Improved Dehydrogenation Properties of LiAlH₄ by Addition of Nanosized CoTiO₃

Abstract

Despite the application of lithium aluminium hydride (LiAlH₄) being hindered by its sluggish desorption kinetics and unfavourable reversibility, LiAlH₄ has received special attention as a promising solid-state hydrogen storage material due to its hydrogen storage capacity (10.5 wt.%). In this work, investigated for the first time was the effect of the nanosized cobalt titanate (CoTiO₃) which was synthesised via a solid-state method on the desorption behaviour of LiAlH₄. Superior desorption behaviour of LiAlH₄ was attained with the presence of a CoTiO₃ additive. By means of the addition of 5, 10, 15 and 20 wt.% of CoTiO₃, the initial desorption temperature of LiAlH₄ for the first stage was reduced to around 115–120 °C and the second desorption stage was reduced to around 144–150 °C, much lower than for undoped LiAlH₄. The LiAlH₄-CoTiO₃ sample also presents outstanding desorption kinetics behaviour, desorbing hydrogen 30–35 times faster than undoped LiAlH₄. The LiAlH₄-CoTiO₃ sample could desorb 3.0–3.5 wt.% H₂ in 30 min, while the commercial and milled LiAlH₄ desorbs <0.1 wt.% H₂. The apparent activation energy of the LiAlH₄-CoTiO₃ sample based on the Kissinger analysis was decreased to 75.2 and 91.8 kJ/mol for the first and second desorption stage, respectively, lower by 28.0 and 24.9 kJ/mol than undoped LiAlH₄. The LiAlH₄-CoTiO₃ sample presents uniform and smaller particle size distribution compared to undoped LiAlH₄, which is irregular in shape with some agglomerations. The experimental results suggest that the CoTiO₃ additive promoted notable advancements in the desorption performance of LiAlH₄ through the in situ-formed AlTi and amorphous Co or Co-containing active species that were generated during the desorption process.

1. Introduction

Hydrogen is a clean and renewable energy carrier and has become the most preferred alternative to fossil fuels for the future energy system. The advancement of secure and effective solid-state hydrogen storage that operates at low temperature was the focus in developing practical hydrogen as a potential energy carrier, offering a high capacity of hydrogen and excellent reversibility. Recently, alanates have been considered favourable materials for solid-state hydrogen storage with high hydrogen capacity and operating at low temperatures compared to other hydrides [1,2]. Among the alanates, lithium aluminium hydride (LiAlH₄) has attracted considerable research interest. LiAlH₄ offers 10.5 wt.% hydrogen content [3]. Below 250 °C, LiAlH₄ is able to liberate a theoretical capacity of 7.9 wt.% H₂ [4,5]. Basically, LiAlH₄ decomposes to LiH and Al in a three-step reaction as follows [6,7]:

3LiAlH₄ → Li₃AlH₆ + 2Al + 3H₂

(1)

(occurs at 150–175 °C, releasing 5.3 wt.% H₂)

Li₃AlH₆ → 3LiH + Al + 3/2H₂

(2)

(occurs at 180–220 °C, releasing 2.6 wt.% H₂)

3LiH + 3Al → 3LiAl + 3/2H₂

(3)

(occurs at >400 °C, releasing 2.6 wt.% H₂)

However, inadequate reversibility and unfavourable desorption kinetics of LiAlH₄ obstruct its progression to practical utilisation as a solid-state hydrogen storage medium [8]. Furthermore, the 2.6 wt.% H₂ in LiH as in Equation (3) is impracticable for applications owing to its high decomposition temperature (higher than 400 °C) [9]. To overcome these problems, several techniques have been introduced to accelerate the desorption behaviour of LiAlH₄—for example, the use of the nanoconfinement method and high energy ball milling [10,11,12,13], the usage of additives [14,15,16,17] and destabilisation with other hydrides [18,19,20].

The technique of high energy ball milling may exert a colossal effect on the behaviour of solid-state hydrogen storage materials by reducing the particle size that results in a larger surface area, hence shortening the distance of hydrogen diffusion [21,22]. The presence of additives, especially metal oxides, leads to a significant improvement, with a lower decomposition temperature and faster kinetic performance of LiAlH₄ due to enhanced catalytic support [23,24,25,26]. For instance, Li et al. [27] demonstrated satisfactory desorption behaviour of LiAlH₄ with the addition of NiCo₂O₄. The LiAlH₄-NiCo₂O₄ sample could release 4.95 wt.% H₂ at 130 °C in 150 min, while undoped LiAlH₄ only releases <1.0 wt.% H₂ within the same period. Then, Sulaiman et al. [28] showed that LiAlH₄ added with SrFe₁₂O₁₉ began to decompose at temperatures of 80 °C (first stage) and 130 °C (second stage), which are lower than undoped LiAlH₄. They also observed the formation of Fe, LiFeO₂, Sr and Sr-containing species during the heating process, which are advantageous in boosting the desorption properties of LiAlH₄. Furthermore, Li et al. [29] presented a study on the influence of CoFe₂O₄ on the desorption of LiAlH₄. Surprisingly, the performance of LiAlH₄ was significantly enhanced with rapid hydrogen desorption and low dehydrogenation activation energy. The LiAlH₄-CoFe₂O₄ sample could desorb 6.8 wt.% H₂ in 160 min, 6.1 wt.% higher than undoped LiAlH₄. Moreover, the activation energy of the LiAlH₄-CoFe₂O₄ composites decreased by 42.4 kJ/mol (first stage) and 86.1 kJ/mol (second stage) compared to pure LiAlH₄. The enhanced performance of the LiAlH₄-CoFe₂O₄ composite was attributed to the formation of Fe and Co species due to the reaction between LiAlH₄ and CoFe₂O₄. Then, another study revealed that the addition of TiO₂ could notably downshift the decomposition temperature of LiAlH₄ to 60 °C [30]. Desorption kinetics performance at 100 °C demonstrated that a TiO₂-doped LiAlH₄ composite could desorb 5.2 wt.% H₂ in 30 min, whereas the commercial LiAlH₄ desorbed only <0.2 wt.% H₂ in the same time but at a higher temperature (120 °C). The addition of TiO₂ offers significant enhancement of the performance of LiAlH₄ that may be ascribed to the introduction of a high density of defects on the surfaces of the TiO₂ particles throughout the ball milling process.

Inspired by the superior performance of the metal oxide additives, it is motivating to discover the consequence of another metal oxide on the desorption behaviour of LiAlH₄. The metal oxide additives may present outstanding performance compared to metal because of the bulk of oxides and the transition metal ions on the surface that may encounter different crystal fields due to the lack of ions at the surface of these oxides. This allows the electronic 3d state of the ions to be separated and is beneficial for the catalytic activity of transition metal oxides in the adsorption of gas molecules [31]. In this study, nanosized cobalt titanate (CoTiO₃) as an additive was added to LiAlH₄ by means of a ball milling process to ameliorate the desorption behaviour of LiAlH₄. Nanosized approaches are favourable in decreasing the materials’ particle size and reducing the hydrogen diffusion distance, accelerating the rate of hydrogen diffusion and nucleation sites for the desorption process [32]. Moreover, it is believed that nanosized CoTiO₃ could provide a superior performance of LiAlH₄ with low hydrogen release temperature and faster kinetics. Herein, different amounts of CoTiO₃ were milled together with LiAlH₄ to examine the desorption behaviour of LiAlH₄ as well as its kinetics behaviour. Moreover, the possible catalytic mechanism of the desorption process for this system has been discussed in depth.

2. Materials and Methods

Without any prior processing, LiAlH₄ (95% pure) from Sigma Aldrich was used. The nanosized CoTiO₃ was synthesised via the solid-state method, as explained in our previous work [33]. Then, various amounts of CoTiO₃ (5 wt.%, 10 wt.%, 15 wt.% and 20 wt.%) were doped with LiAlH₄ to study its influence on the desorption behaviour of LiAlH₄. The composite was milled together at a speed of 400 rpm for one hour using a planetary ball mill (NQM-0.4).

A Sievert-type pressure-composition-temperature apparatus from the Advanced Materials Corporation was operated to carry out the hydrogen desorption kinetics experiment and the temperature-programmed desorption (TPD). The TPD test was carried out in a vacuum chamber and heated to 250 °C for the purpose of finding the initial desorption temperature. Experiments for the isothermal desorption tests were carried out at 90 °C and 1.0 atm H₂ pressure. Using a Mettler Toledo TGA/DSC 1, differential scanning calorimetry (DSC) measurements were carried out to analyse the thermal characteristics and determine the activation energy. For each characterisation, 6–8 mg of the sample was loaded in a crucible and heated to 300 °C with a flow of argon (50 mL/min) at several heating rates (15 °C/min, 20 °C/min, 25 °C/min and 30 °C/min).

The phase structure of the LiAlH₄-CoTiO₃ sample after milling and after the desorption sample was examined using a Rigaku MiniFlex X-ray diffraction (XRD) system with Cu Kα radiation. The sample was tested at a rate of 2.00°/min over diffraction angles of 20° to 80°. The sample’s morphology and microstructure were then examined with a scanning electron microscope ((SEM; JEOL, Akishima, Tokyo, Japan) (JSM-6350LA)). The Fourier transform infrared (FTIR) spectrometer was tested in the range of 700 and 2000 cm⁻¹ using IR Shimadzu Tracer-100.

3. Results and Discussions

The influence of CoTiO₃ on the initial desorption temperature of LiAlH₄ was characterised by the TPD experiment, as presented in Figure 1. By referring to the figure, the addition of CoTiO₃ results in superior desorption performance of LiAlH₄. Commercial LiAlH₄ begins to release hydrogen at 145 °C (first stage) and 175 °C (second stage) releasing around 7.4 wt.% H₂. After 1 h of milling, the temperature to release hydrogen was similar to the commercial LiAlH₄, demonstrating that the milling process had a negligible influence on the desorption temperature of LiAlH₄. Contrarily, by adding different amounts of CoTiO₃ (5, 10, 15 and 20 wt.%), the hydrogen release initiates at a much lower temperature than commercial and milled LiAlH₄. The LiAlH₄ + 5 wt.% CoTiO₃ sample began to release hydrogen at 120 °C (first stage) and 150 °C (second stage), releasing 6.9 wt.% H₂. Furthermore, increasing the amount of CoTiO₃ to 10 wt.% resulted in a decrease in hydrogen release temperature to 115 °C (first stage) and 145 °C (second stage), releasing 6.2 wt.% H₂. Adding 15 wt.% of CoTiO₃ resulted in a similar temperature for hydrogen release as the LiAlH₄ + 10 wt.% CoTiO₃ sample. For the LiAlH₄ + 20 wt.% CoTiO₃ sample, hydrogen release occurs at 118 °C (first stage) and 144 °C (second stage), releasing 6.0 wt.% H₂. The hydrogen release temperature for all of the CoTiO₃-doped LiAlH₄ samples was much lower for than the undoped LiAlH₄. Notably, with the amount of hydrogen released being slightly reducing compared to the undoped LiAlH_4, the presence of CoTiO₃ provides positive contributions in lowering the temperature for the LiAlH₄ to release hydrogen. The amount of hydrogen released for all of the CoTiO₃-doped LiAlH₄ samples was slightly reduced compared to undoped LiAlH₄ due to the deadweight of CoTiO₃ that does not hold any hydrogen [10,34].

Additionally, Figure 2 demonstrates the desorption kinetic performance of LiAlH₄ with and without CoTiO₃ at 90 °C. As shown in Figure 2, the LiAlH₄ + xCoTiO₃ (x = 5, 10, 15 and 20 wt.%) releases a large amount of hydrogen in a short period of time. Within 30 min, the LiAlH₄-CoTiO₃ sample releases 3.0–3.5 wt.% H₂. However, the undoped LiAlH₄ shows sluggish kinetics, with the ability to release only <0.1 wt.% H₂ within the same period. The addition of CoTiO₃ results in accelerated desorption kinetics which are 30–35 times faster than undoped LiAlH₄. The rapid hydrogen release from the LiAlH₄-CoTiO₃ sample could be related to the creation of surface defects and active materials as a result of a reaction between LiAlH₄ and CoTiO₃. Therefore, it can be concluded that the kinetic behaviour of LiAlH₄ can be enhanced by introducing CoTiO₃. Considering the initial desorption temperature and the desorption kinetics performance, the LiAlH₄ + 10 wt.% CoTiO₃ sample is selected to further examine the effect of CoTiO₃ on the desorption behaviour of LiAlH₄.

The effect of CoTiO₃ on the thermal behaviour of LiAlH₄ was characterised by DSC, as indicated in Figure 3. By referring to Figure 3, the thermal properties of as-milled LiAlH₄ consist of four peaks attributed to the two endothermic and two exothermic peaks. The first exothermic peak (140 °C) was due to the interaction of hydroxyl impurities at the surface of LiAlH₄, while the first endothermic peak (163 °C) is attributed to the melting of LiAlH₄. The second exothermic peak (175 °C) is ascribed to the process of LiAlH₄ decomposing to Li₃AlH₆ and Al, while the second endothermic peak (230 °C) relates to the process of Li₃AlH₆ decomposing to LiH and Al. Upon the addition of CoTiO₃, the number of peaks was decreased to two, namely one endothermic peak and one exothermic peak, respectively, that occur at a lower temperature than for milled LiAlH₄. The exothermic peak (LiAlH₄ decomposition) occurs at 105 °C and the endothermic peak (Li₃AlH₆ decomposition) occurs at 196 °C. Implying that the melt of LiAlH₄ is inhibited after the addition of the nanosized CoTiO₃, the LiAlH₄ + 10 wt.% CoTiO₃ sample begins to decompose prior to melting. These observations are consistent with the previous reports [35,36].

To scrutinise the impact of nanosized CoTiO₃ addition on the desorption activation energy of LiAlH₄, the DSC measurement was performed at several heating rates. The DSC profiles of as-milled LiAlH₄ and LiAlH₄ + 10 wt.% CoTiO₃ at several heating rates are shown in Figure 4. As depicted in the figure, the decomposition peak of LiAlH₄ + 10 wt.% CoTiO₃ occurs at a lower temperature than that of undoped LiAlH₄. The apparent activation energy (E_A) could be fitted from the DSC curve and calculated by Kissinger analysis as follows:

ln[β/T_p²] = −E_A/RT_p + A,

(4)

where β is the heating rate, T_p is the endothermic peak related to the decomposition temperature, E_A is the activation energy, R is the gas constant and A is a linear constant.

Then, the activation energy is determined based on the Kissinger plot of ln [β/T_p²] versus 1000/T_p, as in Figure 5. The activation energy of the undoped LiAlH₄ for the first and second desorption stage are 103.2 and 116.7 kJ/mol, respectively, based on the Kissinger plot. By adding nanosized CoTiO₃, the activation energy of LiAlH₄ was notably decreased to 75.2 and 91.8 kJ/mol for the first and second desorption stage, respectively, being lowered by 27% for the first stage and 21% for the second stage compared to undoped LiAlH₄. The reduced activation energy was in good agreement with the reduced decomposition temperature. From these results, it can be clearly seen that the introduction of CoTiO₃ as an additive remarkably reduced the activation energy of LiAlH₄, demonstrating that CoTiO₃ has a favourable impact on the desorption behaviours of LiAlH₄.

Figure 6 displays the SEM micrograph of commercial LiAlH₄, as-milled LiAlH₄ and LiAlH₄ + 10 wt.% CoTiO₃ sample. By referring to Figure 6a, commercial LiAlH₄ consists of large uniform and blocky-shaped particles. After undergoing the milling process of one hour (Figure 6b), the blocky shape was broken into non-uniform smaller particles, but with some agglomerations. Meanwhile, after the addition of CoTiO₃, as shown in Figure 6c, the particle was transformed to a finer shape and size. The morphology of the composite sample (CoTiO₃-doped LiAlH₄) presents more uniform distributions with fewer agglomerations. The difference in morphology between the milled LiAlH₄ and LiAlH₄ + 10 wt.% CoTiO₃ sample may be due to the presence of an additive that acts as a lubricant and prevents the agglomerations of the sample. These superfine particles could facilitate the rapid hydrogen desorption of LiAlH₄ [32]. As reported previously, metal oxide-based additives have shown good lubricating performance [37,38]. This indicates that the addition of CoTiO₃ is favourable in constraining the agglomeration of the LiAlH₄ and results in a smaller and finer shape of the particle.

Figure 7 presents the particle distribution size of commercial LiAlH₄, as-milled LiAlH₄ and LiAlH₄ + 10 wt.% CoTiO₃ sample evaluated by the Image J software. The average particle size of the commercial LiAlH₄, as-milled LiAlH₄ and LiAlH₄ + 10 wt.% CoTiO₃ sample was calculated to be ~44, 0.5 and 0.3 µm, respectively. The morphological alteration and significant size reduction are responsible for the large surface defects and the expansion of grain boundaries across the composite’s surface [39]. Resulting in more reactions for the nucleation sites and better diffusion channels of hydrogen being achieved with amplified grain boundaries that enhance the desorption behaviour of the LiAlH₄-CoTiO₃ sample with rapid desorption kinetics and lower activation energy, accelerating the rate of hydrogen diffusion [40], the reduction in particle size also results in a shorter hydrogen diffusion path. Similarly, other findings also found that smaller and finer particle distributions demonstrated superior dehydriding performance of LiAlH₄ [36,41].

To discover the reaction mechanism during the milling process, XRD analysis was carried out as presented in Figure 8. Figure 8a shows the XRD peaks of commercial LiAlH₄, and it appears that only peaks of LiAlH₄ were discovered, indicating the high purity of LiAlH₄. For the milled sample (Figure 8b), similar to commercial LiAlH₄, only peaks of LiAlH₄ were detected, suggesting that, similar to a previous report [42], LiAlH₄ maintains high stability throughout the milling process. Meanwhile, for the LiAlH₄ + 10 wt.% CoTiO₃, as shown in Figure 8c, only LiAlH₄ peaks were visible, while the peaks of CoTiO₃ were not identified by the XRD, indicating that the amount of CoTiO₃ was too small to be detected by the XRD. This phenomenon was comparable to a prior work where several additives such as Ti₃C₂ and FeCl₂ were not detected by XRD [2,43].

Figure 9 depicts the FTIR spectra of commercial LiAlH₄, as-milled LiAlH₄ and LiAlH₄ + 10 wt.% CoTiO₃ sample. FTIR analysis was performed to determine the existence of Li₃AlH₆ after the milling process. All three samples exhibit two distinct region modes around 800–900 cm⁻¹, corresponding to the Li-Al-H bending mode, and 1600–1800 cm⁻¹, attributed to the Al-H stretching mode. However, for the LiAlH₄ + 10 wt.% CoTiO₃ sample (Figure 9c), an extra absorbance peak around 1400 cm⁻¹ was identified, indicating the presence of the Al-H stretching mode corresponding to the Li₃AlH₆. This result proves that LiAlH₄ was slightly decomposed to Li₃AlH₆ during the milling process of LiAlH₄ and CoTiO₃.

To explore the specific mechanism and catalytic activity of CoTiO₃ that contributed to the superior desorption behaviours of LiAlH₄, the 10 wt.% and 20 wt.% CoTiO₃-doped LiAlH₄ samples after desorption at 250 °C were analysed using XRD, as depicted in Figure 10. After the desorption process of the 10 wt.% CoTiO₃-doped LiAlH₄ sample (Figure 10a), the main peaks detected are the dehydrogenated products of LiAlH₄, which are LiH and Al, denoting that the complete desorption process of LiAlH₄ occurs. Moreover, the peaks of AlTi were also detected, indicating that there is a reaction between LiAlH₄ and CoTiO₃. However, the peaks of Co or Co-containing species were not identified. The additional characterisation was carried out with 20 wt.% of CoTiO₃ (Figure 10b). Similarly, the main peaks of LiH and Al were observed with the additional peak of AlTi, and the fact that no peaks of Co or Co-containing species were identified after the desorption process may be due to the amorphous form of Co or Co-containing species.

Referring to the results, it is noteworthy to state that the desorption behaviours of LiAlH₄ were improved due to the in situ formation of active species during the heating process. In this study, the formation of AlTi and Co or Co-containing species was believed to contribute to the enhanced desorption behaviours of LiAlH₄. This phenomenon agrees with a previous study that also reported the formation of AlTi after the dehydrogenation process of LiAlH₄ [3,44]. The in situ formation of AlTi is beneficial in ameliorating the desorption behaviour of LiAlH₄ by catalysing the dehydrogenation through reaction in Equations (1) and (2). In addition, the study performed by Wohlwend et al. [45] revealed the superior performance of LiAlH₄ with the addition of Ti-based additives. The Ti atom reacts strongly with the LiAlH₄ surface and reduces the H binding energy. The reduction in binding energy may be a result of the correlation of the charge transfer change between Al and H that results in the accelerated kinetic performance of LiAlH₄. It denotes that Co or Co-containing species also occupy a dominant role in improving the desorption behaviour of LiAlH₄, even though the Co or Co-containing phase was not detected. For instance, a previous study discovered that the presence of Co₂O₃ exhibited a significant enhancement in the desorption of LiAlH₄ compared to undoped LiAlH₄ [46]. Li et al. [29] also indicated that the formation of Co-containing species after the desorption process of LiAlH₄ provides a positive impact on the desorption behaviour of LiAlH₄. Therefore, it is reasonable to infer that the formation of these active species contributes to the notable advancement of the desorption behaviour of LiAlH₄. By shortening the length of diffusion of the reaction ions [29,41], these in situ-formed AlTi and Co or Co-containing species are responsible for boosting the desorption behaviour of LiAlH₄ by taking part as the active sites for the nucleation and creation of the dehydrogenation yield. However, further characterisation using X-ray photoelectron spectroscopy and transmission electron microscopy is needed to evaluate the actual impact and mechanism of the LiAlH₄-CoTiO₃ system.

4. Conclusions

The introduction of nanosized CoTiO₃ that was synthesised via a solid-state method significantly ameliorates the desorption behaviour of LiAlH₄. The LiAlH₄-CoTiO₃ sample began to release hydrogen at around 115–120 °C and 144–150 °C for the first and second desorption stages, respectively, lower than for undoped LiAlH₄. The addition of nanosized CoTiO₃ also results in accelerated desorption kinetics with the ability to desorb hydrogen 30–35 times faster than undoped LiAlH₄. The rapid hydrogen desorption performance of LiAlH₄ when added with CoTiO₃ may be due to the lower activation energy of the LiAlH₄-CoTiO₃ sample, which was calculated to be 75.2 kJ/mol (first stage) and 91.8 kJ/mol (second stage), reduced by 28.0 and 24.9 kJ/mol compared to undoped LiAlH₄. The nanosized CoTiO₃ was also beneficial in the reduction of LiAlH₄ particle size during the milling process. Smaller particle size is favourable in reducing the hydrogen diffusion length, which then results in superior desorption performance. The enhanced desorption behaviour of LiAlH₄ was also due to the synergetic effect of the in situ formation of AlTi and Co or Co-containing species during the heating process of LiAlH₄ and CoTiO₃. By reducing the initial desorption temperature and activation energy, the addition of nanosized CoTiO₃ remarkably ameliorates the desorption behaviour of LiAlH₄, resulting in finer and smaller particles and accelerating the desorption kinetic behaviour. These findings shed light on the preparation of LiAlH₄ hydrogen storage systems for mobile applications.