Enhanced hydrogen storage properties of LiAlH₄ catalyzed by CoFe₂O₄ nanoparticles

Abstract

The catalytic effects of CoFe₂O₄ nanoparticles on the hydrogen storage properties of LiAlH₄ prepared by ball milling were investigated. The onset desorption temperature of the LiAlH₄ + 2 mol% CoFe₂O₄ sample is 65 °C, which is 90 °C lower that of the as-received LiAlH₄, with approximately 7.2 wt% hydrogen released at 250 °C. The isothermal desorption results show that for the 2 mol% CoFe₂O₄ doped sample dehydrogenated at 120 °C, 6.8 wt% of hydrogen can be released within 160 min, which is 6.1 wt% higher than that of the as-received LiAlH₄ under the same conditions. Through the differential scanning calorimetry (DSC) and the Kissinger desorption kinetics analyses, the apparent activation energy, E_a, of the 2 mol% CoFe₂O₄ doped sample is calculated as 52.4 kJ mol⁻¹ H₂ and 86.5 kJ mol⁻¹ H₂ for the first two decomposition processes. This is 42.4 kJ mol⁻¹ H₂ and 86.1 kJ mol⁻¹ H₂ lower compared with the pristine LiAlH₄, respectively, indicating considerably improved dehydrogenation kinetics by doping the CoFe₂O₄ catalyst in the LiAlH₄ matrix. From the Fourier transform infrared spectroscopy (FTIR) and X-ray diffraction (XRD) analyses, a series of finely dispersed Fe and Co species with a range of valence states, produced from the reactions between LiAlH₄ and CoFe₂O₄, play a synergistic role in remarkably improving LiAlH₄ dehydrogenation properties. The rehydrogenation properties of the LiAlH₄ + 2 mol% CoFe₂O₄ sample have also been investigated at 140 °C under 6.5 MPa pressure held for 2.5 h.

---

1. Introduction

As a renewable energy source, hydrogen can be produced from water and biomass without any greenhouse gas emissions. Thus, hydrogen attracts considerable attention from research aiming to solve the fossil fuel depletion problem accompanied by the global environmental issues.^1–3 The prerequisite for widespread hydrogen use as an energy carrier is the development of advanced hydrogen storage materials for safely storing it at high gravimetric and volumetric densities.^4–6

Among numerous possible hydrogen storage materials, lithium aluminum hydride^7–10 (LiAlH₄) is a promising candidate due to its relatively large theoretical hydrogen storage capacity and high potential reversible hydrogenation capability. Theoretically, LiAlH₄ can desorb 10.5 wt% hydrogen upon heating to 420 °C, which make it an ideal hydrogen storage material to meet the U.S. Department of Energy 2015 targets for a viable hydrogen storage system¹¹ with gravimetric density ≥5.5 wt% and volumetric density ≥40 g L⁻¹. Upon heating, LiAlH₄ would gradually release hydrogen, according to the following three steps.¹² The first reaction step (R1) occurs in the 150–175 °C temperature range and releases 5.3 wt% hydrogen:

3LiAlH₄ → Li₃AlH₆ + 2Al + 3H₂ (1)

Then the second reaction step (R2) occurs between 180 °C and 220 °C, releasing 2.6 wt% hydrogen:

Li₃AlH₆ + 2Al → 3LiH + 3Al + 3/2H₂ (2)

The third reaction step (R3) starts to release 2.6 wt% hydrogen above 400 °C:

3LiH + 3Al → 3LiAl + 3/2H₂ (3)

Thus, the dehydrogenation properties of LiAlH₄ are generally analyzed for the first two decomposition reactions due to the high onset and decomposition temperatures, and the low desorbed hydrogen content of the reaction R3 from the practical applications perspective.^13–16

Since Bogdanovic et al.¹⁷ conducted the seminal work in improving the hydrogen storage performance of NaAlH₄ by doping TiCl₃, extensive efforts have been devoted to ameliorate the re/dehydrogenation properties of LiAlH₄ by adding various catalysts to lower its onset dehydrogenation temperature and increase its dehydrogenation kinetics. To date, the documented catalysts for LiAlH₄ can be classified as: (1) pure metals;^1,18–26 (2) carbon-containing species;^21,27–31 (3) metal halides;^{13,19,21,32–42} (4) alloys;^18,20 (5) metal oxides^{14–16,43,44} and (6) other compounds.^45–49 To our knowledge, a partial reversibility can be realized through doping LiAlH₄ with various catalysts.^28,35,44,50 However, the rehydrogenation property was not ideal. From the practical applications perspective, solid-state materials (LiAlH₄, NaAlH₄, MgH₂, etc.) do have the potential to outperform physical methods of storage (cryostorage or high-pressure technologies) through comprehensively considering the safety, environment friendless and cost, which has been reported in many review papers.^51,52 However, it is crucial to find an advanced catalyst, which could not only significantly improve the dehydrogenation, but also rehydrogenation performance of LiAlH₄. Recently the authors have observed the superior effects of Fe₂O₃ and Co₂O₃ nanoparticles on promoting the dehydrogenation properties of LiAlH₄, however, nano-sized Fe₂O₃ and Co₂O₃ failed to produce any reversibility for LiAlH₄.¹⁵ Herein, it is reasonable to speculate that Co ferrite shows a great potential as the catalyst to advance hydrogen storage performance of LiAlH₄.

In this work, the catalytic efficiency of CoFe₂O₄ nanoparticles on the dehydrogenation and reversible hydrogenation properties of LiAlH₄ was evaluated by utilizing a pressure-composition-temperature (PCT) apparatus and differential scanning calorimetry (DSC). The catalytic mechanism of CoFe₂O₄ nanoparticles was demonstrated by analyzing the results of the Fourier transform infrared spectroscopy (FTIR), X-ray diffraction (XRD) and scanning electronic microscopy (SEM). The comparison of the catalytic effects of CoFe₂O₄, Fe₂O₃ and Co₂O₃ catalysts for LiAlH₄ is also presented in this work.

2. Experimental

2.1. Sample preparation

LiAlH₄ (≥95% pure) was purchased from the Sigma Aldrich Co., and CoFe₂O₄ (≥99.99% pure, 20 nm) was prepared by using the sol–gel method. The details of the preparation procedure are given in the previous report.⁵³ All handling of the samples was conducted in a glove box (Mikrouna Co., China) under high-purity argon atmosphere (H₂O: <10 ppm; O₂: <10 ppm) in order to minimize oxidation and humidity. About 1.5 g of LiAlH₄ was mixed with various mole fractions of CoFe₂O₄ nanopowder, and then the mixture was loaded into a stainless steel grinding vial (5 cm in diameter, quenching). After that, ZrO₂ balls (Mohs hardness ≥ 7.5) were added with a ball-to-powder weight ratio of 20 : 1 in the glove box. Finally, the grinding vial with the mixed sample was ball milled for 30 min by using a high energy Spex mill (QM-3B) at a milling rate of 1200 rpm. In order to prevent excess heating and the surface fatigue wear of ball-milling materials, the grinding vial was cooled down for 5 min after milling every 10 min.

2.2. Characterization

The hydrogen storage performance of the as-received and doped LiAlH₄ samples was measured by using a Sieverts-type PCT apparatus (Beijing Nonferrous Metal Research Institute, China). The PCT equipment can be heated up to 600 °C with a maximum hydrogen pressure of 10 MPa. To measure the dehydrogenation properties, 0.3 g sample was loaded into a stainless steel vessel and then heated to 250 °C at 5 °C min⁻¹ heating rate under 0.1 atm pressure. For the rehydrogenation measurements, the samples that completed the first dehydrogenation were directly reheated at 150 °C under 6.5 MPa for 3 h. The de/rehydrogenation amount for all samples was calculated from the pressure changes, and then the values were converted for pure LiAlH₄ with the elimination of various impurities, the detailed calculation formula is as follows,

mat-wt% = H₂ mass/[mass (storage material) + mass (catalyst) + H₂ mass] (4)

All the weight percentage values we talked about in the present paper on materials (pure or doped sample) basis.

In order to investigate the decomposing behavior and calculate the activation energy of both as-received and doped LiAlH₄ samples, DSC measurements were conducted by using NETZSCH STA 449C under a flow of 50 mL min⁻¹ high-purity Ar. Typically, about 5 mg of sample was sealed into a 50 mL alumina crucible in the glove box, and then was heated at different heating rates (6 °C min⁻¹, 9 °C min⁻¹, and 12 °C min⁻¹) from 35 °C to 300 °C, respectively.

The morphology of the as-received and 2 mol% CoFe₂O₄-doped samples were observed by SEM (ZEISS EVO 18, Germany) equipped with the energy dispersive spectroscopy (EDS) detector. Prior to the SEM observations, the samples were prepared inside the glove box, and then transferred to the SEM chamber in order to prevent oxidation and moisture adsorption.

FTIR analysis of the as-received and doped LiAlH₄ samples after ball milling was carried out by using Bruker Vector 22 FTIR spectrometer. The FTIR spectra were recorded between 2000 cm⁻¹ and 750 cm⁻¹ with a spectral resolution of 4 cm⁻¹.

Phase structure characteristics of the as-milled and de/rehydrogenated samples were detected by XRD (MXP21VAHF X-ray diffractometer with CuKα radiation, 40 kV, 200 mA) at room temperature. The X-ray intensity was tested over the 2θ angle ranged from 10° to 90° with a scanning velocity of 0.02° per second.

3. Results and discussion

Fig. 1 displays the non-isothermal desorption curves of the as-received LiAlH₄, as-milled LiAlH₄, and LiAlH₄ doped with 1 mol%, 2 mol%, 3 mol%, and 5 mol% CoFe₂O₄ nanoparticles, heated from 25 °C to 250 °C at a heating rate of 5 °C min⁻¹. As seen in Fig. 1, the as-received LiAlH₄ sample started to release hydrogen at around 155 °C and about 5.0 wt% hydrogen desorbed during the first dehydrogenation step. With increasing temperature, the as-received LiAlH₄ sample entered into the second dehydrogenation step from 200 °C and about 2.5 wt% hydrogen was released at the second dehydrogenation stage. Thus, the total hydrogen release capacity of 7.5 wt% could be obtained when the as-received LiAlH₄ was heated to 250 °C. For the as-milled LiAlH₄ sample, the onset dehydrogenation temperature in the first two dehydrogenation steps deceased by about 21 °C, compared with the as-received LiAlH₄, mainly attributed to the surface activation, introduced to the LiAlH₄ matrix by mechanical milling.^{13–16,24,28,38,39,42,43} Compared with the LiAlH₄ samples without any catalysts doping, the onset desorption temperature of LiAlH₄ doped with CoFe₂O₄ nanoparticles exhibited a remarkable reduction, not only for the first, but also for the second dehydrogenation step. When 1 mol% CoFe₂O₄ nanopowder was added to the LiAlH₄ matrix, the onset dehydrogenation temperature decreased by 75 °C for the first stage and 40 °C for the second stage, compared with the as-received LiAlH₄. The 1 mol% doped sample released 7.4 wt% hydrogen at the first two dehydrogenation steps. By further increasing the content of the CoFe₂O₄ nanoparticles to 2 mol%, the LiAlH₄ + 2 mol% CoFe₂O₄ sample started to release hydrogen at 65 °C and 130 °C for the first two dehydrogenation steps, which decreased by 90 °C and 70 °C, compared with the as-received LiAlH₄, respectively. Overall, 7.2 wt% hydrogen was released for the 2 mol% doped sample. For the hydrogen release content of 1 mol% and 2 mol% doped samples, they are close to the theoretical hydrogen release content of pristine LiAlH₄ (7.5 wt% H₂). For the 3 mol% CoFe₂O₄ doped sample, the onset dehydrogenation temperature further decreased to 61 °C for the first dehydrogenation step, while only 5.5 wt% hydrogen was released during the first two dehydrogenation processes, indicating a drastic reduction in the released hydrogen capacity after doping an excess amount of CoFe₂O₄ nanoparticles. A similar phenomenon was also proposed in previous reports.^{13,14,28,35,37,43,47} However, when 5 mol% of CoFe₂O₄ were added, the LiAlH₄ doped sample started to dehydrogenate at 100 °C, which is much higher than the other contents CoFe₂O₄-doped samples. Meanwhile, the desorption hydrogen content dropped sharply to 3.2 wt% for the first two dehydrogenation steps, which only accounts for 41.7% of the total hydrogen release for pure LiAlH₄. The excessive decrease in the amount of hydrogen release for the LiAlH₄ + 5 mol% CoFe₂O₄ samples contributes to the excessive catalytic effect, leading to the complete decomposition of LiAlH₄ during the high-energy ball-milling process. In the meanwhile, the dehydrogenation process conducted during the heating and desorption process was the second desorption stage only. Fig. 2 shows hydrogen released from LiAlH₄ doped with different amounts of CoFe₂O₄, Fe₂O₃ and Co₂O₃ catalysts, which is nearly close to the theoretical hydrogen release content of the pristine LiAlH₄. However, when the content of every catalyst is higher than a certain value, the amount of hydrogen released sharply decreases. For the CoFe₂O₄ doped LiAlH₄ sample, its hydrogen released amount declined quickly when more than 2 mol% CoFe₂O₄ nanoparticles were added. However, as for the Fe₂O₃ and Co₂O₃ doped LiAlH₄ samples, their hydrogen release content decreases rapidly when the Fe₂O₃ and Co₂O₃ nanoparticles content was more than 5 mol%. CoFe₂O₄ has a stronger catalytic effect on the dehydrogenation properties of LiAlH₄, compared with Fe₂O₃ and Co₂O₃. The LiAlH₄ + 2 mol% CoFe₂O₄ sample exhibits optimal dehydrogenation performance, based on the onset dehydrogenation temperature and hydrogen desorption capacity, and would be utilized to analyze the catalytic effect and the mechanism of the CoFe₂O₄ nanoparticles in the following tests.

Fig. 1 Thermal desorption profiles of the as-received LiAlH₄, as-milled LiAlH₄, and LiAlH₄ doped with 1, 2, 3, and 5 mol% CoFe₂O₄ nanoparticles. The samples are heated to 250 °C at 5 °C min⁻¹ heating rate.

Fig. 2 Hydrogen released from LiAlH₄ doped with different catalysts in the 25–250 °C temperature range.

Fig. 3 shows the isothermal dehydrogenation behavior of the as-received LiAlH₄ at 120 °C and the LiAlH₄ + 2 mol% CoFe₂O₄ at 90 °C, 120 °C and 150 °C, respectively. From the curve (a) in Fig. 3, only 0.7 wt% of hydrogen could be detected within 180 min, indicating a perishing desorption kinetics of pristine LiAlH₄ at 120 °C. However, the dehydrogenation kinetics of LiAlH₄ was significantly enhanced after doping Co ferrite nanopowder. When heated at 90 °C (Fig. 3b), the CoFe₂O₄-doped sample could release 5.1 wt% hydrogen within 160 min, suggesting the first dehydrogenation step completion for LiAlH₄. Furthermore, the 2 mol% doped sample released 6.8 wt% of hydrogen within 160 min at 120 °C (Fig. 3c), which is 6.1 wt% higher compared with the as-received LiAlH₄ for the same heating temperature and time. When further increasing temperature up to 150 °C, only 55 min were required to complete the first two dehydrogenation steps for the LiAlH₄ doped with 2 mol% CoFe₂O₄, as seen in Fig. 2d. Thus it is reasonable to conclude that CoFe₂O₄ exhibits superior catalytic performance and significantly improves the dehydrogenation kinetics of LiAlH₄, which makes it quite attractive for the PEM fuel cell applications.

Fig. 3 Isothermal dehydrogenation kinetics of (a) as-received LiAlH₄ at 120 °C, and LiAlH₄ + 2 mol% CoFe₂O₄ at: (b) 90 °C, (c) 120 °C, and (d) 150 °C. (I) represents the first dehydrogenation step, and (II) presses the second dehydrogenation step.

To further reflect the CoFe₂O₄ nanoparticles excellent catalytic effect of improving the LiAlH₄ isothermal dehydrogenation kinetics and test the practical operating temperature of the PEM fuel cells, Fig. 4 shows isothermal dehydrogenation kinetics of LiAlH₄ doped with CoFe₂O₄, Fe₂O₃ and Co₂O₃ heated at 90 °C. As seen in Fig. 4, the Co₂O₃ and Fe₂O₃ doped samples release 4.0 wt% and 4.4 wt% H₂ in 180 min at 90 °C, while the CoFe₂O₄ doped sample could release 5.1 wt% H₂ within 160 min, indicating that CoFe₂O₄ is superior to Fe₂O₃ and Co₂O₃ in improving the dehydrogenation kinetics of LiAlH₄. This is in good agreement with the hydrogen released amount results of LiAlH₄ doped with these three catalysts (Fig. 2).

Fig. 4 Isothermal dehydrogenation kinetics of LiAlH₄ doped with 2 mol% CoFe₂O₄, 5 mol% Fe₂O₃ and 5 mol% Co₂O₃ heated at 90 °C.

In order to further analyze the dehydrogenation steps of the CoFe₂O₄ doped samples in terms of the exo/endothermic characteristics and to acquire activation energy (E_a) for each dehydrogenation step according to the Kissinger method, Fig. 5 displays the DSC curves of the as-received LiAlH₄ (6 °C min⁻¹) and 2 mol% CoFe₂O₄ doped LiAlH₄ (6 °C min⁻¹, 9 °C min⁻¹ and 12 °C min⁻¹) within the 35–300 °C temperature range, respectively. The as-received LiAlH₄ DSC curve contains four characteristic peaks in the first two dehydrogenation steps (two exothermic and two endothermic peaks). These four thermal characteristic peaks correspond to the interaction of LiAlH₄ with surface hydroxyl impurities at 154.9 °C, melting of LiAlH₄ at 166.4 °C,⁵⁴ decomposition of liquid LiAlH₄ (R1) at 184.5 °C and decomposition of Li₃AlH₆ at 240 °C (R2).³² However, the DSC curve of the CoFe₂O₄ doped LiAlH₄ sample has only two characteristic peaks measured at different heating rates. When heated at a heating rate of 6 °C min⁻¹, the exothermic peak of the doped sample appeared at about 131 °C. Thus the first exothermic peak is attributed to the decomposition of the solid state LiAlH₄, since the CoFe₂O₄ doped LiAlH₄ started to decompose prior to its melting. Then the endothermic peak emerged at 205 °C, corresponding to the dehydrogenation step of Li₃AlH₆. Furthermore, the characteristic temperatures of these two endothermic peaks gradually rise with the increasing heating rate, suggesting that the doped sample has more time to relax at any given temperature and thus the decomposition occurs sooner at a lower temperature when heated at the relatively lower rate. A similar phenomenon is also reported in the DSC results of LiAH₄ doped with various catalysts.^{13–16,18,32,38,40,43,44} Therefore, the dehydrogenation properties of LiAlH₄ were significantly improved by adding CoFe₂O₄ nanoparticles, reflecting the remarkable reduction on the characteristic peak temperature of LiAH₄.

Fig. 5 DSC curves of (a) as-received LiAlH₄, LiAlH₄ + 2 mol% CoFe₂O₄ in the 35–300 °C temperature range and the heating rate of: (b) 6 °C min⁻¹, (c) 9 °C min⁻¹, and (d) 12 °C min⁻¹.

In order to analyze the catalytic mechanism of CoFe₂O₄ nanoparticles on the dehydrogenation properties of LiAlH₄, the apparent activation energy (E_a) of the as-received LiAlH₄ and the CoFe₂O₄-doped LiAlH₄ sample for the first two decomposition steps were calculated by using the Kissinger method,⁵⁵

where β, T_p and R express the heating rate, the peak temperature and the gas constant, respectively. Fig. 6 shows the Kissinger plots for the first and second dehydrogenation steps of the as-received LiAlH₄ and LiAlH₄ + 2 mol% CoFe₂O₄. According to the slope of the line in Fig. 6, the E_a values of the as-received LiAlH₄ for the first two dehydrogenation steps are calculated to be 94.8 kJ mol⁻¹ H₂ and 172.6 kJ mol⁻¹ H₂, respectively. Furthermore, the E_a for the two decomposition reactions of the CoFe₂O₄ doped sample are 52.4 kJ mol⁻¹ H₂ and 86.5 kJ mol⁻¹ H₂, which is 42.4 kJ mol⁻¹ H₂ and 86.1 kJ mol⁻¹ H₂ lower than those of pristine LiAlH₄, respectively. Hence, it is reasonable to conclude that the addition of CoFe₂O₄ nanopowder has effectively lowered the kinetic barrier for the LiAlH₄ decomposition.

Fig. 6 Kissinger plots for the as-received LiAlH₄: (a) the first step and (b) the second step and LiAlH₄ doped with 7 mol% CoFe₂O₄: (c) the first step and (d) the second step.

In order to compare the change of morphology of the powder samples before/after ball milling and show the distribution of constitution elements of catalyst around the LiAlH₄ matrix, Fig. 7 presents the SEM images of the as-received LiAlH₄, ball-milled LiAlH₄ and the 2 mol% doped LiAlH₄ coupled with the elemental maps. As seen in Fig. 7a, the as-received LiAlH₄ sample consists of large irregular polyhedron particles, up to 40 μm in size. However, in Fig. 7b, the morphology of the as-milled LiAlH₄ became as amounts of regular globular particles with diameter ranging from 3 and 10 μm, reflecting a significant decrease in the particle size of LiAlH₄ after ball milling. Fig. 7(c–g) display the SEM images and the corresponding elemental maps of the LiAlH₄ + 2 mol% CoFe₂O₄ sample after mechanical ball-milling for 30 min. Microscopically, the grains of the LiAlH₄ + 2 mol% CoFe₂O₄ sample are fine but inhomogeneous, and the original particles were broken into smaller particles with the average size of about 6 μm by mechanical ball-milling. The tiny particles have a tendency to assemble and form stepped structures. As seen in Fig. 7(d–g), the elemental maps of constituent elements Al, O, Fe, and Co show uniform distribution of these species in the mixture, indicating that the catalyst of CoFe₂O₄ nanopowder could be well mixed with LiAlH₄ matrix through high energy ball milling. There is an existing good contact between the CoFe₂O₄ catalyst and the LiAlH₄ particles, resulting in the significantly enhanced dehydrogenation kinetics of LiAlH₄. Nevertheless, through comparison the elemental map O with that of other constituent elements of CoFe₂O₄ catalyst, it is worth to note that the elemental map of O has more distribution than that of Fe and Co, which is mainly caused by the oxidation during the specimen preparation process and oxygen element introduced from the conducting resin. Therefore, the high density surface defects and well dispersed catalyst introduce a larger amount of reaction nucleation sites the and hydrogen diffusion channels around the LiAlH₄ matrix for the dehydrogenation process, which results in the surface activation and obviously improved dehydrogenation properties of LiAlH₄.

Fig. 7 SEM micrographs of (a) as-received LiAlH₄ and (b) LiAlH₄ + 2 mol% CoFe₂O₄ after ball-milling. (c) SEM micrograph with (d)–(g) corresponding elemental maps of the 2 mol% CoFe₂O₄-doped sample.

IR spectra of the as-received LiAlH₄, as-milled LiAlH₄ and LiAlH₄ doped with 1 mol%, 2 mol%, 3 mol% and 5 mol% CoFe₂O₄ samples after ball milling are compared in Fig. 8. According to ref. 14, 24, 28, 40, 44 and 46, the active infrared vibrations of the Al–H bond for LiAlH₄ distribute at two regions, corresponding to 1600–1800 cm⁻¹ for the Al–H stretching modes and 800–900 cm⁻¹ for the Li–Al–H bending modes. While the active infrared vibrations for Li₃AlH₆ exhibit the Al–H stretching modes in the 1500–1400 cm⁻¹ region.^{14–16,28,44,56} For the CoFe₂O₄ doped LiAlH₄ samples shown in Fig. 8 (curves c–f), their active infrared vibration of the Al–H stretching modes appear at 1473 cm⁻¹, suggesting the existence of Li₃AlH₆ in the doped sample after ball milling. However, no Al–H bond peak of Li₃AlH₆ is found at the same position in the IR spectra of the as-received and as-milled LiAlH₄ (Fig. 8, curves a and b). The absorption intensity of the Li₃AlH₆ peak gradually strengthen with increasing CoFe₂O₄ catalyst content, which indicates that the content of Li₃AlH₆ continuously increases resulting from the decomposition proportion of LiAlH₄ raise with more CoFe₂O₄ catalyst. It is worth to note that the LiAlH₄ IR absorption peak cannot be observed when adding 5 mol% CoFe₂O₄ nanoparticles into the LiAlH₄ matrix, resulting from the 5 mol% CoFe₂O₄ doped sample complete decomposition and Li₃AlH₆ formation during the ball milling process. This phenomenon can be confirmed by the nonisothermal dehydrogenation performance of the 5 mol% doped LiAlH₄ (Fig. 1). Based on the comprehensive IR spectra analysis, it is concluded that the CoFe₂O₄-doped LiAlH₄ decomposition reaction occurs, forming the Li₃AlH₆ phase during the ball-milling process. The decomposition reaction of LiAlH₄ gradually intensifies with the increasing CoFe₂O₄ amount, and the details of the decomposition byproducts would be determined by the following XRD measurements.

Fig. 8 FTIR spectra of (a) as-received LiAlH₄, (b) as-milled LiAlH₄ and (c) 1 mol%, (d) 2 mol%, (e) 3 mol% and (f) 5 mol% CoFe₂O₄ doped LiAlH₄ after ball milling.

The above measurements confirm that some specific stoichiometric reactions between LiAlH₄ and CoFe₂O₄ occur during the ball-milling process. To clarify the phase transforms between LiAlH₄ and CoFe₂O₄ during the ball-milling process, Fig. 9 presents the XRD patterns of the as-milled LiAlH₄ and LiAlH₄ doped with 2 mol%, 3 mol% and 5 mol% CoFe₂O₄ after the ball milling process. In the XRD spectra of the as-milled LiAlH₄ all diffraction peaks correspond to the LiAlH₄ phase, without any additional decomposition products, suggesting that pure LiAlH₄ remains rather stable during the ball milling process.^{13–16,18,40,43,57,58} This point can also be proven by the non-isothermal dehydrogenation properties of the as-received and as-milled LiAlH₄ (Fig. 1), and the FTIR spectra of the as-milled LiAlH₄ (Fig. 8). However, compared with the as-milled LiAlH₄, the XRD patterns of the CoFe₂O₄ doped LiAlH₄ samples do not appear as just physical mixtures of LiAlH₄ and Co ferrite, which is in a good agreement with the FTIR results (Fig. 8). Adding 2 mol% CoFe₂O₄ nanoparticles into the LiAlH₄ matrix by mechanical milling causes weak diffraction peaks of microcrystalline aluminum and Li₃AlH₆ to appear in the XRD pattern. Meanwhile, the diffraction peaks of LiFeO₂ are observed at 41.3°, 44.7° and 34.8°, and the diffraction peaks at 31.3°and 44.8° correspond to AlCo, while the diffraction peaks of Fe₃O₄ are at 44.8°, and 65.1°. However, the CoFe₂O₄ phase could not be detected in the XRD patterns for the doped samples after the ball milling, which demonstrates that the reaction between LiAlH₄ and CoFe₂O₄ occurred during the ball-milling process. A similar phenomenon also appears in LiAlH₄ doped with other documented nanosized catalysts: MnFe₂O₄,¹⁴ Fe₂O₃,¹⁵ NiFe₂O₄¹⁶ and Nb₂O₅,⁴⁴ in which a complete reaction occurs between LiAlH₄ and the catalyst precursor, and subsequently the reaction products act as real catalysts for the succeeding decomposition of LiAlH₄. With increasing the Co ferrite content up to 3 mol%, the diffraction intensity of the decomposition products of Al, Li₃AlH₆, LiFeO₂ and Fe₃O₄ is gradually enhanced. The diffraction intensity of LiAlH₄ conspicuously declines, compared with that of 2 mol% doped LiAlH₄ sample, signifying that LiAlH₄ reacts with CoFe₂O₄, resulting in more LiAlH₄ decomposition during the ball milling process. Surprisingly, the diffraction peaks of LiAlH₄ cannot be observed for the 5 mol% doped sample, and all diffraction peaks correspond to the decomposition products, including LiFeO₂, Fe₃O₄, AlCo, Al and Li₃AlH₆, as seen in Fig. 9. This can be explained by the reaction between LiAlH₄ and CoFe₂O₄, leading to the complete decomposition of LiAlH₄ doped with CoFe₂O₄ during the ball milling process, causing the LiAlH₄ phase disappearance in the 5 mol% CoFe₂O₄-doped sample. In addition, the nano-sized CoFe₂O₄ phase cannot be detected in the XRD patterns of all doped samples, mainly because of the complete reaction between LiAlH₄ and CoFe₂O₄, forming LiFeO₂, AlCo, Al and Li₃AlH₆ phases. In the literature, a similar phenomenon has been reported, where NbF₅−,¹³ MnFe₂O₄−,¹⁴ NiCl₂−,³⁷ TiF₃−,³⁸ and TiO₂−,⁴³ as additives for LiAlH₄ also could not be detected after high energy ball-milling.

Fig. 9 XRD patterns for the as-milled LiAlH₄ and LiAlH₄ + 2 mol%, 3 mol% and 5 mol% CoFe₂O₄ after ball milling.

Fig. 10 displays the XRD patterns of the as-milled LiAlH₄ and 2 mol%, 3 mol% and 5 mol% CoFe₂O₄-doped LiAlH₄ after dehydrogenation at 250 °C. The XRD spectra of dehydrogenated as-milled LiAlH₄ only consists of Al and LiH phases as the dehydrogenation products, demonstrating that the first two dehydrogenation steps of LiAlH₄ have completed upon heating to 250 °C. In contrast, the XRD patterns of the doped samples show the dehydrogenation products containing not only Al and LiH phases, but also LiFeO₂, LiAlO₂, Fe_0.98O and Al_0.52Co_0.48 phases, which is quite different compared with the dehydrogenation products of the as-milled counterpart samples. Moreover, the diffraction peaks of LiFeO₂, LiAlO₂, Fe_0.98O and Al_0.52Co_0.48 phases gradually strengthen with the increasing CoFe₂O₄ amount. With respect to the significantly improved dehydrogenation performance of LiAlH₄ by doping CoFe₂O₄ nanoparticles, in situ formed reaction products may act as the catalyst for the first two dehydrogenation steps of LiAlH₄. Meanwhile, the reactions occurring during the dehydrogenation processes could facilitate the dehydrogenation dynamics of LiAlH₄. These favorable factors together provide a synergetic contribution to the significantly improved dehydrogenation properties of LiAlH₄.

Fig. 10 XRD patterns of the as-milled LiAlH₄ and LiAlH₄ + 2 mol%, 3 mol% and 5 mol% CoFe₂O₄ after dehydrogenation at 250 °C.

The above experimental results demonstrate that the CoFe₂O₄ nanopowder plays an important role in improving the dehydrogenation properties of Li alanate. The reasons leading to the significantly improved dehydrogenation properties, acquired in this work for the CoFe₂O₄-doped samples, could be summarized as follows: first, previous studies have revealed that the reaction thermodynamics could be affected by reducing the grain size.⁵⁹ The smaller particle size and a large number of created surface defects can introduce more reaction nucleation sites and hydrogen diffusion channels for the dehydrogenation process of LiAlH₄. Second, CoFe₂O₄ reacts with LiAlH₄ during the ball-milling process by forming a ternary Li–Fe oxide (LiFeO₂), Al–Co compound (AlCo) and Fe oxide (Fe₃O₄) species, suggesting that Co ferrite can transform into other new Co- and Fe-containing phases by increasing the high local temperature (demonstrated in ESI†) during the ball milling process. After dehydrogenation, the LiFeO₂, LiAlO₂, Fe_0.98O and Al_0.52Co_0.48 phases as the dehydrogenation products appear in the XRD patterns, and the diffraction intensity of these products gradually increases with further CoFe₂O₄ additive amount. These finely dispersed reaction products serve as the active sites for nucleation and growth of the dehydrogenation products, and the diffusion length of the reaction ions is largely shortened. Third, series of reactions between LiAlH₄ and CoFe₂O₄ occur by forming a ternary Li–Fe oxide, Fe oxide and Al–Co phases with a reduced valence state during heating. Thus, it is expected that these reactions could alter the reaction thermodynamics by lowering the enthalpy of the dehydrogenation reaction.⁴⁴ It is reasonable to conclude that the refinement of the LiAlH₄ powder combined with the reactions between LiAlH₄ and CoFe₂O₄ together contribute to the significantly improved dehydrogenation kinetics of LiAlH₄.

In order to comprehensively consider the catalytic effect of nano-sized CoFe₂O₄ for LiAlH₄, Fig. 11 shows the rehydrogenation results of the 2 mol% doped sample at 140 °C under 6.5 MPa pressure, followed by the subsequent desorption at 250 °C. After complete dehydrogenation during the first two reactions heated up to 250 °C, the sample was rehydrogenated at 140 °C under 6.5 MPa pressure. It is obvious that the rehydrogenation properties of the CoFe₂O₄ doped sample reach 0.15 wt% H₂ resorbed for the given conditions. Meanwhile, in order to confirm the rehydrogenation effect, Fig. 11 also provides the XRD pattern of the LiAlH₄ + 2 mol% CoFe₂O₄ sample after resorbing hydrogen for the given conditions in 2.5 h. The XRD spectra of the rehydrogenated sample shows almost identical results with the dehydrogenated sample, except for the appearance of few Li₃AlH₆ peaks, indicating that the second decomposition reaction of LiAlH₄ may be partially reversible by the catalytic effects of Co- and Fe-containing products. However, further study of hydrogen storage reversibility of the dehydrogenated LiAlH₄ is still underway.

Fig. 11 Rehydrogenation of LiAlH₄ + 2 mol% CoFe₂O₄ sample and its corresponding XRD pattern after hydrogen resorbtion at 140 °C under 6.5 MPa H₂ for 2.5 h.

4. Conclusions

In summary, the dehydrogenation properties of LiAlH₄ catalyzed by CoFe₂O₄ nanoparticles have been substantially improved compared with pure Li alanate powder. The onset desorption temperature of the 2 mol% CoFe₂O₄ doped LiAlH₄ sample is 65 °C, resulting in 90 °C decrease, compared with the as-received LiAlH₄. The rehydrogenation properties of the 2 mol% CoFe₂O₄ doped LiAlH₄ are inferior for the tested conditions, with 0.15 wt% H₂ resorbtion. The isothermal dehydriding kinetics shows that the LiAlH₄ + 2 mol% CoFe₂O₄ sample can release 6.8 wt% of hydrogen in 160 min under 0.1 MPa pressure, which is 6.1 wt% higher than that of the pristine LiAlH₄ under the same conditions (time, temperature and pressure). Furthermore, through the differential scanning calorimetry and the Kissinger desorption kinetics analyses, the apparent activation energy, E_a, of the 2 mol% CoFe₂O₄ doped sample are calculated to be 52.4 kJ mol⁻¹ H₂ and 86.5 kJ mol⁻¹ H₂ for the first two decomposition reactions, which are 42.4 kJ mol⁻¹ H₂ and 86.1 kJ mol⁻¹ H₂ lower than those of the pristine LiAlH₄, respectively. Based on the FTIR and XRD analyses of the doped samples, a series of reactions occurred between LiAlH₄ and CoFe₂O₄ during the ball-milling process, forming Al, Li₃AlH₆, LiFeO₂, Fe₃O₄, and Fe₃O₄ as decomposition products. These reactions proceeded upon heating, and the LiFeO₂, LiAlO₂, Fe_0.98O and Al_0.52Co_0.48 phases appeared. These in situ formed decomposition products, coupled with the reactions, play a synergistic role in remarkably improving dehydrogenation properties of LiAlH₄. From the conducted experiments it is reasonable to conclude that CoFe₂O₄ nanoparticles play a critical role in the significantly improved LiAlH₄ dehydrogenation performance.

Acknowledgements

The authors thank the National High-Tech R&D Program (863 Program) of China (2006AA05Z132) for financial support of this research. Fuqiang Zhai thanks the China Scholarship Council (CSC) for the scholarship.

References

1. L. G. Li, Q. F. Gu, Z. W. Tang, X. W. Chen, Y. B. Tan, Q. Li and X. B. Yu, J. Mater. Chem. A, 2013, 1, 12263–12269.
2. M. Ismail, Y. Zhao, X. B. Yu and S. X. Dou, RSC Adv., 2011, 1, 408–414.
3. R. Din, X. H. Qu, P. Li, L. Zhang, M. Ahmad, M. Z. Iqbal, M. Y. Rafique and M. H. Farooq, RSC Adv., 2012, 2, 4891–4903.
4. P. Sridechprasat, L. Phuirot, P. Rangsunvigit, B. Kitiyanan and S. Kulprathipanja, Energies, 2012, 5, 3691–3700.
5. F. Y. Cheng, Z. L. Tao, J. Liang and J. Chen, Chem. Commun., 2012, 48, 7334–7343.
6. H. Reardon, J. M. Hanlon, R. W. Hughes, A. G. Jopek, T. K. Mandalac and D. H. Gregory, Energy Environ. Sci., 2012, 5, 5951–5979.
7. A. Borgschulte, A. Jain, A. J. Ramirez-Cuesta, P. Martelli, A. Remhof, O. Friedrichs, R. Gremaud and A. Züttel, Faraday Discuss., 2011, 151, 213–230.
8. J. C. Fallas, W. M. Chien, D. Chandra, V. K. Kamisetty, E. D. Emmons, A. M. Covington, R. Chellappa, S. A. Gramsch, R. J. Hemley and H. Hagemann, J. Phys. Chem. C, 2010, 114, 11991–11997.
9. Y. B. Tan and X. B. Yu, RSC Adv., 2013, 3, 23879–23894.
10. D. Lacina, L. Yang, I. Chopra, J. Muckerman, Y. Chabal and J. Graetz, Phys. Chem. Chem. Phys., 2012, 14, 6569–6576.
11. http://www1.eere.energy.gov/hydrogenandfuelcells/storage/pdfs/targets_onboard_hydro_storage_explanation.pdf.
12. A. Andreasen, T. Veggea and A. S. Pedersena, J. Solid State Chem., 2005, 178, 3672–3678.
13. M. Ismail, Y. Zhao, X. B. Yu and S. X. Dou, Int. J. Hydrogen Energy, 2010, 35, 2361–2367.
14. F. Q. Zhai, P. Li, A. Z. Sun, S. Wu, Q. Wan, W. N. Zhang, Y. L. Li, L. Q. Cui and X. H. Qu, J. Phys. Chem. C, 2012, 116, 11939–11945.
15. Z. L. Li, P. Li, Q. Wan, F. Q. Zhai, Z. W. Liu, K. F. Zhao, L. Wang, S. Y. Lü, L. Zou, X. H. Qu and A. A. Volinsky, J. Phys. Chem. C, 2013, 117, 18343–18352.
16. P. Li, Z. L. Li, Q. Wan, F. Q. Zhai, X. Q. Li, X. H. Qu and A. A. Volinsky, J. Phys. Chem. C, 2013, 117, 25917–25925.
17. B. Bogdanovic and M. Schwickardi, J. Alloys Compd., 1997, 253, 1–9.
18. V. P. Balema, V. K. Pecharsky and K. W. Dennis, J. Alloys Compd., 2000, 313, 69–74.
19. J. Chen, N. Kuriyama, Q. Xu, H. T. Takeshita and T. Sakai, J. Phys. Chem. B, 2001, 105, 11214–11220.
20. V. P. Balema, J. W. Wiench, K. W. Dennis, M. Pruski and V. K. Pecharsky, J. Alloys Compd., 2001, 329, 108–114.
21. M. Resan, M. D. Hampton, J. K. Lomness and D. K. Slattery, Int. J. Hydrogen Energy, 2005, 30, 1413–1416.
22. X. P. Zheng, P. Li, F. Q. An, G. Q. Wang and X. H. Qu, Rare Met. Mater. Eng., 2008, 37, 400–403.
23. N. Mehraj-ud-din, R. Sami-ullah, C. S. So, S. W. Hwang, A. R. Kim and K. S. Nahm, Int. J. Hydrogen Energy, 2009, 34, 8937–8943.
24. R. A. Varin and L. Zbroniec, J. Alloys Compd., 2010, 506, 928–939.
25. H. W. Langmi, G. S. McGrady, X. F. Liu and C. M. Jensen, J. Phys. Chem. C, 2010, 114, 10666–10669.
26. X. F. Liu, H. W. Langmi, S. D. Beattie, F. F. Azenwi, G. S. McGrady and C. M. Jensen, J. Am. Chem. Soc., 2011, 133, 15593–15597.
27. K. L. Hima, B. Viswanathan and M. S. Srinivasa, Int. J. Hydrogen Energy, 2008, 33, 366–373.
28. R. Din, L. Zhang, P. Li and X. H. Qu, J. Alloys Compd., 2010, 508, 119–128.
29. L. H. Kumar, C. V. Rao and B. Viswanathan, J. Mater. Chem. A, 2013, 1, 3355–3361.
30. C. P. Hsu, D. H. Jiang, S. L. Lee, J. L. Horng, M. D. Gerb and J. K. Chang, Chem. Commun., 2013, 49, 8845–8847.
31. W. C. Hsu, C. H. Yang and W. T. Tsai, Int. J. Hydrogen Energy, 2014, 39, 927–933.
32. A. Andreasen, J. Alloys Compd., 2006, 419, 40–44.
33. J. R. Ares Fernandez, K. F. Aguey-Zinsou, M. Elsaesser, X. Z. Ma, M. Dornheim, T. Klassen and R. Bormann, Int. J. Hydrogen Energy, 2007, 32, 1033–1040.
34. Y. Suttisawat, P. Rangsunvigit, B. Kitiyanana, N. Muangsinb and S. Kulprathipanjac, Int. J. Hydrogen Energy, 2007, 32, 1277–1285.
35. X. P. Zheng, P. Li, I. S. Humail, F. Q. An, G. Q. Wang and X. H. Qu, Int. J. Hydrogen Energy, 2007, 32, 4957–4960.
36. Y. Kojima, Y. Kawai, M. Matsumoto and T. Haga, J. Alloys Compd., 2008, 462, 275–278.
37. T. Sun, C. K. Huang, H. Wang, L. X. Sun and M. Zhu, Int. J. Hydrogen Energy, 2008, 33, 6216–6221.
38. S. S. Liu, L. X. Sun, Y. Zhang, F. Xu, J. Zhang, H. L. Chu, M. Q. Fan, T. Zhang, X. Y. Song and J. P. Grolier, Int. J. Hydrogen Energy, 2009, 34, 8079–8085.
39. R. A. Varin and L. Zbroniec, J. Alloys Compd., 2011, 509S, S736–S739.
40. Z. B. Li, S. S. Liu, X. L. Si, J. Zhang, C. L. Jiao, S. Wang, S. Liu, Y. J. Zou, L. X. Sun and F. Xu, Int. J. Hydrogen Energy, 2012, 37, 3261–3267.
41. J. Fu, L. Röntzsch, T. Schmidt, M. Tegel, T. Weißgärber and B. Kieback, Int. J. Hydrogen Energy, 2012, 37, 13387–13392.
42. X. P. Zheng, J. J. Zheng, Q. H. Ma, S. L. Liu, X. Feng, X. B. Lin and G. Xiao, J. Alloys Compd., 2013, 551, 508–511.
43. M. Ismail, Y. Zhao, X. B. Yu, I. P. Nevirkovets and S. X. Dou, Int. J. Hydrogen Energy, 2011, 36, 8327–8334.
44. R. Din, X. H. Qu, P. Li, L. Zhang and M. Ahmad, J. Phys. Chem. C, 2011, 115, 13088–13099.
45. X. P. Zheng and S. L. Liu, J. Alloys Compd., 2009, 481, 761–763.
46. M. Ismail, Y. Zhao, X. B. Yu, A. Ranjbar and S. X. Dou, Int. J. Hydrogen Energy, 2011, 36, 3593–3599.
47. S. S. Liu, Z. B. Li, C. L. Jiao, X. L. Si, L. N. Yang, J. Zhang, H. Y. Zhou, F. L. Huang, Z. Gabelica, C. Schick, L. X. Sun and F. Xu, Int. J. Hydrogen Energy, 2013, 38, 2770–2777.
48. L. Li, F. Y. Qiu, Y. J. Wang, Y. N. Xu, C. H. An, G. Liu, L. F. Jiao and H. T. Yuan, Int. J. Hydrogen Energy, 2013, 38, 3695–3701.
49. L. Li, Y. N. Xu, Y. Wang, Y. J. Wang, F. Y. Qiu, C. H. An, L. F. Jiao and H. T. Yuan, Dalton Trans., 2014, 43, 1806–1813.
50. X. F. Liu, S. D. Beattie, H. W. Langmi, G. S. McGrady and C. M. Jensen, Int. J. Hydrogen Energy, 2012, 37, 10215–10221.
51. U. Eberle, M. Felderhoff and F. Schüth, Angew. Chem., Int. Ed., 2009, 48, 6608–6630.
52. S. Orimo, Y. Nakamori, J. R. Eliseo, A. Züttel and C. M. Jensen, Chem. Rev., 2007, 107, 4111–4132.
53. M. George, S. S. Nair, K. A. Malini, P. A. Joy and M. R. Anantharaman, J. Phys. D: Appl. Phys., 2007, 40, 1593–1602.
54. M. McCarty, J. N. Maycock and V. R. P. Verneker, J. Phys. Chem., 1968, 72, 4009–4014.
55. H. E. Kissinger, Anal. Chem., 1957, 29, 1702–1706.
56. S. S. Liu, Y. Zhang, L. X. Sun, J. Zhang, J. N. Zhao, F. Xu and F. L. Huang, Int. J. Hydrogen Energy, 2010, 35, 4554–4561.
57. R. A. Varin, L. Zbroniec, T. Czujko and Z. S. Wronski, Int. J. Hydrogen Energy, 2011, 36, 1167–1176.
58. P. B. Amama, J. T. Grant, P. J. Shamberger, A. A. Voevodin and T. S. Fisher, J. Phys. Chem. C, 2012, 116, 21886–21894.
59. J. Lu, Y. J. Choi, Z. Z. Fang, H. Y. Sohn and E. Ronnebro, J. Am. Chem. Soc., 2009, 131, 15843–15852.

---

Footnote

† Electronic supplementary information (ESI) available: XRD pattern for the as-milled LiAlH₄ doped with CoFe₂O₄ by using hand-milling method is provided in order to explain the fact that temperature is driving force for the reaction between LiAlH₄ and CoFe₂O₄. See DOI: 10.1039/c4ra00841c

---