Improved corrosion resistance of SS316L in molten nitrates containing chloride ions: Pre-oxidation treatment

Introduction

Solar energy, as an entirely clean and renewable energy source, possesses numerous advantages, such as its inexhaustibility, wide distribution and high safety. These characteristics render it a crucial tool for achieving sustainable development and addressing global energy and environmental challenges. Concentrated solar power (CSP) systems exhibit outstanding energy storage and load-following capabilities, making them suitable for integration with wind and photovoltaic power generation to enhance the stability of electric load operation. ¹

Molten salt energy storage is a fundamental technology in CSP, and the currently prevalent storage medium is solar salt (NaNO₃: KNO₃ = 60: 40 (wt-%)), which exhibits high heat capacity, high thermal conductivity and low viscosity. However, molten solar salt exhibits low energy density, high melting point and low working temperature, which significantly constrain the power generation efficiency and cost of CSP.^2,3 Furthermore, the performance of solar salt is susceptible to various impurities, imposing specific purity requirements on nitrate that also contribute to increased costs.^4,5 Researches have demonstrated that the addition of sodium nitrite to create Hitec salt (NaNO₃: KNO₃: NaNO₂ = 7: 53: 40 (wt-%)) can effectively reduce the melting point of the salt to 142 °C, albeit at the expense of diminished thermal stability, with a decomposition temperature of 500 °C. ⁶ The addition of carbonates is also considered a method to enhance the performance of molten salt. Wu et al. ⁷ proposed LiNaK carbonate (LiCO₃: NaCO₃: KNO₃ = 32.1: 33.4: 34.5 (wt-%)), which increased the thermal stability temperature of the molten salt to 800 °C, with a melting point of 400 °C. By incorporating 29 mol% Cl⁻ into the solar salt (NaNO₃: KNO₃: NaCl: KCl = 60: 40: 13.18: 16.82 (wt-%)), the thermal stability temperature of the molten salt could be elevated to 603 °C, resulting in a slight reduction in the melting point and an increase in energy density to 3.4 J/(gres, ⁸ which could satisfy the requirement of 2.25 J/(gch could satisfy the requirement of 2.25elting point and an incr. ⁹

However, the introduction of Cl⁻ enhances the corrosiveness of the molten salt towards iron-based alloys. In pure solar salt, iron-based alloys typically undergo the following reactions (M represents Fe, Cr, Ni etc.):

Li et al.^10,11 have indicated that the introduction of Cl⁻ as impurities into solar salt would exacerbate the corrosion of iron-based austenitic stainless steels, such as SS316L, through three aspects ¹² : (1) induce active oxidation; (2) exacerbate the corrosion of molten solar salt and (3) facilitate the decomposition of oxides. For iron-based stainless steel, the presence of 1% NaCl in solar salt (M = Fe, Cr) leads to the following reactions:

Chloride undergoes a reaction with oxygen, leading to the generation of additional O²⁻ ions and promoting active oxidation. Simultaneously, it forms chlorides with the Fe and Cr elements in the alloy, which volatilise at elevated temperatures and escape outward, exacerbating the adhesion of the corrosion layer and rendering it more susceptible to detachment. Subsequently, the escaped gaseous chlorides give rise to relatively loose oxides on the outer surface of the corrosion layer due to higher O²⁻ ion concentration. The aforementioned reactions accelerate the corrosion rate of iron-based stainless steel in solar salt containing Cl⁻, leading to continuous delamination of the corrosion layer and a reduction in its operational lifespan. Moreover, particulates of exfoliated corrosion layers may circulate within the heat-carrying fluid in the conduits, potentially leading to pipe blockages and causing structural damage to the pipe alloy, thereby hastening the stress-induced cracking of the corrosion layer. ¹³ After being immersed in pure solar salt for 500 h, the weight loss of SS316L was only 0.03 mg/cm². However, when 1 wt-% NaCl was added, the mass loss was significantly increased to 1.8 mg/cm². ¹⁰ According to the research of Javier et al., for low-carbon steel A516 Gr70, after being immersed in molten salt containing 1.2 wt-% NaCl for 1581 h, the mass loss was 11.4 mg/cm²; while under 3 wt% NaCl conditions, the mass loss was further increased to 38.5 mg/cm². ¹⁴ Thus, it can be seen that as the concentration of chloride ions increases, the corrosiveness of molten salt to iron-based alloys are significantly enhanced. For the case of prolonged immersion, the influence of chloride ion concentration in solar salt on alloy corrosion decreases, which is believed to be related to the consumption of Cl⁻.

In order to enhance the performance of structural materials in chloride-containing molten nitrate salts, it is common to consider reducing the Cl⁻ content in the molten salt. However, this does not align with the research objective of this experiment. Additionally, the addition of corrosion inhibitors or the use of inert atmospheres (such as N₂) can improve the performance of the molten salt to a certain extent. By optimising the alloy composition and selecting nickel-based alloys with better corrosion resistance (such as Inconel 600, Inconel 625, Inconel 800H and Haynes 230), the corrosion resistance of the material can be significantly enhanced. Huang and co-authors ¹⁵ compared the corrosion behaviours of nickel-based alloy GH3539 and SS347 in an air environment at 600 °Cs of nickel-based alloy GH3539 and SS347 in ₃: KNO₃: NaCl: KCl = 60: 40: 13.18: 16.82 (wt-%). After being immersed in the molten salt for 200 h, the results showed that the mass loss of SS347 was ∼16.22 mg/cm², while the nickel-based alloy GH3539 only showed a slight increase in mass, ∼0.11 mg/cm². Thus, it can be seen that the nickel-based alloy has excellent corrosion resistance in this environment and can effectively provide protection, but it also comes with relatively high costs. Additionally, applying coatings is a method that can effectively reduce costs while improving the corrosion resistance of the material. However, for complex-shaped workpieces, there are still certain challenges in preparing uniform and reliable coatings. Pre-oxidation treatment may be a simple, effective method that can enhance the corrosion resistance of iron-based stainless steel. This method has the advantages of being easy to operate, having high economic efficiency, and being able to uniformly form an oxide film on the surface of the workpiece. Pre-oxidation treatment not only enhances the material's high-temperature oxidation resistance, but also effectively reduces the corrosion caused by molten salt by isolating the direct contact between the matrix and the molten salt environment. Especially during the initial exposure stage, the stable oxide layer formed by pre-oxidation can significantly reduce the reaction rate between the alloy and the molten salt, thereby exerting a good protective effect. Liu et al. ¹⁶ carried out pre-oxidation treatment on T91 steel, forming a double-layer pre-oxidation film on the surface with an outer layer rich in Fe and an inner layer rich in Cr, significantly enhancing the corrosion resistance of the alloy in the liquid lead-bismuth eutectic (LBE). Additionally, Taehoon et al. ¹⁷ pre-oxidated the iron-based high-aluminium alloy ACES, effectively reducing the corrosion degree of the alloy in the molten NaCl–KCl environment.

Therefore, the effect of the pre-oxidation process on the corrosion behaviour of SS316L in molten solar salt, adding chloride salts, is investigated in this article. The compositions of the oxide layer on SS316L following air atmosphere annealing at various temperatures and the corrosion resistance of SS316L before and after pre-oxidation in a molten salt mixture of NaNO₃–KNO₃–NaCl–KCl are discussed. Furthermore, the corrosion mechanism of iron-based austenitic stainless steel SS316L in NaNO₃–KNO₃–NaCl–KCl is thoroughly explored.

Experimental procedures

Materials and reagents

Analytical-grade NaCl (99.9%), KCl (99.9%), NaNO₃ (99.0%) and KNO₃ (99.0%) salts provided by Shanghai Titans Scientific Co., Ltd were selected. The experimental salt was dried at 200 °C for at least 24 h, and then dried again at 200 °C for another 24 h after being weighed with a specific proportion of NaNO₃: KNO₃: NaCl: KCl = 60: 40: 13.18: 16.82 (wt-%) and mixed. SS316L was selected as the experimental alloy for this study, and its composition is detailed in Table 1. The specimens were laser-cut into pieces measuring 15 mm × 10 mm × 2 mm, with a 2 mm diameter hole drilled in each sample for suspension. The samples were sequentially polished with 600-mesh and 1000-mesh silicon carbide sandpaper, followed by ultrasonic cleaning in deionised water and ethanol to eliminate surface impurities. Subsequently, the dried sample was weighed using an analytical balance with a precision of 0.1 mg.

OPEN IN VIEWER

Table 1.

Chemical composition of SS316L (wt-%).

	C	Ni	Mo	Si	Mn	Fe	Cr	S	P
SS316L	0.014	12.58	2.28	0.60	0.80	66.57	17.14	0.0073	0.013

Results

Pre-oxidation treatment

After oxidation at 750 °C, 800 °C and 850 °C for 5 h, the XRD patterns of SS316L are depicted in Figure 1. At three temperatures, Fe₃O₄ and Cr₂O₃ are the primary oxidation products of SS316L. Furthermore, the oxide peak area increases with temperature, indicating progressive oxidation. At 850 °Cre the primary oxidation₂O₄ as an additional phase is observed. Upon exposure to 750 °Cas an additional phase is observed. Upon exposure to 750, the oxide peak area increases wixide layer is formed at both 800 °C and 850 °Ca

OPEN IN VIEWER

Figure 1.

The X-ray diffractometer (XRD) patterns of SS316L after oxidation (a) at 750 °C, (b) 800 °C and (c) 850 °C for 5 h, respectively.

The surface morphologies of SS316L after oxidation for 5 h at 800 °C and 850 °C are depicted in Figure 2. It can be seen that SS316L exhibits a smooth appearance after oxidation at 800 °C for 5 h, with a uniform oxide layer covering the surface and devoid of any discernible scratches. When the oxidation temperature is raised to 850 °C, with a uniform oxide layer covering the surface and devoid of any discernible scratches. When the oxidation temperature is raised to 8an be attributed to the grain coarsening caused by high-temperature annealing, and the scratches that occurred during polishing are not completely covered. Figure 3 shows the cross-sectional morphology of SS316L stainless steel after oxidation for 5 h at 750 ^oC, 800 ^oC and 850 ^oC, respectively. After oxidation at 800 ^oC, a uniform and dense oxide layer is formed on the surface of SS316L stainless steel, mainly composed of chromium oxide (Cr₂O₃) and iron(III) oxide (Fe₃O₄), presenting continuous and dense microstructural characteristics. According to the data in Figure 3, the thickness of the oxide layer formed after oxidation at these three temperatures is 0.34, 2.70 and 4.07 μa, respectively. Therefore, considering the uniformity and density of the oxide layer, the pre-oxidation process is finally selected to be carried out for 5 h at 800 ^oC, and this condition is used for the preparation of pre-oxidation samples in the subsequent corrosion tests.

OPEN IN VIEWER

Figure 2.

Surface morphologies of SS316L after oxidation (a, b) at 800 °C and (c, d) 850 °C for 5 h.

OPEN IN VIEWER

Figure 3.

Cross-sectional morphologies of SS316L after oxidation at 800 °C for 5 h.

Corrosion kinetics

The corrosion kinetic curves of untreated and pre-oxidised SS316L in molten NaNO₃–KNO₃–NaCl–KCl at 600 °C under an air atmosphere for 200 h are depicted in Figure 4. Untreated SS316L exhibits continuous weight loss during 200 h, with a mass loss of ∼16.63 mg/cm². This result is much greater than the mass loss of 316L alloy after being immersed in 60 wt-% NaNO₃ + 40 wt-% KNO₃ molten salt and 60 wt-% NaNO₃ + 40 wt-% KNO₃ + 1 wt-% NaCl molten salt for 500 h, which were ∼0.30 and 1.80 mg/cm1, respectively. Even compared with the mass loss of 316L alloy after being immersed in 60 wt-% NaNO₃ + 40 wt-% KNO₃ + 1 wt-% NaCl molten salt for 3000 h (about 5.00 mg/cmut, ¹² this result is still significantly higher. It can be seen that the addition of Cl⁻ significantly aggravates the corrosion of 316L alloy in molten solar salt, and this effect becomes more obvious with the increase of Cl⁻ concentration. This may be attributed to the presence of Cl⁻ ions in the molten salt, which can expedite the oxidation of alloy elements and compromise the adhesion between the corroded surface layer and the substrate, ultimately resulting in exfoliation of the corroded layer in scale form. After pre-oxidation treatment, SS316L exhibits a marginal decrease in mass during the initial 20 h, followed by a gradual increase over time. After 200 h corrosion, the final mass loss of the pre-oxidised SS316L is only about −0.72 mg/cm². This result is significantly lower than the mass loss of 16.63 mg/cms result is significantly lower than  h. Suggesting the potential formation of substances within the molten salt environment. The modest magnitude of mass alteration indicates a certain level of stability for the pre-oxidised SS316L in molten NaNO₃–KNO₃–NaCl–KCl.

OPEN IN VIEWER

Figure 4.

Corrosion kinetic curves of untreated and pre-oxidised SS316L in molten NaNO₃–KNO₃–NaCl–KCl at 600 °C for 200 h.

Corrosion products analysis

The XRD patterns of untreated and pre-oxidised SS316L in molten NaNO₃–KNO₃–NaCl–KCl at 600 °C under an air atmosphere for 200 h are shown in Figure 5. The corrosion products of untreated and pre-oxidised SS316L are similar, primarily comprising iron oxides such as Fe₂O₃, Fe₃O₄, spinel FeCr₂O₄ and NiFe₂O₄. Nevertheless, the intensity of the oxide product peaks in the corroded samples under pre-oxidation treatment is notably elevated, suggesting enhanced crystallinity and increased content of the products. This phenomenon may be related to the formation of a relatively compact oxide layer on the substrate surface.

OPEN IN VIEWER

Figure 5.

The X-ray diffractometer (XRD) patterns of untreated and pre-oxidised SS316L after corrosion in molten NaNO₃–KNO₃–NaCl–KCl at 600 °C for 200 h.

The cross-sectional morphologies and EDS analysis results of untreated and pre-oxidised SS316L after being corroded for 200 h at 600 °C in molten NaNO₃–KNO₃–NaCl–KCl are presented in Figures 6 and 7. The XRD analysis (Figure 5) and EDS results (Figures 6 and 7(a) and (c)) reveal that the corrosion layer of untreated SS316L primarily comprises FeCr₂O₄ and Fe₂O₃ particles on the outer surface, with a contiguous corrosion layer beneath, which is stratified into two distinct layers. The upper layer predominantly consists of Fe₂O₃, while the lower layer is principally composed of FeCr₂O₄ and NiFe₂O₄ with a spinel structure. Localised corrosion occurs at the interface between the substrate and the corrosion layer, which can be ascribed to the penetration of Cl⁻. After pre-oxidation treatment, a continuous and uniform three-layered corrosion product that exhibits excellent adhesion to the alloy is formed on the SS316L, with an approximate thickness of 14.22 μm. Based on the XRD analysis (Figure 5) and EDS results (Figures 6(b1) to (b3) and 7(c1) to (c5)), it is evident that the outer corrosion layer predominantly comprises Fe₂O₃, while the intermediate corrosion layer is primarily composed of Fe₃O₄. Furthermore, the inner corrosion layer exhibits a similar constitution to the untreated SS316L, mainly consisting of FeCr₂O₄ and NiFe₂O₄ with a spinel structure. Two distinct chromium-depleted corrosion layers on the outer surface of the pre-oxidised SS316L may be ascribed to the weight loss phenomenon of the pre-oxidised sample during the initial 20 h (Figure 4). Furthermore, a chromium depletion layer is indeed formed on the sample surfaces after 20 h of molten salt corrosion following pre-oxidation treatment (Figure 7(b1) to (b5)). This also indicates that the weight loss phenomenon observed during the initial 20-hour stage (Figure 4) is related to chromium dissolution.

OPEN IN VIEWER

Figure 6.

The cross-sectional morphologies and energy dispersive X-ray spectroscopy (EDS) analysis of untreated (a to c) and pre-oxidised (d to f) SS316L after corroded for 200 h at 600 °C in molten NaNO₃–KNO₃–NaCl–KCl.

OPEN IN VIEWER

Figure 7.

Cross-sectional morphologies and the corresponding mapping scanning EDS results of SS316L stainless steel after corrosion in molten NaCl–KCl–NaNO₃–KNO₃ salt at 600 °C: untreated 316L for 200 h (a1 to a5), pre-oxidation 316L for 20 h (b1 to b5), pre-oxidation 316L for 200 h (c1 to c5).

Discussion

In molten NaNO₃–KNO₃–NaCl–KCl, untreated SS316L demonstrates continuous and significant mass loss and the corrosion kinetics curve follows a parabolic trend. The presence of Cl⁻ negatively impacts the adhesion of the corrosion layer, resulting in facile delamination. Partially detached NiFe₂O₄ is observed at the interface between FeCr₂O₄, NiFe₂O₄, and the outer layer of Fe₂O₃. The corrosion mechanism is illustrated in Figure 8. At the initial stage of corrosion, due to the active oxidation of Cl⁻ (reaction 4), the partial pressure of O²⁻ in the molten salt is lower than that in pure solar salt (approximately 4–6). This lower oxygen partial pressure is sufficient to cause the rapid oxidation of Fe and Cr elements on the alloy surface to Fe₂O₃ and Cr₂O₃. As a result of the active oxidation mechanism,^19,20 the presence of Cl⁻ in the molten salt leads to an increased release of O²⁻ and Cl₂ through reaction 4, thereby promoting rapid growth of the oxidation layer. Moreover, as Cr₂O₃ dissolves in the molten salt, there is continuous external diffusion of Cr, resulting in a more porous and loosely structured oxide layer on the alloy surface. Thus, as depicted in Figure 7(b), the outer oxide layer primarily consists of Fe₂O₃. As the reaction progresses, Cl₂ permeates through the porous oxide layer on the surface and reaches the interface between the alloy and corrosion layer, where pCl₂ governs the reaction kinetics. At this interface, Fe and Cr selectively form chlorides. The lower Gibbs free energy of CrCl₂ makes chlorine gas more thermodynamically favourable to react with Cr to form CrCl₂, as opposed to reacting with Fe to form FeCl₂. Due to the relatively high saturated vapour pressure of CrCl₂ and FeCl₂, they volatilise at 600 °C in a molten salt environment and escape through the corrosion layer, leading to a weakening adhesion of the corrosion layer and facilitating its delamination. At this juncture, when the pO₂ is elevated on the exterior of the corrosion layer, O²⁻ undergoes a reaction to generate metal oxides such as Fe₂O₃, Cr₂O₃ and Cl₂ on the outer surface of the corrosion layer, thereby inducing rapid growth of the oxide layer on its outer side.

OPEN IN VIEWER

Figure 8.

The corrosion mechanisms of untreated SS316L (a to c) and pre-oxidised SS316L (d and e) in molten NaNO₃–KNO₃–NaCl–KCl.

In contrast to bare SS316L, a pre-formed continuous and dense oxide layer comprising Fe₃O₄ and Cr₂O₃ on SS316L after heat treatment at 800 °C for 5 h makes the sample exhibit favourable corrosion resistance in molten NaNO₃–KNO₃–NaCl–KCl. In the initial stage, the pre-oxidised samples exhibit a slight mass loss, which can be attributed to the dissolution of chromium and its oxides in the highly alkaline nitrate. The reactions are as follows:

In addition, Bonk et al. detected CrO₄²⁻ in the chlorinated solar salt that had corroded AISI312 alloy, confirming the dissolution of chromium oxides.^21,22 During the corrosion process, as the outer layer of chromium dissolved, iron oxides rapidly formed on the alloy surface, effectively hindering the outward diffusion of chromium and its oxides, thereby suppressing further weight loss. An outer layer with low chromium and high iron content was formed. Leading to the subsequent development of spinel structures comprising FeCr₂O₄ and NiFe₂O₄ at the interface between the alloy and the Fe oxides. ²³ The final corrosion layer primarily consists of an outer layer of Fe₂O₃, a middle layer of Fe₃O₄ and an inner layer of FeCr₂O₄ and NiFe₂O₄.

The pre-formed Fe₃O₄ oxide layer creates a barrier to prevent Cl₂ from entering the corrosion layer and reducing its adhesion, while external diffusion of Fe outside the oxide layer leads to the formation of a new Fe₂O₃ oxide layer. This process effectively inhibits peeling of the corrosion layer. Under these conditions, the addition of Cl⁻ only triggers an active oxidation mechanism, resulting in an increase in O²⁻ content in the molten salt. Meanwhile, at the interface between the alloy and the Fe₃O₄ oxide layer, O²⁻ reacts with Fe, Cr and Ni to generate metal oxides such as FeO, Cr₂O₃ and NiO. Concurrently, while FeCr₂O₄ and NiFe₂O₄ spinel structures are formed through the following reactions ²⁴ :

A new corrosion layer, consisting of FeCr₂O₄ and NiFe₂O₄ spinel, is formed between the alloy and the Fe₃O₄ oxide layer. The ordered distribution of cations in the spinel layer hinders the migration path of iron ions, restricts the movement of iron vacancies, and thereby reduces the total number of vacancies. At the same time, the spinel structure can decrease the number of oxygen ion vacancies in the oxide layer, inhibiting the formation and expansion of cavities, and thereby reducing the absorption of oxygen and effectively lowering the intensity of the oxidation reaction. This process diminishes the impact of active oxidation.^25,26 This further improves the corrosion resistance of SS316L in the molten salt mixture of NaNO₃–KNO₃–NaCl–KCl. Ultimately, a three-layer corrosion layer structure is developed on the surface of SS316L, comprising an inner layer of FeCr₂O₄ and NiFe₂O₄ spinel, a middle layer of Fe₃O₄ oxide and an outer layer of Fe₂O₃ oxide.

Conclusions

This study investigates the corrosion resistance of SS316L before and after pre-oxidation in NaNO₃–KNO₃–NaCl–KCl molten salt using static corrosion methods. The research findings indicate that the optimal oxidation condition for SS316L is 800 °Cl molt h, characterised by the formation of a dense oxide layer composed of Fe₃O₄ and Cr₂O₃. The chloride ions in molten nitrates significantly increase in the corrosion rate of SS316L, mainly by promoting oxidation and diminishing the adhesion of the oxide layer. Pre-oxidation treatment significantly reduced the corrosion rate of SS316L in the molten salt and no large-scale peeling of oxide layers occurred during direct corrosion. A continuous stable three-layer structure is formed on the substrate surface, where the inner layer is a FeCr₂O₄ and NiFe₂O₄ spinel, the middle layer is a Fe₃O₄ oxide layer and the outer layer is a Fe₂O₃ oxide layer. Reactive pre-oxidation treatment enhances the corrosion resistance of SS316L in molten salts by facilitating the formation of a compact chromium-iron oxide layer. This article provides a feasible new idea for the corrosion prevention design of the next-generation CSP equipment. However, the enhancement effect of pre-oxidation treatment on the corrosion resistance of 316L stainless steel in a dynamic molten salt environment still requires further study. In addition, combining the pre-oxidation technology with methods such as adding protective coatings and introducing inert gases is expected to further reduce the corrosion rate of iron-based stainless steels in high-chlorine molten salts.