Applied Surface Science 329 (2015) 104–115

Contents lists available at ScienceDirect

Applied Surface Science

journal homepage: www.elsevier.com/locate/apsusc

Effect of O2 on corrosion of 3Cr steel in high temperature and high

pressure CO2 –O2 environment
Xueqiang Lin, Wei Liu ∗ , Fei Wu, Chuanchuan Xu, Juanjuan Dou, Minxu Lu
Corrosion and Protection Center, Institute for Advanced Materials and Technology, University of Science and Technology Beijing, Beijing 100083, PR China

a r t i c l e i n f o a b s t r a c t

Article history: The development of corrosion scale of 3Cr steel in an aqueous environment containing 95% CO2 –5% O2
Received 8 October 2014 was characterized and analyzed by scanning electron microscopy (SEM), energy-dispersive spectrometry
Received in revised form (EDS), X-ray diffraction (XRD) and in situ electrochemical impedance spectroscopy (EIS). The results show
10 December 2014
that the corrosion of 3Cr steel was promoted by dissolved oxygen. Fe(OH)3 precipitation formed at local
Accepted 16 December 2014
areas, leading to the nonuniform distribution of Cr element and pitting corrosion. With the increase of cor-
Available online 25 December 2014
rosion time, Cr-rich area formed at the bottom of pits, where the pitting corrosion process was inhibited.
The importance of Fe(OH)3 and Cr-rich area on the formation of corrosion pits and pits development was
Keywords:
3Cr steel
emphasized.
SEM © 2014 Elsevier B.V. All rights reserved.
EIS
O2 corrosion
CO2 corrosion
Pitting corrosion

1. Introduction oil, high-temperature multi-component thermal fluent containing

small amounts of O2 needs to be injected into the oil wells [13–15].
CO2 corrosion of 3Cr steel is a research focus in petroleum indus- The water containing both dissolved oxygen (DO) and CO2 was
try for its high CO2 corrosion resistance, which is about 2.5–40 more corrosive than water containing only a single component of
times higher than that of carbon steel and lower maintenance cost, DO or CO2 [16,17]. CO2 corrosion was dominant at low DO levels,
which is about 1.5 times less than that of conventional carbon however, the corrosion was considerably enhanced at higher DO
steel [1–4]. Some researchers considered that 3% Cr content in steel levels (>1 ppm) [18]. McIntire [19] proposed that CO2 was carried
could significantly improve the density and adhesion of the corro- out as a catalytic role in corrosion of iron at normal pH levels by
sion product scale formed in CO2 environment [2,5]. Cr element providing a pathway for dissolution of the initially formed corro-
could cause the transition of corrosion product from crystalline to sion scale from the passive layer and further oxidation to ␥-FeOOH,
amorphous state, which results in denser corrosion scale and lower which was the final observed corrosion scale, with the existence
susceptibility to localized corrosion [2]. The Cr-enrichment in cor- of higher concentrations of DO. Further, John [11] illustrated that
rosion scale could prevent corrosive anions such as Cl− , CO3 2− and trace levels of DO affected the hydration rate of CO2 and might
HCO3 − from penetrating into the corrosion scale and reduce the promote the hydrogen evolution reaction. Previous results showed
anion concentration in the interface between the corrosion scale that in the CO2 environment with high O2 concentration, Fe2+ was
and steel matrix [5–8]. easily oxidized to Fe3+ , and the corrosion product of N80 steel is
However, O2 is inevitably introduced into the oil and gas produc- composed of FeCO3 , Fe2 O3 , and Fe3 O4 . The protective capability
tion through chemical treatments, the reinjection of CO2 , inhibitors of this kind of corrosion product scale was weaker than FeCO3
and produced water, etc. [9–12]. Moreover, with the develop- corrosion scale [20]. Although the problem of CO2 corrosion has
ment and application of thermal extraction of offshore heavy crude been widely studied, very few studies have been carried out for
the influence of DO to CO2 corrosion of 3Cr steel. Thus, the current
study is aimed at investigating the corrosion behavior of 3Cr steel in
high temperature and high pressure CO2 environments containing
∗ Corresponding author. Tel.: +86 13681205463; fax: +86 10 62334410.
O2 .
E-mail address: [email protected] (W. Liu).

http://dx.doi.org/10.1016/j.apsusc.2014.12.109
0169-4332/© 2014 Elsevier B.V. All rights reserved.
 X. Lin et al. / Applied Surface Science 329 (2015) 104–115 105

2. Experimental Ag/AgCl electrode filled with 0.1 mol/L potassium chloride solu-
tion. The frequency range for EIS measurement was from 100 kHz
2.1. Materials and solutions to 10 mHz with a 5 mV amplitude (peak to zero) at open circuit
potential. The EIS data were fitted and analyzed by ZSimpWin soft-
The chemical composition of 3Cr steel with the microstruc- ware.
ture of acicular ferrite, was shown in Table 1. The specimens were
machined into the size of 10 mm × 10 mm × 3 mm, and then ground 3. Results
with SiC paper up to grade 800 #. Before the corrosion experi-
ment, the specimens were mounted with epoxy resin with the area 3.1. Corrosion rate and morphologies
of 1 cm2 exposed to the environment. Then the specimens were
cleaned by distilled water and ethanol, and then dried with cool air. The average corrosion rates of 3Cr steels in two aqueous envi-
The test solution simulating the formation water in one gas field, ronments containing CO2 and 95% CO2 –5% O2 after 120 h of
was made up with analytical grade reagents and deionized water immersion were 0.22 mm/y and 1.36 mm/y, respectively, which
(18 M cm in resistivity), the chemical composition of which was indicated that the corrosion rate increased significantly with the
shown in Table 2. existence of O2 . The macroscopic morphologies of 3Cr specimens
before and after the removal of corrosion scale were shown in Fig. 2.
2.2. Corrosion test 3Cr steel suffered severe localized corrosion with amount of pits on
the surface in an aqueous environment containing 95% CO2 –5% O2 .
An autoclave was used to simulate the static corrosion envi- In contrast, 3Cr steel presented general corrosion morphology in
ronment under high temperature and high pressure condition, as CO2 environment, which indicated that the addition of O2 reduced
shown in Fig. 1. The corrosion tests were performed at the temper- the localized corrosion resistance of 3Cr steel. The integrity of the
ature of 60 ◦ C and the total pressure of the mixed gas was 1 MPa corrosion scale could be destroyed when 5% O2 was introduced into
i.e. pO2 = 0.05 MPa and pCO2 = 0.95 MPa and pure CO2 environ- CO2 environment.
ment was used for comparison. Before the autoclave was heated,
the solution was firstly deaerated by pure N2 for 24 h and then 3.2. Characteristic of corrosion scale
continuously bubbled with mixed gas at least for 4 h. The test dura-
tion was 4 h, 8 h, 24 h, 48 h and 120 h, respectively. Six specimens Fig. 3 shows the surface morphologies of 3Cr steel after corro-
were used for each test. Among them, three specimens were for sion in the aqueous environments containing CO2 and 95% CO2 –5%
weight loss measurement. Two specimens were for surface and O2 for 120 h. The difference in the surface morphologies of corro-
cross-sectional morphology observation, and the last one for X-ray sion scales forming in CO2 environment and CO2 –O2 environment
diffraction analysis. were significant. Guo [21] found that the corrosion scale on the
Before weight loss tests, the specimens were weighed using surface of 3Cr steel forming in CO2 environment had a single Cr-
electronic analytical balance with an accuracy of 10−4 g, and the rich amorphous structure before 517 h. According to Guo’s result,
weight values were recorded as the original weight (W0 ). Then the the corrosion layer shown in Fig. 3a and b was amorphous and
specimens were mounted with epoxy resin and cleaned by distilled also continuous and tough, which covered the whole specimen
water and ethanol, and dried by cool air. After the corrosion tests, surface. The cracking of the inner layer was induced by the dehy-
the corroded specimens were taken out from the epoxy resin. The dration effect [22]. In contrast, the corrosion scale of 3Cr steel
corrosion scale were removed according to ASTM G1-03 standard forming in an aqueous environment containing 95% CO2 –5% O2 had
(a chemical cleaning procedure was shown in Table 3). Then the a duplex layer structure, as shown in Fig. 3c and e, respectively.
specimens were rinsed and dried, finally reweighed to obtain the The loose and porous outer layer consisted of scattering granular
final weight (Wt ). The corrosion rate (CR ) was calculated according corrosion product (Fig. 3d), which provided little corrosion pro-
to Eq. (1): tection. After the outer layer of the granular scale was removed
with scotch tape, the inner layer appeared, as shown in Fig. 3e. On
8.76 × 104 × (W0 − Wt )
CR = (1) the surface of inner layer, some damaged regions were observed
St
and many cracks and needle-like scale appeared in the damaged
where W0 and Wt were the original and final weight of specimens, regions.
g, respectively; t was the immersion time, h;  was the density of
steel, g/cm3 ; and S was the exposed surface area, cm2 . 3.3. Macroscopic morphologies

2.3. Morphology observation and composition analyses Fig. 4 shows the macroscopic morphologies of 3Cr steel in an
aqueous environment containing 95% CO2 –5% O2 before and after
The surface and cross-sectional morphologies of corrosion the removal of corrosion scale within 120 h. Before 24 h, the macro-
scale were characterized by scanning electron microscopy (SEM, scopic morphologies of specimens were characterized as general
LEO-1450). Corrosion scale of corroded specimens were analyzed corrosion, whereas typical localized corrosion was observed after
using X-ray diffraction (XRD, CuK␣,  = 0.154 nm, Rigaku). Energy- 48 h and 120 h. The appearance of localized corrosion was mainly
dispersive spectrometry (EDS, Kevex SuperDry) was employed to determined by the corrosion scale. For example, at 48 h, there
analyze the composition of the corrosion scale. were several reddish brown protuberances on the corrosion scale
(Fig. 4c), and pits were found at the positions where the pro-
2.4. Electrochemical tests tuberances covered after removing the product (Fig. 4d). This
phenomenon could also be found at 120 h. It seemed that the
During the corrosion tests, electrochemical measurements were positions covered with protuberances were corroded relatively
performed using a three-electrode cell. The schematic of autoclave quicker, resulting in localized corrosion, which was in accordance
for in situ electrochemical measurements was shown in Fig. 1. The with the study of Yu [12] and John [11].
working electrode (WE) was 3Cr steel with a tested surface area Fig. 5 shows the average corrosion rate of 3Cr steel in an aqueous
of 1 cm2 . The counter electrode (CE) was a platinum plate. The ref- environment containing 95% CO2 –5% O2 at various corrosion times.
erence electrode (RE) was a high temperature and high pressure The results showed that the corrosion rate decreased rapidly with
106 X. Lin et al. / Applied Surface Science 329 (2015) 104–115

Table 1
Chemical composition of the tested 3Cr steel.

Elements C Si Mn S P Cr Ni Mo Fe

Wt.% 0.16 0.22 0.51 0.002 0.009 3.02 0.05 0.35 Bal.

Table 2
Chemical composition of the test solution.

Composition MgCl2 6H2 O CaCl2 Na2 CO3 NaHCO3 Na2 SO4 NaCl KCl

Content (mg/L) 73.8 49.9 383.6 1022.7 7.1 543.2 1.8

Fig. 1. Schematic of the autoclave used for electrochemical measure.

Table 3
Chemical cleaning procedure for removal of corrosion scale.

Solution Time Temperature Remarks

◦
500 mL hydrochloric acid (sp gr 1.19) 3.5 g hexamethylene 10 min 20–25 C Longer times may be required in certain instances
tetramine reagent water to make1000 mL

the increasing corrosion time within the initial 8 h, reduced slowly specimen surface was covered with corrosion scale, which further
from 8 h to 48 h, and then changed slightly from 48 h to 120 h. decreased the corrosion rate of the steel within 8–48 h. After cor-
Combined with the corrosion rate and macroscopic morphologies rosion for 48 h and 120 h, most of the regions were covered with
(Fig. 4), it could be seen that the surface of 3Cr steel rapidly dis- corrosion scale. These corrosion scales could inhibit the corrosion
solved and the corrosion scale rapidly formed on the surface of process, and thus the corrosion rate changed slightly. However, a
the steel, which could inhibit the corrosion process, leading to a large number of pits occurred within 48–120 h, as shown in Fig. 4d
rapid decrease in corrosion rate. With the corrosion proceed, the and f.

Fig. 2. Macroscopic morphologies of 3Cr steel corroded in an aqueous environment containing 95% CO2 –5% O2 at 120 h: (a) with corrosion scale; (b) without corrosion scale.
 X. Lin et al. / Applied Surface Science 329 (2015) 104–115 107

Fig. 3. Surface morphologies of corrosion scale of 3Cr steel at 120 h: (a) CO2 environment; (b) position A in Fig. 3a; (c) outer layer in an aqueous environment containing 95%
CO2 –5% O2 ; (d) position B in Fig. 3c; (e) inner layer in an aqueous environment containing 95% CO2 –5% O2 ; (f) position C in Fig. 3e.

3.4. Morphologies of outer layer by scotch tape, the inner layer could be observed. At 4 h, the inner
layer nearly covered the entire surface, and there was no apparent
Fig. 6 shows morphologies of the outer layer of 3Cr steel in an damaged area on the surface (Fig. 8a). However, cracks of two dif-
aqueous environment containing 95% CO2 –5% O2 at various corro- ferent degrees occurred in the inner layer, which was induced by
sion times. At 4 h, a small amount of granular scale scattered on the the dehydration effect. The Cr content in the serious cracking region
surface of 3Cr steel, leaving most parts of the inner layer exposed to A (Fig. 8a) was much lower than that in the slight cracking region B
the corrosive electrolyte (Fig. 6a). At 8 h, the quantity of the granular (Fig. 8a). It could be confirmed that the degree of the cracks in inner
scale increased, but the inner layer was still visible (Fig. 6b). At 24 h, layer was related to the nonuniform content of Cr in the inner layer.
the size and quantity of granular scale increased obviously with At 8 h, the inner layer was slightly damaged in local region, and
higher surface coverage (Fig. 6c). At 48 h, the size and quantity of the size of damaged region was approximately 36 ␮m (Fig. 8b). At
granular scale continued increasing, and a large number of granular 24 h, the size of damaged region increased to approximately 46 ␮m
scales piled up in partial region (Fig. 6d). At 120 h, the granular scale (Fig. 8c). At 48 h, the size of damaged region obviously increased up
almost covered the entire specimen surface (Fig. 6e). It could be to approximately 120 ␮m with a lot of cracks at the internal posi-
seen that the outer granular scale scattered on the inner layer, and tion of the damaged regions (Fig. 8d). At 120 h, a large damaged
the inner layer remained visible at all experimental periods. Mean- region occurred in the inner layer and the damaged regions were
while, the size and quantity of granular scale increased obviously connected together (Fig. 8e). Corrosion product formed in an aque-
with increasing corrosion time, and the loose and porous outer layer ous environment containing 95% CO2 –5% O2 with non-uniform Cr
provided little corrosion resistance to corrosive electrolyte. distribution was different from that formed in CO2 environment
The EDS analysis at region A in Fig. 6e showed that the main with uniform Cr distribution (Fig. 9a and b). The localized corro-
content of granular scale formed in an aqueous environment con- sion occurs due to the protection decrease of the low Cr regions in
taining 95% CO2 –5% O2 were Fe, O and Cr (0.23 wt.%), and the atomic the inner layer with Cr nonuniform distribution.
content of O was approximate 1.5 times as high as that of Fe. Addi- Fig. 10 was the XRD spectra of the inner corrosion scale of 3Cr
tionally, XRD results of the outer layer (Fig. 7b) showed that the steel in an aqueous environment containing 95% CO2 –5% O2 with
outer layer mainly consisted of granular Fe2 O3 , and the peaks of immersion of 4 h and 120 h. At 4 h, the inner layer was mainly
FeCO3 crystals came from the inner layer. The granular Fe2 O3 was consisted of crystalline FeCO3 and a small amount of iron oxide.
formed by the dehydration of Fe(OH)3 . EDS analysis showed that the main content of region A in Fig. 8a
was Fe, O, C (atomic ratio was 5: 1: 9) and Cr (4.50 wt.%) while the
3.5. Microscopic morphologies of inner layer main content of region B in Fig. 8a was Fe, O, C (atomic ratio was
1: 1: 3) and Cr (17.31 wt.%), which indicated that Fe and O contents
Fig. 8 shows the morphologies of inner layer of 3Cr steel in an of region A were higher than that of region B. According to the EDS
aqueous environment containing 95% CO2 –5% O2 at various corro- and XRD results of inner layer (Fig. 10a), it could be concluded that
sion times. After the outer layer of granular-Fe2 O3 was removed the corrosion scale of region A were iron oxides. Iron oxide scale
108 X. Lin et al. / Applied Surface Science 329 (2015) 104–115

Fig. 4. Macroscopic morphologies of 3Cr steel within 120 h in an aqueous environment containing 95% CO2 –5% O2 : (a) 24 h, with corrosion scale; (b) 24 h, without corrosion
scale; (c) 48 h, with corrosion scale; (d) 48 h, without corrosion scale; (e) 120 h, with corrosion scale; (f) 120 h, without corrosion scale.

formed in local area and a trace of element Cr formed in this sit-

uation. It was suggested [23] that the corrosion scale containing
low content of Cr provided little protection, so that localized corro-
sion often occurred in these regions. Fig. 10b shows the XRD result
of inner layer forming in the aqueous environment containing 95%
CO2 –5% O2 at 120 h. It could be seen that the corrosion scale in inner
layer also contained iron oxides as the same as those formed at 4 h.

3.6. Cross-section morphologies

Fig. 11 shows the cross-section morphologies of the corrosion

scale formed on the surface of 3Cr steel in an aqueous environment
containing 95% CO2 –5% O2 at various corrosion times. At 4 h and
8 h, a thin continuous product formed on the surface of 3Cr steel
and no apparent pits at the substrate, as shown in Fig. 11a and b.
However, at 24 h, a shallow-disk shape pit appeared with 33 ␮m
in width and 9 ␮m in depth, as shown in Fig. 11c. Moreover, at
48 h, the size of the pit increased quickly in width to 210 ␮m and
Fig. 5. Corrosion rate of 3Cr steel within 120 h in an aqueous environment contain-
slowly in depth to 20 ␮m, as shown in Fig. 11d. At 120 h, the pits
ing 95% CO2 –5% O2 .
in the interface between the substrate and corrosion scale showed
 X. Lin et al. / Applied Surface Science 329 (2015) 104–115 109

Fig. 6. Microscopic morphologies of outer layer forming on the surface of 3Cr steel in an aqueous environment containing 95% CO2 –5% O2 : (a) 4 h; (b) 8 h; (c) 24 h; (d) 48 h;
(e) 120 h.

a trend to interconnecting each other and dark scale with higher Two equivalent circuits were used to quantify the EIS electro-
content of Cr (22.43 wt.%) formed at the bottom of pits, as shown in chemical parameters, which are shown in Fig. 13. In the equivalent
Fig. 11e. Cr should exist as Cr(OH)3 or Cr2 O3 due to low solubility circuits, Rs is solution resistance, Rct is charge transfer resistance,
product of Cr(OH)3 , and Cr2 O3 was the compound from dehydra- which is related to polarization resistance of cathodic and anodic
tion of Cr(OH)3 [3,21,24,25]. This amorphous Cr(OH)3 inhibited the reactions at the electrode surface, RL is inductance resistance, L
development of pits in depth so that the pitting corrosion could is inductance and Cf is corrosion scale capacitance. The porosity
only develop along the horizontal direction, leading to the pits characteristic of the corrosion scale was manifested by the cor-
interconnected together. rosion scale morphologies in Fig. 6, so Rpo is used to represent
corrosion scale resistance in the equivalent circuit. Qdl is constant
3.7. EIS results phase element, which demonstrates the capacitive characteristic
of electric double layer. It has non-integer power dependence on
Fig. 12 shows the EIS plots of 3Cr steel in an aqueous envi- the frequency, but deviates from the ideal capacitance behavior
ronment containing 95% CO2 –5% O2 after 0 h, 4 h, 8 h, 24 h, 48 h due to surface inhomogeneity, roughness and adsorption effects. W
and 120 h of corrosion time. At the initial stage of corrosion (0 h), is Warburg diffusion impedance element, which reflects diffusion
two time constants were identified, a capacitive semicircle in the barrier of corrosive electrolyte through the corrosion scale pores to
high–medium frequency range and an inductive loop in the low reach the electrode surface.
frequency range (Fig. 12a and b). The capacitive semicircle charac- The elements values obtained by fitting of EIS data according to
terized the active state of the interface when the steel was exposed the equivalent circuits are listed in Table 4 and the change curves
to the solution. The inductive loop in Fig. 12a at low frequency of corrosion scale resistance Rpo and charge transfer resistance Rct
range was related to the local adsorption of OH− on the steel surface with the increasing immersion time are shown in Fig. 14, respec-
forming intermediate product FeOHads [26–28]. tively. Fig. 14 indicated the same change trends of Rpo and Rct with
With the increasing corrosion time (from 4 h to 120 h), the immersion time, which were highly consistent with the surface and
low frequency inductive semicircle disappeared, and Nyquist plots cross-section SEM morphologies of 3Cr steel at different immersion
exhibited a capacitive semicircle at high frequency, a capacitive times. Firstly, Rpo and Rct increased from 0 h to 8 h, then decreased
semicircle at middle–low frequency and small diffusion impedance from 8 h to 48 h, and finally increased from 48 h to 120 h. During the
at low frequency (Fig. 12c–f). The high frequency capacitive semi- initial immersion time from 0 h to 8 h, the corrosion scale formed
circle was related to the corrosion scale, the middle–low frequency on the surface of 3Cr steel and with the increase of its compact-
capacitive semicircle was related to cathodic and anodic reactions ness and integrity (Fig. 11a and b), Rpo and Rct increased as a result
at the electrode–electrolyte interface, and the small low frequency (Fig. 14). Then, after the immersion time of 8 h, the integrity of cor-
diffusion impedance was related to corrosive electrolyte diffusion rosion scale decreased due to the corrosion scale damage at local
barrier. areas, where corrosion pits appeared (Fig. 11c and d), leading to
110 X. Lin et al. / Applied Surface Science 329 (2015) 104–115

Fig. 7. EDS and XRD patterns of surface corrosion scale of 3Cr steel in an aqueous environment containing 95% CO2 –5% O2 at 120 h: (a) EDS at position A in Fig. 6e; (b) XRD
for outer layer.

the decrease of Rpo and Rct from 8 to 48 h. Furthermore, from 48 to corrosion scale compact, oxygen molecules of large size were
120 h, new dark corrosion scale with higher Cr content formed at difficult to diffuse through the corrosion scale. Therefore, the
the bottom of the pits under the local damaged areas of corrosion appearance of Warburg impedance should be mainly related to the
scale, which restrained the corrosion process. Therefore, Rpo and Rct barrier effect of Cr-rich corrosion scale on oxygen diffusion [29].
increased again. Nevertheless, the corrosion of 3Cr steel in the CO2 –O2 environment
Moreover, the values of Warburg impedance from 4 h to 120 h was still mainly controlled by electrochemical polarization.
were small and exhibited no obvious change with the immersion
time, as shown in Table 4. It indicated that corrosive electrolyte dif-
fusion through the corrosion scale was a little retarded. From the 4. Development mechanism of corrosion pits
values in Table 4 and EIS plots, the corrosion should be dominantly
controlled by electrochemical polarization and slightly by diffusion. From the above results and analyses, the development mech-
Due to electric neutrality, the main way for Oxygen molecules to anism of corrosion pits of 3Cr steel in an aqueous environment
get to the electrode was diffusion, and for Cr-enrichment made the containing 95% CO2 –5% O2 was further proposed in Fig. 15.
Table 4
Electrochemical parameters fitted from the measured EIS data.

Corrosion time (h) Rs ( cm2 ) RPO ( cm2 ) Cf (F cm−2 ) Rt ( cm2 ) Qdl (−1 sn cm−2 ) Qdl -n W ( s5 cm−2 ) Rl ( cm2 ) L (H cm2 )

0 20.68 43.71 2.72E−3 0.81 122.2 381.6

4 20.92 36.51 1.77E−7 122.3 1.55E−3 0.85 0.19
8 21.36 40.74 5.35E−8 215.1 4.63E−4 0.81 0.18
24 22.10 37.30 1.94E−7 152.4 6.03E−4 0.85 0.12
48 20.61 34.25 2.11E−7 91.9 3.41E−3 0.81 0.25
120 20.26 38.91 4.79E−8 127.4 1.30E−3 0.85 0.16
 X. Lin et al. / Applied Surface Science 329 (2015) 104–115 111

Fig. 8. Microscopic morphologies of inner layer formed on the surface of 3Cr steel in an aqueous environment containing 95% CO2 –5% O2 : (a) 4 h; (b) 8 h; (c) 24 h; (d) 48 h;
(e) 120 h.

Fig. 9. Microscopic morphologies of corrosion scale formed on the surface of 3Cr steel in CO2 saturated formation water (60 ◦ C, CO2 0.95 MPa): (a) surface; (b) cross-section.

Fig. 10. XRD patterns of inner corrosion scale of 3Cr steel in an aqueous environment containing 95% CO2 –5% O2 : (a) 4 h; (b) 120 h.
112 X. Lin et al. / Applied Surface Science 329 (2015) 104–115

Fig. 11. Cross-section microscopic morphologies of corrosion scale formed on the surface of 3Cr steel in an aqueous environment containing 95% CO2 –5% O2 : (a) 4 h; (b) 8 h;
(c) 24 h; (d) 48; (e) 120 h.

Fig. 15a showed the initial forming period of corrosion scale of and corrosion pits appeared. Numerous scholars demonstrated
3Cr steel. During the initial immersion period from 0 h to 8 h, a [21,31–33] that Cr enrichment in corrosion scale greatly increased
thin corrosion scale formed on the steel surface. The inner layer the protection effect of corrosion scale. At local areas, Fe(OH)3
was mainly consisted of two kinds of parts, one part is the mix- formed instead of Cr(OH)3 , leading to nonuniform Cr distribution
ture of FeCO3 and Cr-rich compounds, and the other part was in the corrosion scale. The corrosion scale with low Cr content had
a small amount of Fe(OH)3 . Cr should exist as Cr(OH)3 or Cr2 O3 poor protection so that the local steel surface covered with Fe(OH)3
due to low solubility product of Cr(OH)3 , and Cr2 O3 was the com- was prone to be corroded. Therefore, the local steel surface was cor-
pound from dehydration of Cr(OH)3 [30]. FeCO3 and Cr(OH)3 or roded more severely than other areas, corrosion pits formed and the
Cr2 O3 formed as the main corrosion scale in the inner layer as inner layer was damaged. With the formation of pits below Fe(OH)3 ,
Eqs. (2)–(4). Oxygen in the solution attended cathodic reaction the protection effect of the corrosion scale decreased (Fig. 14) and
and oxidized Fe2+ into Fe3+ at some local areas. Fe(OH)3 could the corrosion rate reduced (Fig. 5).
easily deposit at the local areas due to its pretty low solubility Fig. 15c shows the pits developing period of 3Cr steel. During the
(KspFe(OH)3 = 4.0 × 10−38 ) as Eq. (5) [9]. At local Fe(OH)3 exist- immersion period from 24 h to 48 h, due to the damage of corrosion
ing areas, little Cr(OH)3 could form because the solubility product scale, more porous Fe(OH)3 appeared at the local inner layer of the
of Cr(OH)3 (KspCr(OH)3 = 6.3 × 10−31 ) is much lower than that of damaged corrosion scale on the pits. Because of different protection
Fe(OH)3 . Therefore, element Cr distributed nonuniformly in the effect of corrosion scale, a corrosion cell formed between the dam-
inner layer, which could be found in EDS results (Fig. 8a). With the aged area and the undamaged area. The damaged area with low
thickness increase of the corrosion scale and the diffusion barrier of protection effect was anode, and the undamaged area of mixed
Cr-rich areas, the protection effect of the corrosion scale increased Cr enrichment and FeCO3 with high protection effect was cath-
(Fig. 14) and therefore, the corrosion rate reduced (Fig. 5). ode. The corrosion cell could cause acidification inside the pits and
accelerate pits anode dissolution [34]. The corrosion scale resis-
Fe2+ + CO3 2− → FeCO3 (2)
tance Rpo and charge transfer resistance Rct reduced in the period
3+ + (Fig. 14), and the corrosion rate reduced obviously (Fig. 5), which
Cr + 3H2 O → Cr(OH)3 + 3H (3)
were related to the corrosion scale damage and pits development.
2Cr(OH)3 → Cr2 O3 + 3H2 O (4) Fig. 15d shows the pits inhibiting period of 3Cr steel. During
2+ + the immersion period from 48 h to 120 h, the nearby pits intercon-
4Fe + 10H2 O + O2 → 4Fe(OH)3 + 8H (5)
nected and Cr-rich area appeared at the bottom of the pits, which
Fig. 15b shows the pits forming period of 3Cr steel. During the prohibited the pits development. Cr-rich area would reduce cor-
immersion period from 8 h to 24 h, the inner layer was damaged rosive ions transport through corrosion scale and improved the
 X. Lin et al. / Applied Surface Science 329 (2015) 104–115 113

Fig. 12. EIS plots measured on the 3Cr steel in an aqueous environment containing 95% CO2 –5% O2 within 120 h: (a), (c) and (e) Nyquist diagram; (b), (d) and (f) Bode diagram.

Fig. 13. Electrochemical equivalent circuits for fitting the experimental impedance data of 3Cr steel in an aqueous environment containing 95% CO2 –5% O2 : (a) 0 h; (b)
4–120 h.

cation selectivity of corrosion scale [30]. The corrosion scale could corrosion scale with Cr-rich area could prohibit oxygen from dif-
block anions such as Cl− , CO3 2− , and HCO3 − penetrating through fusing into the pits, and at the same time, oxygen inside the pits
the corrosion scale and reduce the anionic concentration at the could be consumed by cathodic reaction and oxidation reaction of
interface between the corrosion scale and the steel. In addition, the Fe2+ . Cr-rich area at the bottom of pits was analyzed to be caused
114 X. Lin et al. / Applied Surface Science 329 (2015) 104–115

Fig. 14. Curves of corrosion film resistance Rpo and charge transfer resistance Rct of 3Cr steel as a function of immersion time in an aqueous environment containing 95%
CO2 –5% O2 .

Fig. 15. Schematic of the development mechanism of corrosion pits of 3Cr steel in an aqueous environment containing 95% CO2 –5% O2 : (a) corrosion scale forming period;
(b) pits occurring period; (c) pits developing period; (d) pits inhibiting period.

possibly by the decrease of oxygen inside the pits. From the transformed to Fe2 O3 and FeOOH in the corrosion scale. Cr-rich
surface and the cross-section observation of the specimens area formed at the bottom of pits prevented the development of
(Figs. 8e and 11e), there should exist the synergy of two aspects, pits and made them develop along the steel surface. Fe(OH)3 and
one was corrosion protection effect of Cr-rich area at the bottom of Cr-rich areas played an important role on the formation of corrosion
pits and the other was possibly corrosion promotion effect of pH pits and pits development.
value reduction at the top of pits due to hydrolytic action of Fe3+
from oxidation of anodic dissolution product Fe2+ . The Cr-rich area Acknowledgement
at the bottom of the pits prohibited the pits from developing deeply
into the steel and anodic dissolution reaction was prone to occur The work is funded by National Key Technologies R&D Program
at the top of the pits. Consequently, the pits developed along the of Ministry of Science and Technology of the People’s Republic
steel surface and connected with nearby pits. In the period, due to of China under the contract No. 2011BAE25B03 and the authors
the protection of Cr-rich area, corrosion scale resistance Rpo and acknowledge the assistance.
charge transfer resistance Rct manifested the increase with immer-
sion time (Fig. 14), and the reduction of corrosion rate stopped and
References
showed slight change (Fig. 5).
[1] D.S. Carvalho, C.J.B. Joia, O.R. Mattos, Corrosion rate of iron and iron–chromium
alloys in CO2 medium, Corros. Sci. 47 (2005) 2974–2986.
5. Conclusions
[2] C. Chen, M.X. Lu, D.B. Sun, Z.H. Zhang, W. Chang, Effect of chromium on the pit-
ting resistance of oil tube steel in a carbon dioxide corrosion system, Corrosion
The corrosion scale of 3Cr steel forming in an aqueous environ- 61 (2005) 594–601.
ment containing 95% CO2 –5% O2 exhibited a duplex-layer structure. [3] L. Xu, S. Guo, W. Chang, T. Chen, L. Hu, M. Lu, Corrosion of Cr bearing low alloy
pipeline steel in CO2 environment at static and flowing conditions, Appl. Surf.
The loose and porous outer layer consisted of scattering granular Sci. 270 (2013) 395–404.
Fe2 O3 , which provided little corrosion protection. The inner layer [4] M.B. Kerman, D. Edmonds, J.C. Gonzales, G.L. Turconi, L. Scoppio, G. Dicken,
mainly consisted of large areas of FeCO3 and Cr enrichment as well Development of superior corrosion resistance 3% Cr steels for downhole appli-
cations, in: Corrosion’ 2003, NACE, Paper no. 116, 2003.
as small areas of Fe2 O3 and FeOOH, which affected protection of [5] H. Takabe, M. Ueda, Corrosion resistance of low Cr bearing steel in sweet and
corrosion scale obviously. sour environments, in:, in: Corrosion’ 2002, NACE, Paper no. 41, 2002.
In the CO2 –O2 environment, dissolved oxygen damaged the cor- [6] R. Nyborg, A. Dugstad, Mesa corrosion attack in carbon steel and 0.5% chromium
steel, in: Corrosion’ 1998, NACE, Paper no. 29, 1998.
rosion scale integrity of 3Cr steel by precipitation of Fe(OH)3 at [7] B. Kermani, T. Perez, C. Morales, L. Pigliacampo, J.C. Gonzales, G.L. Turconi, Win-
local area, which caused the nonuniform distribution of Cr element, dow of application and operational track record of low carbon 3Cr steel tubular,
leading to pitting corrosion and high corrosion rate. Fe(OH)3 can in: Corrosion’ 2006, NACE, Paper no. 113, 2006.
 X. Lin et al. / Applied Surface Science 329 (2015) 104–115 115

[8] Q. Wu, Z. Zhang, X. Dong, J. Yang, Corrosion behavior of low-alloy steel con- [22] T. Chen, L. Xu, M. Lu, W. Chang, L. Zhang, Study on factors affecting low Cr alloy
taining 1% chromium in CO2 environments, Corros. Sci. 75 (2013) 400–408. steels in a CO2 corrosion system, in: Corrosion’ 2011, NACE, Paper no. 74, 2011.
[9] R.L. Martin, Corrosion consequences of oxygen entry into sweet oilfield fluids, [23] J. Sun, W. Liu, W. Chang, Z. Zhang, Z. Li, T. Yu, M. Lu, Characteristics and for-
in: SPE’ 2001, SPE Permian Basin Oil and Gas Recovery Conference, SPE. 70029, mation mechanism of corrosion scales on low-chromium X65 steels in CO2
2001. environment, Acta Metall. Sin. 45 (2009) 84.
[10] R.L. Martin, Corrosion consequences of oxygen entry into oilfield brines, in: [24] R. Xu, G. Wu, X. Yang, X. Zhang, Z. Wu, G. Sun, G. Li, P.K. Chu, Corrosion behavior
Corrosion’ 2002, NACE, Paper no. 270, 2002. of chromium and oxygen plasma-modified magnesium in sulfate solution and
[11] D. John, B. Kinsella, S. Bailey, R.D. Marco, Flow dependence of carbon dioxide simulated body fluid, Appl. Surf. Sci. 258 (2012) 8273–8278.
corrosion rates and the interference of trace dissolved oxygen, in: Corrosion’ [25] J. Sun, W. Liu, W. Chang, Z. Zhang, Z. Li, T. Yu, M. Lu, Characteristics and for-
2007, NACE, Paper no. 315, 2007. mation mechanism of corrosion scales on low-chromium X65 steels in CO2
[12] B. Yu, D.Y. Li, A. Grondin, Effects of the dissolved oxygen and slurry velocity on environment, Acta Metall. Sin. 45 (2009) 84–90.
erosion–corrosion of carbon steel in aqueous slurries with carbon dioxide and [26] J. Zhang, Z.L. Wang, Z.M. Wang, X. Han, Chemical analysis of the initial corrosion
silica sand, Wear 302 (2013) 1609–1614. layer on pipeline steels in simulated CO2 -enhanced oil recovery brines, Corros.
[13] M. Zenghua, S. Yongtao, L. Haitao, Corrosion and protection of oil pipe Sci. 65 (2012) 397–404.
steel P110 in multicomponent thermal fluid, Chem. Eng. Oil Gas 41 (2012) [27] G.A. Zhang, Y. Zeng, X.P. Guo, F. Jiang, D.Y. Shi, Z.Y. Chen, Electrochemical
583–585. corrosion behavior of carbon steel under dynamic high pressure H2 S/CO2 envi-
[14] Z. Xingang, L. Yuanyuan, Y. Weigang, L. Chao, Z. Jie, Multihot fluid corrosion and ronment, Corros. Sci. 65 (2012) 37–47.
protection of petroleum and natural gas mining and gathering system, Corros. [28] F. Farelas, M. Galicia, B. Brown, S. Nesic, H. Castaneda, Evolution of dissolu-
Prot. 30 (2009) 316–340. tion processes at the interface of carbon steel corroding in a CO2 environment
[15] S. Yongtao, F. Chaoyang, Y. Xiulan, Corrosion analysis of oil pipe in injection studied by EIS, Corros. Sci. 52 (2010) 509–517.
exploitation of high temperature multicomponent thermal fluid, Chem. Eng. [29] L. Hao, S. Zhang, J. Dong, W. Ke, Evolution of atmospheric corrosion of MnCuP
Oil Gas 41 (2012) 408–410. weathering steel in a simulated coastal-industrial atmosphere, Corros. Sci. 59
[16] T.S. Jackson, P.E. Martin, R.S. Posey, Surface response design to study the effect (2012) 270–276.
of low oxygen levels on carbon dioxide corrosion environments, in: Corrosion’ [30] C. Chen, M. Lu, D. Sun, Z. Zhang, W. Chang, Effect of chromium on the pitting
2009, NACE, Paper no. 574, 2009. resistance of oil tube steel in a carbon dioxide corrosion system, Corrosion 61
[17] C. Durr, J.A. Beaver, Effect of oxygen on the internal corrosion of natural gas (2005) 594–601.
pipelines, in: Corrosion’ 1996, NACE, Paper no. 612, 1996. [31] A. Ikeda, M. Ueda, CO2 Corrosion Behaviour of Cr-Containing Steels, European
[18] S. Wang, Effect of Oxygen on CO2 Corrosion of Mild Steel, a Master’s thesis, Ohio Federation of Corrosion, 1994, pp. 59–93.
University, 2009. [32] L. Paolinelli, S. Simison, T. Pérez, The influence of steel microstructure chem-
[19] G. McIntire, J. Lippert, J. Yudelson, The effect of dissolved CO2 and O2 on the ical composition and precorrosion on CO2 corrosion inhibitor efficiency, in:
corrosion of iron, Corrosion 46 (1990) 91–95. Corrosion’ 2007, NACE, Paper no. 311, 2007.
[20] J. Zhang, W. Liu, X. Lin, S. Dong, S. Lu, C. Yang, T. Wang, M. Lu, Corrosion behavior [33] A.M. Huntz, V. Bague, G. Beauplé, C. Haut, C. Sévérac, P. Lecour, X. Longaygue, F.
and mechanism of N80 steel under high temperature and high pressure CO2 –O2 Ropital, Effect of silicon on the oxidation resistance of 9% Cr steels, Appl. Surf.
coexisting condition, in: Corrosion 2013, NACE, Paper no.2479, 2013. Sci. 207 (2003) 255–275.
[21] S. Guo, L. Xu, L. Zhang, W. Chang, M. Lu, Corrosion of alloy steels containing 2% [34] S.A.M. Refaey, Inhibition of chloride pitting corrosion of mild steel by sodium
chromium in CO2 environments, Corros. Sci. 63 (2012) 246–258. gluconate, Appl. Surf. Sci. 157 (2000) 199–206.