Cobalt catalysts carbon number distribution

The carbon number distribution of Fischer-Tropsch products on both cobalt and iron catalysts can be clearly represented by superposition of two Anderson-Schulz-Flory (ASF) distributions characterized by two chain growth probabilities and the mass or molar fraction of products assigned to one of these distributions.7 10 In particular, this bimodal-type distribution is pronounced for iron catalysts promoted with alkali (e.g., K2C03). Comparing product distributions obtained on alkali-promoted and -unpromoted iron catalysts has shown that the distribution characterized by the lower growth probability a, is not affected by the promoter, while the growth probability a2 and the mass fraction f2 are considerably increased by addition of alkali.9 This is... 

For iron catalysts in general, the incorporation of 1-alkenes is negligible, and that of ethene is much lower than that of cobalt.1617 Therefore, for all published carbon number distributions for iron catalysts, a strict representation by two superimposed ASF distributions is obtained. Examples are given by Schliebs and Gaube,7 Dictor and Bell,8 and Huff and Satterfield.10 Also, the old experiments of the Schwarzheide tests are well represented by this model.7... 

Under intrinsic-kinetic conditions the carbon number distribution of products from a reduced fused magnetite catalyst Is not significantly affected by wide variations In H, and CO concentration and mass-transfer resistances have no noticeable effect, as would be expected. To the extent that other selectlvltles, such as oxygenate product composition, are governed by H and CO concentrations In the liquid, we would similarly expect to observe effects caused by mass transfer, although this was not done here. Likewise with other catalysts, such as cobalt, which appear to be more sensitive to reaction conditions and to secondary reactions, more marked effects from significant mass transfer reactions are anticipated. 

A value of c equal to 0.3, previously used to describe FT selectivity data on Ru catalysts (4), was also chosen here to describe the behavior of cobalt catalysts. This equation for hydrocarbon diffusion in melts reflects the strong influence of molecular size in reptation and entanglement models of transport in such systems (IJ6). Our model also requires the input of intrinsic values for jSn (given by the asymptotic j8r), jSo, j8r, and j8s, measured independently. After such parameters are specified, the model yields a non-Flory carbon number distribution of increasingly paraffinic hydrocarbons that agrees well with our experimental observations (Fig. 16). 

Fig. 18, Support effects on carbon number distribution for cobalt catalysts (473 K, H2/ CO = 2,1, 2000 kPa, 50-60% CO conversion).

However, the detailed description of the FT product distribution together with the reactant conversion is a very important task for the industrial practice, being an essential prerequisite for the industrialization of the process. In this work, a detailed kinetic model developed for the FTS over a cobalt-based catalyst is presented that represents an evolution of the model published previously by some of us.10 Such a model has been obtained on the basis of experimental data collected in a fixed bed microreactor under conditions relevant to industrial operations (temperature, 210-235°C pressure, 8-25 bar H2/CO feed molar ratio, 1.8-2.7 gas hourly space velocity, (GHSV) 2,000-7,000 cm3 (STP)/h/gcatalyst), and it is able to predict at the same time both the CO and H2 conversions and the hydrocarbon distribution up to a carbon number of 49. The model does not presently include the formation of alcohols and C02, whose selectivity is very low in the FTS on cobalt-based catalysts. 

---