Nitrogen fixation with renewable electricity: plasma catalysis as alternative for small-scale ammonia synthesis?
Kevin Rouwenhorst is a PhD student in the research group Catalytic Processes and Materials (CPM). Supervisor is prof.dr.ir. L. Lefferts from the Faculty of Science & Technology (S&T).
This PhD thesis presents an investigation into plasma-catalytic ammonia synthesis in a dielectric barrier discharge (DBD) plasma reactor. Ammonia (NH3) can be synthesized from hydrogen gas (H2) and nitrogen gas (N2). Ammonia has current applications as intermediate for the fertilizer industry and the chemical industry. About 45% of the current pure hydrogen demand in industry is used for ammonia production. Future applications of ammonia include its use as zero-carbon fuel and as hydrogen carrier. Thus, ammonia may play a significant role in a decarbonized energy landscape, as discussed in Chapter 1.
Currently, ammonia is produced via the Haber-Bosch process. In this process, hydrogen and nitrogen gas are compressed to 100-450 bar. Thereafter, the feed gas is combined with a recycle stream and the mixture is heated to 400-500°C. The gas mixture is then converted to ammonia over an iron-based catalyst with multiple promoters, until equilibrium conversion is achieved. The typical single pass conversion of hydrogen and nitrogen in this process is 15-20%. After the reactor, the gas mixture is cooled down to room temperature (or below) to liquefy the produced ammonia. The gas mixture is then recycled together with the feed gas.
The Haber-Bosch process has its limitations for renewable ammonia synthesis. Firstly, renewable electricity sources such as solar and wind are fluctuating, while the Haber-Bosch process typically relies on a continuous production. Secondly, the iron-based catalyst is only active for ammonia synthesis at elevated temperatures of around 400-500°C. This is due to the strong triple nitrogen-nitrogen bond for atmospheric nitrogen, which needs to be broken on the catalyst in order to synthesize ammonia.
This PhD thesis investigates whether pre-activation by a non-thermal plasma can loosen the nitrogen-nitrogen bond before reacting on the catalyst, thereby allowing for operation under milder conditions. A non-thermal plasma can be operated in a dielectric barrier discharge (DBD) plasma reactor at near-ambient temperatures. Highly energetic electrons are formed through electric fields and collisions with molecules, such as molecular nitrogen. The bulk of the gas will remain at near-ambient temperatures. Low temperature operation of the plasma reactor implies that the equilibrium shifts towards ammonia formation, implying the process can be operated at ambient pressure. Another benefit of the plasma reactor is that it is fed with electricity, and the mild operating conditions imply that the reactor can be easily started up and shut down. An overview of previous research on plasma-catalytic ammonia synthesis, as well as a current understanding of reaction mechanisms involved can be found in Chapter 2.
The bulk of this PhD thesis focuses on mechanistic understanding of reactions relevant for plasma-catalytic ammonia synthesis at various plasma intensities. First, plasma-catalytic ammonia synthesis at relatively mild plasma intensities is discussed, e.g. at specific energy inputs below 1 kilojoule per liter. In previous research, it was postulated that the activation barrier for nitrogen dissociation on the catalyst could be reduced by pre-activating the nitrogen molecules in a plasma. This could possibly result in a higher ammonia production on the catalyst. In Chapter 3, this is demonstrated with experimental data for ruthenium-based catalysts that vibrational or electronic activation of nitrogen molecules in the plasma increases the ammonia production rate via a lower activation barrier for nitrogen dissociation. In Chapter 4, a mechanism is suggested for the reduced activation barrier for nitrogen dissociation on ruthenium-based catalysts.
Plasma-catalytic ammonia synthesis at higher plasma intensities are discussed thereafter, e.g. at specific energy inputs in the order of 10 kilojoules per liter. Under these conditions, part of the nitrogen molecules is activated by the electrons in the plasma to such an extent that the molecules are broken in the plasma to form nitrogen radicals. These nitrogen radicals are highly reactive, and open new reaction pathways for plasma-catalytic ammonia synthesis.
In Chapter 5, the contributions of plasma chemistry, plasma catalysis, and thermal catalysis are evaluated for ruthenium-based catalysts. It is found that plasma chemistry reactions increase with increasing plasma power due to the formation of nitrogen and hydrogen radicals in the plasma. Also, ammonia can be formed via nitrogen radicals reacting with the ruthenium surface, resulting in plasma-catalytic ammonia synthesis. This becomes relevant at temperatures where ammonia can desorb from the ruthenium surface and when surface hydrogenation becomes sufficiently fast. Thermal-catalytic ammonia synthesis becomes relevant under the same conditions like without plasma, e.g. above 300°C. Plasma-catalytic ammonia synthesis reactions are typically faster than the thermal-catalytic ammonia decomposition reaction, resulting in ammonia synthesis beyond thermal equilibrium above 450°C.
Reaction mechanisms with nitrogen radicals are further investigated in Chapter 6. Various transition metals are tested for their activity, namely cobalt, copper, palladium, platinum, ruthenium, and silver. All catalysts are found to have a similar activity for plasma-catalytic ammonia synthesis. This is different from traditional thermal-catalytic ammonia synthesis, where the activity among transition metal catalysts varies by orders of magnitude. It is found that a dominant reaction step for plasma-catalytic ammonia synthesis is the Eley-Rideal reaction between a nitrogen radical from the plasma and a surface-adsorbed hydrogen atom on the catalyst surface.
One of the limitations of plasma conversions is that not only the reactants are activated by the plasma, but also the products. This limits the energy efficiency of plasma-catalytic ammonia synthesis. A zeolite 4A adsorbent was introduced to the plasma reactor. Ammonia can be removed in situ from the plasma zone via adsorption on the zeolite. Ammonia is located inside the pores of the adsorbent, where the plasma cannot penetrate to decompose the ammonia. As discussed in Chapter 7, the introduction of the adsorbent improves the energy yield for plasma-based ammonia synthesis by a factor two.
The techno-economic feasibility of plasma-catalytic ammonia synthesis is discussed in Chapter 8. The process is compared to a Haber-Bosch ammonia synthesis process, and emerging technologies for small-scale ammonia synthesis. It is found that plasma-catalytic ammonia synthesis is not economically feasible, even for the optimal plasma activation van stikstof. The reason for this is the exothermic nature of the ammonia synthesis reaction from hydrogen and nitrogen, implying that any plasma activation of molecules ends up as a heat loss. In Chapter 9, it is concluded that research on plasma conversions should focus on endothermic reactions, such as nitrogen oxide synthesis from air. Alternatively, ammonia synthesis at mild operating conditions may be achieved through more active catalysts and solid sorbents for ammonia separation.
