- Email: [email protected]
Hydrogen solubility and permeability of V–W–Mo alloy membrane for hydrogen separation and purification
Journal of Alloys and Compounds 580 (2013) S386–S390
Contents lists available at SciVerse ScienceDirect
Journal of Alloys and Compounds journal homepage: www.elsevier.com/locate/jalcom
Hydrogen solubility and permeability of V–W–Mo alloy membrane for hydrogen separation and purification H. Yukawa a,⇑, C. Tsukada a, T. Nambu b, Y. Matsumoto c a
Department of Materials Science and Engineering, Graduate School of Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, Aichi 464-8603, Japan Department of Materials Science and Engineering, Suzuka National College of Technology, Shiroko-cho, Suzuka, Mie 510-0294, Japan c Department of Mechanical Engineering, Oita National College of Technology, Maki, Oita 870-0152, Japan b
a r t i c l e
i n f o
Article history: Available online 26 March 2013 Keywords: Hydrogen permeable membrane V–based alloy Hydrogen solubility Hydrogen embrittlement Hydrogen permeability DBTC
a b s t r a c t The alloying effects of molybdenum (Mo) into V–W alloy on the solubility of hydrogen, the resistance to hydrogen embrittlement and the hydrogen permeability are investigated in a fundamental manner. It is found that the addition of Mo into V–W alloy decreases the hydrogen solubility. As a result, the applicable hydrogen pressures at the DBTC (the ductile-to-brittle transition hydrogen concentration, which is about 0.2 (H/M) for vanadium alloys) increases. In fact, about 0.15 MPa (673 K) to 0.6 MPa (773 K) of hydrogen pressures can be applied to V–5 mol%W–5 mol%Mo alloy membrane while keeping the hydrogen concentration less than or equal to the DBTC, which is about twice of hydrogen pressure applicable to V–5 mol%W alloy. Thus, the addition of Mo into V–W alloy improves the resistance to hydrogen embrittlement. In addition, the alloying of Mo into V–W alloy also improves the hydrogen permeability. For instance, the hydrogen permeability of V–5 mol%W–5 mol%Mo alloy is about 4–5 times higher than that of Pd–25mass%Ag alloy at 673–773 K. Ó 2013 Elsevier B.V. All rights reserved.
1. Introduction Mass production of high purity hydrogen gas efficiently is necessary for the future clean energy society. Hydrogen permeable membranes are important materials for hydrogen separation and purification [1,2]. They are also important for the hydrogen production by membrane reactors. For example, Pd-based alloy membranes (e.g., Pd–Ag and Pd–Cu membranes) are widely used practically for these purposes. Recently, there has been a great demand for the development of new hydrogen permeable alloys to substitute for currently used Pd-based alloys, in order to reduce the material cost as well as to improve the hydrogen permeability [3–6]. Group 5 metals such as niobium (Nb) and vanadium (V) are less expensive and exhibit higher hydrogen permeability than palladium [7], so they are ones of the most promising metals for hydrogen permeable membranes. However, there is still a large barrier to the practical use due to their poor resistance to hydrogen embrittlement. Recently, the mechanical properties of niobium in hydrogen gas atmosphere at high temperature have been investigated by the in situ small punch (SP) test method [8]. It is found that the ductile-to-brittle transition occurs drastically at the hydrogen concentration around 0.20 (H/M) at the temperature range between 573 K
⇑ Corresponding author. Tel.: +81 527893233. E-mail address: [email protected] (H. Yukawa). 0925-8388/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jallcom.2013.03.144
and 773 K. This fact suggests that the resistance to hydrogen embrittlement of niobium will be improved by keeping the hydrogen concentration below this critical value during the practical hydrogen permeation. From these results, the present authors have proposed a concept for alloy design of Nb-based hydrogen permeable membrane in order to satisfy both high hydrogen permeability and strong resistance to hydrogen embrittlement [9]. Following the concept, Nb–W, Nb–Ru and Nb–W–Mo alloy membranes have been designed and developed, that possess excellent hydrogen permeability without showing any evidence of hydrogen embrittlement [10,11]. Recently, the concept has been applied to vanadium system [12]. The ductile-to-brittle transition hydrogen concentration (DBTC) for pure vanadium and V-based alloys is found to be about 0.2 (H/M), similar to that for pure niobium [8]. Then, the alloying effects of tungsten (W) on the hydrogen solubility, the resistance to hydrogen embrittlement and the hydrogen permeability have been investigated [12]. It is shown that the designed V–5 mol%W alloy membrane exhibits higher hydrogen permeability and stronger resistance to hydrogen embrittlement than Nb-based alloy membranes. For example, no cracking due to hydrogen embrittlement occurs for V–5 mol%W alloy membrane during hydrogen permeation under 0.3 MPa of hydrogen pressure at 773 K, which is about three times higher pressure applicable to Nb–5 mol%W–5 mol%Mo alloy membrane at the same temperature [11,12].
S387
H. Yukawa et al. / Journal of Alloys and Compounds 580 (2013) S386–S390
In this study, the alloying effects of molybdenum (Mo) into V–W alloy on the hydrogen solubility are investigated in a fundamental manner in order to improve further the resistance to hydrogen embrittlement. The hydrogen permeability of the V–W–Mo alloy membrane is also investigated.
Hydrogen pressure, P / MPa
V-5W-5Mo
2. Experimental procedure 2.1. Sample preparation The purity of the raw materials used in this study is 99.9 mass% for vanadium (Taiyo Koko Co., Ltd.) and 99.95 mass% for tungsten and molybdenum (Rare Metallic Co., Ltd.). V–5 mol%W–5 mol%Mo alloy is prepared by using tri-arc furnace in a purified argon gas atmosphere. According to the X-ray diffraction analysis, the alloy is confirmed to be composed of a single solid solution phase with simple BCC (bodycentered cubic) crystal structure.
0
10
-1
10
V-5W
[12]
0 .01 MPa
Pure V
[14]
-2
10
673K -3
10
2.2. Hydrogen pressure–composition–isotherm (PCT) measurement In order to examine the hydrogen solubility, the pressure–composition–isotherms (PCT) are measured by using a Sieverts-type apparatus. A small piece of the alloy is set into a cell and then evacuated with turbo molecular pump (TMP) system. Subsequently, it is heated up to 773 K, and then high purity hydrogen (99.99999% purity) of about 5 MPa is introduced and cooled down to room temperature. This process is repeated at least three times prior to the PCT measurement in order to activate the sample surface for the hydrogen absorption and desorption reactions to take place smoothly. The PCT curves are measured at 673–773 K up to about 5 MPa.
1
10
0.0
0.1
0.2
0.3
0.4
0.5
0.6
Hydrogen content, content C / (H/M) Fig. 1. PCT curve for V–5 mol%W–5 mol%Mo alloy measured at 673 K. PCT curves for pure V [14] and V–5 mol%W alloy [12] measured at 673 K are also drawn in the figure for comparison.
1
10
723K
The hydrogen permeation tests are performed at 673–773 K in order to evaluate the hydrogen permeability. Disk-shaped specimens of about / 12 mm in diameter with a thickness of about 0.5 mm are prepared. They are polished mechanically by using alumina abrasive paper followed by the final polishing with 0.3 lm Al2O3 powders. Subsequently, pure palladium of about 200 nm in thickness is deposited at 573 K on both sides of the sample surfaces by using an RF magnetron sputtering apparatus. This palladium layer on the surface protects the sample from the oxidation. It also acts as a catalyst for hydrogen dissociation reaction and subsequent dissolution into metal to take place smoothly. The disk sample is set into the hydrogen permeation apparatus and then evacuated with TMP. Subsequently, it is heated up to the measuring temperature, and then a high purity hydrogen gas is introduced into both sides of the specimen. The hydrogen pressure applied to the inlet side is determined from the corresponding PCT curve at each temperature so that the hydrogen concentration does not exceed the critical value of the ductile-to-brittle transition hydrogen concentration, i.e., 0.2 (H/M), to avoid the sample cracking due to hydrogen embrittlement. The outlet pressure is fixed to be 0.01 MPa. Then, the hydrogen fluxes, J, permeated through the disk samples are measured by the conventional gas permeation method using a mass flowmeter. A detailed explanation of the permeation test is given elsewhere [13].
Hydrogen pressure, P / MPa
2.3. Hydrogen permeation test
773K
673K
0
10
0.6 MPa 0. 3 MPa 0.15 MPa
-1
10
-2
10
DBTC of V
-3
Ductile
[8]
Brittle
10
0.0
0.1
0.2
0.3
0.4
0.5
0.6
Hydrogen content, C / (H/M) Fig. 2. PCT curves for V–5 mol%W–5 mol%Mo alloy measured at 673–773 K. The ductile-to-brittle transition hydrogen concentration (DBTC) for vanadium alloys [8] is indicated by a broken line in the figure.
3. Results and discussion 3.1. Alloying effects of Mo on the hydrogen solubility of V–W alloy The PCT curve for V–5 mol%W–5 mol%Mo alloy measured at 673 K is shown in Fig. 1. For comparison, the results for pure vanadium reported by Veleckis and Edwards [14] and for V–5 mol%W alloy reported by Yukawa et al. [12] are also drawn in the figure. As is evident from the figure, the PCT curve shifts toward the left and upper region by the addition of 5 mol% of molybdenum into V–5 mol%W alloy, indicating that the amount of dissolved hydrogen under a given hydrogen pressure (e.g., 0.01 MPa) decreases by the addition of molybdenum. Fig. 2 shows the PCT curves for V–5 mol%W–5 mol%Mo alloy measured at 673–773 K. As shown in Fig. 2, the PCT curve also shifts toward the left and upper region with increasing temperature. As a result, the equlilibrium hydrogen pressure at the DBTC increases with increasing temperature. For example, about 0.15 MPa, 0.3 MPa and 0.6 MPa of hydrogen pressures can be applied to V–5 mol%W–5 mol%Mo alloy membrane at 673 K, 723 K and 773 K, respectively, which is about twice
of hydrogen pressure applicable to V–5 mol%W alloy membrane [12]. 3.2. Hydrogen permeability of V–W–Mo alloy membrane The hydrogen diffusion in metal membrane is generally the rate-limiting process of the total reaction of hydrogen permeation through it. Then, the Fick’s law:
J ¼ D
@C @x
ð1Þ
is commonly applied to the metal membrane to discuss the property of it, where J is the hydrogen flux, D is the diffusion coefficient and @C/@x is the concentration gradient across the membrane. Applying the Fick’s law to a membrane with a thickness of d, the Eq. (1) can be modified as follows:
J¼D
C in C out d
ð2Þ
S388
H. Yukawa et al. / Journal of Alloys and Compounds 580 (2013) S386–S390
other hand, as shown in Fig. 1, the hydrogen concentration at the outlet side is smaller for V–5 mol%W–5 mol%Mo alloy than V– 5 mol%W alloy at the same pressure, e.g., 0.01 MPa. Then, the hydrogen concentration difference between the inlet and outlet sides, Cin Cout, is increased by the addition of molybdenum into V–W alloy. As a result, higher hydrogen flux can be obtained for V–W–Mo alloy than V–W alloy, following Eq. (2). In fact, the hydrogen flux for V–5 mol%W–5 mol%Mo alloy measured at 773 K under the pressure condition of inlet/outlet = 0.6/0.01 MPa is higher than that for V–5 mol%W alloy measured under the pressure condition of inlet/outlet = 0.3/0.01 MPa. In addition, it is about four times higher than that for Pd–25 mol%Ag alloy estimated for 773 K under the same pressure condition, i.e., inlet/outlet = 0.6/ 0.01 MPa. Similar results are also obtained at 723 K and 673 K as shown in Fig. 3b and c, respectively. It is important to note here that the hydrogen diffusivity is also an important factor controlling the hydrogen permeability as expressed by Eq. (2). The investigation about the alloying effects on the hydrogen diffusion coefficient under the practical condition of hydrogen permeation is now underway. After the hydrogen permeation test, the system is evacuated. Then the gas leak test is performed using helium (He) in order to check the existence of any cracks on the disk sample. A photo image of the sample after the hydrogen permeation test is shown in the inset of Fig. 3a. It is confirmed that there is no evidence of crack on the sample due to hydrogen embrittlement. Thus, V–5 mol%– 5 mol%Mo alloy membrane possesses high hydrogen permeability
120
(a)
ol H m-1 S-1 Hydrogen flux, 10 J d / mo
773K
100
80
[12]
V-5W (0.3/0.01)
60
100
(b)
723K
80
60
0.3 / 0.01
0.1 / 0.01
0 15 / 0 0.15 0.01 01
0 05 / 0 0.05 0.01 01
6
6
ol H m-1 S-1 Hydrogen flux, 10 J d / mo H
where Cin and Cout are the hydrogen concentration at the inlet and outlet sides of the membrane. The steady-state hydrogen fluxes, J, are measured at 673–773 K for V–5 mol%W–5 mol%Mo alloy membrane. It is divided by the inverse of the sample thickness, 1/d, in order to estimate the normalized hydrogen flux, J d. It is noted here that atomic hydrogen flux (mol H m1 s1) is evaluated in this study, which is twice as large as gaseous hydrogen flux (mol H2 m1 s1). The time dependences of the normalized hydrogen flux, J d, during the measurement at 673–773 K are shown in Fig. 3a–c. The inlet and outlet hydrogen pressures for each measurement are indicated in the legend of the figure as (inlet/outlet (MPa)). For comparison, the corresponding values of V–5 mol%W alloy measured at 773 K [12] and Pd–25 mol%Ag alloy estimated from Ref. [15] are indicated in the figures with arrows. As shown in Fig. 3, the hydrogen flux is stable and nearly constant during each measurement. The value of the hydrogen flux changes depending on the applied hydrogen pressures. As shown in Fig. 3a, the hydrogen flux is higher for V–5 mol%W alloy than V–5 mol%W–5 mol%Mo alloy under the same pressure condition, i.e., inlet/outlet = 0.3/0.01 MPa, meaning that the hydrogen permeability is decreased by alloying molybdenum with V–W alloy. However, when considering the ductile-to-brittle transition hydrogen concentration, DBTC, higher hydrogen pressure can be applied to V–5 mol%W–5 mol%Mo alloy and the maximum hydrogen concentration at the inlet side become the same, i.e., 0.2 (H/M), for both V–5 mol%W and V–5 mol%W–5 mol%Mo alloys. On the
40
[15]
Pd-25Ag (0.6/0.01)
0.6 / 0.01 0.56 / 0.01 0.5 / 0.01
0.3 / 0.01 0.1 / 0.01 0.05 / 0.01
20
0 0
100
200
300
500
400
600
40
20 Pd-25Ag [15] (0.3/0.01)
0
700
100
0
200
300
400
500
600
700
Time , t / s
70
(c)
60
673K
50 40
6
ol H m-1 S-1 Hydrogen flux, 10 J d / mo
Time , t / s
30
0.15 / 0.01 0.10 / 0.01
0.05 / 0.01
20 Pd-25Ag [15] (0.15/0.01)
10 0 0
100
200
300
400
500
600
700
Time , t / s Fig. 3. Time dependences of the normalized hydrogen flux, J d, during the measurements at (a) 773 K, (b) 723 K and (c) 673 K. The inlet and outlet hydrogen pressures for each measurement are indicated in the legend as (inlet/outlet (MPa)). A photo image of the sample after hydrogen permeation test is shown in the inset of the figure.
S389
H. Yukawa et al. / Journal of Alloys and Compounds 580 (2013) S386–S390
120
Hydrogen flux, 10 J d / mol H m-1 S-1
673K 0.3
723K
773K
DBTC of V [8]
V-5W-5Mo
100
80
60
6
0.2
0.0
0.2
0.3 MPa
0.0
0.6 MPa
0.1 0.15 MPa
Hydrogen content, C / (H/M)
0.4
0 .4
0 .6
673K
40
723K 773K
20
0
0 .8
1. 0
0
1. 2
100
200
300
Δ
Square root of hydrogen pressure , P / MPa1/2
400
500
600
700
800
Pa1/2 P/P
Fig. 5. Correlation between the normalized hydrogen flux, J d, and the difference pffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffi of the square root of hydrogen pressures, P inlet P outlet , at 673–773 K.
Fig. 4. Correlation between the square root of hydrogen pressure and the hydrogen content for V–5 mol%W–5 mol%Mo alloy measured at 673–773 K. -6
-1 -1
3.3. Analysis of hydrogen permeation coefficient When the Sieverts’ law is valid in the metal-hydrogen system, the hydrogen flux expressed by Eq. (2) can be written as follows:
J ¼DK
pffiffiffiffiffiffi pffiffiffiffiffiffiffiffi Pin Pout d
Permeability, e o H 2 m s Pa a φ / mol
together with stronger resistance to hydrogen embrittlement than V–5 mol%W alloy.
-1/2
10
ð3Þ
where K is the Sievert’s solubility constant, Pin and Pout are the inlet and outlet hydrogen pressures. The product of D and K is defined as the hydrogen permeation coefficient, /(=D K), and commonly used as a measure to show the hydrogen permeability of metals and alloys. The PCT curves shown in Fig. 2 are redrawn in Fig. 4 in which the hydrogen concentration is plotted as a function of the square root of hydrogen pressure. It is evident that the Sieverts’ law is almost valid for V–5 mol%W–5 mol%W alloy at 673–773 K below the DBTC, i.e., 0.2 (H/M). Then, the hydrogen fluxes shown in Fig. 3 are analyzed in view of the difference of the square root of hydrogen pressure, and the results are shown in Fig. 5. As shown in the figure, there is almost a linear relationship between the normalized hydrogen flux, J d, and the difference of the square root of pffiffiffiffiffiffi pffiffiffiffiffiffiffiffi hydrogen pressure, Pin Pout , for each measuring temperature, indicating that the hydrogen permeation reaction through the V–5 mol%W–5 mol%Mo alloy membrane takes place following Eq. (3). Then, the hydrogen permeation coefficients, /, are estimated from the slope of the lines shown in Fig. 5 and the results are plotted in Fig. 6 as a function of the inverse of temperature. For comparison, the hydrogen permeation coefficients for Pd–25mass%Ag [15], V–15 mol%Ni [16] and Nb–30 mol%Ti–30 mol%Ni [17] alloys are also presented in the figure. As shown in Fig. 6, the hydrogen permeability of V–5 mol%W–5 mol%Mo alloy is about 4–5 times higher than that for Pd–25mass%Ag, V–15 mol%Ni and Nb–30 mol%Ti– 30 mol%Ni alloys at 673–773 K. As shown in Fig. 6, the hydrogen permeation coefficient for V–5 mol%W–5 mol%Mo alloy estimated by the present permeation method increases with decreasing temperature at 673–773 K. This trend is in good agreement with that for pure vanadium calculated from the hydrogen solubility and diffusivity [7]. It is noted here that the temperature dependence of the hydrogen permeation
-7
10
V-15Ni
[16]
-8
10
Pd-25Ag [15] Nb40 - Ti 30 - Ni 30
(Duplex phase alloy)
-9
[17]
10
1.2
1.4
1.6
1.8
1000/T / K
2.0
2.2
2.4
-1
Fig. 6. Temperature dependence of the hydrogen permeation coefficient for V– 5 mol%W–5 mol%Mo alloy. The reported values for Pd–25mass%Ag [15], V– 15 mol%Ni [16] and Nb–30 mol%Ti–30 mol%Ni [17] alloys are also plotted in the figure for comparison.
coefficient for V-based alloy is different from that for Nb-based alloy. For example, the hydrogen permeation coefficient for Nb– 5 mol%W–5 mol%Mo alloy estimated from the same permeation method with the present study decreases with decreasing temperature at 673–773 K [18].
4. Summary The alloying effects of molybdenum on the hydrogen solubility and the hydrogen permeability of V–W alloy are investigated quantitatively. The hydrogen solubility decreases by the addition of molybdenum into V–W alloy or by increasing temperature. As a result, the resistance to hydrogen embrittlement is improved by reducing the dissolved hydrogen concentration in the alloy. In fact, V–5 mol%W–5 mol%Mo alloy possesses excellent hydrogen permeability without showing any hydrogen embrittlement when used under appropriate permeation conditions.
S390
H. Yukawa et al. / Journal of Alloys and Compounds 580 (2013) S386–S390
Acknowledgement This research was supported in part by the Japan Society for the Promotion of Science (JSPS). References [1] [2] [3] [4]
S.N. Paglieri, J.D. Way, Sep. Purif. Methods 31 (2002) 1–169. V. Mordkovich, Y.K. Baichtock, M. Sosna, Platinum Met. Rev. 36 (1992) 90–97. R. Buxbaum, A. Kinney, Ind. Eng. Chem. Res. 35 (1995) 530–537. C. Nishimura, M. Komaki, S. Hwang, M. Amano, J. Alloys Comp. 330–332 (2002) 902–906. [5] K. Hashi, K. Ishikawa, T. Matsuda, K. Aoki, J. Alloys Comp. 368 (2004) 215–220. [6] T. Nambu, K. Shimizu, Y. Matsumoto, R. Rong, H. Yukawa, M. Morinaga, I. Yasuda, J. Alloys Comp. 446–447 (2007) 588–592. [7] S.A. Steward, Lawrence Livemore National Laboratory Reports, UCRL-53441, 1983.
[8] Y. Matsumoto, H. Yukawa, T. Nambu, Metall. J. LXIII (2010) 74–78. [9] H. Yukawa, T. Nambu, Y. Matsumoto, N. Watanabe, G.X. Zhang, M. Morinaga, Mater. Trans. 49 (2008) 2202–2207. [10] N. Watanabe, H. Yukawa, T. Nambu, Y. Matsumoto, G.X. Zhang, M. Morinaga, J. Alloys Comp. 477 (2009) 851–854. [11] Y. Awakura, T. Nambu, M. Matsumoto, H. Yukawa, J. Alloys Comp. 509S (2011) S877–S880. [12] H. Yukawa, T. Nambu, M. Matsumoto, M. Morinaga, J. Alloys Comp. 509S (2011) S881–S884. [13] T. Nambu, N. Shimizu, H. Ezaki, H. Yukawa, M. Morinaga, J. Jpn. Inst. Met. 69 (2005) 841–847. [14] E. Veleckis, R.K. Edwards, J. Phys. Chem. 73 (1969) 683–692. [15] E. Serra, M. Kemali, A. Perujo, D.K. Ross, Metall. Mater. Trans. A 29A (1998) 1023–1028. [16] C. Nishimura, M. Komaki, S. Hwang, M. Amano, Int. J. Hydrogen Energy 330– 332 (2002) 902–906. [17] K. Ishikawa, S. Tokui, K. Aoki, Intermetallics 17 (2009) 109–114. [18] H. Yukawa, T. Nambu, Y. Matsumoto, Defect Diffusion Forum 333 (2013) 61– 71.
Contents lists available at SciVerse ScienceDirect
Journal of Alloys and Compounds journal homepage: www.elsevier.com/locate/jalcom
Hydrogen solubility and permeability of V–W–Mo alloy membrane for hydrogen separation and purification H. Yukawa a,⇑, C. Tsukada a, T. Nambu b, Y. Matsumoto c a
Department of Materials Science and Engineering, Graduate School of Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, Aichi 464-8603, Japan Department of Materials Science and Engineering, Suzuka National College of Technology, Shiroko-cho, Suzuka, Mie 510-0294, Japan c Department of Mechanical Engineering, Oita National College of Technology, Maki, Oita 870-0152, Japan b
a r t i c l e
i n f o
Article history: Available online 26 March 2013 Keywords: Hydrogen permeable membrane V–based alloy Hydrogen solubility Hydrogen embrittlement Hydrogen permeability DBTC
a b s t r a c t The alloying effects of molybdenum (Mo) into V–W alloy on the solubility of hydrogen, the resistance to hydrogen embrittlement and the hydrogen permeability are investigated in a fundamental manner. It is found that the addition of Mo into V–W alloy decreases the hydrogen solubility. As a result, the applicable hydrogen pressures at the DBTC (the ductile-to-brittle transition hydrogen concentration, which is about 0.2 (H/M) for vanadium alloys) increases. In fact, about 0.15 MPa (673 K) to 0.6 MPa (773 K) of hydrogen pressures can be applied to V–5 mol%W–5 mol%Mo alloy membrane while keeping the hydrogen concentration less than or equal to the DBTC, which is about twice of hydrogen pressure applicable to V–5 mol%W alloy. Thus, the addition of Mo into V–W alloy improves the resistance to hydrogen embrittlement. In addition, the alloying of Mo into V–W alloy also improves the hydrogen permeability. For instance, the hydrogen permeability of V–5 mol%W–5 mol%Mo alloy is about 4–5 times higher than that of Pd–25mass%Ag alloy at 673–773 K. Ó 2013 Elsevier B.V. All rights reserved.
1. Introduction Mass production of high purity hydrogen gas efficiently is necessary for the future clean energy society. Hydrogen permeable membranes are important materials for hydrogen separation and purification [1,2]. They are also important for the hydrogen production by membrane reactors. For example, Pd-based alloy membranes (e.g., Pd–Ag and Pd–Cu membranes) are widely used practically for these purposes. Recently, there has been a great demand for the development of new hydrogen permeable alloys to substitute for currently used Pd-based alloys, in order to reduce the material cost as well as to improve the hydrogen permeability [3–6]. Group 5 metals such as niobium (Nb) and vanadium (V) are less expensive and exhibit higher hydrogen permeability than palladium [7], so they are ones of the most promising metals for hydrogen permeable membranes. However, there is still a large barrier to the practical use due to their poor resistance to hydrogen embrittlement. Recently, the mechanical properties of niobium in hydrogen gas atmosphere at high temperature have been investigated by the in situ small punch (SP) test method [8]. It is found that the ductile-to-brittle transition occurs drastically at the hydrogen concentration around 0.20 (H/M) at the temperature range between 573 K
⇑ Corresponding author. Tel.: +81 527893233. E-mail address: [email protected] (H. Yukawa). 0925-8388/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jallcom.2013.03.144
and 773 K. This fact suggests that the resistance to hydrogen embrittlement of niobium will be improved by keeping the hydrogen concentration below this critical value during the practical hydrogen permeation. From these results, the present authors have proposed a concept for alloy design of Nb-based hydrogen permeable membrane in order to satisfy both high hydrogen permeability and strong resistance to hydrogen embrittlement [9]. Following the concept, Nb–W, Nb–Ru and Nb–W–Mo alloy membranes have been designed and developed, that possess excellent hydrogen permeability without showing any evidence of hydrogen embrittlement [10,11]. Recently, the concept has been applied to vanadium system [12]. The ductile-to-brittle transition hydrogen concentration (DBTC) for pure vanadium and V-based alloys is found to be about 0.2 (H/M), similar to that for pure niobium [8]. Then, the alloying effects of tungsten (W) on the hydrogen solubility, the resistance to hydrogen embrittlement and the hydrogen permeability have been investigated [12]. It is shown that the designed V–5 mol%W alloy membrane exhibits higher hydrogen permeability and stronger resistance to hydrogen embrittlement than Nb-based alloy membranes. For example, no cracking due to hydrogen embrittlement occurs for V–5 mol%W alloy membrane during hydrogen permeation under 0.3 MPa of hydrogen pressure at 773 K, which is about three times higher pressure applicable to Nb–5 mol%W–5 mol%Mo alloy membrane at the same temperature [11,12].
S387
H. Yukawa et al. / Journal of Alloys and Compounds 580 (2013) S386–S390
In this study, the alloying effects of molybdenum (Mo) into V–W alloy on the hydrogen solubility are investigated in a fundamental manner in order to improve further the resistance to hydrogen embrittlement. The hydrogen permeability of the V–W–Mo alloy membrane is also investigated.
Hydrogen pressure, P / MPa
V-5W-5Mo
2. Experimental procedure 2.1. Sample preparation The purity of the raw materials used in this study is 99.9 mass% for vanadium (Taiyo Koko Co., Ltd.) and 99.95 mass% for tungsten and molybdenum (Rare Metallic Co., Ltd.). V–5 mol%W–5 mol%Mo alloy is prepared by using tri-arc furnace in a purified argon gas atmosphere. According to the X-ray diffraction analysis, the alloy is confirmed to be composed of a single solid solution phase with simple BCC (bodycentered cubic) crystal structure.
0
10
-1
10
V-5W
[12]
0 .01 MPa
Pure V
[14]
-2
10
673K -3
10
2.2. Hydrogen pressure–composition–isotherm (PCT) measurement In order to examine the hydrogen solubility, the pressure–composition–isotherms (PCT) are measured by using a Sieverts-type apparatus. A small piece of the alloy is set into a cell and then evacuated with turbo molecular pump (TMP) system. Subsequently, it is heated up to 773 K, and then high purity hydrogen (99.99999% purity) of about 5 MPa is introduced and cooled down to room temperature. This process is repeated at least three times prior to the PCT measurement in order to activate the sample surface for the hydrogen absorption and desorption reactions to take place smoothly. The PCT curves are measured at 673–773 K up to about 5 MPa.
1
10
0.0
0.1
0.2
0.3
0.4
0.5
0.6
Hydrogen content, content C / (H/M) Fig. 1. PCT curve for V–5 mol%W–5 mol%Mo alloy measured at 673 K. PCT curves for pure V [14] and V–5 mol%W alloy [12] measured at 673 K are also drawn in the figure for comparison.
1
10
723K
The hydrogen permeation tests are performed at 673–773 K in order to evaluate the hydrogen permeability. Disk-shaped specimens of about / 12 mm in diameter with a thickness of about 0.5 mm are prepared. They are polished mechanically by using alumina abrasive paper followed by the final polishing with 0.3 lm Al2O3 powders. Subsequently, pure palladium of about 200 nm in thickness is deposited at 573 K on both sides of the sample surfaces by using an RF magnetron sputtering apparatus. This palladium layer on the surface protects the sample from the oxidation. It also acts as a catalyst for hydrogen dissociation reaction and subsequent dissolution into metal to take place smoothly. The disk sample is set into the hydrogen permeation apparatus and then evacuated with TMP. Subsequently, it is heated up to the measuring temperature, and then a high purity hydrogen gas is introduced into both sides of the specimen. The hydrogen pressure applied to the inlet side is determined from the corresponding PCT curve at each temperature so that the hydrogen concentration does not exceed the critical value of the ductile-to-brittle transition hydrogen concentration, i.e., 0.2 (H/M), to avoid the sample cracking due to hydrogen embrittlement. The outlet pressure is fixed to be 0.01 MPa. Then, the hydrogen fluxes, J, permeated through the disk samples are measured by the conventional gas permeation method using a mass flowmeter. A detailed explanation of the permeation test is given elsewhere [13].
Hydrogen pressure, P / MPa
2.3. Hydrogen permeation test
773K
673K
0
10
0.6 MPa 0. 3 MPa 0.15 MPa
-1
10
-2
10
DBTC of V
-3
Ductile
[8]
Brittle
10
0.0
0.1
0.2
0.3
0.4
0.5
0.6
Hydrogen content, C / (H/M) Fig. 2. PCT curves for V–5 mol%W–5 mol%Mo alloy measured at 673–773 K. The ductile-to-brittle transition hydrogen concentration (DBTC) for vanadium alloys [8] is indicated by a broken line in the figure.
3. Results and discussion 3.1. Alloying effects of Mo on the hydrogen solubility of V–W alloy The PCT curve for V–5 mol%W–5 mol%Mo alloy measured at 673 K is shown in Fig. 1. For comparison, the results for pure vanadium reported by Veleckis and Edwards [14] and for V–5 mol%W alloy reported by Yukawa et al. [12] are also drawn in the figure. As is evident from the figure, the PCT curve shifts toward the left and upper region by the addition of 5 mol% of molybdenum into V–5 mol%W alloy, indicating that the amount of dissolved hydrogen under a given hydrogen pressure (e.g., 0.01 MPa) decreases by the addition of molybdenum. Fig. 2 shows the PCT curves for V–5 mol%W–5 mol%Mo alloy measured at 673–773 K. As shown in Fig. 2, the PCT curve also shifts toward the left and upper region with increasing temperature. As a result, the equlilibrium hydrogen pressure at the DBTC increases with increasing temperature. For example, about 0.15 MPa, 0.3 MPa and 0.6 MPa of hydrogen pressures can be applied to V–5 mol%W–5 mol%Mo alloy membrane at 673 K, 723 K and 773 K, respectively, which is about twice
of hydrogen pressure applicable to V–5 mol%W alloy membrane [12]. 3.2. Hydrogen permeability of V–W–Mo alloy membrane The hydrogen diffusion in metal membrane is generally the rate-limiting process of the total reaction of hydrogen permeation through it. Then, the Fick’s law:
J ¼ D
@C @x
ð1Þ
is commonly applied to the metal membrane to discuss the property of it, where J is the hydrogen flux, D is the diffusion coefficient and @C/@x is the concentration gradient across the membrane. Applying the Fick’s law to a membrane with a thickness of d, the Eq. (1) can be modified as follows:
J¼D
C in C out d
ð2Þ
S388
H. Yukawa et al. / Journal of Alloys and Compounds 580 (2013) S386–S390
other hand, as shown in Fig. 1, the hydrogen concentration at the outlet side is smaller for V–5 mol%W–5 mol%Mo alloy than V– 5 mol%W alloy at the same pressure, e.g., 0.01 MPa. Then, the hydrogen concentration difference between the inlet and outlet sides, Cin Cout, is increased by the addition of molybdenum into V–W alloy. As a result, higher hydrogen flux can be obtained for V–W–Mo alloy than V–W alloy, following Eq. (2). In fact, the hydrogen flux for V–5 mol%W–5 mol%Mo alloy measured at 773 K under the pressure condition of inlet/outlet = 0.6/0.01 MPa is higher than that for V–5 mol%W alloy measured under the pressure condition of inlet/outlet = 0.3/0.01 MPa. In addition, it is about four times higher than that for Pd–25 mol%Ag alloy estimated for 773 K under the same pressure condition, i.e., inlet/outlet = 0.6/ 0.01 MPa. Similar results are also obtained at 723 K and 673 K as shown in Fig. 3b and c, respectively. It is important to note here that the hydrogen diffusivity is also an important factor controlling the hydrogen permeability as expressed by Eq. (2). The investigation about the alloying effects on the hydrogen diffusion coefficient under the practical condition of hydrogen permeation is now underway. After the hydrogen permeation test, the system is evacuated. Then the gas leak test is performed using helium (He) in order to check the existence of any cracks on the disk sample. A photo image of the sample after the hydrogen permeation test is shown in the inset of Fig. 3a. It is confirmed that there is no evidence of crack on the sample due to hydrogen embrittlement. Thus, V–5 mol%– 5 mol%Mo alloy membrane possesses high hydrogen permeability
120
(a)
ol H m-1 S-1 Hydrogen flux, 10 J d / mo
773K
100
80
[12]
V-5W (0.3/0.01)
60
100
(b)
723K
80
60
0.3 / 0.01
0.1 / 0.01
0 15 / 0 0.15 0.01 01
0 05 / 0 0.05 0.01 01
6
6
ol H m-1 S-1 Hydrogen flux, 10 J d / mo H
where Cin and Cout are the hydrogen concentration at the inlet and outlet sides of the membrane. The steady-state hydrogen fluxes, J, are measured at 673–773 K for V–5 mol%W–5 mol%Mo alloy membrane. It is divided by the inverse of the sample thickness, 1/d, in order to estimate the normalized hydrogen flux, J d. It is noted here that atomic hydrogen flux (mol H m1 s1) is evaluated in this study, which is twice as large as gaseous hydrogen flux (mol H2 m1 s1). The time dependences of the normalized hydrogen flux, J d, during the measurement at 673–773 K are shown in Fig. 3a–c. The inlet and outlet hydrogen pressures for each measurement are indicated in the legend of the figure as (inlet/outlet (MPa)). For comparison, the corresponding values of V–5 mol%W alloy measured at 773 K [12] and Pd–25 mol%Ag alloy estimated from Ref. [15] are indicated in the figures with arrows. As shown in Fig. 3, the hydrogen flux is stable and nearly constant during each measurement. The value of the hydrogen flux changes depending on the applied hydrogen pressures. As shown in Fig. 3a, the hydrogen flux is higher for V–5 mol%W alloy than V–5 mol%W–5 mol%Mo alloy under the same pressure condition, i.e., inlet/outlet = 0.3/0.01 MPa, meaning that the hydrogen permeability is decreased by alloying molybdenum with V–W alloy. However, when considering the ductile-to-brittle transition hydrogen concentration, DBTC, higher hydrogen pressure can be applied to V–5 mol%W–5 mol%Mo alloy and the maximum hydrogen concentration at the inlet side become the same, i.e., 0.2 (H/M), for both V–5 mol%W and V–5 mol%W–5 mol%Mo alloys. On the
40
[15]
Pd-25Ag (0.6/0.01)
0.6 / 0.01 0.56 / 0.01 0.5 / 0.01
0.3 / 0.01 0.1 / 0.01 0.05 / 0.01
20
0 0
100
200
300
500
400
600
40
20 Pd-25Ag [15] (0.3/0.01)
0
700
100
0
200
300
400
500
600
700
Time , t / s
70
(c)
60
673K
50 40
6
ol H m-1 S-1 Hydrogen flux, 10 J d / mo
Time , t / s
30
0.15 / 0.01 0.10 / 0.01
0.05 / 0.01
20 Pd-25Ag [15] (0.15/0.01)
10 0 0
100
200
300
400
500
600
700
Time , t / s Fig. 3. Time dependences of the normalized hydrogen flux, J d, during the measurements at (a) 773 K, (b) 723 K and (c) 673 K. The inlet and outlet hydrogen pressures for each measurement are indicated in the legend as (inlet/outlet (MPa)). A photo image of the sample after hydrogen permeation test is shown in the inset of the figure.
S389
H. Yukawa et al. / Journal of Alloys and Compounds 580 (2013) S386–S390
120
Hydrogen flux, 10 J d / mol H m-1 S-1
673K 0.3
723K
773K
DBTC of V [8]
V-5W-5Mo
100
80
60
6
0.2
0.0
0.2
0.3 MPa
0.0
0.6 MPa
0.1 0.15 MPa
Hydrogen content, C / (H/M)
0.4
0 .4
0 .6
673K
40
723K 773K
20
0
0 .8
1. 0
0
1. 2
100
200
300
Δ
Square root of hydrogen pressure , P / MPa1/2
400
500
600
700
800
Pa1/2 P/P
Fig. 5. Correlation between the normalized hydrogen flux, J d, and the difference pffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffi of the square root of hydrogen pressures, P inlet P outlet , at 673–773 K.
Fig. 4. Correlation between the square root of hydrogen pressure and the hydrogen content for V–5 mol%W–5 mol%Mo alloy measured at 673–773 K. -6
-1 -1
3.3. Analysis of hydrogen permeation coefficient When the Sieverts’ law is valid in the metal-hydrogen system, the hydrogen flux expressed by Eq. (2) can be written as follows:
J ¼DK
pffiffiffiffiffiffi pffiffiffiffiffiffiffiffi Pin Pout d
Permeability, e o H 2 m s Pa a φ / mol
together with stronger resistance to hydrogen embrittlement than V–5 mol%W alloy.
-1/2
10
ð3Þ
where K is the Sievert’s solubility constant, Pin and Pout are the inlet and outlet hydrogen pressures. The product of D and K is defined as the hydrogen permeation coefficient, /(=D K), and commonly used as a measure to show the hydrogen permeability of metals and alloys. The PCT curves shown in Fig. 2 are redrawn in Fig. 4 in which the hydrogen concentration is plotted as a function of the square root of hydrogen pressure. It is evident that the Sieverts’ law is almost valid for V–5 mol%W–5 mol%W alloy at 673–773 K below the DBTC, i.e., 0.2 (H/M). Then, the hydrogen fluxes shown in Fig. 3 are analyzed in view of the difference of the square root of hydrogen pressure, and the results are shown in Fig. 5. As shown in the figure, there is almost a linear relationship between the normalized hydrogen flux, J d, and the difference of the square root of pffiffiffiffiffiffi pffiffiffiffiffiffiffiffi hydrogen pressure, Pin Pout , for each measuring temperature, indicating that the hydrogen permeation reaction through the V–5 mol%W–5 mol%Mo alloy membrane takes place following Eq. (3). Then, the hydrogen permeation coefficients, /, are estimated from the slope of the lines shown in Fig. 5 and the results are plotted in Fig. 6 as a function of the inverse of temperature. For comparison, the hydrogen permeation coefficients for Pd–25mass%Ag [15], V–15 mol%Ni [16] and Nb–30 mol%Ti–30 mol%Ni [17] alloys are also presented in the figure. As shown in Fig. 6, the hydrogen permeability of V–5 mol%W–5 mol%Mo alloy is about 4–5 times higher than that for Pd–25mass%Ag, V–15 mol%Ni and Nb–30 mol%Ti– 30 mol%Ni alloys at 673–773 K. As shown in Fig. 6, the hydrogen permeation coefficient for V–5 mol%W–5 mol%Mo alloy estimated by the present permeation method increases with decreasing temperature at 673–773 K. This trend is in good agreement with that for pure vanadium calculated from the hydrogen solubility and diffusivity [7]. It is noted here that the temperature dependence of the hydrogen permeation
-7
10
V-15Ni
[16]
-8
10
Pd-25Ag [15] Nb40 - Ti 30 - Ni 30
(Duplex phase alloy)
-9
[17]
10
1.2
1.4
1.6
1.8
1000/T / K
2.0
2.2
2.4
-1
Fig. 6. Temperature dependence of the hydrogen permeation coefficient for V– 5 mol%W–5 mol%Mo alloy. The reported values for Pd–25mass%Ag [15], V– 15 mol%Ni [16] and Nb–30 mol%Ti–30 mol%Ni [17] alloys are also plotted in the figure for comparison.
coefficient for V-based alloy is different from that for Nb-based alloy. For example, the hydrogen permeation coefficient for Nb– 5 mol%W–5 mol%Mo alloy estimated from the same permeation method with the present study decreases with decreasing temperature at 673–773 K [18].
4. Summary The alloying effects of molybdenum on the hydrogen solubility and the hydrogen permeability of V–W alloy are investigated quantitatively. The hydrogen solubility decreases by the addition of molybdenum into V–W alloy or by increasing temperature. As a result, the resistance to hydrogen embrittlement is improved by reducing the dissolved hydrogen concentration in the alloy. In fact, V–5 mol%W–5 mol%Mo alloy possesses excellent hydrogen permeability without showing any hydrogen embrittlement when used under appropriate permeation conditions.
S390
H. Yukawa et al. / Journal of Alloys and Compounds 580 (2013) S386–S390
Acknowledgement This research was supported in part by the Japan Society for the Promotion of Science (JSPS). References [1] [2] [3] [4]
S.N. Paglieri, J.D. Way, Sep. Purif. Methods 31 (2002) 1–169. V. Mordkovich, Y.K. Baichtock, M. Sosna, Platinum Met. Rev. 36 (1992) 90–97. R. Buxbaum, A. Kinney, Ind. Eng. Chem. Res. 35 (1995) 530–537. C. Nishimura, M. Komaki, S. Hwang, M. Amano, J. Alloys Comp. 330–332 (2002) 902–906. [5] K. Hashi, K. Ishikawa, T. Matsuda, K. Aoki, J. Alloys Comp. 368 (2004) 215–220. [6] T. Nambu, K. Shimizu, Y. Matsumoto, R. Rong, H. Yukawa, M. Morinaga, I. Yasuda, J. Alloys Comp. 446–447 (2007) 588–592. [7] S.A. Steward, Lawrence Livemore National Laboratory Reports, UCRL-53441, 1983.
[8] Y. Matsumoto, H. Yukawa, T. Nambu, Metall. J. LXIII (2010) 74–78. [9] H. Yukawa, T. Nambu, Y. Matsumoto, N. Watanabe, G.X. Zhang, M. Morinaga, Mater. Trans. 49 (2008) 2202–2207. [10] N. Watanabe, H. Yukawa, T. Nambu, Y. Matsumoto, G.X. Zhang, M. Morinaga, J. Alloys Comp. 477 (2009) 851–854. [11] Y. Awakura, T. Nambu, M. Matsumoto, H. Yukawa, J. Alloys Comp. 509S (2011) S877–S880. [12] H. Yukawa, T. Nambu, M. Matsumoto, M. Morinaga, J. Alloys Comp. 509S (2011) S881–S884. [13] T. Nambu, N. Shimizu, H. Ezaki, H. Yukawa, M. Morinaga, J. Jpn. Inst. Met. 69 (2005) 841–847. [14] E. Veleckis, R.K. Edwards, J. Phys. Chem. 73 (1969) 683–692. [15] E. Serra, M. Kemali, A. Perujo, D.K. Ross, Metall. Mater. Trans. A 29A (1998) 1023–1028. [16] C. Nishimura, M. Komaki, S. Hwang, M. Amano, Int. J. Hydrogen Energy 330– 332 (2002) 902–906. [17] K. Ishikawa, S. Tokui, K. Aoki, Intermetallics 17 (2009) 109–114. [18] H. Yukawa, T. Nambu, Y. Matsumoto, Defect Diffusion Forum 333 (2013) 61– 71.