Chapter III · REFINING, BLEACHING, AND DEODORISING FATS AND OILS

Physical Methods—Washing, freezing, filtration, treatment with charcoal and fuller’s earth, steaming. Chemical methods—Caustic soda, sodium carbonate and silicate, alkaline earths, ozone, hydrosulphites, sodium bisulphite, organic peroxides.

Although, as was pointed out in Chapter II., too much stress cannot be laid upon the importance of using only the freshest materials of the best possible quality in the manufacture of edible fats, yet even these frequently require a certain amount of preliminary treatment in order to bleach, deodorise, or refine them and render them palatable for human food. Numerous processes, both physical and chemical, have been devised, and in many cases patented, for these different purposes, and the following is a summary of the more important ones.

Physical Methods.—Among the physical methods employed may be mentioned washing with hot water, the removal of suspended matter by settling or filtration, and of excess of stearin by subjection to low temperatures, bleaching by filtration through animal charcoal or fuller’s earth, and deodorising by injecting steam either at atmospheric pressure, in vacuo, or in presence of an indifferent gas.

The impurities in freshly expressed oil which are partly in suspension and partly in solution, consist chiefly of dirt, fragments of vegetable fibre, and mucilaginous and albuminous substances. A portion of them rapidly deposits when the oil is allowed to stand, and the upper liquid may then be drawn off from the sediment and subjected to filtration or to further refining processes.

In some cases a simple filtration, after standing for a short time, is sufficient to render the oil brilliant, but special treatment is necessary when a large proportion of albuminous matter is present, since such oils either pass through the filter without becoming bright, or if a closer filtering medium is used, the pores of the filter speedily become clogged.

Various methods are employed to coagulate or precipitate albuminous matters before filtration, such as dry heat, or the introduction of fine jets of steam, or the addition of a small quantity of insoluble powder (e.g. fuller’s earth or kieselguhr), which as it subsides attracts and carries down simultaneously the particles of the gum-like mucilage.

The formation of an insoluble precipitate within the oil answers the same purpose. Thus, in Linde’s process a small quantity of milk is introduced and the mixture heated so as to coagulate the casein, the subsidence of which removes at the same time the substances that cause turbidity in the oil.

Other substances, such as solutions of tannin, are used in the same way in refining oils for technical purposes, but are inadmissible in the case of edible oils.

In the removal of dissolved impurities, alkali solutions, milk of lime, or magnesia are reagents in common use, while dilute sulphuric acid is employed to clarify linseed and certain fish oils for industrial uses.

To facilitate the purification of oils by washing with water, Dubovitz has recently recommended the addition to the water of aluminium sulphate, in the proportion of about ¾ oz. to 220 gallons per degree of hardness. This forms with the lime in the water a bulky, colloidal precipitate, which serves to bleach and clarify the oil.

Removal of Stearin.—When certain oils are exposed to a low temperature they become turbid and in some cases give a white deposit. This consists of glycerides of the more solid fatty acids, and is known as “stearin,” and its formation is frequently regarded as objectionable, notwithstanding the fact that when the oil is gently heated the stearin is redissolved.

In the preparation of the best edible oils, therefore, a process of chilling followed by filtration is often employed in order to remove part of the stearin, and the oils thus treated may then be exposed to a low temperature without giving a further deposit.

Oils treated in this way are commonly known as “winter oils,” and those which will only keep brilliant at the ordinary summer temperature are termed “summer oils.”

The solid fat separated in this way from cotton-seed oil is known as cotton oil stearin, although it is quite free from stearic acid. A similar process is employed for the separation of cocoanut and palm-nut oils into their respective stearins and oleins (see Chocolate Fats).

Methods of Filtration.—The types of filter press used in the filtration of oils are very varied. A common form consists of a hydraulic press containing a series of communicating plates with rims raised so as to form a space into which filter cloths may be fitted.

In other forms of apparatus the oil is introduced from below into a chamber, and rises upwards through the filtering medium into a compartment in which a partial vacuum has been created. Or methods of centrifugal filtration may be employed, as in apparatus in which the oil is introduced into a revolving chamber with a perforated wall round which is wrapped a filtering cloth. The oil is flung against the wall of the revolving chamber in a fine state of division, and passes through the filtering medium into an outer chamber, whence it can be drawn off.

The materials used as filtering media include sand, kieselguhr, Spanish clay, fuller’s earth, animal charcoal, paper pulp, and a mixture of wool and vegetable fibres disintegrated into a pulp.

Chemical Methods.—The number of chemicals employed in refining oils and fats for edible purposes is necessarily very limited. It is of course very objectionable to employ any reagent which is poisonous, though the use of barium oxide has been patented by Rocca (Fr. Pat. 325,381, 1902), and processes involving the use of mineral acids are, in general, inadmissible, as they spoil the flavour of the oil. The chief reagents employed are caustic soda (sodium hydroxide), sodium carbonate, sodium silicate, calcium or magnesium oxide, ozone, hydrosulphites, formaldehyde-sulphoxylates, and organic peroxides.

Caustic Soda.—Of the alkaline refining process the treatment of cotton-seed oil with a solution of sodium hydroxide or potassium hydroxide is the best example. The oil is mechanically agitated with the alkali solution, which usually has a specific gravity of about 1·10, either with or without the aid of heat, and the mixture then allowed to stand until it separates into two layers, the lower of which contains a sediment of impurities. This is drawn off, and the treatment repeated, but this time with a more dilute solution of alkali, and finally the oil thus clarified is washed with water to remove the excess of alkali.

In this process of refining not only are the albuminous and resinous matters precipitated, but the free fatty acids in the oil are neutralised, and accordingly freshly prepared cotton-seed oil is almost neutral in its reaction, and has a bland taste. The treatment also removes a large proportion of the colouring matter separated from the seed in the expression of the oil, and changes the dark-brown colour of the crude product to a light-golden tint.

The residue left from the refining of cotton-seed and other oils is a thick deposit containing the impurities in a concentrated form, together with a considerable proportion of oil; such residues are known as “foots,” and are utilised in the manufacture of soap.

The amount of caustic soda required depends on the degree of acidity of the oil, which it should be just sufficient to neutralise. The acidity of the oil is therefore first determined, as described in Chapter VIII., and the quantity of caustic soda calculated which will neutralise the given bulk of oil to be treated. The following is a more detailed account of the process:—

The calculated quantity of alkali is dissolved in water, the solution diluted to 12 or 15° Tw. (8° or 10° B.), and one-third of it added, either in a fine stream, or through a sprinkler, to the oil contained in a steam-jacketed tank. The mixture is now heated first to 100° F., and then gradually to 120° F., the whole being well agitated mechanically, or by blowing a current of air through a pipe inserted to the bottom of the tank.

After about fifteen minutes the agitation is stopped, and the oil allowed to rest for some time, preferably overnight, to allow the soap and impurities to settle down to the bottom, whence they may be drawn off. This treatment is then repeated a second and third time, with the same quantity of caustic soda solution, but usually of weaker strength, and in exactly the same manner as just described, after which a clear, yellow oil should be obtained.

The agitation with air must not be unduly prolonged, as this tends to oxidise the oil, raising its specific gravity and refractive index, and also injuring its flavour.

Treatment with caustic soda solution is also frequently employed for refining other vegetable oils, notably cocoanut oil. In all cases the principle is the same, viz., combination of the alkali with the free fatty acids to form soap, which on settling carries down with it colouring matter and other impurities. Only a weak solution, of say 12° Tw. (8° B.), should be used, and the quantity added should not be more than sufficient to neutralise the free acid, since otherwise some of the neutral oil may be saponified.

Sodium Carbonate.—This may be employed instead of caustic soda to neutralise the free fatty acids of an oil or fat. In practice, however, it is less frequently used by itself for refining purposes, though there is less risk of saponification of the neutral oil if a slight excess is added with this reagent than with caustic soda.

A process has been patented in France by G. Muller (Fr. Pat. 334,366, 1903) for the treatment of cacao butter with sodium bicarbonate. The fat is heated with sodium bicarbonate and water, then cooled with constant agitation until it congeals, allowed to stand for twenty-four hours, and finally subjected to a process of pressing and kneading. The fat thus treated is claimed to be softer and less brittle.

Sodium Silicate.—This is an alkaline salt, and its action is very similar to that of sodium carbonate. Its use for bleaching oils and fats has been patented by Godard (Eng. Pat. 22,085, 1903), who mixes the oil with sodium silicate, separates the soap formed, and then deodorises the neutral oil by means of steam in a fine state of division.

Alkaline Earths.—Lime and magnesia are sometimes used for removing the free fatty acids from oils and fats, insoluble calcium and magnesium soaps being formed. In Rocca’s patent, to which reference has already been made (vide supra), the oil is first neutralised with caustic soda or sodium carbonate, decanted from the resulting soap, a small quantity of strong acid added to decompose any soap remaining, and the oil finally neutralised with lime, magnesia, or baryta.

Fresenius (Eng. Pat. 19,171, 1902) neutralises the oil with caustic soda, lime, or magnesia, under a pressure of 2 or 3 atmospheres, either in the presence of carbon to prevent oxidation, or according to a later process in an atmosphere of an inert gas. The increased pressure is claimed to facilitate the separation of the soap emulsion.

Bleaching of Oils.—The colouring matter of crude oils consists of chlorophyll, which gives them a greenish tinge, or of substances frequently of a resinous nature, which impart a brown colour.

In the case of some oils, such as olive oil, the natural greenish tint is allowed to remain, but the dark colour of certain other crude oils has to be reduced before the product is saleable.

As was mentioned above, treatment with alkali removes from cotton-seed oil a large proportion of the dark colouring matter at the same time as the constituents that cause turbidity.

The methods in which fuller’s earth or milk is used to refine oils have also some effect in producing a filtrate of lighter colour, while a treatment with freshly prepared animal charcoal is effective as a decolorising process in some cases.

Charcoal.—Bleaching with charcoal may be effected by mixing the oil with 1 to 5 per cent. of animal charcoal, in a granular form, warming for a short time, and filtering through a filter press. The bleaching action of animal charcoal, attributed by Knecht to the presence of nitrogen compounds, is greater in the presence of acid.

Crude charcoal requires preliminary treatment before use for bleaching purposes. It should be well boiled, first with pure water, and then after the addition of sufficient sodium carbonate or hydroxide to render it alkaline. It is next washed free from alkali, and boiled for twelve hours with four times its weight of a mixture of equal parts of commercial hydrochloric acid and water, after which it is washed free from acid, dried, and burned in closed vessels. A good bleaching charcoal is thus obtained.

Fuller’s Earth.—Fuller’s earth (aluminium magnesium hydrosilicate) should be dehydrated by roasting prior to use, in order to secure the best results. The quantity required varies from 2-5 per cent. for cocoanut, palm-kernel, and olive oils, to upwards of 10 per cent. for arachis and cotton-seed oils, and the oil should be thoroughly mixed with the reagent and maintained at a temperature of say 100° F. for about fifteen minutes, and then filtered through a filter press. The fuller’s earth retains about 80 per cent. of the oil, which may be extracted by means of a solvent, the latter distilled off, and the recovered oil treated with a fresh quantity of fuller’s earth. The spent fuller’s earth may be “regenerated” by heating it up to 400-500°. Fuller’s earth is frequently used in America for the filtration of “premier jus.”

According to Hirzel (Chem. Rev. Fett- u. Harz-Ind., 1904, 116-118; 145-146), the earthy flavour sometimes remaining after contact with fuller’s earth may be removed by washing with 10 per cent. of a 10 per cent. solution of brine, and by adding 1 to 1½ per cent. of powdered, dry sodium bicarbonate.

Godard (Eng. Pat. 22,086, 1903) carries out the agitation with reagents in vacuo in order to prevent oxidation of the oil.

No very definite details as to the quantity of charcoal or fuller’s earth to be used, the temperature to which the oil should be raised in contact with them, and the time it should be maintained thereat, can be given, as these vary considerably for different oils, and can only be determined by actual experiments with the particular oil it is required to bleach.

The use of finely divided alumina, bauxite, or magnesite previously ignited at a low temperature, has also been patented for decolorising oils, and a process for the recovery of spent decolorising materials (which consists in mixing them with salt water, heating the mixture to about 85° C. with sulphuric acid, then treating it with sodium carbonate and mechanically agitating it for a short time) has been patented by the Soc. Anon. Huilerie et Savonnerie de Lurian (Fr. Pat. 499,915, 1909). The carbon dioxide liberated by this treatment carries the oil to the surface.

Methods in which chlorine or bleaching powder are employed are only applicable for bleaching oils intended for the making of soap or other technical purposes, owing to the reagent attacking the glycerides and imparting a flavour. For similar reasons Watts’ bichromate process of bleaching palm oil is also unsuitable for the treatment of fat intended for food.

Methods of oxidation, either by means of hot air passed through the heated oil or fat, or by means of ozone or ozonised air, under controlled conditions, are employed both to obtain paler oils and to remove substances of unpleasant odour.

The bleaching of beeswax by exposing it in thin strips to the action of air and sunlight has long been practised, and Japan wax is also rendered nearly white by similar means.

The use of artificial light in place of sunlight is claimed in several patents for bleaching oils, such as cotton-seed oil. The oil is made to pass across a transparent surface through which are transmitted the rays from a powerful arc light.

Ultra-violet rays, such as those emitted by the mercury vapour lamp, are employed as the bleaching agent in similar processes.

Ozone.—The use of ozone as a bleaching agent has long been known, but its application to the treatment of oils and fats has hitherto met with little success. It has been the subject of many patents, among which may be mentioned that of Andreoli (Eng. Pat. 14,570, 1898), who makes use of the joint action of ozone and a hypochlorite solution, and that of J. Harris (Eng. Pat. 22,430, 1906). Harris first ozonises the oil at a temperature of 100° to 180° F. for fifteen to thirty minutes, until there is an appreciable rise—say 0·5 per cent.—in the free fatty acids. The treatment with ozone is then stopped, and the oil neutralised with alkali, preferably an aqueous solution of caustic soda at about 15° Tw. (10° B.). The impurities are next separated, and the oil dried by means of a current of air while warm, after which it may be still further refined by heating it with 5 per cent. of aluminous earth, followed by filtration.

This latter process seems to be rather a mistaken one, as from experiments made by one of the authors (S.), the best results are obtained when the ozonisation is only carried to such a point that there is no appreciable increase in the free acidity of the oil.

Recent experiments by him on the bleaching and deodorising action of pure ozone free from nitrogenous compounds, as obtained with the Ozonair apparatus, have given results which are most promising. Palm oil, even of the crudest description, has been most effectively and cheaply bleached by merely passing a strong current of ozonised air through it, and the colour of certain other vegetable oils has been distinctly improved by this treatment; while in the case of a very acid sample of cocoanut oil, the fat was not only made whiter in colour, but had also almost entirely lost its characteristic odour. Hence there seems to be a likelihood of the successful employment of the ozone process for this purpose in the future.

Hydrosulphites or “hyposulphites.”—Sodium hydrosulphite, obtained by digesting a concentrated solution of sodium bisulphite with zinc dust or turnings, is a powerful reducing agent, and has been patented as a means of bleaching oils and fats by Metz and Clarkson (Eng. Pat. 11,983, 1906). It is particularly applicable to maize oil, which may be bleached by agitating 200 parts of the oil with 600 parts of water and 15 parts of sodium hydrosulphite for ten hours in a closed vessel, allowing the mixture to stand for thirty-two hours, and then separating the oily layer.

Sodium hydrosulphite formaldehyde, which was shown by Baumann, Thesmar, and Frossard in 1904 to be commercially a mixture in almost equal proportions of sodium sulphoxylate formaldehyde and sodium bisulphite formaldehyde, is also recommended for the purpose, the oil being heated to 70° C. with this reagent in a closed vessel, and then allowed to stand.

The formaldehyde-sulphoxylates are now sold under a variety of fancy names, the sodium compound being termed Rongalite C. and Hydraldite C., and the zinc salt Decroline.

Sodium Bisulphite.—This salt, which possesses the property of combining with aldehydes, has been utilised in some cases for deodorising oils and fats, with good results.

Organic Peroxides.—The Vereingte Chem. Werke have recently patented the use of various organic peroxides (such as those of benzoyl, acetyl, and acetone, together with the oxidation products obtained by the action of nitrogen tetroxide on organic compounds) for the bleaching of oils and fats. The oil is heated to 100° C. with about 0·2 per cent. of the peroxide, and allowed to stand in a warm place for a short time, until bleaching has taken place.

There is now a large number of per-salts available, e.g., persulphates, percarbonates, perborates, which are being utilised for various bleaching purposes, but so far they do not appear to have been employed to any appreciable extent in the treatment of oils and fats, though ammonium persulphate (palidol) is now being used for bleaching soap in the pan.

Deodorisation of Fats.—The odours of oils and fats are due to the presence of small quantities of volatile substances, either derived from the vegetable substance as in the case of maize and wheat oils, or formed by slight decomposition of the oil itself.

It is chiefly in connection with cocoanut oil, which forms the basis of so many vegetable lards and butters, that the problem of deodorisation of a fresh fat has to be faced.

According to the recent research of Haller and Lassieur (Comptes Rend., 1910, 150, 1013), the unpleasant odour of commercial cocoanut oil must be attributed, partly to a process of decomposition of the glycerides, with the liberation of fatty acids (caproic, caprylic, capric acids, etc.) with a pronounced odour; and partly to the presence of certain substances, which include methyl-heptyl ketone and methyl-nonyl ketone. Traces of these may also be detected in the refined fat, and account for its odour when heated. The substances to which the odour is due may be obtained in the form of an essence (échappés) by distilling the cocoanut oil in a current of superheated steam.

Speaking broadly, two methods only have so far been discovered, or, at any rate, published, for the deodorisation of cocoanut oil—(1) washing out of the odoriferous bodies with alcohol, and (2) their volatilisation by treatment with steam. Other unpublished processes are doubtless used by some manufacturers, but their secrecy is very jealously guarded.

The washing of cocoanut oil with alcohol of course removes free fatty acids, which are soluble therein. This method of treatment was first introduced by Chevreul, and has been also utilised by Schlinck, who employed a joint process of washing with alcohol and treatment with charcoal. A special apparatus for its application was patented by Urbain and Feige (Fr. Pat. 361,966, 1905), consisting of a series of vessels so arranged that the oil descends from vessel to vessel, and meets hot alcohol circulating in the opposite direction, so that the more impure oil comes in contact with the more impure alcohol. The temperature is maintained at 70° C. during the treatment, and means are provided for evaporating and condensing the alcohol after extraction is complete.

An improvement upon this patent has been since protected by the patentees, in which sufficient alkali is added to combine with the free fatty acids prior to the extraction with alcohol.

With so volatile a substance as alcohol, there is necessarily a considerable loss by evaporation during working, and the cost of alcohol in this country is too high for the process to be economical.

The treatment of cocoanut oil with steam in order to deodorise it dates back to 1882, when a process was patented in Germany by Jeserich and Meinert, in which the volatile fatty acids were first driven off by passing high pressure steam at 6 to 8 atmospheres into the fluid oil, with constant stirring for about two or three hours, after which steam was shut off, and the non-volatile fatty acids separated by addition of 0·25 per cent. of calcined magnesia, the magnesium soap formed rising to the surface, whence it could be skimmed off.

Several modifications of this process have since been patented. Klimont (Eng. Pat. 3164, 1902) neutralises the oil with alkali prior to treatment with superheated steam, and then removes non-volatile impurities by either raising the temperature of the oil to above 100° C., treating it with calcium or magnesium hydroxide, or extracting it with some solvent such as acetone, which dissolves the oil and precipitates the impurities.

The Fabrique de Prod. Chim. de Thann et de Mulhouse treat the oil with steam in the absence of air, either in vacuo, or in an atmosphere of an indifferent gas.

In another method of deodorisation cocoanut oil is freed from its more fluid constituents (as in the preparation of chocolate fats, q.v.) and the fatty acids eliminated in the form of calcium salts by treatment with lime; sodium silicate is employed in another process as the means of removing the fatty acids. Yet however carefully the odoriferous substances may have been removed, the readiness with which cocoanut oil undergoes hydrolysis, and the nature of the fatty acids contained in its glycerides, renders this fat particularly liable to acquire an odour again, after having been exposed for a short time to the air.

Treatment of Rancid Fats.—Although the practice is to be condemned, it is not uncommon for fats that have become slightly rancid to be subjected to treatment to eliminate the compounds to which rancidity is due, and thus enable the fat to be blended with fresh fats.

In the development of the changes understood by the term “rancidity,” the glycerides undergo hydrolysis with the liberation of fatty acids, which are in turn decomposed or oxidised, with the formation of various compounds such as fatty anhydrides, aldehydes, and hydroxy acids.

These changes are brought about under the influence of light and atmospheric oxidation, and are probably promoted by enzymic action or by the influence of micro-organisms when albuminous substances have been left in the fat.

The removal of free fatty acids is frequently sufficient to make the oil appear fresh. This may be effected by mechanically agitating the fat with a suitable proportion of milk of lime or magnesia, which combine with the free fatty acids to form insoluble calcium or magnesium soap, which can readily be separated by filtration.

Neutralisation with caustic alkali (as in the refining of cotton-seed oil) or with a solution of sodium silicate, also effects the removal of free fatty acids, but the latter reagent is liable to produce an obstinate emulsion when employed on a large scale.

In another process the rancid oil is treated with a suitable proportion of precipitated chalk, and is subsequently filtered while hot through a layer of animal charcoal.

Reference has already been made to the method of dissolving out the free fatty acids by means of alcohol. Oils in which a process of hydrolysis has once started are liable, after removal of the free fatty acids, not to keep so well as freshly expressed products.

Rancidity may be present in an oil without the liberation of free fatty acids, though, as a rule, the acid value of a fat will afford some index of the degree of rancidity, since the development of acidity and of rancidity are often simultaneous.

For the removal of the aldehydic compounds formed in the changes that occur, a treatment with a strong solution of sodium bisulphite is used in a process described by Nagel, while volatile products are expelled by heating the oil in a current of steam at gradually increasing temperatures.

Other reagents used for sweetening rancid oils are a solution of ordinary salt, dilute (1 per cent.) sulphuric acid, a solution of myrrh in methyl alcohol, etc.

As a rule, the different processes are employed successively, and followed by filtration and drying of the oil.

The tendency of an oil to become rancid depends largely upon the proportion of volatile fatty acids and unsaturated fatty acids it contains. Thus butter and cocoanut oil readily turn rancid, whereas beef stearin and cacao butter will keep for a long time unaltered.

Any oil or fat is best protected from rancidity by keeping it in the dark in a vessel from which all air is excluded. Rancidity will take place in the dark, but much less rapidly than when the fat is exposed to light. This is due to the accelerating effect of light upon the action of oxygen. Heat has also a pronounced influence in promoting chemical changes in oils, and cold has a retarding influence, although it does not stop the oxidation.

The odour of rancid fats that do not (like cocoanut oil) contain volatile fatty acids is due to the formation of aldehydes or of esters, the latter being probably produced, in some cases at all events, by the action of certain micro-organisms upon the liberated glycerin.

Rancidity is accompanied by a decrease in the iodine value of the fat, which is due to the absorption of oxygen by the unsaturated bonds of the liquid fatty acids. A determination of this value may therefore, in some cases, give indications of the freshness of a fat.

A more promising method, however, is to distil the fat in a current of steam, and to estimate the amount of aldehydes in the distillate.