Stable Charging Conditions for Low Hydrogen Concentrations in Steels

Transcription

1 , pp Stable Charging Conditions for Low Hydrogen Concentrations in Steels Takuya HARA Kimitsu R&D lab., Nippon Steel Corporation, 1 Kimitsu, Kimistu, Chiba, Japan. hara.takuya@nsc.co.jp (Received on July 15, 2011; accepted on November 4, 2011) The time dependence of hydrogen concentrations in steel was investigated using immersion, galvanostatic-charging, and potentiostatic-charging methods in various environments in order to establish a method for stable hydrogen entry into steel. Stable hydrogen charging conditions can be obtained under galvanostatic-charging in a deaerated buffer solution of a ph more than 5.0. Hydrogen can also stably enter into steel in immersion tests such as those in the environment of a ph more than 5.0 with poisonous substances such as H 2S and ammonium thiocyanate. A different hydrogen concentration can be accurately obtained by an immersion test with different concentrations of poisonous substances such as H 2S or ammonium thiocyanate or with a different ph, and it is also obtained by applying the different low current densities under galvanostatic-charging. KEY WORDS: hydrogen; hydrogen entry; hydrogen permeation; immersion; cathodic-charging. 1. Introduction Hydrogen atoms enter into steel due to corrosion in immersed environments. In such cases, the cathodic reactions on the steel surface consist of hydrogen evolution reactions. The loss of the positive charge of oxonium ions results in the formation of adsorbed hydrogen atoms on the steel surface. This reaction is referred to as a Volmer reaction. Two adsorbed hydrogen atoms are combined and a hydrogen molecule is formed on the steel surface. This reaction is referred to as a Tafel reaction. Most of the adsorbed hydrogen atoms are emitted from the surface as hydrogen gas molecules, but a part of them are absorbed into steel. Generally, the hydrogen concentration in steel is greatly affected by the environment. Hydrogen concentrations in steel increase with the decrease of ph. 1) Hydrogen concentration in steel is enhanced in environments where hydrogen sulfide (H 2S), cyanogens (CN), and arsenic (As) etc., which are also known as poisonous substances, exist together. It is reported that these substances, otherwise known as siteblocking elements, suppress the reaction rate of Tafel and Volmer reactions through interaction between the electron density of the steel and these substances. 2,3) Hydrogen embrittlement, such as delayed fracture, may occur when hydrogen concentrations in steel increase. Hydrogen embrittlement sometimes occurs under stress less than yield strength, and as a result, a serious accident may occur. Therefore, it is necessary to reduce the entry of hydrogen into steel, to apply material with low susceptibility to hydrogen embrittlement, or to reduce the applied stress, such as residual stress, in order to prevent hydrogen embrittlement. Hydrogen embrittlement occurs if hydrogen concentrations in steel exceed the threshold hydrogen concentration for hydrogen embrittlement. Based on this concept, substantial research has been carried out regarding hydrogen embrittlement. 4 6) Thus, the susceptibility of hydrogen embrittlement has been evaluated, comparing the hydrogen concentrations in steel and the threshold hydrogen concentration of delayed fracture. In this study, the hydrogen entry behavior into steel is studied using immersion, galvanostatic-charging, and potentiostatic-charging methods. Hydrogen concentrations in steel were measured using the thermal desorption analysis after stable continuous hydrogen charging over certain durations. The methods to evaluate the threshold hydrogen concentration for delayed fracture is proposed using a tensile tests, such as a constant load tensile test after hydrogencharging and cadmium plating. 6) On the other hand, the crack susceptibility of steel to hydrogen embrittlement is sometimes evaluated by a constant load tensile test during continuous-charging. 7) In this case, stable hydrogen entry into steel is required for hydrogen embrittlement to occur from the surface. The accurate threshold hydrogen concentration of delayed fracture is not possible to be measured if hydrogen concentrations in steel scatter or fluctuate at the surface. Hydrogen concentrations in steel can be obtained accurately if the stable hydrogen charging condition of continuous-charging can be established. The threshold hydrogen concentration in steel of delayed fracture can also be measured accurately after stable hydrogen entry into steel during continuous-charging. In high strength steel, the threshold hydrogen concentration of delayed fracture is low, and it is commonly less than 0.1 ppm or lower. Therefore, it is very important to establish a stable hydrogen charging method for low hydrogen concentrations in order to evaluate the hydrogen embrittlement property of high strength steel. 8) In order to establish a hydrogen charging condition by which small amounts of hydrogen stably enter into steel, hydrogen entry behavior was investigated using the hydro ISIJ 286

2 gen permeation method under the immersion, galvanostaticcharging, and potentiostatic-charging methods in various environments. 2. Experimental Procedure 2.1. Materials Table 1 shows the chemical compositions of the tested steel specimens. For the tests, 0.04% carbon steel (X65-1, X65-2, and X120), 0.06% carbon steel (X100), and 0.25% carbon steel (VC-1, VC-2, and D) were used. The tensile strengths of these steel specimens are also shown in Table 1. The tensile strengths of the X65-1 and X65-2 steel specimens correspond to that of API (American Petroleum Institute) grade X65, which is more than 535 MPa, while the tensile strengths of the X100 and X120 steel specimens correspond to those of API grade X100 and X120, which are more than 760 MPa and more than 915 MPa, respectively. All steel specimens were melted in a 300-ton LD convertor and then hot-rolled into plates. The microstructure of the X65, X100, and X120 steel specimens is dominantly bainite. The VC steel specimens contain 0.25% carbon and 1.0% vanadium. These steel specimens are melted in a 300-kg VIM in a laboratory and rolled into plates. The VC-1 steel specimens are heat-treated at C (1 323 K) and tempered at 350 C (623 K), while the VC-2 steel specimens are heat-treated at C (1 323 K) and tempered at 650 C (923 K). The tensile strength of the VC-1 and VC-2 steel specimens is MPa. The microstructure consists of tempered martensite. Steel D with a tensile strength of 620 MPa is commercially purchased. The microstructure is ferrite and pearlite. The specimens subjected to the hydrogen permeation method were taken from the quarter portion of these seven plates. These specimens, with 20 mm in length, 50 mm in width, and 1 mm in thickness, were cut, machined, and electro-polished. The hydrogen charging conditions of the immersion, galvanostatic-charging, and potentiostatic-charging methods were summarized in Table 2. When galvanostatic-charging was used, two kinds of deaerated solutions were used. A liter of a buffer solution concocted of 1 mol/l acetic acid and 1 mol/l sodium acetate solutions containing 3 mass% NaCl was adjusted to ph 5.4 with 3 g/l of NH 4SCN. Several applied current densities were selected from 0.02 to 0.5 ma/cm 2 (0.2 to 5.0 A/m 2 ). The second test solution filled in the left side consisted of a deaerated sodium hydroxide solution containing 3 mass% NaCl with a ph of Several applied current densities were selected from 0.01 to 0.1 ma/cm 2 (0.1 to 1.0 A/m 2 ). When the potentiostatic-charging method was applied, a deaerated 1 mol/l acetic buffer solution containing 3 mass% NaCl and 3 g/l of NH 4SCN was used. Applied potential was selected either 0.75 V or 0.80 V SCE (Saturated Calomel Electrode). When an immersion test was carried out, two kinds of solutions were used. The first test solution was consisted of a 1 mol/l acetic buffer solution containing 5 mass% NaCl, saturated with H 2S under partial pressure of 0.1 MPa, of which the ph was adjusted to 3.0, 5.0, or 5.3. The second test solution was consisted of a deaerated 1 mol/l acetic buffer solution at a ph of 5.4 containing 3 mass% NaCl with the addition of 0.3 g/l, 3 g/l, 10 g/l, and 50 g/l NH 4SCN. Deaeration by bubbling with N 2 gas was continued during the test. The test temperature was set to 25 C (298 K), and the test duration was four days or one week. The permeation current density was measured, and the results are presented in terms of hydrogen permeability, Per, as defined by: Per = J L... (1) where J refers to the measured current density, L the test 2.2. Hydrogen Permeation Method A hydrogen permeation test 9) was carried out to evaluate the hydrogen penetration performance of the steel specimens under various environments. Figure 1 schematically shows the constitution of the hydrogen permeation system. Cathodiccharging was applied to the surface on the left side cell, in which a corrosive solution mixed with certain poisonous substances, such as hydrogen sulfide (H 2S) or ammonium thiocyanate (NH 4SCN), was poured. The current detected by the oxidation of hydrogen was measured on the right-side cell filled with a 1 mol/l NaOH solution. Ni plating was applied to the specimen surface facing to the right-hand side. 9,10) Fig. 1. Schematic illustration of the hydrogen permeation method. Table 1. Chemical composition and tensile properties of the tested steel. (mass%) C Si Mn Mo B V TS, MPa X X X X VC VC D Method Table 2. Galvanostaticcharging Potentiostaticcharging Immersion Hydrogen charging conditions and test solutions. Test conditions 1) 3%NaCl + 1 mol/l acetic buffer solution + 3 g/l NH 4SCN (ph:5.4) 0.02 to 0.5 ma/cm 2 (0.2 to 5.0 A/m 2 ) 2) 3%NaCl + NaOH (ph:12.5) 0.01 to 0.1 ma/cm 2 (0.1 to 1.0 A/m 2 ) 3%NaCl + 1 mol/l acetic buffer solution + 3 g/l NH 4SCN (ph:5.4) 0.75 to 0.80 V vs SCE 1) 5%NaCl + 1 mol/l acetic buffer solution saturated with H 2S of 0.1 MPa (ph: 3.0 to 5.3) 2) 3%NaCl + 1 mol/l acetic buffer solution to 50 g/l NH 4SCN (ph:5.4) ISIJ

3 Fig %C, 0.06%C, and 0.25%C steel specimens under galvanostatic-charging of 0.2 A/m 2 (0.02 ma/cm 2 ) in a buffer solution with 3 g/l NH 4SCN. (a) X65-1 (0.04%C) and X100 (0.06%C) steel specimens, (b) VC (0.25%C) steel specimens. Fig %C (X65-1) and 0.06%C (X100) steel specimens under galvanostatic-charging of 2.0 A/m 2 (0.2 ma/cm 2 ) in a buffer solution with 3 g/l NH 4SCN. Fig %C (X65-1 and X120) steel specimens under galvanostatic-charging of 5.0 A/m 2 (0.5 ma/cm 2 ) in a buffer solution with 3 g/l NH 4SCN. specimen thickness. The hydrogen permeation area was accurately measured to be 2.1 to 2.3 cm Results 3.1. Time Dependence of Hydrogen Permeability in Galvanostatic-charging Figure 2 shows the time dependence of the hydrogen permeability of the 0.04% carbon steel specimen (X65-1), the 0.06% carbon steel specimen (X100), and the 0.25%C specimens (VC-1 and VC-2), under galvanostatic-charging of 0.02 ma/cm 2 (0.2 A/m 2 ) in a buffer solution, with 3 g/l NH 4SCN. The hydrogen permeability of the steel except for VC-2 reached A/m immediately after the test started and then showed the same value for minutes. By contrast, there was some incubation time for the buildup of hydrogen permeability only for the VC-2 steel, which is due to the hydrogen trap sites caused by precipitated vanadium carbides. The hydrogen permeability of all the steel specimens at a steady state was almost the same value irrespective of the carbon content and tensile strength. Figure 3 shows the time dependence of the hydrogen permeability of the 0.04% carbon steel (X65-1) and 0.06% carbon steel (X100) specimens under galvanostatic-charging at 0.2 ma/cm 2 (2.0 A/m 2 ), in a buffer solution with 3 g/l NH 4SCN. The hydrogen permeability of both steel samples reached about A/m immediately after the test started and then showed almost the same value for minutes. However, a slight fluctuation was observed in the hydrogen permeability of both steel specimens. Figure 4 shows the time dependence of the hydrogen permeability of the 0.04% carbon steel specimens (X65-1 and X120) under galvanostatic-charging at 0.5 ma/cm 2 (5.0 A/m 2 ) in a buffer solution with 3 g/l NH 4SCN. The hydrogen permeability of both steel samples reached about A/m immediately after the test started, and then showed almost the same value for minutes. A slight fluctuation in hydrogen permeation was observed. The reason why hydrogen permeability showed slight fluctuation is considered to be the influence of the adsorption of hydrogen gas on the surface of the test specimen. Therefore, applying galvanostaticcharging at low current density is preferable Time Dependence of Hydrogen Permeability in Potentiostatic-charging Figure 5 shows the time dependence of the hydrogen permeability of the 0.25% carbon steel specimens (VC-1 and VC-2) under potentiostatic-charging at 0.75 V (SCE) and 0.80 V (SCE) in a buffer solution with 3 g/l NH 4SCN. Corrosion potential stayed in between 0.70 V (SCE) and 0.72 V (SCE). The hydrogen permeability of both steel specimens reached A/m immediately after the test started under potentiostatic-charging at 0.75 V (SCE), and then decayed with increasing test time. The hydrogen permeability of both steel specimens reached 1.2 to ISIJ 288

4 Fig %C (VC-1 and VC-2) steel specimens under potentiostatic-charging of 0.75 V and 0.80 V vs SCE in an a buffer solution with 3 g/l NH 4SCN. (a) 0.75 V vs SCE, (b) 0.80 V vs SCE. Fig. 6. Time dependence of the current density of the 0.25%C (VC-1 and VC-2) steel specimens under potentiostatic-charging of 0.75 V and 0.80 V in a buffer solution with 3 g/l NH 4SCN. (a) 0.75 V vs SCE, (b) 0.80 V vs SCE. A/m immediately after the test started under potentiostaticcharging of 0.80 V (SCE) and then decayed with increasing test time. Thus, hydrogen permeability decayed with time under potentiostatic-charging at 0.75 and 0.80 V (SCE). Figure 6 shows the time dependence of current density under potentiostatic-charging of 0.75 and 0.80 V (SCE). Cathodic current density also decayed with time under potentiostatic-charging at 0.75 and 0.80 V (SCE). The reason why hydrogen permeability and cathodic current density decayed with time is not clear although it may be pointed out that there are some influences that electrochemical reactions such as reduction of passive film or rust parallelly occur. This is the future subject Time Dependence of Hydrogen Permeability in the Immersion Method Figure 7 shows the time dependence of the hydrogen permeability of the 0.04% carbon steel specimens (X65-1 and X65-2) obtained by immersion test method in H 2S saturated solutions with an H 2S partial pressure of 0.1 MPa and a ph of 3.0. The hydrogen permeability of both steel specimens was as high as 8.0 to A/m immediately after the test started, and then decayed with time, to reach the constant value of A/m after minutes. The reason why hydrogen permeability decayed with time is that the corrosion rate is lowered with time through the formation of iron sulfide (FeS). 11) Figure 8 shows the time dependence of the hydrogen permeability of the 0.04% carbon steel (X65-1) obtained by the same method in an H 2S saturated solution with an H 2S partial pressure of 0.1 MPa and ph of 5.0 and 5.3. The hydrogen permeability of both steel Fig %C (X65-1 and X65-2) steel specimens in an H 2S saturated solution with an H 2S partial pressure of 0.1 MPa and a ph of 3.0. specimens at ph of 5.0 and 5.3 stand still at around and A/m, respectively. Thus, the immersion test method in the solution saturated with H 2 S partial pressure with 0.1 MPa at ph of 5.0 to 5.3 is plausible. However, hydrogen cannot enter into steel in an H 2 S saturated solution with an H 2 S partial pressure of 0.1 MPa if the ph exceeds 5.5, due to the stability of the formation of iron sulfide (FeS) in an H 2 S solution environment. 4) Therefore, it is better to perform immersion test using a saturated sour environment with an H 2 S partial pressure less than 0.1 MPa. Figure 9 shows the time dependence of the hydrogen permeability of the 0.25% carbon steel specimens (VC-1 and VC- 2) obtained by immersion tests in an ammonium thiocyanate ISIJ

5 solution in which the concentration of ammonium thiocyanate was 3 g/l. The hydrogen permeability of VC-1 steel specimens reached the saturated value of A/m immediately after the test started, while that of VC-2 showed some incubation time. This trend was the same as the one which was observed in galvanostatic-charging method as in Fig. 2(b) Method to Charge Small Amount of Hydrogen into Steel for Evaluation of Threshold Hydrogen Concentration At first, the conditions by which hydrogen enters into Fig %C (X65-1) steel specimens in an H 2S saturated solution with an H 2S partial pressure of 0.1 MPa and a ph of 5.0 and 5.3. steel through galvanostatic-charging were studied. Figure 10 shows the time dependence of hydrogen permeability under the cathodic-charging of different current densities in a deaerated alkali solution. 12) Hydrogen permeability in an alkali solution showed almost constant values. Based on this result, constant hydrogen concentrations can enter into steel under galvanostatic-charging in an alkali solution. A different hydrogen concentration can be obtained accurately under galvanostatic-charging with selecting the applied current density. The conditions by which hydrogen can enter into steel by the immersion test are also studied. Figure 11 shows the time dependence of hydrogen permeability in a 3 mass% acetic buffer solution with a different ammonium thiocyanate concentration from 3 to 50 g/l. Hydrogen permeability showed an almost constant value soon after immersion, irrespective of the ammonium thiocyanate concentration. Although incubation time is existing in the case of VC-2 steel, saturated hydrogen permeability values are essentially the same regardless of hydrogen trapping kinds. Therefore, the immersion test method is convenient to accurately control hydrogen concentration in steel. It should be noted that it is not possible to avoid the corrosion film formation on the steel surface. The results of hydrogen entry into steel under the galvanostatic-charging, potentiostatic-charging, and immersion methods, were summarized in Table 3. Hydrogen permeability showed a constant value irrespective of test time under galvanostatic-charging at lower current density than 0.02 ma/cm 2 (0.2 A/m 2 ). Hydrogen permeability decayed with increasing time Fig %C (VC-1 and VC-2) steel specimens in an ammonium thiocyanate solution. Fig %C steel specimen (D) under the cathodic-charging of a different current density in an alkali solution. Fig. 11. Time dependence of hydrogen permeability in an acetic buffer solution with different ammonium thiocyanate concentrations changing in the range of 0.3 to 50 g/l. (a) VC-1 (0.25%C) steel specimens, (b) VC-2 (0.25%C) steel specimens ISIJ 290

6 Table 3. Method Summary of the results of stable hydrogen entry into steel from the cathodic-charging and immersion methods. Condition of constant hydrogen entry into steel Galvanostaticcharging Potentiostaticcharging Immersion 1) Applied to a deaerated buffer solution of ph of 5 or higher with low current density 2) Applied to a deaerated alkali solution Not found in this study Applied to a buffer solution with poisonous substances such as NH 4SCN in weak acidic or neutral environments under potentiostatic-charging at all applied potentials. The potentiostatic-charging methods by which hydrogen stably enters into steel were not found in this study. Hydrogen permeability decayed during an immersion test in acidic environments with ph less than 3.0 because the corrosion rate was decreased by the formation of corrosion product film. Therefore, the methods by that hydrogen can stably enter into steel during immersion tests in a solution of a ph more than 5.0 with poisonous substances such as H 2S and ammonium thiocyanate are recommended. Fig. 12. Correlation between hydrogen permeability and hydrogen overpotential in acidic and neutral environments with the present data obtained from the potentiostatic-charging, galvanostatic-charging, and immersion methods. (The black and red marks show the immersion test, the blue marks show the potentiostatic-charging, and the green marks show the galvanostatic-charging.) 4. Discussion Correlation between hydrogen permeability and hydrogen overpotential Figure 12 represents the correlation between hydrogen permeability and hydrogen overpotential in acidic and neutral environments 13) together with the present data obtained using the potentiostatic-charging, galvanostatic-charging, and immersion methods. Overpotentials, from H 2O/H 2 equilibria, which is defined as hydrogen overpotential for here were estimated from a potentio-dynamic cathodic polarization curve during galvanostatic-charging. In this plot, a unit of hydrogen permeability was converted from an electronic charge amount to an atomic hydrogen flux. The data points followed a linear relationship in spite of the different chemical compositions, strengths, and microstructures of steels for the immersion tests as shown by the solid line. For potentiostatic-charging and galvanostatic-charging, the correlation between hydrogen permeability and hydrogen overpotential had a different relationship from the immersion methods as shown in dashed line. With hydrogen overpotential less than 0.18 V, the same relationship was observed regardless of hydrogen charging methods. 12) It is thought that excessive applied potential changes the hydrogen evolution reaction rate or the rate-controlling step of the hydrogen evolution reaction. The difference of the hydrogen entry conditions under cathodic-charging and under natural immersion must be clarified as future subjects. 5. Conclusions In order to establish a method to allow small amounts of hydrogen to stably enter into steel, the time dependence of the hydrogen concentrations in steel was investigated using galvanostatic-charging, potentiostatic-charging, and immersion methods in various environments. The main conclusions are as follows. Hydrogen concentrations can stably enter into steel under the galvanostatic-charging of a low current density in a deaerated buffer solution of a ph more than 5.0. Hydrogen can also stably enter into steel in immersion tests such as those in the environments of ph more than 5.0, with poisonous substances such as H 2S and ammonium thiocyanate. A different hydrogen concentration can be accurately obtained by an immersion test with different concentrations of poisonous substances such as H 2S or ammonium thiocyanate, or with a different ph, and it also can be obtained by applying the different low current densities under galvanostaticcharging. Therefore, the threshold hydrogen concentration of delayed fracture can be accurately obtained using the continuous-charging method. Acknowledgement I am grateful to Dr. Kihira from Nippon Steel Corporation for his valuable advice and to this paper. REFERENCES 1) T. Murata: 78, 79th Nishiyama Memorial Seminar, ISIJ, Tokyo, (1981), 227 (in Japanese). 2) E. Protopopoff and P. Marcus: J. Chim. Phys., 88 (1991), ) G. Jerkiewicz, J. Jborodzinski, W. Chrzanowski and B. E. Conway: J. Electrochem. Soc., 142 (1995), No. 11, ) T. Hara and H. Asahi: Quantitative Study on Conditions of HIC Occurrence for X65 Linepipe Steels Proc. 16th Int. Corrosion Cong., Beijing, China, (2005), 19. 5) H. Asahi, M. Ueno and T. Yonezawa: Corrosion, 50 (1994), No. 7, ) S. Yamazaki: Tetsu-to-Hagané, 83 (1997), No. 7, 42 (in Japanese). 7) NACE standard TM Item No.21212, Nace Int., Houston, Texas. 8) H. Asahi, S. Yamazaki and D. Hirakami: ISIJ Int., 43 (2003), No. 4, ) S. Yoshizawa, T. Tsuruta and K. Yamakawa: Boshoku-Gijutsu (presently Zairyo-to-Kankyo) 24 (1975), 511 (in Japanese). 10) S. Yoshizawa and K. Yamakawa: Boshoku-Gijutsu (presently Zairyoto-Kankyo) 24 (1975), 365 (in Japanese). 11) H. Ogawa: 133th Nishiyama Memorial Seminar, ISIJ, Tokyo, (1990), 17 (in Japanese). 12) T. Hara and T. Tarui: Zairyo-to-Kankyo, 59 (2010), 173 (in Japanese). 13) T. Hara: Zairyo-to-Kankyo, 60 (2011), No. 5, 259 (in Japanese) ISIJ