Format And Type Fonts CHEMICAL ENGINEERING TRANSACTIONS VOL. 39, 2014 A publication of The Italian Association of Chemical Engineering www.aidic.it/cet Guest Editors: Petar Sabev Varbanov, Jiří Jaromír Klemeš, Peng Yen Liew, Jun Yow Yong Copyright © 2014, AIDIC Servizi S.r.l., ISBN 978-88-95608-30-3; ISSN 2283-9216 DOI:10.3303/CET1439201 Please cite this article as: Rangsunvigit P., Sridechprasat P., Kitiyanan B., Kulprathipanja S., 2014, Effects of TiCl3, TiO2, ZrCl4, and ZrO2 on hydrogen desorption of MgH2 and its reversibility, Chemical Engineering Transactions, 39, 1201-1206 DOI:10.3303/CET1439201 1201 Effects of TiCl3, TiO2, ZrCl4, and ZrO2 on Hydrogen Desorption of MgH2 and Its Reversibility Pramoch Rangsunvigit* a , Pattaraporn Sridechprasat a , Boonyarach Kitiyanan a , Santi Kulprathipanja b a The Petroleum and Petrochemical College, Chulalongkorn University, Bangkok, Thailand b UOP, A Honeywell Company, Des Plaines, Illinois, USA Pramoch.r@chula.ac.th A thorough investigation on hydrogen desorption of MgH2 and its reversibility with the focus on effects of a catalyst (TiCl3, TiO2, ZrCl4, and ZrO2) were reported. The experiments were carried out using a Sievert’s type apparatus. The reaction kinetics and the phase transformation behaviours were examined using a differential scanning calorimeter (DSC) and an X-ray diffractometer (XRD). It was found that the catalysts play an important role in facilitating smaller MgH2 crystallite size formation during ball-milling and accelerating hydride formation during the subsequent hydrogen absorption cycles resulting in the hydrogen desorption and reversibility enhancement of MgH2. Doping with ZrCl4 resulted in the smallest MgH2 crystallite size. That, in turn, resulted in the lowest hydrogen desorption temperature and the highest amount of released hydrogen in the first desorption. However, TiO2 was identified as the best catalyst because doping with TiO2 provided the highest amount of reversible hydrogen (~10% capacity drop from initial cycle). 1. Introduction Hydrogen is regarded as the most promising alternative energy source because it is abundance and produces virtually water as a product from fuel cells. However, to use hydrogen as a fuel for vehicles, it requires a very compact and safe storage system (Conte et al., 2004). As compared with traditional hydrogen storage, e.g. compressed hydrogen gas and liquid hydrogen, solid-state hydrides are preferable because it has high gravimetric and volumetric hydrogen densities, and it is also reversible at appropriate conditions (Schüthet al., 2004). This statement is in agreement as reported by Züttel (2004). David (2005) also reported that the solid-state hydrides or metal hydridesare the most important characteristics of materials for hydrogen storage. Magnesium hydride (MgH2) is one of potential solid-state hydride materials because it can reversibly store 7.6 wt% H2 and has relatively high volumetric hydrogen density (106 kg H2 m -3 ) with reasonable cost (Lu et al., 2009). However, it releases hydrogen at a high temperature and posses slow hydrogen desorption and absorption kinetics (Sakintuna et al., 2007). Typical methods to destabilize MgH2 for hydrogen desorption are ball-milling and doping with catalyst (Xie et al., 2009). Ball-milling is the simplest method that would potentially reduce crystallite size of MgH2, increase defects, and decrease hydrogen diffusion distance within MgH2 bulk (Fátayet al., 2005). These results in the faster hydrogen desorption kinetics and higher amount of desorbed hydrogen. Oelerichet al. (2001a) revealed that the catalystdoping would also enhance the hydrogen dissociation and diffusion in MgH2 by synergically accelerating the desorption kinetics. And the effects of catalyst doping are confirmed by Hanada et al. (2006) and Varin et al. (2011). There are many types and forms of a catalyst such as transition metal and transition metal compound of oxides and chlorides. Their effects on the hydrogen storage improvement of MgH2 were reported by several research groups. For the use of a transition metal catalyst, Liang et al. (1999) found that doping 5 mol% 3d-transition metals nanocomposite (Ti, V, Mn, Fe, and Ni) did not change the enthalpy and the entropy of formation of MgH2, but it dramatically reduced the activation energy or increased the reaction kinetics of MgH2. MgH2 doped with Ti exhibited the most rapid absorption kinetics. Zaluska et al. (1999) 1202 also reported some enhancement of hydrogen storage properties by transition metal doping in the ball- milled MgH2. The doping of 5 wt% Zr significantly reduced the hydrogen desorption temperature of MgH2, while the doping of 8 wt% V markedly enhanced the hydrogen absorption kinetics. In the case of transition metal compounds, Malka et al. (2010) found that the lowest desorbed temperature was achieved by doping MgH2 with either 7 wt% TiCl3 or ZrF4 and different metal compounds, e.g. chloride and halide, provide different desorption temperatures. Typically, a transition metal compound catalyst (e.g. oxide and nitride of a transition metal) posses better catalytic activity than a transition metal catalyst due to their smaller particle sizes and higher surface areas (Barhordarian et al., 2003). In addition, the transition metal compound catalyst has multiple valence states, which provides larger catalytic effects. Oelerich et al. (2001b) found that MgH2 doped with V2O5 yielded the fastest hydrogen desorption, followed by the doping of VN, VC, and V. Furthermore, the doping of V did not show any kinetics improvement. Bhat et al. (2006) showed that doping a catalyst resulted in significant particle size reduction, and MgH2 doped with NbCl5 and CaF2 showed better catalytic activity than doping with Nb2O5. However, to our knowledge, the reversibility of such MgH2 doped with transition metal compounds has not yet been reported. In this work, effects of ball-milling and transition metal compound catalysts, including TiCl3, TiO2, ZrCl4, and ZrO2, on the hydrogen desorption temperature, the amount of released hydrogen, and the reversibility of MgH2 were investigated. Phase transformation, calculated crystallite size, and activation energy of the samples were also reported. 2. Experimental The starting metal hydride, MgH2 (90 % and Mg 10 %, Acros Organics), and catalysts, TiCl3 (prepared by vacuum drying of 12% TiCl3 solution in hydrochloric acid), TiO2 (Degüssa P25), ZrCl4 (99.99 %, Sigma Aldrich), and ZrO2 (98.5 %, Riedel-de Haën), were used without further purification. All sample handlings were done in a glovebox filled with liquid nitrogen to prevent contamination from air and moisture. MgH2 samples were prepared from a 1:0.03 molar ratio of MgH2 and a catalyst using a centrifugal ball mill (Retsch S100) at a speed of 300 rpm under nitrogen atmosphere for 5 h. The weight ratio of stainless steel balls (316SS) to powder was 60:1. The hydrogen desorption temperature and capacity as well as the reversibility of the sample were examined in the Sievert’s type apparatus. Approximately, 0.3 g of a milled sample was packed into the stainless steel vial reactor, where two calibrated K-type thermocouples (Cole Parmer) were placed to monitor and control the temperature throughout the desorption. A pressure transducer (Cole Parmer, model 68073-68074) was employed to monitor the pressure change. The hydrogen desorption was carried out from room temperature to 400C with a heating rate of 2C min -1 under 0.1 MPa H2 initial pressure. The temperature was held at 400C until no pressure change in the system was observed. The released hydrogen in the hydrogen desorption was calculated using the real gas law. It should be pointed out that the amount of a doped catalyst was not taken into account when the amount of released hydrogen was calculated. After the hydrogen desorption, the sample was compressed under 6.0 MPa H2 and 350C for 12 h for the hydrogen absorption. Both hydrogen desorption and absorption were repeated to investigate the reversibility of the sample. Sample identification was performed using a Rigaku X-ray diffractometer (XRD, model DMAX 2200 HV) at room temperature over a range of 25 to 75 with Cu K radiation (40 kV, 30 mA). The crystallite size of the sample was averaged from the first four highest peak intensities, calculated using the Scherrer equation. The activation energy of the sample was determined from the Kissinger method. The hydrogen desorption temperature at the maximum peak (TP) of each experiment was measured using a differential scanning calorimeter (DSC 822, Mettler Toledo). About 1.0 mg of a milled sample was placed into an Al crucible. The crucible was heated from 100 to 400 C using different heating rates of 2, 5, 8, and 10 C min -1 . 3. Results and discussion Effects of ball-milling and the catalyst on phase transformation and crystallite size of MgH2 were examined by XRD. As illustrated in Figure 1, MgH2 appears to be the major phase in all samples. Differences in the height and width of the MgH2 peaks indicate different quantities and crystallite sizes of MgH2 after ball- milling and doping with different catalysts. In addition, the XRD pattern of the as-received sample (Figure 1(a)) indicates the presence of Mg and Mg(OH)2 impurities. The diffraction peaks of these impurities disappear after ball-milling, while there are new MgO peaks observed for all un-doped and doped samples. This would possibly be caused by (i) the decrease in the crystallite sizes of impurities, (ii) the reaction between remaining oxygen and oxide in the system to form thermodynamically stable MgO, and (iii) the decomposition of Mg(OH)2 to MgO, which corresponds to the report of Hout et al. (2001). They showed 1203 that the high energy produced during ball-milling reduced the decomposition temperature of Mg(OH)2 to MgO. The MgH2 peaks after ball-milling (or un-doped MgH2) are lower and broader than those of the as- Figure 1: XRD patterns of (a) as-received MgH2, (b) un-doped MgH2, MgH2 doped with (c) TiCl3, (d) TiO2, (e) ZrCl4, and (f) ZrO2 after ball-milling for 5 h received sample, indicating partial decomposition and reduction in the crystallite size of MgH2, . The calculated crystallite size using Scherrer equation shows that the size of MgH2 is reduced from 46.12 to 35.57 nm after ball-milling (Uvarov and Popov, 2007). Except those of the ZrO2 doped sample, the XRD patterns of the doped MgH2 samples (Figures 1(c)-(f)) possess only diffraction peaks of MgH2 and MgO without those of the catalysts. The calculated MgH2 crystallite sizes (after ball-milling) of ZrCl4, TiO2, TiCl3, and ZrO2 doped samples are 20.76, 21.55, 24.06, and 28.12 nm, , which are smaller than that of the un-doped sample. This indicates that doping MgH2 with the catalysts results in smaller MgH2 crystal formation, and different catalysts seem to reduce the MgH2 crystallite sizes in different extents. Hydrogen desorption temperature, amount of desorbed hydrogen, and reversibility of the MgH2 samples were investigated by using a Sievert’s type apparatus over a temperature range of 50-400 C. The hydrogen desorption profiles in the first to forth desorption of un-doped MgH2 and MgH2 doped with TiCl3, TiO2, ZrCl4, and ZrO2 after ball-milling for 5 h are shown in Figures 3(a)-(e). The total amounts of released hydrogen and desorption temperatures are compared in Figures 2(b)-(e). For the first desorption cycle, all samples initially release hydrogen at approximately 100 C (Figure 3). This low desorption temperature may be possibly from MgH2 crystal distortion created during ball-milling. A significant release of hydrogen can be noticed at a temperature higher than 300 C (Figure 2(b)), while doping with catalysts seems to facilitate the release at a lower temperature. It can be clearly seen that the desorption temperature is reduced from 370 C for the un-doped sample to 325, 335, 340, and 360 C for the samples doped with ZrCl4, TiO2, TiCl3, and ZrO2, . Figure 2(b) shows an interesting trend of the desorption temperature as a function of MgH2 crystallite size, which is consistence with the observation on the hydrogen desorption activation energy. That is MgH2 doped with ZrCl4 possesses the smallest crystallite size and the lowest activation energy; hence, it releases hydrogen significantly at the lowest temperature among the investigated MgH2 samples. The catalysts seem to increase the total amount of released hydrogen for the first desorption cycle. Figure 2(c) shows that the total amount of desorbed hydrogen of MgH2 (un-doped sample) is increased from 4.8 to 5.5, 5.4, 5.2, and 5.0 wt% when MgH2 is doped with ZrCl4, TiO2, TiCl3, and ZrO2, but it does not reach the theoretical value of 7.6 wt%. The figure reiterates the correlation between the desorbed hydrogen amount from the first desorption cycle and MgH2 crystal size as previously observed for the activation energy and desorption temperature. It is believed that one important role of the catalyst on the first 1204 hydrogen desorption cycle is to facilitate smaller hydride crystallite size formation after ball-milling rather than directly accelerate the hydrogen desorption kinetics. As smaller MgH2 crystal would have relatively low hydrogen desorption barriers, e.g. hydrogen diffusion from core to outer surfaces, hydrogen desorption can then take place at a lower temperature with faster kinetics, and higher released hydrogen. Hydrogen desorption behaviours in the subsequent desorption cycles of all samples are illustrated in Figures 2(d)-(e) and Figures 3(a)-(e). The results show that all the samples release hydrogen at a slightly lower temperature, while there is no hydrogen desorption at about 100C. This is an expected result for the reformulated hydride crystal under equilibrium condition without mechanical force applied. Even though the hydrogen desorption temperature of subsequent cycles still follows the initial hydride crystallite size, the amount of desorbed hydrogen shows otherwise. It was found that the total amounts of released hydrogen from MgH2 doped with TiO2 and ZrO2 are significantly higher than those of TiCl3 and ZrCl4 doped samples. Nevertheless, all catalyst doped samples still provide higher hydrogen desorption amount than un-doped MgH2. It is interesting to further note that even though un-doped MgH2 has more Mg to re-absorb hydrogen, the total hydrogen desorption amounts from the subsequent cycles of un-doped MgH2 is the lowest. This evidence suggests that the catalysts also play a role in accelerating the hydride formation during hydrogen re-absorption process. The employed oxide catalysts seem to better catalyze the hydride formation than the chlorides. In conclusion, the employed catalysts in this study seem to play at least two important roles on hydrogen desorption/adsorption of MgH2. The first role is to facilitate smaller MgH2 crystallite size formation during ball-milling, which helps to expedite the desorption kinetics for all desorption cycles. Besides, the catalysts also play a role in the reversibility of MgH2 by accelerating hydride formation that enhances the subsequent cycles absorption capability. TiO2 is identified as the best among the employed catalyst that can provide significant by faster desorption kinetics, lower desorption temperature (~20 - 35 C), and higher hydrogen desorption amount (~15 %) than those of the un-doped sample, while still possesses the hydrogen absorption reversibility (~10 % capacity drop from initial cycle). Figure 2: Relationship between crystallite size of MgH2 and (a) activation energy, (b) desorption temperature, (c) total amount of desorbed hydrogen in the first desorption, (d) desorption temperature, and (e) total amount of desorbed hydrogen in the subsequent desorption 1205 Figure 3:Hydrogen desorption profiles of (a) un-doped MgH2, MgH2 doped with (b) TiCl3, (c) TiO2, (d) ZrCl4, and (e) ZrO2 after ball-milling for 5 h (I) first, (II) second, (III) third, (IV) forth desorption, and (IV) desorption temperature 4. Conclusions In this work, effects of ball-milling and a catalyst (TiCl3, TiO2, ZrCl4, and ZrO2) on the hydrogen storage properties of MgH2 were reported. It was found that the employed catalysts can enhance both hydrogen released and reversibility of MgH2 by facilitating smaller MgH2 crystallite size formation during ball-milling and accelerating hydride formation during the subsequent hydrogen absorption cycles. Among the employed catalyst, TiO2 was identified as the best that provided significantly faster desorption kinetics, lower desorption temperature (~20-35 C), and higher hydrogen desorption amount (~15 %) than those of the un-doped sample, while still possessed the hydrogen absorption reversibility (~10 % capacity drop from initial cycle). 1206 Acknowledgements This work was supported by Royal Jubilee Ph.D. Program (Grant No. PHD/0249/2549), Thailand Research Fund;the Ratchadaphiseksomphot Endowment Fund of Chulalongkorn University(RES560530021-CC); Center of Excellence on Petrochemicals, and Materials Technology; The Petroleum and Petrochemical College, Chulalongkorn University, Thailand; and UOP, A Honeywell Company, USA. References Barhordarian G., Klassen T., Bormann R., 2003, Fast hydrogen sorption kinetics of nanocrytallite Mg using Nb2O5 as catalyst, Scripta Mater., 49, 213-217. Bhat V.V., Rougier A., Aymard L., Darok X., Nazrib G., Tarascon J.M., 2006, Catalytic activity of oxides and halides on hydrogen storage of MgH2.J. Power Sources, 159(1), 107-110. Conte M., Prosini P.P., Passerini S., 2004, Overview of energy/hydrogen storage: state-of-the-art of the technologies and prospects for nanomaterials, Mater. Sci. Eng. B, 108, 2-8. David E., 2005, An overview of advanced materials for hydrogen storage, J. Mater. Process. Tech.,162- 163, 169-177. Fátay D., Révész Ă., Spassov T., Particle size and catalytic effect on the dehydriding of MgH2,2005, J. Alloy. Compd., 399(1-2), 237-241. Hanada N., Ichikawa T., Hino S., Fujii H., 2006, Remarkable improvement of hydrogen sorption kinetics in magnesium catalyzed with Nb2O5, J. Alloy. Compd., 420(1-2), 46-49. Hout J., Liang G., Schulz R., 2001,Mechanically alloyed metal hydride systems,Appl. Phys. A, 72, 187- 195. Liang G., Huot J., Boily S., Van Neste A., Schulz R., 1999, Catalytic effect of transition metals on hydrogen sorption in nanocrystalline ball milled MgH2-Tm (Tm=Ti, V, Mn, Fe and Ni) systems, J. Alloy. Compd., 292(1-2), 247-252. Lu H.B., Poha C.K., Zhang L.C., Guo Z.P., Yu X.B., Liu H.K., 2009, Dehydrogenation characteristics of Ti- and Ni/Ti-catalyzed Mg hydrides, J. Alloy. Compd., 481(1-2), 152-155. Malka I.E., Czujko T., Bystrzycki J., 2010, Catalytic effect of halide additives ball milled with magnesium hydride, Int. J. Hydrogen Energy, 35(4), 1706-1712. Oelerich W., Klassen T., Bormann R., 2001a, Metal oxides as catalysts for improved hydrogen sorption in nanocrystalline Mg-based materials, J. Alloy. Compd., 315(1-2), 237-242. Oelerich W., Klassen T., Bormann R., 2001b, Comparison of the catalytic effects of V, V2O5, VN, and VC on the hydrogen sorption of nanocrystalline Mg, J. Alloy. Compd., 322, L5-L9. Sakintuna B., Lamari-Darkrimb F., Hirscher M., 2007, Metal hydride materials for solid hydrogen storage: A review, Int. J. Hydrogen Energy, 32(9), 1121-1140. Schüth F., Bogdanović B., Felderhoff M., 2004, Light metal hydrides and complex hydrides for hydrogen storage, Chem. Commun., 2249-2258. Uvarov V., Popov I., 2007, Metrological characterization of X-ray diffraction methods for determination of crystallite size in nano-scale materials, Mater. Charact., 58(10), 883-891. Varin R.A., Zbroniec L., Polanski M., Bystrzycki J., 2011,A review of recent advances on the effects of microstructural refinement and nano-catalytic additives on the hydrogen storage properties of metal and complex hydrides, Energies, 4, 1-25. Xie L., Liu Y., Zhang X., Qu J., Wang Y., Li X., 2009,Catalytic effect of Ni particles on the desorption kinetics of MgH2nanoparticles,J. Alloy. Compd., 482, 388-392. Zaluska A., Zaluski L., Ström–Olsen J.O., 1999,Nanocrystalline magnesium for hydrogen storage, J. Alloy. Compd., 288(1-2), 217-225. Züttel A., 2004, Hydrogen storage methods, Naturwissenschaften, 91(4), 157-172.