The analysis of fatty acid-containing compounds requires previously their hydrolysis or saponification, the separation of the non-acidic constituents (when necessary) and the liberation of the acids from the mixture of soap.

Frequently, the lipids are transesterified to produce esters in a single step.

Alkaline hydrolysis of glycerides

The process which involves the production of soap is commonly referred to as saponification.
Discovered and described in detail by Chevreul in 1823, the process is yet run nearly in its original form. In his famous work, Hilditch (1956) recommends for most natural fats saponifying 100 parts by weight of fat with a solution of 30 parts of KOH in 500 parts of alcohol (95-100%), boiling under reflux for 3 hours, followed by removal of most of the alcohol by distillation. Ethanol is generally satisfactory but other solvents have been recommended. A small amount of water must be present to affect rapid and complete saponification of glyceride oils.
Henriques R (Z angew Chem 1895, 8, 721; 1896, 9, 221) reported that fats can be saponified readily at room temperature. The process may be carried out by dissolving 260 g of KOH in 250 ml water. After some minutes, a liter of the oil to be saponified is added to this solution. Ethanol (10 ml) is then added. This process was applied to fats containing easily altered fatty acids (i.e. conjugated fatty acids, highly unsaturated).

Separation of unsaponifiable matter

After saponification, approximately half of the alcohol is removed under vacuum, the residue being diluted with water and the unsaponifiable matter (hydrocarbons, tocopherols, sterols, etc ) extracted by shaking with an organic solvent, diethyl ether or petroleum ether.

Recovery of acids from soaps

Fatty acids are liberated by the addition of mineral acid, usually a 10% excess of sulfuric acid or HCl. When very short-chain acids are present, the alcohol is removed prior acidification under vacuum with the addition of sufficient water to keep soaps in solution. If the amount of acids is important, the long-chain acids are separated from the aqueous medium by allowing the mixture to stand in the cold.

Hydrolysis of waxes

Many waxes can be saponified by dissolving 2-5 g of wax in 30 ml of alcohol and adding 50 ml of 0.5 M KOH and refluxing the mixture for 1 or more hours. For some wax (carnauba wax) it was recommended to use xylene instead of ethanol.

After their separation, fatty acids may be purified by one of these procedures :

DISTILLATION
SALT-SOLUBILITY METHODS
LOW-TEMPERATURE CRYSTALLISATION
LIQUID CHROMATOGRAPHY
GAS-LIQUID CHROMATOGRAPHY
UREA-FATTY ACID COMPLEXES

Fractional distillation at reduced pressures was initially used to separate mixtures of fatty acids and esters derived from natural fats. The early distillation apparatus were not very efficient. Thus, natural mixtures were first separated into chemically similar groups prior to distillation.

The oldest record of a distillation process dates about 3600 B.C.. It was found at Tape Gowra in Mesopotamia, the instrument was 48 cm high and 53 cm in largest diameter. The distillation pot had a capacity of about 40 liters and the distillate collecting ring held about 2 liters. It was probably used to make perfumes (Great Chemists, E Farber ed, Interscience, NY 1961). As a science, distillation and fractional distillation had to await the discovery of the physical laws of Dalton (1766-1844) and Raoult (1830-1901). After that, the science of distillation developed at an amazing rate.

Distillation apparatus made its appearance in the chemical laboratory in the early part of the 19^th century, according to Underwood AJV ( Trans Inst Chem Engrs (London) 1932, 10, 112-152). An anonymous publication (Ind Eng Chem, News Ed 1935, 13, 140) attributes the invention of the distillation column to Cellier-Blumenthal and Derosene in France; Thorpe JF (Thorpe's Dictionary of appl Chem, vol 1, 1937) attributes it to Coffey in England. These bubble-plate towers were used primarily for the commercial distillation of spirits. An early laboratory distillation unit used for the fractionation of fatty acid esters is illustrated in Reilly's book (Distillation, Methuen, London, 1936). A similar instrument and its operation are described by Hilditch (The constitution of natural fats, 1940, pp374-6).The earliest laboratory column were simply open tubes and spiral-type column was introduced by Warren in 1864 (Mem Am Acad Arts Sci, 1867, 9, 121). The period 1900-1930 was one during which marked advances were made. One of the most important was the development of highly efficient packed columns. Columns containing rotating fractionation sections were described later by Podbielniak in 1935. This device was modified with a thin metal strip rotates at high speeds (1000-2500 rpm) in the glass column (spinning-band column). Baker (Ind Eng Chem, Anal Ed 1940, 12, 468) described a 18-ft column, 0.6 cm in diameter with a separation efficiency equivalent to 70 theoretical plates. The head pressure was 1 mm Hg. This device was used by Privett OS et al (J Am Oil Chem Soc 1959, 36, 443) to distill several fractions of methyl esters of pork liver lipid. One sample of esters was separated into 38 distillate fractions with chain lengths from C14 to C22. 19 specific fatty acids could be identified and quantified (amounts from 0.2 to 35.4% plus traces of 8 other fatty acids, C10-C15, C17 and C19.

The history of fractional distillation of fats and oils may be divided roughly into 2 periods, one preceding and one following the year 1930. Laboratory vacuum distillation was not introduced until 1869 (Dittmar W et al., J Chem Soc 1869, 22, 446; Kekule A et al., Ber 1872, 5, 906). Between 1880 and 1903 Krafft carried out distillations of fatty acids, esters and aldehydes under high vacuum. Separations of pure palmitic and stearic acids for the synthesis of glycerides were reported in 1903 (Krafft F, Ber 1903, 36, 4336) and by Kreiser H et al. (Ber 1903, 36, 2766). A few years later (Haller A et al., Compt Rend 1906, 143, 803) reported the fractional distillation of the methyl esters derived from coconut oil, but without quantification. In 1906, Bull H (Ber 1906, 39, 3570) fractionated methyl esters from cod liver oil and isolated 9-eicosanoic acid. Elsdon GD calculated the composition of coconut oil and palm kernel oil from analytical data obtained by fractional distillation (Analyst 1913, 38, 8-11; 1914, 39, 78). Between 1920 and 1930, analytical fractionation of methyl esters was applied in determining the composition of several oils and fats : peanuts (Jamieson GS et al., J Am Chem Soc 1921, 43, 2696), sunflower (Jamieson GS et al., J Am Chem Soc 1922, 44, 2952), soybeans (Baughman WF et al., J Am Chem Soc 1922, 44, 2947), olives (Jamieson GS et al., J Oil Fat Ind 1925, 2, 40; 1927, 4, 63) etc. In 1923, Brown JB et al. (J Am Chem Soc 1923, 45, 1289) published results of investigations of the unsaturated acids from menhaden oil. The same technique was applied by Brown JB with beef brain lipids (J Biol Chem 1929, 83, 783; 1930, 89, 167).

Following 1930, the objective of ester distillation was to obtain fractions containing "no more than two adjacent homologous saturated and two unsaturated members "(Hilditch Biochem J 1934, 28, 779) and having chain lengths differing by two carbon atoms. Later, with the development of spinning-band columns the objective was to produce fractions of the same chain length irrespective of the degree of unsaturation. This new concept is exemplified by the isolation and identification of 32 compounds from wool wax (Weitkamp AW et al., J Am Chem Soc 1945, 67, 447).
Technical improvements led to molecular distillation of vitamin D (Bills CE et al., J Biol Chem 1938, 126, 241), esters from fish oil (Komori S et al., J Chem Soc Japan, Ind Chem Sect 1951, 54, 225), cholesterol (Hickman KC Ind Eng Chem 1940, 32, 1451) and tocopherols (Hickman Ind Eng Chem 1940, 32, 1451).
Farmer et al (J Soc Chem Ind London 1938, 57, 24) applied molecular distillation to the fractionation of methyl esters from fish oils. They were able to separate DHA in a high state of purity (> 99%). Schuwirth K (Z physiol Chem 1943, 277, 147) distilled the methyl esters of the unsaturated fatty acids of brain phospholipids.

Saturated and unsaturated fatty acids form salts with metallic ions, whose solubilities in water and organic solvents vary with the nature of the metallic ion and the chain length, degree of unsaturation, and other characteristics of the acid radicals.

The oldest and most widely used method of metallic salt separation depends on the differential solubility of the lead salts or soaps of fatty acids in ether. This method was introduced by Gusserow CA in 1828 (Arch Pharm 1828, 27, 153) and improved later by Varrentrapp F (Ann Chem Liebigs 1840, 35, 196).

One of the more important modifications was introduced by Twitchell (Ind Eng Chem 1921, 13, 806) who substituted ethanol for diethyl ether. Both methods were used for analytical purposes but they are preferable for preparative purpose. The solid or saturated acids are regenerated from the insoluble lead salts by boiling the latter with dilute HCl. After cooling, the cake of solid acids is separated and the aqueous layer is extracted with ether. Complete and sharp separation between saturated and unsaturated acids is impossible. The method is not reliable for the separation of mixtures containing unsaturated acids of chain length greater than C18 or saturated acids of chain length of C14 or shorter. It may be said that the method is applicable with vegetable oils, but not with the seed fats of the Palmae, the rapid drying oils, the cruciferous oils, castor oil, marine oils, and oils with trans-fatty acids.

We give below the procedure of "The precipitation of solid fatty acids with lead acetate in alcoholic solution" as it was written by Twitchell in 1921:

Crystallization of acids and esters from solvents at low temperatures was introduced in the late 1930's at the time that precise fractional distillation was being developed. It was widely applied for the separation of fatty acids and monoesters, and also for the separation of glycerides and other lipid substances. A review can be found from Brown JB et al. (Prog Chem fats and other lipids, vol 3 Pergamon 1955, pp 57-94) who first applied low-temperature crystallization for the separation of fatty acids.

The firt reports on the separation of fatty acids were made in 1937 (Brown JB et al., J Am Chem Soc 1937, 59, 3-6; 6-8). Previous attempts were made (Holde D et al., Ber 1901, 34, 2402) in separating the glycerides of olive oil.

Hilditch used extensively this method to fractionate glycerides of natural fats. Various modifications were made with respect to solvents, concentration of fats or glycerides, and temperatures. Low-temperature crystallization has been applied for about 25 years and as an alternative to the lead salt method. It was useful to prepare concentrates of specific unsaturated fatty acids from raw stocks.
A review can be found from Markley KS (Fatty acids, Markley ed., Interscience Pub, 1964, part 3, 1983-2123).

Michael Tswett (1872-1920) is credited with developing the first concepts and techniques of chromatography as they are now known. Thus, the separation of plant pigments gave rise to the first chromatographic studies. Tswett was aware that his results possessed significance beyond the mere separation of pigments, as is evident from his first communication. Chemists were not attracted by methods which enable to handle only milligrams of materials.

Kuhn R et al. (Z Physiol Chem 1931, 197, 141) rediscovered chromatography in the course of studies on carotenoids. The subsequent applications of chromatography with fatty acids and lipids in general will wait until 1950 to find some references in this field.

The first chromatographic separations of fatty acids were concerned with short-chain acids and were reported by Smith EL (Biochem J 1942, 36, 22) and Elsden SR (Biochem J 1946, 40, 252). Adsorption chromatography was used by White MF et al. (J Am Chem Soc 1948, 70, 4269) to purify methyl arachidonate on alumina, by Herb SF (J Am Oil Chem Soc 1951, 28, 505-7) to separate several unsaturated fatty acids on silicic acid as it is now currently used.

The reversed polarity partition chromatography was first used by Boldingh J (Experientia 1948, 4, 270) to separate fatty acids on sheets of filter paper impregnated with latex to support the non-polar phase. Powdered latex was later used for column partition chromatography (Boldingh J, Rec Trav Chim 1950, 69, 247), and aqueous acetone or methanol as mobile phase to separate long-chain fatty acids. Besides rubber, or glyceride oils, silicone oil or synthetic polymers were used in paper and column chromatography.

Reversed polarity partition was used by Wittenberg JB (Biochem J 1957, 65, 42) to separate the even-numbered C6-C12 acids. Hirsch J (Fed Proc 1959, 18, 246 abst) used acetone + water for separation of C4-C20 acids, Stoffel W et al (J Lipid Res 1960, 1, 139) used this technique to separate C18 and C20 acids from menhaden oil.

The first separation using reversed phase column was reported by Boldingh J (Rec Trav Chim 1950, 69, 247) who used latex powder saturated with benzene or peanut oil as the stationary phases and methanol + acetone or methanol + water as mobile solvents.

The history of the gas-liquid chromatography of fatty acids dates from 1952, when James AT et al. (Biochem J 1952, 50, 679) published a description of separations of free fatty acids from 1 to 12 carbon atoms by this technique. Later, they discovered that a previous methylation of fatty acids improved their volatility and their separation at temperatures below 250°C (James AT et al., Biochem J 1956, 63, 144-152). These classic papers contain yet valid descriptions of the principles of gas chromatography and may serve as an introduction for those needing basic information on that important technique. It is noticeable that the first publication of James and Martin was not only the first description of gas-liquid chromatography of fatty acids, but corresponds also to the first description of that technique. These authors used Apiezon stopcock grease (silicon grease) as the stationary phase incorporated into a solid powdered support (Celite 545) and packed into a glass column. At that time it was not possible to analyze a complex mixture since the use of increasing temperatures was foreseen but impossible to apply with the detector included in the instruments. A major advance was the introduction by Orr CA et al. (J Am Chem Soc 1958, 80, 249) of a liquid polyester phase as stationary phase able to conveniently separate saturated and unsaturated fatty acids. Later, the use of capillary columns with very thin films of various polar phases enabled to profundly improve fatty acid analyses (Horning EC et al., Anal Chem 1963, 35, 526).

UREA-FATTY ACID COMPLEXES

The first description of the urea-fatty acid complexes by Bengen F in 1940 (German Patent Appl. OZ 12438, march 18, 1940) revolutionized fat chemistry for about 20 years. For a long time this technique was efficient in preparing relatively purified fatty acids from fats and vegetal oils. During his research on milk, Bengen showed that urea forms in water, methanol, and ethanol well-defined crystalline inclusion compounds with straight-chain compounds but not with branched-chain or cyclic compounds (Bengen F, Experientia 1949, 5, 200). He generalized this phenomenon to hydrocarbons, fatty acids, esters, alcohols, aldehydes, and ketones. In addition to Bengen's group, several researches were conducted by petroleum companies on the separation of linear and branched-chain hydrocarbons. These inclusion compounds are combinations of two or more molecules, one of which is contained within the crystalline framework of the other. The stability of urea complexes increases with increasing chain length. Furthermore, it was observed that the greater the degree of unsaturation the greater the deviation in behavior from the corresponding normal saturated compound. Thus, at a constant chain length, saturated fatty acids form urea complexes preferentially to monounsaturated, mono- preferentially to di-unsaturated, and so on. With this approach, oleic (Swern D et al., J Am Oil Chem Soc 1952, 29, 614), linoleic and linolenic acids (Swern D et al., J Am Oil Chem Soc 1953, 30, 5) were isolated and purified (80-95%) from natural sources on a Kg basis.

As an example, we give below a brief description of the procedure used by Swern and co-workers for preparing linoleic acid from a natural oil.

To a solution of 880 g of urea in 2.7 l of hot methanol, 1 Kg of safflower oil acids (containing 78 % linoleic acid) is rapidly added and the mixture is heated and stirred until complete dissolution. After cooling overnight at room temperature, the precipitate is filtered off and discarded. The filtrate is evaporated and several volumes of warm water containing 30 ml of 6 M HCl is added. The insoluble oil is dissolved in hexane and washed several times with water to remove urea. Evaporation of the solvent yields 790 g of linoleic acid. The authors distillated under vacuum this oil and obtained 610 g of 95% linoleic acid (and 5 % oleic acid).

A simple procedure which can be easily used at a laboratory level is described in this site.

The formation of urea complexes was also used to reduce the free fatty acid content of natural glycerides to 1 % or less. Moreover, Schlenk H et al. (J Am Chem Soc 1950, 72, 5001) were the first to separate unoxidized from oxidized fatty acids by precipitation of the former as urea complexes. Similarly, oxidatively produced polymers may be separated as they do not form urea complexes. Abu-Nasr AM et al. (J Am Oil Chem Soc 1954, 31, 16) applied the same process to enrich several fish oils in polyunsaturated fatty acids. This was the basis of the purification of methyl eicosapentaenoate and docosahexaenoate. Extensive data from the literature can be found in the Schlenk's review (Progress in the chemistry of fats and other lipids, Vol 2, Pergamon Press, NY, 1954, pp. 243-267)

Aylward F et al. (J Appl Chem 1957, 7, 583) have shown that saturated 1-monoglycerides were able to form urea complexes, whereas 2-monoglycerides do not.

As before the development of gas-liquid chromatography (1960) the chemical identification of long-chain fatty acids remained a complex problem, it is worthy of note that the formation of urea complexes and their dissociation temperature were used to define the length of the carbon chain (Knight HB et al. Anal Chem 1952, 24, 1331). A review on the physical description of urea inclusion compounds may be found in the paper of Hayes DG (Inform 2002, 13, 781).

A remarkable property of urea complexes of unsaturated fatty acids is their resistance toward autoxidation (Schlenk H et al. J Am Oil Chem Soc 1950, 72, 5001).

Owing to its low cost, low toxicity, and simplicity, urea fractionation has been most frequently used for the purification of fatty acids from fish oil (Ratnayake WMN et al., Fat Sci Technol 1988, 90, 381) and from several vegetal oils (blackcurrant oil : Traitler H et al., J. Am Oil Chem Soc 1988, 65, 755; borage oil : Shimada Y et al., J Am Oil Chem Soc 1988, 75, 1539; linseed oil : Swern D et al., J Am Oil Chem Soc 1953,  30, 5; rapeseed oil : Hayes DG et al., J Am Oil Chem Soc 1998, 75, 1403; cyclopropene fatty acids : Fehling E et al., J. Am Oil Chem Soc 1998, 75, 1757; conjugated fatty acids : Ma DWL et al., J Am Oil Chem Soc 1999, 76, 729).
A review of the main applications of urea fractionation may be found in a review by Hayes DG (Inform 2002, 13, 832).

 )