The influence of calcareous deposits on hydrogen uptake and embrittlement of API 5CT P110 steel

Corrosion Science 118 (2017) 178–189

Contents lists available at ScienceDirect

Corrosion Science journal homepage: www.elsevier.com/locate/corsci

The influence of calcareous deposits on hydrogen uptake and embrittlement of API 5CT P110 steel Leonardo Simoni ∗ , Júlio Queiroz Caselani, Leandro Brunholi Ramos, Roberto Moreira Schroeder, Célia de Fraga Malfatti PPGE3M/DEMET, Federal University of Rio Grande do Sul − UFRGS,Av. Bento Gonc¸alves, 9500, Porto Alegre, RS, Brazil

a r t i c l e

i n f o

Article history: Received 26 September 2016 Received in revised form 25 January 2017 Accepted 5 February 2017 Available online 8 February 2017 Keywords: A. Low alloy steel B. Hydrogen permeation B. SEM C. Cathodic protection C. Hydrogen embrittlement C. Hydrogen permeation

a b s t r a c t The effect of calcareous deposits on hydrogen uptake and embrittlement of API 5CT P110 steel was investigated using the electrochemical hydrogen permeation and slow strain rate techniques. A deposit with two distinct layers was formed at −1000 mVSCE , comprising an initial Mg-rich layer followed by a Ca-rich layer whereas at −1500 mVSCE the deposit was porous and rich in Mg. The formation of calcareous deposits did not significantly alter hydrogen uptake and embrittlement at −1000 mVSCE whereas at −1500 mVSCE they were increased. A mechanism was proposed to explain the different roles of calcareous deposits in hydrogen absorption and embrittlement. © 2017 Elsevier Ltd. All rights reserved.

1. Introduction

solubility product is exceeded, such as calcium carbonate (3) and magnesium hydroxide (4), can precipitate.

Increasing energy demand has driven the development and exploration of new oil and gas fields around the world, many of which are offshore. Many of the materials used in offshore exploration adopt the cathodic protection technique to minimise or prevent corrosion in a range of structures such as pipelines and well casings [1–3]. When cathodic protection is used, as structures are cathodically polarised, oxygen reduction (1) and water dissociation (2) reactions occur on the metal surface. O2 + 2H2 O + 4e− → 4OH−

(1)

2H2 O + 2e− → H2 + 2OH−

(2)

In either case, hydroxyl ions are produced resulting in increased pH in the solution adjacent to the metal surface. Consequently carbonate ion concentration is enhanced at the metal/solution interface due to the modification of inorganic carbonic equilibrium. Thus, as a result of the higher interfacial pH and higher carbonate ion concentration at the interface, inorganic compounds whose

∗ Corresponding author. E-mail address: [email protected] (L. Simoni). http://dx.doi.org/10.1016/j.corsci.2017.02.007 0010-938X/© 2017 Elsevier Ltd. All rights reserved.

Ca2+ + CO3 2− → CaCO3(ppt)

(3)

Mg2+ + 2OH− → Mg(OH)2(ppt)

(4)

It has been shown that the critical pH for the precipitation is ∼7.5 for CaCO3 and ∼9.5 for Mg(OH)2 (brucite) precipitation [4,5]. Mixed deposits of calcium carbonate and magnesium hydroxide are generally called calcareous deposits. It is known that these deposits decrease the corrosion rate and the current demand of cathodic protection by progressively decreasing the active metal surface and hindering the oxygen transport to the metal surface [6–11]. The composition and morphology of calcareous deposits is dependent on many factors, such as the applied potential, the concentration of dissolved oxygen, the pH, the pressure, the temperature and the flow of the electrolyte, among others [8,12,13]. Barchiche et al. [10] performed X-ray diffraction (XRD) analysis of the outer layers of calcareous deposits formed on rotating steel electrodes in the 10–30 ◦ C range and suggested that for potentials between −900 and −1100 mVSCE the deposits were composed by aragonite whereas a combination of aragonite and brucite was found at −1200 mVSCE . They also observed that the deposit was composed exclusively by brucite at potentials more negative than −1300 mVSCE . However, it has been established that below the aragonite crystals observed

L. Simoni et al. / Corrosion Science 118 (2017) 178–189

on the outer surface there exists a Mg-rich inner layer, presumed to be magnesium hydroxide [4,5,14]. The hydrogen evolution induced by the applied cathodic potential in certain circumstances results in hydrogen embrittlement of the protected material. It is known that hydrogen evolution occurs in stages, and can lead to entry of hydrogen into the metal lattice [15–17]. Once it has been absorbed, hydrogen can be found either in normal interstitial sites or trapped in defects such as grain boundaries, vacancies and interfaces [18]. These hydrogen traps play an important role in hydrogen diffusion and distribution in steels, and are usually classified as irreversible (traps with high activation energy) and reversible (traps with low activation energy). The absorbed hydrogen has a significant impact on a material’s mechanical and microstructural properties. Depending on the conditions and the material, several mechanisms of hydrogen embrittlement have been proposed. The most cited ones are HELP (hydrogen enhanced localised plasticity) and HEDE (hydrogen enhanced decohesion). The former is based on slip softening due to enhanced dislocation movement in some crystallographic planes at crack tip caused by atomic hydrogen [19,20], while the latter supposes that hydrogen decreases the cohesive force between atoms in the metal lattice, at grain boundaries and at interfaces [18,21]. Thus it results in the nucleation of microcracks, resulting in fracture. The relation between calcareous deposit formation, hydrogen uptake and embrittlement is still not clear. Some researchers have studied and discussed the effect of deposit formation on hydrogen absorption using electrochemical hydrogen permeation techniques but there is no consensus concerning their results. After electrochemical hydrogen permeation tests with galvanostatic charging, Chyn Ou and Wu [22] concluded that calcareous deposit formed either by Mg(OH)2 or by CaCO3 decrease the hydrogen uptake. Olsen and Hesjevik [23] also concluded that the presence of calcareous deposits decreased hydrogen entry in supermartensitic stainless steels by a factor of two or three. Turnbull and Hinds [24] reported that the calcareous deposits formed in artificial seawater in permeation tests at −1000 mVSCE decreased the hydrogen uptake on supermartensitic stainless steel in comparison with tests carried out in 3.5% NaCl solution. However the authors concluded that the difference was not significant and should not be relied upon. Recently, results obtained by Smith and Paul [25] also showed a slightly lower hydrogen uptake (measured by thermal desorption) in artificial seawater in comparison with 3.5% NaCl solution at −1100 mVSCE . However, Lucas and Robinson [26] and Hamzah and Robinson [27] observed the opposite behaviour of calcareous deposits, i.e., hydrogen absorption was higher in seawater than in 3.5% NaCl solution at different cathodic potentials (from −850 mVSCE to −1300 mVSCE ). In addition, Zucchi et al. [28] used the slow strain rate technique (SSRT) to study the hydrogen embrittlement of duplex stainless steel. They attributed an increase in elongation in tests performed at −1200 mVSCE compared with −1000 mVSCE and -900 mVSCE to the formation of calcareous deposits on the specimens surfaces. Increases or decreases in elongation in SSRT tests compared with tests performed in air are usually used as a parameter to measure hydrogen embrittlement or stress corrosion cracking susceptibility [28,29]. Thus, these studies indicated a decrease in susceptibility to hydrogen embrittlement caused by the precipitation of calcareous deposit. Despite these published studies, there are still some unclear points in the understanding of this phenomenon, and there remain no focused studies on the calcareous deposit effect on the correlation between hydrogen uptake and hydrogen embrittlement. In addition, there are almost no studies to date concerning the hydrogen effect on API 5CT grade P110 steel, a well casing material that is often used in offshore oil and gas industries. As this material is usually cathodically protected the formation of calcareous

179

Fig 1. Microstructure of API 5CT grade P110 steel composed of tempered martensite, upper bainite and dispersed carbides, observed by optical microscopy.

deposits on its surface is expected. The main objective of the present work is to investigate the effect of calcareous deposits on hydrogen uptake and embrittlement of API 5CT P110 steel. To accomplish this, electrochemical hydrogen permeation tests and SSRT were used in ASTM D1141 artificial seawater, in artificial seawater without calcium and magnesium ions, and in 3.5% NaCl solution. 2. Experimental procedures 2.1. Preparation of specimens and solutions The specimens were cut from an API 5CT grade P110 steel seamless pipe with an external diameter of 170 mm and a thickness of 12.7 mm supplied by Petrobras S.A. The chemical composition of this steel is given in Table 1 and the microstructure (composed by tempered martensite, upper bainite and dispersed carbides) is shown in Fig. 1. The specimens for electrochemical permeation tests were cut by wire electroerosion into samples of 22 × 25 × 1 mm3 . The surfaces were then prepared as follows: 1. Mechanical wet grinding of the surface in contact with the cathodic cell down to 600 grit SiC emery paper. The surface in contact with the anodic cell was wet ground down to 1200 grit SiC emery paper; 2. Electrodeposition of a thin palladium coating (less than 0.1 ␮m thick) onto the surface in contact with the anodic cell. The electrodeposition was carried out in a 28% NH4 OH + 5 g L−1 PdCl2 solution with a cathodic current density of 2 mA cm−2 for 90 s, as described by Manolatos and Jerome [30]. The importance of the palladium coating has previously been discussed in several studies [31–33]; 3. Degreasing with acetone and ethanol. The specimens for SSRT were machined according to NACE TM 198-2004 for subsize specimens [34]. Prior to each test, the surface of the specimens was sequentially wet-ground down to 600 grit SiC emery paper and degreased with acetone and ethanol. Note that the surface preparation for the SSRT specimens and for the cathodic side of the electrochemical hydrogen permeation specimens is the same, in order to have the same finish on the surface for hydrogen entry.

180

L. Simoni et al. / Corrosion Science 118 (2017) 178–189

Table 1 Chemical composition of API 5CT grade P110 steel (wt.%). C

Si

Mn

P

S

Cr

Mo

Ni

Al

Co

Nb

Ti

V

Fe

0.236

0.259

0.957

0.013

0.0025

0.969

0.271

0.018

0.049

0.0026

0.0059

0.0063

0.0056

97.19

2.2. Test solutions Tests were performed in three solutions based on seawater that are shown in Table 2. One of the solutions is ASTM D1141 artificial seawater [35]. The second solution contained the same compounds specified in ASTM D1141 artificial seawater but excluding CaCl2 and MgCl2 . Finally, a 3.5% NaCl solution was used. The choice of these solutions is based on the fact that the presence of magnesium and calcium ions in seawater can lead to the formation of calcareous deposits on specimen surfaces under certain conditions. Thus, the formation of calcareous deposits on the specimen surfaces is expected in tests performed in ASTM D1141 artificial seawater, but not in the other solutions, as they do not contain these ions. 2.3. Electrochemical Electrochemical hydrogen permeation In order to verify the surface effect caused by calcareous deposits on hydrogen absorption in API 5CT P110 steel, electrochemical hydrogen permeation tests were carried out. An electrolytic cell with two compartments was used, as proposed by Devanathan and Stachurski [36]. Both compartments had one saturated calomel electrode (SCE) and one platinum counter electrode. Samples were placed between the two compartments, i.e. between the charging cell (entry of hydrogen) and the oxidation cell (exit of hydrogen). Circular areas of 1.88 cm2 and 0.63 cm2 were exposed to the electrolyte on the charging cell and the oxidation cell, respectively. This area relationship was used in order to decrease the error caused by lateral diffusion, as reported on literature [37,38]. The tests were carried out at room temperature (22.0 ± 1.0 ◦ C). The oxidation cell was filled with approximately 300 mL of 0.1 M NaOH solution deaerated by nitrogen purging for five hours before testing and continuing throughout the test. Prior to hydrogen charging, the membrane was depleted of residual hydrogen by holding the exit surface at a potential of +200 mVSCE until the current density was reduced to 100 nA cm−2 . The charging cell was then filled with one of the solutions shown in Table 2 and a constant cathodic potential was applied. The measurements performed in this work were obtained at two cathodic potentials: −1000 mVSCE and −1500 mVSCE . The former potential is usually applied in the cathodic protection of offshore pipelines. It is believed that the calcareous deposits on steel surfaces formed by polarising steel at this potential in seawater consist of a Mg(OH)2 inner layer covered by CaCO3 crystals [4]. The potential −1500 mVSCE corresponds to an extreme situation of overprotection and to a high hydrogen overpotential, which may result in calcareous deposits mainly formed by Mg(OH)2 [10] and intense hydrogen evolution. The mathematical relationship derived from Fick’s law allows the calculation of the apparent hydrogen diffusion coefficient from the hydrogen permeation test results using the following equation:

Dapp =

L2 Mt

Where Dapp is the apparent diffusion coefficient, L is the specimen thickness and M a constant depending on the time value t chosen in the diffusion transient. The most commonly used values for M are 15.3 and 6. The former corresponds to the elapsed time tb (Time-breakthrough) measured by extrapolating the linear portion of the rising permeation transient and the latter corresponds to the elapsed time tlag (Time-lag) to achieve a value of J(t)/Jss = 0.63. In

the present work the apparent diffusion coefficient was calculated using the time breakthrough methodology. 2.4. SSRT SSRT tests were carried out with a constant crosshead speed of 0.0015 mm min−1 (equivalent to a nominal strain rate of 1 × 10−6 s−1 ) at room temperature (22.0 ± 1.0 ◦ C). The specimen, a saturated calomel electrode (SCE), and a Pt counter electrode were inserted into a cell to perform the tests. The cell was then filled with one of the testing solutions illustrated in Table 2. A constant cathodic potential was applied throughout the test: −1000 mVSCE or −1500 mVSCE . The tests were also carried out in air in order to access the mechanical properties without corrosive environment and applied potential. At the end of the test, the fracture surface was examined by scanning electron microscopy (SEM) in order to verify the fracture mechanisms. Tests carried out in artificial seawater resulted in the formation of calcareous deposits on the specimen surfaces. Thus, after SSRT a section was cut from the specimen and its morphology and chemical composition were evaluated by analysis of the top surface by SEM and EDS, respectively. A further section was embedded in epoxy resin to evaluate its cross section, which was wet ground down to 4000 grit SiC emery paper and polished with a 1.0 ␮m diamond paste. The crystal structures of the calcareous deposits were investigated by XRD using a Philips Diffractometer, model X’Pert MPD with a graphite monochromator and fixed anode operated at 40 kV and 40 mA using Cu-K␣ (␭ = 1.5406) radiation. Susceptibility to hydrogen embrittlement was evaluated by normalised elongation, i.e. the ratio between percentage elongation of the specimen to fracture in the test solution and in air. The lower the normalised elongation, the higher the degree of embrittlement. 3. Results and discussion 3.1. Electrochemical hydrogen permeation Fig. 2(a) shows the average the permeation curves for each solution at −1000 mVSCE . A delay in diffusion is observed in tests carried out in artificial seawater in comparison with other conditions, which is probably associated with the formation of calcareous deposits. Consequently, the apparent diffusion coefficient is lowest at −1000 mVSCE among all tests performed, as shown in Table 3. The average steady state current and the hydrogen average flux are also lower in seawater than in other solutions. However, as shown in Table 3, only a small decrease is observed. Thus, the formation of calcareous deposits under these conditions seems to have an impact on hydrogen uptake even if it is not a large reduction. On the other hand, tests performed in 3.5% NaCl solution showed the highest steady state current and the highest apparent diffusion coefficient. Tests carried out in artificial seawater without Ca2+ and Mg2+ showed intermediate behaviour. It is known that the presence of impurities on the surface can affect the hydrogen uptake in metals due to the occupation of certain adsorption sites and the modification of the adsorption energy [15,16,39]. Thus, the adsorption of other ionic species besides hydrogen such as Mg2+ , Ca2+ , CO3 2− and OH− , and the precipitation of calcareous deposits on the electrode’s surface might be the main mechanism responsible for these differences, since it occurs simultaneously with hydrogen reduction on the surface and decreases

L. Simoni et al. / Corrosion Science 118 (2017) 178–189

181

Table 2 Composition of solutions used in the study. Salt

Artificial Seawater (ASTM D1141-90) pH 8.2

Artificial Seawater without Ca2+ and Mg2+ pH 8.2

Concentration −1

NaCl MgCl2 Na2 SO4 CaCl2 KCl NaHCO3 KBr H2 BO3 SrCl2 NaF

3.5% NaCl solution pH 6.5

Concentration −1

−1

Concentration −1

gL

mol L

gL

mol L

g L−1

mol L−1

24.53 5.20 4.09 1.16 0.695 0.201 0.101 0.027 0.025 0.003

4.20 × 10−1 5.46 × 10−2 2.88 × 10−2 1.05 × 10−2 9.32 × 10−3 2.39 × 10−3 8.49 × 10−4 4.37 × 10−4 1.58 × 10−4 7.14 × 10−5

32.19 0 4.09 0 0.695 0.201 0.101 0.027 0.025 0.003

5.51 × 10−1 0 2.88 × 10−2 0 9.32 × 10−3 2.39 × 10−3 8.49 × 10−4 4.37 × 10−4 1.58 × 10−4 7.14 × 10−5

35.00 0 0 0 0 0 0 0 0 0

5.99 × 10−1 0 0 0 0 0 0 0 0 0

Fig. 2. Hydrogen permeation curves of API 5CT grade P110 steel in different solutions at (a) −1000 mVSCE ; (b) −1500 mVSCE .

the quantity of available adsorption sites and consequently the hydrogen entry. These results are in accordance with the work of Turnbull and Hinds [24] and Smith and Paul [25] in which they record a small decrease in hydrogen uptake in artificial seawater compared with 3.5% NaCl solution. However, the former authors concluded that this difference was not significant. Fig. 2(b) shows the average of the permeation transients for each solution at −1500 mVSCE . The steady state currents for all solutions at this potential were higher than those at −1000 mVSCE , as expected since the hydrogen overpotential is higher at more negative potentials. There was no significant difference in the calculated apparent diffusion coefficient in tests performed at −1500 mVSCE (as shown in Table 3). The steady state current and consequently the hydrogen flux were higher at this potential in the tests carried out in artificial seawater than in artificial seawater without Ca2+ and Mg2+ and in 3.5% NaCl solution. The latter two solutions showed very similar behaviour. It is interesting to observe that at

this potential the steady state current was higher in artificial seawater than in the other solutions. This indicates an opposite effect of calcareous deposit formation on hydrogen uptake at this potential compared with the effect observed at −1000 mVSCE , which seemed to decrease the steady state current. This role inversion is possibly related to competition between the surface effect caused by calcareous deposits and the resultant hydrogen overpotential as a function of the local pH conditions. Surface factors such as the presence of calcareous deposits affect the reactions that occur on surface and can disturb hydrogen adsorption, absorption and evolution [15,16,39]. On the hand, the buffering reaction of the bicarbonate/carbonate system and the precipitation of Mg(OH)2 , which result in the formation of calcareous deposits, tend to buffer the local pH. The buffering process caused by Mg(OH)2 precipitation has been shown to be more effective, provided that its solubility limit is reached [40]. However the two processes in combination provide more efficient

Table 3 Steady state current, hydrogen flux and apparent diffusion coefficient for tests carried out in different solutions. Test Solution

Artificial seawater

3.5% NaCl solution

Artificial seawater without Ca2+ and Mg2+

Artificial seawater

3.5% NaCl solution

Artificial seawater without Ca2+ and Mg2+

Potential [mVSCE ] iss [␮A cm−2 ] Dapp [10−7 cm2 s−1 ] Jss [10−11 mol s−1 cm−2 ]

−1500

−1500

−1500

−1000

−1000

−1000

4.67 ± 0.22

2.68 ± 0.60

2.87 ± 0.55

0.83 ± 0.10

1.15 ± 0.21

1.04 ± 0.09

6.65 ± 1.10

6.56 ± 0.29

5.30 ± 0.77

2.27 ± 0.04

4.68 ± 0.50

3.83 ± 0.25

4.85 ± 0.23

2.78 ± 0.63

2.97 ± 0.57

0.86 ± 0.10

1.20 ± 0.22

1.08 ± 0.09

182

L. Simoni et al. / Corrosion Science 118 (2017) 178–189

Fig. 3. Stress-Strain curves of API 5CT grade P110 steel in different solutions at (a) −1000 mVSCE ; (b) −1500 mVSCE .

Fig. 4. SEM images of the fracture surface after SSRT in air (a) macroscopic view; (b) microscopic view of region A (arrows indicate dimples).

Fig. 5. SEM images of the fracture surface after SSRT in artificial seawater at −1000 mVSCE (a) macroscopic view; (b) microscopic view at region A.

local buffering of the solution than either process individually. The cathodic reactions on the working electrode surface result in local pH enhancement. However, if the above-mentioned buffering reactions occur, a less alkaline local pH is expected. Thus, as the hydrogen overpotential decreases with increasing pH, a higher overpotential is expected in artificial seawater than in 3.5% NaCl solution due to the buffering reactions that occur in artificial seawater and lead to a less alkaline local pH. The difference in hydrogen

overpotential between artificial seawater and 3.5% NaCl solution has already been emphasised by Hartt [41] in the discussion with Turnbull and Hinds [24,41]: the local pH in the NaCl 3.5% solution should be one or more units higher than in artificial seawater. Turnbull and Ferris [40] verified that the crack tip pH on fatigue cracks in artificial seawater at −1000 mVSCE is about 10.7, while it is about 12.1 in 3.5% NaCl solution under the same conditions. A similar behaviour to that in 3.5% NaCl solution is expected in

L. Simoni et al. / Corrosion Science 118 (2017) 178–189

183

Fig. 6. SEM images of the fracture surface after SSRT in artificial seawater without calcium and magnesium ions at −1000 mVSCE (a) macroscopic view; (b) microscopic view at region A.

Fig. 7. SEM images of the fracture surface after SSRT in 3.5% NaCl solution at -1000 mVSCE (a) macroscopic view; (b) microscopic view of region A.

artificial seawater without Ca2+ and Mg2+ , since the buffering reaction of the bicarbonate/carbonate system is only partially effective, as reported by Turnbull and Ferris [40]. Thus, the increase in the steady state hydrogen flux in artificial seawater can be related to the less alkaline local pH, and consequently to a higher hydrogen overpotential in comparison with the other solutions. It has already been shown that changes in the solution pH have great influence on stress corrosion cracking and hydrogen embrittlement phenomena [42], possibly related to the fact that at higher pH OH− ions compete with H+ at metal surface adsorption sites [43]. The surface effects caused by calcareous deposit formation at −1000 mVSCE and −1500 mVSCE might not be same, since there are significant differences in the deposits formed at each potential, as discussed further below. It seems that at −1000 mVSCE the surface effect caused by calcareous deposit precipitation compensates the effect caused by the greater hydrogen overpotential associated with the lower local pH in artificial seawater, thus resulting in lower hydrogen uptake in comparison with the other solutions. However, at −1500 mVSCE the hydrogen overpotential is so important that the surface effect caused by the calcareous deposits is not so significant in view of the difference in hydrogen overpotential between the solutions (due to buffering reactions that occur in artificial seawater). Besides that, throughout the tests at −1500 mVSCE , it was observed that calcareous deposits on the electrode surface was continuously detached due to the formation of hydrogen bubbles.

The proposed mechanism can help understanding the divergent results recorded in the literature about the role of calcareous deposits on hydrogen uptake, as mentioned in the introduction. The differences are probably associated with the different conditions in which such deposits were formed, as it is known that these deposits are complex and that many factors influence their formation and consequently their chemical composition, morphology and properties: dissolved oxygen, temperature, salinity, pH, substrate, applied potential, flow conditions, and pressure, among others [12,13]. Thus, if these deposits show different characteristics it is expected that the surface effects might not be same and that competition with the hydrogen overpotential may result in different behaviours. 3.2. SSRT Fig. 3(a) and (b) shows stress-strain curves, obtained by SSRT, performed in artificial seawater, in artificial seawater without Ca2+ and Mg2+ , and in 3.5% NaCl solution at −1000 mVSCE and at −1500 mVSCE , respectively. The elongation of tensile tests performed in air was about 12%, which is larger than all elongations obtained in the presence of the solutions and with applied cathodic potentials (as shown in Fig. 3(a) and (b)). The decrease in ductility of the tested specimens under these conditions is mainly related to the effect of hydrogen on their mechanical properties. There is no sig-

184

L. Simoni et al. / Corrosion Science 118 (2017) 178–189

Fig. 8. SEM and EDS element mapping top analysis of calcareous deposits formed in artificial seawater at −1000 mVSCE on the SSRT specimen surface (corresponding to approximately 36 h of formation).

nificant difference between the stress-strain curves obtained in the different solutions at −1000 mVSCE (Fig. 3(a)). This indicates that calcareous deposit formation does not have significant influence on hydrogen embrittlement of the studied material. Fig. 4 shows the fracture surface of specimens tested in air: macroscopically a typical ductile fracture is observed with an inner fibrous zone, an intermediate radial zone and an outer shear zone. On the microscopic level there is evidence of ductile fracture micromechanisms in the form of microvoid coalescence dimples. The fracture surfaces of specimens tested in artificial seawater, in artificial seawater without Ca2+ and Mg2+ and in 3.5% NaCl solution at −1000 mVSCE showed different fracture micromechanisms (Figs. 5–7 ). The fracture surfaces of specimens tested under all three conditions are very similar: cracks are nucleated on the specimen surface and are propagated in a brittle manner up to a certain distance. Evidence of quasi-cleavage micromechanism and secondary cracks is observed in this brittle propagation region. The secondary cracks observed are believed to be related to glide-plane decohesion, which occurs along tempered martensite lathes [44]. After the brittle propagation, a ductile fracture mode gradually took place leading to the final fracture of the specimens. In this region there was evidence of shallow shear dimples. SEM/EDS analysis of calcareous deposits formed on SSRT specimens in artificial seawater at −1000 mVSCE (which corresponds to approximately 36 h of formation) is shown in Figs. 8 and 9 (Top analysis) and in Fig. 10 (Cross-section analysis). An Mg-rich inner layer and a Ca-rich outer layer are observed. In calcareous deposits formed in Mg2+ containing solutions, calcium carbonate precipitates exclusively in the form of aragonite, one of the allotropic forms of CaCO3 . Aragonite has a needle-like morphology, as shown in Fig. 9. XRD measurements (Fig. 11) confirmed that the Mg-rich inner layer consists of Mg(OH)2 (Brucite) and that the Ca-rich outer layer consists of aragonite (CaCO3 ). The top (Fig. 9) and cross-

Fig. 9. Morphology of calcareous deposits formed in artificial seawater at −1000 mVSCE on the SSRT specimen surface (corresponding to approximately 36 h of formation): aragonite crystals nucleate on an Mg-rich inner layer.

sectional analysis (Fig. 10) revealed aragonite needles on top of the Mg(OH)2 inner layer. The distinction between these two layers is clearly observed in the cross-sectional analysis, consisting of a brucite layer of about 2.2 ␮m that grows at the steel/solution interface and aragonite needles nucleating and growing above this inner layer; in the tested conditions (approximately 36 h until specimen rupture) these crystals were approximately 20 ␮m thick and did not completely cover the steel surface. Tests performed at −1500 mVSCE showed a decrease in elongation in comparison with tests at −1000 mVSCE . However, at this potential, there was a significant difference in elongation between the tests carried out in different solutions: tests performed in

L. Simoni et al. / Corrosion Science 118 (2017) 178–189

185

Fig. 10. SEM and EDS element line cross-sectional analysis of calcareous deposits formed in artificial seawater at −1000 mVSCE on the SSRT specimen surface (corresponding to approximately 36 h of formation).

Fig. 11. XRD analysis (␭-Cu-K␣ radiation) of calcareous deposits formed in artificial seawater at −1000 mVSCE (corresponding to approximately 36 h of formation). A shows the diffraction lines of aragonite, B those of brucite and a-Fe indicates those of the steel substrate.

artificial seawater resulted in higher embrittlement than in tests performed in seawater without Ca2+ and Mg2+ , or in 3.5% NaCl solution. There was no significant difference in elongation between the latter two solutions.

Fractography analyses of specimens tested in artificial seawater, in artificial seawater without Ca2+ and Mg2+ and in 3.5% NaCl solution are shown in Figs. 12–14 . Under all conditions surface cracks were initiated and propagated in the same way as those observed at −1000 mVSCE . However there were some differences in the extent of brittle propagation regions, which were greater at −1500 mVSCE than at −1000 mVSCE. In general, the fracture surfaces of specimens tested at −1500 mVSCE showed transgranular propagation by quasi-cleavage micromechanisms, with secondary cracks between martensite lathes. Additionally, particularly in artificial seawater at −1500 mVSCE , some intergranular facets were observed in these regions (Fig. 12). In general, fracture surfaces in different solutions and at different potentials show similar characteristics, with transgranular propagation of cracks nucleated on specimen surfaces. These fractures are very different from the ductile fractures observed in tests performed in air. Indeed, hydrogen can induce a change of fracture micromechanism from microvoid coalescence to quasicleavage/cleavage or intergranular [44]. SEM/EDS analysis of calcareous deposits formed on SSRT specimens in artificial seawater at −1500 mVSCE (which corresponds to approximately 24 h of formation) is shown in Figs. 15 and 16 (Top analysis) and in Fig. 17 (Cross-sectional analysis). The top surface analysis of the calcareous deposits shows an irregular and porous deposit mainly composed of magnesium (Fig. 15). The detection of

Fig. 12. SEM images of the fracture surface after SSRT in artificial seawater at −1500 mVSCE (a) macroscopic view; (b) microscopic view of region A.

186

L. Simoni et al. / Corrosion Science 118 (2017) 178–189

Fig. 13. SEM images of the fracture surface after SSRT in artificial seawater without calcium and magnesium ions at −1500 mVSCE (a) macroscopic view; (b) microscopic view of region A.

Fig. 14. SEM images of the fracture surface after SSRT in 3.5% NaCl solution at −1500 mVSCE (a) macroscopic view; (b) microscopic view of region A.

iron by element mapping is believed to be due to the penetration of the electron beam to the steel substrate and not necessarily to the presence of iron in the deposits. Deposits formed only by Mg(OH)2 have already been reported in the work of Barchiche et al. in which they relate that at potentials more negative than −1300 mVSCE the calcareous deposits consist only of Mg(OH)2 [10]. At higher magnification (Fig. 16(a)), the powdery and porous characteristic of the deposit is clearly observed. The high intensity of hydrogen bubble formation at this potential was detrimental to deposit adherence; some parts of the deposit were detached from the specimen surface. Their morphology was analysed (Fig. 16(b)) and it was found to be very similar to the deposits that remained on the steel surface. XRD analysis of calcareous deposits detached from the specimen surface and of calcareous deposits that remained on the surface (which had been carefully removed from it) is shown in Fig. 18. It confirms that the deposit is mainly composed by Brucite (the presence of NaCl is related to its crystallisation from the electrolyte). The cross-sectional analysis of the deposit formed at −1500 mVSCE shows an irregular thickness (Fig. 17). Beyond that, it is not as clearly identified as the deposit formed at −1000 mVSCE . This is caused by the penetration of the resin into deposit pores, which hampers its observation. Fig. 19 shows the correlation between the hydrogen embrittlement susceptibility of API 5CT P110 steel (obtained by SSRT and evaluated by normalised elongation) and the steady state hydro-

gen flux (obtained by Electrochemical Hydrogen Permeation). An inverse linear proportionality between normalised elongation and steady state hydrogen flux is clearly observed. This provides evidence for the strict correlation between diffusible hydrogen amount and hydrogen embrittlement susceptibility that has already been shown in studies on other steels [29,45–47]. 4. Conclusions The effect of calcareous deposits on hydrogen embrittlement of API 5CT grade P110 steel was investigated using electrochemical hydrogen permeation and SSRT. The calcareous deposits formed during tensile tests and specimen fracture surfaces were analysed using SEM/EDS and XRD. The results can be summarised as follows: 1. Competition between the surface effect caused by calcareous deposits and the hydrogen overpotential is mainly responsible for the effect of calcareous deposits on hydrogen uptake and hydrogen embrittlement; 2. The formation of calcareous deposits in artificial seawater at −1000 mVSCE causes a modest decrease in the steady state hydrogen flux compared with other solutions at the same potential. However, it does not statistically alter the normalised elongation obtained by SSRT tests;

L. Simoni et al. / Corrosion Science 118 (2017) 178–189

187

Fig. 15. SEM and EDS element mapping top analysis of calcareous deposits formed in artificial seawater at −1500 mVSCE on the SSRT specimen surface (corresponding to approximately 24 h of formation).

Fig. 16. Morphology of calcareous deposits formed in artificial seawater at −1500 mVSCE (corresponding to approximately 24 h of formation): (a) on the SSRT specimen surface; (b) detached from the specimen surface.

Fig. 17. SEM and EDS element line cross-sectional analysis of calcareous deposits formed in artificial seawater at −1500 mVSCE on the SSRT specimen surface (corresponding to approximately 24 h of formation).

188

L. Simoni et al. / Corrosion Science 118 (2017) 178–189

Fig. 18. XRD analysis (␭-Cu-K␣ radiation) of calcareous deposits formed in artificial seawater at −1500 mVSCE (corresponding to approximately 24 h of formation). B indicates the diffraction lines of brucite and N indicates those of NaCl.

Fig. 19. Normalised elongation as a function of steady state hydrogen flux for the different tested conditions.

3. The steady state hydrogen flux in artificial seawater at −1500 mVSCE is higher than in other solutions at the same potential and it is responsible for the higher degree of embrittlement observed; 4. Normalised elongation was found to have an inverse linear proportionality to steady state hydrogen flux; 5. Calcareous deposits formed at −1000 mVSCE consist of an Mg(OH)2 inner layer and an outer layer of aragonite needles. However, the deposit formed at −1500 mVSCE is porous, powdery and mainly composed of Mg(OH)2 . Acknowledgements Authors would like to thank the National Council for Scientific and Technological Development (CNPq-Brazil) for the financial support and Petrobras S.A. for the supplied material. References [1] M.E. Parker, E.G. Peattie, Pipe Line Corrosion and Cathodic Protection: A Practical Manual for Corrosion Engineers, Technicians, and Field Personnel, Gulf Publishing Company, 1984.

[2] S. Lorenzi, T. Pastore, T. Bellezze, R. Fratesi, Cathodic protection modelling of a propeller shaft, Corros. Sci. 108 (2016) 36–46, http://dx.doi.org/10.1016/j. corsci.2016.02.035. [3] W.H. Hartt, Frank Newman Speller Award: cathodic protection of offshore structures-history and current status, Corrosion 68 (12) (2012) 1063–1075, http://dx.doi.org/10.5006/0010-9312-68.12.1063. [4] A. Neville, A.P. Morizot, Calcareous scales formed by cathodic protection – an assessment of characteristics and kinetics, J. Cryst. Growth 243 (2002) 490–502, http://dx.doi.org/10.1016/S0022-0248(02)01532-4. [5] W.H. Hartt, S.W. Smith, J.S. Luo, R.U. Lee, T.Y. Chen, Chemistry – structure – property interrelationships for calcareous deposits as stand alone coatings, Off. Nav. Res. – Anu. Rep. (1988) 1–21, http://dx.doi.org/10.1177/ 002248717002100212. [6] K.E. Mantel, W.H. Hartt, T.Y. Chen, Substrate, surface finish, and flow rate influences on calcareous deposit structure, Corrosion 48 (1992) 489–500, http://dx.doi.org/10.5006/1.3315965. [7] W.H. Hartt, C.H. Culberson, S.W. Smith, Calcareous deposits on metal surfaces in seawater—a critical review, Corrosion 40 (1984) 609–618, http://dx.doi. org/10.5006/1.3581927. [8] J.S. Luo, R.U. Lee, T.Y. Chen, W.H. Hartt, S.W. Smith, Formation of calcareous deposits under different modes of cathodic polarization, Corrosion 47 (1991) 189–196. [9] S. Elbeik, A.C.C. Tseung, A.L. Mackay, The formation of calcareous deposits during the corrosion of mild steel in sea water, Corros. Sci. 26 (1986) 669–680, http://dx.doi.org/10.1016/0010-938X(86)90032-6. [10] C. Barchiche, C. Deslouis, D. Festy, O. Gil, P. Refait, S. Touzain, B. Tribollet, Characterization of calcareous deposits in artificial seawater by impedance techniques: 3 – deposit of CaCO3 in the presence of Mg(II), Electrochim. Acta 48 (2003) 1645–1654, http://dx.doi.org/10.1016/S0013-4686(03)00075-6.

L. Simoni et al. / Corrosion Science 118 (2017) 178–189 [11] C. Rousseau, F. Baraud, L. Leleyter, M. Jeannin, O. Gil, Calcareous deposit formed under cathodic protection in the presence of natural marine sediments: a 12 month experiment, Corros. Sci. 52 (2010) 2206–2218, http:// dx.doi.org/10.1016/j.corsci.2010.02.047. [12] R.H. Heidersbach, R. Baxter, J.S. Smart, M. Haroun, Cathodic Protection, in Corrosion, ASM Metals Handbook, 13, ASM International, 1992, pp. 2440–2450. [13] D.A. Shifler, Understanding material interactions in marine environments to promote extended structural life, Corros. Sci. 47 (2005) 2335–2352, http://dx. doi.org/10.1016/j.corsci.2004.09.027. [14] C. Barchiche, C. Deslouis, O. Gil, P. Refait, B. Tribollet, Characterisation of calcareous deposits by electrochemical methods: role of sulphates, calcium concentration and temperature, Electrochim. Acta 49 (2004) 2833–2839, http://dx.doi.org/10.1016/j.electacta.2004.01.067. [15] J.O. Bockris, E.C. Potter, The mechanism of the cathodic hydrogen evolution reaction, J. Electrochem. Soc. 99 (1952) 169, http://dx.doi.org/10.1149/1. 2779692. [16] J.O. Bockris, H. Mauser, The kinetics of the evolution and dissolution of hydrogen at electrodes, Can. J. Chem. 37 (1959) 475–488. [17] C.D. Kim, B.E. Wilde, The kinetics of hydrogen absorption into iron during cathodic hydrogen evolution, J. Electrochem. Soc. 118 (1971) 202–206, http:// dx.doi.org/10.1149/1.2407967. [18] R.A. Oriani, Hydrogen embrittlement of steels, Annu. Rev. Mater. Sci. 8 (1978) 327–357. [19] H.K. Birnbaum, P. Sofronis, Hydrogen-enhanced localized plasticity-a mechanism for hydrogen-related fracture, Mater. Sci. Eng. A 176 (1994) 191–202, http://dx.doi.org/10.1016/0921-5093(94)90975-X. [20] C.D. Beachem, A new model for hydrogen-assisted cracking (hydrogen embrittlement), Metall. Trans. 3 (1972) 441–455, http://dx.doi.org/10.1007/ BF02642048. [21] S. Lynch, Hydrogen embrittlement phenomena and mechanisms, Corros. Rev. 30 (2012) 105–123, http://dx.doi.org/10.1515/corrrev-2012-0502. [22] K.C. Ou, J.K. Wu, Effect of calcareous deposits formation on the hydrogen absorption of steel, Mater. Chem. Phys. 48 (1997) 52–55, http://dx.doi.org/10. 1016/S0254-0584(97)80076-4. [23] S. Olsen, S.M. Hesjevik, Hydrogen embrittlement from CP on supermartensitic stainless steels – recommendations for new qualification methods, (2004) Corrosion paper no. 4546. [24] G. Hinds, A. Turnbull, Technical note: does calcareous scale formation on cathodically protected steel affect hydrogen uptake? Corrosion 61 (2005) 835–837, http://dx.doi.org/10.5006/1.3280652. [25] W. Smith, S. Paul, Natural deposit coatings on steel during cathodic protection and hydrogen ingress, Coatings 5 (2015) 816–829, http://dx.doi.org/10.3390/ coatings5040816. [26] K.A. Lucas, M.J. Robinson, The influence of lattice hydrogen content on the hydrogen-assisted cracking of high strength steel, Corros. Sci. 26 (1986) 705–717, http://dx.doi.org/10.1016/0010-938X(86)90034-X. [27] R.B. Hamzah, M.J. Robinson, Hydrogen absorption resulting from the simulated pitting corrosion of carbon-manganese steels, Corros. Sci. 27 (1987) 971–979, http://dx.doi.org/10.1016/0010-938X(87)90063-1. [28] F. Zucchi, V. Grassi, C. Monticelli, G. Trabanelli, Hydrogen embrittlement of duplex stainless steel under cathodic protection in acidic artificial sea water in the presence of sulphide ions, Corros. Sci. 48 (2006) 522–530, http://dx.doi. org/10.1016/j.corsci.2005.01.004. [29] T. Depover, K. Verbeken, The effect of TiC on the hydrogen induced ductility loss and trapping behavior of Fe-C-Ti alloys, Corros. Sci. (2016), http://dx.doi. org/10.1016/j.corsci.2016.07.013 (in press). [30] P. Manolatos, M. Jerome, A thin palladium coating on iron for hydrogen permeation studies, Electrochim. Acta 41 (1996) 359–365, http://dx.doi.org/ 10.1016/0013-4686(95)00379-7.

189

[31] P. Manolatos, C. Duret-Thual, M. Le Coze, E. Bollinger, The electrochemical permeation of hydrogen in steels without palladium coating. Part I: interpretation difficulties, Corros. Sci. 37 (1995) 1785–1796, http://dx.doi.org/ 10.1016/0010-938X(95)00080-4. [32] P. Manolatos, C. Duret-Thual, J. Le Coze, M. Jerome, E. Bollinger, The electrochemical permeation of hydrogen in steels without palladium coating. Part II: study of the influence of microstructure on hydrogen diffusion, Corros. Sci. 37 (1995) 1785–1796, http://dx.doi.org/10.1016/0010-938X(95)00080-4. [33] M.A.V. Devanathan, Z. Stachurski, The mechanism of hydrogen evolution on iron in acid solutions by determination of permeation rates, J. Electrochem. Soc. 111 (1964) 619–623, http://dx.doi.org/10.1149/1.2426195. [34] NACE, NACE TM198: 2004 – Slow Strain Rate Test Method for Screening Corrosion-resistant Alloys (CRAs) for Stress Corrosion Cracking in Sour Oilfield Service, 1998, pp. 1–20. [35] ASTM, Standard practice for the preparation of substitute ocean water – D 1141-93 (Reapproved 2003), Am. Natl. Stand. 98 (2003) 98–100, http://dx.doi. org/10.1520/D1141-98R13. [36] M.A.V. Devanathan, Z. Stachurski, The adsorption and diffusion of electrolytic hydrogen in palladium, Proc. R. Soc. London (1962) 90–102. [37] R.B. Hutchings, D.H. Ferriss, A. Turnbull, Ratio of specimen thickness to detection area for reliable hydrogen permeation measurement, Natl. Phys. Lab. (1993) 1–7. [38] A. Traidia, A.M. El-Sherik, H. Attar, Recommended specimen dimensions and boundary conditions for measurement of hydrogen permeation in thick carbon steel plates, Corrosion 71 (2015) 585–597, http://dx.doi.org/10.5006/ 1327. [39] R.N. Iyer, Analysis of hydrogen evolution and entry into metals for the discharge-recombination process, J. Electrochem. Soc. 136 (1989) 2463, http://dx.doi.org/10.1149/1.2097429. [40] A. Turnbull, D.H. Ferriss, Mathematical modelling of the electrochemistry in corrosion fatigue cracks in structural steel cathodically protected in sea water, Corros. Sci. 26 (1986) 601–628, http://dx.doi.org/10.1016/0010938X(86)90027-2. [41] W.H. Hartt, Discussion: does calcareous scale formation on cathodically protected steel affect hydrogen uptake? Corrosion 62 (2006) 947–949, http:// dx.doi.org/10.5006/1.3278232. [42] M. Bonis, J.L. Crolet, Practical aspects of the influence of in situ pH on H2S-induced cracking, Corros. Sci. 27 (1987) 1059–1070, http://dx.doi.org/10. 1016/0010-938X(87)90098-9. [43] Z.A. Foroulis, Electrochemical behavior and corrosion of iron in aqueous sulfidic solution, Mater. Corros. Korros. 31 (1980) 463–470, http://dx.doi.org/ 10.1002/maco.19800310606. [44] B. Somerday, P. Sofronis, R. Jones, Effects of Hydrogen on Materials: Proceedings of the 2008 International Hydrogen Conference (2008). [45] S. Pillot, L. Coudreuse, 2 – Hydrogen-induced disbonding and embrittlement of steels used in petrochemical refining, in: Gaseous Hydrogen Embrittlement of Materials in Energy Technologies, Woodhead Publ. Ser. Met. Surf. Eng., Woodhead Publishing, 2012, pp. 51–93, http://dx.doi.org/10.1533/ 9780857093899.1.51. [46] Y.F. Cheng, Analysis of electrochemical hydrogen permeation through X-65 pipeline steel and its implications on pipeline stress corrosion cracking, Int. J. Hydrogen Energy 32 (2007) 1269–1276, http://dx.doi.org/10.1016/j.ijhydene. 2006.07.018. [47] F. Zucchi, G. Trabanelli, C. Monticelli, V. Grassi, SCC inhibition of a C-steel in acidic 5% NaCl solutions in the presence of thiosulphate, Corros. Sci. 42 (2000) 505–515, http://dx.doi.org/10.1016/S0010-938X(99)00104-3.