Improvement of hydrogen-storage properties of Mg by reactive mechanical grinding with Fe2O3

Improvement of hydrogen-storage properties of Mg by reactive mechanical grinding with Fe2O3

Journal of Alloys and Compounds 396 (2005) 264–268 Improvement of hydrogen-storage properties of Mg by reactive mechanical grinding with Fe2O3 Ik-Hyu...

317KB Sizes 0 Downloads 51 Views

Journal of Alloys and Compounds 396 (2005) 264–268

Improvement of hydrogen-storage properties of Mg by reactive mechanical grinding with Fe2O3 Ik-Hyun Kwon a , Jean-Louis Bobet b , Jong-Soo Bae c , Myoung-Youp Song a, ∗ a

Division of Advanced Materials Engineering, Research Center of Advanced Materials Development, Engineering Research Institute, Chonbuk National University, 664-14 1ga Deogjindong Deogjingu Jeonju, Chonbuk 561-756, Republic of Korea b Institut de Chimie de la Mati` ere Condens´ee de Bordeaux (ICMCB), CNRS UPR 9048, Universit´e de Bordeaux I, 87 Avenue du Dr. A. Schweitzer, 33608 Pessac Cedex, France c Nanopowder Materials Group, Materials Technology Department, Korea Institute of Machinery and Materials, 66 Sangnamdong, Changwon, Kyungnam, Republic of Korea Received 6 December 2004; accepted 20 December 2004 Available online 9 February 2005

Abstract We tried to improve the H2 -sorption properties of Mg by mechanical grinding under H2 (reactive mechanical grinding) with Fe2 O3 under various milling conditions. The sample Mg–10 wt.%Fe2 O3 prepared by milling at a revolution speed of 250 rpm for 24 h has the best hydrogenstorage properties. It absorbs 5.05 wt.% hydrogen at 593 K under 12 bar H2 for 60 min at the first cycle. Its activation is accomplished after three hydriding–dehydriding cycles. The activated sample absorbs 4.22 wt.% hydrogen at 593 K, 12 bar H2 for 10 min. The reactive grinding of Mg with Fe2 O3 increases the H2 -sorption rates by facilitating nucleation (by creating defects on the surface of the Mg particles and by the additive), by making cracks on the surface of Mg particles and reducing the particle size of Mg and thus by shortening the diffusion distances of hydrogen atoms. Hydriding–dehydriding cycling also increases the H2 -sorption rates by creating defects on the surface of the Mg particles, and by making cracks on the surface of Mg particles and reducing the particle size of Mg. © 2005 Elsevier B.V. All rights reserved. Keywords: H2 -sorption properties of Mg; Fe2 O3 addition; Reactive mechanical grinding; Hydriding and dehydriding rates; BET surface area

1. Introduction Magnesium has many advantages for a hydrogen-storage material: large hydrogen-storage capacity (7.6 wt.%), low cost and abundance in the earth’s crust. But its hydriding and dehydriding kinetics are very slow [1]. A lot of work to ameliorate the reaction kinetics of magnesium with hydrogen has been done by alloying certain metals with magnesium [2–10], by mixing metal additives with magnesium [11], by plating nickel on the surface of magnesium [12]. Song [13] reviewed the kinetic studies of the hydriding and the dehydriding reactions of Mg. Many works do not agree with one another on the rate-controlling step(s) for hy∗

Corresponding author. Tel.: +82 63 270 2379; fax: +82 63 270 2386. E-mail address: [email protected] (M.-Y. Song).

0925-8388/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2004.12.036

driding or dehydriding of magnesium. However, there is no contradiction in the points that the hydriding and dehydriding reactions of Mg are nucleation-controlled under certain conditions and progress by a mechanism of nucleation and growth, and that the hydriding rates of Mg are controlled by the diffusion of hydrogen through a growing Mg hydride layer. The hydriding and dehydriding kinetics of Mg can be improved, therefore, by a treatment such as mechanical alloying [14–16] which can facilitate nucleation by creating many defects on the surface and/or in the interior of Mg, or by the additive acting as active sites for the nucleation, and shorten diffusion distances by reducing the effective particle sizes of Mg. Bobet et al. [17] reported that mechanical alloying in H2 (reactive mechanical grinding) for a short time (2 h) is an effective way to improve strongly the hydrogen-storage proper-

I.-H. Kwon et al. / Journal of Alloys and Compounds 396 (2005) 264–268

265

ties of both magnesium and Mg + 10 wt.%Co, Ni or Fe mixtures. The oxides are brittle, and thus they may be pulverized during mechanical grinding. The added oxides and/or their pulverization during mechanical grinding may help the particles of magnesium to become finer. In this work, we chose Fe2 O3 as the oxide to add to Mg. Mg–10 wt.%Fe2 O3 alloys were prepared by mechanical grinding under H2 (reactive mechanical grinding) in a planetary ball mill under various conditions (revolution speeds and milling times), and their hydrogen-storage properties were investigated.

2. Experimental Fig. 1. X-ray powder diffraction patterns of Mg–10 wt.%Fe2 O3 as prepared by milling at various revolution speeds for different times.

Pure Mg powder (particle size 50 ␮m) was mixed with 10 wt.%Fe2 O3 (total weight = 8 g) in a stainless steel con-

Fig. 2. Microstructures observed by SEM for the sample Mg–10 wt.%Fe2 O3 as prepared by milling at various revolution speeds for different times (a) 200 rpm, 1 h; (b) 250 rpm, 2 h; (c) 250 rpm, 4 h; (d) 250 rpm, 6 h; (e) 250 rpm, 8 h; (f) 250 rpm, 12 h; and (g) 250 rpm, 24 h.

266

I.-H. Kwon et al. / Journal of Alloys and Compounds 396 (2005) 264–268

tainer (with 21 hardened steel balls, total weight = 72 g) closed in a hermetic way. The average particle size of Fe2 O3 was smaller than 5 ␮m. All the handlings were performed in a glove box under Ar in order to prevent oxidation. The mill container was then filled with high-purity hydrogen gas (≈10 bar). The disc revolution speed was 200 and 250 rpm. The hydriding apparatus has been described previously [18]. The absorbed or desorbed hydrogen quantity was measured as a function of time by a volumetric method. X-ray diffraction (XRD) analysis was carried out for the as-milled powders and for the samples after hydriding–dehydriding cycling. The microstructures were observed by SEM (scanning electron microscope). BET (Brunauer, Emmett and Teller) surface areas of the as-milled powders were measured at 77 K using a Micromeritics ASAP 2020 automatic analyzer. The samples were outgassed at 573 K for 6 h before measuring the surface areas. Fig. 3. Ha versus t curves at 593 K under 12 bar H2 for the first cycle of Mg–10 wt.% Fe2 O3 prepared at various revolution speeds for different times.

3. Results and discussions Fig. 1 shows the X-ray (Cu K␣) powder diffraction patterns of Mg–10 wt.%Fe2 O3 as prepared by milling at a revolution speed of 200 rpm for 1 h and at the revolution speed of 250 rpm for 2, 4, 6, 8, 12 and 24 h. The XRD patterns are similar and the peaks become broader as the milling time increases. The peaks for the sample prepared at 250 rpm for 24 h are, in particular, very weak and broad. We believe that the peak broadening results from the decrease in the particle size as the milling time increases. The samples as prepared contain Mg, Fe2 O3 and a small amount of MgH2 . Fig. 2 gives the microstructures observed by SEM for the sample Mg–10 wt.%Fe2 O3 as prepared by milling at a revolution speed of 200 rpm for 1 h and at the revolution speed of 250 rpm for 2–24 h. The samples contain small particles and large particles. As the milling time increases, the number of small particles increases. The sample prepared at 250 rpm for 24 h has very small particles. Fig. 3 shows the variations of weight percentage of hydrogen Ha absorbed at 593 K under 12 bar H2 for the first cycle by Mg–10 wt.%Fe2 O3 prepared at various revolution speeds for different times. The percentage of absorbed hydrogen Ha is expressed with respect to the sample weight. The sample prepared by milling at 200 rpm for 1 h does not absorb hydrogen under these conditions. The samples milled at 250 rpm have the higher hydriding rates as the milling time increases. The sample milled at 250 rpm for 24 h has the highest hy-

driding rates. It absorbs 5.05 wt.% hydrogen at 593 K under 12 bar H2 for 60 min. We studied the hydrogen-storage properties of the Mg–10 wt.%Fe2 O3 milled at 250 rpm for 24 h in detail. Fig. 4 shows Ha versus t curves at 593 K and 12 bar H2 for the activated Mg–10 wt.%Fe2 O3 prepared under various milling conditions. The sample milled at 200 rpm for 1 h does not absorb hydrogen. The samples milled at 250 rpm have the higher hydriding rates as the milling time increases. The sample milled at 250 rpm for 24 h has the highest hydriding rates. It absorbs 4.22 wt.% hydrogen at 593 K, 12 bar H2 for 10 min. Table 1 gives the values of Ha (wt.%) at 593 K, 12 bar H2 as a function of time for the activated samples. Fig. 5 shows the variations of Ha absorbed by Mg–10 wt.%Fe2 O3 milled at 250 rpm for 24 h as a function of time t (min) under 12 bar H2 at 593 K according to the number of hydriding–dehydriding cycles n. As the number of hydriding–dehydriding cycles increases, the hydriding rates increase in the beginning of the hydriding reaction. We believe that the activation is accomplished after the third hydriding cycle. The Ha value after 60 min is 5.20 wt.% at n = 3. Fig. 6 shows the variations with the milling time of the quantity of hydrogen Ha (60 min) absorbed at 593 K under 12 bar H2 after activation and the BET surface area. Ha

Table 1 Ha (wt.%) at 593 K under 12 bar Ha for the activated Mg–10 wt.% Fe2 O3 prepared at various revolutions speeds for different times Time (min)

5 10 20 60

Milling conditions 200 rpm, 1 h

250 rpm, 2 h

250 rpm, 4 h

250 rpm, 6 h

250 rpm, 8 h

250 rpm, 12 h

250 rpm, 24 h

0.00 0.00 0.00 0.00

0.08 0.11 0.19 0.46

0.83 1.08 1.38 2.13

1.74 2.15 2.58 3.43

2.62 3.08 3.51 4.32

2.94 3.44 3.91 4.64

3.31 4.22 4.65 5.04

I.-H. Kwon et al. / Journal of Alloys and Compounds 396 (2005) 264–268

267

Fig. 6. Variations of the quantity of hydrogen Ha (60 min) absorbed at 593 K under 12 bar H2 for 60 min after activation and BET surface area with the milling time. Fig. 4. Ha versus t curves at 593 K under 12 bar H2 for the activated Mg–10 wt.% Fe2 O3 prepared at various revolution speeds for different times.

(60 min) and the BET surface area increase as the milling time becomes longer. The reactive grinding of Mg with Fe2 O3 increases the hydriding rates by facilitating nucleation (by creating defects on the surface of the Mg particles and by the additive) and by reducing the particle size of Mg and thus by shortening the diffusion distances of hydrogen atoms. Fig. 7 shows the Ha versus t curves for the activated Mg–10 wt.%Fe2 O3 at 593 K under various hydrogen pressures. As the hydrogen pressure increases, the hydriding rate becomes higher. From the results of hydriding rates and microstructure, it can be inferred that the hydriding–dehydriding cycling also increases the H2 -sorption rates by facilitating nucleation (by creating defects on the surface of the Mg particles), by making cracks on the surface of Mg particles and reducing the particle

Fig. 7. Ha versus t curves for the activated Mg–10 wt.%Fe2 O3 , prepared at a revolution speed 250 rpm for 24 h, at 593 K under various hydrogen pressures.

size of Mg and thus by shortening the diffusion distances of hydrogen atoms.

4. Conclusions

Fig. 5. Variations of H2 absorbed by Mg–10 wt.%Fe2 O3 , prepared at a revolution speed 250 rpm for 24 h, as a function of time t under 12 bar H2 at 593 K according to the number of hydriding cycles n.

The samples Mg–10 wt.%Fe2 O3 were prepared by mechanical grinding under H2 (reactive mechanical grinding) at various revolution speeds for different times. The sample Mg–10 wt.%Fe2 O3 prepared by milling at a revolution speed of 250 rpm for 24 h has the best hydrogen-storage properties. It absorbs 5.05 wt.% hydrogen at 593 K under 12 bar H2 for 60 min at the first cycle. Its activation is accomplished after three hydriding–dehydriding cycles. The activated sample absorbs 4.22 wt.% hydrogen at 593 K, 12 bar H2 for 10 min. The reactive grinding of Mg with Fe2 O3 increases the H2 sorption rates by facilitating nucleation (by creating defects on the surface of the Mg particles and by the additive), by

268

I.-H. Kwon et al. / Journal of Alloys and Compounds 396 (2005) 264–268

making cracks on the surface of Mg particles and reducing the particle size of Mg and thus by shortening the diffusion distances of hydrogen atoms. Hydriding–dehydriding cycling also increases the H2 -sorption rates by creating defects on the surface of the Mg particles, and by making cracks on the surface of Mg particles and reducing the particle size of Mg. Acknowledgements This research was performed for the Hydrogen Energy R&D Center, one of the 21st Century Frontier R&D Program, funded by the Ministry of Science and Technology. References [1] [2] [3] [4] [5]

A. Vose, Metal Hydrides, U.S. Patent 2944 (1961), p. 587. J.J. Reilly, R.H. Wiswall, Inorg. Chem. 6 (12) (1967) 2220. J.J. Reilly, R.H. Wiswall Jr., Inorg. Chem. 7 (11) (1968) 2254. D.L. Douglass, Metall. Trans. A 6 (1975) 2179. M.H. Mintz, Z. Gavra, Z. Hadari, J. Inorg. Nucl. Chem. 40 (1978) 765.

[6] M. Pezat, A. Hbika, B. Darriet, P. Hagenmuller, French Anvar Patent 7,820,382 (1978); Mater. Res. Bull. 14 (1979) 377. [7] M. Pezat, B. Darriet, P. Hagenmuller, J. Less-Common Met. 74 (1980) 427. [8] Q. Wang, J. Wu, M. Au, L. Zhang, Hydrogen energy progress V, in: T.N. Veziro˘glu, J.B. Taylor (Eds.), Proceedings of the Fifth World Hydrogen Energy Conference, Toronto, Canada, Vol. 3, Pergamon Press, New York, 1984, pp. 1279–1290. [9] E. Akiba, K. Nomura, S. Ono, S. Suda, Int. J. Hydrogen Energy 7 (10) (1982) 787. [10] J.M. Boulet, N. Gerard, J. Less-Common Met. 89 (1983) 151. [11] B. Tanguy, J.L. Soubeyroux, M. Pezat, J. Portier, P. Hagenmuller, Mater. Res. Bull. 11 (1976) 1441. [12] F.G. Eisenberg, D.A. Zagnoli, J.J. Sheridan III, J. Less-Common Met. 74 (1980) 323. [13] M.-Y. Song, J. Mater. Sci. 30 (1995) 1343. [14] M.-Y. Song, E.I. Ivanov, B. Darriet, M. Pezat, P. Hagenmuller, Int. J. Hydrogen Energy 10 (3) (1985) 169. [15] M.-Y. Song, E.I. Ivanov, B. Darriet, M. Pezat, P. Hagenmuller, J. Less-Common Met. 131 (1987) 71. [16] M.-Y. Song, Int. J. Hydrogen Energy 20 (3) (1995) 221. [17] J.-L. Bobet, E. Akiba, Y. Nakamura, B. Darriet, Int. J. Hydrogen Energy 25 (2000) 987. [18] M.-Y. Song, D.S. Ahn, I.-H. Kwon, H.J. Ahn, Met. Mater. Int. 5 (5) (1999) 485.