Using Additives to Control the Decomposition Temperature of Sodium Borohydride

Camille Jubert Tomasso ^{1, 2}, Anne L. Pham ¹, Tracy M. Mattox ^{1, *}, Jeffrey J. Urban ^{1, *}

1. Molecular Foundry, Lawrence Berkeley National Laboratory, One Cyclotron Rd, Berkeley CA 94720, USA
2. Université de Paris, 85 Bd St Germain - 75006, Paris, France

* Correspondence: Tracy M. Mattox, Jeffrey J. Urban

Academic Editor: Alfonso Chinnici

Special Issue: Hydrogen Energy: Sustainable Production, Storage and Utilisation

Received: April 20, 2020 | Accepted: May 13, 2020 | Published: May 19, 2020

Journal of Energy and Power Technology 2020, Volume 2, Issue 2, doi:10.21926/jept.2002009

Recommended citation: Jubert Tomasso C, Pham AL, Mattox TM, Urban JJ. Using Additives to Control the Decomposition Temperature of Sodium Borohydride. Journal of Energy and Power Technology 2020; 2(2): 009; doi:10.21926/jept.2002009.

© 2020 by the authors. This is an open access article distributed under the conditions of the Creative Commons by Attribution License, which permits unrestricted use, distribution, and reproduction in any medium or format, provided the original work is correctly cited.

1. Introduction

There is an ever increasing need to develop energy efficient materials that are environmentally friendly. Hydrogen (H₂) burned with oxygen is incredibly clean, and is becoming a popular zero-carbon emission fuel. Unfortunately, the storage of pure H₂ can be problematic. Storing in its gaseous state is very common, but there are serious concerns about flammability and explosive hazards. Storing H₂ in the liquid state requires ultralow, cryogenic temperatures, making storage expensive while adding the hazards, challenges and expense of cryogen handling [1,2]. There is a strong need to create a regenerative hydrogen storage material that is safe and lightweight enough for easy transportation. Thus, researchers have been exploring H₂ storage methods in the solid state, trying to meet the constraints given by the U.S. Department of Energy (DOE) [3], such as the recommendation to find materials containing > 7 % H₂ (gravimetric weight) with a maximum H₂ release rate under 125 °C.

There have been various strategies employed to develop these solid state materials. Some have focused on H₂ adsorbents, gaining an understanding of the gravimetric and volumetric capacities and binding strengths in materials such as metal organic and covalent organic frameworks [4]. Interestingly, there is often disagreement in this field on the theoretical analysis, such as calculating the H₂ capacity. Other researchers have turned their attentions toward nanostructured hydride materials, whose tiny size and large surface area make it possible to tune its properties independently of their bulk counterparts [5]. Research is ongoing, and it remains to be determined whether the required thermodynamics can be achieved to make metal hydrides a viable H₂ storage material, especially considering that hydrogenation induces mechanical stresses due to volume expansion.

Four different strategies have been applied over the years to reduce the decomposition temperatures of metal hydrides: 1) The nanoconfinement approach, where the complex size is reduced by infiltration into a nanoporous scaffold, decreases the decomposition temperature as a result of an increase of the reaction kinetics [6,7,8,9]. 2) Thermodynamic destabilisation uses additives to alter the decomposition pathway, therefore reducing the decomposition temperature. These additives may include a metal hydride, a complex metal hydride, a metal, or even a metal based compound such as a metal oxide [10,11]. 3) Using a combination of scaffolds can both nanoconfine and thermodynamically destabilise the infiltrated complex metal hydride [12,13,14,15]. 4) Finally, the use of a catalyst increases the reaction kinetics of the metal hydrides, resulting in a lower decomposition temperature [16,17].

This review focuses solely on a complex metal hydride system in the metal borohydride family. Sodium borohydride (NaBH₄) is a stable, inexpensive solid containing 10.6% (by weight) H₂ that has been a popular material of study for energy applications [18,19,20,21,22,23]. At the start of the century, scientists attempted to make use of a hydrolysis process to make NaBH₄ a feasible H₂ source. The clean concept was to simply react NaBH₄ with water via the following reaction [24]:

NaBH₄ + 2H₂O → NaBO₂ + 4H₂

where the sodium metaborate (NaBO₂) could eventually be recycled back into NaBH₄ by using a catalyst such as magnesium hydride [25,26]. Unfortunately, this process met with multiple failures due to solubility issues and low yields, so was abandoned by the DOE in 2007 [27,28].

In more recent years, researchers have turned to non-aqueous, solid state methods to generate H₂ from NaBH₄. When heated, NaBH₄ releases 133 ± 3 J K⁻¹ mol⁻¹ of H₂ along with solids of sodium (Na) and boron (B), with a small amount of sodium borides and sodium hydride [2]. Furthermore, the bulk of the decomposed by-product is recyclable according to theoretical calculations, where the solids remaining after dehydrogenation can reform NaBH₄ via the addition of more H₂ gas:

Na(s) + B(s) + 2H₂(g) → NaBH₄(s)

Unfortunately, the metal to ligand charge transfer in NaBH₄ induces high thermodynamic stability [29], so NaBH₄ does not decompose until about 535 °C at 1 bar of H₂ [2,30]. This is significantly higher than the desired <125 °C proposed by the DOE. As a result, research efforts have turned toward developing methods to weaken the NaBH₄ structure by using additives to destabilize the bonds and decrease the temperature of decomposition, thereby making H₂ more accessible. For example, using zirconium chloride [31] or titanium fluoride [32] as additives reduces the onset temperature of dehydrogenation to about 300 °C, while cobalt, copper and nickel nanoparticles reduce this onset to as low as 200 °C [9,17,33,34]. Unfortunately, dehydrogenation starting at such low temperatures is not a guarantee that the sample will decompose and completely release large amounts of H₂ at low enough temperatures to become an acceptable fuel source. Regardless, many researchers’ results show promise for the method of using additives to destabilize NaBH₄, and there is a need for ongoing work to develop a better understanding of how additives incorporate into NaBH₄ in order to improve future materials design.

Figure 1 compiles the known decomposition temperatures of NaBH₄ resulting from the use of various additives. Though there are no obvious trends for controlling decomposition through metal size, anion selection, oxidation state, etc., there is a wide range of accessible decomposition temperatures available using additives to lower the kinetics. Current methods have been able to reduce the decomposition temperature as low as 379 °C. Though this does not quite reach the targets set by DOE, NaBH₄ may still be a relevant H₂ storage material if a better additive can be found.

Figure 1 Compiled onset decomposition temperatures of NaBH₄ with additives, listing additive; decomposition in °C [15,30,31,32,33,35,36,37,38,39,40,41,42,43,44,45]. Note that the method of data collection varies between samples, which influences the onset decomposition temperature.

The aim of this review is to provide information about recent research involving various additives to improve the kinetics of NaBH₄ for hydrogen storage, focusing only on additives that have been found to successfully decrease the decomposition temperature of NaBH₄. Note that publications reporting only the H₂ release temperature without mention of decomposition are not included in this review. Though we have made every effort to capture all relevant research, the findings here are not exhaustive. This work is organized in sections by metal additive, explaining how the various additives destabilize and reduce the decomposition temperature of NaBH₄. All of the research presented here used ball milling (BM) as the driving force to incorporate additives with NaBH₄. Interestingly, the experimental methods for both synthesis and decomposition analysis varied greatly between groups. The molar ratios of additives used, milling rate and total mill time, breaks taken throughout the milling process to cool the material, and in some cases post-milling heating to anneal differed , which may play a large role in why some additives cause NaBH₄ to decompose at a higher or lower temperature than expected. Furthermore, variances in the methods by which thermal decomposition was measured may have a significant influence on the values reported. Though it is not traditional to include a subsection on synthetic techniques in a review like this one, we felt it important in this instance. Though no one has yet tied together how these various methods influence the final decomposition temperature, it is our hope that having all of this data in one place will help researchers make future decisions regarding their methodology. Our ultimate goal is that this review will help researchers gain a better understanding of how to select additives to more effectively weaken the bonds in NaBH₄, thus improving decomposition and H₂ release to make NaBH₄ a more viable hydrogen storage material.

2. Discussion

This review focuses on the changing decomposition temperatures of NaBH₄ when using the following additives to destabilize the NaBH₄ structure: Ti, Ni, Mn, Fe, Zr, and Na. A brief description of the synthetic methodology and a discussion of results are described below.

2.1 Titanium Additives

2.1.1 Titanium Additive Experimental Procedures

Ti has been incorporated into NaBH₄ through different ball milling techniques with the following methods:

●

Humphries et. al. milled NaBH₄ with Ti (65nm), TiB₂ and TiSiO₄ (2 mol %) for 1 hour at 280 rpm under argon [33]. Decomposition temperature was measured by TGA: heated to 600 °C at 2°/min with Ar purge of 50 mL/min.

●

Mao et. al. milled NaBH₄ with Ti, TiH₂ and TiF₃ (about 4:7 mol %) for 2 hours at 400 rpm under argon [32]. Decomposition temperature was measured by Sieverts apparatus: heated to 500°C at 5 °C/min.

●

Jansa et. al. reported the effect of BM NaBH₄ with the TiF₄ (2 mol %) additive at 300 rpm under argon [30]. Decomposition temperature was measured by TGA: heated to 600 °C at 2 °C/min with Ar purge of 100 mL/min.

2.1.2 Titanium Additive Results and Discussion

Several different titanium (Ti) additives have been found to lower the decomposition temperature of NaBH₄, including hydrides, fluorides, and metal nanoparticles. Though there have been obvious challenges with the formation of intermediates and in developing trends to understand the influence of oxidation states, the fact is that these additives have been found to influence the decomposition and dehydrogenation.

Figure 1 shows that the lower decomposition temperature was obtained with TiSiO₄ (Ti⁺⁴) additive, at 489 °C. Humphries et al. found that metallic Ti and Ti⁺² in TiB₂ decompose at higher temperatures than this (493 °C and 503 °C, respectively). The powder x-ray diffraction (PXD) patterns show that the addition of TiSiO₄ and TiB₂ did not change the diffraction patterns of NaBH₄ during the BM process [33]. However, after milling there was an additional diffraction peak observed for Ti that suggested the presence of titanium metal [38]. Likewise, Mao et al. found that the TiF₃ PXD exhibited two phases [32,38] after BM, suggesting the presence of the solid additive. Looking to the study of milled TiF₃ with Mg(BH)₄ [46], there was a destabilisation of NaBH₄ due to an interruption of its framework. Furthermore, Kalantzopoulos et al. found that the boron nuclear magnetic resonance (NMR) spectra of NaBH₄ with the TiF₃ additive contained NaBF₄ in a low ratio of 1:150 NaBF₄:NaBH₄. It was assumed that NaBF₄ was in the amorphous state as it was not detected by PXD. Its formation can be explained thermodynamically during the warm-up induced by BM. This compound facilitates the dehydrogenation process, which will be discussed later, as it is an intermediate of the decomposition products NaF and H₂ [38]. Unfortunately, it is difficult to compare this additive effect to the others because the decomposition temperature is unknown. These samples are all physical mixtures, but the absence of additional PXD peaks for TiSiO₄ and TiB₂ indicates their possible presence in the amorphous state. Moreover, a similar reaction to the one with TiF₃ occurred during the BM using TiF₄, where NaBF₄ was formed [30]. Surprisingly, despite this substitution reaction, the decomposition temperature of this mixture (522°C) was only slightly lower than that of pure NaBH₄ (532 °C).

A first estimation of capacities is given by hydrogen desorption rate measurements at 440 °C for Ti (65 nm), TiH₂ and TiF₃. Within 400 minutes, the NaBH₄-Ti mixture desorbs 1.65 wt% of H₂, TiH₂ desorbs 1.85 wt%, and TiF₃ releases 3.2 wt% [32]. It should be noted that in the 20:1 mixture of NaBH₄:TiF₃, the first release of H₂ observed was from 300 to 360 °C, with a second step at 380 °C [32]. The more additive-concentrated sample, with a 4:1 ratio of NaBH₄:TiF₃, had lower release temperatures, with a first step from 220 and 320 °C and a second from 320 and 420 °C. Humphries et al. reported in-situ powder PXD patterns (Figure 2) [38]. From room temperature (RT) to 220 °C, the formation of Na₃TiF₆ was detected, whose diffraction intensity decreased until it disappeared at 320 °C, at which point NaF was formed. Thus, the two steps decomposition process of NaBH₄-TiF₃ are related firstly to the dehydrogenation that occurs during the Na₃TiF₆ formation, then to the decomposition of NaBH₄ decreased by the presence of the ternary fluoride [38]. In contrast to this, Mao et al. reported TiB₂ and NaF formation during TiH₂ and TiF₃ decomposition, which was determined to be an active species, but they made no mention of the appearance of Na₃TiF₆. They also commented that TiF₃ had better H₂ desorption and adsorption capacities compared to TiH₂ and Ti.

Figure 2 In-situ powder XRD of NaBH₄/TiF₃ system between 50 and 560 °C with a heating rate of 2 °C/min [38]. Reproduced from Ref. [38] with permission from the PCCP Owner Societies.

The thermogravimetric analysis (TGA) of a NaBH₄ and TiF₄ mixture indicated a mass loss of 0.3 wt% between 100 and 300 °C, and 20% between 300 and 600 °C, with H₂ desorption starting at about 470 °C [30]. It is difficult to compare TiF₄ to TiF₃ because the time exposures to temperatures were not reported, generated species were not specified in the case of TiF₄, and the ratios used differed. However, it is noteworthy that the decomposition of NaBH₄ with TiF₄ seemed particularly efficient, which does not contradict the observation made by Kalantzopoulos et al. about the effect of oxidation state in fluoride-containing additives.

It was observed that high oxidation states generally tend to more strongly decrease melting and decomposition temperatures. Jansa et al. [30] explained that the high coordination of the metal induced a high exchange area between NaBH₄ and the anions of the additive, and the additive cation furnished some vacancies that permitted further reactions. This was noticed using both chloride and fluoride additives milled with MgH₂ [47,48]. For instance, NaBH₄-CeF₄ decomposed at 502 °C whereas NaBH₄-CeF₃ decomposed at 506 °C. Furthermore, NbF₅ (Nb⁺⁵) is notably one of the best destabilizers of NaBH₄. However, some exceptions exist; FeF₂ and CrCl₂ more effectively reduce the decomposition temperature than FeF₃ and CrCl₃ [47]. Electronic structures of metals, depending on their oxidation state, are responsible for creating bonds during milling or heating, which influence the decomposition process. Indeed, it was highlighted earlier that TiF₃ and TiF₄ behaviours during BM, then heating, are very different. In addition, their effect on NaAlH₄ unit-cell dimensions also differ slightly, where the unit-cell parameter using TiF₃ is 5,0231 Å and using TiF₄ it is 5.0235 Å [49].

Ti additives offer much in the way of improving our understanding of how additives influence the decomposition of NaBH₄. Though the lowest decomposition temperature reported reaches only 493 °C, H₂ release has been reported in these mixed materials with an on-set temperature as low as 220 °C. There are many challenges to overcome if researchers pursue the use of Ti additives. The formation of intermediates such as Na₃TiF₆ and NaF make the chemistry of decomposition difficult to study, and raise questions about the potential for recyclability. More controlled studies are needed to gain a solid understanding of the influence of Ti oxidation states as well. The lowest reported decomposition onset temperature was 493 °C when using 65nm Ti nanoparticles as additives. Given the growing popularity of the study of nanoparticles and the well-known fact that nano-sized particles oftentimes have different tuneable properties compared to their bulk-sized counterparts, it would be worthwhile to continue this direction of research to see if smaller nanoparticles would lower the decomposition even further. It’s possible that Ti will not lower the decomposition and dehydrogenation temperatures enough to be feasible for H₂ storage and transportation, but the potential insights gained about oxidation and particle size trends suggest this is a material in need of additional study.

2.2 Nickel Additives

2.2.1 Ni Additive Experimental Procedures

Nickel (Ni) additives have been studied using a traditional BM technique. Humphries et. al. has reported data for traditional milling methods using five different nickel (Ni) additives; Ni nanoparticles (20nm), Ni₃B, NiCl₂, NiF₂ and Ni on Si/AlO₃. A 2 molar % concentration of additives were BM at 280rpm with NaBH₄ under argon atmosphere for 1 hour at 280 rpm, resulting in the final mixed products [33]. The decomposition temperature was measured by TGA: heated to 600 °C at 2° /min with an Ar purge of 50 mL/min.

2.2.2 Ni Additive Results and Discussion

Unlike the Ti additive, pure nickel (Ni) nanoparticles used as traditional additives through BM do not reduce the decomposition temperature better than their halogenated counterparts. It is fortunate that BM the Ni additives with NaBH₄ does not result in the creation of many boron derivatives, which is promising for H₂ cyclability, but chloride additives do tend to create Na(BH₄)_1-xCl_x. Though halogens have been mentioned, it remains to be seen how the oxidation state of Ni influences decomposition.

Both halide additives, NiCl₂ and NiF₂, appear to be the best destabilizers of nickel compounds just behind the mixed material Ni on Si/AlO₃. NiCl₂ decreases the NaBH₄ decomposition temperature to 460 °C and NiF₂ to 453 °C. The other additives examined, Ni₃B and Ni particles, with a much lower oxidation state (Ni⁰), were nevertheless effective as they lowered the decomposition temperature of NaBH₄ to 462 °C and 483 °C, respectively. This temperature distribution is consistent with observations made earlier with the Ti additives with regard to oxidation states. Humphries et al.’s first PXD study affirms that no reaction took place during the BM using each Ni additive, and no borane derivatives were generated, which favours reversibility [33]. Another experiment using more concentrated samples (with 20 and 25 molar % of additive) reported substitution reactions when using chloride derivatives such as NiCl₂ and TiCl₂. The appearance of Na(BH₄)_1-xCl_x during BM was brought to light by PXD and FTIR [50]. This phenomena, for which the substitution ratio depends on the transition metal of the additive, has also been observed using a NaCl additive [36], which will be discussed in the sodium (Na) section below.

According to Humphries et. al., the only gas released during heating of these mixed materials was H₂. The group studied desorption by cycling pressure-composition-temperature (PCT) and found Ni₃B and NiCl₂ to be the most effective Ni additives (Table 1). After the first cycle, the desorption using NiCl₂ at 514 °C reached up to 0.25 wt%, and 1.48 wt% at the second cycle, and with Ni₃B at 460 °C the desorption was up to 1.45 wt%. It was revealed by PXD that Ni additives form Ni₃B during heating. Given the Ni₃B efficiency at destabilizing NaBH₄ during desorption, it was logical to observe encouraging results with other Ni components. The particular destabilizer effect of the chloride-based nickel compound could be correlated to the appearance of Na(BH₄)_1-xCl_x during BM, but it is not sufficient to explain its best hydrogen desorption during the second cycle. Nevertheless, NiCl₂ seems to show the most promise for H₂ storage applications, but with the disadvantage that it desorbs only at high temperatures. On the other hand, NiF₂ is advertised as being less effective but it has been analysed at a temperature 50 °C lower than that of chloride whereas NaBH₄-NiF₂ has just a 7 °C lower decomposition temperature. Furthermore, after a few dehydrogenation/rehydrogenation cycles, formation of metallic Na from the decomposition of NaBH₄ and NaH from the hydrogenation of metallic Na were observed in each case. The general results of Humphries et al. showed that chloride and boride are better than fluoride additives in decreasing the decomposition temperature of NaBH₄, as observed in Figure 1.

Table 1 Pressure-composition-temperature hydrogenation and dehydrogenation cycling of NaBH₄ milled with additives Ni, NiCl₂, Ni on Si/Al₂O₃, Ni₃B, and NiF₂ [33].

It is fortunate that heating NaBH₄-Ni mixtures produce very few boron derivatives. Though this is promising for recyclability, there are many challenges yet to overcome. For example, fluorinated Ni reduces the decomposition temperature but increases the H₂ desorption temperature. There is also a need to study NaBH₄ systems with varying Ni oxidations states, as this property may be important in lowering dehydrogenation, but has yet to be studied. Additional work is needed and encouraged to determine whether Ni additives may be the key to making H₂ more accessible in NaBH₄.

2.3 Manganese and Iron Additives

2.3.1 Mn and Fe Additive Experimental Procedures

Several groups have studied Mn and Fe as additives with NaBH₄. Below are the methods that have been successful in reducing the NaBH₄ decomposition temperature:

●

Humphries et al. ball-milled 25nm Fe nanoparticles for 1h at 280 rpm in a 2 molar % proportion with NaBH₄ [33]. Decomposition temperature was measured by TGA: heated to 600 °C at 2 °/min with Ar purge of 50 mL/min.

●

Jansa et al. took a different approach and BM FeCl₃ with NaBH₄ in the ratio 1:4 under argon during 3 hours, with 1:100 powder to stainless balls ratio, and MnCl₂ was BM under the same conditions with an additive/NaBH₄ ratio of 1:3 [50]. Decomposition temperature was measured by TGA: heated to 600 °C at 2 °/min with Ar purge of 50 mL/min.

●

Kalantzopoulos et al. also incorporated Fe additives, where 4NaBH₄:1FeF₃ and 4.8NaBH₄:1MnF₃ were BM at 400 rpm under argon for 3 hours, under 1 bar of argon gas [38]. Decomposition temperature was determined via TGA by heating the sample between 30 and 600 °C with a heating rate of 21 °C and an argon gas flow of 50 mL/min.

●

Sing et al. stirred MnCl₂ with a mortar and pestle with NaBH₄ in 20ml diethyl ether for 1 hour at 35 °C under nitrogen gas until the solvent evaporates totally. MnCl₂ was used in 10 wt%, 20 wt%, 30 wt%, 40 wt%, 50 wt% proportion in NaBH₄ [39]. The decomposition temperature was measured by TGA heating from ambient to 300 °C with a heating rate of 5 °C/min under a nitrogen gas flow rate of 50 mL/min.

2.3.2 Mn and Fe Additive Results and Discussion

Similar to what was discussed above, manganese (Mn) and iron (Fe) additives also generate boron derivatives during the milling process, and the use of chloride-based additives produces the same substituted Na(BH₄)_1-xCl_x structure reported for NiCl₂, TiCl₂ and NaCl. While these intermediates may pose problems for H₂ cycling, they also offer another means of reducing the decomposition and dehydrogenation temperatures of NaBH₄.

Kalantzopoulos et al. found that a portion of MnF₃ was reduced into MnF₂, which lead to the formation of NaBF₄ during milling with NaBH₄ [38]. Boron NMR showed a NaBH₄/NaBF₄ ratio of 40:1. The same product was generated in a higher amount (10NaBH₄:NaBF₄) when FeF₃ was used, but in an amorphous state undetectable by PXD. In the latter case, FeF₃ was reduced to metallic Fe, which led to the formation of NaF. The presence of NaBF₄ in the MnF₂ sample induced an overlap marked by an asymmetry of the melting peak of the differential scanning calorimetry (DSC) pattern [30]. This was not observed for the FeF₃ sample in which NaBF₄ was amorphous.

Residual gas analysis found that the sample containing Na-MnF₃ released H₂ in three steps. The hydrogen desorption started at a low temperature (135 °C) due to the high destabilization effect of MnF₂ (Figure 3) [38]. The first step occurred between 160 and 230 °C, when NaBH₄ reacted with MnF₂ to form the ternary fluoride NaMnF₃ with a release of 2.0 wt%. Then, between 230 and 280°C, 1.5 wt% was released. Finally, after 390 °C, 1.8 wt% was desorbed. For the Na-FeF₃ sample, from 161 to 350 °C, 1.7 wt% was released, with a high H₂ desorption rate between 250 and 350 °C. This was consistent with the in-situ PXD signal of NaBH₄ which started decreasing at 270 °C, and was followed by 1.5 wt% desorption between 350 and 490 °C. Given these dehydrogenation results, MnF₃ may be slightly better at destabilizing NaBH₄ than FeF₃. However, the decomposition temperatures of NaBH₄ with MnF₃ and FeF₃ are very similar, 483 °C and 498 °C, respectively (Figure 1).

Jansa et al. reported that during the BM of the chloride-containing metals MnCl₂ and FeCl₃, the BH₄^- ligand was partially substituted by Cl^- to form Na(BH₄)_1-xCl_x [50], as described for the NiCl₂, TiCl₂ and NaCl additives. The substitution was higher using FeCl₃ than MnCl₂. Nevertheless, NaBH₄ containing MnCl₂ had a lower H₂ release temperature below 230 °C. This is a result of other species being formed, such as amorphous Na–Mn-borohydride. Furthermore, this mixture desorbed a total of 5.7 wt% H₂. It was found by Singh et al. that the best concentration to lower the decomposition temperature was 20% MnCl₂ in NaBH₄, while NaBH₄ with FeCl₃ released H₂ and decomposed at 437 °C with a maximum H₂ desorption of 0.2 wt% between 50 and 600 °C. [39]

Figure 3 Temperature-programmed desorption with residual gas analysis from room temperature up to 575 °C of as-milled samples of a) pure NaBH₄, b) 4NaBH₄ + TiF₃, c) 4.8NaBH₄ + MnF₃, and d) 4NaBH₄ + FeF₃ [38]. Reproduced from Ref. [38] with permission from the PCCP Owner Societies.

The use of Fe and Mn additives to reduce the decomposition temperature of NaBH₄ has meet with some success. Interestingly, when compared to Ti additives, the opposite trend of halogen size occurs, with the Cl having the lower decomposition temperature (reaching 437 °C with FeCl₃ and 498 °C with FeF₃). This prohibits researchers from drawing all-encompassing conclusions about halogens attached to metal additives. It’s possible the trend here differs because of the multi-step process of dehydrogenation, or that the boron derivatives must be taken into account when attempting to draw conclusions about best additive selections. Additional research is needed to determine whether or not Mn and Fe additives may have a different outcome when incorporated as nanoparticles, and if other types of additives containing these metals might prevent the formation of the boron derivatives during BM.

2.4 Zirconium Additives

2.4.1 Zr Additive Experimental Procedures

Two zirconium (Zr) additives have been used in an attempt to improve the H₂ desorption of NaBH₄:

●

Kumar et al. ball-milled NaBH₄ with 1 mol % ZrCl₄ using stainless steel balls under 0.1 MPa argon gas for two hours, with a 30 minute break to cool the sample half way through the milling process [31]. Decomposition temperature was measured by TGA: heated to 600 °C at 2 °C/min with an Ar purge of 50mL/min.

●

Jansa et al. ball-milled NaBH₄ with 2 mol % ZrF₄ at 300rpm under an argon atmosphere for one hour [30]. Decomposition temperature was measured by TGA: heated to 600 °C at 2 °C/min with an Ar purge of 100 mL/min.

2.4.2 Zr Additive Results and Discussion

It is possible to reduce the decomposition temperature of NaBH₄ using halogenated Zr additives. Unlike the previous additives discussed, the chlorinated ZrCl₄ decomposes NaBH₄ at a significantly lower temperature than the fluorinated ZrF₄, with a difference of more than 100 °C. Researchers have yet to determine why the different halogens yield such drastically different outcomes, but it is promising that work has been done to determine how the BM process changes the chemistry of these additives.

Studies involving ZrCl₄ have found that this additive interacts with the surfaces of magnesium borohydride and magnesium hydride to improve H₂ release while reducing to ZrCl₃/ZrCl₂ and nano-sized Zr metal [51,52]. When added into the NaBH₄ system, a similar change of the precursor occurs. The thermodynamic calculations of Kumar et al. discovered that ZrCl₄ reacts with NaBH₄ during BM via the following two pathways [31]:

4 NaBH₄ + ZrCl₄ → 4NaCl + 4B + Zr + 8H₂ (g)

2 NaBH₄ + ZrCl₄ → 2NaCl + 2B + ZrCl₂ + 4H₂ (g)

The ZrCl₄ reduces partially to ZrCl₂ and completely to metallic Zr. Kumar et. al reported that the ZrCl₄ additive reduces the crystallite size of NaBH₄ during BM, making NaBH₄ more reactive as a result of the increased surface area. Interestingly, though it was clear that ZrCl₄ reacted with NaBH₄, no NaCl was detected by PXD as a byproduct upon completion of the BM process. This problem with detection was either due to small concentrations or else the NaCl was amorphous, so not measurable by this technique.

TGA results comparing pure NaBH₄ (p-NaBH₄) and ZrCl₄-containing NaBH₄ (c-NaBH₄) are shown in Figure 4 [31]. The decomposition and dehydrogenation of the milled p-NaBH₄ began at 475 °C, with the sample losing more than 60% of its weight by 600 °C. This weight loss reportedly was a result of both hydrogen desorption and the evaporation of Na metal, the latter of which does not bode well for recyclability of this material. The ZrCl₄ additive, on the other hand, had a significant impact on the decomposition temperature, reducing the onset temperature to below 300 °C, with most of the H₂ content being released by 500 °C.

Figure 4 TGA of pure-NaBH₄ (p-NaBH₄) and ZrCl₄-containing NaBH₄ (c-NaBH₄) with corresponding hydrogen evolution observed by MS (inset) [31]. Reprinted from Int J Hydrogen Energ., Vol 42, Kumar S, Jain A, Miyaoka H, Ichikawa T, Kojima Y., Study on the thermal decomposition of NaBH₄ catalyzed by ZrCl₄, 22432-22437, Copyright 2017, with permission from Elsevier.

The Fourier-transform infrared spectroscopy (FTIR) data showed no changes between pure and ZrCl₄-containing NaBH₄, while x-ray photoelectron spectroscopy (XPS) confirmed that there were notable changes to the oxidation state of the Zr additive. In fact, all XPS peaks shifted to lower energy after NaBH₄ and ZrCl₂ were milled together, with energy changes of the Zr3d electrons pointing to the Zr⁴⁺ oxidation shifting to Zr²⁺ after the reaction. Overall, the data reported found that ZrCl₄ reduced to ZrCl₂, but it is unclear if metallic Zr was formed or if ZrCl₂ was in any way responsible for the effect observed.

Unlike other metal examples listed in this review, the fluorinated ZrF₄ additive did not reduce the decomposition temperature of NaBH₄ nearly as much as the chlorinated ZrCl₄, reaching only 503 °C compared to 400 °C. The ZrF₄ additive also lost its diffraction pattern after BM, corresponding to the formation of metallic Zr and NaBF₄. It remains to be seen if cyclability of H₂ is possible with this complex, but given the high temperature required to decompose NaBH₄ with ZrF₄, it is likely that this mixed material is not worth pursuing.

Boron derivatives remain an ongoing challenge in the quest to find materials that will work for H₂ cycling, and Zr additives are not immune to this problem. However, the ability of the ZrCl₄ additive to decrease the decomposition so low may be one of the keys to designing a better material. Though going nano in size doesn’t always solve the problem, it would be interesting to see how Zr nanoparticles influence the decomposition of NaBH₄, and to see if more complex Zr additives have a different outcome.

2.5 Sodium Additives

2.5.1 Na Additive Experimental Procedures

Several groups have reported the use of Na additives with NaBH_4, with varying results as to changes in decomposition. Below are the methods reported that were successful:

●

Chater et al. mixed NaBH₄ with NaNH₂ in a 1:1 molar ratio and heated at 190 °C for 12 h under argon [53]. For a NaBH₄-NaNH₂ (1:1 molar ratio) high temperature study, NaNH₂-NaBH₄ was milled and heated to 220 °C for 1 h, then the mixture annealed at 70 °C for 10 days to reach the low temperature phase. Decomposition temperature was measured using a homebuilt temperature programmed desorption apparatus coupled to a quadrupole mass spectrometer. The sample was heated under a 100 mL/min flow of argon up to 350 °C at a rate of 2 °C/min.

●

Wu et al. BM a mixture of NaNH₂ and NaBH₄ (2:1 molar ratio) under argon at 300 rpm from 1 to 32 h [40]. Decomposition temperature was measured by TG-DTA: heated to 450 °C at 10 °C/min with an Ar purge of 10 ml/min.

●

Olsen et al. changed concentrations using a NaCl additive. A mixture of NaBH₄ and NaCl with molar ratios 9:1, 3:1, 2:1, 1:1, 1:2, 1:3, 1:9 were ball milled under argon at 500 rpm for 24 hrs. Samples were taken out after 2, 6, and 12 h of milling [35]. Decomposition temperature was measured by DSC: heated to 600 °C, at 2 °C/min.

●

Ravnsbaek et al. also used NaCl as an additive. NaBH₄ with NaCl were BM with respective molar ratios of 0.5:0.5 and 0.75:0.25 under argon 60 times with 2 min of milling followed by 2 min breaks to prevent overheating. After BM, 0.5g of the mixture was transferred to an Al₂O₃ crucible, placed inside a sealed argon filled quartz capillary, and annealed in a furnace at 300 °C for 3 days [36]. Decomposition temperature was measured by Sieverts apparatus: heated to 450 °C at 0.5 °C/min.

●

Rude et. al. mixed a 0.1:1 ratio of NaBF₄ to NaBH₄, BM 15 times under argon with 2 min milling followed by 2 min breaks to cool. Decomposition temperature was measured by TGA/DSC heated to 450 °C at 5 °C/min. [54]

2.5.2 Na Additive Results and Discussion

Sodium (Na) as an additive for NaBH₄ has mainly been explored in two forms, NaNH₂ and NaCl. As discussed already in this review, the use of chloride-based additives, such as NaCl, produces the same substituted Na(BH₄)_1-xCl_x structure reported for NiCl₂, TiCl₂ and NaCl, while NaNH₂ has shown to produce ammonia emissions. Despite this, the low decomposition and dehydrogenation temperatures for both NaCl and NaNH₂ show promise as additives as hydrogen storage materials.

NaBH₄-NaNH₂ is a composite hydrogen storage material (composed of an amide and a borohydride) which has a theoretical hydrogen storage of more than 7 wt% and the gas product obtained after heating is mainly H₂. NaBH₄-NaNH₂ was studied because hydrogen generation performances of composite materials such as NaBH₄-NaNH₂ is considered potentially superior to pure borohydrides or amides. The melting point of NaNH₂ is 210 °C and the boiling point is 405 °C. Wu et al. heated NaNH₂ to 420 °C and found that NaNH₂ decomposes in two stages. The first stage, prior to the melting point, the weight loss is only about 1.5 wt% [40]. The second stage, after the melting point, the weight loss is 11.6 wt%. Based on TG-DTA data for NaNH₂-NaBH₄ (2:1 molar ratio), this composite’s decomposition mechanism appears to be influence by NaNH₂, also having two stages, starting at 124 °C and ending at 420 °C with a total weight loss of 7.52 wt%. Wu et al. found that the thermal decomposition of NaNH₂_NaBH₄ (2:1) was best controlled below 400 °C [5].

Through a different study, it was found that ball milling a 1:1 ratio of NaBH₄ to NaNH₂ produced a new complex, Na₂(BH₄)(NH₂). This new compound desorbed hydrogen at about 290 °C after its melting point at about 220 °C [53]. This new compound’s high-temperature and low-temperature phase was studied and from the high temperature XRD and differential thermal analysis, NaNH₂ was found to react with NaBH₄ at 157 °C with good stability up to its melting point (219 °C), with no detectable structure change from 219 °C to 300 °C. Above 300 °C, the dehydrogenation reaction took place [41].

Interestingly, changing the molar ratio from 1:1 to 2:1 of NaNH₂/NaBH₄ potentially produced Na₃(BH₄)(NH₂)₂, which has an onset decomposition temperature of 267 °C, with a 6.85 wt% of hydrogen released by 400 °C [40]. Adding CoNiB to this Na₃(BH₄)(NH₂)₂ mixture reduced the onset of dehydrogenation as low as 200 °C with 5.05 wt% of hydrogen released upon heating to 500 °C [53].

Another Na compound, NaCl, was studied by Olsen et al. as an additive for NaBH₄. NaCl was chosen for its potential to alter the thermodynamic properties of borohydride materials through anion substitution[35]. The powder X-ray diffraction (PXD) of NaBH₄_NaCl with a 1:1 molar ratio showed that NaCl dissolves faster in NaBH₄ than vice versa and that 12 hours of milling is sufficient to reach a single phase for mixtures containing less than 50% NaBH₄ , while more than 50% required 24 hours to reach full homogeneity.

Ravnsbaek et al. found that annealing NaBH₄_NaCl after ball milling was better than ball milling alone. Synchrotron radiation powder X-ray diffraction (SR-PXD) data showed that the annealed NaBH₄_NaCl dissolved NaCl into NaBH₄ at about 240 °C. It was concluded from the SR-PXD data investigating of the decomposition pathway that at higher temperatures (~415 °C), the rate of diffusion is increased, and a homogeneous solid solution is formed before the NaBH₄-part decomposed at ~530 °C [36].

Rude et. al. combined a sodium tetrafluoroborate (NaBF₄) additive with NaBH₄ to make a NaBH₄–NaBF₄ composite, using PXD to directly observe the substitution of fluorine for the BH₄^- group of NaBH₄ [54]. New resonances found in ¹⁹F magic-angle spinning NMR supported DFT calculations and confirmed the formation of a complex ion, BF₂H₂, after BM. The H-F exchange destabilized the NaBH₄ structure enough to reduce the decomposition temperature to as low as 300 °C. Though other research groups have used different fluoride-based metal additives, the use of Na prevents confusion of the cations in the system and made it possible to observe the direct H-F exchange. Given the dramatic decrease in decomposition it is certainly worth pursuing other fluorine-based metals to determine whether the combination of a F exchange as well as the appropriate metal might enable the NaBH₄ system to come closer to approaching DOE’s guidelines for H₂ release. Furthermore, it is known that anion substitution can be readily accommodated in borohydrides [55], and there is certainly a need to pursue these additives further with the goal of decreasing the decomposition temperature of NaBH₄.

Including the NaCl additive with NaBH₄ not only reduced the NaBH₄ decomposition to 415 °C, but yielded a H₂ release of 2.6 wt%, which is 90% of the calculated capacity. It was unfortunate that NaBO₂ formed during annealing, as this raises questions about cyclability and prevents the mixed material from reaching 100% of the calculated capacity. However, reaching 90% H₂ release is still an impressive feat and is deserving of further exploration. In contrast, NaNH₂ as an additive can dehydrogenate at 300 °C with a decomposition temperature of 400 °C, but it produces ammonia emissions, which create additional concerns [31]. It is intriguing that milling NaNH₂ with CoNiB reduces the dehydrogenation temperature to 200 °C with 5.05 wt% hydrogen release and no ammonia emissions [30,31,53]. NaBH₄ containing NaCl and NaNH₂ are promising materials compared to mixtures using other additives in this review due to their low decomposition and dehydrogenation temperatures and ability to release a large amount of their H₂. Add to this how promising the addition of CoNiB is to these mixtures, and it is clear that Na additives deserve further exploration.

3. Conclusions

Controlling the decomposition temperature of NaBH₄ with additives has met with a wide range of successes, with the decomposition onset temperature ranging from 522 °C with TiF₄ to as low as 300 °C with H-F exchange in NaBH₄-NaBF₄. Unfortunately, researchers have yet to find a common trend to allow them to destabilize NaBH₄ closer to room temperature. Some additive studies have found that fluorides are better than chlorides at weakening NaBH₄, especially when it is confirmed that the fluoride can substituted into the BH₄ group, or in additive mixtures involving MgH₂. Ni and Ce additives with fluorides also create a lower decomposition temperature in NaBH₄, but not enough work has been done to determine whether this is uniformly true for all metals. It’s also unclear if there is a trend where metal nanoparticle additives reduce the decomposition more than their complimentary halogenated metals. For example, Ti nanoparticle additives create a significantly lower decomposition than TiF₄, yet the reverse is not true when comparing Ni nanoparticles to NiF₂.

Researchers have yet to find a reliable method to weaken NaBH₄ enough to make it a viable option for H₂ storage, and much more progress is needed to meet DOE’s guidelines to release H₂ below 125 °C. It is clear that in order to make a cyclable H₂ storage material of NaBH₄, it will be important to find additives that do not create boron derivatives during milling or annealing, a common challenge in the majority of the additives discussed in this review. Our hope is that compiling information on these additives will enable researchers to gain a better understanding of additive incorporation and to determine how a new system might be designed to make NaBH₄ a viable H₂ fuel source. It’s possible that creating a NaBH₄ mixture in-situ would be better than ball milling, or that different types or sizes of nanoparticles could help catalyze NaBH₄. There is still a lot that is unknown about NaBH₄ and additional research is needed before abandoning it as a potential energy source of the future.

Acknowledgments

This work was supported by the Molecular Foundry at Lawrence Berkeley National Laboratory, a User Facility supported by the Office of Science, Office of Basic Energy Sciences, of the U.S. Department of Energy (DOE) under Contract No. DE-AC02-05CH11231. The authors also gratefully acknowledge research support through the Hydrogen Storage Materials – Advanced Research Consortium (HyMARC), from the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, Fuel Cell Technologies Office, under Contract Numbers DE-AC0494AL85000 and DE-AC52-07NA27344.

Author Contributions

All authors contributed to the writing of this review.

Funding

Funding was provided by the U.S. Department of Energy: Office of Basic Energy Sciences, U.S. Department of Energy (DOE), Contract No. DE-AC02-05CH11231; Office of Energy Efficiency and Renewable Energy, Fuel Cell Technologies Office, Contract Numbers DE-AC0494AL85000 and DE-AC52-07NA27344

Competing Interests

The authors have declared that no competing interests exist.