HAROLD V. ATWELL – "DEVELOPMENT OF IRON CATALYST PROCESSES"

H. V. Atwell. Regarding iron catalysts, the only information which I picked up personally was from Lurgi at Frankfort, and that has already been reported. To supplement that for this occasion, I have gone through what microfilms were already available to us and picked up some information that others may not have the opportunity to find yet. That survey was also incomplete because so few of the microfilms have been distributed and the reading of them is so tedious. A very good summary of information covered by other C.I.O.S. reports was prepared by Dr. Storch for the Fischer-Tropsch Process Summary Report.

For a considerable length of time there was a period of competition in which I. G. and Lurgi particularly were trying to develop something to match the Ruhrchemie cobalt catalyst operation. Following that came a period of cooperation for the purpose of conserving cobalt and to develop a commercial process which could be used in the war emergency. The I. G. people were apparently the first to work on the iron catalysts, beginning in about 1935. They developed the Michael process of Duftschmidt in which hydrocarbon oil circulated as a coolant over a fixed bed of catalyst. The details for this are given in the Fischer-Tropsch Summary Report No. 1, page 25 ff.

In the microfilms we find references to the I. G. precipitated iron catalyst containing iron, copper, potassium, magnesium and kieselguhr, called their "standard" catalyst. In Reel 26, Item "N", there is a report covering the components of this I. G. precipitated catalyst in considerable detail. The I. G. Record subsequently pointed out that precipitated iron catalysts give better yields of paraffins, whereas the fused or sintered iron catalysts give predominantly olefins and alcohols. The addition of copper, according to I. G., was believed to serve chiefly to stabilize the hexagonal Fe₂C iron carbide which they said was previously unidentified. Another interesting development reported in a memorandum by I. G. describes the use of calcium fluoride as the promoter for the purpose of making lower boiling hydrocarbons.

The information obtained from Lurgi concerned chiefly the recycling process with a fixed bed of iron catalysts under middle pressure conditions. They evidently explored that process extensively on pilot plant scale and got what they considered very satisfactory results. The catalyst which they regarded as best was reported to us in detail: composition 100 iron, 25 copper, 9 alumina, 2 potassium oxide, 30 kieselguhr. From such a catalyst they claimed to have obtained yields of 170 grams per cubic meter including the liquefied gas. Dr. Herbert of Lurgi stated that so far as he knew the so-called fluidized technique had never been tried for the Fischer-Tropsch process.

The T. A. C. Summary Report contained information regarding the research at Kaiser Wilhelm Institute where a catalyst containing 2 to 3 percent of copper, ½ to 1 percent of alkali, and the balance iron, was represented as the best they could develop. From information obtained from I. G. and Lurgi, and Ruhrchemie, it looks as if this a rather incomplete disclosure. They stated that the use of copper was chiefly to give a catalyst of reproducible properties and the introduction of alkali was chiefly valuable for lowering the molecular weight of the product.  They also stressed the fact that the reduction conditions for these catalyst were extremely important, perhaps as important as the initial compositions themselves. Papers by Roelen, 12 and 13 September, 1940, (T.O.M. Reel 33, Bag 3440, Item VI) describe Ruhrchemie’s fundamental work on iron catalysts as of that date in great detail. These include a very comprehensive survey of the effect of variables in synthesis with iron catalysts. In July 1942 (T.O.M. Reel 35, Bag 3452, Dec. 1) Ruhrchemie reported recycle operations over an iron catalyst which were already giving very good operation. Numerous microfilm documents reported Ruhrchemie’s interest in iron catalysts effective at lower temperatures. It is probably generally realized that all of the early work with iron catalysts were don’t at 250 degrees Centigrade and higher, which apparently was a natural characteristic of such a catalyst and it was not until the importance of replacing cobalt with iron in commercial unites became acute that the limitation imposed by this high temperature was brought out. It was about the same time that intensive cooperative work began. In all the commercial units, cooling was by indirect heat transfer to boiling water, and the temperature was controlled by the pressure on the boiler. If the required catalyst operating temperature went up the temperature of the coolant and the pressure on the cooling zone went up accordingly, and mechanical limitations became very severe.

Extensive cooperative research work was directed at the problem of finding an iron catalyst that was more active and could be operated closer to the 200 degrees C. which the cobalt catalysts required. Considerable success had been obtained at Ruhrchemie, using temperatures of 220 to 225 degrees C., although I found no record that they had actually gotten as low as 200 degrees. Regarding the Ruhrchemie’s activities along this line, we have found in the microfilm a rather complete record of the work during the year 1944, in T.A.C. Reels 33 and 34, Document 31. On Reel 42, Document 22, there will be found very complete records of experiments 650 to 808 covering the period from 28 January 1944, to 19 October 1944, and presumably representing about the last work Ruhrchemie did on iron catalyst, and also some of the best results they obtained. I have picked out a couple of catalysts (Experiments 709 and 735) which appeared to be among the best they developed, and I will include in the record a translation of the preparation and testing of those particular catalysts.

W. C. Schroeder. Do you want to submit that as written or read?

H. V. Atwell. I don’t think we need to read them here, we can provide copies if desired.

W. C. Schroeder. All right, fine.

H. V. Atwell. I might add that practically all of this work was confined to catalysts of this general composition: 100 parts of iron, 5 parts of copper, 10 parts of calcium oxide, 35 parts kieselguhr, and a small amount of potassium as a promoter. The variations were in the kieselguhr concentration, potassium concentration, and the methods of preparation. From a number of these catalysts they obtained results that were quite comparable with the best that they had obtained previously with cobalt, except for the fact that the operation temperature was still a little higher.

Early in May 1943 the Reichsamt arranged for cooperative tests on iron catalysts from several different companies to see if something could be found that could be put into commercial use to meet the prospective shortage of cobalt. The tests continued through the better part of 1943, with the final report being issued in July 1944. The actual operations were conducted by Brabag, and Brabag issued a report summarizing the data. I have copies a summary table from that report with some supplemental data from other sources and will insert it in the record. The catalysts were submitted by Kaiser Wilhelm Institute, Lurgi, Brabag, I. G., and Rheinpreussen. The CO conversions ranged from 70 to 88 percent and the yields on any basis were almost as good as would be expected from cobalt. Unfortunately, we do not have analyses of all of these catalysts, but we have what appears to be the analyses of the Brabag catalyst containing iron, copper, and zinc; and the Rheinpreussen catalyst containing iron, copper, and calcium. We were told by Lurgi that I. G. submitted their synthetic ammonia catalyst for these test but we have not yet found any I. G. statement to that effect. As far as we know the iron catalysts had not been put in commercial use in Germany but the results of these cooperative Brabag tests indicate they were very close to that point.

There is a conspicuous lack of information regarding C-ray patterns, surface areas, etc., which we are so much interested in regarding catalysts in general. If anyone has run across information of that sort, I would like to have the reference.

W. C. Schroeder. Thank you, Mr. Atwell, we would like very much to have all that data you have in table form.

I wonder if I might briefly summarize some of the statements made in this discussion. One was that the Germans had not made any extensive study of the physical properties of the catalyst, and they rather thought that such a study would not yield valuable results that could be readily coordinated with the use of the catalyst in practice. Dr. Faragher expressed that viewpoint, and also express the opinion that he tended to agree with it. Mr. Atwell, on the other hand, voiced the opinion that perhaps such studies were desirable if they were pursued in the right direction. It was also stated that Martin from Ruhrchemie had expressed the belief that the cobalt catalyst was about as desirable as the iron catalyst and that it had as many advantageous properties as the iron catalyst. He, however, seemed to be the only man in Germany who expressed that opinion, and even in Ruhrchemie there was a definite belief that the iron catalyst had possibilities.

W. A. Horne. May I say here, if there was a difference of opinion in Ruhrchemie, they were still going ahead with the constructed plan for manufacture of the iron and calcium catalyst as Mr. Atwell mentioned for use in Fischer-Tropsch.

A. R. Powell. I think that Mr. Horne can give a good description of catalyst manufacture.

W. A. Horne. I can’t give it in a half hour, but I will say that I think it comes about from the fact that the iron apparently has a shorter life than these precipitated types and a shorter life because it could not be regenerated. It, therefore, requires considerable larger volumes of preparation. From the same production, there would be a more expensive operation on iron, but weight for weight, the iron is cheaper.

H. V. Atwell. As the Chairman mentioned on that point, Ruhrchemie had run one iron catalyst for 1100 hours with no sign of decline in activity, and the Lurgi people, from their experience with life tests, have made the prediction that their iron catalyst would operate for a year. I think the picture with respect to life in general is pretty favorable to the iron catalyst.

Paul K. Kuhne. I would like to ask Dr. Storch if he agrees with the expressed opinion about the lack of correlation between physical properties of the catalysts, and the catalyst’s activity or effectiveness.

H. H. Storch. I suppose that as far as surface area measurements are concerned, we have had very little success. We have had some success with activated carbon monoxide studies. So far as the magnetic studies are concerned, we are quite enthusiastic about them.

COPY

Ruhrchemie Oberhausen-Holten

8 June 1944

Experiment 709

100 Fe, 5 Cu, 10 CaO, 30 Kieselguhr (Soda precipitation) + 3 per cent K₂CO₃

Impregnation

Preparation of Catalyst PN-31 in Catalyst Plant –

Hot nitrate solution (1.8 Kg Fe and corresponding amounts of Cu and Ga in 50 liters H20) is run into a boiling (Kochende) soda solution (6.4 Kg in 50 liters H20) is run into a boiling (Kochende) soda solution (6.4 Kg soda in 50 1. water) 540 gm.  Kieselguhr is added. After stirring for a short time the mixture is filtered with suction. The cake is washed ten times with 10 liters of "condensate." The cake is then slurried with 100 liters of water and again subjected to vacuum filtration. The cake is mixed with 572 cc. potash solution (95 g. K₂CO₃ per liter) in a kneading mixer for 30 minutes. The paste is spread on plates and dried at 110^o C. for 16 hours, and then screened through a 16 mm. Sieve.

Reduction in 6 Liter Reduction Chamber (R 69)

The reductor, previously filled with nitrogen, is charged with 6 liters of raw (catalyst) granules. A H2-N2 mixture (elsewhere stated to contain 75 per cent H₂-ed) at 300^oC. is passed through for 1 hour (35 cubic meters). The apparatus is purged with N₂, and the catalyst cooled with HS-N₂ stream. The catalyst is discharged in a N₂ atmosphere. Shrinkage: 14 per cent. Degree of reduction: 75 per cent (acetic acid method).

Testing of the Catalyst; Oven MR-5

The oven is charged with 2.6 kg of catalyst (about 5 liters), and the catalyst is wet with cetane. The pressure is raised to 10 atm with water gas and the oven is heated. Gas feed begins at 150^oC. and the temperature is gradually raised. At 200^o C. after 11 hours operation a conversion of 43 per cent is reached. (methane relationship = 8; ratio of CO:H₂ converted = 0.85). To obtain a steady conversion of 62 per cent the temperature must be raised gradually to 221^oC. During the period from 137 to 1113 hours of operation the conversion maintains an average value of 60 per cent. (Mv = 20, X = 0.73). The test was ended only because of the necessity of using the oven for another purpose. No decline in conversion was noted up to the end of the run. No difficulty was experienced in emptying the oven.

Appraisal of the Catalyst

Activity was high and steady (220^oC.; conversion = 60 per cent for 1000 hours) methane production high, but not harmful to the catalyst (Mv = 20). NO tendency to excessive methane production. Ratio of CO:H₂ consumption X = 0.75.

Note: The exact analysis of the raw (catalyst) granules was as follows:

100 Fe, 3.04 cu, 8.1 CaO, 35.4 Kieselguhr.

COPY

Oberhausen-Holton

14 Sept. 1944

Experiment 735

100 Fe, 5 Cu, 10 CaO Kieselguhr (Na₂CO₃ Precipitation)

Impregnated with 1.8 KOH per 100 Fe

Preparation of Catalyst PN-47

Hot nitrate solution (1.8 Kg Fe and the corresponding amounts of Cu and Ca in 50 liters water) is poured into boiling soda solution (6.2 Kg Na₂CO₃ in 50 liters water) PH value 8.8-9.0. 540 gms. Kieselguhr added. Filter with vacuum after brief stirring; wash with 100 liters water. Knead for 30 minutes with 370 cc. KOH solution (100 g. KOH per liter). Spread on plates and dry 12 hours at 100oC. Screen to (make) 15 mm. Granules.

Water Gas Treatment (Ovens 3 and 6)

Charge 5 liters of raw catalyst in oven 6 and 2 liters in oven 3. Heat for 5 hours in CO₂ at 250^oC. Pass through water gas for 24 hours 15 100 liters per hour per liter of catalyst. (Gas analysis given after 18 hours). Cool in H₂ Stream (20 liters per liter of catalyst) for 15 hours. Discharge in N₂ atmosphere into a container filled with CO₂.

Testing of the Catalyst in Oven MR 4 Recycle

Charge 2.90 Kg (about 5 liters) into the oven. Pressure to 10 atm with water gas and heat. Beginning at 100^oC. charge gas at 350 liters per hour. At 202^oC. after 24 hours of operation the conversion amounts to 42 per cent (Mv = 3.2; X = 1.49). By raising the temperature to 215^o C. a conversion of 62 per cent (Mv = 12; X = 1.24) is obtained for a long time. By increasing the temperature to 220^oC. a conversion of 68 per cent is reached (Mv = 14; X = 1.2). The experiment was ended after 740 hours. The paraffin was bright yellow.

Appraisal of the Catalyst

Activity exceptional. Methane formation high (Mv=14). Ratio of CO:H₂ consumption good (X = 1.2). Compare Exp. 738.

(Translator’s Note: Recycle ratio is not specified. Data for Exp. 738 are not found on this reel.)

W. C. Schroeder. The next paper will be on "Research at the Kaiser-Wilhelm Institute at Mulheim," by Vladimir Haensel.