Return to Table of Contents

SECTION I. (b)

(3) K.W.I. (Kaiser Wilhelm Institut, Muelheim) (See reference I(b)/6 and I(b)/7 at end of section). KWI developed a precipitated Fe catalyst using no carrier. The Fe is precipitated as hydroxide from its nitrate solution, washed, filtered, and pressed. In some cases it may also be alkalized. (For exact recipe see below).

It was found that reduction with H₂ reduces the Fe₂O₃ to Fe₃O₄, which does not yet catalyze the FT reaction. Treatment with CO however, gives a highly active catalyst, following the formation of the carbide. This process (“Formierung”) must carried out under very definite conditions.

It was found that operation with high COH₂ gas at atmospheric pressure damaged the catalyst. While at higher pressure no such effect could be established. From this it was concluded that the overcarbidization (“Uebercarbidierung”) at low pressure was due to the low H₂-partial pressure resulting in a lowering of this hydrogenating activity of Fe.

It was also found that the pressure at which the formation of the carbide takes place has an appreciable effect on the life and sustained activity of the catalyst in subsequent MP synthesis. Influence of pressure during formation on subsequent activity is shown below:

Thus a low pressure appears essential for the formation of an active carbide which will retain its activity for many months.

It was also found that for this formation of an active catalyst an optimum temperature exists (other conditions being equal) near 315° C. With lower temperature the activity is definitely inferior, and with higher temperature only a fair activity results.

Extent of “Forming." Aside from temperature and pressure, the space velocity and duration of the forming are also important. Since the reduction and carbide formation, result in the formation of CO₂ this affords a means to control the forming of the catalyst. The latter is considered terminated when CO₂ in the exitgas has passed through a maximum and reached a constant value. Further production of CO₂ then corresponds to deposition of free carbon.

The importance of this forming can be seen from the variation of the subsequent synthesis temperature required for maximum contraction. An Fe catalyst which was put into operation at 15 atm. Without “performing” required initially 290° C for a 45% contraction. At the end of the 5th month this temperature had to be raised to 300° C. A similar catalyst, formed” 24 hours with CO at 1/10 atms. And 255° C, was started at 250° C and at the end of the 16th month gave a contraction of 50% at 260° C.

Summarizing the “Forming” it consists in treatment of the catalyst at pressures below and temperatures above those used in the subsequent synthesis. It is preferred to “form” with pure CO at high space velocity. During this treatment a certain equilibrium is established between the solid phase and the gas phase. The lower the CO₂ content of the gas used in formation, the better the reduction and carbide formation.

Some work was done on the analysis of these Fe catalysts. In particular magnetic measurements are used to determine the extent of the conversion from Fe₂O₃ to Fe₃O₄ to Fe₃C. Two Curie points are involved in these changes which allow a determination of the conversion by comparison with known mixtures using a calibrated apparatus. There are probably many carbides present in the catalyst which make an exact determination of Fe₃C difficult. Furthermore it is probably the loosely bound carbon in the higher carbides which accounts for the activity of the catalyst. However, KWI found their method of determining Fe₃C a good way of predicting the activity of a catalyst. There seems to be about 70-80% Fe₃C in a good, well formed catalyst after prolonged use. After an initial complete carbidization the Fe₃C content is reduced to this figure and remains constant at this level. The deposition of free carbon on the catalyst is not considered a “poisoning” of the latter, but simply a mechanical disability.

The Influence of Alkali on the Catalyst. Alkali was found to increase production of higher boiling hydrocarbons as follows:

The % alkali is wt. % K₂CO₃ based on Fe. Metal. High alkali, however, shortens the life of the catalyst.

Addition of Kieselguhr. Contrary to the experience with cobalt, it appears that Kieselguhr is not required for Fe catalysts. The best results were obtained without a carrier.

Treatment with Hydrogen. It has already been pointed out, that the use of H₂ instead of CO for the “forming” of the catalyst gave poor results.

However, intermediate regeneration over the life of the catalyst gives an immediate but short lived increase in activity. The situation is similar to that encountered with cobalt. The regeneration must however be carried out before the contraction has dropped below 45-50%.

Preparation of Catalyst. A hot solution of iron nitrate is precipitated with soda. It was found that a mixture of ferrous and ferric nitrate gave the best result, pure ferric iron was not satisfactory, with pure ferrous an intermediate.

Copper may be added (½% based on Fe) to give a more reproducible result. The precipitate is washed free of alkali, then slurried in a potash solution, fitered, dried and pelleted. Next the catalyst is “formed” n the manner described above and finally purged with CO₂ for transportation. It is also possible to soak it in wax. The particles are thereby covered and can be exposed to air with losing their activity.

The finished carbided catalyst contains 50 g Fe metal in 100 cc Catalyst. The Fe represents about 60% of the total weight.

Products. The chief characteristic is the olefinicity of the product. The C₃-C₄ cut contained around 70-80% olefins; gasoline (200° EP), 50-60% olefins; diesel oil 10-20% olefined. The wax is practically olefine free.

The olefins and paraffins are mostly normal hydrocarbons. Overall, about 1 carbon atom in 30 is tertiary. Diolefines are absent.

Oxygenated products are characteristic by-products of the Fe synthesis. The yield was given an approximately 13 g/m³ water soluble products. Aside from alcohols, which are the bulk of the oxygenated products, the usual mixture of acids, ester, aldehydes was found, but it appears that not much work was done by KWI on this phase of the process. The oxygen content of the hydrocarbons phase was given as 0.2 to 2% with most of the oxygenated compounds in the low boiling fractions.

It was found that these Fe catalysts would produce some very high boiling alcohols (with chain length similar to those of waxes) if operated at 50 atm., or higher, but the yields are low and the catalyst has a tendency to form carbonyl at these pressures.

Summarizing the results of KWI work on Fe catalysts it may be stated that a catalyst and mode of operation was found which gave the following operational results:

For best results the feed gas preferably contains an excess of CO over hydrogen but this is not a necessity.

It might be mentioned that Kreislauf operation was not considered favorably here. It was felt that the cost was not justified by the increase in olefins available over a once-through basis.

Return to Table of Contents