment time (s) CA Final current density (mAcm-2)

0.5mAcm-2 18V 15min 190.2 0.23

1mAcm-2 18V 15min 75.0 0.22

2mAcm-2 18V 15min 15.7 0.20

18th International Corrosion Congress 2011 Paper 535 - Page 5

3.2 PP Test Results PP scans shown in Figure 6 demonstrated that treatments undertaken using 2 mAcm-2 current density substantially reduced the cathodic and anodic corrosion kinetics. On the contrary, the lower current density treatments (0.5 mAcm-2 and 1 mAcm-

2) seemed to be less effective in suppressing corrosion processes. It is worth noting that none of the anodising treatments prevented the surface from breaking down at the same potential as the control sample.

Table 3 summarises Ecorr, Icorr, and Ebr values from all the PP scans presented in this study. The untreated control specimen had the most negative Ecorr, − 839 mV, whereas the Ecorr was shifted to more noble values for all the anodised specimens. The corrosion rate was substantially reduced, in the case of 2 mAcm-2 anodised specimen, to less than one third of the value in control specimen. However, the other anodised specimens both exhibited higher corrosion current densities Icorr than that of the control. This may reflect the presence of a highly defective coating.

-1.0 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3

-7

-6

-5

-4

-3

-2

-1

0

1

2

log

j mA

cm

-2

E vs SCE/V

Control 0.5 mAcm-2 18V 15min 1 mAcm-2 18V 15min 2 mAcm-2 18V 15min

Figure 6. The effect of current density on PP test results

Table 3. Ecorr, Icorr, and Ebr values (vs SCE), obtained via Tafel extrapolation of PP scans

Pretreatment Ecorr (mV) Icorr (μA/cm-2) Ebr (mV) Control -839 0.080 -691

0.5mAcm-2 18V 15min -746 0.262 -719 1mAcm-2 18V 15min -733 0.207 -707 2mAcm-2 18V 15min -755 0.022 -697

3.3 Morphologies Figure 6 shows the treated (left) and untreated (right) surfaces under an optical microscope. The whole surface was polished down to 0.05 μm with alumina powder before any treatment, in order to give better comparisons of the treated and untreated surfaces. The microstructure is the same with that seen in Figure 1, where intermetallic particles are present in the aluminium metal matrix. The intermetallic particles appeared to be higher than the surrounding surface in OP image, which was likely due to these intermetallic particle having a higher hardness and therefore was more difficult to polish down. Overall there was no significant height difference between the treated and untreated surfaces. It was observed that the anodising treatment altered the surface morphology completely. The treatment seemed to have significantly increased the surface roughness. This observation was confirmed by the OP image (Figure 8).

18th International Corrosion Congress 2011 Paper 535 - Page 6

Figure 7. The optical microscope image of the treated (2mAcm-2 18V 15min) and untreated surface

Figure 8. The OP image of the treated (2mAcm-2 18V 15min) and untreated surface

4. DISCUSSIONS

The anodising routine in the [P6,6,6,14][NTf2] IL, seemed to have produced a protective film on the aluminium alloy, because the current densities were reduced to very low values after the CA treatment, like that seen in aqueous electrolyte systems (18). The fast potential increase in Figure 4 can be associated with the formation of a compact barrier layer on the alloy surface. The rate that the potential was increasing slowed down at around 7 V in all cases. This was probably due to the change in film forming mechanisms that involves pores starting to initiate as the film continued to grow over the compact layer, built up during the first phase of the CP anodising. Different steps in anodising have also been reported in

18th International Corrosion Congress 2011 Paper 535 - Page 7

aqueous electrolyte systems (20). In the CA treatment, the final current decreased with increasing current density (Figure 5), which indicated that the anodising film formed at 2 mAcm-2 was probably more insulating as well as less porous. This would have a strong effect on the corrosion properties of the specimens.

In PP tests, the corrosion properties were evaluated. The substantial reduction in corrosion kinetics, seen in the specimen treated by anodising at 2 mAcm-2 then held at 18 V for 15 minutes, can be attributed to the metal reacting with the IL under the electrical bias and forming an anodising film on the alloy surface. The other anodising routines may have also deposited a surface film, but it was probably defective rather than protective. Therefore, there was no improvement in corrosion resistance. This observation correlated well with the final current density seen in the CA treatment, which is an indication of how defective the film was. The pitting potential remained unchanged for all treated specimens. This implied that whatever the anodising film was, it was either non-uniform or not thick enough to change the nature of the corrosion processes.

This was subsequently confirmed by the morphology observations in Figure 7 and Figure 8. Even the film formed on 2 mAcm-2 treated sample, which has shown improved corrosion resistance, appeared to be non-uniform and extremely thin. Unfortunately, the high roughness of the treated surface prevented easy determination of the thickness using the profilometry technique. The non-uniformity of the film may be caused by the intrinsic heterogeneity caused by the presence of various intermetallic particles. Further investigation is required to develop a better understanding of the film forming mechanisms, for example, whether there are preferential reaction sites for the film deposition to occur.

5. CONCLUSIONS

The CP and CA anodising treatments in [P6,6,6,14][NTf2] IL, seemed to result in the formation of a surface film as indicated by the lower current densities that were achieved after the final treatment step (CA). The higher the applied current density, the less time it took for the potential to reach the limiting potential in CP, and subsequently lower current was achieved in CA treatment and the better (possibly less porous) the film produced.

PP scans revealed that the corrosion potentials Ecorr were shifted to more noble values for all anodising treated specimens, which indicated that the anodising had changed the surface, most likely in the way of film deposition. However, only the 2 mAcm-2 anodised specimen showed significant reduction in corrosion kinetics. In addition, anodising seemed to have no effect on the pitting potential, with all treated specimens exhibiting Ebr in the same range. This can be related to extremely thin or defective films.

The microscope and OP images of the surface gave a direct appreciation of the morphology of the anodising film. There was an obvious change in roughness but no observable thickness change, which may be due to the surface film being extremely thin and preventing the determination of the film thickness.

Future work will focus on developing a better understanding of the film forming mechanism, for example, whether the film formation initiates on the more reactive intermetallic particles. In addition, the effect of various electrochemical parameters on the film morphology will be investigated in order to form a more corrosion resistant surface film.

6. ACKNOWLEDGEMENTS

The authors are grateful for funding support via the Australian Research Council through the Discovery Project scheme. We would also like to thank Mrs. Julie-Anne Latham for helpful discussions and assistance with the initial experimental protocols.

7. REFERENCES

1. Peng C, Yang L, Zhang Z, Tachibana K, Yang Y, Zhao S. Investigation of the anodic behavior of Al current collector in room temperature ionic liquid electrolytes. Electrochimica Acta. 2008;53(14):4764-72. 2. Goldman JL, McEwen AB. EMIIm and EMIBeti on Aluminum Anodic Stability Dependence on Lithium Salt and Propylene Carbonate. Electrochemical and Solid-State Letters. 1999;2(10):501-3. 3. Birbilis N, Howlett PC, MacFarlane DR, Forsyth M. Exploring corrosion protection of Mg via ionic liquid pretreatment. Surface and Coatings Technology. 2007;201(8):4496-504. 4. Forsyth M, Howlett PC, Tan SK, MacFarlane DR, Birbilis N. An Ionic Liquid Surface Treatment for Corrosion Protection of Magnesium Alloy AZ31. Electrochemical and Solid-State Letters. 2006;9(11):B52-B5. 5. Ohno H. Importance and Possibility of Ionic Liquids: John Wiley & Sons, Inc.; 2005. 6. Blanchard LA, Hancu D, Beckman EJ, Brennecke JF. Green processing using ionic liquids and CO2. Nature. [10.1038/19887]. 1999;399(6731):28-9.

18th International Corrosion Congress 2011 Paper 535 - Page 8

7. MacFarlane DR, Huang J, Forsyth M. Lithium-doped plastic crystal electrolytes exhibiting fast ion conduction for secondary batteries. Nature. [10.1038/45514]. 1999;402(6763):792-4. 8. Howlett PC, Brack N, Hollenkamp AF, Forsyth M, MacFarlane DR. Characterization of the Lithium Surface in N-Methyl-N-alkylpyrrolidinium Bis(trifluoromethanesulfonyl)amide Room-Temperature Ionic Liquid Electrolytes. Journal of The Electrochemical Society. 2006;153(3):A595-A606. 9. Howlett PC, MacFarlane DR, Hollenkamp AF. High Lithium Metal Cycling Efficiency in a Room-Temperature Ionic Liquid. Electrochemical and Solid-State Letters. 2004;7(5):A97-A101. 10. Uerdingen M, Treber C, Balser M, Schmitt G, Werner C. Corrosion behaviour of ionic liquids. Green Chemistry. 2005;7(5):321-5. 11. Koch VR, Nanjundiah C, Appetecchi GB, Scrosati B. The Interfacial Stability of Li with Two New Solvent-Free Ionic Liquids: 1,2-Dimethyl-3-propylimidazolium Imide and Methide. Journal of The Electrochemical Society. 1995;142(7):L116-L8. 12. Howlett PC, Zhang S, MacFarlane DR, Forsyth M. An Investigation of a Phosphinate-Based Ionic Liquid for Corrosion Protection of Magnesium Alloy AZ31. Australian Journal of Chemistry. 2007;60(1):43-6. 13. Shkurankov A, Zein El Abedin S, Endres F. AFM-Assisted Investigation of the Corrosion Behaviour of Magnesium and AZ91 Alloys in an Ionic Liquid with Varying Water Content. Australian Journal of Chemistry. 2007;60(1):35-42. 14. Forsyth M, Neil WC, Howlett PC, Macfarlane DR, Hinton BRW, Rocher N, et al. New Insights into the Fundamental Chemical Nature of Ionic Liquid Film Formation on Magnesium Alloy Surfaces. ACS Applied Materials & Interfaces. 2009;1(5):1045-52. 15. Efthimiadis J, Neil WC, Bunter A, Howlett PC, Hinton BRW, MacFarlane DR, et al. Potentiostatic Control of Ionic Liquid Surface Film Formation on ZE41 Magnesium Alloy. ACS Applied Materials & Interfaces. 2010;2(5):1317-23. 16. Howlett PC, Khoo T, Mooketsi G, Efthimiadis J, MacFarlane DR, Forsyth M. The effect of potential bias on the formation of ionic liquid generated surface films on Mg alloys. Electrochimica Acta. 2010;55(7):2377-83. 17. Latham J-A, Howlett PC, MacFarlane DR, Forsyth M. Electrochemical reactivity of trihexyl(tetradecyl)phosphonium bis(2,4,4-trimethylpentyl)phosphinate ionic liquid on glassy carbon and AZ31 magnesium alloy. Electrochimica Acta.In Press, Corrected Proof. 18. Sheasby PG. The Surface Treatment and Finishing of Aluminium and its Alloys. 6th ed: ASM International; 2001. 19. Zhou X, Sheasby PG, Scott BA. Coatings Produced by Anodic Oxidation. In: Tony JAR, editor. Shreir's Corrosion. Oxford: Elsevier; 2010. p. 2503-18. 20. Parkhutik VP, Shershulsky VI. Theoretical modelling of porous oxide growth on aluminium. Journal of Physics D: Applied Physics. 1992;25(8):1258.

8. AUTHER DETAILS

P. Huang is a PhD student at Deakin University, Australia. She has got a bachelor degree in Materials Engineering at Monash University, Australia, and a bachelor degree in Materials Science and Engineering at Wuhan University of Technology, China. She is now working on a discovery project under the supervision of Prof. Maria Forsyth and Dr. Patrick Howlett, which involves passivation of aluminium alloys using ionic liquid coatings.