Improving corrosion resistance of pure Al through ECAP

ECAPed specimen preparation

The material used for ECAP process was cut from cast industrial pure Al ingots, which has the chemical composition of Al–0·15Fe–0·05Si–0·005Mn–0·001Ca. The schematic illustration of the ECAP process is shown in Fig. 1. The billets with the size of 20×20×40 mm were cut by an electrical discharging machine and then pressed for 16 passes with a plunger speed of 0·5 mm s⁻¹ at room temperature. Graphite was used as lubricant to reduce the friction coefficient between the billet and the die inner wall. After each pass, the billet was inverted, and then rotated by 180° on the same axes.

OPEN IN VIEWER

Figure 1

Schematic illustration of ECAP process

Microstructure observation

The samples used for optical microstructure observation were cut parallel to the pressing direction, and then electrolytic etched in 2·5% fluoboric acid (HBF₄) at room temperature. Transmission electron microscopy (TEM, JEM2000EX, Japan) was conducted to observe the grain size, grain boundary and dislocation of the UFG industrial pure Al. The thin foils for TEM were parallel to the pressing direction and prepared by twin jet electropolisher using 10%HClO₄+90% alcohol solution at the temperature of 7·8°C and 20 V electropolish voltage. The electron backscatter diffraction (EBSD) measurements were carried out using FEI Sirion 400 SEM with a field emission gun operated at 15 kV. The thin foils for EBSD measurements were the same as the ones for TEM.

Corrosion tests

Corrosion resistance of the ECAPed pure Al was investigated by two methods, namely, constant immersion test and electrochemical test. The as cast industrial pure Al sample was also studied for comparison. All the experiments were conducted in NaCl solution at room temperature, and the solution was obtained via dissolving AR grade NaCl in distilled water.

For constant immersion tests, the initial specimens were ground with SiC paper up to 1000 grit and then mechanical polished using diamond paste. The prepared specimens were immersed in the 3·5 wt-%NaCl solution for 10 days. After immersion, their surface morphologies were observed by a KH-7700 digital microscope (Hirox, USA) and a scanning electron microscopy (Hitachi S340-N, Japan) respectively.

Electrochemical tests were conducted in 0·01 mol L⁻¹ NaCl solution using a Parstat 2273 advanced potentiostat with a traditional three-electrode system. The system contains a saturated calomel reference electrode and a Pt counter electrode. All samples were cut from the core of UFG billets perpendicular to the pressing direction by electrical discharging machine, and then moulded in the epoxy with an exposed surface of 1 cm². For a good reproducibility, all samples were polished, cleaned with acetone and dried in warm air, and then preimmersed in the solution for 30 min to achieve a stable open circuit potential before all the electrochemical tests. Three kinds of electrochemical tests, namely, electrochemical impedance spectroscopy (EIS), potentiodynamic polarisation and potentiostatic polarisation, were carried out to compare the corrosion resistance of UFG samples with that of as cast samples. For good accuracy, potentiodynamic polarisation was performed at a relative low scan rate of 0·5 mV s⁻¹. The frequency range of EIS tests were from 10 kHz to 10 mHz and the amplitude of sinusoidal potential signal was 10 mV with respect to the open circuit potential.

Microstructures of UFG industrial pure Al

Figure 2 presents optical microstructures of the as cast and UFG industrial pure Al. The original grain size distribution (before ECAP) is about 100-200 μm. Owing to severe plastic deformation during the ECAP process, the UFG sample after 16-pass ECAP has a much finer grain than that of as cast sample. It is rather difficult to distinguish grain boundaries of the UFG sample, except that the intensive plastic flow along the shearing direction can be observed under optical microscope.

OPEN IN VIEWER

Figure 2

Optical microstructures of cast and UFG industrial pure Al

The deformed microstructure of UFG industrial pure Al was investigated in detail by TEM. Figure 3 shows TEM microstructure of the UFG industrial pure Al. According to the equivalent strain equation in the Ref. 15, the sample stored nearly more than 16·0 equivalent strains, which leads to a substantial reduction in grain size, from the size of 200 μm (as cast) to 300-500 nm on average (after ECAP). The imposed strain not only leads to reduced grain sizes but also forms many crystalline defects, such as grain boundaries and dislocations. The ultrafined grains are surrounded by clear and regular shaped boundaries, some of which present equal thickness fringes and thus are characterised high angle grain boundaries.16 The high angle grain boundary is non-equilibrium grain boundary, with high internal stored energy. The severe strain creates lots of intragranular dislocations and also provides enough energy for the dislocation movement during the ECAP process (from the inside of the grains to the vicinity of the grain boundaries).17 Because of the high stacking fault energy of the Al matrix, dynamic recovery should occur during the whole ECAP process. Lots of dislocations could be annihilated, and thus there are nearly free dislocations in the inside of some grains. However, due to the extreme strain imposed, there are still a number of dislocations kept in the vicinity of boundaries and inside of some grains.

OPEN IN VIEWER

Figure 3

Microstructures (TEM) of UFG industrial pure Al

Through TEM observation, one can find the existence of high angle grain boundaries in the UFG samples, are not measureable for accurate statistics. However, EBSD in the SEM can be helpful for the study on the microstructure and precise misorientation of the grain boundaries.18 Figure 4a shows the grain morphology of the UFG sample obtained from EBSD. The EBSD measurement uses misorientation of grain boundaries to distinguish two adjacent grains, and then displays these as different colours. The grain morphology obtained from EBSD is similar to the one obtained using TEM (i.e. elongated grains having size of 300-500 nm and distributed along the shearing direction). Figure 4b is the grain boundary misorientation map of the UFG sample. The grain boundaries having misorientations of 2-5, 5-15 and 15-180° are identified by red, green and blue respectively. As well known, those grain boundaries having a misorientation larger than 15°can be regarded as high angle grain boundaries.19 From the map, it is clear that most of the grain boundaries are marked by blue, which indicates the extremely high fraction of high angle grain boundaries of the UFG industrial pure Al. Figure 4c shows a histogram of the misorientation of grain boundaries in the UFG sample. About 83·5% of the grain boundaries have a misorientation larger than 15°, with some grain boundaries having a misorientation larger than 60°. From the EBSD data, one can believe that the grain boundary structure of industrial purity Al will be severely changed during grain refinement by ECAP, and the resulting fraction of high angle grain boundaries is extremely high.

OPEN IN VIEWER

Figure 4

Misorientation of grain boundaries of UFG industrial pure Al obtained from EBSD

Corrosion resistance in constant immersion tests

Constant immersion test is a most direct and illustrative way to detect corrosion behaviour of a material. Figure 5 shows optical micromorphologies of the as cast and UFG industrial pure Al samples after immersion in 3·5 wt-%NaCl solution for 10 days. It is clear that corrosion of the UFG and as cast industrial pure Al both reveal pitting, which implies the breakdown of the passive film on the sample surface. From the optical micromorphologies, one can clearly find that the UFG sample suffered reduced pitting damage, which presented as fewer and smaller pits compared with the as cast sample. Figure 6 shows SEM morphologies of typical pits in the as cast and UFG industrial pure Al sample surfaces after immersion in 3·5 wt-%NaCl solution for 10 days. The pit size on the as cast sample surface is about 15 μm and is much larger than those from the UFG sample surface (5 μm). From the optical and SEM morphologies, one can find that the UFG industrial pure Al also revealed pitting corrosion behaviour, but the corrosion resistance is obviously improved compared to as cast samples.

OPEN IN VIEWER

Figure 5

Optical pitting morphology of cast and UFG industrial pure Al after immersion in 3·5 wt-%NaCl solution for 10 days

OPEN IN VIEWER

Figure 6

Morphologies (SEM) of typical pits on surface of cast and UFG industrial pure Al samples after immersion in 3·5 wt-%NaCl solution for 10 days

Corrosion resistance in electrochemical tests

Electrochemical tests were performed to get electrochemical data of the UFG industrial pure Al. Electrochemical impedance spectroscopy was conducted in order to study the electrochemical characteristic of the passive film of the industrial pure Al in the NaCl solution. Figure 7 shows Nyquist plots of the impedance spectra of the UFG industrial pure Al in 0·01 mol L⁻¹ NaCl solution. According to the literature,20 passive film formation is the major reaction on the sample surface when Al and Al alloys were immersed in water solution during the initial stage. The passive film is mainly composed of Al₂O₃.H₂O, whose density and stability correlate with the corrosion resistance of Al and Al alloys. From the plots, both samples have single capacitive arc in the middle and low frequencies, and no inductive arc or tail are found. It indicates that the passive films on both sample surfaces are intact, and that no stable pits happen in the initial immersion stage in NaCl solution. Several references (for example, Ref. 21) relate this capacitive arc with metal dissolution in the corrosion process, whose diameter is associated with charge transfer resistance, i.e. corrosion resistance. An increased diameter of the capacitive arc corresponds to improved corrosion resistance. As shown in Fig. 7, the diameter of the capacitive arc of the UFG sample is much larger than that of as cast one. It indicates the larger resistance of the passive film on the UFG sample surface, which can provide better corrosion resistance of the UFG industrial pure Al matrix against Cl⁻. The Randles equivalent circuit, illustrated in the right corner of the Fig. 7, was used to fit the experimental Nyquist plots of the as cast and UFG samples by ZsimpWin commercial software (USA) in order to get circuit parameters polarisation resistance R _p and film capacitance C for the capacitive arc. The R _p value of the UFG sample is 87 kΩ cm², more than twice as high as the value for the as cast sample (36 kΩ cm²). From the larger diameter of the capacitive arc and increased R _p value, the resistance of the passive film on the UFG sample surface is improved compared with that of the as cast material.

OPEN IN VIEWER

Figure 7

Nyquist plots of impedance spectra and selected equivalent circuit of cast and UFG pure Al in 0·01 mol L⁻¹ NaCl solution

Figure 8a shows potentiodynamic polarisation curves of the UFG and as cast industrial pure Al samples tested around their respective stable open circuit potentials. Both samples show a degree of passivity in the tested 0·01 mol L⁻¹ NaCl solution, resulting from the presence of the air formed passive film. The greater initial steepness of the anodic polarisation curve indicates that the UFG industrial pure Al has improved passivity compared to the as cast one. Additionally, as shown by the dashed line in Fig. 8a, the pitting potential E _pit of the UFG sample [about −500 mV(SCE)] is much more positive than that of the as cast sample [about −600 mV(SCE)]. This implies that the UFG industrial pure Al has the lower susceptibility to pitting corrosion. Besides the more positive E _pit value, the UFG sample also has a more noble corrosion potential, E _corr [−670 mV(SCE)] than the as cast material [−703 mV(SCE)]. Herein, from the more positive E _pit and E _corr values, one can believe that the UFG industrial pure Al has improved pitting corrosion resistance than the as cast one.

OPEN IN VIEWER

Figure 8

Polarisation curves of cast and UFG industrial pure Al in 0·01 mol L⁻¹ NaCl solution

The potentiostatic polarisation test was investigated to study pit propagation on the sample surfaces under a fixed potential, by studying the evolution of the current density. The selected potential is −500 mV, which is the pitting potential of the UFG sample. Figure 8b shows potentiostatic polarisation curves of the UFG and as cast industrial pure Al samples. Clearly, both samples present increasing current density during the whole test, which correlates with stable pit propagation and corrosion damage on their surfaces. The as cast sample has a much larger current density during the whole test (about five times larger than that of the UFG sample). In the initial stage, the current density of the as cast sample increases rapidly, which indicates the rapid increase in the number and size of stable pits. During the following stage, the as cast sample presents a relatively lower increase in current density, and this indicates the propagation in the pits depth. In contrast, the UFG sample has a much lower increase in the current density during the whole test. Figure 9 presents SEM pitting morphologies of the as cast and UFG industrial pure Al after potentiostatic polarisation. The arrow marking the white and black spots are corrosion induced pits, in which the white ones are pits covered by a corrosion product. According to the pits number and pits sizes, one can easily find that the UFG sample suffered reduced pitting corrosion damage during the test and shows better pitting corrosion resistance compared to the as cast one.

OPEN IN VIEWER

Figure 9

Pitting morphology (SEM) of cast and UFG industrial pure Al after potentiostatic polarisation test

Effect of ECAP induced crystalline defects on passive film

As well known, the oxide films or other passive films are prone to nucleate at the crystalline defects, such as grain boundaries and dislocations.22 From the present TEM and EBSD observation, ECAP brings to the UFG industrial Al not only significant grain refinement but also lots of crystalline defects with high internal energy, such as large fractions of high angle grain boundaries and dislocations. Those energetic crystalline defects provide more stored energy for the growth of the passive film. When the UFG sample is immersed in aqueous solution, a more rapid growth of the passive film will occur, which will grow to a thicker value than the as cast material. A similar phenomenon can be found in ECAP fabricated UFG pure Ti in the literature.23 This passive film formed on the UFG sample surface will provide better protection against Cl⁻ and present better stability in aggressive solution compared to that of the as cast one. Thus, it will take more time for Cl⁻ to penetrate the thicker and denser passive film resulting in a reduced passive current density and the development of a clear pitting potential in chloride media.

The EIS results verify the effect of energetic crystalline defects on the passive film. Thus, from the single capacitive arc, one can know that the passive films on both sample surfaces are relative intact, and there is no significant anodic dissolution. The UFG sample has much larger diameter of capacitive arc and R _p value compared to the as cast alloy. Those improved electrochemical characters should be caused by the rapider, thicker and denser passive film, which benefits from the large fractions of high angle grain boundaries and dislocations of the UFG Al matrix.

Effect of impurity particles refinement on corrosion resistance of industrial pure Al

Seri and Furumata indicated that Si containing second phase particles are a major cause for pitting corrosion of industrially pure Al.24 Thus, the quantity, size and distribution of the Si containing particles will greatly affect the corrosion resistance of the industrial pure Al. The harmful effects of these particles are as follows. First, they reduce the integrity and insulation of the passive film. The oxide film on the Al matrix is a good insulator while second phase particles (e.g. silicon) must be considered as semiconductor with a reduced thickness of (or indeed no significant) passive air formed aluminium oxide film. Thus, when the second phase impurity particles are distributed on the surface, it inevitably weakens the integrity and insulation of the passive film. Second, Si (and Fe) containing second phase particles have more positive electrode potentials than that of Al matrix, and consequently will tend to act as local cathodes to induce pitting corrosion.

Benefiting from the severe plastic deformation, the morphology and distribution of the second phases and impurity particles will be changed significantly with the grain refinement. Chung's research indicated that second phase impurity particles in 1050 Al alloy were significantly refined (reduced in size) by ECAP, resulting in a distribution that was relatively uniform.25 Here, SEM and EDS spectra measurements were conducted to observe Si containing impurity particles on the surface of the UFG and as cast industrial pure Al, for verifying the effect of ECAP on the refinement of Si containing particles. Figure 10 presents the morphology and EDS spectra of the typical Si containing particles on the as cast and UFG sample surfaces. The EDS spectra show that the observed particles on both surfaces are Si containing and that the size of a typical second phase particle in the UFG sample surface is ∼3 μm, which is much smaller than in the as cast alloy (∼8 μm).

OPEN IN VIEWER

Figure 10

Images (SEM) and spectra (EDS) of typical Si containing impurity particles observed on surface of cast and UFG industrial pure Al

The refinement of the Si containing particles induced by ECAP improves the UFG sample's corrosion resistance. First, it will improve the integrity and insulation of the passive film. Second, it will reduce the susceptibility of pitting corrosion via the classical ‘small cathode, large anode’ mechanism. The elevated pitting potential verifies its effect on enhancing pitting corrosion resistance of the UFG industrial pure Al visually.

Using UFG+post- annealed sample to verify the effect of two factors

As discussed above, ECAP resulted in two factors that enhanced the corrosion resistance of the UFG industrial pure Al. First, benefiting from a high stored energy, the ECAP UFG alloy has a thicker and denser passive film compared with the as cast alloy, and this passive film presents excellent stability in the NaCl solution. Second, the refinement of the Si containing impurity particles reduces the susceptibility of the UFG industrial pure Al to local pitting corrosion.

For better verification of the beneficial effect of the two discussed factors, the UFG sample was post-annealed at 623 K for 6 h and the subsequent change in the microstructure and corrosion resistance were also studied. Figure 11 shows the optical microstructure of the UFG alloy as well as the post-annealed alloy. The microstructure of the latter material was significantly changed due to complete recrystallisation. Thus, the original ultrafine grain, but heavily deformed, microstructure becomes equiaxed with grain size about 10-30 μm. Although post-annealing will completely change the microstructure of the UFG sample, the refined Si containing particles will not be recombined. Therefore, the post-annealed sample retained the refined impurity particles. Figure 12 presents electrochemical characteristic curves of the post-annealed sample in 0·01 mol L⁻¹ NaCl solution, where Fig. 12a is Nyquist plots of the EIS test and Fig. 12b is potentiodynamic polarisation curves. The values of fitted R _p, E _pit and E _corr for the post-annealed sample are 60 kΩ cm², −550 mV(SCE) and −680 mV(SCE) respectively, which remain higher than the as cast alloy, although they are lower than those of the ECAP alloy. Higher values for R _p, E _pit and E _corr indicate that the post-annealed sample retains an improved corrosion resistance compared with the as cast sample. This phenomenon verifies the beneficial effect of the refinement of Si containing impurity particles to the corrosion resistance of UFG sample.

OPEN IN VIEWER

Figure 11

Optical microstructure of 16-pass ECAPed+623K/6 h annealed industrial pure Al

OPEN IN VIEWER

Figure 12

Electrochemical properties of ECAP+annealed industrial pure Al