Comparative electrochemical impedance spectroscopy study in chloride environment of nano-structured bainitic steels prepared by austempering at varied temperatures

Steel of chemical composition Fe-0.86% C-1.4% Si-1.8% Mn-0.47% Al-1.88% Co- 0.23% Mo (in weight %) has been used in the current work to obtain nanobainitic structure after heat treatment. Presence of high carbon in the alloy decreases the transformation temperatures for metastable phases (bainite, martensite) and increases the gap between bainitic start (B_S) and martensitic start (M_S) temperature [8, 20, 21]. This makes it possible to form nanobainite over a range of temperatures (T) such that M_S < T < B_S. Addition of Si has been done to prevent the precipitation of cementite during the transformation [22, 23]. Mn has been added to decrease the B_S temperature and to delay the pearlitic transformation by shifting the time-temperature- transformation (TTT) curves towards a larger time [24]. Alloying elements like Co and Al have been added to accelerate the kinetics of bainite formation [25, 26]. The steel has been produced in a vacuum induction furnace and the ingot has been homogenized at 1000 °C for 48 h followed by hot rolling in multiple passes with 57% reduction in thickness. Multiple steel blocks comprising of nanobainite have been obtained by austenitizing the homogenized sample at 1000 °C for 75 min followed by austempering at 250 °C, 300 °C and 350 °C respectively in a salt bath furnace for 48 h. The isothermal temperatures were selected based on the B_S (400 °C) and M_S (120 °C) determined from empirical equations [17, 27–29]. A difference of 50 °C between the three isothermal transformation temperatures was chosen to get different morphologies and content of bainitic ferrite and retained. The specimens extracted from the different steel blocks have been referred to as NB250, NB300 and NB350 based on the respective austempering temperatures of the steel blocks.

The initial microstructures of all samples have been examined before electrochemical testing using scanning electron microscopy (SEM) and X-ray diffraction (XRD). Small rectangular specimens of size 10 × 5 × 5 mm³ have been cut from each austempered block for metallographic examination. The specimens have been polished to a surface finish of 1 μm using silicon carbide papers and diamond paste followed by chemical etching with 2% Nital. X-ray diffraction (XRD) of all three specimens has been carried out in PANalytical X'pert Pro MPD system with Cu-Kα radiations using a monochromator detector to determine the volume fraction of phases in each specimen. A current of 30 mA and a voltage of 40 kV has been applied during the scan over a 2ϴ range of 40°–120° with a step size of 0.01°. The XRD data has been subsequently analyzed using X'Pert Highscore software and volume fraction of bainite and retained austenite was calculated from integrated intensities of ferrite {110}, {200}, {211}, {220} peaks and austenite {111}, {200}, {220}, {311} peaks respectively [30, 31].

EIS and potentiodynamic polarization have been used to evaluate the corrosion behavior of nanostructured bainitic steel. Specimens of size 5 × 10 × 5 mm³ have been cut from the heat-treated blocks. The samples were mounted in a cold setting resin such that an area of 0.5 cm² was exposed. The exposed surface of sample was polished to 1 μm finish to avoid any micro grooves and rinsed with ethanol and deionized (DI) water prior to electrochemical tests. All the crevices on the exposed surface were sealed using a quick setting epoxy to stop the penetration of electrolyte. A three-electrode set up was used with the bainitic specimen being the working electrode, saturated calomel electrode (SCE) with standard potential of E°_SHE = +0.244 V as a reference and platinum mesh with 1 cm² area as a counter electrode. The electrolyte used in this study was 3.5 wt% NaCl solution. The electrochemical tests were performed on AUTOLAB PGSTAT204 potentiostat. Prior to EIS and potentiodynamic polarization experiments, the samples were kept immersed in the electrolyte for 5 h to attain the open circuit potential (OCP). The EIS was carried out by applying sinusoidal potential of 10 mV amplitude in frequency range from 10⁵ to 10⁻² Hz at the respective OCP values of the specimens. Impedance data was fitted to an equivalent circuit using NOVA 2.1 software. Polarization tests were performed at ambient temperature at a scan rate of 0.166 mV s⁻¹ in range of ±1.5 V_SCE with respect to OCP. Corrosion current density and polarization resistance for the different specimens were calculated using circuit fitting and polarization plot extrapolation values. The polarization resistance and corrosion rate has been calculated by the equations shown below (equations (1), (2)) using corrosion current density and slopes:

$$\begin{eqnarray}&&{R}_{P}\,\left({\rm{\Omega }}\,{{\rm{cm}}}^{2}\right)=\,\displaystyle \frac{{\beta }_{A}\times {\beta }_{C}}{2.3\,\left({i}_{corr}\right)({\beta }_{A}+\,{\beta }_{C})}\end{eqnarray} \tag{ 1 }$$

$$\begin{eqnarray}&&CR\,\left(mpy\right)=\,\displaystyle \frac{0.13\,\times {i}_{corr}\times (E.W)}{d}\end{eqnarray} \tag{ 2 }$$

R_P represents polarization resistance in ohm cm², i_corr is corrosion current density in μA cm⁻², β_A and β_C are slopes expressed in volts per decade for anodic and cathodic regions respectively, CR represents corrosion rate in milli inches per year (mpy), E.W is the equivalent weight of iron and 'd' is density in g cm⁻³. The corrosion current density has been calculated by extrapolating the linear portion of anodic curve to the point where it cuts the corrosion potential. The anodic slope was measured above ±100 mV of E_corr over 1 decade.

The detailed microstructural characterization of the three specimens was done using SEM and XRD. Figure 1 shows the microstructure of nanostructured bainitic specimens after austempering at 250, 300 and 350 °C respectively. Two phases with distinct morphologies can be observed in the micrographs: fine laths of BF separated by films and blocks of RA. Figure 1 also shows the refinement in the morphology of microstructure with a decrease in austempering temperature. Nanobainitic transformation is displacive in nature and doesn't involve diffusion of substitutional alloying elements like Mn, Si, Al, Co and Mo between the parent and product phase [29, 32–34]. However, carbon partitions from BF to RA during the transformation resulting in chemical heterogeneity between the two phases [35–37]. The volume fraction of RA and BF has been calculated from XRD and is shown in table 1.

Zoom In Zoom Out Reset image size

Figure 2 shows the XRD spectra of isothermally transformed specimens prior to immersion test. The intensity of ferrite peaks in figure 2 grows stronger with decreasing austempering temperature and is an indication of a large fraction of bainitic ferrite. The volume fraction of BF and RA in NSB steels is influenced by the austempering temperature for a fixed transformation time. A lower austempering temperature offers a large free energy change from austenite to ferrite promoting the formation of larger fractions of BF in the final microstructure. Moreover, a decrease in austempering temperature results in formation of finer bainitic laths and films of RA.

Zoom In Zoom Out Reset image size

The electrochemical characteristics of all austempered specimens have been analyzed after extracting the data from EIS and potentiodynamic polarization measurements. The results from the EIS tests are shown in the form of a) Nyquist plot and (b) Bode plot for all specimens in figure 3 and 4 respectively. Nyquist curve is plotted between the imaginary (Zʹʹ Ω cm²) and real (Zʹ Ω cm²) impedance component at different excitation frequencies (figure 3).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The shape of the Nyquist plot reflects the electrochemical behavior and the output has been simulated using electrical circuit with components like resistors and capacitors (circuit shown in figure 5). R_S and R_CT represent the solution resistance and charge transfer resistance respectively. The diameter of the semicircle along the x-axis on Nyquist plot represents the charge transfer resistance (R_CT). It can be seen from figure 3 that the diameter of the semicircle decreases with an increase in bainitic transformation temperature signifying lowest value of R_CT for NB350 specimen. The shape of Nyquist plot shown in figure 3 isn't a perfect semicircle. A depressed and distorted semicircle often refers to a non-ideal capacitive behavior and therefore it is appropriate to simulate the response using a constant phase element (Z_CPE). The constant phase element can simulate the electrode surface roughness and heterogeneities as well as the non-uniform charging of double layer formed at the interface of metal and electrolyte [12, 15] The equation for Z_CPE is given below:

$$\begin{eqnarray}&&{Z}_{CPE}=\displaystyle \frac{1}{{Q}_{dl}{\left(j\omega \right)}^{n}}\end{eqnarray} \tag{ 3 }$$

where Q_dl is a constant with units mΩ⁻¹ cm⁻² sⁿ, 'j' is the imaginary number, 'ω' is the angular frequency. The value of n varies between 0 < n < 1; 0 for an ideal resistor and 1 for pure capacitor. The EIS data has been fitted with an equivalent circuit using electrochemical software Nova 2.1 and is shown in figure 5.

Zoom In Zoom Out Reset image size

The modulus of impedance and phase angle is plotted with respect to the frequency in a Bode plot as shown in figure 4. The Bode modulus plot clearly shows the trend for the variation with austempering temperature of charge transfer and solution resistance at low and high frequencies respectively. The metal-electrolyte system can be considered as an electrical circuit and electrons will always follow the path of least impedance. At high frequency, the capacitance is negligible due to its inverse proportionality to frequency and therefore only the value of R_S is accounted. At low frequency, capacitance is very high and therefore, electrons flow through R_CT. This gives the total resistance of R_S + R_CT at low frequency. No distinct variation amongst different specimens is observed in the Bode magnitude plot at high frequency due to similar values of solution resistance however, at low frequency each specimen exhibits a clear variation in the impedance. The value of capacitance cannot be directly determined from either of the plots (Nyquist or Bode) and therefore, circuit fitting is required.

All parameters obtained after circuit fitting are compiled in table 2. It can be seen from table 2 that specimen NB250 shows the highest corrosion resistance which can be observed from the Nyquist plot and Bode magnitude plot as well. The R_CT of NB250 is 42% greater than that of NB350. The value of Q_dl represents the capacitance of the double layer formed at the interface of specimen and electrolyte. The highest Q_dl was observed in NB350 specimen with the lowest 'n'. A slight evidence of two-time constants (horizontal region at the peak of phase -frequency plot for NB250) can be observed in figure 4 for NB250 in Bode magnitude plot. The two peaks however are not properly resolved and hence only a single time constant circuit is fitted for NB250. This behavior is in general a result of accumulated corrosion products on surface acting as the second film. However, in the present case for NB250, the layer is not stable enough to yield a separate capacitive peak on the Bode plot [38].

Potentiodynamic polarization curves in figure 6 and the extracted corrosion potential (E_corr) and corrosion current density (i_corr) in table 3 show that the E_corr moves towards the active side and i_corr increases with austempering temperature. NB350 shows the highest corrosion current density and therefore is more prone to corrosion than NB300 and NB250. Equation (1) has been used to calculate the polarization resistance and table 3 shows that polarization resistance decreases with an increase in austempering temperature. NB250 in chloride solution experiences a polarization resistance of 1.48 kΩ cm² whereas the specimen with the highest fraction of RA (NB350) exhibited lowest resistance of 1.02 kΩ cm².

Zoom In Zoom Out Reset image size