On-board hydrogen production using multifunctional catalytic hollow fibre-based reactors

Abstract

Faced with the imminent depletion of fossil resources and the large environmental impact associated with their use, the 21st century society demands a gradual change from an economy based on fossil sources to one based on sustainable resources and processes compatible with the environment. In this respect, green ammonia (NH3) has been proposed as promising hydrogen (H2) carrier candidate for on-board application. However, since on-board space constraints represent a significant challenge facing the adoption of NH3 as a future fuel carrier, this work focused on the design and assessment of a pioneering technology for on-board H2 production via NH3 decomposition, i.e. multifunctional catalytic hollow fibre-based reactors. Today, the most prominent technology for hydrogen production via NH3 decomposition is a traditional catalytic packed bed reactor (PBR) that uses a precious metal-based catalyst. As a result, this expensive technology is outsized, thereby limiting its applicability to NH3-fuelled vehicles. The use of hollow fibre-based reactors represents a unique opportunity for innovative solutions in the development of a suitable technology for the production of H2 on-board. This is because they can overcome the limitations of traditional PBRs and allow for decreased required catalyst loading and/or the operating temperatures when compared with a PBR.

However, thermodynamic limitations of NH3 decomposition reaction at low temperature (i.e. 𝑥���������������� = 100% at T ≥ 450°C) and the NH3 poisoning effect on the fuel cell installed on H2-fuelled vehicles, represent still a barrier for the adoption of NH3 as a future fuel carrier for vehicular applications. Therefore, to tackle these problems, the development of a hollow fibre palladium-based membrane hollow fibre reactor (MHFR) has been included in this study as well. In this respect, the MHFR would combine both the reaction and separation units integrating a palladium (Pd) membrane with a 100% selectivity to H2. The approach adopted in this work was to develop a MHFR able to achieve a 99.99999% NH3 conversion in order to ensure an NH3 concentration at the exhaust lower than 0.1 ppm. By Le Chatelier’s principle, the membrane allows increasing the NH3 conversion beyond thermodynamic equilibrium limits by continuously removing the produced H2, enabling the membrane reactor to produce high-purity H2 at lower temperatures than conventional catalytic reactors.

This research work has been carried out at three different levels, i) materials development and characterisation (i.e. Chapters 2-5), ii) reactor design and performance studies during the NH3 decomposition reaction (i.e. Chapters 4- 6), and iii) feasibility study of the different reactors for on-board H2 production (i.e. Chapter 6).

Chapter 2 describes the methods used to synthesise and characterise the catalyst supports proposed, i.e. one carbon xerogel (i.e. CX), two activated carbon xerogels (i.e. ACX1h and ACX5h), and two N-doped carbon xerogels (i.e. UCX and NCX). Similarly, the procedures adopted to synthesise and characterise three series of catalysts, i.e. un-promoted and sodium-promoted ruthenium-based catalysts (i.e. Ru and Ru/Na), and cobalt/molybdenumbased catalysts (i.e. Co/Mo), were described in this chapter.

Chapter 3 provides a detailed description of the textural, chemical and structural properties of the carbon xerogels. For example, the nitrogen adsorption and desorption isotherms at -196°C showed that carbon dioxide (CO2) activation treatment of the carbon xerogels lead to two distinct effects, depending on its duration. Furthermore, it was found that the N-doping of carbon xerogels induced a reduction of the specific surface area and total pore volume of both UCX and NCX due to the addition of heteroatoms to their carbon lattice. Likewise, the Temperature Programmed Desorption and X-Ray Photoelectron Spectroscopy experiments confirmed that due to the presence of fewer oxygen surface groups and more nitrogen groups on their surface, both UCX and NCX exhibit a more basic character compared to CX, ACX1h, and ACX5h. Finally, based on their higher burning temperatures, N-doped carbon xerogels were also found to be more thermally stable than their nondoped counterparts.

Ru-based catalysts were extensively described in Chapter 4 of this thesis. Particular attention was paid to how the properties of the different carbon xerogels and the use of Na as catalyst promoter affected the performance of the catalysts during the NH3 decomposition reaction. In this respect, it was found that among all un-promoted catalysts, Ru-NCX exhibited the best performance due to, in part, the higher basicity and electron conductivity of NCX when compared to the other carbon xerogels. The performance studies, together with the Transmission Electron Microscopy results, showed that catalysts with a Ru average particle size higher than 2.5 nm exhibited higher reaction rates. Likewise, it was found that the addition of Na had a positive effect on the performance of all catalysts studied during the NH3 decomposition reaction (i.e. at least 3.7 and 1.5 times higher reaction rates at 450°C after the first and second reaction run, respectively). Finally, regardless of the use of Na, all Ru-based catalysts exhibited high thermal stability and catalyst preservability at the operating conditions (i.e. 450°C, 1 atm), as shown by 10 h reaction experiments and elemental analysis performed after the stability test.

Similarly, Chapter 5 provides a detailed description of the series of Co/Mobased catalysts studied in this work. It was found that Co/Mo-NCX was the most suitable catalyst candidate for the NH3 decomposition reaction and that the optimal metal particle size for Co/Mo-based catalysts is around 2.2 nm. Furthermore, also Co/Mo-based catalysts showed excellent thermal stability, as proven by the constant ammonia conversion achieved during the long-term stability tests (i.e. 100 h for Co/Mo-NCX and 10 h for the other Co/Mo-based catalysts).

The design of the hollow fibre reactors (HFRs) and the hollow fibre membrane reactor (MHFR) used in this work was covered in Chapter 6 and Chapter 7, respectively. In order to develop the hollow fibre reactors, the best performing Ru-based and Co/Mo-based catalysts were deposited into a 4-channelled hollow fibre (HF) substrate by adopting a two steps approach that ensured a homogeneous distribution of the catalyst. More specifically, a single HF unit 10 cm long was used for the deposition of Ru/Na-NCX, whereas a module of 10 HF units 5 cm long was used for the deposition of Co/Mo-NCX. Likewise, for the development of the MHFR, the HF was used as a support of a Pd membrane, and Ru/Na-NCX was packed in contact with the membrane in the shell side of the MHFR. The catalysts performance was assessed during the NH3 decomposition reaction between 100°C and 600°C using HFRs, and between 300°C and 450°C using the MHFR, to prevent damaging the Pd membrane. Furthermore, the thermal stability during the reaction experiments using HFRs was assessed at 450°C.

To conclude, Chapter 6 and Chapter 7 include the feasibility study of the different catalytic reactors presented in this work for on-board H2 production. With this scope, the PBRs, HFRs, and MHFR were compared in terms of volume, catalyst loading, and efficiency. This investigation demonstrated the superiority of the HF-based reactors over the traditional PBRs, not only in terms of remarkably high NH3 decomposition reaction rates but also by their noteworthy advantages in terms of costs, volume occupied, and efficiency.