1 Introduction

Owing to the combined attack of the chloride and acids released during the decomposition of thermoplastics (such as PVC) via overheating [1], the working environments for plastic injection molding are extremely hostile. To extend the working life of plastic injection molds, stainless steels, especially the 13 wt% Cr-type martensitic stainless steels, are more often used as “corrosion resistant plastic mold steel” in plastic mold production. However, such 13 %Cr-type martensitic stainless steels are sensitive to localized corrosion, especially pitting corrosion, in certain Cl− containing environments [2, 3]. In these cases, chloride acts as a depolarizer, breaking the passive layer, and the metal cations exhibit considerable solubility in chloride solutions [4], resulting in the occurrence of pitting corrosion due to the passive film breakdown [5, 6]. Subsequently, the working life of the molds would be reduced and appearance of the plastic products can be damaged. Hence, it is important to enhance the anti-corrosion properties of the mold steels.

Almost all martensitic stainless steel must undergo quenching and tempering process before use to achieve better mirror surface property and corrosion, abrasion, and wear resistance [7, 8]. Generally, the austenitizing temperature mainly affects the microstructure of martensitic steel by strongly influencing the distribution of insoluble carbides and retained austenite in the steel’s matrix. Pits always initiate at some chemical or physical heterogeneity in the metal surface, such as inclusions, second phase particles, and solute-segregated grain boundaries [2, 9]. Once a pit initiates, it is always accompanied by the breakdown and repassivation of passive films on metal surface. A metastable pit can become stable pitting [10], which continuously induces the depassivation of the stainless steel, leading to severe corrosion. In addition, the localized enrichment of Cr caused by Cr-rich carbides would create a Cr-depleted zone in the steel matrix [11, 12] that degrades its corrosion resistance. Therefore, the insolvable carbides in stainless steels’ matrix under different austenitizing temperatures would affect the corrosion performance of stainless steel, either in terms of the initial pitting sites or localized Cr enrichment. However, the effects of microstructures containing carbides on the pitting resistance of the corrosion-resistant plastic mold steels have not been sufficiently studied [13, 14], and understanding them would facilitate the development of durable, high-performance plastic mold steels.

In this study, the microstructures and the electrochemical properties of a typical corrosion-resistant plastic mold steel that annealed and quenched after austenitizing at different temperatures are examined. The results indicate that the microstructures and the electrochemical properties of the MSS are quite different, and the related mechanisms are discussed.

2 Experimental

2.1 Material and heat treatment

The corrosion-resistant plastic mold steel used in present work was 13 %Cr-type martensitic stainless steel (MSS), which has the basic chemical composition (in wt%) of 0.38C, 0.60Mn, 0.90Si, 13.60Cr, 0.30V, and balance Fe. Before austenitizing, the MSSs were annealed at 850 °C for 12 h and cooled to 700 °C in a furnace for 4 h, followed by air cooling. After annealing, they were, respectively, austenitized at 980, 1030, 1080, and 1130 °C for 45 min, followed by oil quenching (OQ). The samples prepared using the aforementioned heat treatments were marked based on their heat treatment procedure, respectively. For instance, the sample of annealed steel was denoted as an “annealed sample,” while one austenitized at 980 °C was denoted as a 980 sample, hereafter.

2.2 Microstructure observation and carbides analysis

X-ray diffraction (XRD) analyses of the annealed and quenched samples were carried out on a Rigaku/SMART Lab X-ray diffractometer, with Cu Kα radiation. Continuous mode with a scan rate of 6°/min and 2θ ranging from 40° to 90° was applied to analyze the main structure of each specimen, while the step mode with a 0.01° step size, 2 s step time, and 47°–52° of 2θ range were used to gather the detailed information on the structure for each heat-treated specimen.

Microstructure of each heat-treated specimen was observed using field emission scanning electron microscopy (FE-SEM) of type LEO-1530. All of the specimens were ground, mechanically polished, and then etched using a particular etchant comprising 25 ml HNO3, 25 ml HCl, and 50 ml distilled water. The distribution of Cr and the chemical composition of carbides in each specimen were examined with an Oxford Instruments energy dispersive spectrometer (EDS) installed on the FE-SEM.

2.3 Electrochemical measurements

A 3-mm-thick plate was cut from a ϕ10 mm MSS rod, and then mounted in epoxy resin, leaving an exposed area around 0.7854 cm2. The samples were ground successively with 400-, 800-, and 1,200-grit SiC papers to obtain a fine surface. The 3.5 % NaCl aqueous solution was prepared by dissolving analytically pure NaCl in 18.25 MΩ cm deionised water as the corrosion medium (pH 5.6–5.8).

The electrochemical measurements were carried out using a CS350-type electrochemical workstation (Corrtest Instrument, China) with a three-electrode cell, which comprised a thin platinum plate as counter electrode (CE), a saturated calomel electrode (SCE) as reference electrode (RE), and the sample prepared using the above method as the working electrode (WE). Strong cathodic polarization at −500 mV (vs. SCE) was firstly conducted for 0.5 h to remove the initial passive film from the surfaces of the WEs. Before the EIS and polarization measurements were taken, the open circuit potentials (OCPs) of all WEs were monitored for 0.5 h, until a stable OCP was achieved.

After a stable OCP was acquired, the EIS measurements were performed with a ±10 mV (vs. OCP) amplitude perturbation and a frequency range from 10−2 to 105 Hz (i.e., 10, 7.94, 6.31, 5.01…1 Hz) and the results were fitted and analyzed by Zview.

The polarization potential was scanned at 0.166 mV s−1 (vs. SCE) from −500 mV (vs. OCP) in room temperature (23–26 °C) to eliminate the uncertainty of original passive film, and was terminated until the current density of the sample surface reaching 0.1 mA cm−2, where the corresponding potential is defined as the breakdown or pitting potential (E pit) of passive films [2]. The electrochemical results of each type sample were averaged based on at least three parallel specimens from the same heat treatment plot.

3 Results and discussion

3.1 Microstructure and precipitates

Surface hardness measurements of experimental steels reveal that the hardness value of annealed sample is 21.3 HRC, which is much lower than that of quenched samples (>56 HRC, see in supplementary materials). Figure 1 shows the XRD spectra of specimens that were annealed and quenched after different austenitizing temperatures. The peaks of the α phase with a BCC structure are presented in all of the samples. However, the axial ratio of martensite and ferrite is so close (10−4–10−5 nm) [15, 16] that the separate peaks of these two phases cannot be detected merely by XRD analysis. According to the surface hardness measurements, the hardness value recorded for the annealed sample is 21.3 HRC, and those of the quenched samples are all higher than 56 HRC. Thus, the α phase detected by XRD in experimental samples can be further indexed as ferrite in annealed sample, and martensite in quenched samples. In addition, with the increase in austenitizing temperature, the γ phase is present in the 1,080 and 1,130 samples (Fig. 1a, b), which can be considered as retained austenite. This may be attributed to the high stability of austenite promoted by the high C content that results from the dissolution of M23C6 carbides at high austenitizing temperatures [17, 18].

Fig. 1
figure 1

XRD profiles of the samples after different heat treatments: a diffraction angles from 40° to 90° and b detail information between diffraction angles from 47° to 52°

Full size image

The XRD spectra in Fig. 1b show the weak peaks within the 2θ range of 47°–52° for the annealed and quenched samples. The diffraction angles fitted by MDI Jade version 5.0 fitting software were 48.514(4)°, 50.881(4)°, and 51.659(5)° for annealed sample; 48.411(20)° and 50.799(30)° for 980 sample, and 48.497(20)°, 50.738(28)° for 1,030 sample—indexed as the different peaks of (440), (531), and (600) plane for Cr-rich M23C6, respectively.

Fitting results reveal that the lattice parameter of Cr-rich M23C6 (FCC) in annealed, 980 and 1,030 samples are 1.061(1), 1.063(1), and 1.064(1) nm, respectively. Based on the measurement precision of the present experiments, the obtained lattice parameters can be regarded as the same, however, they are slightly smaller than the value of 1.0658(1) nm of pure Cr23C6 noted in Erdos’s research [19], perhaps due to the solution of Fe into the pure Cr23C6 [20]. In addition, the peak intensity of Cr-rich M23C6 of the annealed sample was much greater than that of the quenched samples. As the austenitizing temperature increased from 980 to 1,030 °C, the peak intensity of Cr-rich M23C6 decreased; when austenitizing temperature exceeded 1,080 °C, the peaks of Cr-rich M23C6 disappeared and the peak of γ(200) was observable. It turned out that the Cr-rich M23C6 initially precipitated during annealing, then started to dissolve during austenitizing and with the tendency of thoroughly dissolved in steel matrix as the austenitizing temperature increase. Thus, the M23C6 carbides existed in 980 and 1,030 samples could be reasonably recognized as “undissolved M23C6 carbides”.

The microstructures of the annealed sample shown in Fig. 2a, b included ferrite + M23C6 without pearlite, which might be attributable to the reduction of carbon in steel matrix when carbon has dissolved into carbides, in accordance with other reports [21], while martensitic laths could be observed in the quenched samples (Fig. 2c, d, e, and f). There were more carbides in the annealed sample than in the quenched samples. The amount of carbides decreased as the austenitizing temperature increased, such that when it reached 1,130 °C, no carbides could be observed. As was confirmed by XRD analysis, the main carbides observed in annealed, 980 and 1,030 samples are Cr-rich M23C6-type, which exhibit as some round or elongated particles with a size range of 1–2 μm and smooth edges (Fig. 2a, b, c, and d). They were typically noted between the grain boundary and within the martensitic lath, indicating that they precipitated during annealing, and tended to dissolve more in the steel matrix as austenitizing temperature increased. When the austenitizing temperature reached a critical value (1,130 °C for instance), the Cr-rich M23C6 would be totally dissolved in the austenite phase of the steel, and did not precipitate during quenching.

Fig. 2
figure 2

SEM morphology of a annealed sample, b local magnification of carbides in (a), and quenched samples after austenitization at c 980 °C, d 1,030 °C, e 1,080 °C, and f 1,130 °C

Full size image

3.2 Carbides characterization

Following the literature [13, 22, 23], the preferable austenitizing temperature for 13 wt%Cr-type stainless steel is in the range of 1,020–1,050 °C, and the undissolved M23C6 has been proven to gain the same structure parameter after 980 and 1,030 °C austenitization. Therefore, the 1,030 sample was selected here for further study of Cr-rich M23C6. The annealed and 1,130 samples were also selected for comparison. The distribution maps for the elements Cr and Fe using energy dispersive spectrometer (EDS) in the annealed, 1,030 and 1,130 samples are shown in Fig. 3. Clearly, the distributions of Cr and Fe in the annealed and 1,030 sample are non-uniform, whereas in the 1,130 sample they were uniform. Compared with the microstructure morphologies and distribution maps, the Cr-rich regimes in the annealed and 1,030 samples were M23C6 combined with a relative depleted zone of Fe element.

Fig. 3
figure 3

Element distribution maps (mainly of Cr and Fe) via EDS of the micro zone of experimental MSSs: a annealed sample, b 1,030 sample, and c 1,130 sample

Full size image

The composition analysis of the selected micro regions (clearly marked by A, B, C, D, E, F, and G in Fig. 3) was conducted via SEM–EDS. In Fig. 3, points A and B (D and E) represent the M23C6 carbides in annealed (1,030) sample, and points C, F, and G can be recognized as the matrix of all three samples, respectively. The results are shown in Table 1. However, carbon cannot be detected due to its low content in steel and the limited detection accuracy of the spectrometer when applied to such a light element. The Cr content (in weight %) of M23C6 in both the annealed and 1,030 samples was around 48 %, which was higher than that in the matrix (around 13 %). In addition, the Cr content in the matrix of 1,130 samples was similar to that of other samples. However, the Cr distributions seemed to be irrelevant to the precipitating location, as the Cr content of M23C6 precipitated among the grain boundary and martensitic laths are almost the same. Furthermore, the silicon (Si) content in M23C6 is significantly lower than that in matrix, confirming that Si is not a carbide-forming element.

Table 1 Results of composition analysis via EDS (corresponding to the points in Fig. 3)
Full size table

As reported, the M23C6 is always the first phase formed in austenitic stainless steel and some other low-carbon high alloying steels during annealing, usually precipitating at the grain boundary (GBs) due to high interfacial energy [24, 25]. With higher Cr content and relatively lower Fe, the precipitation of M23C6 creates a lack of Cr among the steel matrix, becoming preferential sites for pitting corrosion [26]. However, as was shown in Fig. 3a, the bright contrast of Cr in M23C6 particles indicates a localized enrichment of Cr. The analysis of Cr content distribution through line scanning (see in supplementary materials) across the carbide particles further confirmed that the Cr content gradually increased from the matrix to the carbides, indicating Cr enrichment in the carbide area. However, no Cr-depletion area was found around M23C6, confirming that Cr-rich M23C6 carbides dissolved before quenching.

3.3 Electrochemical performance

Figure 4 illustrates the variation of the open circuit potential (OCP) of experimental stainless that were annealed and quenched in different austenitizing temperatures in the 3.5 wt% NaCl aqueous solution. In all cases, the shapes of the curves are similar, during the first 600 s of immersion, OCP increased dramatically with time and then reached a stationary state. The values of OCP increased with the austenitizing temperature, and some potential transients were observed. Figure 5 shows the potentiodynamic polarization curves of experimental martensitic stainless steels after different heat treatments. The distribution of experimental steels’ pitting potentials, as obtained from the potentiodynamic polarization curves, is shown in Fig. 6. According to Fig. 5, all of the samples exhibited a clear passive region and a sharp increase in current density after the breakdown of passive film in 3.5 wt% NaCl aqueous solution. Several current transients were observed in passive region, especially from the polarization curves of the annealed, 980 and 1,030 samples, which may be related to the nucleation and repassivation of metastable pits of stainless steels that occurred at such active sites as carbides [10, 27]. Figure 6 reveals that the E pit of the annealed sample (−30 mV) was much lower than that of the quenched samples, and E pit increased with the austenitizing temperature; specifically, when austenitizing temperature reached 1,130 °C, the E pit of experimental steel increased to 400 mV, which was much higher than that of other samples austenitized at lower temperature (128–178 mV), this indicates that the pitting resistance of the quenched samples was better than annealed sample, and an increase in austenitizing temperature can enhance the resistance of quenched martensitic stainless steels to pitting corrosion.

Fig. 4
figure 4

The open circuit potential (OCP) of experimental steels in 3.5 wt% NaCl aqueous solution after 0.5 h-long cathodic polarization at −500 mV (vs. SCE) around 25 °C

Full size image
Fig. 5
figure 5

Potentialdynamic polarization curves of experimental stainless steels in 3.5 wt% NaCl at room temperature (around 25 °C)

Full size image
Fig. 6
figure 6

Pitting potential of experimental stainless steels based on potentiodynamic polarization curves

Full size image

Figure 7 shows the Nyquist plots of the experimental steels at OCP under the same experimental condition as the potentiodynamic polarization. It can be seen that the EIS spectra of all the samples exhibited a single capacitive in the high-medium frequency range, which is associated with the double layer capacitance (C dl) and passive layer capacitance (C film) in parallel with the charge transfer resistance. However, since the C film of MSS is much lower than C dl, the total capacitance here is mainly depended on the passive film other than the double electric layer [28–30]. As is shown in Fig. 7, the diameter of the arc of the annealed sample was smaller than that in the quenched samples, and the diameter increased with the austenitizing temperature, indicating that the surface of the quenched samples that austenitized at higher temperatures could form more protective and well-structured passive films due to the improved homogeneity of local areas from reducing the amount of Cr-rich carbides, increasing the corrosion resistance of the martensitic stainless steel in Cl− containing solution.

Fig. 7
figure 7

EIS spectra of experimental steels in 3.5 wt% NaCl (aq) at OCP

Full size image

El-Egamy and Badaway [31] proposed an equivalent circuit model to simulate the behavior of 304 stainless steel in a sodium sulphate solution. This model can be used to fit the electrochemical performance of the present experimental samples. The equivalent circuit is shown in Fig. 8, in which the parallel-connected elements (CPE//R ct) are used to describe the electrochemical process at the substrate/passive film interface.

Fig. 8
figure 8

Equivalent circuits used to fit the experimental impedance data

Full size image

The impedance of CPE can be described by:

$$Z({\text{CPE}}) = Q_{0}^{ - 1} (j\omega )^{ - \alpha },$$
(1)

where Q 0 is the CPE parameters, j is \(\sqrt { - 1}\), ω is angular frequency and α is the dispersion coefficient related to surface inhomogeneity. The impedance resistance of the electrode/passive film/solution system can be calculated by:

$$Z = Z(R_{\text{e}} ) + Z(R_{\text{p}} //{\text{CPE}}) = R_{\text{e}} + \left[ {R_{\text{ct}}^{ - 1} + Q_{0} (j\omega )^{\alpha } } \right]^{ - 1},$$
(2)

where CPE is the constant-phase element, and reflects a distribution of reactivity of passive films, R ct the charge transfer resistance, and R e the Ohmic or solution resistance [32, 33]. Table 2 gives the best fitting results of EIS based on the equivalent circuit model shown in Fig. 8.

Table 2 Fitted parameter values of experimental steels obtained from EIS spectra
Full size table

From Table 2, the α values of quenched stainless steels are in the range of 0.91–0.92, which are higher than for the annealed sample (0.89), it implies that the electrochemical behavior of passive films is mildly deviated from a pure capacitive behavior. When α < 1, the system shows behavior that has been attributed to surface heterogeneity, and the CPE parameter Q cannot represent the capacitance. Some researchers have explored the relationship between CPE parameters and the interfacial capacitance, and their results have shown that the effective capacitance (C eff) is proportional to Q −α. Additionally, Hirschorn [32] proposed that the thickness of oxide films formed on an Nb disk electrode in 0.1 M NH4F solution increases as Q decreases. In our work, when the austenitizing temperature rose from 980 to 1,130 °C, the Q 0 values of the quenched samples decreased from 4.33 × 10−5 to 3.13 × 10−5 Ω−1 cm−2 sα, and the R ct values increased from 1.72 × 105 to 3.13 × 105 Ω cm2, indicating that the increase in corrosion resistance of passive films due to either film thickening or compositional modification was obtained by improving local homogeneity through increasing austenitizing temperature, which is well matched with the results of potentiodynamic polarization experiments.

According to the above analysis, although no poor Cr zone was observed around undissolved Cr-rich M23C6 particles, the enrichment of Cr around undissolved M23C6 particles was clearly detected. It can be inferred that the local areas with different Cr contents would possess different electrode potentials, and the enrichment of Cr in local areas around undissolved Cr-rich carbides would also induce depletion of Cr content in other areas. In addition, it is experimentally proved that the corrosion resistance of austenite is higher than ferrite/martensite, which is resulted from the fact that the austenite contains lower internal stresses and less defects [34, 35]. It is also reported that in duplex stainless ferrite is more chemically active than austenite and the metastable pits always preferentially initiate inside ferrite [36, 37]. Therefore, the retained austenite appeared in the steel matrix when the austenitizing temperature reached 1,080 °C could also enhance the pitting resistance of the MSSs quenched at higher austenitizing temperatures. As a result, reducing the amount of these Cr-rich carbides, which can be promoted by increasing the austenitizing temperature, should improve the corrosion resistance of MSSs.

4 Conclusion

The Cr-rich M23C6 carbides observed in the MSSs matrix precipitated during annealing, and the increase in austenitizing temperature dramatically decreased the amount of unsolved Cr-rich M23C6 within the matrix of the MSS, especially when austenitizing temperature reached 1,130 °C, the M23C6 totally dissolved in the steel and did not precipitate again during quenching. It shows that in the sample quenched at 1,030 °C, there was a localized Cr enrichment area around the undissolved Cr-rich M23C6 carbides, whereas in the sample quenched at 1,130 °C, no Cr enrichment area could be found due to the disappearance of the Cr-rich M23C6 carbides. The present results indicate that the homogeneity of Cr distribution has been improved with the increase in austenitizing temperature.

Increasing the austenitizing temperature could enhance the corrosion resistance of the martensitic stainless steel to both pitting and general corrosion. The improvement in corrosion resistance of the steel was due to greater homogeneity in the distribution of Cr and a reduced amount of Cr-rich carbides. The polarization curve and EIS measurements confirmed that the decrease in M23C6 carbides content among the MSS matrix by increasing the austenitizing temperature allowed more protective and well-structural passive films to form, which enhanced the pitting corrosion resistance of martensitic stainless steel.