Sodium borohydride production and utilisation for improved hydrogen storage

Abstract

Sodium borohydride (NaBH4) has been the subject of extensive investigation as a potential hydrogen storage material. Its advantages include its ability to be stored as a stabilised aqueous solution and hydrolysed catalytically on demand, providing hydrogen safely, controllably, and at mild conditions. However, several drawbacks were identified by the US Department of Energy (DOE) in their “No-Go” recommendation in 2007; namely, the limitations on its gravimetric hydrogen storage capacity (GHSC) and its prohibitive manufacturing cost. This thesis aims to address some of the major issues associated with the use of NaBH4 for hydrogen storage. Efforts have been spread across three key research areas, which were identified in a review of the recent literature: • Synthesising an efficient, durable, and inexpensive catalyst for the hydrolysis of aqueous NaBH4 solution • The design of a hydrogen storage system based on NaBH4 which can overcome the GHSC constraints imposed by the solubility of its hydrolysis by-product, NaBO2 • The development of a novel recycling process for regenerating NaBH4 from NaBO2, with the goal of reducing its production cost to within the DOE’s target range of $2-4 per gallon of gasoline equivalent (gge) To achieve the first goal, a new electroless plating method was developed for preparing cobalt-boron (Co-B) catalysts supported on shaped substrates. This method involves mixing the plating solutions at low temperature (<5 oC), in contrast to previously employed methods, which are generally carried out at room temperature or higher. The new method requires only one plating step to achieve the desired catalyst loading, and has higher loading efficiency than processes requiring multiple plating steps, due to reduced catalyst wastage. The catalysts produced by this method show significantly higher NaBH4 hydrolysis activity than those prepared by conventional methods, with a hydrogen generation rate over 24000 mL/min/g recorded at a temperature of 30 oC and a NaBH4 concentration of 15 wt%. The improved activity of these catalysts was thought to be due to their increased boron content and nanosheet-like morphology, both of which are products of the low temperature preparation method. The stability of these catalysts was examined over several usage and deactivation cycles, which involved long-term exposure to alkaline solution. It was found that the deactivated catalysts were able to recover 80-90% of their initial activity even after 10 cycles, if operated at 40oC or higher. Characterisation of the deactivated catalysts revealed that a layer of NaBO2 had formed on the surface, blocking the catalytically active Co sites. The improved cyclic stability is most likely due to the dissolution of this layer at higher temperatures. A new reactor design for a NaBH4-based hydrogen storage system was developed, with the goal of overcoming the GHSC limitations encountered by conventional system designs. The new design is based on a fed-batch configuration in which solid NaBH4 is added gradually to a container of alkaline solution and reacted as it is needed, eliminating the need for high NaBH4 concentrations. The reactor is separated into high and low temperature zones, each of which serves a different purpose; the hydrolysis of NaBH4 occurs in the former, while crystallisation of the by-product NaBO2 occurs in the latter. This separation helps to prevent the crystallised NaBO2 from blocking and damaging the catalyst, which is a major issue encountered in previous designs. A 2.0 L-scale prototype reactor was built and tested, with a maximum GHSC of 4.10 wt% achieved on a reactants-only basis. This was increased to 4.74 wt% with simulation of water recycling from an attached fuel cell. Finally, a novel process for the low-cost regeneration of NaBH4 was proposed. This process uses organometallic hydrides to produce NaBH4 from an alkoxylated derivative of NaBO2, a new reaction which is reported for the first time in this thesis. The production of NaBH4 was confirmed by XRD, XPS, and NMR analyses, and an unoptimised yield of 45% was obtained. The organometallic hydrides can themselves be regenerated within the process, by thermal decarboxylation of the corresponding organometallic formate. If all intermediates are recycled, the only key input into the process is formic acid, a relatively inexpensive organic reagent. It is anticipated that this process can provide cost and energy savings of an order of magnitude relative to current NaBH4 production methods.