A high-density ammonia storage/delivery system based on Mg(NH₃)₆Cl₂ for $\mathrm{SCR}$–${\mathrm{DeNO}}_x$ in vehicles

Abstract

In this paper, we present a new benchmark for the automobile selective catalytic reduction of ${\mathrm{NO}}_x$: Mg(NH₃)₆Cl₂. This solid complex releases ammonia upon heating and can be compacted into a dense shape which is both easy to handle and safe. Furthermore, the material has a high volumetric ammonia density of up to 93% of that of liquid ammonia. This provides a long lasting ammonia storage ($\approx 20000\mathrm{km}$ of driving per 6.2 L Mg(NH₃)₆Cl₂ for an average medium-sized vehicle). The controlled thermal decomposition of Mg(NH₃)₆Cl₂ was demonstrated. A small reactor with a volume of 785 mL was filled with $\approx 260\mathrm{g}$ of Mg(NH₃)₆Cl₂ yielding a bed density of 331 kg/m³. The reactor was coupled to a buffer with a free volume of roughly 200 mL. A heating wire wrapped around the outside of the reactor supplied the heat-energy. A mass-flow controller was used to simulate a varying ${\mathrm{NO}}_x$ signal. It was demonstrated that it was possible to control the desorption using a simple ON–OFF controller with the buffer pressure as the control variable. Approximately 99% of the ammonia contained in the salt could be desorped and dosed, while maintaining the 5 bars used as the set-point pressure. The low density was improved by compressing the Mg(NH₃)₆Cl₂ powder to a density of 1219 kg/m³, which is very close to the theoretical crystal density of 1252 kg/m³. Temperature programmed desorption showed that the ammonia could easily be desorped by heating the densified material. Stoichiometric calculations have shown, that compared to the current choice of ammonia delivery for mobile ${\mathrm{DeNO}}_x$ (thermal decomposition of a 32.5% wt/wt aqueous urea solution), the high-density Mg(NH₃)₆Cl₂ compound weighs 2.8 times less and takes up 3.1 times less space. This makes Mg(NH₃)₆Cl₂ ideal for use as an ammonia storage compound in both diesel and lean-burn gasoline-driven automobiles.

Introduction

With todays great focus on clean and efficient vehicles, the reduction of harmful emission gases from cars and trucks is now facing several challenges. To increase the fuel economy in gasoline-driven cars a combustion method known as lean-burn (higher than stoichiometric air-to-fuel ratio) has been applied. Unfortunately, the current technology capable of reducing ${\mathrm{NO}}_x$, CO and HC (uncombusted hydrocarbons) in gasoline-driven cars, the so-called three-way catalyst, only functions in a limited area around the stoichiometric air-to-fuel ratio. In lean-burn, there is too much oxygen present for the three-way catalyst to reduce ${\mathrm{NO}}_x$ (however, both CO and HC are readily oxidized). Diesel-cars, which are driven using the lean-burn combustion method, are not equipped with any ${\mathrm{NO}}_x$ reduction catalyst and it is not until recently with the up-coming EURO 4 emission standards (The European Parliament, 1999), that some sort of initiative from the automobile industry has been required. Several methods have been proposed, including the ${\mathrm{NO}}_x$ storage-reduction catalyst (NSR) (Matsumoto, 1997) and the urea-SCR method (Hyundai Motor Co.: Choi et al., 2001), which uses aqueous urea-solutions of 32.5% wt/wt (eutectic concentration) as an ammonia source for the SCR-process.

The NSR-method is highly sensitive toward sulfur in the fuel and optimal sulfur contents are less than 1 ppm. Such low sulfur containing fuel is at the moment unavailable and since the declared goal of the European Union is a fuel with less than 10 ppm sulfur (The European Parliament, 2003), there is no reason to believe, that the NSR method will be a competitive method for the European market. The urea-SCR method, which is the current choice of technology for removing ${\mathrm{NO}}_x$ from diesel-trucks, depends on the decomposition of urea, which is known for its many possible side reactions (Fang and DaCosta, 2003). The 32.5% wt/wt aqueous urea solution has a freezing point of −11 °C, which can present problems during start-up when e.g. a frozen block of urea has to be heated if the surrounding temperature is below this value. Another downside to this method is that the urea-solution needs to be sprayed into the exhaust to facilitate decomposition. Particles have been found to be formed during this process (Koebel et al., 2000), which may present problems in the nozzle (clogging). Obviously, liquid ammonia would have been ideal as the ammonia source instead of urea for the SCR process, but due to the inherent danger of liquid ammonia stored at 8 bars it is not considered safe for mobile applications. In the patent WO 99/01205 (Marko et al., 1999) a system consisting of a container packed with granular Ca(NH₃)₈Cl₂ or Sr(NH₃)₈Cl₂ is used as an ammonia source for the SCR reaction. Ammonia is released through heating of the compound. The theoretical storage capacity is very high (for Sr(NH₃)₈Cl₂ it is approximately 48 kmoles of ammonia per m³ of storage material), but since the compound is only considered stored in granular form, the effective storage capacity is significantly lower. Packing of granular materials rarely gives a void fraction below 45–50%. The high void fraction also decreases the effective thermal conductivity, resulting in a larger time-delay in a delivery system, which is an important control issue. A further draw-back of using Ca(NH₃)₈Cl₂ or Sr(NH₃)₈Cl₂ as the storage materials, is the very high vapour pressure of ammonia at room temperature. For Ca(NH₃)₈Cl₂ at 30 °C the vapour pressure is nearly 1 bar and as high as 9 bar at 80 °C. Sr(NH₃)₈Cl₂ is similarly high in vapour pressure at these temperatures (0.6 and 6.6 bar, respectively), which means there can be a serious safety risk in the event of an accident or rupture of the storage container. In addition, the handling of a material with $\approx 1\mathrm{bar}$ in vapour pressure at room temperature is inherently problematic.

We present in this paper an SCR-based method based on the controlled thermal decomposition of Mg(NH₃)₆Cl₂. This compound has a low ammonia vapour pressure at room-temperature $(\approx 0.2\mathrm{kPa})$ and has a high volumetric ammonia density (very close to liquid ammonia, see Section 2.4). A demonstration system was developed, which consists of an ammonia storage unit (Mg(NH₃)₆Cl₂ salt) coupled to a delivery system (a simple mass-flow controller). The entire storage-delivery system is controlled via a LabView program. The system uses simple ON–OFF control for the heating of the storage container and a mass-flow controller, which follows a sinusoidal-like flow function, in order to emulate a varying ${\mathrm{NO}}_x$ signal found during driving. It will be shown, that the system can easily deliver a transient NH₃ flow, while maintaining a given, somewhat oscillating, ammonia set-point pressure. Finally, using compacted material the demonstrated volumetric ammonia density reaches upto 93% of liquid ammonia, while still maintaining a fast release rate due to generation of nano pores as the ammonia desorption progresses.