Ammonia Decomposition Catalyst with Resistance to Coexisting Sulfur Compounds

Shigeyuki Uemiya¹, Masayuki Uchida^2;*, Hiroshi Moritomi²,

Ryo Yoshiie¹ and Makoto Nishimura¹

A cheap and disposal catalyst will be required for the decomposition of  ammonia in the presence of sulfur compounds. The possibility of iron ore and red mud as the ammonia decomposition catalyst was investigated using pure or diluted ammonia containing hydrogenation hydrogen sulfide. On the other hand, a relative low conversion of ammonia was observed using a nickel-based commercial catalyst for the ammonia decomposition. The deactivation behavior of an iron ore catalyst caused by sulfur poisoning developed on the pretreatment atmosphere before the reaction; namely, the deactivation was observed for the H₂ pretreatment, while the high level of ammonia conversion remained for the CO pre treatment. From the X-ray diffraction pattern of the catalysts, the used iron ore catalysted in the CO atmosphere included FeC_x, which was also included in the red mud that was active for the ammonia decomposition. FeC_x may be responsible to the sulfur poisoning resistance.

Keywords: ammonia, catalyst, decomposition, iron carbide, red mud, sulfur compound

Experimental

Reactions were conducted using a conventional fixed-bed reactor made of a Pyrex glass tube. The catalyst was crushed and sieved to a particle size of 250-500um, and then a weighed amount(6.0g) of the catalyst was packed in the center of the reactor using silica wool and glass beads. Five kinds of iron ore, ROB, MAC, and NAM, red mud(main components are Fe₂O₃, SiO₂，TiO₂ and Al₂O₃), triiron monocarbide reagent(Fe₃C), and iron oxide reagent(Fe₂O₃) were used as the disposal catalysts and furthermore, a nickel-based commercial catalyst for the ammonia decomposition and an iron-chromium-based commercial catalyst for the high temperature water gas shift reaction, donated by Shutai were also used for comparison of the catalystic activity.

The catalyst was heated in a stream of argon and then reduced at 500℃ in a stream of pure H2 or CO before the reactions. The reactions were typically conducted at a temperature of 500℃, atmosphere pressure, a space velocity of 100h-1(NH₃-basis) and using a model reactant gas consisting of  pure Nh3 or diluted NH3 gas consisting of NH₃-50%,H₂S-5000PPm, N₂-balance. The sulfurated catalysts pretreated by 500PPm H₂S at 500℃ were also tested. The produced gas was analyzed using a TCD gas chromatograph, and its flow rate was measured by a soap film type flow meter. The level of ammonia conversion was defined as follows:

Identification of the crystalline phase of the catalyst after the pretreament and reaction was conducted using X-ray diffraction.

Results and Discussion

First, the ammonia decomposition activity of five kinds of iron ores was evaluated. The main products were hydrogen and nitrogen, and a trace amount of methane was also observed in the produced gas. the iron ores gave a slightly different ammonia conversion during the initial stage of the reaction, but after a 1-h reaction, the ammonia conversion reached almost the same level. We had hoped that the catalystic activity depended on the iron content or composition(or producing distinct), but the apparent relation between the catalystic activity and these variables was hardly observed. Thus, ROB was mainly used as the typical iron ore in this study.

Next, four kinds of catalysts pretreated and sulfurated by H₂S were evaluated, and the experiment results using the pure ammonia without H2S as the reactant are shown in Fig.1. Note that the red mud gave the highest activity among the catalysts tested during the initial stage of reaction, namely just after the H₂S pretreatment. On the other hand, a relatively low level of ammonia conversion was obtained for the nickel-based commercial catalyst for ammonia decomposition. Furthermore, is should be noted that the level of ammonia decomposition increased with time on stream for the iron ore and nickel-based commercial catalyst, because they were probably regenerated by the hydrogen and nitrogen, while such regeneration of the catalystic activity could not be observed for the red mud. These results indicate that some iron-based catalysts will have a catalystic activity for ammonia decomposition even in the presence of sulfur compounds, and will be regenerated by the produced hydrogen under appropriate reaction conditions. Assuming that iron is the active component for the decomposition, the iron ore has the possibility to provide a higher catalystic than the red mud, because the total amount of iron species in the iron ore(55.6%) was twice that in the red mud(27.9%). Unfortunately, the catalystic activity of the iron ore  was lower than the red mud. Since the red mud was a complicated to clarify which is active for the ammonia decomposition with the presence of H₂S.

Figure 2 shows the time variation of the ammonia conversion using H₂S(5000PPm)-ammonia as the reactant. The commercial nickel catalyst and iron ore produced a higher activity than the red mud during the reaction, but their activity drastically decreased with time on stream. The nickel catalyst and iron ore may be suitable for ammonia decomposition without sulfur compounds, while with sulfur compounds, the red mud had a relatively low, but stable activity during the reaction. Furthermore, it was experimentally confirmed that the sulfurated iron ore catalyst could be regenerated in a H₂ flow at 500℃, and the ammonia conversion just after the regeneration attained the level during the initial stage of the reaction without H₂S. This result indicates that the deterioration results from the poisoning of H₂S, and not from the sintering of the active sites. It can be concluded that red mud may include species that ammonia decomposition even in a sulfurated atmosphere.

The X-ray patterns of the untreated(before reaction), sulfurated, and used(after reaction) red mud catalysts are  illustrated in Fig.3. Some of the peaks can be attributed to the iron carbides of FeC₈ and Fe₃C. A similar behavior was reported for an iron-based catalyst used for FT synthesis; the Fe catalyst was carbonized by the treatment in the CO atmosphere and gave a relatively stable activity even for the sulfur-inculded reactant in the FT synthesis. The excellent resistance to H₂S poisoning was reported due to the formation of FeC_x.

The deterioration of the iron ore catalyst was investigated by different pretreament conditions of 1-h H₂ reduction, and 1-h and 4-h CO reductions before the reaction. These results are shown in Fig.4. As expected, the deterioration behavior was dependent on the pretreatment atmosphere. The deactivation of the catalyst bu sulfur poisoning was observed for the H₂  pretreatment, while the level of the ammonia conversion was kept constant for 4-h CO pretreatment. From the X-ray diffraction pattern of the CO pretreated iron ore, shown in Fig.5, the iron ore was carbonized to FeC_x, the peaks of which overlapped the peaks of Fe₃C with an orthorhombic phase. These phenomena agreed with the one reported for the FT synthesis.

For the purpose of clarifying the active species with a resistance to H₂S, the catalystic of the Fe₃C reagent for the ammonia decomposition with the presence of H₂S was investigated. The time variation in the ammonia conversion on the Fe₃C reagent with and without pretreatment and Fe₂O₃ reagent is presented in Fig.6 with the result of the red mud for comparison. It is shown that Fe3C itself didn’t have a resistance to H₂S. We assumed that an intermediate phase of  FeC_x during the transformation of   Fe₂O₃ to Fe₃C is the key species for resistance.

Figure 7 shows the relation between the catalystic activity of the Fe₂O₃ reagent and the CO pretreatment time with the result of the H₂-pretreated Fe₂O₃ reagent. It was found that a suitable time of CO pretreatment(4-8h) existed for providing the best resistance against the H₂S poisoning. Here the XRD patterns of  the carbonized Fe₂O₃ reagent(FeC_x) by CO pretreatment with different treatment time, the Fe3C reagent with a hexagonal phase, and red mud recovered after the reaction are shown in  Fig.8. Note that the peak positions were shifted with the CO pretreatment time. As mentioned above in Fig.5, the peaks of a carbonized iron compound(FeC_x) produced by the long CO pretreatment of iron ore overlapped those of a Fe3C reagent with the orthorhombic phase. In this case, the resistance against sulfur compounds was hardly observed. Furthermore, for the Fe₃C reagent with the hexagonal phase, the XRD peaks of which partly overlapped those of the short CO pretreatment samples, also didn’t have a resistance against the sulfur compounds. Therefore, the peak shift caused by the prolonged CO pretreatment  was found to be a result from the structural change in FeC_x from the hexagonal to orthorhombic phase. Namely, the intermediate phase from the hexagonal to orthorhombic transformation may be responsible for the resistance against the sulfur compounds. In fact, the XRD peaks of the red mud recovered after the reaction partly overlapped those of the intermediate phase observed for the sulfur-resistant samples obtained by the suitable CO pretreatment with the time of 4-8h.

4. Conclusion 

Cheap and disposal iron-based catalysts, especially red mud, will be applicable for the decomposition of ammonia in the presence of sulfur compounds, an FeC_x formed during the CO pretreatment had a resistance against the poisoning by sulfur compounds. The intermediate phase of the Fe2O3 reagent from the hexagonal to orthorhombic is speculated to be responsible for the resistance. The red mud residue its importance as cheap and disposal catalysts for the ammonia decomposition in the presence of sulfur compounds. Further study on the catalystic performance of the red mud should be conducted in near future.