Chemical Resistance of Cu-Al-Mn Superelastic Alloy Bars in Acidic and Alkaline Environments

The CAM SEA specimens were prepared by Furukawa Techno Material, Hiratsuka, Japan. The nominal composition of the CAM SEA used in this study was Cu-16.9 at% Al-11.6 at% Mn. The solution treatment was conducted at 900°C, followed by quenching in water, and subsequent aging at 150°C to stabilize the superelastic property. As mentioned previously, the control material used in this study is SS-400 steel, a mild steel conforming to JIS G 3101 (JSA 2015), which is equivalent to Grade E275 of ISO 630:1995 (ISO 1995). SS-400 has a minimum yield stress of 245 MPa and an elongation of 17%. The specific carbon content is not indicated in JIS G 3101 (JSA 2015); and JIS G 3101 (JSA 2015) limits the phosphorus and sulfur content of SS-400 to 0.05% by weight.

Two types of specimens, Type A and Type B (Fig. 1), were prepared for both CAM SEA and SS-400. In addition, another specimen, Type C, was prepared only for CAM SEA. Type A specimens were $4\times 4\times 100\text{-}\mathrm{mm}$ prismatic specimens, Type B specimens were prepared according to ISO 6892-1 (ISO 2016), and the Type C was a $10\times 10\times 5\text{-}\mathrm{mm}$ prismatic specimen. Type A specimens were used to determine the chemical resistance of the metals via microscopy and weight change measurements. Type B specimens were used to measure the changes in the mechanical properties, e.g., yield load and residual strain, due to chemical activity. The Type C specimen was used to examine the potentiodynamic polarization behavior of CAM SEA. Type A and Type B specimens had different shapes because the square cross section of Type A specimens made it easier to perform imaging and calculating the area loss due corrosion/erosion. Furthermore, the specimens did not have to be as large to allow for fixing in a tensile testing machine. On the other hand, in Type B specimens, sufficiently large samples with a reduced gauge length were needed to properly attach the specimens in the testing machines and ensure that the failure occurs in the gauge length of the specimens and not at the gripping locations.

Fig. 2 schematically illustrates one cycle of salt spray and immersion tests for measuring the chemical resistance of CAM SEA and SS-400. Salt spray represents the most realistic exposure mechanism; therefore, it is the most commonly used test method for studying the corrosion resistance of structural metals. Therefore, salt solution spraying, based on ISO 9227:2017 (ISO 2017) (unless otherwise indicated), was adopted here to study the NaCl resistance of metals. Each cycle of the salt spray test was carried out by spraying a 10% by weight NaCl solution on the specimens and then drying for 2 h in a fan oven. In contrast, to better match the actual exposure conditions for the types of applications mentioned previously, immersion tests were conducted rather than solution spraying for other acidic and alkaline solutions. For the immersion tests, one cycle consisted of immersing the specimens in the respective chemical solution and then storing them at 40°C for 1 day, then removing them from the solution and drying at 40°C for 1 day in a fan oven. The corroded bars were washed with an ammonium hydroxide solution to remove the corrosive products after each cycle. The weight change of the specimens was measured after the washing procedures. The chemical solutions used in immersion testing were three types of 5% by weight inorganic acid solutions, namely sulfuric acid, nitric acid, and hydrochloric acid; 5% sodium hydroxide (NaOH) solution as an alkaline solution; and tap water. The authors did not follow any standard in the immersion tests because they were not aware of any standardized test that reflects the real exposure conditions (repetition of immersion and drying) for civil engineering structures. The most prominent corrosive reagents were selected with higher concentrations not prescribed in any standard available. These conditions were selected to be conservative for common applications in civil engineering structures but are highly likely in industrial and chemical plants, including buildings disposing of chemical solvents that may leak from the waste disposal pipelines, in addition to acid rain and other super alkaline materials such as geopolymers. In these tests, nominally 50 cycles and 5 cycles were repeated for each salt spray and immersion test, respectively, of both CAM SEA and SS-400 specimens. However, the exact number of cycles was based on the weight change (corrosion state) of the control SS-400 bars. In most studies, the testing is stopped after a weight change of more than 20%; however, this study conducted tests to extreme conditions until the SS-400 specimens degraded completely.

Three Type A and three Type B specimens each were subjected to cyclic salt spray and immersion testing. For Type A specimens, a weight change measurement was taken after every 10 cycles in the salt spray test and after every cycle in the immersion tests. Upon completing the chemical aging, the specimens were sliced, mounted on conductive resin, and mirror polished. After imaging the whole cross section with an optical microscope, the corroded portion was imaged with a scanning electron microscope (SEM) and analyzed using energy-dispersive X-ray spectroscopy (EDS). The following parameters were used in the SEM and EDS: accelerating $\mathrm{voltage}=15\mathrm{kV}$, $\mathrm{beam}\mathrm{current}=3\mathrm{nA}$, and $\mathrm{spot}\mathrm{size}=30\mathrm{\mu m}$. In the EDS analysis, Cu, Al, Mn, and O were mapped for the CAM SEA specimens, and Fe and O were mapped for the SS-400 specimens.

Type B CAM SEA specimens were subjected to one cycle of tensile loading up to 6% strain prior to and after the chemical aging. After measuring the dimensions of the specimens, they were loaded in displacement control up to 6% strain and unloaded in load control until complete removal of the tensile load. The loading rate was ${10}^{-3}\mathrm{mm}{\mathrm{s}}^{-1}$. The yield load and the residual strain were obtained before and after the chemical resistance tests and compared to evaluate the effect of different solutions on the mechanical properties of the specimens. Prior to each chemical exposure, one Type B SS-400 specimen was tested using the same loading protocol as for the CAM SEA to obtain the yield load and residual strain. After the chemical exposure, three other SS-400 specimens were tested to study the corrosion-induced degradation.

The conditions for the electrochemical study of the Type C CAM SEA specimen were as follows. A platinum counter electrode and a saturated silver–silver chloride reference electrode was used. The working electrode was prepared by embedding the Type C CAM SEA specimen into an epoxy resin mount to leave one exposed face with a surface area of $1{\mathrm{cm}}^2$. The surface was finished by wet polishing and buffing. The test was performed in a cell containing $800{\mathrm{cm}}^3$ 3.5% by weight NaCl at 30°C. A potentiodynamic polarization measurement was performed at a scan rate of $0.33\mathrm{mV}{\mathrm{s}}^{-1}$ and an applied voltage of 0.5 V. ISO 10993-15 (ISO 2019) recommends a scan rate of $1.0\mathrm{mV}/\mathrm{s}$ for potentiodynamic measurements of metals and alloys. The scan rate used by Gojić et al. (2011) and Alfantazi et al. (2009) was $0.5\mathrm{mV}/\mathrm{s}$ and $1.0\mathrm{mV}/\mathrm{s}$, respectively. The scan rate of $0.33\mathrm{mV}/\mathrm{s}$, recommended in JIS G 0577 (JSA 2014) and JIS T 0302 (JSA 2000), was found most applicable for the purposes of this study.

Fig. 3 represents the physical conditions of the tested specimen at the end of the chemical resistance tests. During the salt spray test, the SS-400 specimens started rusting after as few as two cycles, and rust constantly increased until 50 cycles. However, the CAM SEA specimens did not show any rusting. In the immersion tests, CAM SEA specimens had remarkable acid resistance to ${\mathrm{H}}_2{\mathrm{SO}}_4$, ${\mathrm{HNO}}_3$, and HCl solutions, with a slight decrease in cross-sectional area of the specimens after five cycles (Fig. 4). Despite significant resistance of SS-400 specimens to the ${\mathrm{H}}_2{\mathrm{SO}}_4$ and HCl solutions, heavy corrosion occurred in the ${\mathrm{HNO}}_3$ solution, leading to the production of a thick layer of red rust on the surface of the SS-400 specimens. Both CAM SEA specimens and SS-400 had excellent chemical resistance to 5% NaOH alkaline solution, showing no corrosion and no dimensional change. The same phenomenon was observed for the CAM SEA specimens in ${\mathrm{H}}_2\mathrm{O}$, whereas the SS-400 specimens corroded with a thick layer of rust formed on their surfaces.

Fig. 4 shows micrographs of the cross sections of Type A specimens after completion of chemical aging. An insignificant reduction (erosion) of the specimen cross section occurred in both the CAM SEA and SS-400 specimens when subjected to the salt-spray. CAM SEA specimens had remarkably superior chemical resistance compared with the SS-400 specimens in the immersion tests with 5% by weight of ${\mathrm{H}}_2{\mathrm{SO}}_4$, ${\mathrm{HNO}}_3$, and HCl acid solutions. Of the acidic solutions, both CAM SEA and SS-400 specimens had the most severe reaction with ${\mathrm{HNO}}_3$. The average corrosion depths of SS-400 specimens were 2 and 1.5 mm in the HCl and ${\mathrm{H}}_2{\mathrm{SO}}_4$ solutions, respectively, whereas heavy erosion occurred in the entire cross section of the SS-400 specimens in the ${\mathrm{HNO}}_3$ solution. On the other hand, the maximum corrosion depth of CAM SEA specimens in the acidic solutions was about 1 mm, which occurred in the ${\mathrm{HNO}}_3$ solution. The performance of CAM SEA and SS-400 specimens was comparable in the NaOH solution and water, the resistance to both of which was superior to the resistance to other chemicals.

Fig. 5 illustrates the weight change of Type A CAM SEA and SS-400 specimens subjected to chemical aging. To produce these graphs, an average value of the weight changes of the three SS-400 specimens and the three CAM SEA was used. The error bars show one standard deviation above and below the average value. An average increase of 0.3% and 1.2% was observed in the weight of CAM SEA and SS-400 specimens, respectively, at the end of the salt spray test. The SS-400 specimens had a progressive increase in corrosion, leading to an increase in the weight. However, the CAM SEA specimens had limited corrosion.

After five cycles in the immersion test with the 5% acid solutions, the CAM SEA specimens had a weight change of $-0.07\%$, $-11.6\%$, and 0.83% when subjected to 5% by weight ${\mathrm{H}}_2{\mathrm{SO}}_4$, ${\mathrm{HNO}}_3$, and HCl solution, respectively. These values for the SS-400 specimens were $-45.9\%$,$-19.2\%$, and $-36.8\%$, respectively, demonstrating a drastic loss of weight in the SS-400 specimens subjected to acidic solutions. These results indicate that the CAM SEA specimens had superior acid resistance compared with the SS-400 specimens, which had the lowest resistance to 5% ${\mathrm{HNO}}_3$ solution. In 5% by weight alkaline NaOH solution, the CAM SEA and SS-400 specimens had limited weight change, with an average increase of 0.27% and less than 0.1%, respectively, after five cycles of the immersion test. Similarly, the weight change of both types of materials in water was less than 0.4%.

Some specimens increased in weight. This weight increase was due to the formation of a loose surface layer. This might give an impression that the cross section of CAM SEAs increased; however, this surface layer was loosely attached to the base metal and did not contribute to the mechanical strength, as demonstrated in the section “Tensile Tests.” Although it is common to remove the loose surface layer in mass loss tests, this approach was not followed in this paper to clearly differentiate between the chemical composition of the base metal and that of the surface layer as shown later in the section “SEM and EDS Analysis.”

Fig. 6 shows the SEM and EDS-SEM mapping charts for Cu, O, Al and Mn elements of Type A CAM SEA after the chemical aging. The mapping charts show formation of the surface layer with Cu and O components on the CAM SEA specimens due to the salt spray test. As with the salt spray test using NaCl solution, the CAM SEA specimens exposed to 5% by weight of ${\mathrm{H}}_2{\mathrm{SO}}_4$ and HCl had formation of a similar surface layer. However, the oxidation progressed and deeper erosion occurred in the specimens exposed to aggressive acidic solution of 5% by weight of ${\mathrm{HNO}}_3$. Regarding the specimens subjected to the NaOH solution and water in the immersion test, the surface of CAM SEA specimens had a deposit of a surface layer similar to that in the salt spray test. Fig. 7 shows the SEM and EDS-SEM mapping charts for Fe and O elements of SS-400 specimens after completion of the chemical resistance tests. Deep erosion of the surface of SS-400 specimens occurred with the ${\mathrm{H}}_2{\mathrm{SO}}_4$, ${\mathrm{HNO}}_3$ and HCl acid solutions due to aggressive oxidation, showing the poor resistance of SS-400 steel to the acidic solutions, particularly to the ${\mathrm{HNO}}_3$ solution. The results were similar for NaOH and water exposure.

Figs. 8 and 9 compare the load-strain diagrams from tensile testing of CAM SEA and SS-400 specimens before and after the chemical aging. Only one of each set of three specimens is shown in Figs. 8 and 9 for clarity. The other samples had similar behavior. Insignificant degradation of strength and superelasticity occurred in the salt spray test and the submersion tests with both acidic and alkaline solutions, as well as water, except for the case of ${\mathrm{HNO}}_3$ (Fig. 8).

As expected, the load-carrying capacity of all the CAM SEA specimens subjected to the chemical aging was reduced because of a reduction in the cross-sectional area. The reduction of the yield load of CAM SEAs for salt spray and the immersion tests in the NaCl, ${\mathrm{H}}_2{\mathrm{SO}}_4$, ${\mathrm{HNO}}_3$, HCl, and NaOH solutions and in water were on average 9.7%, 5.6%, 35.2%, 9.8%, 11.9%, and 4.7%, respectively. The SS-400 specimens subjected to the salt spray test and the immersion in NaOH solution and water had an insignificant increase in yield load. However, the yield load of SS-400 specimens significantly decreased when subjected to immersion in acidic solutions. The average yield load decreased to 48.6%, 42.1%, and 42.3% as a result of exposure to ${\mathrm{H}}_2{\mathrm{SO}}_4$, ${\mathrm{HNO}}_3$, and HCl solutions, respectively.

Fig. 10 shows the potentiodynamic polarization curves of Type C CAM SEA and SS-400 samples. The corrosion current density $i_{\mathrm{corr}}$ and the corrosion potential $E_{\mathrm{corr}}$ were obtained as $i_{\mathrm{corr}}=3.0\mathrm{\mu A}{\mathrm{cm}}^{-2}$ and $E_{\mathrm{corr}}=-0.23\mathrm{V}$ using the extrapolated Tafel lines. These values were of the same order of magnitude as those of Cu-Al-Ni SMAs reported by Alfantazi et al. (2009), i.e., $i_{\mathrm{corr}}=2.75\mathrm{\mu A}{\mathrm{cm}}^{-2}$ and $E_{\mathrm{corr}}=-0.33\mathrm{V}$. The corrosion parameters for mild steel were $i_{\mathrm{corr}}=30\mathrm{\mu A}{\mathrm{cm}}^{-2}$ and $E_{\mathrm{corr}}=-0.42\mathrm{V}$. Based on the Faraday’s law and the experimentally obtained value of $i_{\mathrm{corr}}$, the erosion rate of CAM SEA and SS-400 were calculated as 0.032 and $0.35\mathrm{mm}/\mathrm{year}$, respectively.

No significant reduction was observed in the weight and strength of CAM SEA [Figs. 5(a) and 8(a), respectively]. Similar observation can be made for the mild steel from these figures. On the other hand, it was observed from visual inspection [Fig. 3(a)] that rusting in mild steel was much more prevalent than that in the CAM SEA. This observation is supported by the potentiodynamic measurements (Fig. 10), in which the corrosion rate of CAM SEA was estimated to be about one-tenth that of mild steel.

The corrosion performance of CAM SEA was found to be close to that of Cu and other Cu-based alloys because corrosion parameters, e.g., corrosion current density and corrosion potential, obtained from the potentiodynamic measurements were of the same magnitude as those for Cu and other Cu-based alloys such Cu-Al-Ni SMAs (Alfantazi et al. 2009; Gojić et al. 2011). Similar to other Cu-Al alloys, the mechanism underlying the superior performance of CAM SEA over the mild steel is the formation of surface layers including Cu, Al, Mn, and O, the existence of which was confirmed from the EDX image for NaCl exposure (Fig. 6). The formation of cuprous oxide, aluminum oxide/hydroxide, manganese oxide, and cuprous chloride in the surface layers was inferred from the literature (Alfantazi et al. 2009; Gojić et al. 2011). In the potentiodynamic polarization curve of CAM SEA (Fig. 10), two peaks and a reduction in the current density, $i$, were observed when the potential, $E$, was positive. This suggests the protective effect of the surface layers mentioned previously, although the slight increase in $i$ for $E>0.25\mathrm{V}$ suggests that the surface layers are not fully passive.

No significant change was seen in the weight and strength of both CAM SEA and mild steel [Figs. 5(e) and 8(e)] after immersion in dilute NaOH solution. From these results, it is concluded that similar to mild steel, CAM SEA is suitable for use as reinforcing elements not only in ordinary portland cement concrete, which has an alkaline environment with a pH value of about 13, but also in geopolymer concrete, the pH value of which could exceed 13.

The formation of a thin surface layer with Cu and O was seen in the EDX image for NaOH exposure (Fig. 6). The black coating observed in the visual inspection [Fig. 3(e)] suggests that the surface layer consisted mostly of CuO. This corrosion mechanism is similar to that of Cu and other Cu-based alloys exposed to dilute alkaline solutions.

After immersion in dilute ${\mathrm{H}}_2{\mathrm{SO}}_4$ and HCl solutions, no significant change in weight and strength occurred for CAM SEA from visual inspection [Figs. 3(b and d)], the weight change measurements [Figs. 5(b and d)], and the tension test results [Figs. 8(b and d)]. On the other hand, significant reduction in weight and strength occurred when CAM SEAs were immersed in dilute ${\mathrm{HNO}}_3$ solution [Figs. 3(c), 5(c), and 8(c)]. These responses are similar to those of Cu. In mild steel samples, severe reduction of weight and strength occurred in all the acidic solutions tested. These results indicate that CAM SEAs have corrosion performance comparable to that of Cu and much better than that of mild steel against the dilute acidic solutions tested in this paper.

Although a surface layer with Cu and O was observed in the acidic solution cases of ${\mathrm{H}}_2{\mathrm{SO}}_4$, ${\mathrm{HNO}}_3$, and HCl (Fig. 6), CAM SEAs were not very resistant to dilute ${\mathrm{HNO}}_3$ solution, which resulted in heavy oxidization. This observation is supported by the formation of a thick layer with Cu and O (Fig. 6).