Mechanisms of partial hydrogen sorption reversibility in a 3NaBH₄/ScF₃ composite

Abstract

A new hydrogen storage composite containing NaBH₄ and a 3d transition metal fluoride, 3NaBH₄/ScF₃, was synthesized via ball milling. The composite shows no reaction during milling and its dehydriding process can be divided into three steps upon heating: (i) partial substitution of H⁻ by F⁻ in NaBH₄ to form NaBH_xF_4−x at the early stage, releasing about 0.19 wt% of hydrogen; (ii) formations of Na₃ScF₆, NaBF₄ and ScB₂ through the reaction between NaBH₄ and ScF₃, with 2.52 wt% of hydrogen release and a dehydriding activation energy of 162.67 kJ mol⁻¹ H₂; (iii) further reaction of residual NaBH₄ and Na₃ScF₆ to form NaF, B and ScB₂, with a dehydriding activation energy of 169.37 kJ mol⁻¹ H₂. The total hydrogen release of the composite reaches 5.54 wt% at 530 °C. The complete dehydrided composite cannot be rehydrogenated while the products after the second dehydriding step can be hydrogenated with an absorption activation energy of 44.58 kJ mol⁻¹ H₂. These results demonstrate that by adding 3d transition metal fluorides into NaBH₄, a partial reversibility in NaBH₄ can be achieved.

---

1 Introduction

Hydrogen is one of the most promising alternative and attractive clean energy sources that can substitute fossil fuels, with sufficient energy density and environment-friendliness.^1,2 Nevertheless, almost a century since the concept of “hydrogen economy” was introduced by Jules Velne,^3,4 it is still challenging to find reliable, flexible and cost-efficient hydrogen media for on-board, stationary and portable applications.^5,6 In the past few decades, great attention has been paid to both hydrogen production technologies and a variety of hydrogen storage methods,^7–10 including the use of different compounds,^11,12 especially complex hydrides, of which borohydrides are typical representative.^13–15 These solid-state hydrogen storage materials offer some advantages over high pressure gaseous storage and low temperature liquid storage, such as high capacity, high safety, and low cost. In particular, alkali metal borohydrides are regarded as possible hydrogen carries, ascribed to their contribute to their high gravimetric and volumetric hydrogen density, together with good stability.¹⁶ However, current research results showed that few technologies regarding the use of metal borohydrides as hydrogen storage carriers are able to fulfill the requirements established by the US Department of Energy.¹⁷

Early investigations into borohydrides have shown that, compared to other borohydrides such as LiBH₄ and KBH₄, NaBH₄ is stable under alkaline conditions, and undergoes hydrolysis through the following reaction:¹⁸

NaBH₄ + 4H₂O → NaBO₂·2H₂O + 4H₂ (1)

However, the use of NaBH₄ as a hydrogen generator through hydrolysis faces several issues related to the catalyst durability, and/or poisoning, as well as storage vessels.^19–21 In contrast, thermal decomposition of NaBH₄ has emerged as potential alternative method for hydrogen storage. J. Urgnani et al. investigated the thermal decomposition behaviors of NaBH₄, and proposed that NaBH₄ would decompose in two steps according to the following reactions:²²

NaBH₄ → Na + B + 2H₂ (3)

For the sake of improving the thermodynamic and kinetic properties of the thermal decomposition of NaBH₄, many strategies have been taken, such as adding catalyst, building thermodynamic destabilization system, nano-engineering and chemical modifications.^13,23–25 For instance, NaBH₄/MgH₂ system could fulfil its decomposition before melting due to the formation of MgB₂ in the system.²⁶ Czujiko et al. suggested that pure Mg could lower down the desorption temperature of NaBH₄ through catalytic effect.²⁷ In addition, the decomposition of NaBH₄ may occur at lower temperatures with some reversibility through the addition of Ni-based catalysts,²⁸ which facilitates hydrogen release and improves the reversibility to some extent.²⁹ Moreover, the chemical reaction that regenerates borohydrides from metal–borides occurs much easier over the regeneration from boron since less energy is required for breaking the chemical bond between B–M (M means metal) relative to the B–B's.³⁰

In previous works, many research works regarding hydrogen storage composite systems have been carried out, and some of them have explored how rare earth element (RE) addition can effect the thermodynamics and kinetics properties of metal borohydrides based systems, i.e., NaBH₄–YF₃,³¹ NaBH₄–ScCl₃,³² LiBH₄–YCl₃,³³ etc. In particular, hydrogen sorption reversibility was achieved in 3NaBH₄/LnF₃ systems with good thermodynamic and kinetic properties.³⁴

Considering that scandium lies in the III B column of the periodic table as the lanthanide elements, we attempt to prepare a new hydrogen storage system, 3NaBH₄/ScF₃, through ball milling method. Our previous investigations on NaBH₄–MF₃ (M = metal) systems proved that the molar ratio of 3 : 1 was the best one, i.e. 3NaBH₄/LnF₃ (Ln = La, Ce, Nd, Gd, Yb),³⁴ 3NaBH₄/YF₃,³¹ 3NaBH₄/PrF₃ and 3NaBH₄/HoF₃.^35,36 If the molar ratio of NaBH₄ to MF₃ is higher than 3 : 1, some NaBH₄ will be left after the desorption due to the incomplete reaction. While if the ratio is less than 3 : 1, MF₃ is excessive, and the overall hydrogen sorption capacity is reduced since MF₃ can neither release nor absorb hydrogen. Consequently, the ratio of NaBH₄ to ScF₃ is set as 3 : 1 in the present work. We conducted a detailed study of hydrogen sorption behaviors of the 3NaBH₄/ScF₃ system, and proposed mechanisms of hydrogen sorption in this composite, depending on experimental analyses.

2 Experimental

2.1 Sample preparation

NaBH₄ (Aladdin Reagent Database Inc., 96%) and ScF₃ (Aladdin Reagent Database Inc., 98%) were used as starting materials without further purification and mixed in the molar ratio of 3 : 1 in a planetary ball miller whose type is QM-1SP2. The stainless steel vessel with 100 ml volume was used to load 0.3928 g of NaBH₄ and 0.3530 g of ScF₃ powders together with 25 stainless steel balls (diameter of 5 mm, average weight of 0.8950 g each). The ball to powder weight ratio is approximately 30 : 1. Ball milling is conducted at a rotation speed of 400 rpm for 180 min. The prearrangement, manipulation and storage of specimen were carried out in an Ar-filled Lab 2000 glove box (Etelux inert gas system Co., Ltd.) in which neither moisture nor oxygen concentration beyonds 10 ppm.

2.2 Characterization

The analyses of phase composition for the ball milled, dehydrogenated and rehydrogenated samples were accomplished using an X-ray diffraction apparatus (D/max 2550VL/PC), equipped with a Cu-Kα radiation source. For XRD tests, composite powder at different states were kept into specific sample holders with arched glass on both sides and airtight PVC tape on the top, to isolate them from air. Meanwhile, the broad peak at around 2θ = 15° in the patterns is caused by the tape. Using a Spectrum Nicolet iS5 produced by Thermo Fisher Scientific Inc., Fourier transform infrared spectroscopy (FTIR) tests were performed on samples with different states in an Ar filled glove box. In principle of volumetric methods, we carried out temperature-programmed-desorption (TPD) measurements on 0.2 g of the 3NaBH₄/ScF₃ composite sample from room temperature to about 530 °C under an initial vacuum condition, with a heating rate R_H of 3 °C min⁻¹. Evaluations of hydrogen absorption performance were implemented for 10 h at various temperatures under around 3.2 MPa H₂ pressure.

Dehydrogenation behaviors of composites were determined by synchronous thermal analyzer (Differential Scanning Calorimetry/Thermal Gravimetry, DSC/TG), in a Netzsch, STA 449 F3 Jupiter equipment, in which R_H = 3 °C min⁻¹, 5 °C min⁻¹, 7 °C min⁻¹ and 10 °C min⁻¹, respectively, starting from room temperature to about 500 °C, under the protection of 0.1 MPa Ar flow. To compare the properties of hydrogen storage performances of different composites, data of weight percent considering hydrogen release and uptake during the tests was assessed based on the samples' initial weight value.

3 Results and discussions

3.1 Dehydrogenation of the 3NaBH₄/ScF₃ composite

The effect of the ScF₃ addition on the hydrogen desorption behaviors of NaBH₄ was examined by DSC measurements at different R_H, namely 3 °C min⁻¹, 5 °C min⁻¹, 7 °C min⁻¹ and 10 °C min⁻¹, as well as TG measurement at the R_H of 10 °C min⁻¹, as Fig. 1 showed.

Fig. 1 DSC curves of 3NaBH₄/ScF₃ composite samples under heating rates of 3 °C min⁻¹, 5 °C min⁻¹, 7 °C min⁻¹ and 10 °C min⁻¹ and TG profile at the heating rate of 10 °C min⁻¹ (a) and the corresponding Kissinger plots for the two major endothermic desorption steps (b).

Two major endothermic peaks upon heating appeared on the DSC curves, indicating that two main desorption steps took place during dehydrogenation. According to the starting point of dehydriding in DSC profiles, the first major dehydrogenation begins at 356 °C with R_H = 3 °C min⁻¹ heating rate condition. In contrast, pure NaBH₄ shows a dehydrogenation temperature of 517 °C under the same condition.³¹ In addition, a subsequent broad endothermic peak was also recorded. TG curve obtained at the heating rate of 10 °C min⁻¹ shows that the total mass loss reaches 4.10 wt% at 500 °C. According to the DSC/TG profiles, the dehydrogenation enthalpies of the first and second major desorption reactions are determined to be 27.43 ± 5 kJ mol⁻¹ H₂ and 29.54 ± 2 kJ mol⁻¹ H₂, respectively.

The apparent dehydrogenation activation energy (E_a) of the 3NaBH₄/ScF₃ sample could be determined using the Kissinger method,³⁷ as described below:

where heating rate (β), peak temperature (T_m), and gas constant (R) show a specific relationship. Table 1 gives the peak temperatures in DSC curves at various R_H obtained from Fig. 1a. The fitting plot displays that ln(β/T_m²) and 1/T_m have good linearity, as shown in Fig. 1b. According to eqn (4), the E_a value is calculated to be 162.67 kJ mol⁻¹ H₂ and 169.37 kJ mol⁻¹ H₂ for the first major desorption step and second major desorption step, respectively.

Table 1 The DSC peak temperatures of the 3NaBH₄/ScF₃ at different heating rates under 0.1 MPa argon atmosphere

Sample	Heating rate/°C min^−1	Temperature of peaks/°C
3NaBH4/ScF3	3	373	424
	5	383	440
	7	392	442
	10	397	452

---

To obtain further information of dehydrogenation process of the target system, TPD measurement was performed on the ball-milled composite with a constant R_H of 3 °C min⁻¹ from ambient temperature to 530 °C and the results are shown in Fig. 2. The results exhibit that 3NaBH₄/ScF₃ composite has an appropriate onset dehydrogenation temperature, which is actually lower than 200 °C in vacuum. Meanwhile, the figure shows that the desorption behavior may be subdivided into three consecutive processes, with a small amount of hydrogen (∼0.19 wt%) released at temperature lower than 310 °C in the initial process, and the latter two steps range from 310 °C to 420 °C, and above 420 °C, releasing about 2.52 wt% and 2.83 wt% of hydrogen, respectively. In comparison to the DSC measurements, the second and third desorption steps shown in TPD profile should correspond to the two major endothermic peaks on DSC curves.

Fig. 2 TPD profile of the 3NaBH₄/ScF₃ sample at a heating rate of 3 °C min⁻¹.

The entire hydrogen release obtained from the experimental process up to 530 °C is 5.54 wt%, which is over 95% of the theoretical hydrogen content. Previous study shows that during dehydrogenation process of pure NaBH₄ under the same condition, only 0.68 wt% weight loss is observed when heated up to 482 °C.¹⁹ Thus, the hydrogen desorption properties of NaBH₄ were significantly promoted by the addition of ScF₃. However, from Fig. 1, there is no endothermic peak present from ambient to 300 °C in DSC curves. C. Bonatto Minella reported a related phenomenon and the difference between DSC analyses and volumetric measurements was ascribed to dissimilar experimental conditions.³⁸

XRD analyses were performed on samples treated under a series of controlled conditions in order to have a better understanding of mechanisms of de/rehydrogenation in the 3NaBH₄/ScF₃ composite, as shown in Fig. 3. The results show that no new product can be found in the sample after ball milling, except NaBH₄ (JCPDS no. 09-0386) and ScF₃ (JCPDS no. 46-1243), which means that only physical mixing takes place during milling process. Before dehydrogenation at various temperatures of 300 °C, 420 °C and 530 °C, the loaded sample container was first placed under a 4.5 MPa H₂ pressure, then followed by a quick temperature rising and heat preservation. At last, the sample chamber was evacuated and powders were treated at the expected temperature for 3 h. After dehydrogenation at 300 °C under vacuum, it can be clearly seen in Fig. 4 that diffraction peaks from NaBH₄ in the 3NaBH₄/ScF₃ composite shifted slightly from high angle side to lower angle side, demonstrating a lattice expansion of NaBH₄ after the first dehydrogenation.

Fig. 3 XRD patterns of the 3NaBH₄/ScF₃ composite after ball milling (a), dehydrogenated at 300 °C (b), dehydrogenated at 420 °C (c), fully dehydrogenated at 530 °C (d) and rehydrogenated at 420 °C (e).

Fig. 4 XRD patterns of 3NaBH₄/ScF₃ composite after ball milling (a) and after dehydrogenation at 300 °C (b) in the range of low diffraction angle.

On the basis of the XRD patterns and using the RIR analysis, the lattice parameters of NaBH₄ in ball milled sample is calculated to be: a = 0.6165 nm, b = 0.6184 nm, c = 0.6153 nm and α = β = γ = 90°. After the first step dehydrogenation at 300 °C, the lattice constants of NaBH₄ changed into a = 0.6185 nm, b = 0.6225 nm, c = 0.6166 nm and α = β = γ = 90°. Such a lattice expansion of NaBH₄ might be attributed to the fact that H⁻ was partially substituted by F⁻ in the unit cell. Similar phenomenon was also observed in some previous works,^29,39–43 for which the lattice expansion was attributed to the formation of an intermediate compound NaBH_xF_4−x. In the present work, the formation of NaBH_xF_4−x occurred at the first desorption step between NaBH₄ and ScF₃, together with releasing small amount of hydrogen, as also seen in the 3NaBH₄/NdF₃, 3NaBH₄/PrF₃, 3NaBH₄/HoF₃ systems,^19,35,36 and was regarded as an energy favorable process in theory.⁴⁴ Upon heating at temperatures higher than 310 °C, along with the reduction of NaBH₄ and disappearance of ScF₃ (Fig. 3c), Na₃ScF₆ appeared in the system, indicating that a major reaction between NaBH₄ and ScF₃ occurred. It has been also indicated by Radovan Cerny et al. that in the case of NaBH₄/ScCl₃ system, Na₃ScCl₆ and NaSc(BH₄)₄ formed as a result of the reaction between NaBH₄ and ScCl₃.³² Chong et al. reported that the reaction occurred at 250 °C between NaBH₄ and HoF₃ could produce NaHo(BH₄)₄ and NaHo₂F₇ phases.³⁶ In the 3NaBH₄/ScF₃ composite, NaSc(BH₄)₄ might form upon heating and then decomposed to ScB₂ and H₂. Based on the XRD analysis, the second dehydrogenation step before 420 °C can be described as:

27NaBH₄ + 20ScF₃ = 8Na₃ScF₆ + 12ScB₂ + 54H₂ + 3NaBF₄ (5)

Such a reaction has a theoretical hydrogen release of 2.53 wt%, close to the value measured from TPD for the second step dehydrogenation. At around 500 °C, the remaining NaBH₄ reacts with Na₃ScF₆ and NaBF₄ to produce NaF, ScB₂ and B, as shown in the indexed XRD pattern of Fig. 3d. Thus, the third dehydrogenation step can be described as follows:

33NaBH₄ + 8Na₃ScF₆ + 3NaBF₄ = 60NaF + 8ScB₂ + 66H₂ + 20B (6)

with a theoretical hydrogen desorption value of 3.08 wt%. This value is also close to what is observed in the TPD analysis for the third dehydrogenation step. According to Garroni et al.,⁴⁵ Na₂[B₁₂H₁₂] is usually a byproduct during desorption of NaBH₄ based composites, which forms in an intermediate step and is still present at the end of reaction. Na₂[B₁₂H₁₂] is found to be a stable byproduct and cannot be re-hydrogenated to NaBH₄, thus is regarded as an unfavorable product for the reversibility.^46,47 However, the Na₂[B₁₂H₁₂] phase was not found in the XRD pattern of the partial or the complete dehydrogenated 3NaBH₄/ScF₃ samples, which means that very small amount or even no such a byproduct was generated during the decomposition of the 3NaBH₄/ScF₃ composite.

3.2 Rehydrogenation in the 3NaBH₄/ScF₃ composite

Fig. 3e shows the XRD pattern of the complete dehydrogenated 3NaBH₄/ScF₃ composite (530 °C for 3 h) that is maintained under a pressure of 3.2 MPa H₂ at 420 °C for 10 h. The pattern shows no change compared to that of the complete dehydrogenated composite, which means that the latter one has no hydrogen absorption ability.

As the complete dehydrogenated 3NaBH₄/ScF₃ composite shows no reversibility, the hydrogen absorption after the second dehydrogenation step was attempted to study the possible reversibility. The profiles of hydrogen absorption are given in Fig. 5a, which are obtained under the condition of 380 °C, 400 °C and 420 °C at 3.2 MPa H₂ pressure for the sample that has gone through dehydrogenation at 420 °C for 3 h. These curves clearly show the reversible hydrogen absorption of the partial dehydrogenated 3NaBH₄/ScF₃ composite: under the pressure of 3.2 MPa H₂, a hydrogenation capacity of 1.59 wt% can be achieved at 420 °C for 8 h, while it can absorb 1.28 wt% at 400 °C and 1.19 wt% at 380 °C, respectively. By contrast, under 3.5 MPa hydrogen pressure, pure NaBH₄ shows no hydrogen absorption at 400 °C.³⁶

Fig. 5 Isothermal hydrogen absorption curves for the partially dehydrogenated 3NaBH₄/ScF₃ composite sample at temperatures of 380 °C, 400 °C and 420 °C (a) and the ln k-1000/T fitting plot (b).

The hydrogenation activation energy (E_ab) is generally utilized to discriminate kinetics of absorption, by analyzing the entire energy barriers of hydrogen absorption process. Based on the Johanson–Mehl–Avrami (JMA) model, the following equation can be used to evaluate absorption kinetics:⁴⁸

ln[−ln(1 − α_A)] = η ln k + η ln t (7)

where α(t) is a function of time t, k is a parameter describing kinetic, η is the Avrami exponent which matches transformation mechanism. Then, the following Arrhenius equation is used to calculate E_ab:where A represents temperature-independent constant, R represents universal gas constant, and T represents the absolute temperature. The scheme of ln k versus 1000/T, which is shown in Fig. 5, displays a good linear relationship. Therefore, the E_ab value obtained from the slope is therefore estimated to be 44.58 kJ mol⁻¹ H₂ for the partially dehydrogenated 3NaBH₄/ScF₃ composite.

To elucidate the mechanism of hydrogen absorption in the partially dehydrogenated 3NaBH₄/ScF₃ composite, XRD analysis is carried out on the rehydrogenated sample and the result is shown in Fig. 6. The 3NaBH₄/ScF₃ sample was dehydrided at 420 °C for 3 h in vacuum and then rehydrogenated at 420 °C for 10 h under the pressure of 3.2 MPa H₂. In Fig. 6, the diffraction peaks from NaBF₄ and ScB₂ phases became weaker and even disappeared along with the increment in peak intensities of NaBH₄ and ScF₃ as compared to those in Fig. 3c. Therefore, the hydrogen absorption in the partial dehydrogenated 3NaBH₄/ScF₃ composite follows exactly the reverse reaction path of the second step dehydrogenation. That is, the rehydrogenation consumes Na₃ScF₆, NaBF₄ and ScB₂, accompanied with the regeneration of NaBH₄ and ScF₃. Compared to the completely dehydrogenated 3NaBH₄/ScF₃ composite, the partially dehydrogenated composite contains Na₃ScF₆ and NaBF₄ phases, indicating that these two phases play the key role for the rehydrogenation.

Fig. 6 XRD pattern of the partially dehydrogenated 3NaBH₄/ScF₃ composite sample after rehydrogenation.

The results of FTIR analyses for the ball-milled 3NaBH₄/ScF₃ sample, sample dehydrogenated at 420 °C for 3 h and corresponding products rehydrogenated at 400 °C for 3 h can be found in Fig. 7. In Fig. 7a, the FTIR spectrum of sample after ball milling has the signatures of B–H stretching band in the position of 2226 cm⁻¹, 2306 cm⁻¹ and 2366 cm⁻¹, and B–Hbending band peak at 1119 cm⁻¹, all of which are supposed to be originated from borohydride. These peaks are considered to be from NaBH₄.¹⁹ However, it should be noted that, the height of those peaks, which represent the intensity of B–H bonds vibration from the [BH₄]⁻ group, gradually become weaker as dehydrogenation reaction proceeds, indicating the decomposition of NaBH₄, as seen at 1121 cm⁻¹, 2221 cm⁻¹, 2338 cm⁻¹ and 2369 cm⁻¹ in Fig. 7b. According to the work of D. Syamala,⁴⁴ peak located at 1065 cm⁻¹ can be marked as [BF₄]⁻asymmetric stretching, indicating the formation of NaBF₄ after the second step dehydrogenation, which is in good agreement with the XRD results (Fig. 3c). In Fig. 7c, the signatures of [BH₄]⁻ bending at 1120 cm⁻¹ and [BH₄]⁻ stretching at 2223 cm⁻¹, 2304 cm⁻¹ and 2359 cm⁻¹ were clearly revealed for the rehydrogenated sample and the intensity of these peaks increased, indicating the regeneration of NaBH₄.⁴⁹ Meantime, the peak from [BF₄]⁻ asymmetric stretching disappeared after rehydrogenation, showing the consumption of NaBF₄ along with rehydrogenation. Peaks at wave numbers between 1330 cm⁻¹ and 1800 cm⁻¹ (Fig. 7a) were subtracted from the unavoidable moisture absorption and atmospheric humidity absorbed by the sample during measurement, while peak located at around 3745 cm⁻¹ was identified as stretching band vibration of O–H.⁵⁰

Fig. 7 FTIR spectra of the 3NaBH₄/ScF₃ composite after ball-milling (a), half dehydrogenated 3NaBH₄/ScF₃ sample (b), rehydrogenation of partially dehydrogenated 3NaBH₄/ScF₃ sample (c).

3.3 Mechanisms of hydrogenation in 3NaBH₄/ScF₃ composite

It is shown in the present work that the hydrogen storage performance of NaBH₄ can be effectively improved by introducing ScF₃ as a reagent. In particular, rehydrogenation can be achieved in the partially dehydrogenated 3NaBH₄/ScF₃ composite. Analyses revealed that both Sc³⁺ cation and F⁻ anion show irreplaceable importance during the re/dehydrogenation processes of the composite. Firstly, F anion can replace H anion in the initial process of dehydrogenation, from NaBH₄ to form NaBH_xF_4−x. Then, Sc cation loses electron to form ScB₂, in which Sc cation has the calculated valence of +4.08,⁵¹ rather than served as a three-valent cation. This is accompanied with the formation of hydrogen gas. Meanwhile, during the second dehydrogenation step, a portion of F anions from ScF₃ incorporate into Na₃ScF₆ crystallites, which might serve as the nucleation center for the growth of other products.

It has been established that the regeneration of NaBH₄ in the NaBH₄–MF_x systems is associated with electronegativity (χ_p) of the metal cations.⁵² Previous works have shown that after adding transition metal fluorides into NaBH₄ based composites, when the Pauling's electronegativity of the transition metal lies in around between 1.23–1.54, hydrogen sorption reversibility has larger thermodynamic tendency to occur.⁵² The χ_p value of Sc³⁺ is 1.415, which lies in such specific range, thus the regeneration of NaBH₄ in the dehydrided 3NaBH₄/ScF₃ system is favorable.⁵³

Using a database of density functional theory,^54–56 the enthalpies of desorption reactions are calculated to be 41.01 kJ mol⁻¹ H₂ for the second dehydriding step, and 43.31 kJ mol⁻¹ H₂ for the final step. These values are comparably higher than the values obtained from DSC/TG analyses, but significant lower than that of pristine NaBH₄ (108 ± 3 kJ mol⁻¹ of H₂).⁵⁷ The differences between the calculated and measured enthalpies can be explained by fact that H⁻ was partially substituted by F⁻ in NaBH₄, which is also observed in other NaBH₄ based systems containing fluorides.⁵⁸ However, the enthalpy for complete dehydrogenation is still fairly high, about 56.97 kJ mol⁻¹ H₂ calculated from DSC analyses. Consequently, the rehydrogenation of the complete dehydrided 3NaBH₄/ScF₃ composite is difficult from the thermodynamic point of view.⁵²

During the rehydrogenation process, the experimental results show that only partial dehydrogenated products (NaBF₄ + ScB₂ + Na₃ScF₆) have reversibility, while the final dehydriding products (NaF + ScB₂ + B) cannot be rehydrogenated. Apart from the thermodynamic factors, this might also be understood from structural similarity between [BF₄]⁻ in NaBF₄ and [BH₄]⁻ in NaBH₄, which may facilitate the regeneration of NaBH₄ through the exchange between H⁻ and F⁻. The structural similarity was also observed between dehydrogenated and rehydrogenated products in the 3NaBH₄/LnF₃ systems, which led to the improved rehydrogenation kinetics in these systems.²⁵ Researchers have also found that substitution reaction could be well understood from the hydride–fluoride isostructure, which has been proposed and confirmed in various hydrides–fluorides compounds having different stoichiometries.^59–61

4 Conclusions

In this study, the 3NaBH₄/ScF₃ composite was prepared through mechanical milling. The behaviors and mechanisms of hydrogen de/absorption of the composite were explained by using TPD, DSC/TG, XRD and FTIR techniques. Following are the summarized results:

(1) TPD and DSC analyses confirmed that NaBH_xF_4−x compound formed at the early dehydriding stage due to the partial substitution of H⁻ anion by F⁻ anion in NaBH₄, releasing about 0.19 wt% of hydrogen. When temperature further increases, Na₃ScF₆, NaBF₄ and ScB₂ formed through the reaction between NaBH₄ and ScF₃ with 2.52 wt% of hydrogen released. Finally, the reaction of residual NaBH₄ with Na₃ScF₆ produces NaF, B and ScB₂, releasing about 2.83 wt% of hydrogen.

(2) The partially dehydrogenated products, Na₃ScF₆, ScB₂ and NaBF₄, can be rehydrogenated to generate NaBH₄ with an activation energy of 44.58 kJ mol⁻¹ H₂. In contrast, the fully dehydrogenated products, NaF + ScB₂ + B, cannot be hydrogenated. The hydrogen sorption reversibility of the partially dehydrogenated composite can be understood through thermodynamic point of view and the structural similarity between [BF₄]⁻ in NaBF₄ and [BH₄]⁻ in NaBH₄.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by National Nature Science Foundation (No. 51771112), the Shanghai Science and Technology Commission under No. 14JC1491600 and Shanghai Education Commission “Shuguang” scholar project (16SG08).

Notes and references

1. W. Li, X. F. Gao, X. G. Wang, D. H. Xiong, P. P. Huang, W. G. Song, X. Q. Bao and L. F. Liu, J. Power Sources, 2016, 330, 156.
2. T. da Silva Veras, T. S. Mozer and A. da Silva César, Int. J. Hydrogen Energy, 2017, 42, 2018.
3. Q. W. Lai, M. Paskevicius, D. A. Sheppard, C. E. Buckley, A. W. Thornton, M. R. Hill, Q. F. Gu, J. F. Mao, Z. G. Huang, H. K. Liu, Z. P. Guo, A. Banerjee, S. Chakraborty, R. Ahuja and K.-F. Aguey-Zinsou, ChemSusChem, 2015, 8, 2789.
4. L. Schlapbach and A. Züttel, Nature, 2001, 414, 353.
5. J. F. Mao and D. H. Gregory, Energies, 2015, 8, 430.
6. W. Li, X. F. Gao, D. H. Xiong, F. Xia, J. Liu, W. G. Song, J. Y. Xu, S. M. Thalluri, M. F. Cerqueira, X. L. Fu and L. F. Liu, Chem. Sci., 2017, 8, 2952.
7. F. Z. Song, W. Li, G. Q. Han and Y. J. Sun, ACS Appl. Energy Mater., 2018, 1, 3.
8. W. Li, D. H. Xiong, X. F. Gao, W. G. Song, F. Xia and L. F. Liu, Catal. Today, 2017, 287, 122.
9. W. Li, X. G. Wang, D. H. Xiong and L. F. Liu, Int. J. Hydrogen Energy, 2016, 41, 9344.
10. P. P. Silva, R. A. Ferreira, F. B. Noronha and C. E. Hori, Catal. Today, 2017, 289, 211.
11. W. Li, X. Gao, D. Xiong, F. Wei, W. G. Song, J. Xu and L. Liu, Adv. Energy Mater., 2017, 7, 1602579.
12. C. Y. Cao, C. Q. Chen, W. Li, W. G. Song and W. Cai, ChemSusChem, 2010, 3, 1241.
13. J. F. Mao, Q. F. Gu, Z. P. Guo and H. K. Liu, J. Mater. Chem. A, 2015, 3, 11269.
14. S. Orimo, Y. Nakamori, J. R. Eliseo, A. Züttel and C. M. Jensen, Chem. Rev., 2007, 107, 4111.
15. P. Chen and M. Zhu, Mater. Today, 2008, 11, 36.
16. D. M. F. Santos and C. A. C. Sequeira, Renewable Sustainable Energy Rev., 2011, 15, 3980.
17. Y. F. Liu, Y. X. Yang, M. X. Gao and H. G. Pan, Chem. Rec., 2016, 16, 189.
18. U. B. Demirci, O. Akdim, J. Andrieux, J. Hannauer, R. Chamoun and P. Miele, Fuel Cells, 2010, 10, 335.
19. L. N. Chong, J. X. Zou, X. Q. Zeng and W. J. Ding, J. Mater. Chem. A, 2013, 1, 3983.
20. M. Yadav and Q. Xu, Energy Environ. Sci., 2012, 5, 9698.
21. G. Moussa, R. Moury, U. B. Demirci, T. Sener and P. Miele, Int. J. Energy Res., 2013, 37, 825.
22. J. Urgnani, F. J. Torres, M. Palumbo and M. Baricco, Int. J. Hydrogen Energy, 2008, 33, 3111.
23. T. J. Frankcombe, Chem. Rev., 2011, 112, 2164.
24. P. E. de Jongh and P. Adelhelm, ChemSusChem, 2010, 3, 1332.
25. J. J. Vajo, T. T. Salguero, A. F. Gross, S. L. Skeith and G. L. Olson, J. Alloys Compd., 2007, 446, 409.
26. A. T. Dinsdale, Calphad, 1991, 15, 317.
27. T. Czujiko, R. A. Varin, Z. Zaranski and Z. S. Wronski, Arch. Metall. Mater., 2010, 55, 539.
28. T. D. Humphries, G. N. Kalantzopoulos, I. Llamas-Jansa, J. E. Olsen and B. C. Hauback, J. Phys. Chem. C, 2013, 117, 6060.
29. J. F. Mao, Z. P. Guo, I. P. Nevirkovets, H. K. Liu and S. X. Dou, J. Phys. Chem. C, 2012, 116, 1596.
30. S. A. Jin, J. H. Shim, Y. W. Cho, K. W. Yi, O. Zabara and M. Fichtner, Scr. Mater., 2008, 58, 963.
31. J. X. Zou, L. J. Li and X. Q. Zeng, Int. J. Hydrogen Energy, 2012, 37, 17118.
32. R. Cerny, G. Severa, D. B. Ravnsbæk, Y. Filinchuk, V. D'Anna, H. Hagemann, D. Haase, C. M. Jensen and T. R. Jensen, J. Phys. Chem. C, 2010, 114, 1357.
33. D. B. Ravnsbæk, Y. Filinchuk, R. Cerny, M. B. Ley, D. Haase, H. J. Jakobsen, J. Skibsted and T. R. Jensen, Inorg. Chem., 2010, 49, 3801.
34. L. N. Chong, J. X. Zou, X. Q. Zeng and W. J. Ding, J. Mater. Chem. A, 2015, 3, 4493.
35. L. N. Chong, J. X. Zou, X. Q. Zeng and W. J. Ding, J. Mater. Chem. A, 2013, 1, 13510.
36. L. N. Chong, J. X. Zou, X. Q. Zeng and W. J. Ding, Int. J. Hydrogen Energy, 2014, 39, 14275.
37. H. E. Kissinger, Anal. Chem., 1957, 29, 1702.
38. C. Bonatto Minella, S. Garroni, C. Pistidda, R. Gosalawit-Utke, G. Barkhordarian, C. Rongeat, I. Lindemann, O. Gutfleisch, T. R. Jensen, Y. Cerenius, J. Christensen, M. D. Baró, R. Bormann, T. Klassen and M. Dornheim, J. Phys. Chem. C, 2011, 115, 2497.
39. L. C. Yin, P. Wang, X. D. Kang, C. H. Sun and H. M. Cheng, Phys. Chem. Chem. Phys., 2007, 9, 1499.
40. R. Gosalawit-Utke, J. M. Bellosta von Colbe, M. Dornheim, T. R. Jensen, Y. Cerenius, C. Bonatto Minella, M. Peschke and R. Bormann, J. Phys. Chem. C, 2010, 114, 10291.
41. H. W. Brinks, A. Fossdal and B. C. Hauback, J. Phys. Chem. C, 2008, 112, 5658.
42. Y. F. Liu, F. H. Wang, Y. H. Cao, M. X. Gao, H. G. Pan and Q. D. Wang, Energy Environ. Sci., 2010, 3, 645.
43. Z. Z. Fang, X. D. Kang, Z. X. Yang, G. S. Walker and P. Wang, J. Phys. Chem. C, 2011, 115, 11839.
44. D. Syamala, V. Rajendran, R. K. Natarajan and S. Moorthy Badu, Cryst. Growth Des, 2007, 7, 1695.
45. S. Garroni, C. Milanese, D. Pottmaier, G. Mulas, P. Nolis, A. Girella, R. Caputo, D. Olid, F. Teixdor, M. Baricco, A. Marini, S. Suriñach and M. D. Baró, J. Phys. Chem. C, 2011, 115, 16664.
46. J. H. Her, W. Zhou, V. Stavila, C. M. Brown and T. J. Udovic, J. Phys. Chem. C, 2009, 113, 11187.
47. R. Caputo, S. Garroni, D. Olid, F. Teixidor, S. Surinach and M. D. Baro, Phys. Chem. Chem. Phys., 2010, 12, 15093.
48. P. S. Rudman, J. Less-Common Met., 1983, 89, 93.
49. T. Sato, K. Miwa, Y. Nakamori, K. Ohoyama, H. W. Li, T. Noritake, M. Aoki, S. I. Towata and S. I. Orimo, Phys. Rev. B, 2008, 77, 104114.
50. A. M. Olsson and L. Salmén, Carbohydr. Res., 2004, 339, 813.
51. R. S. Roth and S. Schneider, National Bureau of Standards, 1972, p. 364.
52. Y. Nakamori, K. Miwa, A. Ninomiya, H. W. Li, N. Ohba, S. I. Towata, A. Züttel and S. I. Orimo, Phys. Rev. B, 2006, 74, 045126.
53. K. Y. Li and D. F. Xue, J. Phys. Chem. A, 2006, 110, 11332.
54. A. Jain, G. Hautier, S. P. Ong, C. J. Moore, C. C. Fischer, K. A. Persson and G. Ceder, Phys. Rev. B: Condens. Matter Mater. Phys., 2011, 84, 045115.
55. A. Jain, S. P. Ong, G. Hautier, W. Chen, W. D. Richards, S. Dacek, S. Cholia, D. Gunter, D. Skinner, G. Ceder and K. A. Persson, APL Mater., 2013, 1, 011002.
56. L. C. Yin, P. Wang, Z. Z. Fang and H. M. Cheng, Chem. Phys. Lett., 2008, 450, 318.
57. P. Martelli, R. Caputo, A. Remhof, P. Mauron, A. Borgschulte and A. Züttel, J. Phys. Chem. C, 2010, 114, 7173.
58. L. N. Chong, J. X. Zou, X. Q. Zeng and W. J. Ding, J. Mater. Chem. A, 2014, 2, 8557.
59. A. Bouamrane, J. P. Laval, J.-P. Soulie and J. P. Bastide, Mater. Res. Bull., 2000, 35, 545.
60. J. P. Soulié, J. P. Laval and A. Bouamrane, Solid State Sci., 2003, 5, 273.
61. Q. Zhou and B. J. Kennedy, J. Solid State Chem., 2004, 177, 654.

---