Temperature Effect on the Kinetic Alumina Layer Growth on 5086 Aluminum Substrate

*LSCMI, Mechanical Engineering Faculty, University of Sciences and Technology Oran Mohamed Boudiaf BP 1505 El M’Naouer 31000, Oran-Algeria. E-mail: bousourai@yahoo.fr **Laboratory of Tribology and Systems Dynamics (LTDS), UMR 5513, 36 Avenue Guy de Collongue, 69134 Ecully, Cedex, France. E-mail: Sophie.Pavan@ec-lyon.f; E-mail: Vincent.Fridrici@ec-lyon.fr; E-mail: Philippe.Kapsa@ec-lyon.fr ***Normandie Univ, Esplanade de la Paix, F-14032 Caen, Cedex 5, France, UNICAEN, CIMAP, F-14050 Caen, France, ENSICAEN, F-14050 Caen, France, CNRS, UMR 6252, F-14050 Caen, France, CEA, UMR 6252, F-14050 Caen, France, E-mail: christophe.poilane@unicaen.fr


Introduction
Nowadays, the surface finishing becomes a greater part of final product and the coating plays an important part to enhance adherence properties.In anodization process, the current flows through an electrolyte in which aluminium sheet is used as anode.A thick film of aluminium oxide is built up on this substrate surface.Various parameters determine the structure and the properties of a formed porous anodic alumina (PAA) [1,2].The layer is uniform and presents a higher corrosion and abrasion resistance to those of the aluminum.The most important electrolytes are sulfuric acid (SAA: H2SO4), chromic acid (CAA: H2CrO4), phosphoric acid (PAO: H3PO4), boric acid (BAAO: H3BO3) and oxalic acid (OAA: HOOC-COOH) [3].A mixed of different acids is used as electrolyte in the aerospace industry.Porous alumina films consist in two layers: a thin barrier of oxide in contact with the metal and a thick porous layer.The anodic oxide exhibits good electrical resistance, which is correlated to the oxide thickness.Aluminum oxide is very resistant to electric current and the current density decreases with time anodizing.An appropriate anodization parameters control allows obtaining good results in terms of porous structure regularity [4].The applied voltage, the current density, the texture, microstructure of aluminum substrate and the temperature affect the pores distribution as a honeycomb during the anodizing process [5][6][7].
Current distribution was modeled by Akolkar et al. [8] in stirred H2SO4 electrolyte at different voltages and temperatures for various anodizing times; they showed that the major resistivity resides in the sub-layer.
Sulka and Parkola [9] obtained good results for anodizing parameters control at 25 V for a temperature of 1°C.The anodic oxide film formed at 58°C exhibits lower dielectric properties compared with the amorphous oxide films at 8°C and 25°C [10].
Several authors showed the interest of temperature on the courant behavior [11,12].A difference, of 20°C between low temperature (0°C) and room temperature, modifies the alumina layer kinetic growth.A local difference of 5°C generates cracks into the oxide layer at the interface [13].Aerts et al. [14] quantified the porosity and the pores dimension with temperature whereas Sulka and Parkola [15] and Araoyinbo et al. [16] analyzed the effect of applied anodizing voltage.Fratila-Apachitei et al. [17] showed the effect of the convective regime on electrode temperature at 65°C as on oxide layer growth.Vrublevsky et al. [18] confirmed the change in the growth mechanism of porous alumina for a tension above 55 V by the break in the curves of anodizing voltage.They used a temperature of 50°C for chemical dissolution of the oxide, and then the temperature reaches rapidly 60°C under free convection which enhances local oxide dissolution resulting in non-uniform thickness.Aerts et al. [19] studied the anode electrochemical behavior, the oxide layer morphology with the temperature variation of anode and electrolytic solution.
Temperature close to the boiling point (96°C), usually undertaken for the sealing process, involves secondary reactions in anodized layer [20][21].
Assessment of mechanical properties such as hardness and Young's modulus is necessary to analyze the mechanical behavior of coating layers under static or dynamic conditions [22].The accurate determination of the elasticity modulus is possible for massive and homogeneous materials with Vickers hardness tests.However, it is often difficult to achieve it for thin films with heterogeneous properties.Nano-indentation tests are used for mechanical characterization of thin film [23][24][25][26].

Sample's preparation
The experiments were conducted with 5086 aluminium alloy (4.9 wt % Mg) as substrate with chemical composition given in Table 1.Samples are cut out with precision blade (ISOM-ET 4000 BUEHLER) under lubrication.Final dimensions of samples are 120 mm x 25 mm x 2.5 mm.Before anodiz-ing, polishing of twelve surface specimens was grounded successively with silicon carbide papers from 120 to 1200, diamond paste 1 µm and alumina 0.4 µm to reach a mirror state with 0.1 μm roughness.Then, specimens were cleaned in an ultrasonic acetone bath during 15 min then rinsed with distilled water.

Anodization
We choose a diluted H2SO4 solution (SAA, 20 % mass) largely used in industry to obtain a hard layer [27].The hard anodization process in sulfuric acid at low temperature easily develops an oxide layer much thicker than natural barrier, up to 150 µm during few hours.Directly built and strongly bonded on the aluminium, alumina is a passive oxide film barrier in the electrolyte with pH range from 4 to 8.
A special electrodes configuration was adopted to minimize the electrolyte resistance.The aluminium anode is placed at 2 mm from the borders lead cathode shaped in V as illustrated in Fig. 1.The anodization SAA was performed in 1 L glass beaker cooled by a circulating machine (Lauda Circulator Programmable C6).Cooling temperatures ranging between 16°C to 21°C are used for specimens at room (RT) and between -1°C to 4°C for the lowest temperatures (LT).A magnetic stirrer at 500 rpm is used to prevent localized temperature concentration.After anodizing, the specimens were only rinsed with distilled water and dried with compressed air without sealing to conserve the layers in their initial states.
The anodization time was set at 30 min, 60 min, 120 min, 180 min, 240 min and 300 min with an initial current of 1.4 A corresponding to voltages of 14 and 18 V respectively for both cooling temperatures.For long anodization times (beyond to 30 min) we consider that the thermal equilibrium is well established between the electrolyte and the electrodes.The solution being vigorously stirred, the difference in temperature between the oxide and the substrate, which presents a high thermal conductivity, is negligible [19].

Results and discussions
Current, temperature and voltage evolution are automatically recorded according to time.The voltage evolution according to times at low and room temperature are gathered in Fig. 2 for an initial courant of 1.4 A. The effect of anodization time is significant on tension at low temperature.We observed a difference ranging from 3V to 8 V during anodization between low and room temperature.The current decreases more rapidly at low temperature than at room temperature as shown in Figs. 3 and 4.This result well agrees with results given in [10].Qualitatively, the Fig. 5 shows a large colors range, each one corresponding to a given thickness and the alumina layer growth with the anodizing time and temperature.Colors ranging are from light to dark gray for tests at room temperature corresponding to layers with thickness lower than 80 µm and from light brown to very dark.At low temperatures the thickness can reach 150 µm.
The specimen cross section was examined by microscopy where the thickness was directly measured, Fig. 6.
Hard anodizing in sulphuric acid at low temperature forms a thicker layers oxide than those at room temperature, Fig 7.
The difference in thickness is particularly observed for time up to 60 min and it is not recommended to continue anodizing after 120 min at room temperature since the oxidation is balanced by a simultaneous ions reduction presents in solution.This drawback is not noted at low temperature which explains the ability to make thick anodized layer by such a way.For the fifth specimen at low temperature (5 hours, LT), we have voluntarily stopped the cooling after 120 min anodizing.Thus, this promotes defects in the oxide film and the current curve is close to the current room temperature, Fig. 3.All current curves versus of time show regularly spaced fluctuations; two cases are considered: 1.A local increase in temperature leads to the dissolution of the oxide and consequently to a decrease in the layer growth kinetics.2. During growth, a difference between the densities of alumina and aluminum substrate, generates cracks during anodizing, Fig. 9, a.Cracks can be deep, Fig. 9, b.Then, the exposure of the substrate aluminum in contact with the electrolyte develops a high current density, because the anodic surface (aluminum) is very little in regard with the large cathode surface (alumina).Fig. 8 shows porosities in the cross section layer at low and room temperature, the effect of cold sulfuric acid electrolyte enhances the pores formation.At the initial anodizing state and at the surface or immediately close to the substrate, Mg5Al8 precipitates generate magnesium cation which migrates across the oxide under electric field faster than aluminium [28].During anodizing formation of alumina and for long time, the precipitates are trapped in the oxide and the magnesium content in Mg5Al8 particles, highly soluble in the acid electrolyte, forms pores through the cross section and in regions called stains [29].
A detailed observation of color evolution at LT shows a slight decrease in blackness for the fifth sample, Fig. 5.
Fig. 10 illustrates the hardness evolution according to anodizing time.The measurement was made on the top of coating surface.The hardness of low temperature alumina is more important, the maximal value is reached at two hours, after which there is no need to extend the anodizing process.For the aim to assess the mechanical proprieties of both aluminium and oxide layer, a nano-indentations was performed on a row according to an angle of 20 °C with oxide-aluminum interface.The distances between imprints into the layer are 15 µm through thickness and 40 µm in the longitudinal direction.The tests were conducted at room temperature on XP nano-indenter with a Berkovich indenter.A maximum load of 450 mN was applied.
Before and after tests nanoindentation, indentations are performed on a reference sample of silica, for two reasons in order to: -Establish the defect of the indenter tip (Berkovich) used for the calculations of hardness and elastic's modulus.
-Check the mechanical properties on a silica sample.The Fig. 11 represents respectively loadingunloading curves, the reduced Young's modulus evolution, hardness and one example of imprint obtained for aluminium subtract.We remark that the substrate exhibits homogeneous mechanical properties.The average hardness and the reduced Young modulus values are, 700 MPa and 72 GPa, respectively [30].Fig. 12 represents the loading-unloading curves, the evolution of reduced Young's modulus and hardness, and the imprint series in the alumina layer.Nano-indentation analysis shows heterogeneity in the mechanical properties during layer growth, Fig. 13.
The reduced Young modulus and the hardness vary respectively from 80 GPa to 40 GPa and from 1 GPa to 4 GPa from the interface to the oxide outer limit.The average hardness and Young's modulus evolution according to depth are presented in Fig. 14.It is clear that the alumina is heterogonous in terms of mechanical properties through the oxide cross section.That is due to defect induced during the pores growth generated by the anodizing process.
The evolution of hardness with the thickness oxide is similar whatever the anodization time.In the Fig. 15, the origin of the depth scale is the aluminium interface.The hardness decreases with the depth but this is particularly more quickly according to RT anodization.

Conclusion
In this study we showed qualitatively the temperature effect and layers color correlation with the thickness.For all anodizing times at room temperature the voltage was constant but the current decreases according to the alumina layer thickness growth.At low temperature, the forced convection of the electrolyte improves the layer's growth and avoids a local temperature rise.In all case, low temperatures enhance layer's thickness growth with high mechanical properties than those the room temperature alumina witch the layer thickness is limited in time.Nanoindentation tests illustrate the heterogeneous nature of alumina layer at room temperature.

Fig. 2 Fig. 3
Fig. 2 Voltage vs anodizing time at low and room temperature

Fig. 7 Fig. 8
Fig. 7 Thickness evolutions against time anodizing for room and low temperature

Fig. 10 Fig. 11 a
Fig. 10 Hardness against hard anodizing time on the layer surface