Exposure Test Performance for Chromium Bearing Steel in Concrete under Coastal Environment

Abstract

The corrosion resistance of Cr bearing steel (CR) in concrete was estimated by the exposure test in a coastal environment. Carbon steel (SM) had a significant corrosion on the surface, and the concrete block had large cracks. On the other hand, CR had little corrosion, and there was no crack on the concrete. In the case of Electrochemical Impedance Spectroscopy (EIS) measurements, the impedance at low frequency region (Z_L) of CR showed much higher values than that of SM. Cr and Si were enriched in inner rust of CR observed by Energy Dispersive Spectroscopy (EDS) measurement. Transmission Electron Microscopy (TEM) showed that nano complex iron oxides containing Cr and Si were formed in inner rust, which suppressed the corrosion of CR in the concrete.

1. Introduction

Recently, the deterioration of concrete structure has been caused by the corrosion of reinforcing steels. Specially, the corrosion of steels in concrete is accelerated by airborne salt from the sea. Besides, there are some reports which show that the road salt decreases the corrosion resistance of steel in concrete structures in winter season. Therefore the stainless steel has been applied to reinforcing rebar,^1,2,3) however, it is not widely used because of its high cost. Although Cr bearing steel has the attractive attention as a high corrosion-resistant steel, only a few studies⁴⁾ have evaluated its corrosion performance.

The corrosion resistance of Cr bearing steel in concrete is thought to depend on the rust (corrosion product) on the surface under actual environment. In the case of the low alloy weathering steel, there are many reports analyzed the rust formation and the effect of alloying element.^{5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)} However, the structure of the rust on Cr bearing steel is not fully understood.

In the previous paper,^20,21) Cr bearing steel have showed high resistance against the corrosion in accelerate tests. Specially, the rust (R_rust) and the corrosion resistance (R_t) of Cr bearing steel were demonstrated to be very high in concrete as compared to SM after the wet and dry test by EIS. Moreover, the nano complex iron oxides containing Cr and Si were detected by TEM in the rust of Cr bearing steel, which could increase the corrosion resistance of this steel.

In this paper, the exposure test performance evaluation of Cr bearing steel in concrete was examined in a coastal environment in Miyako-island in Okinawa prefecture. Besides, the nano structure of the rust of Cr bearing steel in concrete was investigated in detail using TEM – EELS (Electron Energy Loss Spectroscopy). Finally, the relationship between the high corrosion resistance and the rust formation was discussed for the Cr bearing steel in concrete exposed under coastal environment.

2. Experimental Procedure

2.1. Test Sample and Exposure Test

Concrete samples were prepared from ordinary Portland cement with a water/cement ratio of 0.6. The dimension of sample block was 80 × 80 × 80 mm, where G_max (Maximum size of rock) was 15 mm. The position of the steel plates was set at 10 or 20 mm from the test surface for the cover thickness. The chemical composition of Cr bearing reinforcing steel (CR) was 7% Cr - 2% Si–Fe in mass%, additionally, carbon steel (SM) was used for the comparison.

The exposure test was conducted for 2 years at the exposure test site near Irabu Oohashi (bridge) in Miyako-island as shown in Fig. 1. As the test samples were exposed on the tetrapod located at a distance of 4 m from the sea, the sea water was splashed directly on the samples. The annual climate data (temperature, RH (relative humidity), amount of rain) for the test site are shown in Fig. 2. The monthly average temperature and RH during the corrosion test ranged 19–31 C and 65–84%, respectively, and the amount of rain shows high values during 3 seasons in a year, which is identified as a high humidity climate in the subtropical zone.

Fig. 1.

Photos for the (a) Irabu Oohashi and (b) Exposure test site.

Fig. 2.

Changes for the Rain (mm/y), Temperature (°C) and RH (%) in a year at Miyako-shima.

After the exposure test, depth profiles of chloride (Cl) concentration and pH from the test surface in concrete block were measured by using the Ag/AgCl and W/WOx microelectrodes²⁰⁾ as shown in Fig. 3. The concentration of Cl is high at 5 and 10 mm from the surface, and it keeps around 0.2–0.3 M at 20–40 mm. The pH is low around 8.5 at 5–20 mm, and it increases to 9 at 20–40 mm. Thus, Cl concentrations at steel position of 10 and 20 mm are very high, which indicates the very severe corrosion condition in concrete block.

Fig. 3.

Depth profiles of Cl concentration and pH from the test surface in concrete block (sample a and b) after the exposure test.

2.2. Physical Analysis of Rust Formed on the Steel in Concrete

After the exposure test of 2 years, the rust on steels in concrete was measured by the surface analyses. XRD (X-Ray Diffraction) was performed on the rust of steels. The acceleration voltage was 12 kV and a Cu target was used. The scanning range was 10–35 degree at a speed of 2.0 degree/min. An internal standard method was conducted using ZnO as the standard chemicals, which could estimate the rust composition.

The chemical composition in rust was measured by SEM (Scanning Electron Microscopy) - EDS (Electron Dispersing Spectroscopy). The concentration and enrichment of various elements in the rust were investigated in detail. Besides, micro Raman spectroscopy was performed using a 532 nm laser beam, where the frequency region was 4000–200 cm⁻¹ to detect Fe oxides.

The rust on steel was measured at a nanostructure level by using TEM. The rust was cut using FIB (focused ion beam) from the inner rust near to the steel. EELS (Electron Energy Loss Spectroscopy) analysis was performed to identify the chemical state of the elements, such as O, Fe, Cr, and Si. Finally, the nano structure of inner rust was discussed for test steels in concrete.

2.3. Electrochemical Impedance Spectroscopy (EIS) Measurement of the Rust

EIS measurements were conducted on the rust of low alloy steels in concrete after the exposure test of 2 years. EIS measurements were performed for steels in concrete immerged in 0.1 M Na₂SO₄ solution. The same 2 electrodes of exposed steels are used in EIS measurement with a frequency range of 20 kHz to 1 mHz and an applied voltage of 10 mV.

3. Results and Discussion

3.1. Corrosion Behavior of Steels in Concrete at the Exposure Test Site

After the exposure test for 2 years, the steels and concrete blocks are observed as shown in Fig. 4. The carbon steel (SM) has a considerable corrosion on the surface, and the concrete block has large cracks. This behavior is explained that the expansion of the corrosion product (rust) causes the cracks of the concrete block. In contrast, CR has little corrosion on the surface, and there is no crack on the concrete. Thus, it is possible to demonstrate that CR has significantly higher corrosion resistance than that of SM in concrete at the exposure site.

Fig. 4.

Photos for the concrete samples and embedded (1) SM (carbon steel) and CR (Cr steel) after the exposure test.

The EIS spectra for SM and CR at 10 and 20 mm from the surface in concrete after the exposure test are indicated in Fig. 5. The impedance at low frequency region (Z_L) is thought corresponding to the corrosion behavior. Z_L of CR increases as the frequency decreases, besides, the phase shift is around – 60 degree, which shows the Warburg impedance (Z_w) depending on the diffusion of the oxygen. On the other hand, Z_L of SM indicates a very low value, and the phase shift is almost 0 degree, which shows the charge transfer resistance (R_ct). In the case of surface observation of steel in concrete after the exposure test in Fig. 4, CR almost keeps the passive film. Besides, the corrosion products on CR is supposed to stop the corrosion. Thus, the passive film of CR is thought to be kept when Z_w is shown in EIS. However, SM has heavy corrosion in Fig. 4. The passive film of SM is brake down when R_ct is detected in EIS. In this way, from EIS measurement, CR is thought to have much higher corrosion resistance when Z_w is shown in EIS.

Fig. 5.

EIS spectra for SM and CR at (a) 10 mm and (b) 20 mm from the surface after the exposure test.

3.2. Characterization of the Rust Formed on Cr Bearing Steel

The surface analyses are conducted on the rust on CR and SM in concrete after the exposure test for 2 years. Figure 6 shows XRD results of (1) Diffraction patterns and (2) Mass% of Fe oxides for CR and SM. In XRD spectrum of SM, the α, β, γ -FeOOH and Fe₃O₄ are recognized. Moreover, the peak of NaCl is identified, which indicates a strong saline environment. This fact is corresponding to the detection of β -FeOOH in the rust of SM. On the other hand, the XRD spectrum of CR is slightly broader, indicating that rust particles are very fine. A quantitative evaluation is conducted from XRD results using ZnO as the standard chemicals as shown in Fig. 6(2). The amount of α -FeOOH and Fe₃O₄ are high in the rust of SM and CR. However, 70 and 80 mass% of Fe rust cannot be detected using XRD in SM and CR, respectively. Thus the Fe oxide particles are too fine to be examined using XRD, which is shown as Amor. in the Figure.

Fig. 6.

XRD results of (1) Diffraction patterns and (2) Mass% of Fe oxides.

The cross sections of the rusts for CR sample are measured by SEM-EDS after the exposure test in Fig. 7. The concentration of Cr and Si are very high in inner rust. Moreover, as they are enriched at the same position, they are thought to generate the complex oxides. In the result of Cr, the measuring positions for Raman spectroscopy and FIB-TEM are indicated.

Fig. 7.

SEM-EDS results for the rust of Cr steel after the exposure test.

The line scans of content percentages of Cr and Si in the rust are calculated from EDS results. In the figure of Si as shown in Fig. 7, the average (C_av(x0)) of EDS counts of Si is calculated at the position of x0 by using 256 points of data along the vertical (y) axis. Similarly, all C_av are calculated along the horizontal axis (x0 - xn), which is shown as the line scan of contents of Si and Cr in Fig. 8. The line scan of Cr has a peak (I max) in the rust, which is calculated as a value of 12.1 as compared to the base metal of 7.0 mass%. Thus, the Cr content in the rust is very high. Similarly, the line scan of Si has a peak (I max) of 3.6 relative to the base metal of 2.0 mass%. Thus, the Si content in the rust is higher than that of the base metal. These enrichments of Cr and Si are able to prevent the penetration of Cl ions in the rust, which is discussed in the later.

Fig. 8.

Line scan of contents of Cr and Si according to SEM-EDS results in Fig. 7.

As all of Fe oxides in the rust could not be detected using XRD because the oxide particles were too fine. Therefore, the micro Raman spectroscopy was applied to the rust of CR sample. Figure 9 indicates Raman results of the rust formed on CR after the exposure test for 2 years. The experimental positions are corresponding to those in Fig. 7. In inner rust of CR, α, β –FeOOH and Fe₃O₄ are recognized. In outer rust, α –FeOOH and Fe₃O₄ are detected as the shoulders of the spectra. Thus, the same Fe oxides are detected both in inner and outer rusts. The peaks of β –FeOOH are observed to be very high, because the Cl concentration is high in the rust. This behavior corresponds to the results of Cl concentration measured in the concrete block shown in Fig. 3. Moreover, α-FeOOH and Fe₃O₄ are detected both in inner and outer rust. The α-FeOOH and Fe₃O₄ are supposed to act as the protective rust, and β -FeOOH decreases the corrosion resistance. In the actual case in inner and outer rust of CR, both types of the Fe oxides are detected as shown in Fig. 9. Thus, although β -FeOOH exists in the rust, the protective behaviors of α-FeOOH and Fe₃O₄ are excellent, which increases the corrosion resistance of CR. In order to investigate the rust in more detail, the nano level observation of the rust was conducted using FIB- TEM analysis.

Fig. 9.

Raman results of the rust formed on CR (Cr steel) after the exposure test. The experimental positions are shown in Fig. 7.

Figure 10 indicates the FIB – TEM results of the rust at the position near the steel for CR sample as shown in Fig. 7. In the figure, the upper and lower sides show the rust and steel, respectively. Besides, the bright regions in the figure present the thin part and the dark regions present the thick one in the FIB sample. The points in the figure correspond to the results of EDS in Table 1 and TEM-EELS in Figs. 11, 12, 13.

Fig. 10.

FIB – TEM observation taken from the position in Fig. 7 for the rust of CR (Cr steel).

Table 1. EDS results for the points in the inner rust of Cr steel as shown in Fig. 10.

								(atomic%)
Point	Cr	Si	Fe	Na	O	Cl	C	Cr, Si content
1	13	9	5	1	61	0	9	rich
2	13	9	4	0	62	0	11	rich
3	11	7	5	0	58	0	11	rich
4	12	7	4	0	60	0	13	rich
5	9	6	4	5	57	0	13	middle
6	8	6	4	5	57	0	19	middle
7	1	1	21	5	51	0	16	less
8	1	2	25	3	51	0	14	less
9	0	0	0	1	5	0	92	polymer

Fig. 11.

TEM-EELS spectra of O–K for the rust of CR (Cr steel).

Fig. 12.

TEM-EELS spectra of Fe–L for the rust of CR.

Fig. 13.

TEM-EELS spectra of Cr–L for the rust of CR.

Fig. 14.

TEM-EELS spectra of Si–L for the rust of CR.

For the points in Fig. 10, EDS results for the inner rust of CR are shown in Table 1. In the case of point 9, as the content of C is very high, this is the polymer outer region of the rust. Thus, the point 9 is eliminated from the consideration of the rust structure in the followings. Except point 9, as all points show high oxygen in the range of 51–62 atomic%, they are thought as the oxidized states. The point 1 to 4 indicate high Cr of 11–13 and high Si of 7–9 atomic%, which is higher than those of Fe of 4–5 atomic%. Therefore, they are recognized as Cr and Si rich positions in the rust. The point 5 and 6 indicate average contents of Cr of 8–9 and Si of 6 atomic%, which are named positions of middle content of Cr and Si. The point 7 and 8 indicate less Cr of 1 and less Si of 1–2 atomic%, which are named Cr and Si less positions. Besides, they show higher contents of Fe of 21–25 relative to points 1–6 of 4–5 atomic%. According to the contents of Cr and Si, there are 3 types of oxides are recognized as the rich (points 1–4), middle (points 5–6) and less (points 7–8) contents. The following examinations are conducted according to these 3 types of oxides.

TEM-EELS spectra of O–K for the rust of CR in Fig. 10 are shown in Fig. 11. Spectra 1–4 are indicated for Cr and Si rich oxides, and spectra 5–6 are middle, and spectra 7–8 are less content, respectively. There are 2 peaks in O–K spectra at 532 and 540 eV. At lower energy peak of 532 eV, the peaks of middle content indicate the broad shape, which is different from that of rich one. Thus this behavior is believed to be obeyed by the content of Cr and Si in the rust.

TEM-EELS spectra of Fe–L for the rust of CR in Fig. 10 are shown in Fig. 12. Spectra 1–4 are indicated for Cr and Si rich oxides, and spectra 5–6 are middle, and spectra 7–8 are less content respectively. There are 2 peaks in Fe–L spectra of Fe–L₃ at 709 and Fe–L₂ at 722 eV. The spectra 7–8 show very sharp shape because Fe contents are high. However, there are almost the same peak position of Fe–L₃ and Fe–L₂ in all spectra. From the previous paper,²²⁾ these peak positions correspond Fe (III) oxide state. Therefore, the Fe is thought to be involved in the rust as Fe (III) oxide state in inner rust.

TEM-EELS spectra of Cr–L for the rust of CR in Fig. 10 are shown in Fig. 13. Spectra 1–4 are indicated for Cr and Si rich oxides, and spectra 5–6 are middle content respectively. There are 2 peaks in Cr–L spectra of Cr–L₃ at lower energy region and Cr–L₂ at higher one. The peak position of Cr–L₃ are shown at 570 eV in spectra 1–4, however, that are at 568 eV in spectra 5–6. Besides, the peak position of Cr–L₂ are shown at 579 eV in spectra 1–4, however, those are at 577 eV in spectra 5–6. Thus both the peak positions for Cr–L₃ and Cr–L₂ of spectra 5–6 are 2 eV lower than those of spectra 1–4. From the previous paper²³⁾ concerning the chemical shift of spectra, the peak positions for Cr–L₃ and Cr–L₂ at lower energy correspond to Cr (II) oxide state, and those at higher energy are Cr (III) one. In other words, Cr is thought to exist as Cr (III) oxide state in Cr and Si rich oxides, and as Cr (II) one in middle content oxides. This behavior is thought to be based on the Fe oxide states in inner rust. As this information is very important to understand the formation of Fe oxides, which is discussed later.

TEM-EELS spectra of Si–L for the rust of CR in Fig. 10 are shown in Fig. 13. Spectra 1–4 are demonstrated for Cr and Si rich oxides, and spectra 5–6 are middle content respectively. There are 3 peaks in Si–L spectra of Si–L₃, Si–L₂, and Si–L₁ from higher to lower energy region. Although spectra are a little complicated, the shapes of them are almost the same. Thus there is not much difference between spectra of Cr and Si rich oxides and middle content. From the previous paper,²⁴⁾ the shape of these spectra are similar to that of Si oxides. Thus Si is thought to exist as Si oxide state in inner rust.

3.3. Corrosion Resistant Mechanism of Cr Bearing Steel (CR) in Concrete

In the exposure test, CR demonstrated significantly higher corrosion resistance than SM in concrete. SM had a considerable corrosion on the surface, and the concrete block had large cracks. However, CR had little corrosion on the surface, and there was no crack on the concrete. In the case of EIS measurements, the impedance at low frequency region (Z_L) of CR showed much higher values than those of SM. Therefore, the corrosion resistance of CR was indicated very high in concrete at the exposure site.

In the case of exposure test, although CR almost had a passivated surface, there are some corrosion products on the surface. Thus, these corrosion products were believed to stop further corrosion of steel. Thus, the high corrosion resistance of CR was thought to be related to the rust structure of the steel in concrete.

In TEM-EELS measurements, the chemical shift of Cr–L₃ and Cr–L₂ were recognized in inner rust, which corresponded to Cr (II) and Cr (III) oxide state. Moreover, Cr was thought to exist as Cr (III) oxide state in Cr and Si rich oxides, however, it exists as Cr (II) state in middle content oxides. Here, only Fe₃O₄ structure as Fe (II) Fe (III)₂ O₄ can involve bivalent ion of metal (II) such as Cr (II) or Fe (II). Thus, in the case of Cr oxide, Cr (II) was involved in Fe₃O₄ structure in the form of Cr (II) Fe (III)₂ O₄. Actually in the Raman result, Fe₃O₄ was detected in inner rust of CR. Besides, Fe was observed as Fe (III) oxide state in inner rust in TEM – EELS measurement. Thus the bivalent site of Metal (II) in Fe₃O₄ could be occupied only by Cr (II). Therefore, Cr was thought to exist in Cr (II) Fe (III)₂ O₄ in middle content oxides. In fact, as the content of Fe₃O₄ was measured to be lower than 10% in the rust of CR from the XRD measurement, Fe₃O₄ was the spinel oxide which could not detect by XRD. On the other hand, Cr was thought to exist as Cr (III) oxide state in Cr and Si rich oxides. In the Raman result, α – FeOOH was also detected in inner rust of CR. Thus, Cr (III) was thought to be also involved in α – FeOOH, which showed high corrosion resistance. Therefore, both the spinel oxide and α – FeOOH containing Cr and Si could increase the corrosion resistance of the rust.

In the Raman results, β– FeOOH was also detected in inner rust of CR, which was caused by high chloride ions by sea water. Under this severe condition, high corrosion resistance of the rust of CR was able to be kept by the spinel oxide and α – FeOOH layer. Thus even containing β– FeOOH, CR could maintain the high corrosion resistance. From the result of SEM-EDS, contents of Cr and Si were enriched in the rust. Moreover, in the case of TEM-EDS, contents of Cr and Si were very high at the interface of the rust and the steel. Therefore this high Cr–Si layer at the interface was thought to prevent the chloride penetration.

In the previous paper,²⁵⁾ the examination of the selective permeability on the rust of low alloy steels was conducted. Although the rust of carbon steel showed the anion selective permeability, some alloying elements of steel could change it to the cation permeability. Thus, the inner rust containing Cr and Si were thought to show the cation permeability, and prevent the chloride ions from penetrating into the rust. In addition to this chemical prevention, as the inner rust was formed by nano oxides, the penetration of chloride ions were thought to be physically prevented. It was found that CR showed high corrosion resistance by the rust structure which contained high Cr and high Si.

4. Conclusion

Cr bearing steel (CR) demonstrated much higher corrosion resistance than SM in concrete in the exposure test. SM had a considerable corrosion on the surface, and the concrete block had large cracks. In contrast, CR had little corrosion on the surface, and there was no crack on the concrete. EIS measurements showed that the impedance at low frequency region (Z_L) of CR was much higher values than that of SM. In inner rust of CR, Cr and Si were enriched in nano iron oxides, which was believed to increase the corrosion resistance of CR in concrete. Finally, it was found that CR had high corrosion - resistant performance in concrete under the coastal environment.

Acknowledgements

This work was supported by Council for Science, Technology and Innovation (CSTI), Cross-ministerial Strategic Innovation Promotion Program (SIP), “Infrastructure maintenance, renovation and management” (Funding agency: JST).

References

• 1)   R. G.  Duarte,  A. S.  Castela,  R.  Neves,  L.  Freire and  M. F.  Montemor: Electrochim. Acta, 124 (2014), 218.
• 2)   M.  Serdar,  L. V.  Zulj and  D.  Bjegovic: Corros. Sci., 69 (2013), 149.
• 3)   R. D.  Moser,  P. M.  Singh,  L. F.  Kahn and  K. E.  Kurtis: Corros. Sci., 57 (2012), 241.
• 4)   T.  Sung-Ho and  T.  Ujiro: ISIJ Int., 47 (2007), 146.
• 5)   U. R.  Evans: Corros. Sci., 9 (1969), 813.
• 6)   T.  Misawa,  K.  Asami,  K.  Hashimoto and  A.  Shimodaira: Corros. Sci., 14 (1974), 279.
• 7)   I.  Suzuki,  Y.  Hisamatsu and  N.  Masuko: J. Electrochem. Soc., 127 (1980), 2210.
• 8)   M.  Stratmann,  K.  Bohnenkamp and  T.  Ramchandran: Corros. Sci., 27 (1987), 905.
• 9)   J.  Dunnwald and  A.  Otto: Corros. Sci., 29 (1989), 1167.
• 10)   H. E.  Townsend: Corrosion, 57 (2001), 497.
• 11)   M.  Yamashita,  H.  Miyuki,  Y.  Mastuda,  H.  Nagano and  T.  Misawa: Corros. Sci., 36 (1994), 283.
• 12)   T.  Nishimura,  H.  Katayama,  K.  Noda and  T.  Kodama: Corrosion, 56 (2000), 935.
• 13)   T.  Nishimura,  H.  Katayama,  K.  Noda and  T.  Kodama: Corros. Sci., 42 (2000), 1611.
• 14)   T.  Nishimura,  K.  Noda and  T.  Kodama: Corrosion, 57 (2001), 753.
• 15)   T.  Nishimura,  A.  Tahara and  T.  Kodama: Mater. Trans., 42 (2001), 478.
• 16)   M.  Kimura,  H.  Kihira,  N.  Ohta,  M.  Hashimoto and  T.  Senuma: Corros. Sci., 47 (2005), 2499.
• 17)   T.  Nakayama,  T.  Ishikawa and  T.  Konno: Corros. Sci., 47 (2005), 2521.
• 18)   T.  Ohtsuka and  T.  Komatsu: Corros. Sci., 47 (2005), 2571.
• 19)   A.  Nishikata,  F.  Suzuki and  T.  Tsuru: Corros. Sci., 47 (2005), 2578.
• 20)   V.  Raman and  T.  Nishimura: J. Solid State Electrochem., 14 (2010), 1457.
• 21)   T.  Nishimura: ISIJ Int., 55 (2015), 1747.
• 22)   H.  Tan,  J.  Verbeeck,  A.  Abakumov and  G.  Van Tendeloo: Ultramicroscopy, 116 (2012), 24.
• 23)   T. L.  Daulton and  B. J.  Little: Ultramicroscopy, 106 (2006), 561.
• 24)   T.  Nishimura: Mater. Trans., 48 (2007), 1438.
• 25)   K.  Noda,  T.  Nishimura,  H.  Masuda and  T.  Kodama: J. Jpn. Inst. Met., 64 (2000), 148.