Return to TOM Report Table of Contents

T.A.C Report A1ML-1

The OXO Process For Alcohol Manufacture From Olefins

Reported by:
H.M. Weir
Petroleum Administration for War

Information Supplied by the

U.S. Government Technical Oil Mission
Submitted for Distribution by the
Petroleum Administration for War
to the
Technical Advisory Committee
(A Subcommittee of the Technical Committee)
of the Petroleum Industry War Council

July 30, 1945

T.A.C Report A1ML-1

Index

(ew)

T.A.C Report A1ML-1

The OXO Process For Alcohol Manufacture From Olefins

A. Introduction

Early in 1945 a group of petroleum and fuel technologists from both England and the United States investigated the status of oil technology and research in Germany. This mission was carried out under the sponsorship of C.I.O.S. (Combined Intelligence Objectives Subcommittee) and with the cooperation of the British and American armies in the European theater of operations. Among the processes discussed with German personnel at appropriate plants and locations was the so-called OXO process for making higher alcohols from olefins. Through the present tense will be used frequently in this report, it should be understood, of course, that the process is not operating now and has not been since the termination of war.

B. Source and Character of the Data on the OXO Process

The information in this report was assembled from the following sources:

1. Interviews with personnel, for the most part at Ruhrchemie A.G. at Stekrade in the Ruhr, and at I.G. Farben A.G. at Leuna in Saxony. These interviews, as made by mission members, have been summarized for the most part by preliminary reports by E.B. Peck and V. Haensel.

2. Review of Microfilm No.14 of documents seized at the Leuna works of the I.G.

3. Notes and personal observations of the writer.

Microfilms of all seized documents were not available for the writing of this report. If and when finally received, the additional data contained therein can be summarized in an appendix hereto.

The present report covers the essential features of the process. Consistent citation of the source of many of the statements contained herein could not be made, so that names are seldom given. Where deemed useful, reference is made to Microfilm No.14 by page number; e.g., (14-650) for page stamped 00000650 on Microfilm No.14. Except where expressly stated otherwise, the opinions and facts as given are those of the German workers obtained either in interview or by abstracting the documents.

C. General Features

The OXO process is a two-stage process for producing alcohols from olefins, first, treating the mixture of olefins with carbon monoxide and hydrogen to form aldehydes, and second, reducing them by the use of commercial hydrogen. Operation may batch-wise or continuous.

A wide range of pressures may be used in both stages, but the order of 150 to 200 atmospheres is most common. Temperatures likewise may vary considerably, but the aldehyde producing stage is typically operated at about 140º C. and the second stage at about 180ºC.

A standard Fischer-Tropsch catalyst, containing cobalt, thorium oxide, magnesium oxide, and kieselguhr, is used in the first stage and may be carried through the second stage, though cheaper catalysts have been suggested and used in the second stage.

A wide enough range of olefines has been treated, using pure substances or mixtures with or without inert diluents, to justify the statement that the process is applicable to all olefins from the chemical standpoint only. Practical application of the process in Germany was directed mainly to processing olefin mixtures in the range of C11 to C17 carbon atoms. Such fractions were cut directly from the primary standard Fischer synthesis liquids or were cut from the product of low-temperature cracking of Fischer waxes. Selection of the raw material was guided by the desire to produce mixtures of alcohols in the range of C12 to C18 which after sulphonation, gave effective detergent mixtures. Plasticizers were also made by applying the OXO process to C7-C10 olefin mixtures and subsequently forming the pthalic acid ester of the alcohols produced.

D. Development of the OXO Process

The OXO process was developed as a result of joint efforts and sponsorship of Ruhrchemie A.G. and I.G. Farben-industrie. The process was discovered by Ruhrchemie's Dr.Ruelen, or was at least actively studied by him at an early date.

A company called the OXO Gesellschaft g.m.b.H was formed with offices at Oberhausen-Holten. The date of organization of the company and the distribution of the stock is not known, but the latter was presumably held jointly by the above two companies, with some participation by Henkel and Cie. g.m.b.H, Dusseldorf. In any case, the latter company was destined to make the sulphonated products from the alcohols manufactured by OXO Gesellschaft.

In a document dated October 12, 1942 (14-667), the patent situation was summarized as follows:

1 U.S. Patent 2,327,066, August 17, 1943, Otto Ruelen inventor, unassigned, but seized by U.S. Alien Property Custodian presumably corresponds to R 103, 362. Application date April 15, 1939, refers to the filing in D. Reich September 19, 1938.

The German documents are not available for review. In developing the process to a commercial conclusion, both Ruhrchemie and I.G. Farben took an active part, starting about 1940. C.I.O.S. mission members gathered from the interviews, and the review of seized documents confirms, that fact that representatives of both companies were in substantial agreement as to the efficacy and utility of the reactions for the production of higher alcohols.

Laboratory work was done not only at Ruhrchemie and Leuna but also at Ludwigshafen laboratories of I.G. Naturally, these operations were done batch-wise. The Ruhrchemie the process batch-wise on a semi-plant scale, but workers at Leuna translated the process into a continuous semi-plant operation or a small plant operation. The OXO Gesellchaft equipment at Ruhrchemie was also to be operated batch-wise. It appears, however, that continuous operation was finally looked upon by both Ruhrchemie and Leuna as most effective, and that the reason for batch-wise large-scale operation was that original designs could not be changed in wartime. Interview elicited the information that the annual capacity of the OXO plant at Ruhrchemie was nominally 8,000 to 10,000 metric tons of product. However, the plant was very slow in building, due to the war, low priority, etc., and the information through interview was to the effect that the plant had not been operated as a whole at the end of the war, though a few test runs had been made on part of the equipment.

In the semi-plant equipment at Leuna, about 1-1/4 metric tons per day of OXO alcohol were produced during the first four months of 1944. In May only 7 tons were produced and thereafter no production was made. From the tenor of both documents and interviews, it seems certain that lack of raw materials, difficulties due to bombing, poor priority for construction, and other matters reflecting war conditions, rather than on the confidence in the process as a chemical tool, caused the cessation of semi-plant operations and retarded progress of construction by the OXO Gesellschaft.

E. Character Of the Process From The Chemical Standpoint

Raw Materials

a. Olefins

The reactions of the OXO process can be applied to olefins having from 3 to at least 20 carbon atoms. Aromatic olefins may also be processed. Pure olefins may be used or a mixture, both either with, or without, dilution by inerts such as paraffins. The yield is not substantially affected by dilution or admixture and is of the order 0f 90% alcohol at least when aliphatic olefins constitute the feed stock. When diolefins are present, reaction occurs at only one of the double bonds and the yields are poor.

Apparently, almost any source of olefins may be drawn upon for charge materials.

The alcohols which result will have one more carbon atom than the olefin charged, and the choice of charge stocks for the commercial purposes is thus conditioned by the nature of the product desired.

The sources which are mainly considered by the Germans were the primary aliphatic olefins produced by the Fischer Tropsch process, and the olefins with double bond at the end of the straight chain, which could be obtained from mild cracking conditions applied to the F.T. paraffin-gatsch. However, some work was done upon the olefin mixtures which might be obtained from low-temperature carbonization of brown coal. The two former sources, particularly the first, were finally looked upon most favorably as the source of raw material for alcohols for manufacture of detergents.

At the time in 1940 when the production of 40,000 tons per year of alcohols in C10 to C20 range was being discussed, the production of olefins by the Fischer process was visualized by Dr. Martin of Ruhrchemie, according to the scheme diagrammed on the next page.

It will be noted that the Fischer process, using an iron catalyst at 20 atmospheres, was to furnish the basic raw materials. Fractionation of the heavy oil direct from the process (155º - 330º boiling range) was to furnish slightly more than half of the raw materials, and cracking the wax at low temperature and pressure was to supply the rest of the raw material. The mixture was presumed, on the basis of some actual experiments, to consist of about 70% olefins and 30% paraffins.

In the document (14-480) it appears that Dr. Martin of Ruhrchemie presented the above scheme with the statement that an 85% yield of olefins could be made from cracking F.T. paraffin with only 5% loss. He proposed to crack, thermally, at about 250º C. Long chain olefins with the double bonds at the end were stated to be mainly present in the product, but as cracking temperature was raised, the double bonds tended to wander from the end of the chain toward the middle. At a cracking temperature as high as 350º C., the double bonds are almost entirely in the middle of the chain.

While this early procedure envisioned the application of iron catalyst in the synthesis operations, we now know that the application of iron catalysts on a large scale was never attained during the war period. The trend of the development, as read from the documents, was definitely toward the use of primary olefins from the normal Fischer process using cobalt catalysts.

In a document from Leuna dated March 11, 1943 (14-772), the following statements in regard to suitable charge stocks are made: (Translation is approximate and is condensed.)

"Olefins, made according to various synthesis procedures, and a thermally cracked product from Fischer Gatsch were analytically investigated. The degree of the transformation of the olefin content by help of the OXO reaction was determined. In the experiments, products boiling in the range of the detergent alcohol were given greatest weight. The most important results are gathered together below.

"(1) The Oxierung of the olefins, i.e., treatment by the OXO process, succeeded with all materials examined, given more than 95% yield.

"(2) The 'cobalt products' (made by Fischer Tropsch with cobalt catalyst) are substantially more uniform than 'iron products' (made by iron catalysts and Fischer Tropsch) and the former are composed substantially only of olefins, paraffins, and alcohols.

"(3) The 'iron products' contain, in amounts worth mentioning, not only olefins but also acids, esters, and aldehydes. When the OXO process is applied to such materials, they are considerably reduced in complexity of composition."

The above is cited at this point to indicate that as late as the first part of 1943, experiments were still going on to find the proper source of olefins for the process. Consideration was always directed toward obtaining raw olefins of the proper chain length, and it is indicated from the literature that little difficulty was ever experienced in carrying out the OXO process itself. Indeed, laboratory work was said to have been done successfully on such varied materials as:

In interview, the writer noted that the reaction was said to go when elements other than oxygen are in the compound with a double bond. For example, compounds containing nitrogen, sulphur (?), chlorine, etc., may be processed.

b. Gaseous Reagents

Water gas is entirely satisfactory for the first stage of the process, after proper purification to reduce the sulphur content to the order of 2 milligrams per CBM or less. While a ratio of one volume carbon monoxide to one volume of hydrogen is theoretically required for the reaction, this ratio need not be closely maintained, since a large excess of gas is always employed to direct the reaction to completion. The attention to sulphur reduction in the gases is to prevent the well-known poisoning of the catalyst by sulphur. It is absolutely essential to satisfactory carrying out of the process.

Commercial hydrogen also processed to insure low sulphur content is satisfactory agent for the second stage of the process.

c. Catalyst

"Standard" Fischer-Tropsch catalyst generally prepared by Ruhrchemie at Sterkrade was used in most of the experimental work on the OXO process. "Standard" Fischer-Tropsch catalyst generally prepared by Ruhrchemie at Sterkrade was used in most of the experimental work on the OXO process. "Standard" cobalt catalyst was stated in interviews concerning Fischer-Tropsch operation to be 100 parts cobalt, 5 parts thoria, 8 parts magnesia, and from 180 to 200 kieselguhr. While it was normal to keep the iron content of the kieselguhr less than 1%, the documents indicate that I.G. had operated the OXO process successfully with some cobalt catalyst made by them and stated to contain up to 5% iron. The documents reviewed for this report contain no precise information on the method of manufacturing any catalyst, and reference should be had to the report on the Fischer-Tropsch process itself for information in this regard. The standard cobalt catalyst was preferably ground for better contact with the reactants, but this was not always done according to the documents.

The cobalt of the catalyst is partly converted to carbonyl, (Co(CO)₄) in the first stage and this dissolves in the liquid product. Subsequent hydrogenation breaks up the carbonyl leaving the cobalt in a catalytically active form.

F.Chemisty of the Process

The typical reaction which occurs during the process may be symbolized in general terms in the formula below, where R is a radical.

1st Stage

2nd Stage

The migration of the double bond which always occurs to a significant extent simultaneously with aldehyde formation may be indicated as a reaction preliminary to a first stage, thus:

alpha olefins--------------olefins with "middle" bond, i.e.,

CH₃(CH₂) x CH = CH₂------CH₃ (CH₂) x-y. CH = CH (CH₂)y CH₃

After rearrangement, as portrayed, reaction at the double bond in the first stage may follow either of the courses indicated above. The higher the temperature, usually the greater the rearrangement.

While the U.S. Patent mentions Ketone formation and this was alluded to be in interviews the documents reviewed fail to be specific on this point. A viscous oil ("Dickol") stated to be the result of some aldolization and Ketone Reaction, is mentioned as a by-product of 1-10% yield when C7-C10 compounds are processed.

The rearrangement or isomerization of the double bond is an important feature of the reaction, as is the formation of alpha-methyl and indeed of alpha-alkyl branched products. In a rather comprehensive experiment with specially purified dodecylene as raw material (14-638, August 21, 1942), the I.G. laboratories came to the following conclusions:

1. Mixtures of isomeric aldehydes, alcohols, and corresponding acids are obtained even when pure olefin bond is used as raw material.

2. Using an end double bond olefin as raw material, one obtains about 60% branching in the OXO process.

3. The building of branched compounds is due to the fact that double bond isomerization occurs during the OXO reaction. This is caused by cobalt carbonyl.

4. Not only are alpha-methyl compounds formed, but, in general alpha-alkyl branched products are found. Because of this circumstance, the melting point of OXO acids made from the alcohols are substantially lower than the normal fatty acids with the same number of carbon atoms. The marked tendency of the soaps made from OXO alcohols to absorb water reflects this mixture of branched materials.

5. Isomerization and the OXO reaction proceed simultaneously, the latter probably with the greater velocity, so that the formation of branched products in the reaction forming aldehydes is not as great as might be surmised from the laboratory experiments on the isomerization of olefins in the presence of carbon monoxide--but in the absence of hydrogen.

6. Iron carbonyl also isomerizes olefins, but nickel carbonyl has no effect.

Table I ( the next page) gives the experimental conditions used which led to the above conclusions.

The experiment with iron carbonyl had a more basic incentive than appears from what has been said to date. Aside from the OXO process to produce alcohol, the so-called Synol process is necessary.

In brief, the Synol process consists in contacting synthesis, i.e., purified water gas as used in the normal Fischer process, with a specially reduced iron catalyst. A gas pressure of 20-25 atmospheres is maintained and a temperature of 185º-195ºC. is used. The Synol process produces alcohol directly and thus reverts in a sense to the early work of Fischer (1923-1925), or might be looked upon as the OXO process under slightly different conditions applied to nascent olefins in situ. The reaction requires very careful temperature control and a large ratio of recycle gas to fresh feed. Though the major yield is of alcohol in the carbon atom range of C4-C8, some higher molecular weight and technical comparisons were continually drawn by the Germans between the OXO process and the Synol process for making higher alcohols.

The above explanation is necessary to understand fully the importance of the experiments with iron carbonyl as set down in Table I. The conclusions of the German authors in respect to this phase of their experiments were:

7. The fact that iron carbonyl acts to isomerize olefins is of importance in considering the composition and course of this reaction to produce Synol alcohols. Synol alcohol tends to be practically unbranched but the olefins also produced in the Synol process are themselves of branched structure.

It remained to explain the reason for this phenomenon. The German authors thought that in the Synol process alcohols were formed first, after which a dehydrogenation occurred to produce the Synol olefins. These olefins were branched due probably, in view of the teachings of these experiments, to the isomerizing effect of iron carbonyl.

It is important to note that "substantial quantities" of cobalt carbonyl are formed in the first stage of the OXO process. This is dissolved in the reaction mixture, but subsequent hydrogenation "completely destroys' the carbonyl and restores it to the active catalytic condition " (14-524). The continuous process is particularly advantageous in respect to catalyst conservation, therefore.

The aldehydes which are present at the end of the first step of the process can be separated and marketed as such. Indeed, at an early period in the development, production of aldehydes was contemplated from both batch and continuous operations. However, it was found that to react the olefins completely in the first stage, a sufficiently drastic treatment with the water-gas mixture must be employed so that considerable conversion of aldehyde to alcohol is actually obtained in the first stage. The prospect of getting aldehyde as an end product became less attractive to the Germans commercially, as they learned more of the nature of the reaction, including the solution of cobalt compound in the aldehydes just mentioned.

The OXO process is to be looked upon primarily as one for the production of alcohols, and it is not particularly suited to the production of aldehydes. In fact, definite statements were made in interview that when aldehydes were desired, it was probably better to produce the alcohol by the OXO process, then isolate same and use known methods to oxidize it back to the wanted aldehyde.

G. Properties of OXO Products

The main interest in Germany in the OXO process from the commercial standpoint during the years of 1940 onwards was apparently for the production of wetting and washing agents, obtainable by sulphonating alcohol. It was concluded rather early in the development that the mixture constituting OXO alcohols was better than any of the individual alcohols when converted to washing and wetting agents. Further, that the products were better than the small fraction of similar-number-carbon atom alcohols which could be obtained from the Synol process. Nevertheless, the files contained many comparisons, from one standpoint or another, of the OXO alcohols and the Synol alcohols for washing powder purposes.

Typical of the experimental work done upon the OXO products from the standpoint of their suitability for wetting and washing agents is the discussion contained in a letter from I.G. Farben to I.G. Wolfen (14-659, October 12,1942). Extracts of this letter are given here to indicate the type of product made by OXO process. Samples of OXO sulfates were being sent to the I.G. Wolfen with a request for their examination, and enlisting their aid in having the samples further tested by others. To explain the nature of the samples, the following information was given in the letter. The samples were of washing powder made of the OXO alcohols in the range from C12 to C18. The nature of the OXO alcohols before sulfonation is shown in the table below.

OXO Alcohols

The actual composition of the alcohols whose crude inspections are given above was determined as follows:

To obtain these alcohol cuts, the following procedure was used:
Low-temperature thermal cracking of paraffingotsch produced olefins having the following boiling ranges (after fractionation).

The "Oxierung" of these cuts was carried out at 130º C. and 200 atmospheres water-gas pressure in the presence of F.T. cobalt catalyst made by Ruhrchemie. Reduction of the resulting aldehydes was carried out at 180º C. and 200 atmospheres with hydrogen. These were sulfonated with chlorosulfphonic acid "in the ordinary way". Methanol and sodium hydroxide were added to separate the unsulphonated paraffins and other unreacted materials. Physical decantation followed by benzine extrection was employed to discard the unreacted materials. The purified solution of "soaps" in methyl alcohol so obtained was then evaporated and mixed with sodium sulfate, after which the mixture was dried on a drum drier.

The tests on the OXO products obtained by the procedure first described using the raw materials originating as just indicated, were as follows:

Properties and Appearance of Oxosulfates I, II, and III

The resulting oxosulfates were all fine white powders forming colorless and clear solutions when 5% was dissolved in water at room temperature. Since the solution was slightly alkaline to lachmus, a few drops of sulfuric acid were added to the "stock solution".

Stability in Storage

All of the products took up very little water and remained free-flowing powders, when stored for one week at 20ºC. and 65% relative humidity. This was deemed satisfactory.

Wetability

All samples were stated to be excellent, though the figures given are not understood, since the test is not described.

Foaming Properties

Foam Volume, when tested by a liter graduate and sieve plate holding the solution at 35ºC.,was described as good as "Ludwigshafen 387" or "LU 387" and better than "Sekurit", two powders in the market.

Washing Properties

When tested upon both cotton and wool which had been artificially soiled with natural dirt under various specified conditions, the conclusions were that the oxosulfates were substantially as good as the standard "LU 387" or "Sekurit". In a general summary of the properties of the oxosulfates so obtained, the following was stated:

Oxosulfates met requirements of storage stability, wetting power and foam propensity. They are very good both for wool and cotton washing, including "fine washing". It is further concluded that the products should have some of the C12 alcohols, i.e., sulfates with it, since the cleaning properties of the powders with wool or cotton are thereby enhanced.

Plasticizers-Esters

Aside from washing powders, the OXO process was used or planned for use by Germans to make phthalic acid ester of alcohols of 7-10 carbon atoms which was alleged to be an excellent "plasticizer". A fragmentary reference to the use of OXO alcohols for plasticizer is contained in a letter of September 17, 1942 (14-651). It seems, according to this communication, that beginning in April, 1943, Ruhrbenzin planned to deliver, at a regular rate, 6,000 yearly tons of a Kogsin 1 fraction, cut to include 7 to 10 carbon-atom-molecules and containing at least 35% olefin.

"This amount will be sent to Leuna and put through the OXO process, including the reduction. Crude products will then go to Schkapou (site of the rubber manufacturing plant owned by I.G. and near Leuna) and will there be esterified by means of the steam conversion process (the so-called azeotrope process), whereby in place of the usual benzol-xyol mixture to drive off the water the heptane-decane mixture would itself serve the same purpose.

"The separated mixture (about 4,000 tons per year) will be taken back to Leuna and should be sent to the motor fuel section of the plant, but this material could be divided between Holten and Leuna. The palatinols (i.e., phthalic acid esters) will be finished up at Schkapou."

Further reference to phthalic ester production will be found in the next section.

H. The Process From The Engineering Standpoint

Unfortunately, the microfilm reel #14 now available does not depict all of the documents which were originally seized. Some of the unavailable documents are believed to give considerable information on the engineering features of the process. Only fragmentary information can be given here, mostly taken from the preliminary reports.

As previously stated, the OXO Gesellschaft's plant in Ruhrchemie was to operate batch-wise, but the continuous process was more or less agreed upon as most advantageous. Consequently, it will be described in detail later in this report. It is however, of some interest to note the date and kind of information which was being obtained in semi-technical scale at Leuna for the benefit of the design of redesign of larger equipment at the OXO Gesellschaft. The information which was being obtained in semi-technical scale at Leuna for the benefit of the design or redesign of larger equipment at the OXO Gesellschaft. The information is given in about the same detail as appears in the seized documents, since it might serve as a guide for orientation experiment by others on the OXO process.

Semi-Technical Scale Experiments (14-514 to 05516) Letter dated February 5, 1942.

To see what was required for the larger scale work ( at Ruhrchemie), Leuna used the following apparatus in a test reported on above date.

Vertical Reactor 200 mm ID, length 8000 mm. High pressure tube. Cooling surface total 360 cm2/liter of contents comprising welded jacket surface, 210 cm2/liter and tubular surface (tube size not stated) of 150 cm2/liter of contents. Cooling method not clearly stated, but by inference consisted in releasing pressure and thus allowing vaporization of water in jacket and tubes. Temperature measurement made at 0.5, 1.0, and 4.0 meters above bottom flange by thermocouples.

Method Used

(Translation of German follows.) Reactor filled by injecting "Maische" (i.e., suspension of catalyst in liquid) from below against 200 atm. pressure of gas (synthesis gas in plant network). 140 litres introduced cold, then heated by jacket steam to reaction temperature. Gas pressure then reduced to 150 atm. and regulated thereafter by hand operation of valve. Temperature controlled by steam release in part and by steam condensation in condenser above apparatus. Gas flow through reactor held at 60 m3 /hr. or 190 normal liters per cm2 free cross section and hour. Gas heated to temperature before introduction "but effect was insignificant" in the batch operation. Gas flow maintained 65 minutes, but on account of stoppage in outlet line the vessel contents remained in reactor without gas flow (but presumably under temperature and pressure--translator's note) for another 45 minutes.

Charge

" Maische" or liquid suspension of C12 olefin fraction with proportion of 4 kilos Ruhrchemie Contact to 100 litres actual olefin. Ruhrchemie Contact to 100 liters actual olefin. The C12 fraction was sp. gr. 0.762, boiling range 206º-216º C., S.P.L 67%. Olefin content about 65%.

Gas Reagent.

Winkler water gas desulphurized over special "crude" (i.e., coke from low temperature carbonization) to 2 milligr. S per cu. meter gas. Composition of gas:

Conclusions from the experiments were noted for the benefit of Ruhrchemie to be:

1. It is possible to hinder, that is, control, the completion of the reaction by expanding steam out of the heating system. (Presumably the water jacket, etc., which served as heating system at first and later during reaction as cooling facility.)

2. One must carry out the expansion carefully, otherwise the reaction may cease in the bottom part of the vessel.

3. It was not possible to control the reaction by steam release so that temperature was equalized throughout. A difference of 15º C. was noted between the upper and lower thermocouples.

4. Due to cooling, the reaction ceased in the lower part of reaction chamber, and it was accordingly not possible to complete the reaction in the time demanded of 20 minutes (probably specified by someone, based on Ruhrchemie design).

5. In spite of the long time at temperature at 65 minutes with gas passing through the reaction mixture, the conversion was only 80% as determined by distillation. About 30% of the olefins which were transformed were in the form of "thick oil", i.e., were polymerized.

6. It appears, therefore, that the provisions made in the large equipment at Holten (i.e., Sterkrade) to regulate the reaction will indeed make it possible to control same, but it also appears that it will not be possible to hold constant temperature throughout the reactors; hence it will be difficult to find the proper position for temperature controllers......Also, we question if it will be possible to get the production which has been foreseen, especially in view of the difficulties of temperature equalization.

Batch Producing Plant-OXO Gesellschaft

The OXO Gesellschaft's batch process production plant at Ruhrchemie consisted of nine units, each with two reaction vessels for first and second stage. These reaction vessels were all fitted with tubular heat exchange surfaces for control of the reaction. The notes of interview give the following data, which should, however, be checked by drawings. These, it is believed, will be available later in microfilm form.

"Reactors-Intervolume 1.2 m3

Ø O.D. 570 mm
Ø I.D. 400 mm
Lenght-12,000 mm

"Reactor Cooling Surface

31 tubes 38 mm O.D.
Length-8,000 mm"

Operating Conditions

Temperature first stage - 120º-140º C.
Temperature second stage - 180º C.
Pressure both stages - 150 atms.

The operation of the equipment was simple, consisting in the water-gas treatment in Vessel No. 1 then pumping the reaction product containing catalyst into the second vessel, where hydrogenation was caused to take place. Both the water-gas and hydrogen were continuously circulated through the reaction vessels to insure good contact and to keep the solid catalyst suspended.

The exit hydrogen from the second stage in the recirculation circuit was passed through an iron catalyst to reduce the carbon monoxide content arising from solution of carbon monoxide in the liquid and decomposition of cobalt carbonyl from about 2.0% to 0.05%. The hydrogen purified in this way was pumped back into the second stage. No details of the operation are available, but microfilm Reel 14 contains a memorandum showing what was planned for the batch operation production of phthalic acid ester of a C7 to C10 cut. The work was to be done at Leuna by a method presumably quite similar to that projected for the large plant at Ruhrchemie.

The document is translated with rearrangement here because of interesting sidelights which it contains. File memorandum dated October 20, 1942 (14-668):

Sufficient alcohols in the middle range of chain lengths (say from C8 to C12) are not available for the manufacture of Palatinol as plasticizer for rubber in cables to withstand cold temperatures and for shoe soles."

Since the OXO plant in Holten cannot be put into operation before 1943, and since then the production of middle-length chain alcohols will be at the cost of reducing production of alcohols for washing media, we will consider in what follows whether it is possible in the OXO plant at Leuna to make about 2,000 tons per year of middle-length chain alcohols.

Kogasin I from Ruhrchemie will be cut so as to give the boiling range of 80º - 175º C. (C7-C10 hydrocarbons). The raw materials will be delivered to the plant at Leuna in a form and with a boiling range which makes it unnecessary to distill them.

Form of the OXO Process to be Used

The olefin raw material will be mixed with 3% to 4% of finely divided reduced cobalt catalyst, and the suspension then will be continuously treated first with water-gas and then with hydrogen at high pressure at at the required temperatures.

The 'Maische' or suspension will be expanded to atmospheric pressure after the first stage to remove the carbon monoxide before injecting into the hydrogenation stage. Filtration to separate the solid will be done after the hydrogenation stage. The catalyst will then be stirred up with the fresh olefin (for reuse).

The filtrate will generally be clear and of light yellow tone. It will contain the desired alcohols and also unreacted constituents, such as the paraffins, and up to 10% of the original olefins. Perhaps up to 10% of the alcohol formed will be present as higher boiling viscous oil obtained by aldolization and a ketone reaction.

The present picture of the esterification of phthalic acid to make Palatinol is as follows: in spite of the large amount of neutral oils, namely, from 65% by volume in the case of these alcohols from primary olefins, the esters formed are not lost in the neutral oil. Return of the neutral oil after separation of the water is not necessary, and the neutral oil acts very satisfactorily as water-transporting media (translator's note: Reference must be had to azeotropic removal of water in continuous esterification). An addition of 10% of the viscous oil to Palatinol does not act deleteriously.

According to the present viewpoint, the crude filtered product from the second stage of the OXO process can be sent through to esterification. If, for any reason, distillation of the crude product and a concentration of the alcohols are necessary, such distillation can be carried out in the existing distillation plant....."

Throughput Required of the Plant

On the assumption that the alcohol yield will be one-third of the product introduced, the required capacity of the plant will be 3/4 ton of charge stock per hour or 6,000 tons per year. This will mean that 1,050 liters/m of charge stock, sp.gr. 0.72, will have to be introduced into the reactors.

Requirements for the Reaction

Assuming an average molecular weight of 155, an olefin content of 35% for the charge stock, and the introduction of 750 kilograms of liquid/m, the OXO reaction will require per hour 58 m3 of carbon monoxide and 58 m3 of hydrogen. Both figures are reduced to 0ºC. -- 760mm Hg pressure -- as are all subsequent gas figures.

For hydrogenation of the aldehyde in the second stage, there will also be required 58 m3 of hydrogen.

The following amounts of gas will be dissolved from each of the circulating streams per hour: 42 m3 of CO + H2 - first stage, 26 m3 of hydrogen - second stage.

Accordingly the total usage of the two gases will be 160 m3/hr. of water-gas and 85 m3/hr. of hydrogen.

Heat of the Reaction

1. OXO stage- 100 kilogram calories per kilogram of liquid introduced, under the assumption that 40% of the total hydrogenation will occur in the OXO stage, hence 75,000 kilogram calories/hr. for the 750 kilograms introduced.

2. Hydrogenation stage-50 kilogram calories per kilogram of liquid introduced, hence 37,500 kilogram calories/hr. for the 750 kilograms of liquid introduced.

Methanization of the Carbon Monoxide to Purify the Hydrogen Circulating Stream (Second Stage)

Under the assumption that about one-third of the cobalt contained in the catalyst mass will be present as dissolved carbonyl after the OXO reaction, one calculates a content of bound carbon monoxide in the form of Co (CO)4 of 1/5 gram mole per kilogram of product, equals 4.5 liters carbon monoxide per kilogram of product. With the introduction of 750 kilograms of liquid, the methanization reaction, to form methane out of carbon monoxide, must therefore handle an amount of carbon monoxide equal to 3.4 N m3/hr. The heat quantities to be removed by the reaction will be 3,000 kilogram calories/m3 of carbon monoxide, equal to about 10,000 kilogram calories/hr. 0.8 liters of water will be produced per m3 of carbon monoxide; hence in our case 2.75 liters H2O/hr.

Translator's Note:

The very substantial degree in which the Fischer catalyst is formed and reformed in the reaction itself is evident from the above. It will be clear why the plan was to release the pressure from the products after the first OXO stage, thus freeing the mixture of dissolved gases before passing to the hydrogen stage. This materially reduces the load on the purification of circulation hydrogen by removal of carbon monoxide, as just mentioned. The manuscript then considers the amount of loss of product which will be involved in this release of pressure from the reactants.

Product Loss Through Release of Pressure on Products -- After First Stage

Not inconsiderable amount of product loss will result from the release of pressure on the product gases. These losses will be as follows (translator's note: conditioned by the vapor pressure, hence temperature at which the gas is released):

With product gases, from both stages, in the amount of 70 m3/hr., the product losses will be as follows, expressed as a function of temperature.

As may seen by the above tables, it proves to be necessary to extract liquids from the expansion gases, either by cooking the gases or by raising them through an active charcoal plant.

Catalyst Requirement

In a plant of the size of that in ME 458A (translator's note: this is a building number), a circulation volume of 4 m3 of liquid suspension of 'Maische' can be expected. To begin the experiment, catalyst in the amount of 120 kilograms is required for a concentration of 3% catalyst in the liquid. The catalyst will circulate about six times per day with an injection of about 1 m3/hr. of liquid suspension.

If one reckons according to present experience, with a catalyst application of 50 to 60 times, then the total amount of catalyst application of 50 to 60 times, then the total amount of catalyst must be renewed every ten days. Since, however, the catalyst will be renewed continuously a daily introduction of about 12 kilograms of catalyst will be necessary. As a safety measure, it will be advantageous to hold in storage an amount of catalyst equal to the requirements for two to three months. Accordingly, the amount of catalyst necessary will be as follows:

Store house-1 ton catalyst; For initial experiment-120 kilograms catalyst; Daily catalyst requirement-12 kilograms.

These data are on the basis that Ruhrchemie's Fischer synthesis catalyst will be used.

Reduction of Catalyst

(Translator's note: Apparently I.G. was planning to have unreduced catalyst delivered to them, since the following is stated:)

The reduction equipment in Building ME 458A can still be used for the reduction of catalyst. A milling plant will have to be provided for the catalyst.

Flow Sheet of the Continuous Equipment for the OXO Process

The continuous process developed by I.G. was described in interview by aid of a flow sheet given here as Figure 1. It is not clear from the combination of documents and notes of interview whether the I.G. actually had a plant in this form or would suggest this form of plant for a new installation. The writer may mention that it was often difficult to get at the facts in such cases, because bombing damage, rebuilding of plant with slight changes, projection of plant for a new installation. The writer may mention that it was often difficult to get at the facts in such cases, because bombing damage, rebuilding of plant with slight changes, projection of new equipment by engineers, etc., was often probably a more or less confused mass of facts in the minds of the interviewed individuals, themselves.,

In any case, the flow diagram, Figure 1, is more or less self-explanatory.

The notes of interviews state that the first reactor is identical with those used in the batch process (with inner coolers), while the second reactor of the same size does not have coolers but disc and doughnut baffles. The hydrogenation section is the same, except for a converter in the gas recycling line for converting the carbon monoxide to methane. This is claimed to be essential to avoid poisoning of the catalyst. Gas released from the products of the first stage, thus any special arrangement for liquid recovery from the released gas is avoided. The gas release tank and the gas scrubber are not shown in Figure 1.

Figure 2 shows another version of the process equipment. Reaction vessels are identical, each having an annular chamber for upflow of the liquid and catalyst suspension, assisted by gas, and an inner cylinder fitted with cooling tubes, down which the catalyst suspension moves against the rising flow of gases. The inner section was baffled to prevent excessive back-mixing of the catalyst and liquid reactants. Here again the scrubbing apparatus for the released gases is not shown. In interview it was stated that the results of this pilot plant indicated that a throughput of 3 vol. liq. per 1 vol. react. vessel per hour could be had, and still attain 95% reaction of the olefins.

While it is to be hoped that documents which will become available later will enable the design of the equipment to be given more precisely, the writer is of the opinion that the data already given, together with fundamental information available in handbooks, would enable a reasonable design of plant to be projected by competent engineers. It is, however, also obvious that any group planning to use the process in this country, hence planning to utilize different raw materials and catalysts, would necessarily have to build a background of semi-plant experience before a large plant could be constructed. Substantially no data are available at the present time on the materials of construction for these plants. It was definitely stated to the writer that the vessels in the process were copper-lined, and that no particular difficulty was experienced in pumping the catalyst suspension due to corrosion of pumps, glands, and the like. There is, however, no confirmation or contradiction of these points in the documents reviewed.

J. Economics of the Process

The following is copied directly from the preliminary report of Dr. E.B. Peck:

"The economics of the process as calculated by Ruhrchemie is shown below and indicates a cost of alcohols of 71 pfg. per kilo. I.G. corrected this estimate as shown to 77 pfg. per kilo. These costs are based on an olefine cost of 38 pfg./kg. and a capital cost of 6 million R.M. There is now 11 million R.M. invested in this plant by OXO Gesellschaft. The I.G. on the other hand estimate a cost of 60 pfg. per kilo. for the continuous process. Within the above limits the cost of these alcohols lies between 60 and 100 pfg. per kilo. for the continuous process. Within the above limits the cost of these alcohols lies between 60 and 100 pfg. per kilo which, with exchange at 5.2 R.M. per dollar, would be 5 to 9 cents per pound.

Microfilm pages 724-736 present under date of August 4, 1944, a series of assumptions and cost calculative respecting OXO alcohols and the cost of making washing powders thereof. Pages 733 and 734 clearly diagram the chemicals required, and the process steps in such washing powder manufacture. Pages 735-736 tabulate the cost of RM of the intermediate and final products. Interested readers will probably wish to refer to these pages, some of which are, however, not very legible.

In common with all other cost figures under German war conditions, these figures are probably of value only as a rough indication of the general order of costs. The absolute figures are perhaps of much less value than the ratio of the various figures given, one in respect to the other.