Mar-2022

The gauze catalyst for ammonia oxidation — a holistic approach

The gauze catalyst for ammonia oxidation has a unique position among the catalysts of large-scale chemical industry, since, in contrast to the supported catalysts usually used, it consists entirely of noble metal.

Brad Cook & Juergen Neumann
Sabin Metal Corporation, USA

Article Summary

Despite this circumstance and its long history, there have been hardly any developments and advances for this catalyst. This may be primarily due to the fact that from a scientific point of view only partial aspects were in focus, but only in the rarest cases was an attempt made to create a holistic picture of the whole process.

Despite early recognition that the process takes place under a mass transfer regime, this approach is rarely reflected in the scientific literature or considered in the design of the catalyst. Instead, for several decades, research and development has devoted itself to kinetic investigations of the surface reaction on the catalyst and the interpretation of recrystallisation processes and precious metal losses of the catalyst.

Little attention has so far been paid to the correlation with the plant design, which, with its specific reaction environment, places fundamentally different demands on the catalyst. Most of the articles on this topic deal primarily with the effect, not the cause.
This article will try to bring together the interconnected influencing factors of mass transfer limitation and the kinetics of the reaction, as well as the effects of recrystallisation and precious metal losses, into a holistic context.

A brief glance at history
The Ostwald process of ammonia oxidation for the production of nitric acid was only made possible by the large-scale industrial availability of ammonia according to the Haber-Bosch process. In the patent, Ostwald [1] describes the use of platinum as a catalyst in the form of wire nettings in a multi-layer arrangement.

As early as the 1920s, Bodenstein recognised the mass transfer limitation of the catalysed reaction. However, few academic publications are devoted to this topic. The overwhelming number of research activities focused on mechanistic and kinetic investigations of the processes on the catalyst surface, which have only been comprehensively clarified in the last 20 years.
The use of fabrics was a logical consequence of the requirements of the catalyst being able to flow through, as well as generating the lowest possible flow resistance. Up until the beginning of the 1990s, all catalyst gauzes were woven. This changed in the 1990s with the introduction of knitting technology for the production of catalyst gauzes, and knitted gauzes now differ from one another according to the different knitting technology used.

The alloy of the catalyst was initially adapted to the requirements of the process, in later years the price and availability of the different precious metals increasingly influenced the alloy used. Only two alloys have been specifically developed out of necessity. These include the Pt/Rh/Pd ternary alloy with increasing palladium content to reduce N2O emissions in the process, as well as the four-component alloy, which additionally contains tungsten in order to reduce the precious metal losses occurring in the process.

Mass transfer limitation
As stated previously, it was in the 1920s the ammonia oxidation reaction was discovered to proceed under a mass transfer limited regime, which means that the transfer of ammonia from the bulk gas phase to the surface of the catalyst is the slowest step in the course of the reaction.

Two conclusions can be drawn from this finding: On the one hand, the extent of the mass transfer limitation influences the number of necessary gauze layers in the catalyst package and, on the other hand, the extent can be influenced by the structure of the catalyst gauze.

From the knowledge of mass transfer limitation the number of necessary catalyst layers in the catalyst package can be calculated with relative accuracy by means of the ratio of the mass transfer rate and the gas velocity in the catalyst package. Two parameters are in the foreground here: on the one hand the plant load, which determines the flow rate; and on the other hand the pressure, which is found as the decisive variable in the diffusion rate for the ammonia. The figures below show the change in the NH3 diffusion rate (fig. 1) respective the course of the NH3 plant load (fig. 2) as a function of the pressure.

With regard to the mass transfer limitation, switching from woven to knitted catalyst gauzes has been a counterproductive step. While woven gauzes have a height of twice the wire diameter, in knitted gauzes the mesh protrude out of the plane, so that they reach a height of approximately seven times the wire diameter, which means that they have a significantly higher porosity. The porosity can also be translated as the permeability for ammonia, which makes it clear that a knitted structure is not necessarily the first choice for a mass transfer limited process.

The reason for the change to knitting technology is the inflexibility of weaving, with long wire fabrics being made in stock. High precious metal stocks with rising precious metal prices lead to uneconomically high costs, which are solely due to the precious metal. The knitting technology offers the possibility of just-in-time production of the catalyst gauzes, whereby the costs based on the precious metal inventories are reduced to a minimum.

Only Engelhard-Clal with its OMT (Optimized Mass Transfer) concept developed a gauze structure with the so-called “Bispin®” [2] catalyst gauze that promised improved mass transfer. The two images below (fig. 3) [3] show this “Bispin®” gauze in different views, which in principle consists of parallel wires around which a second wire is wound in a spiral shape.

The pictures show the basic structure of a woven fabric, in which the spiral-shaped wire protrudes into the mesh of the fabric and thus explains an improved mass transfer compared to a standard woven gauze. As praiseworthy as it is that Engelhard-Clal has taken on the problem of the mass transfer limitation, the ratio of the relatively high production costs for such a catalyst structure to the benefits must be questioned.

Reaction kinetics and process yield
While the mass transfer limitation is expressed in the reaction rate and ultimately the number of necessary gauze layers in the catalyst pack, the kinetics of the reaction have a direct influence on the selectivity of the reaction and thus on the yield of the process.

In the catalytic ammonia oxidation, three different reactions forming N2, N2O and NO take place in parallel.
NH₃+3/4 O₂ - 12 N2 + 3/2 H₂O
NH₃+4/4  Oâ‚‚  - 1/2  Nâ‚‚O +3/2  Hâ‚‚ O
NH₃+5/4  Oâ‚‚ - NO + 3/2  Hâ‚‚ O
    âˆ†HR = 319.9 kJ/mol
    âˆ†HR = 275.9 kJ/mol
    âˆ†HR = 226.6 kJ/mol
The kinetics provide information about the conditions under which each reaction takes place and at what rate, and thus the expected ratio of the three different reaction products according to the reaction conditions. Using the kinetic parameters from Kraehnert [4], the following picture of the product distribution over the temperature is obtained (fig. 4):
Nitrogen (N2) and nitric oxide (NO) are the main products of ammonia oxidation over platinum. N2 formation prevails between 150 and 400°C while NO formation is favored at higher temperatures.

Linkage of mass transfer limitation and kinetics
Only by linking the mass transfer process with the kinetics of the catalysed reaction will there be realised a holistic picture of the reaction process. The figure below (fig. 5) shows the NO selectivity over the temperature for the individual gauze layers in the catalyst package for the various types of plants from low-pressure to high-pressure operations.

The figure clearly shows the temperature gradient in the catalyst package which increases with the operating pressure of the reaction and which ultimately determines the NO selectivity and thus the product yield of the process. From the figure it can also be seen that only the reduction of this temperature gradient creates both the possibility of a reduction in the number of gauze layers in the catalyst package and an increase in the NO product yield.

Recrystallisation and Loss of Precious Metals
Recrystallisation and precious metal losses from the catalyst are the third element in the holistic view. Looking at the temperature gradient in the catalyst package, the lowest NO selectivity (and thus the highest N2 formation rate) is found in the uppermost catalyst layers. Relative to the higher enthalpy of reaction of the N2 formation, a correspondingly larger amount of reaction heat is released, which results in higher surface temperatures than in the lower catalyst gauze layers.

This temperature is sufficient for precious metal atoms to detach from the metal lattice, migrate over the surface and form larger surface agglomerates on colder surface regions or energetically exposed places. This leads to extensive surface reconstruction through the growth of agglomerates that form steps and terrace surfaces. The figure below (fig. 6) shows such stepped and terrace-shaped agglomerates on the recrystallised catalyst surface in 30,000-fold magnification.

Due to the high surface temperature, the highest levels of precious metal losses due to the formation of gaseous PtO2 are also found on the uppermost gauze layers. This PtO2 is stable down to a temperature of 350 - 450°C. Below this temperature it decomposes into elemental platinum and oxygen.

Such a decomposition of PtO2 is also responsible for the formation of cauliflower-like shaped growths on the catalyst surface [5]. If agglomerates grow far enough on the surface, they come into contact with the relatively cold process gas and the platinum from the decomposition of the PtO2 condenses on these growths, which thereby grow further and form the cauliflower-like structure. The cauliflower-like growths are therefore nothing more than an expression for the temperature difference between the catalyst surface and this surrounding process gas temperature. The size, number and formation of these structures decrease with increasing process gas temperature and correspondingly increasing gauze position in the catalyst package.

The figure below (fig. 7) illustrates this relationship between the position of the gauze layer in the catalyst package and the resulting surface morphology for increasing gauze layer positions in the flow direction of the process gas flow.