Return to Table of Contents

IV.  COBALT CATALYSTS

A. General Considerations

Martin (9) stated that the cobalt catalyst is still the favorite for most purposes, and has not been displaced by the iron catalyst..  The iron catalyst had been proposed largely as a war-time substitute for cobalt the supply of which was short in Germany, due to the war.  If the Ruhrchemie were to project a Fischer-Tropsch plant today, with free access to world market metals, they would recommend the cobalt catalyst, according to Martin.

Martin stated that they had discovered that the life of the cobalt catalyst could be materially lengthened by strict attention to thoroughness of purification of gas from sulphur, and that this life would now average five to six months.  Regarding regeneration of the cobalt catalyst, Martin said there was a loss of four percent per regeneration, but he felt confident this could be decreased to two per cent.  Asked about the most important factor involved in the manufacture of catalyst, Martin stated that the reduction step was the most critical and that the washing step was second in importance only to the reduction.

Martin stated that the catalyst containing the full amount of thorium produced more paraffin and that any substitution of thorium by magnesia decreased this production.  The magnesia was added solely to make the catalyst particle harder and less subject to disintegration to a dust.  It was necessary to reach a compromise between the relative proportion of thoria and magnesia in order to obtain both a reasonably high production and good physical quality in the catalyst.

Pichler (3) also stated that the best normal pressure catalyst is still the one containing cobalt, and when a good F. T. catalyst is required for some new work, the 100 Co:18 ThO2: 100 kieselguhr catalyst is usually selected.  Iron, nickel and cobalt catalysts are, in general, less susceptible to S poisoning than the noble metal catalysts.  The least susceptible is iron itself.  In comparing the three iron group catalyst, nickel appears to be the worst, since even at atmospheric pressure, there is a loss to carbonyls.  Ni-Co mixtures have been used, but that was done to spare cobalt, and cobalt is a better catalyst.

In the medium pressure synthesis, catalysts can be used which contain more kieselguhr, but the differences are small, and essentially the same preparation is used for both normal and medium pressure synthesis.  Earlier experiments with cobalt catalysts containing no kieselguhr indicated that at 20 atms. the catalyst becomes inactive after a few hours.

So far as could be learned all commercial operations in Germany were conducted with the cobalt catalyst.

B. Catalyst Manufacture and Reduction

The cobalt catalyst for all commercial synthesis units in Western Germany was made by Ruhrchemie at Sterkrade-Holten.  When this plant was visited on April 6 and 7, 1945 the information given below (12) was obtained from Dr. Bloechel, who was in charge of cobalt catalyst manufacture and recovery, and from Dr. Herbke, in charge of catalyst reduction.  Supplemental and generally confirmatory information was obtained from Pichler (3) and Alberts (2).

The composition of the standard cobalt catalyst prepared at Sterkrade was given as 100:5:8:180-200 cobalt:Thoria"magnesia:kieselguhr.  This catalyst had been in use since 1938 and replaced the former 100:15:200, cobalt:thoria:kieselguhr catalyst.

Alberts (2) stated that in his view the most important effect of replacing thoria with magnesia was the increased hardness of the resulting catalyst and consequent reduction of dust formation in the ovens which causes bad gas distribution leading to "hot spots".  Other advantages ease of initiating synthesis, less tendency to form methane and carbon and increased life.  It was only since magnesia-containing catalysts had come into use that lives up to 8 months had been achieved at normal pressure.

Alberts stressed very strongly the necessity for trials on the full-scale  as early as possible in the development of new catalysts.  On the laboratory scale, comparatively small differences were detected between ThO2, MgO, and ThO2 + MgO promoted catalysts.  In the full scale ovens, however, very important differences appeared.  He considered that Co-Mgo-Kieselguhr. catalysts were best for large-scale operation.

According to Bloeckel (12) the kieselguhr must contain less than 1% of iron (as tested by refluxing a sample for 1 hour with 25% HNO3) otherwise excessive methane production occurs during the synthesis.  It must also contain less than 0.4 Al2O3 in the raw, uncalcined state, otherwise the catalyst tends to "gel".  To avoid acid washing, the kieselguhr used at Sterkrade is selected to meet the above conditions.

The present material is obtained from deposits near Hannover and is calcined before use, at Sterkrade, at 600-700 C.  After this treatment the total volatile matter, including water, must not exceed 1%.

Alberts (2) stated that iron and calcium in the kieselguhr should be as low as possible, but that physical structure is of particular importance and there is risk of ruining this by acid treatment .  Pichler (3) said that in the early work iron and other impurities were removed by an acid wash but later it was found that the iron content was not of great importance and the technical kieselguhr was used directly.

Due to slight variations in the density of various batches of kieselguhr, the proportion used in the catalyst is varied over the range indicated above (180-200) in order to maintain a constant cobalt content of 80 gm./litre unreduced catalyst granules.  Alberts (2) stated that the use of dense catalysts (i. e., weighing more than 350 gm/litre) led to excessive reaction in the top layers of catalyst with the formation of methane and carbon.

The procedure for cobalt catalyst preparation at Sterkrade was given by Bloeckel (12) as follows, and is illustrated diagrammatically by Figure 1, page 17.

750 1. of a solution of the nitrates in the desired proportions, 100 CO : 5 ThO2 : 10 MgO (2 parts MgO are left unprecipitated) containing 40-41 gm. Co./1. are heated to 100C. in an overhead, stainless-steel tank, 1.  The contents are run into the precipitating tank, 2 fitted with a direct-drive twin-screw stirrer and containing 750 1. of a solution containing 104 gm. Na2CO3/1. also maintained at 100C.  (Pichler (3) stated that essentially the same results are obtained by adding the sodium carbonate solution to the nitrate solution, and that one solution should be added very rapidly to the other.)  The mixture is stirred for 1/2 minute and then dry kieselguhr is added through the hopper, 3 and stirring continued for a further 1 minute.  The slurry is pumped to a standard-type filter-press, 4 and the cake washed with distilled water until the wash-water is neutral as tested by the addition of nitrophenol to 100 ml. wash water + 5 ml. 1/10 N H2SO4.  Catalyst equivalent to 64 Kg.Co requires about 10m3 wash-water and the washing occupies 14-15 minutes.

The washed cake is then dropped into a "masher", 5 situated below the press and is mixed with dust from the screening plant.  (64 Kg. total Co gives 45 Kg. Co as dust).  The resulting cream is then pumped to a rotating suction filter, 6.  The thin cake scraped off the filter drum contains approximately 70% water and falls into the extruder, 7 where rotating arms force the paste through 3mm. diameter holes whence it falls into the drying chamber, 8.  The dryer is a cylindrical vessel 7 m. in diameter and comprises 20 super-imposed stages 20 cm. apart.  The catalyst is swept round each stage by rotating arms and falls down the vessel from stage to stage during a period of 1 - 2 hours.  Drying is effected by steam heat and an air blast.

From the final stage of the dryer the rough granules containing 10% moisture are carried by a conveyor to the vibrating screens, 9, and separated into - over 3 mm., 1-3 mm. (the desired size), fines, and a dust which is sucked away in an air current.  The air stream is filtered through a cloth filter, 10, and then scrubbed with water and steam, 11.  The fine catalyst recovered from the scrubber is not returned to the masher for inclusion in the final catalyst because oxidation tends to convert carbonate to oxide which renders reduction more difficult and because convert carbonate to oxide which renders reduction more difficult and because dirt an dust are apt to become concentrated in this fraction.  This material, which only represents 0.1-0.2% of the whole, is therefore sent to the catalyst regeneration plant and treated as spent catalyst.  The fines from the screens and cloth filters are returned to the masher.  The particles above 3mm. pass to a further set of screen, 12, where rotating arms force the granules through the 3mm.  screen.  The dust and fines from this process are returned to the masher and the 1-3 mm. grade is blended with the stream from the original screens, 9, and is bagged for transport to the adjacent reduction plant via a telfer conveyor.  The bulk density of the granules is 320-350 gm./1. The maximum daily output of the plant is equivalent to 4 metric tons of cobalt.

The flow-scheme for catalyst reduction is shown in Fig. 2, which is based on a diagram prepared by Dr. Herbke, (12).  The reduction vessel comprises a central compartment of square cross-section containing the catalyst in a bed 30-35 cm. deep and 2.1 m2 in area, with top and bottom fittings in the form of truncated pyramids.  the sheet-iron grill (15 cm. cubes) is placed on top of the catalyst bed and sinks into it to a depth of about 10 cm.  This device serves to break up the gas stream entering the top of the reduction vessel.  One charge of catalyst weighs 200-250 kg. and occupies 800 1.  There are 6 of these reduction vessels.

The reduction is effected by passing downwards through the catalyst a rapid stream of ammonia synthesis gas (75% H2 + 25% N2) which is preheated to 460 C. in a tubular heater fired with coke-oven gas.  The effluent gas is re-heated to 300C. and the CO2 present (ca. 2 gm/m3) is converted to methane by passage through a bed of synthesis catalyst in another reduction vessel.  The gas is then cooled, dried by refrigeration and passage through silica gel and is returned to the preheater for the reduction where it meets fresh make-up gas.

The recycle and fresh gas enter the reducer at about 7000 m3/hr. (S.V. 8800).  The period of reduction varies from 40 to 60 minutes, depending on the exact gas velocity which varies according to the number of reduction vessels in use.  The temperature is varied according to the time and gas velocity as shown below:

Time Temperature at Inlet to Reducer Gas Velocity
40 mins. 435 8000 m3/hr.
60 mins. 428 6000 m3/hr.

The temperature is controlled to within 2C.  After reduction is complete the catalyst is cooled to room temperature in a stream of nitrogen, the nitrogen then displaced by CO2 and the contents of the reduction vessel discharged into a Kbel by removing the top cover plate and inverting the vessel.  The vessels were so balanced that this inversion could be accomplished by a hand operated mechanism.  A total of 16 reduction charges are required to fill/one Kbel.  In the reduction process 50-60% of the cobalt is reduced to metal.  If over 60% is so reduced, a less suitable catalyst results.  The extent of the reduction is determined by measuring the volume of hydrogen evolved when the reduced catalyst is treated with acid.  The exact reduction conditions were said to depend on the nature of the kieselguhr used in the catalyst preparation.  Dense catalysts caused difficulties in the reduction.

In partial disagreement with the above, Pichler (3) stated that the reduction of the catalyst is carried out at 365C.  A temperature of 400C. is too high for a catalyst having a Co:kieselguhr ratio of 1:1, while with higher kieselguhr contents a temperature of 400C. and even higher can be employed.  A reduction time of 4-5 hours is suitable for a hydrogen rate of 2 1.per g. of Co in the catalyst.  When longer reduction times are employed, the formation of methane during the synthesis will increase.

According to Alberts (2) the optimum conditions for cobalt catalyst reduction are very difficult to establish  Generally speaking, the lower the reduction temperature the better, but lower temperatures required longer times.  A difference of 10C in the temperature of reduction made an important difference in the activity and life of the catalyst in the over--a difference visible during the first three days of operation.

Return to Table of Contents