US2340104A

US2340104A - Process fob making mixtures of

Info

Publication number: US2340104A
Authority: US
Publication date: 1944-01-25
Anticipated expiration: 1961-01-25

Description

Patented Jan. 25, 1944 PROCESS FOR MAKING MIXTURES OF- UNSATURATED FATTY ACIDS John B. Brown. Columbus, Ohio, assirnor to Ohio State University Research Foundation, Columbus, Ohio, a corporation of Ohio' No Drawing. Application October 4, 1940,

Serial No. 359,768

6 Claims. (Cl. 260-419) My invention relates to a process for making, from fish oils, mixtures of highly unsaturated fatty acids suitable for use in the manufacture of improved drying oils.

When the term fish oil is used in this speclfle cation it is intended to designate all marine oils including marine animal oils such as whale oil.

An object of my invention is to separate from a mixture of fish oil fatty acids, such as is obtainable by hydrolyzing or by saponifying and acidulating a fish oil, a fraction composed largely of the entire series of the highly unsaturated fatty acids of the original mixture relatively free from saturated fatty acids, and also, if desired, relatively free from the more insoluble of the moderately unsaturated fatty acids.=v

Another object is to separate from such a mixture of fish oil fatty acids a fraction or fractions composed largely of the saturated and single double-bond acids suitable for use in soap manufacture.

In general the object is to fractionate fish oil fatty acids predominantly with respect to unsaturation rather than with respect to molecular weight, and in particular to accomplish the above mentioned fractionations at low temperatures, thus avoiding the decomposition and polymerization which accompany fractionation by distillation at high temperature.

Among other products produced is an improved drying oil or composition, suitable for use in the manufacture of paint, vamlsh, enamel, and products of like nature, also a basic intermediate for the synthesis of drying compositions.

Heretofore in the manufacture of drying oils for use in paints and varnishes naturally occurring oils principally of vegetable origin have been employed, with or without treatment designed to modify their drying characteristics; also synthetic drying oils have been employed.

Mixed fatty acids derived from naturally occurringglycerides have, for example, been sub- Jected to fractional distillation to separate the higher molecular weight acids from those of lower molecular weight, and from the resulting fractions the one composed in highest degree of unsaturated fatty acids has been selected and has been reesterified by heating with glycerine to produce synthetic drying oil. Fractional distillation of fatty acids depends on differences in vapor pressure of fatty acids of diflferent molecular weights, such differences depending principally on the number of carbon atoms in the molecule and only to a negligible degree on unsaturation or drying properties. In order successfully to separate highly unsaturated fatty acids suitable for synthesis of drying oils by fractionally distilling a mixture of saturated, moderately unsaturated, and highly unsaturated fatty acids it is necessary to choose a source of mixed fatty acids such that the higher molecular weight acids'of the mixture (or conceivably the lower molecular weight acids of the mixture) include a large proportion of the highly unsaturated and relatively little of the saturated and moderately unsaturat ed acids. No naturally occurring fatty oil is ideal for this purpose. It is at best quite impossible by such a method to obtain a fraction including substantially all of the highly unsaturated acids of the original mixture which is substantially free from all saturated fatty acids and is at least relatively free from moderately unsaturated acids. Furthermore, the high temperature, often 200 to 250 C., required in carrying out this distillation process unavoidably results in a certain amount of decomposition and polymerization, resulting in contamination of the distillate and of the still residue with the products of these undesired reactions.

In the Journal of the American Chemical Society (1937, p. 3 and p. 6; 1938, p. 54 and p. 2734) the present applicant and others published several articles dealing with the isolation of oleic, linoleic, and other fatty acids by crystallization procedures.

The process now disclosed for producing a fatty acid mixture suitable for use in making improved drying oil is a crystallization process which employs as its raw material a relatively low priced and readily available material, namely, fish oil. Furthermore, it employs a method of separating from fish oil mixed fatty acids a fraction composed essentially of the highly unsaturated acids (each containing two or more double bonds) of the original mixture, employing the degree of unsaturation as the principal controlling factor in accomplishing the separation, regardless of molecular weight. Thus I am able to'obtain a higher yield of the unsaturated acids available in the raw material than is possible by fractional distillation. The employment of damagingly high temperatures is avoided. The portion of the original fatty acid mixture not included in the highly unsaturated product may if desired be segregated as one fraction containing both saturated (no double bonds) and moderately unsaturated (one double bond) acids, or .it may if desired be separated as two or more fractions, at least one of these composed largely of saturated and another largely of moderately unsaturated acids. This last named fraction may alternatively, if desired, be included with the highly unsaturated fraction.

In carrying out my process I separate the fish oil mixed fatty acids from raw fish oil by any convenient saponification method, for example by hydrolysis with water with or without a catalyst, or by caustic saponification followed by acidulation of the resulting soaps. I usually then dry the mixed fatty acids, and I then dissolve them in an appropriate amount of a suitable solvent. The preferred concentration depends upon the degree of separation desired and the efficiency of the means employed to separate the crystals formed in the process from the remaining solution. In general. the concentration must be low enough to permit separating the bulk of the solvent containing the bulk of the uncrystallized fatty acids from the mass of crystals. Relatively low concentrations. for example about five to ten grams of fatty acids per 100 milliliters of solvent, generally result in somewhat sharper fractionation under similar conditions than relatively high concentrations. Under suitable conditions, however, very g od results are attainable at relatively hi h concentrations.

In this specification all concentrations unless otherwise stated are expressed in terms of grams of fatty acids per 100 milliliters of solvent, which corresponds roughly to pounds of fatty acids per 12 U. 8. gallons of solvent.

Solvents which I have found to be suitable for my process include acetone. commercial hexane. petroleum ether, methanol, ethanol, amylene, and ethylene dichloride. In general. any easily distillable organic solvents which have a sufficiently low freezing point and are inert to higher fatty acids may be employed, although some such solvents are found to have a more selective action than others in my fractionation procedure. Of the solvents named herein. methanol and ethanol are less desirable than the others in that they have a slight tendency to form methyl and ethyl esters of the fatty acids. When a saturated hydrocarbon is used as a solvent the. products are as a rule about as light or even lighter in color than the original fish oil, whereas when acetone is used the products are usually darker than the original oil. Under equivalent conditions acetone has been found capable of segregating highly unsaturated fatty acids somewhat more effectively than the hydrocarbon solvents herein mentioned.

Having thus prepared a solution of the mixed fatty acids in a suitable solvent, I then cool this solution gradually and preferably with mild agitation to a temperature low enough to throw out of solution the greater part of the saturated fatty acids, which I have found to be very much less soluble than the unsaturated acids in the solvent-fatty acid solution at temperatures ranging from C. downward. When a solution of typical menhaden oil mixed fatty acids of concentration (as measured in the aforementioned terms) in a hydrocarbon solvent is cooled appreciable quantities of saturated fatty acids crystallize out of solution before the temperature reaches l0 C., and by the time it has reached C. about 70% or more of its saturated acids have. precipitated. At 40 C. no significant proportion of saturated fatty acids remains in solution. The high unsaturated acids do not begin to crystallize to any great extent until the temperature reaches at least -40 C. or lower. The greater part of the moderately unsaturated acids come out of solution between about 20 C. and 50 C.

Two principal considerations govern the choice of the crystallization and filtration temperature: the degree of separation desired (which may conveniently be measured in terms of the iodine value of the fatty-acids of the more unsaturated fraction) and the physical problem of separating the crystals from the solution. If the total amount of crystals to be removed is very bulky, or if available means for removing the crystals are such that a relatively large amount of liquor will remain occluded in the mass of crystals, or if it is desired to separate the original fatty acid mixture into three different fractions rather than two, then it is desirable to remove two crops of crystals, the second at a lower temperature than the first. I have separated crystal fractions at as low a temperature as that obtainable with "dry ice," about -78 C. In general, an increasing degree of unsaturation of the fatty acids in the filtrate is obtainable as the fractionation temperature is lowered, within the range 0 to about 80 C., with a correspondingly lower yield of fatty acids in the filtrate.

I prefer long slow cooling of'the solution over a period of about 6 to 16 hours rather than fast cooling, because-of the formation of easily filterable crystals and also more complete precipitation of all fatty acids that are insoluble at the lowest temperature reached. More rapid cooling may be employed, however, with fair satisfaction.

Having cooled the fatty acid solution to the desired temperature, and preferably having kept it at this temperature for some time, I next separate the remaining solution from the crystals, usually by filtration and preferably at about the same temperature as that of the solution, and preferably with the employment of suction on the receiver or pressure on the material being filtered to hasten the filtration and to minimize the amount of occluded solution left with the crystals.

The crystals may then, if desired, be washed with fresh solvent or with a relatively dilute solution of fatty acids in solvent, the washing fluid having previously been cooled to the original crystallization temperature or lower. By thus washing the crystals mother liquor is removed and a higher yield of filtrate acids may be obtained. The mixture of crystals and wash liquor is filtered, and this second filtrate may conveniently be employed as solvent for another batch of fish oil mixed fatty acids. Thus a single portion of solvent may be used both for washing one crop of crystals and for the crystallization of another crop from another batch of mixed acids.

If'two crops of crystals are to be removed from the fatty acid solution the filtrate from the first crystallization step is cooled to an even lower temperature, such that after filtering oi! the second crop of crystals the fatty acid mixture remaining in the filtrate will have the desired degree of unsaturation as measured by its iodine value or polybromide number or by other suitable means. The second crop of crystals may, if desired, be washed in a manner similar to the optional washing step previously described for the first crop.

After thus fractionating the original fatty acid mixture into two or three (or more if desired) fractions, I then warm each fraction separately and remove the solvent by evaporation. The solvent may if desired be recovered by condensing its vapors and collecting the condensate in a suitable receiver.

If the fraction composed largely of highly unsaturated fattyacids is to be converted into triglycerides, I accomplish this step by heating this fraction with glycerine according to well known procedure. As an example of this esterification operation, 88 grams of 99% glycerine were added to 1000 grams of a highly unsaturated fish oil fatty acid fraction having a mean molecular weight of 300 (this being sufficient fatty acids to form triglycerides and to leave an excess of about of the fatty acids unesterifled) in a 3 liter flask which was provided with a stopper through which passed a gas inlet tubeextending nearly to the bottom and a gas outlet tube leading from near the top. Nitrogen was bubbled through the liquids in the flask to produce agitation, to exclude air, and to sweep out moisture formed during the course of the reaction. The flask and its contents were gradually heated during a two hour period to 250 C. or slightly higher, held at this temperature for an hour, and then cooled to room temperature. The reaction mixture was then treated with an aqueous solution of sodium hydroxide to remove free fatty acids. I have found that the resulting mixture of triglycerides has excellent drying properties when applied, as a varnish, in a thin film to the surface of a solid. It dries faster and forms a tougher and more homogeneous film than either a representative sample of the natural fish oil from which the unsaturated fatty acids were derived, or a reesterified sample of the original mixed fatty acids separated from this oil. Furthermore, I find that A by a suitable choice of fractionating conditions a mixture of unsaturated fish oil fatty acids may be segregated whose mixed triglycerides have drying properties about equal to or even superior to linseed oil.

The fraction or fractions composed principally of saturated and moderately unsaturated fatty acids are suitable for use in soap manufacture, as well as for other purposes for which fatty acids are used.

The following examples afford a more detailed understanding of my process:

EXAMPLE 1 One hundred grams of menhaden oil mixed fatty acids, having an iodine value of 196.8 and a polybromide number of 63.0, were dissolved in two liters of commercial hexane, to produce a solution of concentration 5, and this solution was cooled in about 30 minutes from room temperature to l0 C. The crystals which formed were removed by suction filtration and were not washed. The filtrate was further cooled, slowly and with stirring, to C., and a second'crop of crystals was removed by suction filtration. In a similar manner three more crops of crystals were removed, at C., C. and 65" C. The five crystal fractions and the final filtrate fraction were weighed separately, the solvent was removed from each by evaporation, and the solvent-free fractions were again weighed and were then tested for iodine value and in some cases for polybromide number. The results are summarized in Table I. Column 1 shows the weights of the total fractions after evaporation of solvent; column 2 shows the weights of true precipitates (crystals) corrected for weight of fatty acids dissolved in occluded mother liquor, calculated on the assumption that the composition of said occluded mother liquor is the same as that of the corresponding filtrate; column 3 shows the iodine au ios values determined on representative samples of th fractions after evaporation of solvent; column 4 shows the iodine values of the true precipitates, each calculated by correcting the iodine value of the corresponding fraction after evaporation of solvent for the calculated iodine value Table I Wt. in gins.- Iodine values Polyggmide Fraction Total T ue Total Total True fracfract lracp t tion pp tion pp tion p (4) Example 1:

l0 precipitate" 6. 4.8 54. 5 l7. 5 20 precipitate. 14 ll. 8 37. 3 4. 6 30 precipitate" 5 4. 6 23. 4 6. 5 40" precipitate. 4. 5 4. 0 36. 3 9. O 65" precipitate. 4. 5 3. 8 86. 3 64. 0 l0. 0 0 65 filtrate 63. 5 269. 86. l Example 2:

40" filtrate 54. 2 261. 7 Example 3:

40 filtrate 35.3 269.5

In Example 1 the true precipitates of the first four fractions were composed mostly of saturated fatty acids, as is shown by the fact that the iodine values of the precipitated crystals of these fractions (it being understood that all references to iodine values of precipitated crystals mean iodine values of true precipitates, after correcting for iodine value due to mother liquor contamination) were quite low in comparison with the iodine values of unsaturated fatty acids having but one double bond (i. e., 75 to 100, depending upon length of carbon chain). In contrast to these the fifth fraction contained substantial amounts of unsaturated acids as is shown by the sharp and substantial rise in the iodine value of these last precipitated crystals as compared with the average iodine value of all crystals previously precipitated. The final filtrate contained a 63.5%

. yield of fatty acids, predominantly highly unsaturated, having an iodine value and a polybromide number each about 37% higher than that of the original fish oil mixed fatty acids.

EXAMPLES 2 AND 3 solution was cooled progressively and several crystal fractions were removed as the temperature was lowered to 40 C., the first crystals being removed at about 0 C. and the last at 40" C. The yields and iodine values of the final filtrates after evaporation of solvent are shown in Table I.

EXAMPLES 4, 5, 6, 7, arm 8 In each of these five examples portions of the same menhaden oil fatty acids as were used in Examples 1-3 were dissolved in petroleum ether to produce a solution of the concentration shown in Table II. Each or these solutions wa cooled in about 25 to 30 minutes to the temperature shown in the table, and filtered to remove the precipitated crystals. In each case the filtration temperature was about 20 C. The yields and analytical results shown in the table refer to the filtrate fractions after evaporation of solvent.

Table II Concontra- Iodine value Temperature Exsmnrs 9 AND 10 In each of these two examples portions of the same menhaden oil fatt acids as were used in ram 111 Iodine Solvent Yield value Example No.

Percent 52. 9 55. 3

. Acetone... Methanol.

EXAMPLES 11, 12, 13 m 14 In Examples 11 and 12 the fatty acids from a different lot of menhaden oil were used. This mixture of fatty acids had an iodine value of 181.7.

In Example 13 the mixed fatty acids from still another lot of menhaden oil were used. These had an iodine value of 191.7 and a polybromide number of 57.0.

In Example 14 herring oil mixed fatty acids were used, these having an iodine value of 137.6 and a polybromide number of 32.6.

In each of these four examples portions of the mixed fatty acids were dissolved in the solvent shown in Table IV to produce a solution of the 5 concentration-shown, the solution was cooled to the temperature or temperatures shown, and one, two, or three crops of crystals were removed, one for each temperature shown. In each case the assouos was 15 A mixture of highly unsaturated fish oil fatty acid fractions from several previous fractional crystallizations, each of these fractions being composed of the filtrate fatty acids from a 40' C. crystallization, was dissolved in commercial hexane to produce a solution of concentration 15. This solution was fractionated twice, by crystallizing at 50' C. and then at -'70 0., removing crystals each time. A 83% yield of final filtrate fatty acids having an iodine value of 320 was obtained. This degree of unsaturation is far above the highest unsaturation obtainable by direct distillation.

The polybromide number, referred to in these examples, is a figure representing the per cent of ether insoluble bromides obtained by brominatq ing in cold ether a given quantity of a specimen of fatty acids, or by brominating a given quantity of the methyl esters of these fatty acids, according to the method described by Brown and Beal in the Journal of the American Chemical Society 45, p. 1291 (1923). The polybromide number. which is a value analogous to the hexabromide number commonly used in evaluating drying oils, is a semi-quantitative measure of the proportion of very highly unsaturated fatty acids in a sample, and consequently it serves as an index of the drying properties of glycerides made from a given fatty acid mixture or fraction. Saturated fatty acids, and also fatty acids having only one or two double bonds have zero polybromide numbers.

Inasmuch as the moderately unsaturated (one double bond) fatty acids of 22 to 16 carbon atoms have iodine values of about 75 to 100 but have zero polybromide numbers, whereas the very highly unsaturated fatty acids (three or more double bonds) have iodine values in excess of 270 and have polybromide numbers of about 105, it is possible from a knowledge of both of these values to make certain deductions regarding the type of unsaturated acids present in a mixture. The true precipitate in the fifth fraction of Example 1 in Table I, for example, (the precipitate) has an iodine value of 64 and a polybromide number of zero, indicating a predominace of moderately unsaturated fatty acids and an absence of veryhighly unsaturated fatty acids. In other experiments with the same menhaden oil we have found that fatty acids having a definite iodine value but zero polybromide number begin to precipitate from a solution of concentration 10 in commercial hexane between 20 C. and 40" C. This particular lot of menhaden oil contains a smaller proportion of moderately unsaturated fatty acid radicals than many fish 'oils,

When practicing my process in a preferred manner, as in Example 1, the concentration of the solution and the temperatures to which it is subcooiing was relatively rapid, and the filtration 60 .iected are so chosen as to precipitate saturated was performed at 20 C. The yields and anafatty acids, substantially free from unsaturated lytical data as shown in thetable refer to the fatty acids, and the fractional crystallization is final filtrate fractions after evaporation of solcontinu d by lowering the temperature until'sub- .vent stantially all saturated fatty acids are precipi- Table IV i o Iodin ill-Fai $535353 $5 533 Temperatures Yield valu: 331 3;

12 55 so a i 24s 6 ll l llenhad n 01 2- P eu eth m 50:50 236:7 so e 0 12 40 36.6 262.8 230.4 Petroleum ether 10 30, 60, -70 21.7 297.4 91.6

ascends tated, as indicated by a sharp and substantial rise in the iodine value of the last crystals precipitatlng as compared with the average of all crystals precipitated up to that point. Beyond this point the crystallization may be further continued if a final filtrate fraction of very high iodine value is desired; however, the lowering of the temperature is preferably discontinued before any appreciable quantity of fatty acids having more than two double bonds have precipitated-i. e., before the precipitating crystals show a significant polybromide number. When, on the other hand, a very high yield is desired of a fatty acid fraction whose triglycerides have improved drying properties, but which does not have an iodine value in the highest range obtainable, crystallization may be discontinued when the iodine value of the fatty acid fraction in the final residual solution reaches a suitable value, say about 235, as in Examples and 12, or when the polybromide value of this residual solution attains a desired value.

A characteristic of the highly unsaturated fractions of fish oil fatty acids made by my method is that they may be so prepared as to have a polybromide number of 80.0 or higher, even when the polybromide number of the original fish oil mixed fatty acids is as low as 33. No other direct method of fractionating fish oil fatty acids heretofore described, to my knowledge, can produce a fraction having so high a polybromide number and containing highly unsaturated fatty acids having carbon chains of all lengths present in the original oil. A fraction having a polybromide number of over 80.0 and an iodine value of over 250 is relatively free from fatty acids whose glycerides have poor drying properties.

Fish oils have been shown to include unsaturated fatty acid radicals having carbon chains of 14, 16, 18, 20, 22, and 24 'carbon atoms. Highly unsaturated fish oil fatty acid fractions obtained by distillation contain only the higher members of this series and consist predominantly of C or C22 acids. The 014 and Cm acids, and most of the Cu; acids are carefully excluded because distillation fractions which include these lower members of the unsaturated series also include most' of the saturated acids of the original oil. A distinguishing characteristic of my unsaturated filtrate fractions is that unlike unsaturated distillation fractions they include Cis acids, and also 014 acids if the original fish oil contained C14 fatty acid radicals. In other words, my fractions include carbon chains of all lengths present in the original oil.

Besides menhaden oil and herring oil, sardine oil. whale oil and other commercially available marine oils are suitable raw materials for my process.

Having thus described my invention, what I claim and desire to secure by Letters Patent is:

1. In a process for making an improved drying composition from fish oil, the steps of fractionally crystallizlng a solution comprising one part by weight of fish oil fatty acids in at least three parts by weight of a suitable solvent by cooling said solution until substantially all saturated fatty acids and a significant proportion of the fatty acids having one double bond are precipitated; discontinuing said cooling before any appreciable proportion of fatty acids having more than two double bonds are precipitated; removing the precipitated crystals from the residual solution; and separating'the solvent from the fatty acids in said residual solution, whereby a fatty acid fraction whose triglycerides have improved drying properties is produced.

2. The process of claim 1 in which said solution is cooled until substantially all fatty acids having less than two double carbon bonds are precipitated.

3. The process of claim 1 in which the solvent is a volatile hydrocarbon solvent.

4. The process of claim 1 in which the solvent is acetone.

5. The process of claim 1 in which at least two crops of precipitated crystals are removed from the residual solution, each succeeding cropat a lower temperature than the preceding crop.

6. "In a process for making an improved drying composition from fish oil, the steps of fractionally crystalliing a solution comprising one part by weight of fish oil fatty acids in fromabout 3 to about 15 parts by weight of a suitable solvent by cooling said solution to a temperature, between 0 C. and C., which is sufficiently low to precipitate substantially all saturated fatty acids and a significant proportion of the fatty acids having one double bond, as indicated by a pronounced rise in the iodine value of the precipitating crystals as compared with the average iodine value of all crystals precipitated up to that point; discontinuing said cooling before any appreciable proportion of fatty acids having more than two double bonds are precipitated, as indicated by the polybromide number of the precipitated fatty acids; removing the precipitated crystals from the residual solution; and separating the solvent from the fatty acids in said residual solution, whereby a fatty acid fraction whose triglycerides have improved drying properties is produced.

. JOHN B. BROWN.