Corrosion of Cu in Antifreeze Solutions with Triazine- or Triazole-Type Corrosion Inhibitors for 3 Weeks

Abstract

The corrosion behavior of Cu in antifreeze solutions containing 2,4,6-Tris(5-carboxypentylamino)1,3,5-triazine, 2,4,6-Tris(11-carboxyundecylamino)1,3,5-triazine, 1-Aminomethyl(N′,N′-di(2-hydroxyethyl)tolutriazole, or 1-Aminomethyl(N′,N′-di(2-hydroxyethyl)benzotriazole as corrosion inhibitors were examined by immersion test for 3 weeks as well as potentiodynamic polarization tests before and after immersion test. The corrosion rate of Cu was as relatively high as 10⁻⁵ A/cm² in antifreeze solution with the inhibitor (2,4,6-Tris(11-carboxyundecylamino)1,3,5-triazine) with a high molecular weight of 713 for a short time duration compared with antifreeze solutions using the other three types of inhibitors. However, the corrosion inhibition effect of this large molecule became prominent after 2 weeks, reducing the corrosion rate by about four orders of magnitude. Corrosion of Cu in the solution with inhibitors of high molecular weight of 440 or higher decreased gradually with time, while that in the solution with small molecules slightly increased over 3 weeks.

1. Introduction

Antifreeze solution protects engines, radiators, and pumps from corrosion as it controls the temperature of the automotive engines. It consists of glycol, which transfers heat and additives, including corrosion inhibitor, deformer, and antioxidant. Recent industrial trends require a high level of corrosion inhibition for the antifreeze to protect various materials in the circulatory system and to assure long-term quality.

Some of the common corrosion inhibitors, including many amine-type inhibitors and 2-ethylhexanoic acid (2-EHA), need to be replaced because of toxicity [1,2,3,4]. Global interests in sustainable technology urge the development of new corrosion inhibitors that are environmentally benign and highly resistant to corrosion.

The effectiveness of organic corrosion inhibitors depends greatly on the adsorption property. Organic compounds with heterocyclic groups, such as benzotriazole and triazine, make strong chemical coordination with Cu ions [5], and they can successfully cover Cu surfaces. Benzotriazoles also have high thermal stability [6], which is required for uses with heat-exchanging liquids. Triazine has been reported to have active sites on N atoms and aromatic rings [7], facilitating the adsorption of the molecules on a metal surface.

We synthesized triazole- and triazine-type compounds and evaluated their corrosion-inhibiting effects on Cu in antifreeze solutions. The electrochemical corrosion behavior of Cu was measured in fresh solution and in solution aged for 3 weeks with Cu immersed in the solution to assess time-driven changes, considering the long operation time of antifreeze in automobiles.

2. Experimental Procedures

Six test solutions (

Table 1. The compositions of antifreeze solutions used in this study (Inhibitors A, B, C, and D are defined in Table 2).

Contents	Solution no1	Solution no2	Solution no3	Solution no4	Solution no5	Solution no6
Corrosion inhibitor	3 g (0.89 wt.%) A	3 g (0.89 wt.%) B	3 g (0.88 wt.%) B + 0.7 g (0.21 wt.%) C	3 g (0.88 wt.%) B + 0.7 g (0.21 wt.%) D	1 g (0.30) wt.% D	2 g (0.59 wt.%) D
Benzoic acid	25 g	25 g	12.5 g	12.5 g	-	-
Benzotriazole	-	-	12.5 g	12.5 g	25 g	25 g
Sodium molybdate	0.5 g	0.5 g	0.5 g	0.5 g	-	-
Sodium nitrite	-	-	-	-	0.3 g	0.3 g
Ethylene glycol	300 g	300 g	300 g	300 g	300 g	300 g
Water	10 g	10 g	10 g	10 g	10 g	10 g

) were made by adding four kinds of corrosion inhibitors into the commercial ethylene glycol-based antifreeze solutions, which contain molybdate or nitrite, made by DongA Specialty Chemicals Corporation, Republic of Korea. Unfortunately, the detailed compositions of commercial solutions cannot be stated here due to a business issue. Solutions no1~no4 were made of the antifreeze solution with molybdate, and solutions no5 and no6 were made with nitrite.

The corrosion inhibitors in this study were variations of triazole- [6,8,9] and triazine-type [7,10] inhibitors. 2,4,6-Tirs(5-carboxypentylamino)1,3,5-triazine, 2,4,6-Tris(11-carboxyundecylamino)1,3,5-triazine, 1-Aminomethyl(N′N′-di(2-hydroxyethyl)tolutriazole, and 1–Aminomethyl(N′,N′-di(2-hydroxyethyl)benzotriazole were synthesized and added into the antifreeze solutions with the concentration of 0.2~1 wt.%, as described in Table 1. Their molecular structure and molecular weight are shown in

Table 2. The corrosion inhibitors used in this study.

Denomination	Chemical Name	Structure	Molecular Weight
A	2,4,6-Tris(5-carboxypentylamino)1,3,5-triazine		440
B	2,4,6-Tris(11-carboxyundecylamino)1,3,5-triazine		713
C	1-Aminomethyl(N′,N′-di(2-hydroxyethyl)tolutriazole		250
D	1-Aminomethyl(N′,N′-di(2-hydroxyethyl)benzotriazole		236

.

A Cu (99.9 wt.%) sheet of 20 mm × 20 mm × 2 mm size, commercially finished to have a smooth surface, was used as a specimen. The corrosion of specimens was examined by immersion test or by potentiodynamic polarization measurement.

For the immersion tests, the specimen was immersed in a test solution of 250 mL and kept in an oven at 98 °C for 3 weeks. The pH of each solution was monitored during the test. The concentrations of metallic ions were measured every week by ICP-OES (Inductively Coupled Plasma Optical Emission Spectrometry).

Electrochemical potentiodynamic polarization tests were performed with a three-electrode cell equipped with a Cu working electrode, a graphite counter electrode, and Ag/AgCl reference electrode. The polarization behaviors of Cu specimen in the fresh test solutions and of the Cu specimen used in the immersion test in the solution were aged for 3 weeks for the purpose of assessing the persistency of inhibiting effect.

The surfaces of Cu specimens after immersion corrosion tests were analyzed by AFM (Atomic Force Microscopy) for the morphology and the roughness. The phase and composition of corrosion products were examined by XRD (X-ray Diffraction) and XPS (X-ray Photoelectron Spectroscopy), respectively.

3. Results and Discussion

The AFM images for the Cu sheet surface before and after immersion in the test solutions over 3 weeks are shown in Figure 1. The specimen surface before tests (Figure 1a) has shallow lines formed possibly by the grinding process. After tests, Figure 1b–g show different levels of corrosion. Corrosion in solutions no1 and no2 (Figure 1b,c) appears more severe than others, judging from the high roughness and generally damaged morphology. In solutions no3 and no4 (Figure 1d,e), the surfaces were relatively smooth and clean, possibly indicating excellent corrosion inhibition. The specimen immersed in solutions no5 and no6 (Figure 1f,g) featured moderate corrosion damage.

The concentration of inorganic elements in the solutions was measured every week, and the results are shown in

Table 3. The concentration of inorganic elements in the antifreeze solution after Cu immersion, measured by ICP-OES.

Week	Solution	Cu	K	Mo	Na	Si	Ca
1	1	-	11,696	417	159	3	-
	2	-	11,727	428	449	6	-
	3	-	11,567	416	445	7	-
	4	-	11,421	406	400	8	-
	5	16.9	16	4	4272	-	-
	6	13.8	-	5	4237	4	-
2	1	-	11,846	422	153	3	-
	2	37.2	12,321	441	455	4	-
	3	8.6	11,896	426	448	8	-
	4	0.1	11,476	416	427	5	-
	5	37.6	26	4	4322	-	-
	6	40.3	-	5	4409	-	-
3	1	3.4	11,885	430	162	3	-
	2	7.3	12,480	449	468	6	3.2
	3	16.3	11,949	438	456	11	4
	4	5.2	11,327	411	401	6	4.5
	5	87.5	-	5	4526	3	0.8
	6	57.9	11.4	-	4394	2	5.6

. Cu was detected only from solution no5 and solution no6 after 1 week, and from the solutions except solution no1 after 2 weeks, and from all solutions after 3 weeks. K, Mo, and Na are constituents of the antifreeze solution (Table 1), and their concentrations were not changed by time. The concentration of Si and Ca are very low, and these elements are considered to be impurities.

The variation of Cu ion concentration is depicted in Figure 2. The concentration of Cu increased with test time in general, except for solution no2 after 2 weeks. The increasing rate up to 2 weeks in solution no2, no5, and no6 was higher than in solution no1, no3, and no4. Therefore, solution no1, no3, and no4 are thought to inhibit corrosion more effectively than the others until the second week.

After 2 weeks, about 40 ppm of Cu ions were dissolved in solutions no2, no5, and no6. Cu⁺ or Cu²⁺ ions are not thermodynamically stable in water with pH 7~13, referring to the Pourbaix diagram [11]. Cu ions form oxides, i.e., Cu₂O or CuO, by reactions such as 2Cu⁺ + 1/2 O₂ + 2e⁻ → Cu₂O, 2Cu⁺ + O₂ + 2e⁻ → 2CuO, 2Cu²⁺ + 1/2 O₂ + 4e⁻ → Cu₂O, Cu²⁺ + 1/2 O₂ + 2e⁻ → CuO. The concentration of the dissolved Cu ion should be reduced by these reactions, and this is thought to be what happened significantly in solution no2 with high Cu concentration. This result is quite similar to previous work [12], in which the Cu concentration increased up to 2 weeks and then was almost constant up to 3 weeks in the antifreeze solution with the equivalent composition of no2 in this study. The same compositions of solutions with No1 and No2 in this work were used for the previous, but the immersion test was performed at 120 °C in [12] but at 98 °C in this work. The Cu concentration in the two solutions after 3 weeks was lower in this study than in the solutions with the same composition used in the previous work. Therefore, it is thought that the corrosivity of antifreeze solutions is stronger at higher temperatures and the inhibiting action of the inhibitors A and B is more effective at lower temperature. Furthermore, it is notable that the Cu concentration in the solution with inhibitor A was lower than with inhibitor B up to 1 week in the previous study [12] but it became higher by prolonged test. Both works imply that inhibitor B is more effective in suppressing Cu corrosion than inhibitor A for a long time of more than 2 weeks.

However, the Cu concentration in solution no5 and solution no6 did not decrease, although the concentration is similar to that in solution no2. Inhibitor B, contained in solution no2, may be more effective in promoting oxide formation with a high Cu concentration in the solution than inhibitor D of solutions no5 and no6. Otherwise, the pH difference (Figure 3) or the content of molybdate and nitrite in the solution might cause this result. The pH of solution no2 is 9.8 after 2 weeks of test, but the pH of solutions no5 and no6 was below 8.0 at all measurement times. The reaction tendency of Cu ions to form oxide in solution no2 is anticipated to be higher than in solution no5 or no6 because of the high pH value and because of the presence of molybdate stabilizing Cu₂O [13]. In comparison with previous work [12] which used the same constituents with the solutions no5 and no6 but where the concentration of inhibitor D was 0.89 wt.% and the temperature was as high as 120 °C, the Cu dissolution was severer in this study. The Cu concentration in the previous work was only 3.1 ppm after 3 weeks, in contrast to 87.5 ppm in no5 and 57.9 ppm in no6. Therefore, it seems that the concentration of the inhibitor is more crucial than the temperature to corrosion.

pH values (Figure 3) were gradually decreased for all test solutions with the prolonged immersion test. The pH of solutions no1~no4 was 10.7~11.0 before the corrosion test and decreased down to 8.9~9.6. In the case of solution no5 and solution no6, the pH decreased from 8.0 to 7.3~7.6. Corrosion can involve an increase in pH in that the oxygen reduction (O₂ + 4e⁻ + 2H₂O → OH⁻) causes an increase of OH^- ion concentration in the solution. However, oxide formation accompanies a decrease of pH, releasing H⁺ ions (2Cu⁺ + H₂O → Cu₂O + 2H⁺, 2Cu⁺ + H₂O → Cu₂O + 2H⁺). Precipitation of Cu hydroxides (Cu²⁺ + 2OH⁻ → Cu(OH)₂ or 2Cu²⁺ + OH⁻ → Cu₂O + H⁺) also makes the pH decrease. Because Cu under corrosion has higher open-circuit potential than hydrogen redox potential, the pH of the solution during Cu corrosion would not depend on the hydrogen reduction reaction. Therefore, the overall tendency of decreasing pH possibly implies that the formation of corrosion product is progressing in this work.

The pH decreasing rate in molybdate-containing solutions (no1~no4) is slightly faster than in nitrite-containing solutions (no5~no6). Solution no3 had the lowest rate of pH change and the highest Cu concentration in the third week among the solutions no1~no4, possibly indicating the slow formation of oxide or corrosion product compared with others. Figure 1c,d suggest that the addition of inhibitor C reduced Cu dissolution significantly in the presence of 0.88~0.89 wt.% of inhibitor B. However, the effects of inhibitor C would not be prominent after prolonged immersion for more than 3 weeks as the Cu concentration continues to increase (Figure 2). Solution no4 with inhibitor B (0.89 wt.%) and inhibitor D (0.21 wt.%) had similar behavior of pH change to solution no2, implying that inhibitor D is thought to have a limited effect on the rate of corrosion product formation. However, the concentration of Cu ion in solution no4 is much lower than in solution no2 until the second week or solution no3 up to the third week, and Figure 1e shows a very smooth surface. These results suggest that inhibitor D highly effectively suppresses Cu ion dissolution.

Solutions no5 and no6 have only inhibitor D with different concentrations. They showed a relatively small change in pH during the test. Given that the Cu ion concentration in the solution was higher in solutions no5 and no6 than the others at all points of time (Table 3 and Figure 2), the oxide is unstable or the ratio of the number of Cu ions used for oxide formation to the amount released into the solution is low in these two solutions. The pH of solution no6 changes most slowly among the six test solutions. It is deduced from the results for solution no4 that inhibitor D effectively promotes the formation of a protective corrosion product. However, the Cu concentration increased more rapidly than before, and the decrease in solution pH stopped after 2 weeks for solution no5. The inhibiting action might be weak because solution no5 has 0.30% of inhibitor D only. In the case of solution no6, which has 0.59% of inhibitor D, the decrease in solution pH was slower and the concentration of Cu ion in the solution was lower than in solution no5. It shows that the slow formation of corrosion products resulted in high Cu ion concentration in the solution. Consequently, the higher concentration of inhibitor D improved the protectiveness of the surface oxide against corrosion.

The XRD patterns (Figure 4) of specimens after 3 weeks of corrosion test confirm the presence of oxide on the surface except for solution no1, which showed only metallic Cu peaks. The peaks for Cu₂O were weaker than the peaks for metallic Cu for most of the test solutions. However, it is suggested from the relative peak intensity for no2 that the sample corroded in this solution has a thicker oxide layer than other samples.

Figure 5 presents the electrochemical polarization curves of fresh Cu specimens in the fresh solutions (Figure 5a) and of corroded Cu specimens in the solutions used for the immersion test (Figure 5b). Most of the curves obtained with the newly prepared Cu sample and the fresh solutions show the corrosion potential around −0.3~−0.2 V_SCE (Figure 5a). The current density was smaller than 1~2 μA/cm² in the whole test range. The polarization curve measured in solution no2 was exceptional in that the anodic current density did not increase with an increase of applied potential, as typically observed from a passivated surface. However, the current density was higher than the other samples. It means that solution no2 is not highly inhibited but is favorable for oxide formation. The cathodic current density at the potentials below −0.2 V_SCE rarely show the dependence on the potential decrease in no2 solution, indicating protective oxide layer or concentration polarization due to the depletion of oxygen on the Cu surface. This result corresponds to the Cu concentration trends described in Figure 2, which showed fast dissolution of Cu followed by consumption of ions.

The polarization curves (Figure 5b) measured with the specimen and solution that has undergone the immersion test for 3 weeks showed large scattering in corrosion potential, but the overall shapes of the polarization curves are equivalent to each other. Both anodic and cathodic current density increased with the increase in overpotentials in every solution. The corrosion potential after 3 weeks was −0.15~0.08 V_SCE (Figure 6b) and higher than for the as-prepared specimen and solution in general (Figure 6a). The surface oxide formed during 3 weeks of immersion inevitably raised the corrosion potential. However, the corrosion rate was around 10⁻⁸~10⁻⁷ A/cm² and was not markedly changed by the immersion test for most of the solutions (Figure 7). The only exception was no2 as the corrosion rate became about four orders lower after 3 weeks of immersion. It seems to be related to the increase in the corrosion potential, which resulted in the surface oxide grown during the immersion period. Solution no2 has 0.89% of inhibitor B, of which the molecular weight is much larger than the other inhibitors in this study (Table 2). Large molecules cover the metal surface effectively and is highly inhibitive, but their diffusion rate will be quite slow in antifreeze solution. Therefore, the inhibition effect of inhibitor B needs a long time to be activated. Inhibitors C and D, present in solutions no3 and no4 in addition to inhibitor B, must adsorb into the metal surface much faster because they are smaller than B. Consequently, the corrosion inhibition mechanism appears to be dominated by inhibitor C in solution no3 and by inhibitor D in solution no4, as the role of inhibitor B is limited.

The molecular weight of inhibitor A is lower than B and higher than C or D. The corrosion rate of Cu in solution no1 was similar to that in solutions no3~no6 with fresh metal and solution (Figure 7a). However, it slightly decreased after 3 weeks of immersion, unlikely for the cases of solutions no3~no6 which show a little increased corrosion rate after 3 weeks. From these results, it is implied that the molecular weight is crucial for the inhibition efficiency of organic inhibitors, as is known commonly, and more than about 440 of molecular weight is required, limited in the knowledge from this study, for ethylene glycol-based antifreeze solutions.

4. Conclusions

The inhibition effects of triazine- or triazole-type inhibitors on the corrosion of Cu in antifreeze solutions were examined by immersion test for 3 weeks as well as potentiodynamic polarization tests with the fresh specimen and solution or with the specimen and solution used for three weeks immersion tests.

The inhibitor with high molecular weight (2,4,6-Tris(11-carboxyundecylamino)1,3,5-triazine, molecular weight: 713) had limited effect in a short time of corrosion reaction, resulting in relatively high corrosion rate of about 10⁻⁵ A/cm² from the potentiodynamic polarization behavior. However, the corrosion rate was reduced drastically down to about 10⁻⁹ A/cm² after 3 weeks due to the growth of thick surface oxide.

In contrast, the corrosion rate of Cu in the solution containing 1-aminomethyl(N′,N′-di(2-hydroxyethyl)tolutriazole (molecular weight: 250) or 1-aminomethyl(N′,N′-di(2-hydroxyethyl)benzotriazole (molecular weight: 236) was practically not changed or slightly increased after 3 weeks of immersion, even though the corrosion potential of every solution was ennobled by about 0.1~0.25 V.

The addition of 0.21 wt.% of inhibitors with low molecular weight into the solution with 0.88 wt.% 2,4,6-tris(11-carboxyundecylamino)1,3,5-triazine caused electrochemical behavior dominated by the inhibitor of low molecular weight, possibly due to the diffusion rate of inhibitors in antifreeze solutions with high viscosity.

The solution with 2,4,6-tris(5-carboxypentylamino)1,3,5-triazine, which has a molecular weight of 440, showed a slightly lower rate of Cu corrosion than the solutions with 1-aminomethyl(N′,N′-di(2-hydroxyethyl)tolutriazole or 1-aminomethyl(N′,N′-di(2-hydroxyethyl)benzotriazole. The difference between the corrosion rate in the solution with 2,4,6-tris(5-carboxypentylamino)1,3,5-triazine and that with a triazole-type inhibitor was more noticeable after 3 weeks of immersion.

An increase in the concentration of 1-aminomethyl(N′,N′-di(2-hydroxyethyl)benzotriazole from 0.30 wt.% to 0.59 wt.% in the antifreeze solution retarded dissolution of Cu and solution pH change, but did not result in a marked reduction in the corrosion rate.