Effect of synthetic biomass ash on high temperature corrosion behavior of super austenitic stainless steel 904L

The super austenitic stainless steel 904 L was procured from M/s Bharat Aerospace metals, Mumbai, in the form of hot rolled rods of 16 mm diameter. It was solutionized at 1100 °C for 1 h and quenched in water. The chemical composition of the 904 L steel is presented in table 1. The 16 mm diameter rod was machined to 14.5 mm diameter and disc shaped samples of 14.5 mm diameter and 4 mm thickness were prepared for corrosion study. The samples were mechanically polished with emery paper up to #600 grit size and subsequently cleaned with acetone.

2.1.1. Salt spray coating

2.1.2. Hot corrosion test

Before characterization, the samples were cleaned with hot distilled water to remove any unreacted salts and dried in air, subsequently the samples were cleaned ultrasonically in acetone for 10 min to remove any non-adherent layer. The surface morphology and compositional analysis of the corroded samples was carried out by scanning electron microscope (SEM-Carl Zeiss–Sigma) and EDS (Oxford instruments-INCAx). The phases of the corrosion products were analyzed by XRD (Rigaku MiniFlex-600) with Cu-Kα (λ = 0.15406 nm) scanning at the rate of 10°/min. The cross sections of the corroded samples were analyzed by Electron Probe Micro-Analyzer (EPMA) (CAMECA SXFive) to evaluate the distribution of elements from surface to interior of the samples.

Figure 1 shows photographs of the samples hot corroded by the SM1 and SM2 salt mixtures. The spallation of the corroded/oxidized layer can be seen clearly at the higher temperatures. The corroded layer of the sample by SM1 was found intact following exposure at 500, 550 and 600 °C whereas there was spallation from the exposure at 650 °C. On the other-hand, spallation of the corroded surface of the sample by SM2 was found to start at 550 °C and progressively increase up to 650 °C. It was observed that there was only partial melting of SM1 from the exposure of 100 h at 500 °C whereas there was uniform melting of the SM2. Tiny spots of creamy color may be observed all over the surface of the samples exposed at 550 °C–600 °C for 100 h. On the other hand, there are pale yellow spots on the surface of the SM2 coated sample and spallation of the scale from the flat surface as well as exfoliation along the periphery. However, there was no spallation due to SM1; a few spots of salt remnants were observed after 100 h of exposure. The yellow spots may be due to salt reactions with surface oxides and formation of potassium/sodium chromates.

Zoom In Zoom Out Reset image size

Figures 2(a) and (b) show plots of weight gain per unit area (mg cm⁻²) versus time interval of 15 cycles for the samples, coated with SM1 and SM2 respectively, exposed in air atmosphere at 500, 550, 600 and 650 °C for 100 h. It is evident that there is much higher weight gain in the samples coated with SM2 than in those coated with SM1. Thus, it is obvious that there was more rapid formation of oxides and other corrosion products from SM2. The inserts in the plots show that weight gain during the initial period of 5 h was much less than that beyond 5 h of exposure, in the samples coated with of both SM1 as well as SM2. However, the weight gain even during the initial stage of 5 h is relatively higher from SM2 than from SM1. Figures 3(a)–(b) show the variation of square of weight gain per unit area with duration of exposure up to 100 h. There is rapid increase in (ΔW)² during the exposure of 5–25 h at 600 °C and significantly more at 650 °C in the sample coated with SM2, thereafter the increase is much slower during 25–100 h of exposure, and a dual slope behavior is exhibited (figure 3(b)).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

On the other hand there is relatively much less increase in (ΔW)² from the exposure at 500 and 550 °C and the increase is linear with a single slope during the exposure from 5–100 h. The variation of (ΔW)² with the duration of exposure in the sample coated with SM1 (figure 3(a)) is similar to that coated with SM2 (figure 3(b)), however, the increase from SM1 is much less than that from SM2, as the scale of (ΔW)² in figure 3(b) is much higher than that in figure 3(a). The corrosion kinetics may be seen to follow parabolic rate law ΔW² = k_pt + c, where ΔW is weight gain per unit area, t is the exposure time, kp is rate constant and c is a constant. The calculated values of kp are presented in table 3 which are much higher for the samples coated with SM2 than those coated with SM1, above 500 °C. It is also evident that kp values at 600 and 650 °C are much higher corresponding to 5–25 h of exposure than those for 25–100 h of exposure.

XRD patterns of the samples with SM1 and SM2 coatings exposed in air from 500 °C–650 °C are shown in figures 4(a) and (b) respectively. Details of the analyzed peaks are presented in table 4.

Zoom In Zoom Out Reset image size

Spinels of Cr–Ni–Fe and Cr–Mn–Cu oxides are observed along with the corrosives FeS, FeCl₃ and CrS in the samples coated with SM1 and K₂Cr₂O₇ and K₂FeO₄ in those coated with SM2. At 550 °C similar features are observed with formation of spinels of transition elements oxides. Corrosion products namely KCrO₂ and NaCrO₂ are evolved with SM1 salt mixture whereas NaFeO₂ and Na₂Cr₂O₇ are formed with SM2. At 600 °C corrosion from the salt mixture resulted in formation of a new phase Cr₂MnO₄ in both the samples coated with SM1 as well as SM2. Also, there is extensive formation of Na–K–Fe–Cr–O based corrosion products along with Fe–Cr–S–Cl system. At 650 °C, there was formation of K₂CrO₄ and Na₂CrO₄, new phases of corrosion products. Table 5 shows the various corrosion and oxidation reactions at different temperatures of exposure.

The relative inhomogenieties of the surface morphology and development of scales on the samples in respect of the Cl:S ratio in the two salt mixtures were analyzed by SEM-EDS.

The micrographs in figures 5(a), 6(a), 7(a) and (8)(a)–(b) show top surfaces of the samples corroded by SM1 and those in figures 5(b), 6(b), 7(b) and 9(a)–(b) show surface morphology of the samples corroded by SM2, from 100 h of exposures at 500–650 °C. At all the temperatures of the exposure, coating of the both SM1 and SM2 resulted in characteristic morphology of metals oxides and corrosion products over the surface.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Figures 5(a) and (b) show the samples corroded with SM1 and SM2 respectively, from the exposure at 500 °C for 100 h, exhibiting less oxidation/corrosion. The scale formed from SM1 was intact and there was no spallation (figure 5(a)) whereas the coating of SM2 had cracked at some places and also spalled (figure 5(b)).

In figure 6(a), the surface of the sample corroded by SM1 at 550 °C shows small granules of oxide particles, uniformly grown all over the surface. The morphology of the product formed with SM1 at 550 °C shows self-oriented flowers and features like cactus needle, grown in between the oxide granules whereas the sample corroded by SM2 shows oxide morphology of globular/nodular shape (figure 6(b)). This morphology is like a cauliflower and there are globular domes. This type of morphology can be understood from localized blistering of the scale by volatile products and pushing of scale outward [17].

The morphology of the surface resulting from corrosion at 600 °C due to SM1 and SM2 salt mixtures may be seen in figures 7(a) and (b) respectively. It is clear that from SM1 there was formation of globular nodules and also spallation of the scale. The magnified image shows that lamellar structure, with higher content of Fe–Ni–Cr spinel, as revealed by EDS analysis. As shown in figure 7(b), there was spallation of the sample coated with SM2 and exposed at 600 °C. In this condition, from 550 to 600 °C, oxides are almost similar but differ in morphology. Oxide nodules may be seen to be irregular in shape (figure 7(b)).

Figure 8(a) shows the morphology of the layer lying beneath the spalled layer of the sample coated with SM1 and exposed at 650 °C. Oxides granules are uniformly distributed all over the surface beneath the spalled layer. Figure 8(b) shows morphology of the scale which did not spall. The unspalled layer shows nodular/irregular shape of oxide nodules of Cr–Ni spinels.

Figure 9(a) shows formation of dense filament like features of Ni–Cr–Fe spinel in the region without spallation. Further, at 650 °C there is formation of two types of regions: one with filament like features and the other without such features. Figure 9(b) shows the region where spalled layer is not present and the region beneath the spalled layer, with cuboidal shaped Fe-Ni spinels.

Cross sections of the corroded samples were analyzed by SEM-EDS and EPMA.

Figures 10(a) and (b) show SEM-micrographs of cross sections of the samples corroded by SM1 exposed at 500 and 550 °C whereas figures 10(c) and (d) show the features corroded by SM1 at 600 and 650 °C respectively. The intact layer of the oxide in figure 10(a) reveals negligible corrosion and absence of internal corrosion. From the EDS analysis it was found that there was no depletion of elements with increment in temperature from 550 to 600 °C, however, there was internal attack as well as depletion and cracking along and also beneath the scale. This internal corrosion appears to be higher at 600 °C than that at 550 °C where the scale interface is intact to the base material. From the EDS analysis over the cross section in figures 10(c) and (d), it was observed that there was higher depletion of Fe–Cr.

Zoom In Zoom Out Reset image size

Figures 11(a) and (b) show SEM-micrographs of cross sections of the samples corroded at 500 and 550 °C by SM2. It may be seen that there was greater attack from the SM2 as compared with that from SM1. SEM micrographs of the samples corroded at 600 and 650 °C are shown in figures 11(c) and (d) respectively, where it may be observed that severe corrosion had occurred due to SM2 and there was porous region beneath the scale interface. The corrodants locally attacked the base metal through the porous scale. This region may be presented as a layer of porous γ-Fe₂O₃/Fe₃O₄, and the internal part of the depleted zone is in the form of chromium oxide with the other constituents and showed high amount of chromium by EDS analysis. The depth of the corroded layer was ≈16–18 μm in the sample coated with SM1 and ≈31–33 μm with SM2 (table 7).

Zoom In Zoom Out Reset image size

Figures 12(a)–(b) and 13(a)–(b) show EPMA analysis of cross sections of the samples corroded with SM1 and SM2 at 650 °C for 100 h. The severity of corrosion from SM2 with 60:40 (Cl:S) ratio is quite evident and there is also internal attack. A three layer scale, with a thick depleted region can be seen between the internal scale at the substrate and the outer scale on the surface of the sample with SM2 coat as compared with the much thinner scale resulting from SM1. Moreover, after 100 h of exposure, it can be seen from figure 12(a) that the layer formed from SM1 constituted of dual phase layer. Figure 12(b) shows distribution maps of different elements (Cr, Ni, Fe, Mn, Mo and O). In case of the sample with SM2, multiphase layer can be seen in figure 13(a)–(b). From the color contrast in the EPMA maps in figure 13(b), the distribution of respective elements in the scale and the substrate may clearly be seen. The relative width of the scale and the depletion zone in the samples with SM1 and SM2 can be seen in the figures 14(a) and (b) respectively along with concentration of the elements in these zones. According to line scan analysis, in figure 14(a) the scale consists of iron and oxygen and the depletion zone constitutes of Cr–Fe–O as well as some amount of Ni close to the base metal. On the other hand figure 14(b) shows that the scale resulting from SM2 is much wider and consists of Fe, Cr and O and the large depletion zone shows presence of Cr, Ni, Fe and O.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

It may be seen from the figure 2 that there was little weight gain from the exposure at 500 °C due to non fusion of the coated salt mixtures, in agreement with the earlier observation made by Daniel et al [2]. At 550 °C also there is weight change, more or less of similar trend of linear increase, however, it was relatively more in the sample coated with SM2 (figure 2(b). At 600 and 650 °C there was rapid increase in the weight gain from 5–25 h, beyond which the gained weight increased very slowly whereas it increased progressively at 500 and 550 °C. Since there was no melting of the salt mixtures at 500 °C, the XRD phase analysis showed formation of in situ oxides and in parallel the ex situ diffusion [18] of salt showed less severe corrosive nature of oxides. The much lower value of kp corresponding to exposure of the samples coated with SM1 and SM2, from 25–100 h may be attributed to decrease in the amount of corrodants due to their high reactivity/instability (table 8) and formation of protective layer over the spalled surface region. On the otherhand at lower temperature of 500 °C–550 °C the reactivity of corrodants is much less and their stability is higher, hence a single slope behavior is exhibited (figure 3). The much higher level of the (ΔW)² versus time plot in figure 3(b) at 650 °C as compared to that at 600 °C may be understood from the difference in the hot corrosion rate which nearly 15 times higher at 650 °C than that at 600 °C during 5–25 h of exposure (table 3).

The reactions occurring under this condition are shown in table 4. During the initial 5 h of exposure there is much less increase in the weight as seen from the inserts, beyond 5 h only there is rapid increase in the weight gain up to 25 h of exposure, and parabolic behavior is exhibited (figures 3(a)–(b)).

It is well known that alkali chlorides accelerate the process of corrosion [19], while the alkali sulfates lower the effects of alkali chlorides in biomass combustion [20]. The cross section of the sample coated with SM1 and exposed at 500 °C did not show any porous layer because of insufficient melting of the salts.

The scale formed was intact and there was no spallation. On exposure at 550 °C the non-protective feature of the oxide scale may be observed from its porous nature [8]. The difference in surface of the samples corroded with salt mixtures at 550 °C from that at 500 °C is due to more penetration of salt mixture in to base material at 550 °C. From the weight change at 600 °C it is seen that weight gain was due to formation of chromia layer as per the reactions between (Cr, Fe)₂O₃ matrix and the molten salts (table 5). A new specie formed at 600 °C with SM1 was Cr₂MnO₄, in agreement with the minimum amount of Mn depleting with chromia layer [21].

In particular at 600 and 650 °C (figures 2 and 3), this behavior reflects rapid growth of oxides induced by chlorination and subsequent rapid oxidation by the process of the so called active oxidation. Presence of FeCr₂O₄, Cr₂O₃ on the exposed surface indicates rapid formation of oxides which is in line with the previous observation made on Sanicro-28 [22]. The cross section of the sample following 100 h of exposure at 600 °C shows that the formed scale tended to separate out from the base metal. This kind of the scale layer was also observed at 650 °C. The cause of separation of the scale from the base material is the mechanical stresses induced during cyclic corrosion. At 650 °C there was faster rate of corrosion with higher weight gain resulting from the formation of porous and non-protective layer. From phase analysis it is confirmed that this nature is due to the corrosives such as K₂Cr₂O₇ and Na₂Cr₂O₇. The above two eutectic mixtures enhanced the rate of corrosion very rapidly.

SEM-EDS analysis of cross section of the sample (figure 11) shows that at 650 °C a significant amount of oxides and corrosion products got embedded beneath the scale. These oxides were loosely bonded and spalled. The growth of the other oxides can be seen clearly from the surface morphology (figures 8 and 9).

The surface morphology of the samples corroded with salt SM2 in the temperature range 500 °C–650 °C, clearly reveals variation in the size of oxides. There is far less spallation from exposure at 500 °C–550 °C whereas it is much higher at 600 °C and highest at 650 °C in agreement with the observation of Grabke [23] where high alloy steel was found spalled as well as there was porous scale of Fe₂O₃ at the oxide interface due to high chloride content and escape of high volatile chlorides. Due to high chlorine content in the salt mixture and penetration of Cl^- in to the scale resulted in formation of NiCl₂, FeCl₂, FeCl₃ and CrCl₃. The inward diffusion of Cl^- ions and outward diffusion of alloying elements resulted in formation of a thick depleted zone at the interface between the substrate and the scale. These phases in this zone were characterized by the XRD and SEM-EDS analysis and found to be due to chlorides. These chlorides continuously evaporated and iron chloride got oxidized to Fe₃O₄ in air on the top at outer surface of the scale. Thus, it can be called active oxidation of Fe₂O₃ in the presence of chlorine which acts as catalyst and accelerates the process of corrosion at 600 °C–650 °C forming cracks/splat boundaries, grain boundary grooving or fissuring. The outward diffusion of FeCl₂(g) through Fe₂O₃(s) gave rise to an inverse proportion to thickness of the oxide. Effect of SO₂ in the atmosphere on the active oxidation beneath the deposits was in the form of CrS. A modest increase of oxidation rate was observed with high Cl-content rather than S-content. From the literature also, it is clear that sulfur minimizes the effect of chlorine attack from 500 °C–650 °C up to 100 h in air atmosphere [24, 25].

The results obtained by XRD/SEM-EDS and EPMA, clearly indicate that at 500 and 550 °C oxide scale remained almost similar in the two regions and spinel formed. There was formation of spinel which comprises mainly of the M₂O₃ and M₃O₄ oxides. Where, M is the transition element such as Fe, Cr, Ni, Mn, Mo and Cu. From figures 5 and 6(a)–(b) the amount of corrosives evaluated from EDS analysis confirmed the presence of the phases such as FeCl₃, FeS and CrS. Moreover, the potassium salts oxides are K₂Cr₂O₇, K₂FeO₄ and KCrO₂ and the sodium based oxides NaCrO₂, Na₂Cr₂O₇ and NaFeO₂ were found from the EDS analysis. With increase in temperature of the sample coated with SM1 to 600 °C and 650 °C, there was no change in potassium products while in the sample with SM2 coating K₂Cr₂O₇ got converted to K₂CrO₄ and Na₂Cr₂O₇ to Na₂CrO₄ respectively [26–28].

EPMA analysis of the cross section and the line scans of the samples exposed at 650 °C show that there was a significant outward diffusion of Fe and Ni from the base metal to the scale. A thick crust of oxides rich in Fe and Ni can be seen in figure 13. It is also important to note that at lower temperature of 500 °C–550 °C the alloy surface showed pits and uneven interface, whereas at the higher temperature of 600 °C–650 °C the alloy surface of the sample was uniformly eaten away by the corrodants and a layered structure formed comprising of several laminar cracks.

4.3.1. Significance of alkali metal, chlorides and sulphates

4.3.2. Correlation of EPMA analysis with free energy data