Abstract

Two materials currently of interest for onboard lightweight hydrogen storage applications are sodium aluminum hydride (NaAlH4), a complex metal hydride, and carbon aerogels (CAs), a light porous material connected by several spherical nanoparticles. The objectives of the present work have been to investigate the synthesis, characterization, and hydrogenation behavior of Pd-, Ti- or Fe-doped CAs, NaAlH4, and MgH2 nanocomposites. The diameters of Pd nanoparticles onto CA’s surface and BET surface area of CAs were 3–10 nm and 700–900 m2g−1, respectively. The H2 storage capacity of metal hydrides has been studied using high-pressure TGA microbalance and they were 4.0, 2.7, 2.1, and 1.2 wt% for MgH2-FeTi-CAs, MgH2-FeTi, CAs-Pd, and 8 mol% Ti-doped NaAlH4, respectively, at room temperature. Carbon aerogels with higher surface area and mesoporous structures facilitated hydrogen diffusion and adsorption, which accounted for its extraordinary hydrogen storage phenomenon. The hydrogen adsorption abilities of CAs notably increased after inclusion of metal hydrides by the “hydrogen spillover” mechanisms.

1. Introduction

Hydrogen is recognized as a clean fuel because of its almost complete combustion in air with notably high energy delivery ability. Recently, owning to the need for efficient and safe H2 storage, enormous attention has been paid on the advantages of synthesizing composite metal hydrides [13]. Complex metal hydride composites in the form of ABH4, where A is an alkali metal and B is a group III element, have been widely studied in solution as proton acceptors to enhance H2 adsorption abilities [47]. NaAlH4 is different from any other metal hydrides and borohydrides with similar structures since it is capable to reversibly store H2 after doping with transition metals (e.g., Ti, Fe, or Zr) [810]. Ti-doping also improves significantly the kinetic and cycling performance of NaAlH4 at moderate temperatures around 373 K and ambient pressures. All these properties make it suitable as mobile, lightweight H2 storage materials for potential application on a hydrogen fuel cell [1123]. Practically, achievable H2 storage capacity for Ti-NaAlH4 is only 3.7 wt%, falls short of the theoretical value of 5.6 wt% for NaAlH4 [2023].

Magnesium and its alloys are potential hydrogen storage material because of their very high hydrogen storage abilities. MgH2 is also relatively light and cheaper when compared with other metal hydrides. Magnesium is often alloyed with transition metals (e.g., Fe, Ti, Zr, Zn, or Mn) to increase the adsorption and desorption rates [2429]. Similarly, FeTi is a viable compound with a hydrogen adsorption capacity of around 1.90 wt%. However, activation process of FeTi is necessary to get rid of the TiO2 layer, which would otherwise hinder the hydrogenation/dehydrogenation. Consequently, higher temperatures and pressures are frequently required to achieve reproducible, maximum hydrogen adsorption/desorption capacities in the compounds [30].

Palladium-doped monolithic carbon aerogels (CAs) show a high surface area and pore volume. These characteristics are strongly influenced by the chemical nature, distribution and dispersion of the palladium particles [3133]. Furthermore, addition of 1 wt% Pd can be effective for the hydrogen adsorption capacity of FeTi to achieve its theoretical hydrogen storage capacity of 1.9 wt% [3133]. As carbon materials such as graphite or CAs have been shown to reduce the hydrogen adsorption and desorption temperatures of Mg [34], it is not unreasonable to expect that CA addition could also be able to improve the hydrogen adsorption behavior of the FeTi system. There is a further advantage of using Pd-doped CAs as it has been shown that an addition of 1 wt% Pd in the FeTi can effectively enhance its hydrogen adsorption capacity [3133].

X-ray absorption near edge structure (XANES)/extended X-ray absorption fine structure (EXAFS) spectroscopy offers a powerful tool to understand the oxidation states and fine structures of Ti atoms in the microstructure of metal hydride composites [3538]. Thus, a thorough study of the location of Ti species in metal hydrides may be helpful to explain the enhancement of hydrogen adsorptive kinetics by identifying the compounds in the hydrides. XANES/EXAFS spectroscopy is also an excellent technique for characterizing the valency and local structure of Ti species in a complex Ti-metal hydride with short-range orders [3741]. These studies are conducted to identify the optimum hydrogen storage capacity for hydrogen storage system. In this work, synthesis and characterization, and advantages/disadvantages of different metal-hydride systems such as Mg, TiFe, or Na-based metal hydrides, complex hydrides compounds, were investigated by field-emission scanning electron microscopy (FE-SEM), X-ray diffraction (XRD), and XANES/EXAFS spectroscopy. In addition, the adsorptive H2 storage capacity was also studied using a high-pressure thermogravimetric analyzer.

2. Experimental

2.1. Preparation of Metal-Doped Carbon Aerogels

Metal-loaded organic aerogels were prepared following the method developed by Pekala [42] for resorcinol-formaldehyde aerogels. In a typical procedure, CA was synthesized by first stirring vigorously resorcinol (20.70 g), Na2CO3 (0.04 g), and equal amount of deionized water in a polypropylene jar at room temperature until resorcinol dissolved. Formaldehyde (30.94 g of purity 36.5%) was then added to the solution, followed by 0.01 M PdCl2 (Reagent ACS R.D.H., Germany) to acquire the doping of palladium. The solution was left at room temperature for 24 h, then at 323 K for the next 24 h, and finally at 365 K for another 72 h to obtain a dark black monolith. The product was rinsed with acetone three times and soaked overnight at room temperature, yielding the hydrogel materials. After drying at room temperature, the hydrogels were isolated by heating in an oven at temperatures from 273 to 1,100 K under an Ar atmosphere for several hours to obtain the Pd-doped CAs.

2.2. Preparation of Metal Hydrides

As received TiCl3 (Aldrich, 99.99%, anhydrous) was used as the catalyst precursor. Crystalline NaAlH4 (Fluka, 99.9%) was purified by tetrahydrofuran (THF) (Aldrich, 99.9%, anhydrous) solution and vacuum-dried while being filtered through 0.7 μm filter paper to remove any residual Al powder and other solids from the solution. The purified NaAlH4 was mixed with the catalyst precursor in THF to produce a Ti-doped sample containing up to 20 mol% metal (relative to either Na or Al). The THF was evaporated while the NaAlH4 and catalyst precursor were mixed manually for about 30 min, using a mortar and pestle, or until the sample appeared to be completely dry. A Ti-NaAlH4 sample was then ball-milled for 2 h using a SPEX 8000 high-energy mixer/mill at a milling speed of 1,000 rpm, and the weight ratio of ball to powder was 10 : 1. 0.5–1.0 gram of sample was subjected to ball-milling each time in a 65 cm3 SUS 304 vial by ZrO2 balls of 1.3 cm in diameter. Fe- and Ti-doped MgH2 was prepared following Yao et al. [43], where MgH2 was obtained by heating Mg powders (Alfa Aesar) to 400°C in a hydrogen storage rig. 90 wt% conversion to MgH2 was achieved according to the volumetric calculations. 1.5–2.0 grams of nanophase iron powder (100 nm), Ti (Aldrich Co), and MgH2 samples were mixed in the composition of MgH2-5.0 wt% (2Fe + Ti, molar ratio) and mechanically ball-milled by using a ZrO2 ball mixer/mill (model SPEX 8000). The samples were subjected to ball-milling as described above for 2 h, using a weight ratio of ball to powder of 8 : 1 The powder was then ball-milled again with 5.0 wt% of CAs for an additional 10 h at a milling speed of 1,000 rpm. All procedures described above were carried out in a flowing argon gas glove box to prevent the metal hydride samples from oxidation. The samples were then used for hydrogen adsorption property tests and microstructure characterization.

2.3. Hydrogen Adsorption Measurements

The hydrogen isotherms were measured gravimetrically at different temperatures using a method previously described by Eddaoudi et al. [44]. A Cahn Thermax 500 microgravimetric balance with a sensitivity of 1 μg was used to measure the change in mass of samples suspended within a glass enclosure under a certain atmosphere. A pressure sensor, with a range of 0 to 68 atm (at 1,000°C) and sensitivity of 0.011 atm, was used to measure the hydrogen pressure in the chamber. Samples were outgassed overnight until a constant mass was attained; these varied from 0.2 to 2.0 g. Prior to admittance of argon gas, the entire chamber and manifold were evacuated overnight. The system was purged at room temperature three times with the argon gas before cooling to 77 K, and gases were passed through a molecular sieve trap immersed in liquid nitrogen to remove any condensable impurities or moisture before being exposed to the sample. Pressures were measured with the range covering 1 to 30 atm. Hydrogen was added incrementally, and data points were recorded when no further change in mass was observed. The dynamic hydrogen adsorption data on Pd-, Ti-, Fe-doped CAs, NaAlH4, or MgH2 batch adsorption experiments were also conducted at room temperature (298 K) and different pressures. The adsorbed amount of hydrogen was calculated after the buoyancy correction.

2.4. Characterization of Carbon Aerogels and Metal Hydrides

The morphologies, crystallinity, particle size distribution, and the compositions of as-synthesized metal hydrides and metal-doped CAs were determined by FE-SEM equipped with EDS (Hitachi, S-4700 Type II) and HR-TEM (Zeiss 10C). XRD (MAC Science, MXP18) was performed to identify the phases and crystallinities in the samples. Complex of metallic hydride and CAs was scanned from 20 to 80° () with a scan rate of 4° () min−1, and monochromatic radiation was used. The recorded specific peak intensities and values can be further identified by a computer database system (JCPDS). The surface areas of the CAs were measured by BET (Brunauer-Emmett-Teller) nitrogen adsorption (Micromeritics ASAP 2010 Instrument). For the BET surface area measurement, the samples were scraped from the sample tube substrate and powdered so as to avoid any influence from the steel tube. Prior to measurement, all samples were degassed at 423 K for 1 h. For the calculation of the BET surface areas, the relative pressure range from 0.05 to 0.2 was used. The pore radius distribution was determined by Barrett, Joyner, and Halenda (BJH) method.

The XANES/EXAFS spectra were collected at BL16A1 and BL01C1 (SWLS) beamlines at the National Synchrotron Radiation Research Center (NSRRC) of Taiwan. The electron storage ring was operated with the energy of 1.5 GeV and a current of 100–200 mA. A Si(111) double crystal monochromator (DCM) was used for providing highly monochromatized photon beams with energies of 1 to 15 keV (BL16A1) and 5 to 30 keV (BL01C1) and resolving power of up to 7,000. Data were collected in fluorescence or transmission mode with a Lytle ionization detector [39] for Ti (4,966 eV) and Pd (24,350 eV) K-edge experiments at room temperature. The photon energy was calibrated by characteristic preedge peaks in the absorption spectra of Ti standards. The raw absorption data in the region of 50 to 200 eV below the edge position were fit to a straight line using the least-square algorithms [3740]. The XANES was extended to energy of the order of 50 eV above the edge. The -weighted and EXAFS spectra were Fourier transformed to space over the range between 2.5 and 12.5 . The EXAFS data were analyzed by using the UWXAFS 3.0 program and FEFF 8.2 codes [3741].

3. Results and Discussion

3.1. Morphology of Metal-Doped Carbon Aerogels and Metal Hydrides

The HR-TEM image and FE-SEM micrographs of metal hydride and CAs complexes synthesized are shown in Figure 1. Figure 1(a) is an HR-TEM observation of the Pd-doped aerogel structure consisting of spherical primary Pd(0) nanoparticles. It can be seen that palladium particles of size 3 to 10 nm were uniformly distributed throughout the gel. FE-SEM micrograph in Figure 1(b) reveals the pellet-like or irregularly shaped Ti-NaAlH4 crystalline nanoparticle, with diameters approximately 50–80 nm. Correspondingly, Fe and Ti nanoparticles with an average diameter of 50–80 nm were dispersed uniformly on the surface of the doped MgH2 surface as shown in Figure 1(c).

3.2. X-Ray Powder Diffraction

The structure of the compound was confirmed by structural refinement of X-ray diffraction patterns shown in Figure 2. The intensive peaks appearing at small 2 angles were characteristics of porous materials which possess numerous pores or cavities. Whereas Figure 2(a) shows the characteristic peaks of 2 = 27.46, 31.75, 45.53, 56.54, 66.29, 75.36, and 84.06°, respectively, which indicates that the structure of Ti-NaAlH4 composites might not be affected by Ti added during ball-milling. From the intensity of (200) peak, the minimum grain size was measured to be around 60 nm [1720]. This grain particle size was calculated from Scherrer’s equation (with Warren’s correction for instrumental broadening) applied to the half-height of the maximum intensity diffraction peak. Metallic Na3AlH6 and Al were consistent with the characteristic peak at 32.92 and 38.6°, respectively [24]. From the FE-SEM measurement and XRD patterns, the titanium compound distributed on the surface of the NaAlH4 can be confirmed.

The X-ray diffraction patterns of the Pd-doped CAs indicate that the Pd nanoparticles have a crystalline structure (Figure 2(b)). According to FE-SEM micrographs, the well dispersion of metallic Pd nanoparticles in the CAs apparently facilitated the porous structure. The XRD patterns of Pddoped CAs show those two characteristic sharp peaks with higher intensities around 40.12 and 46.77°, indicating a largely crystalline structure. The mean particle size was about 5–8 nm from HR-TEM observation that may help the mesoporous structure of CAs to adsorb hydrogen. The diffraction angles at around 20 and 44°, corresponding to the (002) and (101) diffraction peaks of graphite, respectively, are represented in Figure 2(b). In addition, hexagonal structure of the CAs started to become disordered with the heat treatment, owning to the progressive graphitization of the sample. Figure 2(c) represents the XRD patterns of FeTi-doped MgH2 structure. The peak of the Fe sample at 44.6° indicates that the decomposition of Fe onto the metal hydrides, whereas characteristic peaks of -MgH2 were observed at 27.81, 31.23, 39.7, 54.57, 65.45, and 75.54°, respectively. Thus, -MgH2 morphology was dominated in the sample, and the metallic Mg with -MgH2 structure represents the peaks at 36, 57.76, and 25.67°, respectively [43].

3.3. XANES and EXAFS

Figures 3(a)–3(e) show the preedge XANES spectra of Ti K-edge (4,966 eV) in nanophase Ti-NaAlH4 nanoparticles and anatase-typed TiO2 standard [45]. These spectra exhibit an absorbance feature of 4,966 eV for the 1s to 3d transition, which is forbidden by the selection rule in case of perfect octahedral symmetry. The sharp feature at 4,987 eV, due to the dipole allowed of 1s to electron transition, indicates the existence of Ti(IV). The intensity of the 1s to transition is proportional to the population of Ti(IV) of the Ti-NaAlH4 complexes. Oxygen atoms are the major atoms coordinated to the central Ti atoms in the Ti-NaAlH4 complexes.

The reversible hydrogenation of NaAlH4 and dehydrogenation to NaH and Al occurred in two steps summarized as follows [16].

Remarkably, the addition of the Ti species lowers the activation energy of the adsorption/desorption processes and enhanced the kinetic rates of these reactions by several orders of magnitude [4650]. However, the exact mechanism still required further investigations. Notably, the hydrogen-depleted multicomponent products resulting from (2) were found to form large crystallites of the complex metal hydride upon the reverse reaction in the presence of Ti during hydrogenation [4547]. This is an important clue to understand the role of Ti where it implies a concerted migration of metal atoms over large distances. Local structural arrangements without the involvement of large-scale atomic migration would instead result in an amorphous or nanocrystalline structure.

The main scientific issues concerning the chemical forms (or speciation) of active species ultimately depend on molecular-scale structure and properties. Basic understanding at this scale is essential for further understanding the catalytic behaviors of Ti-NaAlH4 complexes synthesized at different mol% values of Ti. Generally, Ti K-edge EXAFS spectroscopy in Figure 4 can provide the information on the Ti atomic arrangement of catalysts in terms of bond distance, bond angle, coordination number, kind of near neighbors, and thermal or static disorders. Experimentally, the Ti species with a central Ti atom have a Ti–O bond distance of and a coordination number of 4.07, in a Ti-NaAlH4 composite after ball-milling for 2 h using a high-energy mixer/mill. Comparatively, however, a rutile-typed TiO2 has a central Ti atom with a Ti–Ti and a Ti–(O)–Ti bonds, and their bond distances are and , respectively [2932, 40]. Since the differences of mass density () and band structures between anatase- and rutile-typed TiO2, bond distances of Ti-O are and , respectively [914].

The EXAFS data shown in Figure 4 reveals that the peak intensity of first-shell Ti–O bond increased with decreasing mol% values of Ti-dopants for a Ti-NaAlH4 composite after ball-milling for 2 h using a high-energy mixer/mill. These results may offer a further explanation on the higher oxidation states of Ti species for a Ti-NaAlH4 complex after ball-milling at higher rates and temperatures. Moreover, each Ti(IV) ion in the framework of nanophase Ti-NaAlH4 complex crystals is surrounded by six O2− ions [914]. Since the bond lengths of the first and second shells for Ti–O bonding are and , similarly with the coordination numbers of four and two, respectively, an octahedral structure of the Ti-NaAlH4 composites was found using EXAFS measurement [45]. As shown in Table 1 and Figure 5, the Ti species with a central Ti atom have a Ti–O bond distance of and a coordination number of 1.83, in a Ti-NaAlH4 composite in 8 mol% Ti-NaAlH4 metal hydride composites (more than 2 mol%) after the hydrogen adsorption processes. It can be seen that all Ti species in Ti-NaAlH4 metal hydride composites can be reduced in the hydrogen adsorption processes and be kept in the metallic Ti form.

3.4. N2 Gas Adsorption Isotherms Analyses

Nonspecific physical adsorption of the CAs was carried out to measure the total surface area and pore size distribution for Pd-doped CAs, as shown in Figure 6. The surface area, pore size distribution, and crystalline diameter of CAs were calculated according to the adsorption data summarized in Table 2. A large surface area was generally observed for Pd-doped CAs and beneficial characteristic for a great variety of applications. The microcrystalline structure of Pd on the CAs is confirmed from the HR-TEM analysis (Figure 1(a)), resulting in an increase of the surface area. The nitrogen adsorption-desorption isotherms exhibit a hysteresis behavior, indicating the microstructural nature of the porous specimens. Moreover, the type IV hysteresis isotherm obtained is represented in Figure 6(a). The adsorption hysteresis was observed in the region of a relative pressure above 0.4. The specific surface areas of CAs and Pd-doped CAs are 687 and 725 m2 g−1, with the pore volumes of 0.62 and 0.64 cm3 g−1, respectively, as shown in Table 2. It also indicates that these resulting factors of surface enhancement depended on the doping of the Pd on the CAs and consisted of higher porous structure [49, 50]. These results may be caused by the dispersion of the Pd on the CAs surface and that indicated the metallic dispersion would be helpful for adsorption criteria.

The intensive peak of the Pd(0) K-edge region around 24,368 eV and the stronger peak at 24,392 eV were due to the orbital 3d5s or 4p hybridization caused by unoccupied bands, as shown in Figure 7. These quantified the doping of Pd on the surface of carbon aerogels. Therefore, in the absorption where the orbital 3d5s or 4p hybridization was intensive, the p-like density of states was notably enhanced and the absorption increased like the second absorption peak in the XANES/EXAFS spectra of the Pd standards (Figures 7(b) and 8(a)). The structural parameters of the Pd powder and the palladium doped CAs, obtained from the best fit to the EXAFS data, are shown in Table 1. The coordination numbers and bond lengths also were well compared with the Pd metal and palladium-doped CAs crystal structure data. The Debye-Waller factors (Δ) were less than 0.015 (), which indicated that the palladium confirmatively has center Pd atoms coordinated by Pd–Pd bonding. The bulk Pd metal possessed Pd–Pd bond distance of and a coordination number of 9.59. For comparison, fresh CAs–Pd clusters were found to have a Pd–Pd bond distance of 2.12  with a coordination number of 6.18. This shortening of the bond distance with a lower coordination number was caused by random motion of surface atom on the small Pd particles and was responsible for the increase in the surface area of the palladium doped CAs samples, whereas hydrogen reduced bulk CA-doped Pd cluster consisted with the Pd–Pd bond distance of 2.73  and a coordination number of 9.23. This revealed that the Pd nanoparticles were notably activated in the reduction process [48]. Moreover, the shortening of the bond distance with coordination number probably was caused by the random motion of surface atom on the small Pd particles and therefore increases the surface area of the Pd-doped CAs samples. Furthermore, this result also revealed that the Pd nanoparticles were well dispersed on the surface of CAs, which might improve the amount of hydrogen storage efficiency significantly.

3.5. Hydrogen Gas Adsorption Isotherms Analyses

Comparisons of the hydrogen adsorption of metal hydride and doped metal hydride with carbon aerogel are given in Figures 9(a)–9(e). Under a typical reaction condition of 298 K and 1–30 atm, the hydrogen adsorption curves of pure MgH2 (Figure 9(e)) were consistent with lowest adsorption criteria, whereas the FeTi with CA-doped MgH2 (Figure 9(a)) possessed the highest hydrogen adsorption capacity of 4.02 wt%. Mesoporous structure of the CAs, which has a very high surface area as confirmed from the BET nitrogen isotherm analysis, improved the hydrogen adsorption behavior. This is inline with the finding of Imamura et al. [34], who reported that the uptake of hydrogen in a C-Mg composite took place at a significantly lower temperature than that in pure magnesium. Similarly, Jung et al. [51] suggested that the faster adsorption kinetics of MgH2 systems may result from the very high defect densities of the damaged MgH2 surfaces caused by the ball-milling of Mg with hard and small oxide particles. Defects provide hydrogen an easy path to go into Mg powders. Therefore, the hydrogen adsorption capacity of 2.7 wt% was obtained with the addition of FeTi nanoparticles into the MgH2 samples by ball-milling (Figure 9(b)). Thus, minute particles of FeTi were found to be the most useful catalyst to improve the hydrogen adsorption properties of MgH2. Nitrogen adsorption by the CAs or Pd-doped CAs material clearly shows reversible type IV isotherms at each of the activation stages, indicative of permanent porosity in Figure 6(a). The selectivity of H2 adsorption over Pd–doped CAs was 1.6 wt%, as shown in Figure 9(c). Due to the doping of Pd on CAs, access of hydrogen was easier in the vacancies created by the mesoporous structure with a higher surface area that was confirmed with the BET nitrogen isotherms measurement. In addition, this suggests that an antidispersive force plays an important role in case of Pd-doped CAs samples and the expected specific interaction with Pd active sites. Thus, this work confirmed that metal doping by Pd on CAs may improve the hydrogen adsorption criteria. An intriguing observation concerns the dependence of the rehydrogenation kinetics on the amount of Ti doping. Moreover, as shown in Figure 9(a), it was observed that the hydrogen uptakes of FeTi-CAs-MgH2 systems could be enhanced markedly by the phenomenon of “hydrogen spillover” via a simple technique for building carbon bridges [52]. One of the possible reasons is that the hydrogen molecules do not readily dissociate on Mg surface. Experimentally, the catalytic effect was found on hydrogen adsorption of mixing transition metal of Ti atoms into Mg hydride powder during ball-milling. The likely mechanism proposed for the FeTi-CAs-MgH2 hetero-structures can be also summarized concisely: adsorption of hydrogen on the Ti surface, dissociation of hydrogen and chemisorption of atomic hydrogen on the surface, (3) migration of atomic hydrogen onto the CAs support, and finally (4) chemisorptive spillover onto the MgH2 substrate. In addition, spillover is facilitated through the use of a carbon bridging compound between the Ti/CAs complex and the MgH2 substrate. Typically, an 8 mol% Ti-dop onto the NaAlH4 (Figure 9(d)) revealed the maximum hydrogen adsorption capacity of 1.2 wt% within this class of adsorptive materials. In addition, perfect single crystal TiO2 was inert toward reaction with H2 [51]. However, H2 was absorbed by TiO2-surfaces that contained a higher density of defects in the crystal structure. Therefore, surface of MgH2 in the presence of the metal-doped Fe and Ti with CAs created a high defect density that introduced a higher hydrogen adsorption behavior in the adsorption or desorption processes.

4. Conclusions

The synthesis, characterization, and H2 adsorption capacity of CAs/metallic hydride nanocomposites as a catalyst were investigated in the present work. Experimentally, the H2 storage capacity of metallic samples was conducted and measured by a TGA microbalance method. In addition, fine structures and crystallinity of metallic hydride were identified by BET nitrogen adsorption isotherms, HR-TEM, FE-SEM/EDS, XRD, and XANES/EXAFS. The CAs were very effective in improving the hydrogen storage capacity of the Fe-, Ti-doped MgH2 samples with the “hydrogen spillover” route. Metallic dispersion such as Pd nanoparticles onto the CAs may improve the hydrogen adsorption abilities. Higher surface area from CAs and the defect criteria of the MgH2 surface due to the doping of the metallic particle caused to improve the hydrogen adsorption capacities. On the other hand, Ti-doping onto the sodium aluminum hydride can only improve slightly the hydrogen storage capabilities.

Acknowledgment

The authors gratefully acknowledge the financial support of National Science Council (Contract no. NSC-95-2623-7-155-011) of Taiwan.