The effect of microstructure on the initial corrosion behavior of low carbon steel in simulated coal solution

The service life of weathering steels in wagon body was determined by their corrosion resistance directly. This study investigated the influence of microstructure on the initial corrosion behavior of low carbon steels, systematically. The initial corrosion behavior of ferritic-bainitic (F+B) steel, bainitic (B) steel and ferritic-pearlitic (F+P) steel are thoroughly analyzed using coal leach solution immersion test, macroscopic and microcircuit electrochemical methods. The results revealed that F+B steel exhibited the highest corrosion resistance, with the potential of M-A islands surpassing that of ferrite. The initial corrosion initiates from the dissolution of the ferrite matrix, followed by detachment of the M-A islands. The potential of M-A islands is higher than that of bainitic ferrite lath, and the corrosion originates from ferrite dissolution in B steel. Moreover, F+P steel exhibited the largest potential difference between pearlitic and ferrite, leading to initiation of corrosion from the pearlitic corrosion of internal ferrite. Additionally, the multiphase characteristics of P in F+P steel exacerbates their corrosion susceptibility. Overall, the influence of microstructure on the initial corrosion behavior of low carbon steels can be attributed to the potential difference between different phases. Ferrite is the preferentially dissolved phase due to its negative potential difference.


Introduction
Weathering steel is a kind of low-alloy high-strength steel with a carbon content less than 0.2 wt%.Its service life can reach 2 to 8 times than that of general carbon steel [1,2].One of the key advantages of weathering steel is its high corrosion resistance without the need for expensive anti-corrosion coatings.It also possesses excellent mechanical and welding performance, making it widely used in various applications, such as body steel for railroad wagons.During the coal transportation process, water is often sprinkled to prevent environmental pollution and coal loss caused by dust.The sulfur element in coal reacts with water, leading to the generation of sulfuric acid and sulfurous acid [3].Therefore, the railroad coal wagons are exposed to not only the water and atmospheric conditions, but also the coal leaching solutions.The weathering steel are prone to corrode in acidic environments due to direct contact between the wagon body and the coal block.
The mechanical properties and corrosion resistance of steel play a crucial role in determining the service life of the vehicles.Achieving a harmonious balance of good toughness and corrosion resistance necessitates the establishment of reasonable microstructure and properties [3].The railroad wagon body steel has undergone three distinct stages of development.Initially, plain carbon steel (Q235) was utilized in the 1950s and 1960s, characterized by low strength and poor corrosion resistance.This was followed by the adoption of ordinary weathering steel (09MnCuPTi series) in the 1980s, which offered improved corrosion resistance but had lower tensile strength of approximately 300 MPa.Subsequently, high-strength weathering steel (Q450NQR1, S500AW) has been employed since the beginning of the century and commonly utilized in truck bodies weighing over 70 tons.While this steel variant boasts high strength and toughness despite its ability to accommodate current demands for heavy loads, it still falls short of corrosion resistance in comparison to the 09MnCuPTi series [4].Consequently, the present steel has not met the Ministry of Railways' requirement for a 25-year service life.
The corrosion behavior was determined by the surface structure and unevenness microstructure in the metal.Especially, the corrosion was prone to occurred in the combined effect of SO 4 2− and Cl − in the coal environment.Therefore, the acid environment caused by coal leaching solution was the may reason induced the steel corrosion.Apart from the acid environment, [5,6] solid solution alloying elements, the microstructure and inclusions in steel, will significantly influence the corrosion behavior of the steel during the initial exposure to corrosive media.Therefore, a systematic study on the corrosion behavior of steel in coal leaching solutions has great significance in developing weathering steel.
Currently, extensive researches have been conducted on the corrosion behavior of weathering steels [7].It has been discovered that both the initial microstructure of the steel and the rust layer formed during the corrosion process have an impact on the corrosion resistance [8,9].For instance, the spacing between cementite and pearlite lamella can degrade corrosion resistance of ferritic-pearlitic (F+P) steels [10] due to the preferential dissolution of eutectic and lamellar ferrite surrounding and within the pearlite region.In ferritic-bainitic (F+B) steels, the M-A island in bainite promoted the dissolution of ferrite (BF) matrix in bainite [11].The corrosion resistance of purely bainitic steels was superior than F+P steels [12].The martensite promoted the dissolution of ferrite in acidic solutions [13,14].Furthermore, the formation of a dense rust layer can effectively impede the penetration of aggressive particles due to the formation of α-FeOOH and γ-FeOOH [15].
Research on the corrosion behavior of weathering steel primarily focuses on the role of alloying elements and the protective properties of the rust layer.However, there is a lack of comprehensive studies on the initial corrosion behavior and influencing mechanism of different types of microstructures in weathering steels.Moreover, the relationship between the microstructural characteristics of weathering steels and their initial corrosion behavior remains unclear.Specifically, there is a lack of unified understanding regarding the corrosion resistance and mechanism of high-strength Sb-contained steels.Optimizing microstructure and phase ratio of train carriages steels and regulating their structure types to obtain train carriages steels with excellent corrosion resistance is a valuable research pursuit.
The author's team has developed a novel high-strength stainless steel by microalloying with Sb element, which confirmed that the addition of Sb improved the comprehensive performance of stainless steel for car body [16].The structure of the surface rust layer, the physical phase type, and the surrounding environment are crucial factors in determining its corrosion resistance [17].In this paper, the initial corrosion behavior of this new Sb-contained steel in the simulated coal leaching solution will be analyzed.Different microstructures were obtained through different post-roll cooling processes.The relationship between different initial microstructures and initial corrosion behavior will be analyzed.This investigation will provide theoretical support for regulation of microstructure and design of new weathering steel.

Experimental
The stainless steel was prepared by Shougang Institute of Technology and their chemical composition is shown in table 1.The steel was heated to 1200 °C and held for 2 h, followed by 7 passes of hot rolling (total deformation of 88.6%).The final rolling temperature was about 860 °C to 840 °C.After rolling, three cooling processes (figure 1) were employed to achieve different initial microstructures.The microstructures of F+B, B, and F+P steels with three different microstructure types are displayed in figure 2. F+B steel with granular bainite (GB) was obtained by air cooling to room temperature.B steel which was characterized by a combination of lath bainite (LB) and granular bainite (GB) was obtained by water quenching to room temperature.F+P steel with a mixture of ferrite and pearlite (F+P) was obtained by controlled cooling in resistance furnace.
These three samples were immersed in simulated coal solutions composed of 3.5 wt% NaCl, 0.25 wt% MgCl 2 , 1 wt% CaCl 2 , 0.25 wt% KCl, and 10 wt% Na 2 SO 4 .The pH value was adjusted to 4.0 by addition of HCl.Subsequently, the samples were immersed in simulated coal solution for 5, 10, 30, 60, and 120 min.After each respective immersion period, the samples were taken out and placed in a descaling solution (500 ml HCl + 500 ml H 2 O + 3.5 g hexamethylenetetramine) for descaling.They were then rinsed with anhydrous ethanol and blown dry.The corrosion morphology was characterized in three dimensions using a scanning electron  OLS4100).Electrochemical experiments were conducted on an electrochemical workstation (Bruker Icon) to determine the polarization curves and impedance spectra of the weathering steels and the materials after immersion for different periods.Micro-area electrochemical analyses were performed using an atomic force microscope (RST5200F) to measure the potential difference between different phases.

Results and discussions
During the immersion experiments, it was observed that corrosion pits started to appear on the surface of steel after 5 min.Then the steel entered into a fully corroded state after 30 min.To illustrate the initial corrosion  behavior, the corrosion surface after immersing in the simulated coal solutions for 5, 30, and 60 min were detected, respectively.The semi-in situ test results of the corrosion morphology and 3D morphology of the descaled surface are depicted presented in figures 3-5.

Initial corrosion morphology of different steels
Figure 3 shows the SEM and 3D morphology of F+B steel immersed in the simulated coal leaching solution.The red straight line represented the position of roughness measurement.Figure 3(a) illustrates the morphology after 5 min of immersion, showing evident corrosion occurrences.Numerous small and shallow corrosion pits have emerged at the junction of the M-A island and ferrite.Notably, the M-A island exhibited a slightly higher elevation than the ferrite, indicating a slower corrosion rate for M-A island compared to the ferrite.After 30 min of immersion (figure 3(b)), more stripes and granular M-A islands were visible on the sample's surface.The  corrosion points were situated at the junction of M-A islands and ferrite, closer to the ferrite phase.The size of the corrosion pits was larger than those observed at 5 min, displaying a pronounced convexity and a pit depth of Rz, which was about 0.176 μm (figures 3b1 and b2).Thus the corrosion of ferrite was exacerbated.The number of corrosion pits on the F+B steel surface escalated with increasing the immersion time to 60 min (figure 3(c)).
The Rz value was reached to 1.301 μm.A phenomenon emerged where adjacent small corrosion pits merged into larger pits or corrosion strips.In summary, both ferrite and M-A islands can be dissolved in the coal leaching solution in F+B steel.The dissolution rate of ferrite was notably faster than that of the M-A islands, particularly at the boundary between these two phases.
The corrosion microstructure of B steel immersed in different periods are presented in figure 4. As was depicted in figure 4(a), when immersed for 5 min, B steel experienced slight corrosion, with corrosion points primarily located at the junction of M-A islands and ferrite between the BF lath packets.The depth of the corrosion pits, as indicated by Rz, was only 0.17 μm (figure 4(a1)).With extending the immersion time to 30 min (figure 4(b)), it can be observed that the corrosion pits became deeper and larger in size.These deeper corrosion pits originated from the detachment of smaller M-A islands and subsequent to the corrosion of the surrounding ferrite.After 60 min of immersion (figure 4(c), the lath-like bainitic ferrite became less apparent, and the corrosion points distributed more uniformly compared with the steel immersed for 30 min.
Figure 5 depicts the SEM and 3D morphology of F+P steel after removing the surface rust layer through immersion in the simulated coal solution for different periods (5, 30, and 60 min).After immersing for 5 min (figure 5(a), F+P steel exhibited slight corrosion with corrosion points primarily located within the pearlite and at the junction of pearlite and ferrite.The 3D morphology test (figures 5(a1) and (a2)) results indicated that the depth of corrosion points was 0.111 μm.It can be obtained that selective corrosion occurred with extending the immersion time to 30 min (figure 5(b)).This resulted in the appearance of gaps between the carburized lamella that were no longer in the form of dense lamella.This phenomenon was associated with the corrosion of ferrite between the lamella.In addition, the ferrite within the pearlitic structure continues to corrode, with the pearlitic lamella becoming more prominent, and the space between the lamella continuing to increase.After 60 min of immersion (figure 5(c)), the carburized lamella was more pronounced and protruding compared to 30 min.At this point, the ferrite outside the pearlite also began to corrode, resulting in a dent next to the pearlite.

Electrochemical performance of initial corrosion steels
In an acidic environment, a distinct layer of corrosion products that influence the corrosion process will be formed.To evaluate the electrochemical behavior of different steel in the early stages of corrosion, these three steels (F + B, B, and F + P) were immersed in simulated coal solutions for varying durations (0, 8, and 32 h).Electrochemical analyses of polarization curves were conducted, and the results are presented in figure 6.The fitted data of self-corrosion current density and self-corrosion potential derived from the polarization curves are illustrated in figure 7. To further elucidate the evolution of the electrochemical corrosion properties of the steel,   order of F+B > B > F+P.This suggested that F+B steel exhibited the highest corrosion resistance, followed by B steel, and F+P steel with the lowest corrosion resistance [18].
The Bode diagram is typically used to analyze corrosion resistance by evaluating the low-frequency modulus value and the high-frequency phase angle value.The low-frequency modulus value represented the resistance of the rust layer against erosive fluid, with a higher value indicating better protection and corrosion resistance [19].From the AC impedance spectra in figures 6(d)-(f), it can be observed that the Nyquist curves exhibited a similar shape, characterized by a smaller compressed semicircular arc in the medium/high frequency region and a larger semicircular arc in the low-frequency region.This indicated that the electrochemical mechanisms of the three types of steel were similar.However, the F+B steel showed the largest radius of the capacitive arcs, indicating the best corrosion resistance.F+P steel performed the lowest corrosion resistance.The magnitude of the low-frequency modulus in the frequency-modulus plots (figures 6(d′)-(f′)) followed the order of F +B > B > F+P which was consistent with the Nyquist plots.Additionally, the high-frequency phase angle of F +B steel was higher than that of B steel, and the smallest phase angle was observed for F+P steel, which aligned with the Nyquist plot results.
From figure 7, it was evident that the corrosion current of F+B steel kept at lower level and the corrosion potential consistently remained the highest.This indicated that F+B steel possessed superior corrosion resistance.The corrosion potential of all three steels decreased significantly after 8 h of immersion in the simulated coal leaching solution, which can be attributed to the formation of corrosion products.The corrosion products layer acted as a barrier, impeding the transport of corrosive substances and ions.However, the corrosion potentials and corrosion currents of F+B and B steels exhibited a slight increase with increasing immersion period, suggesting that the corrosion products provided a certain level of protection.The elevation of corrosion potential and corrosion current was more pronounced in B steel.Conversely, the corrosion potential of F+P steel exhibited a continuous decreasing trend, indicating poor protection of the loose rust layer formed on its surface.On the other hand, the dense rust layer on the surface of B steel provided better protection.
The equivalent circuit fitting values are presented in table 2. It revealed that the charge transfer resistances (R ct ) of F+B, B, and F+P steels underwent a process of decreasing, then increasing, and finally stabilizing.This pattern suggested that corrosion led to the formation of corrosion products, and the protective rust layer ultimately stabilizes the corrosion rate.Notably, the value of R ct was the highest for F+B steel.Therefore, the electrochemical analysis results indicated that F+B steel exhibited the optimal corrosion resistance which was in conjunction with the immersion test results.In F+B steel (figure 9(a)), the potential of M-A island was approximately 11 mV, while the potential of the ferrite was about −12mV.The significant potential difference (23 mV) in the corrosive medium environment driven the formation of primary corrosion cell.The M-A island in F+B steel acted as the cathode and ferrite acted as the anode.The corrosion initiated from the ferrite adjacent to the M-A island and accelerated the formation of a corrosion budding point.For the LB in B steel (figure 9(b)), the M-A island exhibited a higher potential (about 25 mV) than the substrate (−5 mV).In case of F+P steel (figure 9(c)), the lamellar carburite exhibited a high potential (about 7 mV) in F+P steel, while the ferrite exhibited a lower potential of approximately −15 mV.Furthermore, the pearlite in F+P steel (figure 9(d)) demonstrated a higher potential of 10mV, and the ferrite demonstrated a lower potential of about −30 mV, resulting in a notable potential difference between the lamella (40 mV).Therefore, it was evident that the corrosion potential of ferrite consistently remained negative which indicated that ferrite was the initial site of corrosion in the steel.

Mechanism of influence of microstructures on initial corrosion behavior
It is well known that multi-phase alloys are more prone to corrosion than single-phase alloys.During the initial stages of corrosion in acidic solutions, the carburite and ferrite phases are particularly susceptible to protocell corrosion.In the present study, the F+P steel had two distinct phases, F and P. The LB in B steels consisted of lath-like BF and M-A islands.It was evident that differed from F, the intricate multiphase microstructures of P and LB phases will cause galvanic coupling corrosion, resulting in heightened electrochemical activity and greater susceptibility to corrosion.The effect of initial corrosion behavior of different steels was schematic presented in figure 10.
In F+B steel, the M-A island served as cathode and accelerated the dissolution of BF matrix.Subsequently, the number of detached carburizers was minimal and uniformly distributed.In the initial stage of corrosion, corrosion pits emerged later and a small number of pits was existed in the surface area.As the corrosion time prolonged, the density of corrosion pits exhibited in a growing trend.However, the density of corrosion pits in F +B steel remained relatively lower and discontinuous compared with B steel.
The initial corrosion behavior of low carbon steel in coal leaching solutions revealed that the primary corrosion of B steel was due to the fine carburite distributed in ferrite matrix in the form of grains or strips.The density of primary cells caused by arburite and ferrite was higher than that of F+B and F+P steels, which resulted in the preferential corrosion of B steel.As the immersion time progressed, the number of corrosion pits in B steel matrix significantly increased, and the corrosion products became more interconnected.This led to the full corrosion state within 30 min.Subsequently, the formation of a dense rust layer hampered the contact between acidic solution and substrate, thereby reducing the corrosion process of the micro-area primary cell [13,20,21].F+P steel formed a complex primary battery in the corrosive environment.Initially, F served as the anode and P as the cathode.The corrosion started from the F which adjacent to P. Subsequently, the F with a lower potential in P preferentially dissolved and the carburizer was protected by its higher potential.Finally, the carburizer next to F began to corrode until the outer layer of F was completely corroded.However, the distribution of carburizer in F was relatively more dispersed which resulted in the formation of initial corrosion pits in F.Moreover, the area of exposed carburizer gradually increased and a porous layer formed on the surface.The loose corrosion layer failed to protect the internal structure.In addition, during high-temperature heat treatment process, the transformation of P led to an uneven spatial distribution of alloying elements and resulted in unstable electrochemical activity.Consequently, F+P steel exhibited the poorest corrosion resistance in the present study.

Conclusions
The microstructures exerted a significant influence on the corrosion resistance of low carbon steels.The main conclusions are as follows.
(1) The corrosion process of F+B, B, and F+P steels, prepared through various post-roll cooling processes, are primarily governed by charge transfer process.The charge transfer resistance follows the order of F +B > B > F+P, while the corrosion current density follows the order of F+B < B < F+P.Therefore, F+P steel has the lowest corrosion resistance.
(2) The potential difference within the microstructure plays a crucial role in determining the corrosion resistance of steels.F+P exhibits a large potential difference (40 mV) between its two phases, whereas the potential difference in the F+B microstructure is only 23 mV.F+B steel demonstrates superior corrosion resistance, while F+P steel exhibits the lowest corrosion resistance.The initial corrosion in F+B steel initiates on the ferrite substrate adjacent to M-A island.The B steel was corroded at interface between lamella beam BF and M-A island.The corrosion initiated from the within the pearlite in F+P steel.
(3) Differed from F phase, the complex microstructure of P and LB present induce galvanic coupling corrosion, resulting in higher electrochemical activity and greater propensity for corrosion.The ferrite and carburite within the pearlite form a corrosion protocell, where the ferrite gradually erodes as the anode, and the carburite lamella becomes more prominent.Eventually, a corrosion primary cell formed between the carburizer and external ferrite of pearlite, leading to the gradual erosion of the ferrite adjacent to the pearlite.

Figure 1 .
Figure 1.Schematic diagram of heat treatment profiles.

Figure 7 .
Figure 7. Self-corrosion current density and self-corrosion potential after fitting to the polarization curve.

3. 3 .
Microcircuit electrochemistry properties of microstructures In order to elucidate the corrosion resistance of different microstructures in low carbon steels, Scanning Kelvin Probe Force Microscopy (SKPFM) was employed to acquire micro-area electrochemical information at the interfaces and the test results are depicted in figure 9.The surface potential diagrams in figure 9 illustrated the potential distribution of different microstructures, including GB microstructure in F+B steel (figure 9(a)), LB microstructure in B steel (figure 9(b)), and F and P microstructure in L+P steel (figures 9(c) and (d)), respectively.