Direct Synthesis of NaBH₄ Nanoparticles from NaOCH₃ for Hydrogen Storage

Abstract

Hydrogen is regarded as a promising energy carrier to substitute fossil fuels. However, storing hydrogen with high density remains a challenge. NaBH₄ is a potential hydrogen storage material due to its high gravimetric hydrogen density (10.8 mass%), but the hydrogen kinetic and thermodynamic properties of NaBH₄ are poor against the application needs. Nanosizing is an effective strategy to improve the hydrogen properties of NaBH₄. In this context, we report on the direct synthesis of NaBH₄ nanoparticles (~6–260 nm) from the NaOCH₃ precursor. The hydrogen desorption properties of such nanoparticles are reported as well as experimental conditions that lead to the synthesis of (Na₂B₁₂H₁₂) free NaBH₄ nanoparticles.

1. Introduction

Sodium borohydride (NaBH₄) has attracted significant interest as a potential hydrogen storage material, owing to its high hydrogen gravimetric capacity (10.8 mass%) [1,2]. However, the practical application of NaBH₄ is difficult to achieve, because of its stability both in terms of hydrogen kinetics and thermodynamics. Typically, the dehydrogenation temperature of NaBH₄ is above 500 °C and the kinetics of both the hydrogen release and uptake are too slow for practical use. Moreover, the inherent phase segregation of NaBH₄ and partial evaporation of Na upon hydrogen release lead to poor hydrogen reversibility [3].

In order to enhance the hydrogen storage properties of NaBH₄, the approach of nanosizing, i.e., reduction of particle size below 100 nm, has proven to lead to some improvements in the thermodynamics and kinetics. This is believed to result from several factors including an increase in the specific surface area and shorter hydrogen diffusion paths [4,5]. The typical method to downsize borohydrides is through their nanoconfinement in porous scaffolds [6,7], but the effective storage capacity achievable by such an approach remains limited, owing to the difficulty of filling the nanopores and the “dead volume and mass” of the host scaffold [4,8]. An alternative method is through the synthesis of isolated nanoparticles and their stabilization as core-shell nanostructures [3,9]. However, to date, this approach is significantly limited by the lack of synthetic methods for readily making NaBH₄ nanoparticles of controlled properties.

Thus far, the synthesis methods for preparing isolated nanosized NaBH₄ particles are rather limited [3,5]. A common method to synthesize nanoparticles is by growth in solution, but for NaBH₄, this is not without challenges. NaBH₄ is a strong reducing agent, and thus, many of the common reagents used for the synthesis and stabilization of conventional nanoparticles are out of scope. In addition, the solvents that can be used to synthesize NaBH₄ nanoparticles are rather limited owing to the lack of solubility and/or reactivity of NaBH₄ [2,3]. Oftentimes, solubilizing solvents form adduct compounds with NaBH₄, and this brings additional complexity in terms of synthesizing solvent free nanoparticles [10]. Via a wet chemistry approach, Christian et al. reported that NaBH₄ nanoparticles could be precipitated in solution with a size ranging from 2 to 60 nm via anti-precipitation with tetra-n-butylammonium bromide (TBAB) as a surfactant [1,5]. Solvent evaporation has also been reported as an effective method for the synthesis of borohydride nanoparticles. For example, Lai et al. have synthesized LiBH₄ and NaBH₄ nanoparticles with Ti doped by solvent evaporation [11] and we have also successfully synthesized LiBH₄ nanoparticles of controllable sizes with various surfactants [5]. However, this approach leaves excessive amounts of surfactant around the nanoparticles. This significantly limits the effective hydrogen storage capacity and contaminates the hydrogen release with fragments of decomposing surfactants at elevated temperatures [5]. Removing such surfactants could be an option, but these are needed to minimize the surface energy of nanoparticles during their nucleation and growth.

A major drawback of current synthetic approaches is that they all rely on a source of NaBH₄ as the starting material to lead to the formation of nanoparticles, and this is an expensive approach given the current cost of NaBH₄ (95 USD$ kg⁻¹) [2]. In this respect, the development of direct approaches for making NaBH₄ nanoparticles from common precursors is more desirable. However, this is not without difficulties. As summarized in Table S1, the known synthetic methods of NaBH₄ have drawbacks, such as: a) low yields, b) high temperatures and/or hydrogen pressures required, and c) low purity. Furthermore, direct production of nanoparticles from these synthetic methods is not evident, and to date, only Lai et al. reported on a thermal decomposition of sodium trimethyl borate (NaB(OCH₃)₄) to make NaBH₄ nanoparticles. In this case, the approach first involves the formation of NaB(OCH₃)₄ through the reaction of NaH with trimethyl borate (B(OCH₃)₃) via reaction (1). This is then followed by the thermal decomposition of NaB(OCH₃)₄ in tetrahydrofuran to precipitate NaBH₄ (reaction (2)) [3]:

NaH + B(OCH₃)₃ → NaBH(OCH₃)₃

(1)

$$4\mathrm{NaBH}({\mathrm{OCH}}_3{)}_3\stackrel{65-70°\mathrm{C}}{\to }{\mathrm{NaBH}}_4+3\mathrm{NaB}({\mathrm{OCH}}_3{)}_4$$

(2)

However, the method was found to be difficult to control, the yield of NaBH₄ was low, and the by-products of the reaction was difficult to remove [9].

In this work, we report on the development of two routes for making isolated NaBH₄ nanoparticles through the reaction of NaOCH₃ with B₂H₆ gas according to reaction (3):

3NaOCH₃ + 2B₂H₆ → 3NaBH₄ + B(OCH₃)₃

(3)

Considering that B(OCH₃)₃ has a low boiling point (68 °C) [12,13] and it is soluble in diethyl ether, in which NaBH₄ is not soluble [14,15,16], B(OCH₃)₃ as a by-product can be removed by heating up or by diethyl ether washing. Thus, we investigated both routes for making NaBH₄ nanoparticles, and in first instance, reaction (3) was carried out at 100 °C to facilitate the evaporation of B(OCH₃)₃, while B₂H₆ was reacted with NaBH₄. To facilitate the dispersion of the reaction by-product, the feasibility of reaction (3) in solvents was also investigated. The morphology, purity and hydrogen properties of the NaBH₄ nanoparticles obtained through these routes are reported. In particular, the main hydrogen release temperature of the NaBH₄ nanoparticles synthesized by both methods has been reduced by more than 100 °C in comparison to commercial NaBH₄.

2. Materials and Methods

Except the solid-gas synthesis of NaBH₄ nanoparticles, all the other experiments were performed under an inert atmosphere in an argon filled LC-Technology glove box (O₂ and H₂O < 1 ppm).

2.1. Materials

NaOCH₃ (95%), NaBH₄ (98%), ZnCl₂ (≥98%), tetra-n-butylammonium bromide (TBAB, >98%), methanol (anhydrous, 99.8%), and octadecylamine (ODA, ≥99%) were purchased from Sigma-Aldrich and hexane (99%, HPLC grade) was purchased from Honeywell. NaOCH₃ was purified by solubilization in a large amount of anhydrous methanol and recrystallization of the centrifuged solution was done by evaporating the methanol. ZnCl₂ was further dried at 120 °C overnight on a Schlenk line under vacuum (0.01 MPa). Hexane was dried with activated molecular sieve (4 Å) prior to use. The other chemicals were used as received.

2.2. Synthesis of the Diborane (B₂H₆) Precursor

NaZn₂(BH₄)₅ as the precursor of B₂H₆ was synthesized by ball milling NaBH₄ and ZnCl₂ (2:1 molar ratio) in a stainless-steel vial (25 mL) containing a single stainless-steel ball (15 mm diameter). A Retsch MM301 mill operated at a frequency of 20Hz was used for this. The milling process was done twice for 10 min.

2.3. Synthesis of the NaBH₄ Nanoparticles via Solid-Gas Reaction

The reaction was undertaken in an in-house built Sievert apparatus with a customized sample holder which had two compartments separated by a stainless-steel mesh (20 μm porosity). 100 mg of purified NaOCH₃ was placed on the top of the mesh and variable amounts (0.2, 0.3, 0.4, 1.0 g) of NaZn₂(BH₄)₅ were placed at the bottom of the sample holder. The reaction was carried out at 100 °C overnight. The materials synthesized by this solid-gas reaction (~72 mg) are denoted NaBH₄-SGX, where X is the molar ratio of NaOCH₃ and the B₂H₆ generated, X = 1.2, 1.8, 2.4, 5.9.

2.4. Synthesis of the NaBH₄ Nanoparticles via Suspension in Hexane and Reaction with B₂H₆

NaOCH₃ nanoparticles were first synthesized by a solvent evaporation method. To this aim, 0.2 M of purified NaOCH₃ and 0.01 M TBAB acting as surfactant were dissolved in 10 mL methanol at room temperature. The solution was then placed in a Biotage V-10 evaporator and the methanol was evaporated at a controlled rate to trigger the nucleation and growth of NaOCH₃ nanoparticles. The evaporation was carried out at 50 °C, at a centrifugal speed of 3000 rpm. The resulting NaOCH₃ powered was then further dried overnight on a Schlenck line at 40 °C and under a vacuum level of 0.01 MPa to remove any methanol residue.

70 mg of as-synthesized NaOCH₃ nanoparticles were then suspended in 10 mL hexane with 0.01 M octadecylamine acting as surfactant in a round flask. Variable amount of NaZn₂(BH₄)₅ (0.5, 1, and 2 g) were placed in a separate round flask on a hot plate at 100 °C to generate B₂H₆. It is important to mention that the amount of B₂H₆ is much larger than that used in the solid-gas reaction. This is to compensate for the weak airtightness of the rubber seal on the glassware. Both round-bottomed flasks were sealed by a rubber septum and connected by a cannula to allow for the B₂H₆ generated from the decomposition of NaZn₂(BH₄)₅ to react with the NaOCH₃ nanoparticles in suspension. After reaction overnight, the resulting material was separated from the hexane by centrifugation, washed three times with diethyl ether to remove the by-product B(OCH₃)₃, and dried on a Schlenck line at room temperature under vacuum at 0.01 MPa. The materials synthesized for this suspension gas reaction (~42 mg) are denoted NaBH₄-SUGX, where X is the molar ratio of NaOCH₃ and the B₂H₆ generated, X = 4.2, 8.5, 16.9.

2.5. Characterization

The morphologies of the materials were determined by Transmission Electron Microscopy (TEM) on a Philips CM200 operated at 200 kV. The materials were dispersed in cyclohexane followed by ultrasonication and then dropped onto a carbon coated copper grid. The copper grid was then transferred to the TEM in a quick manner as to minimize air exposure. The mean particle size and distribution were collected from a minimum of 300 particles from a range of TEM images.

Crystalline phases were determined by X-Ray Diffraction (XRD) on a Philips X’pert Multipurpose XRD system operated at 40 mA and 45 kV with a monochromated Cu Kα radiation (λ = 1.541 Å) over a 2θ range from 10 to 70°. The materials were protected against oxidation from air by a Kapton foil.

Fourier Transform Infrared Spectroscopy (FTIR) was conducted on a Bruker Vertex 70 V. The materials were mixed with KBr in the glove box and loaded in an air-tight chamber fitted on a Harrick-Scientific Praying Mantis Diffuse Reflectance Infrared Fourier Transform Spectroscopy (DRIFT) accessory. The spectra were collected at room temperature from 600 to 3400 cm⁻¹ over 124 scans with a resolution of 1 cm⁻¹.

The hydrogen desorption properties of the materials were determined by Mass Spectrometry (MS) with an Omnistar instrument from Pfeiffer. The thermal stability of the surfactants was investigated by simultaneous Thermogravimetric Analysis/Differential Scanning Calorimetry (TGA/DSC) with a TGA/DSC1 from Mettler Toledo coupled to the MS. The characterizations were conducted under an argon flow at 20 mL min⁻¹ from 20 to 700 °C with a heating rate of 10 °C min⁻¹ and alumina crucibles (70 mg) were used.

The carbon content of the synthesized materials (NaBH₄-SG1.8 and NaBH₄-SUG8.5) that may result from remaining traces of NaOCH₃ or B(OCH₃)₃ was determined with a total carbon analyzer (Multi N/C2100, Analytic Jena, Germany). The materials were dissolved in 10 mL of Milli-Q water and the resulting solution was filtered before total carbon analysis. Based on this analysis, both NaBH₄-SG1.8 and NaBH₄-SUG8.5 (which correspond to the optimum reaction conditions) were found to contain ~2.0 % of carbon per weight of material.

3. Results and Discussion

3.1. NaBH₄ Nanoparticles Synthesized via Solid-Gas Interaction

3.1.1. Morphology of the As-Synthesized Materials

The morphology of the as-synthesized material was characterized by TEM (Figure 1). By exposing the NaOCH₃ particles (Figure S1) to increasing amounts of B₂H₆, (from molar ratio 1.8 to 5.9), the morphology of the material synthesized evolved from large undefined particles of ~262 nm to spherical like particles of 6 nm. This reduction in size is probably due to a fragmentation of the NaOCH₃ particles with B₂H₆ (reaction 3), and the subsequent release of gaseous B(OCH₃)₃ at 100 °C. This phenomenon has also been found to occur in reactions involving some gas release [17,18].

3.1.2. Composition of the As-Synthesized Materials

XRD and FTIR characterizations were performed to confirm the compositions of the as-synthesized material. From XRD analysis (Figure 2 and Figure S2), it is apparent that when the molar ratio of NaOCH₃ and B₂H₆ is above 1.8, all the NaOCH₃ is reacted because only diffraction peaks related to thermodynamically stable phase of NaBH₄ were observed. FTIR analysis further confirmed this trend (Figure 3 and Figure S3). As the molar ratio evolved from 1.2 to 1.8, the vibrational peaks corresponding to NaOCH₃ (Figure S3) disappeared, and this is in agreement with the XRD results (Figure 2). The peak at ~43 ° is assigned to the stainless steel sample holder.

A typical FTIR spectrum of NaBH₄ would display vibrational peaks associated with the BH₄⁻ group in the region of 2410–2200 cm⁻¹ corresponding to the stretching mode of BH₄⁻, whereas the bending modes of BH₄⁻ are to be found around 1120 cm⁻¹ [19,20].

For NaBH₄-SG1.2, the additional peaks observed at 2582, 1194, 1077, and 994 cm⁻¹ were assigned to the C-O stretching modes associated with NaOCH₃ [21,22,23,24,25,26,27,28,29]. Such vibrations were not observed at higher molar ratio of NaOCH₃ and B₂H₆ and this confirmed that above a ratio of 1.2, NaOCH₃ was fully reacted. However, when the molar ratio reached 2.4 and 5.9, a new vibration appeared at 2490 cm⁻¹ (Figure 3). This vibration was assigned to the stretching mode of B₁₂H₁₂²⁻ [11,30]. Accordingly, traces of Na₂B₁₂H₁₂ have been generated via the reaction of the newly formed NaBH₄ with the excess of B₂H₆. Thus far, there have been no reports regarding the direct formation of Na₂B₁₂H₁₂ via the reaction between NaBH₄ and B₂H₆. Only the formation of NaB₃H₈ has been reported to occur according to (4) [31,32]:

NaBH₄ + B₂H₆ → NaB₃H₈ + H₂

(4)

Chen et al. reported that NaB₃H₈ can decompose to NaBH₄, Na₂B₁₂H₁₂, Na₂B₁₀H₁₀ at 150 °C [33]. Given that our synthesis was performed at 100 °C, one possibility is that the Na₂B₁₂H₁₂ phase observed by FTIR is the result of the decomposition of NaB₃H₈.

Finally, the broad peaks between 1420–1260 cm⁻¹ that were observed in all the spectra of the as-synthesized NaBH₄ nanoparticles were assigned to the stretching modes of B-O bond [26]. These vibrations are most likely the result of remaining traces of B(OCH₃)₃ in the as-synthesized NaBH₄ nanoparticles. This was also confirmed by the presence of C-H stretching modes from B(OCH₃)₃ in all the synthesized materials (Figure 3).

3.2. NaBH₄ Nanoparticles Synthesized from NaOCH₃ Nanoparticles Reacted in Suspension

Although NaBH₄ nanoparticles have been successfully obtained by the solid-gas method, the approach did not lead to nanoparticles of well-defined morphology. In order to better control the particle size and synthesize isolated NaBH₄ nanoparticles, we investigated the possibility to first synthesize NaOCH₃ nanoparticles by solvent evaporation method, and then suspended these NaOCH₃ nanoparticles in hexane with octadecylamine as a stabilizer to retain the morphology during reaction with B₂H₆. After reaction at room temperature, the as-synthesized NaBH₄ nanoparticles were collected by centrifugation and washed by diethyl ether to remove the solubilized B(OCH₃)₃ residue [14,15,16]. Figure 4a,b shows the typical TEM images of the as-synthesized NaOCH₃ nanoparticles obtained by solvent evaporation method, and thus, such an approach was found to be effective in leading to nanoparticles of NaOCH₃ with a mean particle size of 16 nm.

3.2.1. Morphology of the As-Synthesized NaBH₄

After the reaction of the suspended NaOCH₃ nanoparticles with B₂H₆, analysis by TEM of the resulting material revealed polydispersed particles with a mean size of 14 nm (NaBH₄-SUG4.2, Figure 4c,d) and 12 nm (NaBH₄-SUG8.5, Figure 4e,f, and NaBH₄-SUG16.9, Figure 4g,h). This is slightly smaller than the particle size of NaOCH₃ (Figure 4a,b), and this reduction in particle size could be the result of elemental loss in the form of B(OCH₃)₃ that is released during the synthesis (reaction 3).

3.2.2. Composition of the As-Synthesized Materials

Figure 5 shows the XRD patterns of the as-synthesized NaBH₄ obtained from the suspension-gas reaction. Different to the cubic phase of NaBH₄ synthesized by solid-gas reaction, the as-synthesized NaBH₄ exhibited peaks of both tetragonal and cubic NaBH₄. The peaks of NaOCH₃ at ~11.5°, 30.7°, and 31.9° in NaBH₄-SUG4.2 disappeared when the molar ratios of NaOCH₃ and B₂H₆ were 8.5 and 16.9. This indicated that all the NaOCH₃ was fully reacted with B₂H₆ to produce NaBH₄.

The FTIR spectra of the NaBH₄ synthesized by suspension-gas reaction also showed the typical stretching modes and bending modes of the BH₄⁻ groups in the range of 2400–2220 cm⁻¹ and at ~1120 cm⁻¹, respectively (Figure 6). Additional vibrational peaks assigned to C-H were observed between 2970–2680 cm⁻¹ (stretching modes), ~1470 cm⁻¹ (bending modes) and 780–710 cm⁻¹ (stretching modes) [22].

Compared with the FTIR spectrum of pristine octadecylamine (Figure S4), the stretching modes of the primary NH₂ shifted from 3331, 3254 cm⁻¹ to ~3280, 3135 cm⁻¹ upon stabilization of the NaBH₄ nanoparticles. In addition, the multiple peaks of NH₂ rocking and wagging modes at 940–750 cm⁻¹ in pristine octadecylamine [34,35,36] appeared as a weak peak at ~880 cm⁻¹. This indicates that the interaction of NaBH₄ and octadecylamine was through its head group (NH₂).

FTIR analysis also proved the presence of the unreacted NaOCH₃ in NaBH₄-SUG4.2. For example, the peaks at around 2582, 1194, 1082, and 994 cm⁻¹ in NaBH₄-SUG4.2 were assigned to the stretching modes of C-O in NaOCH₃ (Figure 6) [21,22,23,24,29]. These peaks were not observed in NaBH₄-SUG8.5 and NaBH₄-SUG16.9 by FTIR, and this proves that at NaOCH₃ and B₂H₆ molar ratios higher than 8.5, NaOCH₃ was fully reacted. However, a significant difference of the suspension-gas reaction as compared to the solid-gas reaction carried at 100 °C is the absence of any vibrations related to the formation NaB₁₂H₁₂ in NaBH₄-SUG16.9, despite the excess of B₂H₆. This indicates that NaB₁₂H₁₂ cannot be formed at room temperature. This may be due to the high energy required for its formation, e.g., Δ_fH₀ = −1086.196 kJ mol⁻¹ for its cubic phase, Δ_fH₀ = −1086.381 kJ mol⁻¹ for its monoclinic phase [37].

3.3. Hydrogen Desorption Properties

The hydrogen desorption properties of the as-synthesized NaBH₄ were characterized by TGA/DSC and MS. Figure 7 and Figure 8 show the TGA, DSC, and MS of NaBH₄-SG1.8 and NaBH₄-SUG8.5, which correspond to the optimum NaOCH₃ and B₂H₆ molar ratios for both reaction paths according to the XRD and FTIR analyses.

The TGA analysis (Figure 7a) of NaBH₄-SG1.8 first showed a mass loss of 3.5% between 280–440 °C. This is correlated to hydrogen released as observed by simultaneous MS analysis (Figure 7b). Within this temperature range, no obvious peak was observed by DSC. There is also no obvious evidence of other gases released from NaBH₄-SG1.8.

For NaBH₄-SUG8.5, the mass loss of 7.1 % from 175–240 °C resulted from the decomposition of octadecylamine (Figure 8 and Figure S5) [5,38,39], as confirmed by the endothermic peaks observed at 215 °C (Figure 8a) and octadecylamine fragments observed by MS (Figure 8b). It should be noted that in this case, the mass loss observed by TGA is much larger than the carbon amount determined by the total carbon analysis, because octadecylamine was removed by filtration before the carbon analysis [40]. At low temperatures, i.e., 170 °C, the small peak of hydrogen release observed by MS was assigned to the recombination of the protic hydrogen (NH^δ+) in the octadecylamine with the hydridic hydrogen (BH^δ-) from NaBH₄. At higher temperatures, i.e., between 240 and 505 °C, the additional mass loss of 2.1% was assigned to hydrogen release as indicated by MS (Figure 8b).

The endothermic peaks observed for NaBH₄-SG1.8 (Figure 7a) at 484 °C and NaBH₄-SUG8.5 at 472 °C (Figure 8a) were assigned to the melting of the as-synthesized NaBH₄ nanoparticles. Compare to the melting point of commercial NaBH₄ (506 °C, Figure S6a), this indicates that the reduction in particle size influences the melting behavior of NaBH₄ particles, and this result is in agreement with previous observations made for almost all free nanoparticles [41,42,43]. Furthermore, from the TGA/DSC/MS analysis, it can be observed that the temperatures at which the main hydrogen release start have been significantly reduced to 452 °C for NaBH₄-SG1.8 and 384 °C for NaBH₄-SUG8.5 in comparison to commercial NaBH₄ at 586 °C (Figure S6). This is believed to be attributed to the reduction in particle size of the NaBH₄ nanoparticles synthesized.

At temperatures >500 °C, the large mass losses of 17.7% for NaBH₄-SG1.8 (Figure 7a) and 24.6% for NaBH₄-SUG8.5 (Figure 8a) are believed to be the result of additional hydrogen release in conjunction with Na evaporation owing to the low melting point of Na (97.5 °C) and a high vapor pressure of 10⁻³ MPa at 500 °C [44,45].

4. Conclusions

We have synthesized NaBH₄ nanoparticles with different sizes ranging from 6 to 260 nm through the direct reaction of NaOCH₃ with B₂H₆. The particle size decreased as the amount of B₂H₆ increased during the solid-gas reaction. However, the formation of Na₂B₁₂H₁₂ was found to occur when NaOCH₃ was exposed to excessive amounts of B₂H₆ during the reaction at 100 °C. In contrast, reacting NaOCH₃ nanoparticles in suspension with B₂H₆ at room temperature did not lead to the formation of Na₂B₁₂H₁₂, and the purity of the resulting NaBH₄ was ~98.0%. The start of hydrogen release from our NaBH₄ nanoparticles is much lower than that of commercial NaBH₄ (by at least 100 °C) and this was attributed to the nanosize nature of the materials synthesized. These results not only demonstrate the possibility to synthesize NaBH₄ nanoparticle via a direct synthesis route with monodispersed particles, in particular through the reaction of NaOCH₃ suspended in solution, but also the possibility to reduce the temperature at which hydrogen is released. Additional modification through a core-shell approach, for example, to confine NaBH₄ and its elements upon hydrogen release, could lead to better hydrogen properties in particular in terms of hydrogen reversibility by avoiding elemental segregation. This will be at the core of future work.