P. Spathis, E. Papastergiadis, G. Stalidis, G. Papanastasiou
Dept. of Chemistry, University of Thessaloniki,54006, Thessaloniki, Greece
The aim of the present work is the study of corrosion and stress corrosion cracking behaviour of 1050 Al-Alloy anodised in a 3M H2SO4 anodising bath in the presence of various carboxylic acids, saturated or not, with different numbers of carboxylic, methyl or hydroxyl groups. The investigation was carried out by SCC tests and electrochemical measurements. The SCC behaviour of anodised 1050 Al-Alloy was found to vary with anodising conditions, stress level and the differences in the structure and the carbonic chain of the additives. Anodic coatings prepared in 3M H2SO4 without any additives, did not protect the bare alloy. Addition to the anodising solution of any of the carboxylic acids used resulted in protection of the alloy, with better protective properties in the case of malonic acid and a high stress level. For the interpretation of the results, SEM micrographs and IR spectra of the surfaces of the specimens were obtained. Properties of the anodic coatings such as thickness, packing density, coating ratio and roughness were also studied. The anodic coatings formed in a electrolytic bath with additives present were found to be less porous, more compact and rougher, having better anticorrosive and mechanical properties. An explanation of the mechanism of the effect of carboxylic acids on oxidation and SCC behaviour of Al-Alloys is the absorption of these compounds on the metal surface. Corrosion and stress corrosion cracking behaviour is better in the case of saturated bicarboxylic acids. The presence of a methyl group improves the protective properties of the oxide but an increase in their number has an opposite effect.
Keywords: corrosion, SCC, protection, aluminium, anodic coatings, bicarbonic acids.
The SCC behaviour of Al-alloys of the 1XXX series is rarely studied because the high strength Al-alloys, eg. of the 7XXX series, are more important for most industrial applications. Neverthless, Al-alloys of the 1XXX series are widely used in many commercial applications and therefore studies of these alloys are useful for estimating the effects of the various constituents of high strength alloys on their corrosion behaviour.
Studies on behaviour of pure aluminium  in saline environments and in stress corrosion conditions, conclude that as with other Al-Alloys [2,3], the cracking is due to an atomic hydrogen absorption that causes crack growth and increases dislocation activity. Also, depending on the applied potential, different processes become prominent and the SCC at OCP is caused by pit formation and crack initiation of pits .
Some Al-alloys used in practice are anodically oxidized, for protection against corrosion and SCC. It has been shown that during anodising, a porous Al2O3 film with a cellular structure up to a thickness of 36 mm is formed, with cells oriented parallel to the direction of gravity this resulting from the rapid evolution of oxygen in the opposite direction. It was found that the protective properties of oxides change with thickness and structure, being better against corrosion than unanodised aluminium but worse against mechanical stress [5,6,7].
Many studies report [8-14] the effect of the addition of various organic compounds on the corrosion behaviour of metals, mainly in the corrosive environment, but these may also, in the case of aluminium, be in the anodising bath during the electrolytic preparation of the oxides. The explanation given for this effect is the absorption of compounds on the metal surface or the formation of some complexes on it.
In earlier work [15-17] the effect on corrosion and stress corrosion cracking behaviour of anodised 1050 Al-Alloy of the addition in an anodising bath of 3M or 4M H2SO4 of some inorganic compounds , bicarboxylic acids in 3M H2SO4  or carboxylic acids in 4M H2SO4  was examined.
In order to further investigate this effect, aim of the present work is the study of corrosion and stress corrosion cracking behaviour of 1050 Al-Alloy anodised in a 3M H2SO4 anodising bath in the presence of various carboxylic acids, saturated or not, with different numbers of carboxylic, methyl or hydroxyl groups. The investigation was carried out by SCC tests and electrochemical measurements.
The Al-Alloy tested was pure aluminium (Al > 99.5%), ASTM 1050, H38 and the corrosive environment was a 3M NaCl solution.
The anodic coatings on Al-Alloy surface were prepared electrolytically in a bath of 3 M H2SO4 with an addition of 0.015 M of the following carboxylic acids: acetic, malonic, succinic, phthalic, glutaric, adipic, fumaric, citric.
The anodising current density was 600 A/m2, at 25�C and the coatings thickness was estimated (6) to 10 mm. The real thickness of the coatings was confirmed by direct microscopic examination of the cross sections of the anodised specimens.
The specimens were cut from a plate of 0.30 mm thickness and have a middle section of reduced width where the cross section was 1 mm2. The total exposed area of the specimens was 1.28 cm2 and the rest was masked with insulating varnish to eliminate any parasitic interactions during testing (Fig. 1). The specimens were stressed directly by loading at a definite stress level and the time to failure (TTF) was derived. Experiments at three different stress levels (4.17, 6.25, 8.33 kg/mm2) were carried out. During the test, the corrosive environment was renewed at a constant rate of 70 ml/h. A galvanostatically controlled anodic current of 5 A/m2 was impressed during the test. Six specimens were tested under each set of conditions.
Figure 1. a) Shape and dimensions of specimens (AA'-BB': anodizing area for the case of anodized specimens). b) Specimens position during SCC tests. (Click the figure for an enlarged view)
To ensure that the SCC mechanism was the predominant factor for the failure of the specimens during the SCC tests, the following factors were included in the design of this experiment. The applied stresses σ were in the region where SCC is defined (threshold stress = 1.75 kg/mm2 < s = 4.17, 6.25, 8.33 kg/mm2 < yield point = 9.41 kg/mm2). The impressed anodic current densities to accelerate the SCC phenomenon were low: 5 A/m2. The present authors took into account the results of a study of the influence of the pure electrochemical dissolution during SCC tests of aluminium alloys in saline water  where it was found that "when testing under anodic SCC conditions and the tests are carried out in the region where a true SCC mechanism is valid, anodic dissolution is not the predominant factor" . These results were obtained from TTF measurements of specimens under the same experimental conditions as in this work. The specimens were at first pre-exposed under the same conditions of applied anodic current without load, and then loaded to fracture.
The specimens were also cut from a plate of the same thickness (0.30 mm), the dimensions were 1 cm x 5 cm and the total exposed area was 2 cm2 (the rest was also masked with insulating varnish).
The cyclic anodic potentiodynamic polarization measurements were carried out with standard methods [19-21], without any agitation or renewal of the solution, at a slow scan rate of 0.6 V/h. Before starting the potential scan, the specimens were immersed in the test solution for 1 h to reach a steady state of equilibrium (open circuit corrosion potential, Ecor). A potentiostat-galvanostat (Bank PGS-81) with a scan generator (Bank VSG-72) and a X-Y recorder, platinum counter electrode and saturated calomel reference electrode (SCE) were used.
Scanning electron microscopy (SEM) was used to study the surface of the specimen. The SEM experiments were carried out with a JEOL, JSM-840 A Scanning Microscope, connected with a Energy Dispenser Spectrometer - EDS - (LINK, AN 10/55S).
For the calculation of the packing density, the weight of the anodic coating, its thickness and the oxidised area of the specimen were measured. The packing density was calculated as the ratio of weight to volume of each anodic coating.
The coating ratio, i.e., the coulombic efficiency for the formation of porous oxide films on Al, was calculated by the ratio of the weight of oxide formed to the weight of aluminium consumed .
The roughness of surface of the anodic coatings was measured with a perthometer (PERTHEN C 5D with a tracer drive unit PVK) and the roughness factor that was used in our experiments was the mean arithmetic value of all the distances of the roughness diagram from its central line.
Infrared Reflection spectrum (IR)
The Infrared Reflection spectrum method was used for the determination of the qualitative composition of the anodic coatings. Spectra were taken with an infrared spectrometer with Fourier transformation (FFT-IR, Bruker IFS 113, lamp Globag I, beamsplitter KBr, detector DTGS with window KBr and 5000-400cm-1 measurement area).
In Figure 2, SCC results (increase % of TTF of anodised against bare specimens) of samples anodised in 3M H2SO4 and with an addition of 0.015 M of various carboxylic acids are shown. The SCC behaviour was found to vary with anodising conditions and stress level. Anodic coating prepared in 3M H2SO4 without any additives did not protect the bare alloy, decreasing TTF. Addition to the anodising solution of any of the carboxylic acids used resulted in protection of the bare alloy. The increase of TTF was greater for the saturated bicarboxylic acids and shows a maximum value in the case of malonic acid, the protective properties being better as stress level increases.
Figure 2. SCC results (increase % of TTF of anodised against bare specimens) of anodised in 3M H2SO4 1050 Al-Alloy and with an addition of 0.015 M of carboxylic acids: (1): 3M H2SO4, (2): 3M H2SO4 + acetic acid, (3): 3M H2SO4+ malonic acid, (4): 3M H2SO4 + succinic acid, (5): 3M H2SO4 + phthalic acid, (6): 3M H2SO4 + glutaric acid, (7): 3M H2SO4 + adipic acid, (8): 3M H2SO4 + fumaric acid, (9): 3M H2SO4 + citric acid. (Click the figure for an enlarged view)
In Figure 3 the potentiodynamic polarization curves of bare or anodised specimens in 3M H2SO4 and with an addition of 0.015 M of malonic or citric acids are shown. From these results it follows that in all cases the Al-Alloy suffers localized corrosion. This is indicated by Epit having a more positive value than Ecor and also from the presence of a hysteresis loop between the forward and reverse scans [23,24]. In all cases the Al-Alloy is not passivated, as passive regions do not appear for potential values more positive than Epit and reverse currents are always higher than forward currents. It is also observed that the addition of carboxylic acids did not significantly influence Ecor but shifted Epit in the noble direction (-610 mV for malonic acid) relative to the bare alloy (-680 mV) and decreased the anodic current indicating less susceptibility to localized corrosion in the free corrosion potential regions. In the absence of any carboxylic acid, Epit shifted in the opposite direction (-720 mV for 3 M H2SO4).
Figure 3. Anodic potentiodynamic polarization curves of bare or anodised in 3M H2SO4 1050 Al-Alloy for various additions of 0.015 M of carboxylic acids: (1): bare Al, (2): 3M H2SO4, (3): 3M H2SO4+ malonic acid, (4): 3M H2SO4 + citric acid. (Click the figure for an enlarged view)
The infrared reflection spectra (IR) are shown in Figure 4. The peaks observed in the areas 3500-3200cm-1 and 1625 cm-1 indicate the presence in the oxides of hydroxyl groups and absorbed water. The peak at 1460cm-1, observed in all coated specimens, corresponds with the C-O bond. The presence of this peak in the case of the oxide prepared in 3M H2SO4 without any additives it is due to the chemical absorption of CO2 from the environment onto the oxide. The peak at 1200cm-1 corresponds with the S-O bond of the sulphide ions that are absorbed in the oxide during anodization and the peaks in the 1080cm-1, 970cm-1, 940cm-1 and 570cm-1 are due to vibrations of the Al-O bond. The detection of such bonds is in accordance with the results of other works , where the spectroscopic study of the aluminium oxide prepared in H2SO4 showed a triple structure for the oxide, composed mainly from coatings of aluminium hydroxyl, aluminium hydroxide and hydrated aluminium oxide. The presence of carbonate and sulphate ions was also detected (Fig. 5).
Figure 4. Infrared Reflection spectrum (IR) of bare or anodised in 3M H2SO4 1050 Al-Alloy for various additions of 0.015 M of carboxylic acids: (1): bare Al, (2): 3M H2SO4, (3): 3M H2SO4+ malonic acid, (4): 3M H2SO4 + citric acid. (Click the figure for an enlarged view)
Figure 5. Schematic presentation of the structure of anodic coated in H2SO4 aluminium surface.
Both corrosion and especially stress resistance of the coating are better when prepared in the presence of malonic acid as shown from the SEM micrographs of SCC tested specimens, coated in 3 M H2SO4 + 0.015 M malonic acid baths (Fig. 6).
Figure 6. SEM micrographs of SCC tested specimens at s = 6.25 kg/mm2: a) bare specimens, b) anodised in 3M H2SO4, c) anodised in 3M H2SO4 + 0.015M malonic acid, d) anodised in 3M H2SO4 + 0.015M acetic acid, e) anodised in 3M H2SO4 + 0.015M glutaric acid, f) anodised in 3M H2SO4 + 0.015M fumaric acid. (Click the figure for an enlarged view)
Table 1: Physical properties of anodic coating prepared in 3 M H2SO4with and without additives
3M H2SO4+0.015M malonic acid
3M H2SO4+ 0.015M citric acid
The measurements of the physical properties of the anodic coatings shown in Table 1 indicate that the addition of carboxylic acids during anodising decreases thickness and increases packing density of the coatings resulting in the formation of a less porous oxide layer. The coating ratio decreases in the presence of additives, while roughness increases. These results can be attributed to the lower presence of SO3 in the oxide film (6.48% in presence of malonic acid and 7.36% without the additive), due to lower incorporation of anions into the oxide structure during anodising in presence of the additives and also to the possible dissolution of the outer surface of the oxide during anodising . An explanation of the mechanism of the effect of carboxylic acids on oxidation, corrosion and SCC behaviour of Al-Alloys is the absorption of these compounds on the metal surface. The different behaviour of the coatings prepared in the presence of various carboxylic acids must be explained by the differences in the structure and the hydrocarbon chain of these compounds. Corrosion and stress corrosion cracking behaviour is better in the case of bicarboxylic acids and worse in the cases of one (acidic acid) or three (citric acid) carboxylic groups present. It is also better in the case of saturated acids than in the presence of double bonds (maleic, fumaric, phthalic acids). The presence of methyl groups improves the protective properties of the oxide, which are worse in the absence of these (oxalic acid) or in the presence of hydroxyl groups (tartaric acid). Increase of the number of the methyl groups from one to four (malonic, succinic, glutaric, adipic acids) makes the protective properties of the oxide worse.
The SCC behaviour of anodised 1050 Al-Alloy was found to vary with anodising conditions, stress level and the differences in the structure and the carbonic chain of the additives. Anodic coatings prepared in 3M H2SO4 without any additives, did not protect the bare alloy. Addition to the anodising solution of any of the carboxylic acids used resulted in protection of the alloy, with better protective properties in the case of malonic acid and a high stress level.
The addition of carboxylic acids during anodising decreases thickness and coating ratio and increases packing density and roughness of the coatings, resulting in the formation of a less porous oxide layer.
An explanation of the mechanism of the effect of carboxylic acids on oxidation and SCC behaviour of Al-Alloys is the absorption of these compounds on the metal surface. Corrosion and stress corrosion cracking behaviour is better in the case of saturated bicarboxylic acids.
The presence of a methyl group improves the protective properties of the oxide but an increase in their number has an opposite effect.
1. Bond, G.M., I.M. Robertson and H.K. Birnbaum, Acta Metall. 36,8 (1988), pp. 2193-2197.
2. Burleigh, T.D., Corrosion J., 47, 2, 1991, pp. 89-98.
3. Holroyd, N.G.H., Environment. Induced Cracking of Metals, Proceedings, 1989, pp. 311-345.
4. Lunarska, E. and Z. Szklarska-Smialowska, Corrosion J., 43, 7, 1987, pp. 414-423.
5. Skoulikidis, Th. and A. Karageorgos, Br. Cor. J., 15, 1, 1980, pp. 41-43.
6. Skoulikidis, Th. and P. Spathis, Br. Cor. J., 17, 1982, pp. 79-83.
7. Skoulikidis, Th. and J. Colios, Aluminium, 63, 6, 1987, pp. 619-623.
8. Moussa, M.N.H., et al., Corrosion Science, Vol. 16, 1976, p. 379.
9. Cotton, J.B., J.R. Scholes, Br. Corr. J. Vol. 2, Jan 1967, p. 1.
10. Poling, G.W., Corrosion Science, Vol. 10, 1970, p. 359.
11. Lyberatos, J., L. Kobotiatis, Corrosion, Vol. 47, 11, Nov. 1991, p. 820.
12. Monticell, C., et all., Corrosion Science, Vol. 37, No 7, 1991, p. 693.
13. Jeong Yong Soo et al., Kumsok Pyomyon Choli, 17 (3), 1984, p. 78.
14. Gryzlov V.N., Khim. Khim Tekhnol (30), 7, 1987, p. 72.
15. Spathis, P., E. Papastergiadis, G. Stalidis, G. Papanastasiou, to be published.
16. Spathis, P., E. Papastergiadis, G. Stalidis, G. Papanastasiou, Advances in Corrosion Protection by Organic Coatings II;The Electrochemical Society, Proceedings Volume, 95-13, 309, 1995.
17. Spathis, P., E. Papastergiadis, G. Stalidis, G. Papanastasiou, to be published.
18. Karageorgos, A., Proc. 5th Int. Cong. Marine Corrosion and Fouling. Barcelona, p. 4, 1980.
19. ASTM G1-81.
20. ASTM G5-82.
21. ASTM G61-78.
22. Diggle, J.W., et al., J. Electrochemical Society. 115, 1968, p. 365.
23. Baboian, R., and G. Haunes, Electrochemical Corrosion Testing, ASTM, STP 727, 1981, pp. 274-282.
24. Siebert, O. Laboratory Corrosion Tests and Standards, ASTM, STR 866, 1985, pp. 65-90.
25. Fondeur, F. and J.L. Koening, J. Adhesion, Vol.40, 1993, pp. 189-205.