Simpson (Fr. Pat. 364,587) has attempted to accelerate further the decomposition by subjecting oils or fats to the simultaneous action of heat and electricity. Superheated steam is pa.s.sed into the oil, in which are immersed the two electrodes connected with a dynamo or battery, the temperature not being allowed to exceed 270 C.

II. _Action of Enzymes._--It was discovered by Muntz in 1871 (_Annales de Chemie_, xxii.) that during germination of castor seeds a quant.i.ty of fatty acid was developed in the seeds, which he suggested might be due to the decomposition of the oil by the embryo acting as a ferment.

Schutzenberger in 1876 showed that when castor seeds are steeped in water, fatty acids and glycerol are liberated, and attributed this to the hydrolytic action of an enzyme present in the seeds. No evidence of the existence of such a ferment was adduced, however, till 1890, when Green (_Roy. Soc. Proc._, 48, 370) definitely proved the presence in the seeds of a ferment capable of splitting up the oil into fatty acid and glycerol.

The first experimenters to suggest any industrial application of this enzymic hydrolysis were Connstein, Hoyer and Wartenburg, who (_Berichte_, 1902, 35, pp. 3988-4006) published the results of a lengthy investigation of the whole subject. They found that tallow, cotton-seed, palm, olive, almond, and many other oils, were readily hydrolysed by the castor-seed ferment in the presence of dilute acid, but that cocoa-nut and palm-kernel oils only decomposed with difficulty. The presence of acidity is essential for the hydrolysis to take place, the most suitable strength being one-tenth normal, and the degree of hydrolysis is proportional to the quant.i.ty of ferment present. Sulphuric, phosphoric, acetic or butyric acids, or sodium bisulphate, may be used without much influence on the result. Butyric acid is stated to be the best, but in practice is too expensive, and acetic acid is usually adopted. The emulsified mixture should be allowed to stand for twenty-four hours, and the temperature should not exceed 40 C.; at 50 C. the action is weakened, and at 100 C. ceases altogether.

Several investigators have since examined the hydrolysing power of various other seeds, notably Braun and Behrendt (_Berichte_, 1903, 36, 1142-1145, 1900-1901, and 3003-3005), who, in addition to confirming Connstein, Hoyer and Wartenburg's work with castor seeds, have made similar experiments with jequirity seeds (_Abrus peccatorius_) containing the enzyme abrin, emulsin from crushed almonds, the leaves of _Arctostaphylos Uva Ursi_, containing the glucoside arbutin, myrosin from black mustard-seed, gold lac (_Cheirantus cheiri_) and crotin from croton seeds. Jequirity seeds were found to have a stronger decomposing action on lanoline and carnauba wax than the castor seed, but only caused decomposition of castor oil after the initial acidity was first neutralised with alkali. Neither emulsin, arbutin nor crotin have any marked hydrolytic action on castor oil, but myrosin is about half as active as castor seeds, except in the presence of pota.s.sium myronate, when no decomposition occurs.

S. Fokin (_J. russ. phys. chem. Ges._, 35, 831-835, and _Chem. Rev.

Fett. u. Harz. Ind._, 1904, 30 _et seq._) has examined the hydrolytic action of a large number of Russian seeds, belonging to some thirty different families, but although more than half of these brought about the hydrolysis of over 10 per cent. of fat, he considers that in only two cases, _viz._, the seeds of _Chelidonium majus_ and _Linaria vulgaris_, is the action due to enzymes, these being the only two seeds for which the yield of fatty acids is proportional to the amount of seed employed, while in many instances hydrolysis was not produced when the seeds were old. The seeds of _Chelidonium majus_ were found to have as great, and possibly greater, enzymic activity than castor seeds, but those of _Linaria_ are much weaker, twenty to thirty parts having only the same lipolytic activity as four to five parts of castor seeds.

The high percentage of free acids found in rice oil has led C. A. Brown, jun. (_Journ. Amer. Chem. Soc._, 1903, 25, 948-954), to examine the rice bran, which proves to have considerable enzymic activity, and rapidly effects the hydrolysis of glycerides.

The process for the utilisation of enzymic hydrolysis in the separation of fatty acids from glycerine on the industrial scale, as originally devised by Connstein and his collaborators, consisted in rubbing a quant.i.ty of the coa.r.s.ely crushed castor seeds with part of the oil or fat, then adding the rest of the oil, together with acidified water (N/10 acetic acid). The quant.i.ties employed were 6-1/2 parts of decorticated castor beans for every 100 parts of oil or fat, and 50 to 60 parts of acetic acid. After stirring until an emulsion is formed, the mixture is allowed to stand for twenty-four hours, during which hydrolysis takes place. The temperature is then raised to 70-80 C., which destroys the enzyme, and a 25 per cent. solution of sulphuric acid, equal in amount to one-fiftieth of the total quant.i.ty of fat originally taken, added to promote separation of the fatty acids. In this way three layers are formed, the one at the top consisting of the clear fatty acids, the middle one an emulsion containing portions of the seeds, fatty acids and glycerine, and the bottom one consisting of the aqueous glycerine. The intermediate layer is difficult to treat satisfactorily; it is generally washed twice with water, the washings being added to glycerine water, and the fatty mixture saponified and the resultant soap utilised.

The process has been the subject of a considerable amount of investigation, numerous attempts having been made to actually separate the active fat-splitting const.i.tuent of the seeds, or to obtain it in a purer and more concentrated form than is furnished by the seeds themselves. Nicloux (_Comptes Rendus_, 1904, 1112, and _Roy. Soc.

Proc._, 1906, 77 B, 454) has shown that the hydrolytic activity of castor seeds is due entirely to the cytoplasm, which it is possible to separate by mechanical means from the aleurone grains and all other cellular matter. This active substance, which he terms "lipaseidine," is considered to be not an enzyme, though it acts as such, following the ordinary laws of enzyme action; its activity is destroyed by contact with water in the absence of oil. This observer has patented (Eng. Pat.

8,304, 1904) the preparation of an "extract" by triturating crushed castor or other seeds with castor oil, filtering the oily extract, and subjecting it to centrifugal force. The deposit consists of aleurone and the active enzymic substance, together with about 80 per cent. of oil, and one part of it will effect nearly complete hydrolysis of 100 parts of oil in twenty-four hours. In a subsequent addition to this patent, the active agent is separated from the aleurone by extraction with benzene and centrifugal force. By the use of such an extract, the quant.i.ty of alb.u.minoids brought into contact with the fat is reduced to about 10 per cent. of that in the original seeds, and the middle layer between the glycerine solution and fatty acids is smaller and can be saponified directly for the production of curd soap, while the glycerine solution also is purer.

In a further patent Nicloux (Fr. Pat. 349,213, 1904) states that the use of an acid medium is unnecessary, and claims that even better results are obtained by employing a neutral solution of calcium sulphate containing a small amount of magnesium sulphate, the proportion of salts not exceeding 0.5 per cent. of the fat, while in yet another patent, jointly with Urbain (Fr. Pat. 349,942, 1904), it is claimed that the process is accelerated by the removal of acids from the oil or fat to be treated, which may be accomplished by either washing first with acidulated water, then with pure water, or preferably by neutralising with carbonate of soda and removing the resulting soap.

Lombard (Fr. Pat. 350,179, 1904) claims that acids act as stimulating agents in the enzymic hydrolysis of oils, and further that a simple method of obtaining the active product is to triturate oil cake with its own weight of water, allow the mixture to undergo spontaneous proteolytic hydrolysis at 40 C. for eight days, and then filter, the filtrate obtained being used in place of water in the enzymic process.

Hoyer, who has made a large number of experiments in the attempt to isolate the lipolytic substance from castor seeds, has obtained a product of great activity, which he terms "ferment-oil," by extracting the crushed seeds with a solvent for oils.

The Verein Chem. Werke have extended their original patent (addition dated 11th December, 1905, to Fr. Pat. 328,101, Oct., 1902), which now covers the use of vegetable ferments in the presence of water and manganese sulphate or other metallic salt. It is further stated that acetic acid may be added at the beginning of the operation, or use may be made of that formed during the process, though in the latter case hydrolysis is somewhat slower.

Experiments have been carried out by Lewkowitsch and Macleod (_Journ.

Soc. Chem. Ind._, 1903, 68, and _Proc. Roy. Soc._, 1903, 31) with ferments derived from animal sources, _viz._, lipase from pig's liver, and steapsin from the pig or ox pancreas. The former, although it has been shown by Kastle and Loevenhart (_Amer. Chem. Journ._, 1900, 49) to readily hydrolyse ethyl butyrate, is found to have very little fat-splitting power, but with steapsin more favourable results have been obtained, though the yield of fatty acids in this case is considerably inferior to that given by castor seeds. With cotton-seed oil, 83-86 per cent. of fatty acids were liberated as a maximum after fifty-six days, but with lard only 46 per cent. were produced in the same time. Addition of dilute acid or alkali appeared to exert no influence on the decomposition of the cotton-seed oil, but in the case of the lard, dilute alkali seemed at first to promote hydrolysis, though afterwards to r.e.t.a.r.d it.

Fokin (_Chem. Rev. Fett. u. Harz. Ind._, 1904, 118-120 _et seq._) has attempted to utilise the pancreatic juice on a technical scale, but the process proved too slow and too costly to have any practical use.

_Rancidity._--The hydrolysing power of enzymes throws a good deal of light on the development of rancidity in oils and fats, which is now generally regarded as due to the oxidation by air in the presence of light and moisture of the free fatty acids contained by the oil or fat.

It has long been known that whilst recently rendered animal fats are comparatively free from acidity, freshly prepared vegetable oils invariably contain small quant.i.ties of free fatty acid, and there can be no doubt that this must be attributed to the action of enzymes contained in the seeds or fruit from which the oils are expressed, hence the necessity for separating oils and fats from adhering alb.u.minous matters as quickly as possible.

_Decomposition of Fats by Bacteria._--Though this subject is not of any practical interest in the preparation of fatty acids for soap-making, it may be mentioned, in pa.s.sing, that some bacteria readily hydrolyse fats.

Schriber (_Arch. f. Hyg._, 41, 328-347) has shown that in the presence of air many bacteria promote hydrolysis, under favourable conditions as to temperature and access of oxygen, the process going beyond the simple splitting up into fatty acid and glycerol, carbon dioxide and water being formed. Under anaerobic conditions, however, only a slight primary hydrolysis was found to take place, though according to Rideal (_Journ.

Soc. Chem. Ind._, 1903, 69) there is a distinct increase in the amount of free fatty acids in a sewage after pa.s.sage through a septic tank.

Experiments have also been made on this subject by Rahn (_Centralb.

Bakteriol_, 1905, 422), who finds that _Penicillium glauc.u.m_ and other penicillia have considerable action on fats, attacking the glycerol and lower fatty acids, though not oleic acid. A motile bacillus, producing a green fluorescent colouring matter, but not identified, had a marked hydrolytic action and decomposed oleic acid. The name "lipobacter" has been proposed by De Kruyff for bacteria which hydrolyse fats.

III. _Use of Chemical Reagents._--Among the chief accelerators employed in the hydrolysis of oils are sulphuric acid and Twitch.e.l.l's reagent (benzene- or naphthalene-stearosulphonic acid), while experiments have also been made with hydrochloric acid (_Journ. Soc. Chem. Ind._, 1903, 67) with fairly satisfactory results, and the use of sulphurous acid, or an alkaline bisulphite as catalyst, has been patented in Germany. To this cla.s.s belong also the bases, lime, magnesia, zinc oxide, ammonia, soda and potash, though these latter substances differ from the former in that they subsequently combine with the fatty acids liberated to form soaps.

_Sulphuric Acid._--The hydrolysing action of concentrated sulphuric acid upon oils and fats has been known since the latter part of the eighteenth century, but was not applied on a practical scale till 1840 when Gwynne patented a process in which sulphuric acid was used to liberate the fatty acids, the latter being subsequently purified by steam distillation. By this method, sulpho-compounds of the glyceride are first formed, which readily emulsify with water, and, on treatment with steam, liberate fatty acids, the glycerol remaining partly in the form of glycero-sulphuric acid. The process has been investigated by Fremy, Geitel, and more recently by Lewkowitsch (_J. Soc. of Arts_, "Cantor Lectures," 1904, 795 _et seq._), who has conducted a series of experiments on the hydrolysis of tallow with 4 per cent. of sulphuric acid of varying strengths, containing from 58 to 90 per cent. sulphuric acid, H_{2}SO_{4}. Acid of 60 per cent. or less appears to be practically useless as a hydrolysing agent, while with 70 per cent. acid only 47.7 per cent. fatty acids were developed after twenty-two hours'

steaming, and with 80 and 85 per cent. acid, the maximum of 89.9 per cent. of fatty acids was only reached after fourteen and fifteen hours'

steaming respectively. Using 98 per cent. acid, 93 per cent. of fatty acids were obtained after nine hours' steaming, and after another seven hours, only 0.6 per cent. more fatty acids were produced. Further experiments have shown that dilute sulphuric acid has also scarcely any action on cotton-seed, whale, and rape oils.

According to Lant Carpenter, some 75 per cent. of solid fatty acids may be obtained from tallow by the sulphuric acid process, owing to the conversion of a considerable quant.i.ty of oleic acid into isoleic acid (_vide_ p. 12), but in the process a considerable proportion of black pitch is obtained. C. Dreymann has recently patented (Eng. Pat. 10,466, 1904) two processes whereby the production of any large amount of hydrocarbons is obviated. In the one case, after saponification with sulphuric acid, the liberated fatty acids are washed with water and treated with an oxide, carbonate, or other acid-fixing body, _e.g._, sodium carbonate, prior to distillation. In this way the distillate is much clearer than by the ordinary process, and is almost odourless, while the amount of unsaponifiable matter is only about 1.2 per cent.

The second method claimed consists in the conversion of the fatty acids into their methyl esters by treatment with methyl alcohol and hydrochloric acid gas, and purification of the esters by steam distillation, the pure esters being subsequently decomposed with superheated steam, in an autoclave, with or without the addition of an oxide, _e.g._, 0.1 per cent. zinc oxide, to facilitate their decomposition.

_Twitch.e.l.l's Reagent._--In Twitch.e.l.l's process use is made of the important discovery that aqueous solutions of fatty aromatic sulphuric acids, such as benzene- or naphthalene-stearosulphonic acid, readily dissolve fatty bodies, thereby facilitating their dissociation into fatty acids and glycerol. These compounds are stable at 100 C., and are prepared by treating a mixture of benzene or naphthalene and oleic acid with an excess of sulphuric acid, the following reaction taking place:--

C_{6}H_{6} + C_{18}H_{34}O_{2} + H_{2}SO_{4} = C_{6}H_{4}(SO_{3}H)C_{18}H_{35}O + H_{2}O.

On boiling the resultant product with water two layers separate, the lower one consisting of a clear aqueous solution of sulphuric acid and whatever benzene-sulphonic acid has been formed, while the upper layer, which is a viscous oil, contains the benzene-stearosulphonic acid. This, after washing first with hydrochloric acid and then rapidly with petroleum ether, and drying at 100 C. is then ready for use; the addition of a small quant.i.ty of this reagent to a mixture of fat (previously purified) and water, agitated by boiling with open steam, effects almost complete separation of the fatty acid from glycerol.

The process is generally carried out in two wooden vats, covered with closely fitting lids, furnished with the necessary draw-off c.o.c.ks, the first vat containing a lead coil and the other a bra.s.s steam coil.

In the first vat, the fat or oil is prepared by boiling with 1 or 2 per cent. of sulphuric acid (141 Tw. or 60 B.) for one or two hours and allowed to rest, preferably overnight; by this treatment the fat is deprived of any dirt, lime or other impurity present. After withdrawing the acid liquor, the fat or oil is transferred to the other vat, where it is mixed with one-fifth of its bulk of water (condensed or distilled), and open steam applied. As soon as boiling takes place, the requisite amount of reagent is washed into the vat by the aid of a little hot water through a gla.s.s funnel, and the whole is boiled continuously for twelve or even twenty-four hours, until the free fatty acids amount to 85-90 per cent. The amount of reagent used varies with the grade of material, the smaller the amount consistent with efficient results, the better the colour of the finished product; with good material, from 1/2 to 3/4 per cent. is sufficient, but for materials of lower grade proportionately more up to 2 per cent. is required. The reaction appears to proceed better with materials containing a fair quant.i.ty of free acidity.

When the process has proceeded sufficiently far, the boiling is stopped and free steam allowed to fill the vat to obviate any discoloration of the fatty acids by contact with the air, whilst the contents of the vat settle.

The settled glycerine water, which should amount in bulk to 50 or 60 per cent. of the fatty matter taken, and have a density of 7-1/2 Tw. (5 B.), is removed to a receptacle for subsequent neutralisation with milk of lime, and, after the separation of sludge, is ready for concentration.

The fatty acids remaining in the vat are boiled with a small quant.i.ty (0.05 per cent., or 1/10 of the Twitch.e.l.l reagent requisite) of commercial barium carbonate, previously mixed with a little water; the boiling may be prolonged twenty or thirty minutes, and at the end of that period the contents of the vat are allowed to rest; the water separated should be neutral to methyl-orange indicator.

It is claimed that fatty acids so treated are not affected by the air, and may be stored in wooden packages.

_Hydrochloric Acid._--Lewkowitsch (_Journ. Soc. Chem. Ind._, 1903, 67) has carried out a number of experiments on the accelerating influence of hydrochloric acid upon the hydrolysis of oils and fats, which show that acid of a specific gravity of 1.16 has a very marked effect on most oils, cocoa-nut, cotton-seed, whale and rape oils, tallow and lard being broken up into fatty acid and glycerol to the extent of some 82-96 per cent. after boiling 100 grams of the oil or fat with 100 c.c. of acid for twenty-four hours. The maximum amount of hydrolysis was attained with cocoa-nut oil, probably owing to its large proportion of the glycerides of volatile fatty acids. Castor oil is abnormal in only undergoing about 20 per cent. hydrolysis, but this is attributed to the different const.i.tution of its fatty acids, and the ready formation of polymerisation products. Experiments were also made as to whether the addition of other catalytic agents aided the action of the hydrochloric acid; mercury, copper sulphate, mercury oxide, zinc, zinc dust, aluminium chloride, nitrobenzene and aniline being tried, in the proportion of 1 per cent. The experiments were made on neutral lard and lard containing 5 per cent. of free fatty acids, but in no case was any appreciable effect produced.

So far this process has not been adopted on the practical scale, its chief drawback being the length of time required for saponification.

Undoubtedly the hydrolysis would be greatly facilitated if the oil and acid could be made to form a satisfactory emulsion, but although saponin has been tried for the purpose, no means of attaining this object has yet been devised.

_Sulphurous Acid or Bisulphite._--The use of these substances has been patented by Stein, Berge and De Roubaix (Germ. Pat. 61,329), the fat being heated in contact with the reagent for about nine hours at 175-180 C. under a pressure of some 18 atmospheres, but the process does not appear to be of any considerable importance.

_Lime._--The use of lime for the saponification of oils and fats was first adopted on the technical scale for the production of candle-making material, by De Milly in 1831. The insoluble lime soap formed is decomposed by sulphuric acid, and the fatty acids steam distilled.

The amount of lime theoretically necessary to hydrolyse a given quant.i.ty of a triglyceride, ignoring for the moment any catalytic influence, can be readily calculated; thus with stearin the reaction may be represented by the equation:--

CH_{2}OOC_{18}H_{35} CH_{2}OH | | 2CHOOC_{18}H_{35} + 3Ca(OH)_{2} = 3Ca(OOC_{18}H_{35})_{2} + 2CHOH | | CH_{2}OOC_{18}H_{35} CH_{2}OH stearin milk of lime calcium stearate glycerol

In this instance, since the molecular weight of stearin is 890 and that of milk of lime is 74, it is at once apparent that for every 1,780 parts of stearin, 222 parts of milk of lime or 168 parts of quick-lime, CaO, would be required. It is found in practice, however, that an excess of 3-5 per cent. above the theoretical quant.i.ty of lime is necessary to complete the hydrolysis of a fat when carried on in an open vessel at 100-105 C., but that if the saponification be conducted under pressure in autoclaves the amount of lime necessary to secure almost perfect hydrolysis is reduced to 2-3 per cent. on the fat, the treatment of fats with 3 per cent. of lime under a pressure of 10 atmospheres producing a yield of 95 per cent. of fatty acids in seven hours. The lower the pressure in the autoclave, the lighter will be the colour of the resultant fatty acids.

_Magnesia._--It has been proposed to subst.i.tute magnesia for lime in the process of saponification under pressure, but comparative experiments with lime and magnesia, using 3 per cent. of lime and 2.7 per cent. of magnesia (_Journ. Soc. Chem. Ind._, xii., 163), show that saponification by means of magnesia is less complete than with lime, and, moreover, the reaction requires a higher temperature and therefore tends to darken the product.

_Zinc Oxide._--The use of zinc oxide as accelerating agent has been suggested by two or three observers. Poullain and Michaud, in 1882, were granted a patent for this process, the quant.i.ty of zinc oxide recommended to be added to the oil or fat being 0.2 to 0.5 per cent.

Rost, in 1903, obtained a French patent for the saponification of oils and fats by steam under pressure in the presence of finely divided metals or metallic oxides, and specially mentions zinc oxide for the purpose.

It has also been proposed to use zinc oxide in conjunction with lime in the autoclave to obviate to some extent the discoloration of the fatty acids.

Other catalytic agents have been recommended from time to time, including strontianite, lead oxide, caustic baryta, aluminium hydrate, but none of these is of any practical importance.

_Soda and Potash._--Unlike the preceding bases, the soaps formed by soda and potash are soluble in water, and const.i.tute the soap of commerce.

These reagents are always used in sufficient quant.i.ty to combine with the whole of the fatty acids contained in an oil or fat, though doubtless, by the use of considerably smaller quant.i.ties, under pressure, complete resolution of the fatty matter into fatty acids and glycerol could be accomplished. They are, by far, the most important saponifying agents from the point of view of the present work, and their practical use is fully described in Chapter V.