Carbon formation and catalyst deactivation

One of the most critical issues in developing catalytic reformers, especially for the reforming of hydrocarbon fuels, is the risk of carbon deposition on the catalyst surface and consequent catalyst deactivation. Carbon formation can occur in several regions of the steam reformer where hot fuel gas is present. Natural gas for example will decompose when heated in the absence of air or steam at temperatures above 650 °C via pyrolysis reaction as shown in Equation 2.4. 

Similar pyrolysis reaction can also occur during reforming of higher hydrocarbons. In fact, higher hydrocarbons tend to decompose more easily than methane and therefore the risk of carbon formation is even higher with vaporized liquid petroleum fuels than with natural gas. Another source of carbon formation is from 

There are several ways to reduce the risk of carbon formation in steam reforming reactions. Some of the approaches include  

The conventional nickel-based catalysts could be modified by adding oxide promoters such as potassium, lanthanam, cerium, and molybdenum in the catalyst formulations. It is believed that the added promoters improve the dispersion of nickel metal on the catalyst surface, thereby reducing the chance of carbon accumulation. Noble metals such as Pd, Pt, Ru, and Ir have been found to be more carbon tolerant as the solubility of carbon is less in these metals.54-57 However, they are more expensive than nickel-based catalyst, and as a consequence, they are less attractive for large-scale commercial applications. Alloying of nickel with other base metals such as Cu, Co, or noble metals such as Au, Pt, and Re has also been found to decrease 

As has been discussed in Section 2.3.6, the use of an adiabatic prereformer also helps in alleviating the risk of carbon formation in the steam reforming of higher 

---

Carbon formation and catalyst deactivation have also been observed in the SRE. Approaches similar to those discussed above for the steam reforming of hydrocarbons are also employed to suppress the carbon formation in ethanol reforming as well.1 7... 

Catalysts for coal liquefaction require specific properties. Catalysts of higher hydrogenation activity, supported on nonpolar supports, such as tita-nia, carbon, and Ca-modified alumina, are reasonable for the second stage of upgrading, because crude coal liquids contain heavy polar and/or basic polyaromatics, which tend to adsorb strongly on the catalyst surface, leading to coke formation and catalyst deactivation. High dispersion of the catalytic species on the support is very essential in this instance. The catalyst/support interactions need to be better understood. It has been reported that such interactions lead to chemical activation of the substrate 127). This is discussed in more detail in Section XIII. 

The steam to carbon ratio (S/C ratio) is the ratio of the moles of steam to atoms of carbon in the reformer feed. The S/C ratio, in conjunction with temperature and pressure, affects hydrogen yield, H2/CO ratio of the syngas product and methane conversion. The minimum S/C ratio for methane is about 1.7. However, excess steam is required to prevent carbon formation, avoid catalyst deactivation and adjust product Hj/CO [31. As a result, actual S/C ratios for steam reforming of methane are typically between 3.5 and 5.0. 

Nickel catalysts were used in most of the methanation catalytic studies they have a rather wide range of operating temperatures, approximately 260°-538°C. Operation of the catalytic reactors at 482°-538°C will ultimately result in carbon deposition and rapid deactivation of the catalysts (10). Reactions below 260°C will usually result in formation of nickel carbonyl and also in rapid deactivation of the catalysts. The best operating range for most fixed-bed nickel catalysts is 288°-482 °C. Several schemes have been proposed to limit the maximum temperature in adiabatic catalytic reactors to 482°C, and IGT has developed a cold-gas recycle process that utilizes a series of fixed-bed adiabatic catalytic reactors to maintain this temperature control. 

The catalysts used in the asymmetric isomerization of allylamines are very susceptible to water, oxygen, and carbon dioxide, and significant deactivation is observed by the presence of donor substances that include NEt4, COD, 27, and 28. Unfortunately, commercial production of 24 is usually accompanied by formation of 0.5-0.7% of 28 and a thorough pretreatment of the substrate 24 is required for the reaction system to attain high turnover numbers (TON), especially when Rh(L2)(5-BINAP)]+BF4 is used as the catalyst.42... 

The outside tubeskin temperature was taken to be identical to that generated in the previous simulation. The input data were also identical. Radial process temperature profiles are given in Figure 7. The ATg between the bed centerline and the wall amounts to 33°C, which is not excessive and permits the radially averaged temperature to be accurately simulated by means of the one dimensional model with "equivalent" heat transfer parameters, as discussed above. The methane conversion at the wall never differed more them 2% absolute from that in the centerline of the bed. The more detailed description which is possible by the two dimensional model would only be required if thermodynamic s predict possible carbon formation, and therefore catalyst deactivation, at locations different from those simulated by the one dimensional model. 

This work, on the other hand, is directed at understanding the kinetics of carbon formation and the species responsible for deactivation. In order to understand what is happening during deactivation, it is first necessary to have an understanding of the reaction kinetics. While reaction kinetics have been reported for iron Fischer-Tropsch catalysts, operating conditions were not sufficiently general and the catalysts were different from those used in this work. Hence, it was necessary to obtain reaction parameters over a wide range of conditions to meet the needs of this study. These parameters were obtained under conditions of little or no deactivation as well as under conditions for which deactivation plays a major role so that effects due to reaction and deactivation could be separated. 

Figure 9 shows rapid KDN deactivation of Co-Mo catalysts on both alumina and on carbon. The expectation that the carbon catalyst would deactivate quickly because it has a larger median pore diameter was observed. However, deactivation of the Co-Mo on alumina catalyst in Figure 9 was much faster than the Ni-Mo on alumina catalyst in Figure 8. An explanation for these differences may involve both the chemical coanposition of the catalyst surface as well as the diffusion path length. However, the deactivation of the 3.2 mm Co-Mo on alumina catalyst in Figure 9 was much faster than the 3.2 mm Ni-Mo on almnina catalyst in Figure 8. Since Ni-Mo is often considered to be a better hydrogenation catalyst more hydrogenation of adsorbed carbonaceous species, less coke formation, and less deactivation might be expected.

A review is given of the use of spectroscopic techniques to investigate the MTG process. Examples are quoted of catalyst characterization, investigation of the first carbon-carbon bond formation, alkene oligomerization and catalyst deactivation through coke formation. 

Although Pt and Cu supported on activated carbon catalysts exhibited promising catalytic properties in the hydrodechlorination of I2DCP [3], little work has been done on fundamental research relevant to these catalysts. For instance, adsorption data, which is an essential aspect in catalysis, are hardly available. Catalyst deactivation is a major problem that hampers the application of these catalysts in industry. The effect of coke formation on catalyst deactivation needs to be clarified and the kinetics of coke formation modeled before an industrial process can be designed. 

A Model for Reforming on Ni Catalyst with Carbon Formation and Deactivation... 

However, a direct link between catalyst properties and kinetics is missing in Aparicio s model (6) and our previous work (7). The present work deals with the development of a microkinetic model for syngas production including carbon formation and deactivation, where the activation energies are related to C-Ni, 0-Ni and H-Ni bond strengths determined by the BOC-MP method. 

The experiments were performed in the Tapered Element Oscillating Microbalance (TEOM) reactor (7,8), in which carbon formation and deactivation could be measured simultaneously by coupling with on-line GC analysis. The dry reforming of methane was studied on an industrial Ni (11 wt%)/(Ca0)a-Al203 catalyst at temperatures of 500 °C and 650 °C, total pressures of 0.1 MPa and 0.5 MPa and a CO2/CH4 ratio of 1. The BET surface area of the catalyst was 5.5 m /g, and the Ni surface area 0.33 m /g. The detailed experimental procedures were similar to that reported previously (7). 

Chen, D., Lodeng, R., Omdahl, K., Anundskas, A., Olsvik, O., and Holmen, A. A model for reforming on Ni catalyst with carbon formation and deactivation. Studies Surface Sci. Cataly. 139, 93-100, 2001. 

---