Effect of Additional Sulfide and Thiosulfate on Corrosion of Q 235 Carbon Steel in Alkaline Solutions

This paper investigated the effect of additional sulfide and thiosulfate on Q235 carbon steel corrosion in alkaline solutions. Weight loss method, scanning electron microscopy (SEM) equipped with EDS, X-ray photoelectron spectroscopy (XPS), and electrochemical measurements were used in this study to show the corrosion behavior and electrochemistry of Q235 carbon steel. Results indicate that the synergistic corrosion rate of Q235 carbon steel in alkaline solution containing sulfide and thiosulfate is larger than that of sulfide and thiosulfate alone, which could be due to redox reaction of sulfide and thiosulfate. The surface cracks and pitting characteristics of the specimens after corrosion were carefully examined and the corrosion products film is flake grains and defective. The main corrosion products of specimen induced by S and S 2 O 3 2− are FeS, FeS 2 , Fe 3 O 4 , and FeOOH.The present study shows that the corrosion mechanism of S and S 2 O 3 2− is different for the corrosion of Q235 carbon steel.


Introduction
The use of high-sulfur bauxite makes Bayer solutions contain high content of S 2− in recent years.Therefore, in Bayer liquor of alumina extraction besides Al(OH) 3 and caustic, other substances are impurities, including Na 2 SO 3 , Na 2 SO 4 , Na 2 S 2 O 3 , Na 2 S and other trace impurities, and so forth [1].The content of four kinds of sulfide is 0.4 g⋅L −1 , 0.8 g⋅L −1 , 1.2 g⋅L −1 , and 1.6 g⋅L −1 , respectively.Among them, Na 2 S 2 O 3 and Na 2 S are corrosion activators to steel corrosion, which remain the most serious forms of metal failure [2,3].However, as an important equipment material, carbon steel has been widely used and plays a significant role in alumina industry which is used to dissolution equipment, dilution tank, Bayer mother liquor evaporation equipment, and so forth.With the progress of production, the corrosion of sulfides has been recognized as a serious problem, which results in a great economic loss and corrosion cracking and pitting of steel equipment has been experienced in different alumina industry.So, the study on corrosion of steel equipment and control is a very critical issue.
The mechanism of steel corrosion and formation of the passive film have been always studied by many scholars [4][5][6].The role of sulfide in the highest concentrations of alkaline solution like white liquor has been studied on the corrosion behavior of carbon steel [7][8][9].But in the higher concentrations of caustic solution sulfur-contained species like alumina industry the research of the role of sulfide on steel corrosion is too sparse and corrosion behavior of 16MnR low alloy steel in sulfide-containing Bayer solutions was researched by Xie and Chen [3].The effect of S 2− , NaOH, and Al 2 O 3 on the corrosion of 16MnR low alloy steel had been studied by polarization curves and electrochemical impedance spectroscopy (EIS).But the effect of the synergy effect of S 2− and S 2 O 3 2− on Q235 carbon steel corrosion has not been discussed in detail.
Moreover, these corrosion experiments in literature have been performed using mostly conventional static stainless steel autoclave, dynamic and static immersion, and so on.The author will use the salt fog box to discuss corrosion of Q235 carbon steel in alkaline solutions containing sulfide and thiosulfate [10].The aim is to use the accelerated corrosion of the salt fog box.The purpose of the study is to evaluate corrosion behavior of Q235 carbon steel through additional sulfide and thiosulfate to alkaline solutions and corrosion experiments are made in the salt fog box.The effect of S 2− and S 2 O 3 2− on corrosion behavior of steel and mechanisms will be discussed in detail by employing weight loss measurements, scanning electron microscopy (SEM) equipped with an energy dispersive spectroscopy (EDS), X-ray photoelectron spectroscopy (XPS), and electrochemistry.The fundamental purposes are to gain the corrosion characteristics of Q235 carton steel in sulfur-containing alkaline solutions and to provide fundamental information for steel corrosion in sulfur-containing Bayer mother liquor and hence to provide insight into the understanding of the corrosion behavior and mechanism of Q235 carbon steel in alumina industry of the Southwest China.

Experimental
2.1.Specimens and Test Solutions.The chemical compositions (wt%) of Q235 carbon steel used in this study are listed in Table 1.The steel coupon was machined from steel with the 15 mm × 15 mm × 2.5 mm square type.Prior to experiment, the coupons were rinsed with distilled water, polished orderly with number 180, 240, 360, and 600 silicon carbide paper and degreased with acetone, and finally dried with cold air.
The sulfur-containing alkaline solutions contained 115 g⋅L −1 NaOH, 0, 5 g⋅L −1 S 2− , and 0, 4, 5 g⋅L −1 S 2 O 3 2− at 328 K.The solutions were prepared by dissolving NaOH, Na 2 S⋅9H 2 O, and Na 2 S 2 O 3 ⋅5H 2 O into 1000 mL of deionized water, respectively.The pH of the solutions is 14 by laboratory test, so H + can barely exist in such strong alkaline solution.All reagents are analytically pure.

The Corrosion Experiment.
All of the corrosion experiments were carried out for 120 h at 328 K (maximum temperature of the salt fog box) in the salt fog box (GB/T 24195-2009).The samples were taken out one by one at regular intervals to analyze the corrosion behavior.The test solutions were supplemented daily.After being taken out from the salt fog box, these coupons were rinsed using distilled water and dried with cold air.
Prior to each corrosion experiment, the coupon was weighed to sensitivity level of 0.0001 g using an analytical balance (AR1140/C) for the original weight ( 1 ) and its area () was measured.Before weight loss measurements, the corrosion products were removed using the chemical products cleanup method (the 500 mL of hydrochloric acid + 500 mL deionized water + 10 g six methyl tetramine solutions) (GB/T 6074-1992) [11].Finally, the sample was weighed again in order to obtain the final weight ( 2 ).
The corrosion rate was calculated using weight loss measurement and the formula to calculate is given as where  is corrosion rate (mm/a),  1 is the quality of the sample before the corrosion test (g) with an accuracy of ±0.0001 g,  2 is corrosion sample after the removal of corrosion products quality (g) with an accuracy of ±0.0001 g,  is the sample surface area (cm 2 ) with an accuracy of 0.01 cm 2 ,  is corrosion time (h) with an accuracy of 1 h, and  is the density of test steel (g/cm 3 ).

Surface Analysis.
The morphology and composition of Q235 carbon steel surface were observed with scanning electron microscopy (SEM) (SUPRA 40) and energy dispersive spectroscopy (EDS) (AZ tec.).The composition of the passive films was analyzed by means of PHI Quantum 2000 Scanning ESCA Microprobe Xrays photoelectron spectrometer (XPS).Monochromatized Al Ka radiation (1486.6 eV) was used as the excitation source.After being exposed to the solutions for different concentrations of S 2− and S 2 O 3 2− , the sample was dried and transferred to the XPS instrument for analysis.The binding energy values were calibrated with reference to the C 1s peak at 284.8 eV.The data processing and Photoelectron analytical peak would be parsed using Multipeak 8.0; the narrow scan spectra were fitted with XPSPeak 4.1 software.

Electrochemical Measurement.
Electrochemical testing was carried out in a conventional three-electrode cell at 338 K.The temperature of 338 K is to simulate the alumina dilution process.Using a saturated calomel electrode (SCE) as the reference electrode, platinum electrodes are auxiliary electrode and Q235 steel as working electrode.Before electrochemical experiment the sample was immersed for 120 h in the solutions.The work area of Q235 steel is 2.25 cm 2 , nonworking parts sealed with epoxy resin.The cell was placed in a water bath to achieve the test temperature.Potentiodynamic polarization curves measurements were performed at a potential scan rate of 1 mV/s.The potential range was from −1.5 V to 0.5 V.The EIS measurements were carried out at OCP over the frequency from 100 kHz to 10 mHz by 5 points per decade.All potentials reported in this paper were measured with respect to the SCE.All the tests were repeated to obtain reproducibility of results.

Corrosion Rate.
Weight loss measurement is a more reliable way to get the corrosion rate compared with other methods [12].Table 2 shows the corrosion rate (mm/a) of Q235 carbon steel based on weightlessness corrosion experiment in the alkaline solutions with different concentrations of S 2− and S 2 O 3 2− at 328 K.The corrosion rate of steel increased significantly in the presence of S 2 O 3 2− or S 2− , and the corrosion rate much increased when it is synergy corrosion.Thus it can be seen that S 2− and S 2 O 3 2− were corrosion When the concentration of S 2 O 3 2− is 5 g⋅L −1 , the loss of electrons of S 2− and S 2 O 3 2− inhibits the iron loss of electrons.So, the corrosion rate decreases as follows: This is consistent with what was previously reported that thiosulfate is corrosive activator [3] and inconsistent with the conclusions of Xie et al. [13].So, whether or not the effect of thiosulfate on corrosion of Q235 carbon steel is to accelerate or hinder which mainly depends on the concentration of sulfide when they are synergy corrosion [14].Besides, in strong alkaline medium, the solutions have a lot of OH − and hinder the secondary hydrolysis reaction of S 2− ion; therefore, the concentration of HS − is higher, is helpful to generate the sulfide film, and thus decreases the corrosion rate of Q235 carbon steel in the initial stage of corrosion [15].

Surface Morphology and Composition.
Figure 1 shows SEM images of corrosion products film on the specimen surfaces and the corresponding EDS in alkaline solutions with different concentrations S 2− and S 2 O 3 2− at 328 K after 120 hours of corrosion.It can be seen that the corrosion products cover unevenly the surface.
As shown in Figure 1, a large amount of corrosion is formed on steel surface with the additives of different concentrations S 2− and S 2 O 3 2− ; the corrosion is the most serious in the presence of S 2− (Figure 1(c)) and the presence of S 2− and S 2 O 3 2− (Figures 1(d) and 1(e)) [16].As can be seen from Figures 1(c), 1(d), and 1(e), there are selective corrosion and pitting with carbon steel.The corrosion surface with a loose and porous structure is easy to fall off, which cannot protect the metal matrix.The corrosion of Q235 carbon steel is slight in pure alkaline solution and steel surface has some defects and scratches (Figure 1(a)).Therefore, it is thought that the corrosion of Q235 carbon steel preferentially takes place in defects and scratches, mainly caused by polishing [17], whose surface roughness and surface residual have resulted in the faster corrosion rate.2, the anodic process exhibits active-passive-transpassive behaviors, which indicates that the surface of specimen formed compact corrosion products film at later stage and the anodic iron dissolution and the transformation of iron compounds.This could be due to the fact that the corrosion process involved transformation of corrosion products, always maintaining the surface active [18].In alkaline solutions regardless of the concentrations of S 2− and S 2 O 3 2− , Q235 carbon steel exhibits corrosion potential ( corr ) near −1.20 V versus SCE in the active state.At the potentials −1.0 V versus SCE, the curves exhibit a humped shape, revealing at least two reactions including one for the exposed steel anodically dissolved into the electrolyte to form FeO 2 − ions and another one attributed to the oxidation of S 2 O 3 2− to higher forms, as described by Zou and Chin [19].
The determination of corrosion parameters ( corr ,   ,   ,  corr ,   , and   ) could provide more information about the overall corrosion process.Table 4 shows the corrosion potentials ( corr ), Tafel slopes (  and   represent anodic and cathodic, resp.), corrosion currents ( corr ), critical current density (  ), and polarization resistance (  ) obtained from Tafel fitting of the polarization curves data.The higher critical current density suggests that corrosion of steel increases.When S 2− concentration is 5 g⋅L −1 , the presence of S 2 O 3 2− caused a considerable increase in the critical current density  (Table 4) which shows that S 2 O 3 2− is activator; however, the increase of S 2 O 3 2− concentration decelerated the corrosion of Q235 carbon steel when the concentration of S 2− is 5 g⋅L −1 [14], reflecting that S 2− played a more important role in the corrosion of Q235 carbon steel.Anodic Tafel slopes,   (Table 4), in S 2− -containing solutions are very similar (about 120 mV), regardless of the concentration of S 2 O 3 2− .
However, the slopes of the anodic branch,   (Table 4), are considerably dependent on the concentration of S 2 O 3 2− .This fact, along with higher values of these slopes compared with those from the anodic branch, indicates the complex nature of the oxidation process.The change of   is the opposite trend to the change of   (Table 4).The shape of cathodic polarization curve of S 2 O 3 2− is different from S 2− , which shows that the International Journal of Corrosion 5 Potential, E we (V)  3 indicate that the overall impedance of steel in alkaline solution without S 2− is the largest, but then it decreases in the presence of S 2− , suggesting an activation of the corrosion processes, as expected in the presence of sulfides [2].In the presence of S 2− , the overall impedance reduces when the S 2 O 3 2− concentration increases from 0 to 4 g⋅L −1 , after increasing the S 2 O 3 2− concentration being 5 g⋅L −1 , and makes the surface more sensitive to S 2− and S 2 O 3 2− attack.These data of EIS testes in Figure 3 further support the results from potentiodynamic polarization curves in Figure 2. Deconvolution of the phase angle curves of Bode plots in Figure 3 indicate that two overlapped time constants are found for Q235 carbon steel in solutions in the presence of S 2− and S 2 O 3 2− .The experimental data was then fitted by -View, and the results are shown in Figure 3, which is in accordance with some references [20,21].As shown, the fitted curves are in good agreement with corresponding experimental data.To further understand the processes and mechanisms of Q235 carbon steel corrosion, electrical equivalent circuit was constructed (see Figure 4).In the model,   denotes the resistance between the working electrode and reference electrode,   and   represent, respectively, the resistance and capacitance of the oxidation film or corrosion products film, and  ct and  dl are the charge transfer resistance and constant phase element (CPE), respectively.The reason of  for capacitors in analysis of impedance spectra is that the passive film is not considered as a homogeneous layer but rather as a defective layer.The impedance of  is defined as where  0 is the admittance,  is angular frequency in rad/s,  2 = (−1), and  is an exponential term which always lies between 0.5 and 1.0.When  = 1, the CPE describes an ideal capacitor.When  = 0.5, the CPE represents a Warburg impedance with diffusional character.For 0.5 <  < 1, the CPE describes a distribution of dielectric relaxation times in frequency space.From the fitted values of electrical equivalent circuits (listed in Table 5), one can find that after the corrosion product film is generated,  1 and  2 decreases with the concentration of S 2 O 3 2− , suggesting that the corrosion of the steel became more and more serious [22].The change of   also corresponds to the change of polarization results (Table 4).Equivalent circuit mode shows that passive film does not completely cover the metal surface, which is due to surface roughness and other reasons that can cause the dispersion effect.The passive film is not a uniform layer, but a porous film of characteristic.Results are consistent with SEM-EDS analysis.

Surface Analysis of Specimens by XPS.
Figure 5 shows the decomposition of peaks for Fe2p3, S2p, and O1s of the passive films formed on Q235 carbon steel.Peak decompositions of Fe2p3, S2p, and O1s core level spectra are in agreement with published results [23][24][25].In the O1s spectrum, a broad peak and two narrow peaks are observed at about 528-536 eV in the solutions.The O1s spectrum can be curve-fitted with two peaks.The peak at 529.92 eV is attributed to O 2− in oxides.The peak at 531.5 eV can be attributed to oxygen in FeOOH.The Fe2p3 spectrum can be curve-fitted with four peaks.The peaks at 706.91, 708.19, 710.49, and 712.82 eV correspond to FeS 2 , Fe 3 O 4 , and FeS.The formation of a stable layer (Fe 3 O 4 ), which can inhibit the diffusion of corrosive species, will be  beneficial to improve the corrosion resistance of specimen.The S2p spectrum can be curve-fitted with four peaks.The peaks at 163.02, 164.17, 161.84, and 162.99 eV correspond to FeS 2 and FeS.Table 6 shows XPS peak position and peak result analysis of the corrosion scale of Q235 carbon steel, which indicates that the corrosion scale of Q235 carbon steel mainly consisted of FeS, FeS 2 , Fe 3 O 4 , and FeOOH.

Figure 2 :
Figure 2: Potentiodynamic polarization curves of Q235 carbon steel in solutions with different concentrations of S 2− and S 2 O 3 2− at 338 K.

Figure 3 :
Figure 3: EIS of Q235 carbon steel in alkaline solutions with different concentrations of S 2− and S 2 O 3 2− solutions.(a) Nyquist plot and amplificatory plot and (b) Bode plot (phase angle versus frequency).

Figure 4 :
Figure 4: Equivalent circuit for EIS of Q235 steels in alkaline.

Table 2 :
Corrosion rate of Q235 carbon steel in alkaline solutions.

Table 3 :
Element distributions of corrosion products with different concentrations of sulfide and thiosulfate at 328 K (wt.%).S 2− and 4 g⋅L −1 S 2 O 3 S 2− and 0 g⋅L −1 S 2 O 3 Table 3 lists the distribution of elements with Q235 carbon steel for different concentrations of S 2− and S 2 O 3 2− in solutions at 328 K. From Table 3 it can be seen that the oxygen content is largest in the absence of S 2− and the sulfur content is largest in the absence of S 2 O 3 2− , which decreases later with the concentration of S 2 O 3 2− .The results show that the oxidation film is generated at the steel surface mainly in the absence of S 2− , and the corrosion products are sulfide in the absence of S 2 O 3 2− .The compactness and stability of oxidation film are higher than those of sulfide.Results of EDS indicate that the corrosion mechanism of Q235 carbon steel S 2− and S 2 O 3 2− is different.