Effect of Environmental Factors on Hydrogen Absorption into Steel Sheet under a Wet-dry Cyclic Corrosion Condition

Abstract

The effects of temperature and chloride deposition on hydrogen absorption into steel were evaluated during wet/dry cyclic corrosion using a temperature-compensated hydrogen absorption monitoring system based on the electrochemical hydrogen permeation method. Hydrogen absorption into steel was detected through the measurement of hydrogen permeation currents during the wet periods under the wet/dry cyclic corrosion. The enhancement of hydrogen absorption was mainly caused by the changes in the solution chemistry during the wetting and drying periods, with a decrease in pH due to the hydrolysis reaction of Fe³⁺ at high Cl⁻ concentration. Hydrogen absorption into steel increased with increasing temperature and chloride deposition. The reasons for the increment of hydrogen absorption are considered that enhancement of the hydrogen evolution reaction with temperature and that the corrosion potential shifted to less noble by increase in the electrolyte thickness with increasing chloride deposition. Based on these results, the amount of absorbed hydrogen map effected by these factors under atmospheric corrosion environment was described.

1. Introduction

High-strength steel is increasingly being applied in the automotive field to reduce body weight and further improve collision safety.¹⁾ However, the application of high-strength steel is limited in certain environments by hydrogen embrittlement.²⁾ Generally, susceptibility to hydrogen embrittlement is related to not only mechanical properties such as the strength of the steel,^3,4) and applied stress,^5,6) but also environmental factors such as temperature and chloride deposition, all of which can affect amount of hydrogen that can be absorbed into steel.^7,8) Since hydrogen absorption into steel is mainly caused by wet/dry cyclic corrosion, it is necessary to evaluate the hydrogen absorption behavior of steel in a corrosive atmospheric environment.^9,10)

Generally, corrosion rate of steel in corrosive atmospheric corrosion environments can be predicted from Eq. (1),¹¹⁾ which was standardized in ISO 9223: 2012 as:   

$$\begin{array}{l}{\mathrm{r}}_{\text{corr}}=1.77{\mathrm{P}}_{\text{d}}^{0.52}exp\left(0.020\mathrm{R}\mathrm{H}+{\mathrm{f}}_{\text{st}}\right)\\ \hspace{1em}\hspace{1em}+0.102{\mathrm{S}}_{\text{d}}^{0.62}exp\left(0.033\text{RH}+0.040\mathrm{T}\right)\end{array}$$(1)

  

$$\begin{array}{l}{\mathrm{f}}_{\text{st}}=0.150\left(\mathrm{T}-10\right)\mathrm{\hspace{0.17em} } \text{when}\mathrm{\hspace{0.17em} } T\le 10,\\ {\mathrm{f}}_{\text{st}}=-0.054\left(\mathrm{T}-10\right)\mathrm{\hspace{0.17em} } \text{when}\mathrm{\hspace{0.17em} } \mathrm{T}>10\end{array}$$

where r_corr is the average corrosion rate (μm/year) of iron for 1 year of exposure, P_d is the SO₂ concentration (mg/(m² day)), RH is the relative humidity, f_st is the temperature term, S_d is the airborne salt content (mg/(m² day)), and T is the temperature (°C).

As shown in Eq. (1), r_corr can be divided into two terms: one relates to SO₂ concentration and the other to airborne salt content. Given that atmospheric SO₂ concentration has decreased in recent years, the corrosion rate of steel in an atmospheric environment is mainly associated with the amount of airborne salt. Therefore, the key environmental factors are temperature, relative humidity, and the amount of airborne salt.

To evaluate the corrosion rate of steel in an automobile body, it is necessary to consider these environmental factors where automobiles are in motion, because environmental factors such as temperature, relative humidity, and salt deposition change depending on the moving area and movement conditions of the automobile, e.g., the amount of salt deposition originating from snow-melting salt during movement. Therefore, to determine the hydrogen absorption behavior of steel sheet in moving automobiles, it is important to evaluate and clarify the corrosion conditions, quantity of absorbed hydrogen, and hydrogen absorption mechanism of steel in consideration of environmental factors such as temperature, humidity, and the amount of salt deposition under these conditions.

In a previous study,¹²⁾ we developed a temperature-compensated hydrogen absorption monitoring system based on the electrochemical hydrogen permeation method, which continuously evaluates hydrogen absorption behavior in the moving automobile environment, and investigate hydrogen absorption into steel in moving and parked automobiles on the basis of in situ measurement results. As a result of the corrosion reaction, the amount of absorbed hydrogen increased due to wetting of the steel surface by road water^13,14) and pH changes. The amount of absorbed hydrogen changed with temperature and relative humidity even at the daily scale and during parking, whereas snow-melting salt adhered to the steel during automobile movement.¹⁴⁾

Thus, it is necessary to consider the conditions under which the steel surface is corroded as a result of wetting by road water during movement,^15,16) as well as the atmospheric conditions, such as temperature and relative humidity, that cause changes in the thickness of the water film and the solution composition while the vehicle is parked. Although the characteristics of hydrogen absorption behavior in the automobile driving environment have largely been clarified, the amount of salt (chloride) deposition has not been quantified, and the effects of temperature and relative humidity on the amount of absorbed hydrogen have not been examined sufficiently.

The effects of temperature and salt deposition on hydrogen absorption under different relative humidity conditions have been previously reported by Omura et al.¹⁷⁾ However, to analyze hydrogen absorption behavior in a moving automobile, it is necessary to quantitatively evaluate the effects of corrosion factors on hydrogen absorption behavior under wet/dry cyclic corrosion, and in moving and parked conditions. It is important to evaluate the hydrogen absorption behavior of high-strength steel that is used as the inner sheets of the automobile, where salt is easily deposited in environments with high salt deposition.

This study evaluated how environmental factors influence hydrogen absorption based on continuous monitoring of hydrogen absorption¹²⁾ as temperature and salt deposition changed under cyclic corrosion conditions in the laboratory.

2. Experimental Procedure

2.1. Sample

The material used in this study was 270-MPa grade cold-rolled steel sheet (thickness: 0.75 mm); Table 1 lists the chemical composition. For hydrogen absorption measurements, the sample was prepared as follows. First, the steel sheet was sheared into small samples (65 mm × 65 mm²), and one side of each sample surface was mechanically polished with successive abrasive SiC papers up to 2000 grit. The small sample was then immersed in a mixture of 6-mL hydrofluoric acid and 94-mL hydrogen peroxide for chemical etching. The thickness of the sheet was finally reduced to approximately 0.7 mm by 50 μm after chemical etching. Then, Pd plating with a thickness of ca. 400 nm was conducted electrochemically on one surface at a constant current density of 3 mA/cm² and 40°C for 480 s in a commercial Pd plating solution (K-Pure Pd, Kojima Chemical Co., Ltd.).

Table 1. Chemical composition of the steel sheet used in this study (mass%).

C	Si	Mn	P	S	Fe
0.02	0.01	0.15	0.01	≦0.01	bal.

First, we determined the hydrogen diffusion coefficient of the steel using an electrochemical hydrogen permeation method.^18,19,20) Samples were prepared following the procedure described above. The hydrogen diffusion coefficient at ambient temperature, D_T, was evaluated using breakthrough-time method,²¹⁾ which is calculated as follows:   

$${\mathrm{D}}_{\text{T}}=\frac{{\mathrm{L}}^2}{15.3{\mathrm{t}}_{\text{b}}}$$(2)

where t_b is the average breakthrough time obtained from a hydrogen permeation current in the potential range from –1 V (SCE: calomel electrode) to –1.2 V, and L (cm) is the sample thickness. The relationship between D_T (cm²/s) and temperature T (K) of the sample can be expressed as follows:   

$${\mathrm{D}}_{\text{T}}=1.52\times {10}^{-6}exp\left(\frac{-20}{\mathrm{R}\mathrm{T}}\right)$$(3)

where R is the gas constant (8.314 J/(mol K)).

2.2. Wet/dry Cyclic Corrosion Test

Figure 1 presents a schematic of the wet/dry corrosion cycle used in this study. This test cycle consisted of two processes: application of an aqueous NaCl (salt) solution of predetermined concentration onto the sample surface and drying of the salt solution by changing the relative humidity at a constant temperature.

Fig. 1.

Schematic drawing of the wet-dry cyclic corrosion conditions used in this study. (Online version in color.)

During salt application, aqueous NaCl solution (0.8, 8, or 27 wt%) was sprayed onto the surface of steel sheet in the hydrogen absorption monitoring system, for a final deposition weight of 37 g/m². Thus, we controlled the amount of salt deposition onto the sample surface to 0.3, 3, or 10 g/m² NaCl. The amount of salt deposition in each wet/dry corrosion cycle was kept constant by washing the steel sheet surface with pure water prior to salt application. The salt solution was applied twice per week, just after the beginning of the dry period of the test cycle.

In the wet/dry cyclic corrosion test, relative humidity was 30% and 90% in the dry and wet periods, respectively. The duration of each period was 2 h, and the transition period between dry and wet periods was also 2 h. Thus, the total duration of one wet/dry cycle was 8 h. During the wet/dry corrosion cycles, a constant temperature of 10, 30, or 50°C was maintained. The test period was 28 days.

2.3. Determination of Absorbed Hydrogen

The amount of absorbed hydrogen into a sample was determined from the hydrogen permeation current using a hydrogen absorption monitoring system¹²⁾ during wet/dry corrosion cycles (Fig. 2).

Fig. 2.

Schematic drawing of the hydrogen absorption monitoring system. (Online version in color.)

The hydrogen absorption monitoring system (Fig. 2) consists of four three-electrode electrochemical cells (surface area exposed to electrolyte solution: 1.1 cm²). The prepared sample was fixed at the top of the cells, such that the Pd-coated surface faced the hydrogen-detection-side cell, where 0.1 kmol/m³ NaOH was introduced. Initially, NaOH was deaerated by nitrogen bubbling for > 24 h. After the NaOH solution was introduced, the sample surface of the hydrogen-detection-side cell was left at +0.20 V vs. Ir/IrOx until residual current density reached < 40 nA/cm² at room temperature.

To compensate for the effects of temperature fluctuation on the permeation current, the residual current of a part of the steel sheet where no corrosion occurred was measured on one cell (i.e., reference cell, Fig. 2); the steel surface on the reference cell was completely covered with silicon sealant to prevent corrosion, and the hydrogen permeation currents, i_p, of the steel on other cells were obtained by subtracting the residual current from that measured in steel exposed to the corrosive environment.¹³⁾

2.4. Corrosion Mass Loss Evaluation

The amount of steel weight lost during the wet/dry cyclic corrosion test was measured. Weight loss was obtained after removing the corrosion products according to the procedure described in ISO 8407: 1991 C. 3.5.²²⁾

3. Experimental Results

3.1. Hydrogen Absorption Behavior in a Wet/dry Corrosion Environment

Changes in i_p measured in steel at various temperatures at a salt deposition of 10 g/m² during the wet/dry cyclic corrosion tests revealed that i_p repeatedly increased and decreased in every wet/dry cycle (Fig. 3). The changes in i_p measured at each temperature generally peaked within 1 week after the start of the test and then decreased gradually. This behavior was similar at all temperatures, but the magnitude of i_p increased with temperature.

Fig. 3.

Time variations of hydrogen permeation current of steel measured at different temperatures: (a) 10°C, (b) 30°C, and (c) 50°C. (Online version in color.)

Figure 4 presents the time-variations of i_p for two wet/dry cycles measured at each temperature at a salt deposition of 10 g/m². The i_p transient results revealed that i_p changed synchronously with relative humidity, with peaks during both wetting and drying periods at approximately 60%RH. In each cycle, the i_p peaks were higher in the drying period than in the wetting period; this trend was more pronounced as temperature decreased. During the wet period, i_p remained nearly constant, and increased with increasing the temperature.

Fig. 4.

Time variations of hydrogen permeation current measured of steel for 2 wet-dry corrosion cycles at different temperatures: (a) 10°C, (b) 30°C, (c) 50°C. Figure (d) changes in temperature and relative humidity.

Figure 5 presents time-variations of i_p measured at 10°C and various salt depositions of 0.3, 3, and 10 g/m²; i_p peaked in the wetting and drying periods at approximately 60%RH, and increased with increasing salt deposition.

Fig. 5.

Time variations of hydrogen permeation current of steel measured at 10°C for 2 wet-dry corrosion cycles at different chloride deposition conditions: (a) hydrogen permeation current, and (b) changes in temperature and relative humidity.

These results suggest that changes of i_p are related to steel corrosion during the wetting, wet, and drying periods; thus, hydrogen absorption into steel is increased by the enhancement of steel corrosion.

3.2. Amount of Hydrogen Absorption

As shown in Figs. 3, 4, and 5, changes of i_p measured during the wet/dry cycles corresponded to changes in hydrogen absorption into steel. The surface concentration of the absorbed hydrogen, C_ab was calculated according to Fick’s first law as follows:   

$${\mathrm{C}}_{\text{ab}}=\frac{{\mathrm{i}}_{\mathrm{p}}\mathrm{L}}{{\mathrm{D}}_{\text{T}}\mathrm{F}\mathrm{d}}$$(4)

where i_p is the hydrogen permeation current calibrated by temperature compensation (μA/cm²), L is the sheet thickness (cm), F is the Faraday constant (96485 C/mol), and d is the steel density (7.85 g/cm³). In this study, we considered the temperature dependence of D_T.

The amount of absorbed hydrogen into the steel is critical, because crack initiation and propagation due to hydrogen embrittlement are more likely to occur as hydrogen content increases.⁵⁾ Thus, although i_p changes with wet/dry corrosion cycles, hydrogen embrittlement is induced once the absorbed hydrogen concentration exceeds a critical value for crack initiation and propagation.²³⁾ In this study, C_ab, which was obtained from the maximum value of i_p in our tests, was treated as the maximum surface hydrogen concentration of the steel (C_ab-max), and this depends on environmental factors such as temperature and salt deposition.

Figure 6 presents the relationship between temperature and C_ab-max described by Eq. (4). C_ab-max increased with increasing temperatures and salt deposition in the corrosion test cycle.

Fig. 6.

Relationship between the surface concentration of absorbed hydrogen estimated from the peak value of i_p and temperature.

3.3 Evaluation of Corrosion Mass Loss

Figure 7 presents time variations of corrosion mass loss of steel at various temperatures at a constant salt deposition of 3 g/m². The corrosion mass losses of steel increased linearly over time, such that the steel corrosion rates were constant. Similar trends were observed at different salt deposition. Figure 8 presents the relationship between the corrosion rate and temperature at various salt depositions; the corrosion rate increased as both temperature and salt deposition increased. Temperature appeared to have a stronger effect than salt deposition on the steel corrosion rate during wet/dry cycles.

Fig. 7.

Changes of corrosion mass loss of steel during the wet-dry cyclic corrosion test at the chloride deposition of 3 g/m².

Fig. 8.

Dependency of the corrosion rate of steel at different chloride depositions on temperature.

4. Discussion

4.1. Effect of Relative Humidity on C_ab

In atmospheric environments, the steel corrosion reaction occurs under a thin electrolyte solution film, and hydrogen absorption into steel occurs as shown in Fig. 9. The anodic reaction of iron dissolution is expressed by Eq. (5). The main cathodic reaction of atmospheric steel corrosion is the oxygen reduction reaction (ORR), expressed by Eq. (6). Simultaneous cathodic reduction of H⁺ can also occur to form H_ad on the corroding steel, as expressed by Eq. (7), and some of the adsorbed hydrogen atoms, H_ad, can be absorbed into the steel, as expressed by Eq. (8). Therefore, the absorbed hydrogen concentration, C_ab, depends on the kinetics of the hydrogen evolution reaction (Eq. (7)) and the hydrogen absorption reaction. The hydrogen evolution reaction is more important, because its kinetics are associated with the corrosion potential of the steel and pH, where the hydrogen absorption kinetics is also accelerated.   

$$\text{Fe}\to {\text{Fe}}^{2+}+2{\text{e}}^{-}$$(5)

  

$${\text{O}}_2+2{\text{H}}_2\text{O}+4{\text{e}}^{-}\to 4{\text{OH}}^{-}$$(6)

  

$${\text{H}}^{+}+{\text{e}}^{-}\to {\text{H}}_{\text{ad}}$$(7)

  

$${\text{H}}_{\mathrm{a}\mathrm{d}}\to {\text{H}}_{\mathrm{a}\mathrm{b}}$$(8)

Fig. 9.

Schematic drawing of the reaction scheme of hydrogen absorption into steel during aqueous corrosion.

As shown in Figs. 4 and 5, i_p changes with relative humidity; i_p peaks were observed at approximately 60%RH in both the wetting and drying periods. During the wet period, i_p remained relatively high. These results indicate that the hydrogen evolution reaction occurs at a relative humidity of > 60%RH, upon the formation of a liquid film. Due to moisture absorption caused by capillary action in iron rust pores²⁴⁾ and the formation of FeCl₂, the liquid film is formed at lower deliquescent humidity than that of NaCl, which is 75%RH.²⁵⁾ Harada et al.²⁶⁾ examined the hydrogen permeation of steel at various relative humidity settings, and observed the maximum hydrogen permeation current at relative humidity from 60 to 70%, which is consistent with our results.

During the wetting and drying periods, the corrosion reaction of steel occurred under a very thin electrolyte solution film, where anodic dissolution of iron was enhanced because ORR is facile due to the thinness of the oxygen diffusion layer. Therefore, dissolved Fe²⁺ accumulated in the solution film. Dissolved Fe²⁺ is then oxidized to Fe³⁺ according to Eq. (9), and most of the Fe³⁺ is deposited as Fe(OH)₃ due to its very low solubility. Simultaneously, H⁺ is formed by the hydrolysis reaction shown in Eq. (10), and pH decreases, which can accelerate the hydrogen evolution rate. This H⁺ becomes the adsorbed hydrogen atom, H_ad, via the cathodic reduction reaction, as expressed by Eq. (8), and is partially absorbed into the steel sheet , as shown in Eq. (9).   

$$2{\text{Fe}}^{2+}+\frac{1}{2}{\text{O}}_2+{\text{H}}_2\text{O}\to 2{\text{Fe}}^{3+}+2{\text{OH}}^{-}$$(9)

  

$${\text{Fe}}^{3+}+{\text{H}}_2\text{O}\to \text{Fe}{\left(\text{OH}\right)}_3+3{\text{H}}^{+}$$(10)

The value of i_p is synonymous with the surface hydrogen concentration, C_ab, which is absorbed into the steel during corrosion; its variation depends on the thickness and composition of the water film on the steel sheet surface. Figure 10 presents a schematic diagram of changes in thickness of the electrolyte film and the related solution chemistry during wet/dry cycles. As the relative humidity changed, the thickness of the electrolyte film changed depending on the salt deposition.^27,28) Thus, an electrolyte film with high Cl⁻ concentration formed on the surface at the beginning of the wetting period or immediately prior to dry-up of the electrolyte film at approximately 60%RH, just as the pH of the electrolyte film^29,30) decreased due to the hydrolysis reaction of Fe³⁺. Therefore, C_ab increased at approximately 60%RH during the wetting and drying periods.

Fig. 10.

Schematic drawing of the changes of electrolyte composition and thickness during a wet-dry cyclic corrosion test.

As shown in Fig. 5, the i_p peak was higher in the drying period than in the wetting period. This may be associated with the difference in solution chemistry of the electrolyte film formed in the drying and wetting periods. In the drying period, the corrosion reaction of steel continued after the wet period, resulting in the formation of Fe³⁺. During the drying stage, the electrolyte film thickness continued to decrease. In contrast, during the wetting period, the corrosion reaction of steel started on a completely dry surface; thus, pH decreased during the drying stage.

From the results of the corrosion rate shown in Fig. 8, the dissolution rate of Fe was estimated at 3.6×10⁻⁴ mol/(m²/min) at 50°C and a salt deposition of 3 g/m², assuming a constant corrosion rate throughout the period. Since the solubility product of Fe(OH)₃ is approximately 10⁻³⁸,³¹⁾ the Fe³⁺ concentration was estimated at 6×10⁻⁵ mol/L at pH 3. Thus, Fe³⁺ concentration approached saturation from the initial wetting stage during the wetting period, when the water film thickness was about 10 μm. The peak current of the hydrogen permeation current was largely the same, because the composition of the water film was the same during both the wetting and drying periods.   

$${\text{Fe}}^{3+}+3{\text{OH}}^{-}\to \text{Fe}{\left(\text{OH}\right)}_3\hspace{1em}\text{p}{\mathrm{K}}_{\text{sFe}{\left(\text{OH}\right)}_3}=38$$(11)

The effects of relative humidity and salt deposition on C_ab in the wet/dry cyclic corrosion environment influenced C_ab, because the hydrogen formation reaction rate changed as the pH of the water film was affected by the changing water film thickness on the steel sheet surface. During the wet period, Cl⁻ concentrations decreased as the water film became thicker than those during the wetting and drying periods; C_ab values were smaller than those in both the wetting and drying periods, although C_ab values were relatively high.

4.2. Effects of Temperature and Salt Deposition on C_ab

Regarding the effect of temperature, i_p peaked at approximately 60%RH during the wetting and drying periods at all temperatures and the peak value increased as temperature increased, as shown in Fig. 4. It can be said that the increase of i_p with temperature was caused by the acceleration of hydrogen evolution reaction together with the acceleration of iron dissolution and oxygen reduction reaction.

Regarding the effect of salt deposition, the changes in i_p measured at different salt depositions showed the same behavior for wet/dry corrosion cycles, and its value in the drying period, i_p increased with increasing the salt deposition, as shown in Fig. 5. The amount of salt deposition is related to the electrolyte film thickness formed in the wet condition, and not to the electrolyte composition. Generally, as the electrolyte film thickness becomes larger than several 10 μm, the corrosion rate of steel decreases.³²⁾ This is because oxygen reduction reaction rate decreases with increasing the electrolyte film thickness. Additionally, the corrosion potential of steel shifts in the less noble direction with increasing the electrolyte film thickness. Therefore, as the salt deposition increases, the hydrogen evolution reaction tends to be accelerated due to the shift of the corrosion potential in the less noble direction. This can cause the increase of i_p with increasing the salt deposition.

As discussed above, it was considered that the corrosion environment factors such as relative humidity, salt deposition, and temperature affected each other to change i_p or C_ab.

4.3. Relationship between C_ab-max and Environmental Factors

Generally, it is said that hydrogen embrittlement of steel in a corrosive environment occurs when the absorbed hydrogen concentration exceeds a critical value.²³⁾ In this study, from the maximum value of i_p measured during the corrosion test period, the surface hydrogen concentration, C_ab-max was determined as the maximum concentration of the absorbed hydrogen in the corrosive environment. The C_ab-max should be a benchmark, since hydrogen embrittlement of steel is more likely to be induced once the C_ab of steel exceeds a critical value.²³⁾ Figure 11 presents a contour map of C_ab-max for steel in corrosion environments. As shown in the figure, the C_ab-max increases with increasing temperatures and salt deposition, suggesting that the potential occurrence of hydrogen embrittlement in steel can be predicted in such environments.

Fig. 11.

Amount of absorbed hydrogen map effected by temperature and chloride load in wet-dry cyclic corrosion environment.

5. Conclusions

This study investigated the effects of environmental factors (temperature and salt deposition) on C_ab of steel in wet/dry cycles using a temperature-compensated hydrogen absorption monitoring system. The results are as follows.

(1) The C_ab of steel during the wet/dry cycles changed synchronously with relative humidity and peaked around the transition period between dry and wet.

(2) C_ab increased with temperature, due to enhancement of the hydrogen evolution reaction with temperature. C_ab increased with salt deposition, due to a change in the solution chemistry of the electrolyte film during the wet/dry cycles.

(3) A hydrogen absorption map of C_ab-max for a range of temperatures and salt depositions was proposed. This map can be applied to estimate the potential occurrence of hydrogen embrittlement of steel in atmospheric environments.

References

• 1)   R. P.  Krupitzer,  D. M.  Urban and  D. A.  Witmer: Galvatech ’04, Conf. Proc., AIST, Warrendale, PA, (2004), 31.
• 2)   S.  Matsuyama: Okurehakai (Delayed Fracture), Nikkan-Kogyo Shimbunsya, Tokyo, (1989), 26.
• 3)   T.  Kushida,  H.  Matsumoto,  N.  Kuratomi,  T.  Tsumura,  F.  Nakasato and  T.  Kudo: Tetsu-to-Hagané, 82 (1996), 297 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.82.4_297
• 4)   T.  Kushida: ISIJ Int., 43 (2003), 470. https://doi.org/10.2355/isijinternational.43.470
• 5)   J. P.  Hirth: Metall. Trans. A, 11 (1980), 861. https://doi.org/10.1007/BF02654700
• 6)   K.  Takashima,  Y.  Yoshioka,  K.  Yokoyama and  Y.  Funakawa: ISIJ Int., 58 (2018), 173. https://doi.org/10.2355/isijinternational.ISIJINT-2017-315
• 7)   S.  Li,  Z.  Zhang,  E.  Akiyama,  K.  Tsuzaki and  B.  Zhang: Corros. Sci., 52 (2010), 1660. https://doi.org/10.1016/j.corsci.2010.02.005
• 8)   Y.  Toji,  S.  Takagi,  M.  Yoshino,  K.  Hasegawa and  Y.  Tanaka: Tetsu-to-Hagané, 95 (2009), 887. https://doi.org/10.2355/tetsutohagane.95.887
• 9)   S.  Misawa: Okurehakaikaimei-no-Shintenkai (New Development of Delayed Fracture Elucidation), ISIJ, Tokyo, (1997), 82.
• 10)   E.  Akiyama,  K.  Matsukado,  M.  Wang and  K.  Tsuzaki: Corros. Sci., 52 (2010), 2758. https://doi.org/10.1016/j.corsci.2009.11.046
• 11)  ISO 9223: 2012, Corrosion of metals and alloys - Corrosivity of atmospheres - Classification, determination and estimation.
• 12)   E.  Tada,  F.  Zheng,  A.  Nishikata,  T.  Tsuru,  S.  Ootsuka,  H.  Nakamaru,  Y.  Sugimoto and  S.  Fujita: CAMP-ISIJ, 25 (2012), 509 (in Japanese).
• 13)   S.  Ootsuka,  S.  Fujita,  E.  Tada,  A.  Nishikata and  T.  Tsuru: Corros. Sci., 98 (2015), 430. https://doi.org/10.1016/j.corsci.2015.05.049
• 14)   S.  Ootsuka,  E.  Tada,  A.  Nishikata,  S.  Fujita and  T.  Tsuru: Tetsu-to-Hagané, 103 (2017), 27 (in Japanese). https://doi.org/10.2355/tetsutohagane.TETSU-2016-068
• 15)   S.  Suzuki,  D.  Mizuno and  S.  Fujita: GALVATECH ’07, ISIJ, Tokyo, (2007), 705.
• 16)   D.  Mizuno,  S.  Suzuki,  S.  Fujita and  N.  Hara: Corros. Sci., 83 (2014), 217. https://doi.org/10.1016/j.corsci.2014.02.020
• 17)   T.  Omura,  T.  Kushida,  T.  Kudo,  F.  Nakasato and  S.  Watanabe: Zairyo-to-Kankyo, 54 (2005), 61 (in Japanese). https://doi.org/10.3323/jcorr1991.54.61
• 18)   M. A. V.  Devanathan and  Z.  Stachurski: Proc. R. Soc. Lond., Ser. A, 270 (1962), 90. https://doi.org/10.1098/rspa.1962.0205
• 19)   M. A. V.  Devanathan,  Z.  Stachurski and  W.  Beck: J. Electrochem. Soc., 110 (1963), 886.
• 20)   M. A. V.  Devanathan and  Z.  Stachurski: J. Electrochem. Soc., 111 (1964), 619.
• 21)   T.  Kushida: Zairyo-to-Kankyo, 49 (2000), 195 (in Japanese). https://doi.org/10.3323/jcorr1991.49.195
• 22)  ISO 8407: 1991, Corrosion of metals and alloys - Removal of corrosion products from corrosion test specimens.
• 23)   S.  Yamasaki and  T.  Takahashi: Tetsu-to-Hagané, 83 (1997), 460 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.83.7_460
• 24)   T.  Ishikawa and  T.  Nakayama: Zairyo-to-Kankyo, 52 (2003), 140 (in Japanese). https://doi.org/10.3323/jcorr1991.52.140
• 25)   H.  Masuda: Shinku, 44 (2001), 868 (in Japanese). https://doi.org/10.3131/jvsj.44.868
• 26)   H.  Harada,  M.  Omoda,  S.  Ootsuka,  T.  Kawano,  H.  Kajiyama and  T.  Tsuru: CAMP-ISIJ, 28 (2016), 165 (in Japanese).
• 27)   W.  Oshikawa,  T.  Shinohara and  S.  Motoda: Zairyo-to-Kankyo, 52 (2003), 293 (in Japanese). https://doi.org/10.3323/jcorr1991.52.293
• 28)   T.  Shinohara: J. Surf. Sci. Soc. Jpn., 36 (2015), 4 (in Japanese). https://doi.org/10.1380/jsssj.36.4
• 29)   M.  Takahashi: Corros. Eng. (Jpn.), 23 (1974), 625 (in Japanese). https://doi.org/10.3323/jcorr1974.23.12_625
• 30)   S.  Ajito,  E.  Tada,  A.  Ooi and  A.  Nishikata: ISIJ Int., 59 (2019), 1659. https://doi.org/10.2355/isijinternational.ISIJINT-2019-113
• 31)   V. L.  Snoeyink and  D.  Jenkins: Water Chemistry, John Wiley & Sons, New York, (1980), 266.
• 32)   Y.  Shi,  E.  Tada and  A.  Nishikata: J. Electrochem. Soc., 162 (2015), No. 4, C135. https://doi.org/10.1149/2.0101504jes