Ammonia catalysts

Emmett P H and Brunauer S 1937 The use of low temperature van der Waals adsorption isotherms in determining the surface area of iron synthetic ammonia catalysts J. Am. Chem. See. 59 1553-64... 

Since the first synthesis of ammonia, catalyst development and chemical reaction engineering have been instrumental in the creation of the chemical process industry. As a result, catalytic processes have contributed much to the realization of prosperous civilizaticm. In the future, catalytic processes are expected to fulfill important roles in petroleum refining, diemical processing and environmental preservation. However, at present, many catalytic processes discharge large amounts of byproducts and consume large amounts of auxiliary raw materials. 

Figure 7.21. Fraction of unoccupied sites, and of sites occupied by atomic nitrogen and NH, as a function of reactor length on a potassium-promoted iron ammonia catalyst at 673 K,...

The gas reacts over the ammonia catalyst in an exothermic process at 450-500 °C, leading to an exit concentration of ammonia of about 15-19%. The ammonia is extracted by condensation and the unreacted gas recycled to the reactor. A fraction is purged to prevent the accumulation of inert components. The ammonia condensation is not complete, meaning that the real inlet gas of the reactor already contains several percent of ammonia. 

In this exercise we shall estimate the influence of transport limitations when testing an ammonia catalyst such as that described in Exercise 5.1 by estimating the effectiveness factor e. We are aware that the radius of the catalyst particles is essential so the fused and reduced catalyst is crushed into small particles. A fraction with a narrow distribution of = 0.2 mm is used for the experiment. We shall assume that the particles are ideally spherical. The effective diffusion constant is not easily accessible but we assume that it is approximately a factor of 100 lower than the free diffusion, which is in the proximity of 0.4 cm s . A test is then made with a stoichiometric mixture of N2/H2 at 4 bar under the assumption that the process is far from equilibrium and first order in nitrogen. The reaction is planned to run at 600 K, and from fundamental studies on a single crystal the TOP is roughly 0.05 per iron atom in the surface. From Exercise 5.1 we utilize that 1 g of reduced catalyst has a volume of 0.2 cm g , that the pore volume constitutes 0.1 cm g and that the total surface area, which we will assume is the pore area, is 29 m g , and that of this is the 18 m g- is the pure iron Fe(lOO) surface. Note that there is some dispute as to which are the active sites on iron (a dispute that we disregard here). 

I. The Use of Multicomponent Catalysts before the Development of the Ammonia Catalyst... 

The reasoning which led the author to make this first shot in the dark regarding the usefulness of combinations of solid compounds as ammonia catalysts was as follows If we assume that a labile iron nitride is an interminate in the catalytic ammonia synthesis, every addition to the iron which favors the formation of the iron nitride ought to be of advantage. In other words, the hypothesis was used that surface catalysis acts via the formation of intermediate compounds between the catalyst and one or more of the reactants. An experimental support for this theory was the fact that a stepwise synthesis via the formation and successive hydrogen reduction of nitrides had been carried out with calcium nitrides (Haber), and cerium nitrides (Lipski). Later, the author found molybdenum nitride as being the best intermediate for such a stepwise synthesis. 

We can readily understand these setbacks today if we consider the high sensitivity of iron as an ammonia catalyst toward numerous catalyst poisons. In those early years, this fact was unknown to us. Specifically, no one suspected the harm which is done to the catalyst by substances such as sulfur and sulfur compounds. Even Haber had never discussed a catalyst poisoning by impurities, because he had been able, apparently to avoid the presence of catalyst poisons in his small scale experiments. 

The main emphasis was laid, in this initial work, on Haber s catalysts, e.g., osmium and uranium compounds, as well as on a series of iron catalysts. Some other metals and their compounds which we tested are, as we know today, less accessibble to an activation by added substances than iron. Therefore, they showed no improvement or only small positive effects if used in the form of multicomponent catalysts. Finally, the substances which we added to the metal catalysts in this early stage of our work were mostly of the same type as those which had proved to favor the nitride formation, e.g., the flux promoting chlorides, sulfates, and fluorides of the alkali and alkaline earth metals. Again, we know today that just these compounds do not promote, but rather impair the activity of ammonia catalysts. 

After we had become convinced once again that added substances were of essential importance for the ammonia catalyst, the work in this field developed more and more into an increasingly systematized search. 

A support-effect such as demonstrated in the classical case of platinum asbestos is of negligible importance with ammonia catalysts except with osmium and ruthenium. 

A brief survey of the positive or negative effects of various elements (viz., of their oxides) on the activity of iron as an ammonia catalyst follows ... 

In February 1912, the author prepared for C. Bosch a survey on the characteristics of the more promising ammonia catalysts which had been tested so far. The following remarks may be of interest ... 

Molybdenum In its pure form, without additions, it is the most efficient catalyst of all the easily obtainable and reducible substances, and it is less easily poisoned than iron. It catalyzes in another way than iron, insofar as it forms analytically easily detectable amounts of metal nitrides (about 9% nitrogen content) during its catalytic action, whereas iron does not form, under synthesis conditions, analytically detectable quantities of a nitride. In this respect, molybdenum resembles tungsten, manganese and uranium which all form nitrides during their operation, as ammonia catalysts. Molybdenum is clearly promoted by nickel, cobalt and iron, but not by oxides such as alumina. Alkali metals can act favorably on molybdenum, but oxides of the alkali metals are harmful. Efficiency, as pure molybdenum, 1.5%, promoted up to 4% ammonia. 

Osmium This metal had already been found by Haber to be an excellent ammonia catalyst. Its activity is further increased by alkali metal oxides, especially by potassium hydroxide. As the pure metal 2% ammonia, promoted, 4%. 

Still during our experiments with ammonia catalysts, other technically important reactions were studied which promised to be accessible to the catalytic method, and particularly to the influence of multi-component catalysts. As this further development has been repeatedly described elsewhere, only a short survey will be presented here. 

Strangely enough, a combination similar to the ammonia catalyst, iron oxide plus alumina, yielded particularly good results (32). Together with Ch. Beck, the author found that other combinations such as iron oxide with chromium oxide, zinc oxide with chromium oxide, lead oxide with uranium oxide, copper oxide with zirconium oxide, manganese oxide with chromium oxide, and similar multicomponent systems were quite effective catalysts for the same reaction (33). 

However, the hydrogenation catalysts can be promoted also by compounds which are ineffective for the iron ammonia catalysts, e.g., by silica and silicates, silicofluorides, borates, and phosphates. An interesting type of hydrogenation catalysts was found in the form of zeolites which by ion exchange were impregnated with heavy metal salts. 

It is strange and typical for the erratic path of some laboratory work, that this final concept of the promoter action on ammonia catalysts was just opposite to our initial working hypothesis according to which flux promoters were considered to be essential for good catalytic activity. 

A favorable combination of valence forces of both components seems to be the basic principle of the nickel-molybdenum ammonia catalyst. It has been found (50) that an effective catalyst of this type requires the presence of two solid phases consisting of molybdenum and nickel on the one hand and an excess of metallic molybdenum on the other. Similar conditions prevail for molybdenum-cobalt and for molybdenum-iron catalysts their effectiveness depends on an excess of free metal, molybdenum for the molybdenum-cobalt combination and iron for the molybdenum-iron combination, beyond the amounts of the two components which combine with each other. A simple explanation for the working mechanism of such catalysts is that at the boundary lines between the two phases, an activation takes place. In the case of the nickel-molybdenum catalyst, the nickel-molybdenum phase will probably act preferentially on the hydrogen and the molybdenum phase on the nitrogen. 

Whereas some knowledge has been obtained about the working mechanism of ammonia catalysts (51), this does not apply to the same extent to catalysts used for many other processes. However, a few typical cases of multicomponent catalysts have been investigated both in the author s laboratory and by others. The main conclusion to be drawn from these studies is that it would be wrong to seek one universal explanation for the promoter effects in solid catalysts. As outlined above, structural as well as chemical effects may cause the improvements which are observed after certain substances have been added to a given catalyst. 

Aika, K. and A. Ncilscn Ammonia, Catalysts and Manufacture. Springcr-Vcrlag... 

From these results, it is concluded that, in a fully reduced catalyst, FeAl204 is not present furthermore, the aluminum inside the iron particle is present as a phase that does not contain iron (e.g., A1203), and this phase must be clustered as inclusions 3 nm in size. These inclusions may well account for the strain observed by Hosemann et al. From the Mossbauer effect investigation then, the process schematically shown in Fig. 17 was suggested for the reduction of a singly promoted iron synthetic ammonia catalyst. Finally, these inclusions and their associated strain fields provide another mechanism for textural promoting (131). 

Kummer and Emmett 311) studied the chemisorption of carbon monoxide on an iron ammonia catalyst by the same method and found a partial mixing of the desorbed gas, as if the surface consisted of a heterogeneous complex of homogeneous parts. Similar results were obtained by Eischens 312). 

The chemisorption of nitrogen on an iron ammonia catalyst was also studied by Emmett and Kummer 313), who found that the surface behaved as if it were of a homogeneous character. 

---