Mixed-metal amide catalysts for ammonia decomposition applications

Our current reliance on fossil fuels is highly unsustainable due to their limited supply and environmental impact. There is an urgent global need to revolutionise the way in which we source and store our energy. Hydrogen is undoubtedly going to play a role in re-designing the transportation sector in particular, but there are significant barriers preventing its widespread implementation. The reversible chemical storage of hydrogen as ammonia provides all the benefits of hydrogen as a fuel, whilst also having characteristics enabling some of the challenges faced by hydrogen to be overcome. It is the catalytic cracking of NH3 to release the intrinsic H2, enabling the provision of hydrogen on demand for use in fuel cells or combustion engines that is arguably the most viable solution to developing a sustainable energy system. The main reason the potential of ammonia as a carbon free hydrogen vector at the point of use has yet to be realised is largely due to the unavailability of an efficient, cheap and effective means for accessing its hydrogen content at scale. Earth abundant Group 1 and 2 metal amides are among the most highly active and effective NH3 decomposition catalysts at modest temperatures and have matched or surpassed the performance of optimised transition metal based catalysts. A progression of different systems have been studied to investigate the ability of mixed-metal amides to tailor the active form and activity of the catalyst towards decomposing ammonia. A wide range of mixed-metal amides can be successfully synthesised by straightforward ball-milling of the constituent single metal amides in the appropriate ratios. The structure and behaviour of these high purity samples was then investigated under inert gas at a range of temperatures through both ex-situ laboratory based, and in-situ neutron diffraction experiments. Correlating changes in the structures of these mixed-metal amides with variations in the mass of the sample as a function of temperature have enabled a complete description of the behaviour and dynamics of the systems to be obtained. Analysis of the low temperature behaviour of Li3Na(NH2 )4 and Li3K(NH2 )4 under argon, showed initially NaNH2 or KNH2 melts and separates out of the structure via nonstoichiometric Li3R(1-x)(NH2 )(4-x) (R=Na or K). There is then a small temperature range for which the amides are present as separate components, before the solid LiNH2 dissolves into the liquid NaNH2 or KNH2 to form an inter-mixed molten phase with a composition approaching that of the constituent amides. LiNa2 (NH2 )3 and LiK2 (NH2 )3 cannot support this non-stoichiometry as they are line phases so they simultaneously separate and melt in one step, but ultimately form analogous inter-mixed molten phases. NaK2 (NH2 )3 does not separate on melting and instead forms molten NaK2 (NH2 )3 which is stable to at least 300 °C, and only separates above this temperature into a molten phase approaching NaNH2+ 2KNH2 . At high temperatures it is the decomposition of the two amides within the inter-mixed molten phases that is observed, and there is a degree of interaction between these two concurrent processes. It is ultimately H2 and N2 release that is observed, in ratios dependent on those of the constituent single metal amides, rather than NH3 which is seen for the decomposition of LiNH2 alone. The decomposition temperatures of the single metal amides, when part of this combined melt phase are also significantly lower than those seen during their individual decomposition. A thorough analysis of the phase space between the two end members of the Li-K-N-H and Na-K-N-H systems led to the identification of two new phases, nominally “LiK(NH2 )2 ” and “Na3K(NH2 )4 ”. The synthesis of a novel triple mixed-metal amide with a composition of Li6NaK(NH2 )8 has been proven and all these species are close to being crystallographically characterised. Imides generally have higher melting points than their corresponding amides, so an imide which forms at low temperatures and remains solid over a wide range of operating conditions would be an ideal candidate for a practical NH3 decomposition catalyst. K2Mg(NH2 )4 and Na2Mg(NH2 )4 are metastable kinetic products and do not readily separate, instead displaying a single melt temperature. They are stable to much higher temperatures before they decompose to novel molten mixed amide-imide phases, KMg(NH2 )(NH) or NaMg(NH2 )(NH) which crystallise out of solution at 367 °C and 413 °C respectively and do not themselves decompose until over 500 °C. It is the correlation of the behaviour of each of these mixed-metal amides under inert gas, with their catalytic activity towards NH3 that is crucial to improving their performance and understanding how to optimise NH3 decomposition conditions. The initial crystal structure of the mixed-metal amides does not affect the NH3 decomposition ability of the intra Group 1 mixed-metal catalysts, as these have separated and melted into a molten phase of the constituent metal amides at the temperatures required for NH3 decomposition. LiNH2 decomposes to the imide under NH3 whereas NaNH2 and KNH2 decompose to the respective metals. It is therefore Li2NH, Na, K that are the active forms of the catalysts and it is the stoichiometry to which these are present, that ultimately affects the overall NH3 decomposition ability of the system. However, K2Mg(NH2 )4 and Na2Mg(NH2 )4 decompose to KMg(NH2 )(NH) and NaMg(NH2 )(NH) which are the active forms of the catalysts for these systems. Mixed-metal amide systems can out-perform single metal amides across the entire temperature range. LiNa2 (NH2 )3 and LiK2 (NH2 )3 are currently the best low temperature catalysts and Li3Na(NH2 )4 and Li3K(NH2 )4 show the best conversion at high temperatures to date. It is the imide forming component which appears crucial to optimising the NH3 decomposition ability. This provides evidence that there is added value of the mixed-metal amides as NH3 decomposition catalysts compared to single metal amides. The percentage recovery of the catalyst is crucially important in order to maintain efficiency and improve lifetimes. Na and K are volatile so the Li-Na-N-H and Li-K-N-H systems suffer from poor containment and reversibility of such materials is clearly limited under these conditions. The NH3 decomposition ability of the K-Mg-N-H and Na-Mg-N-H systems is entirely different to that of the Group 1 mixed-metal amides, but significantly worse. However, K2Mg(NH2 )4 and Na2Mg(NH2 )4 show the highest catalytic mass recovery rates of close to 90 %, due to stabilisation provided by the amide-imide complexes. Future research efforts will therefore identify elements which combine the important properties of high activity towards NH3 decomposition with high mass recovery. This is likely to occur within a system which has an unstable amide and a stable imide or amide/imide in order to optimise the best catalysts for commercial applications.