This invention relates to improved food fats, particularly of the variety used in chocolate confectionery and to methods of making such fats.

The present invention provides a method of improving fat fractions containing symmetrical disaturated glycerides which are separated from vegetable and animal fats for use as a substitute for cocoabutter, the outstanding melting characteristics of which are attributable to the presence of these glycerides. The invention provides a process for the preparation of fat fractions containing symmetrical disaturated glycerides by fractional crystallisation of edible fat comprising a mixture of saturated and unsaturated C, g-C, _fi fatty acid glycerides combined with interesterification m which an interesterification catalyst selectively active in the 1- and 3-positιons only is used to randomise the fatty acid residues of the glycerides in the fat at these positions only.

Fractionation is more or less effective to remove trisaturated glycerides of a high-melting solid fraction, and also di- and tri-unsaturated glycerides as a lower- melting fluid fraction, from disaturated glycerides in fats, although complete separation may not be economically feasible and the isomeric, asymmetric disaturated

OMP

glycerides are not normally separable at all by this means. The effect of the selective enzyme treatment in the process of the invention is to improve the fat fraction required by increasing the amount of symmetrical disaturated glycerides it contains and/or reducing the amount of asymmetric disaturated glycerides present with it. In doing so other glycerides may be increased in amount but being either more highly unsaturated or less unsaturated, are readily separated by fractionation.

One effect of the invention may therefore be to improve the symmetry of fats, expressed as the ratio of the symmetrical to the asymmetrical disaturated glycerides present in the fat, i.e. GUG.GGU, without introducing new fatty acid residues. It can be calculated mathematically that this ratio will be increased in a fat consisting of the two isomers which is subjected to the selective enzyme treatment of the invention, provided that the ratio is initially at least 1:2 molar. Generally speaking, this minimum ratio is substantially exceeded m vegetable fats, on account of the preference of unsaturated fatty acids to occupy the 2-posιtion on the glycerides present. Nevertheless, palm oil contains sufficient of the asymmetric disaturated glycerides to be an embarrassment, affecting as it does the melting qualities of even the most carefully fractionated palm mid-fractions which are used m cocoabutter substitute fats. The invention improves palm mid-fractions by decreasing the amount of asymmetric disaturated glycerides present.

Shea oil on the other hand contains very little asymmetric disaturated glycerides and the principal effect of the enzyme step in the application of the process of the invention to shea fractions is to increase the amount of the symmetrical iso er by conversion of the asymmetric GU- . Whether by this or the former mechanism described,

the effect of the interesterification is to enhance the symmetry of the fat. In the subsequent fractionation of the interesterified fat a fraction is recovered rich in symmetrical disaturated glycerides.

Fats or fat fractions subjected to the enzyme treatment should contain as little 2-saturated glycerides ^' as possible, i.e. P-, UGU and GGU.

As proposed by Richardson and later confirmed by Van der Wal (J. Am. Oil Chem. Soc , _3_7, pages 18 to 20), the fatty acid residues of a vegetable fat are distributed in the glycerides of the fat in a 1-, 3-random -2-random pattern. The fatty acid residues in 1- and 3-positions are therefore already randomly distributed and the application of the invention to vegetable fats includes a preliminary step in which this random distribution in the 1- and 3- positions is changed, as for example by fracti ^* bnation. Fractionation for this purpose may be carried out to recover the symmetrical disaturated glycerides in a fluid fraction substantially free from tπsaturated glycerides, or in a crystallised fraction substantially free from di- and tπ-unsaturated glycerides. Any of the unsaturated fractions obtained may be subjected to the selective enzyme treatment step of the invention followed by fractionation. Alternatively, a mid-fraction rich in the disaturated glycerides may be recovered and this and/or the unsaturated fractions separated from it then subjected to the enzyme treatment followed by refractionation.

Animal fats do not appear to follow the behaviour of vegetable fats in the persistence with which the 2-position is occupied by unsaturated fatty acid residues, nor in the random distribution of the fatty acid residues m the 1- and 3-positions. Accordingly, substantially greater proportions of the asymmetric isomer occur in the

OMP

disaturated glycerides of animal fats and a preliminary treatment affecting the distribution of fatty acids in the 1- and 3-positions is not normally necessary for the fat to be susceptible to changes in glyceride composition by the enzyme treatment step of the invention. Nevertheless, fractionation before this treatment may be desirable as before, preferably to remove from a disaturated fraction, a more highly unsaturated fluid fraction and more particularly, an upper-melting trisaturated glyceride fraction. As before, the enzyme treatment may then be applied to any of the unsaturated fractions obtained.

From whatever source they are obtained, fats containing both saturated and unsaturated fatty acids are rearranged in the enzyme treatment step of the invention by converting monosaturated glycerides to symmetrical disaturated glycerides (2 SU_ = SUS + U..) and by improving the ratio of symmetrical to asymmetric disaturated glycerides (2 SSU •= USU + SSS).

The present invention provides a process for the preparation of an improved confectionery fat , characterised by the fact that a fat consisting substantially of saturated and unsaturated C. _fi and C, „ fatty acid glycerides is rearranged by contact with an enzyme selectively active in the 1- and 3-posιtιons only and the rearranged fat is fractionated to recover a fraction comprising symmetrical disaturated C,, and C, glycerides.

Suitable vegetable fats include palm oil, shea, sal, pentadesma, aceituno, phulwara and mango kernel oils. In addition to these vegetable fats, which consist predominantly of ^c _τ g/ ^c _{τ o} fatty acids, suitable animal fats include tallow. Lard may contain too much asymmetric glyceride to be suitable, according to its source. Fat

fractions which may be treated according to the invention include, for example, palm, shea and sal oleines, palm and tallow mid-fractions and cottonseed stearine. While the fats consist substantially of C, _β and C. _Q fatty acid glycerides, minor amounts of other fatty acids may be present whether of shorter- or longer-chain, length. The fats of the lauric group, e.g. palm kernel and coconut oils, are however not included within the scope of the invention. The symmetry for which fats are prized confer excellent melting characteristics and compatibility with cocoabutter, and refers to the presence of saturated fatty acid residues in the two outer or 1, 3-positions of glycerides, whether or not they possess the same chain length.

Suitable enzymes for catalysing the rearrangement in the 1- and 3-positions only are provided by lipase enzymes. These include those produced by Aspergillus niger, Rhizopus arrhizus, Rhizopus japonicus and Rhizopus niveus, Mucor javanicus and Mucor miehei. Such enzymes require the presence of water to initiate and sustain their activity and catalyse the reversible hydrolysis and esteπfication reactions. The reaction is therefore carried out m the presence of small amounts of water, i.e. below 10%, preferably less than 5%, and more particularly not more than 1% by weight of the fat to minimise hydrolysis. In general, a balance must be struck to provide the minimum water activity consistent with maintaining the activity of the enzyme. As to what this may be, catalysts vary substantially m their demands for water and variants may be found and used with minimal water requirements, thereby improving the process by diminishing the extent of hydrolysis. The reaction is also preferably carried out in the presence of a non-aqueous, substantially water-immiscible solvent for example hexane having a relatively low boiling point, i.e. below 100 ^β C to

facilitate its removal from the interesterified fat. The interesterification reaction is also preferably carried out m accordance with our co-pending patent application A.194 with the provision of means to control the water activity A of the reaction system. A suitable process for rearranging fat in the presence of a lipase selective- in the 1- and 3-posιtions only, is described in our patent applications A.171 and A.189. The amount of catalyst is preferably from 0.1 to 10%, particularly 0.1 to 7H by weight of the fat. The catalyst is preferably supported on a Celite or other inert support as described in these Applications and may be dispersed in the reaction mass after being activated by contact with water. Alternatively the catalyst after activation may be supported in a fixed bed. The reaction terniperature is preferably from 0 to

60°C, more preferably 10 to 40 ^β C. Higher temperatures tend to inactivate these enzyme catalysts.

After the interesterification rearrangement is complete, the reaction mass is separated from the catalyst and any solvent present removed prior to the further recrystallisation step. Alternatively, sufficient solvent may be left with the fat for the ref actionation step. Preferably the fractionation is effected from an organic solvent, e.g. acetone, hexane or nitropropane. The fractionation may alternatively be carried out in the absence of solvent or by the so-called Lanza process.

EXAMPLE 1

Shea oil was fractionated from acetone at 5°C to recover a solid stearine fraction containing principally disaturated glycerides and a fluid oleme fraction which was recovered from the solvent, refined and neutralised. The oleine fraction was dissolved in twice its weight of petroleum ether of boiling range 100 to 120°C and the

solution percolated at 40 ^β C at a flow rate of 10.4 mls/hr through alternate layers of Celite and Celite-lipase catalyst packed in a bed in an upright reactor tube. The layers were previously wetted with water and the catalyst layers comprised Aspergillus mger lipase and were prepared m accordance with our patent Case A.189. A total amount of 7 grams of lipase was used.

The product of the first 24 hours was rejected to minimise hydrolysis in the final product which was then collected ^• for 7 days. After removing the solvent, a second stearine fraction in 17% yield based on the neutralised, refined shea oleine was crystallised from a solution of the interesterifled material in 3 times its weight of acetone maintained for 30 minutes at 5°C, the precipitated crystals being washed twice with acetone. For comparison a second solid stearine fraction was similarly recovered at the same temperature in 24% yield from the shea oleine without interesterification.

In the accompanying Table 1 glyceride compositions are given for shea oleine before ar. after interesterification and the solid stearine fractions crystallised from each.

TABLE 1 - Triglyceride compositions

Triglycerides wt % sss SOS sso SLnS soo Others

Shea oil 0.0 —47. — 5.6 34.3 12.1 Shea oleine 0.0 12.9 0.9 5.5 58.0 22.6 Stearine from shea oleine 0.0 41.4 1.6 7.9 38.2 11.0 Interesterified shea oleine 0.0 19.0 1.1 5.3 43.7 31.5

Example 1

Stearine from mteres¬ terified shea oleine 1.8 83.6 1.8 9.1 2.9 0.7

Typical stearine from shea oil 1.0 82.8 1.1 6.3 6.9 1.6

Table 2 shows the solids content of 20°C of the _ stearine product according to the invention and of a blend of equal parts with a typical palm mid-fraction, together with corresponding data for a typical commercial shea stearine product. From the data it is clear that the stearine from interesterified shea oleme is a useful confectionery fat and that as a result of the interesterification reaction a substantial amount of recoverable, symmetrical disaturated glyceride is generated. By contrast, the control test product contained only half the amount of SOS present in the interesterified stearine product of the invention and is very unsatisfactory as a confectionery fat.

TABLE 2

Pulse NMR solids content of stearine (%) at 20°C

Stearine from interes- terified shea oleine 86.7 83.2 72.9 54.6 28.2 3.1

50:50 blend of stearine from interesterified shea oleine and palm 66.6 59.9 42.9 18.4 5.0 1.2 mid-fraction Typical shea stearine 84.2 83.2 80.4 76.3 66.2 6.1 50:50 blend of typical shea stearine and palm 68.6 63.9 49.2 24.1 7.4 0.6 mid-fraction

EXAMPLE 2

In this Example a neutralised and refined mid- fraction of palm oil used m cocoabutter substitutes was enzymatically interesterified as described in Example 1, with however, a flow rate of 21.0 ml/hr. The lipase was derived from Rhizopus japonicus.

After removing the solvent the interesterified fat was fractionated from a solution in 1.6 times its weight of hexane, maintained for 90 minutes at 8 ^C C. The solid, stearine fraction containing principally trisaturated glycerides was filtered and washed twice with hexane. After combining the filtrate and washings the solvent was removed from the fluid, oleine fraction which was recovered m 75.9% yield based on the interesterified fat.

In Table 3, glyceride compositions are given for the palm mid-fraction starting material, the intermediate interesterified fat therefrom and its oleine fraction. From this data it will be seen that the product from this

example contains marginally more symmetrical disaturated glycerides (SOS/SLnS) but substantially less asymmetric disaturated glyceride (SSO).

In all the Tables S = C _lg + C, _g saturated fatty acid residues, O = oleyl and Ln = Imoleyl.

TABLE 3 - Triglyceride composition

sss SOS SSO SLnS SOO Others

Palm oil 7.0 35.7 4.4 8.6 27.5 16.8

Palm mid-fraction 2.4 76.9 7.3 8.3 3.9 1.2 Interesterified palm mid-fraction 8.6 72.6 1.8 7.0 7.8 2.2

Oleine from interes¬ terified palm mid- 1.0 79.4 2.2 7.5 7.6 2.3 fraction

EXAMPLE 3

Example 1 was repeated on another neutralised and refined mid-fraction of palm oil as used in cocoabutter substitutes. A flow rate of 25 ml/hr was used and the lipase was derived from Rhizopus japonicus.

After removing the solvent the interesterified fat was analysed, redi≤solved in 4 times its weight of acetone, and fractionated after maintaining for 2 hours at 20°C. The solid, stearine fraction containing principally tπsaturated glycerides was filtered and washed with acetone. After combining the filtrate and wash the solvent was removed from the fluid, oleine fraction which was recovered in 89% yield based on the interesterified fat.

Following removal of the solvent the oleine fraction was further fractionated from a solution m 7.5 times its

weight of acetone, maintained for 30 minutes at 5°C. The solid, stearine fraction from this second fractionation containing principally SOS glycerides was filtered and washed twice with acetone. The solvent was removed from the solid, stearine fraction of this second fractionation by evaporation in vacuo to give a mid-fraction in 59% yield based on the original interesterified fat.

In Table 4, glyceride compositions are given for the palm mid-fraction starting material, the intermediate interesterified fat therefrom and the mid-fraction derived from the interesterified fat. From these data it will be seen that the product from this Example contains more symmetrical disaturated glycerides (SOS/SLnS) and substantially less asymmetrical disaturated glycerides (SSO) than the starting palm mid-fraction.

TABLE 4 - Triglyceride Compositions

SSS SOS SSO SLnS SOO Others

Palm mid-fraction 4.0 73.7 7.4 8.5 4.8 1.6 Interesterified palm mid-fraction 10.5 62.9 2.5 8.4 11.7 3.9

Mid-fraction ex. interesterified palm 0.8 88.8 0.5 5.5 4.1 0.4 mid-fraction

In Table 5, a clear improvement in the compatibility of the product of this Example upon admixture with cocoa¬ butter over the original palm mid-fraction is demonstrated by reference to the solids content of such mixtures at 20°C as measured by pulse NMR. The softening effect observed in blends of palm mid-fraction with cocoabutter which is due to the presence of the asymmetrical disaturated glycerides

(SSO) in the palm mid-fraction is not seen in the equivalent blends incorporating the product of this Example.

T BLE_5_ _z _N _2g _Values

100% PMF 25 50 75 100% CB

Control 71.8 68.3 64.6 76.8 82.5

Ex. 3 71.8 72.2 71.5 78.5 82.7

EXAMPLE 4 ^"

Tallow was fractionated from a solution in 3 times its weight of acetone, maintained for 1 hour at 1 ^C C. The solid, stearine fraction containing principally trisaturated glycerides was filtered and washed with acetone. After combining the filtrate and wash the solvent was removed from the fluid, oleine fraction which was recovered m 83% yield based on the tallow.

Following removal of the solvent the oleme fraction was further fractionated from a solution m 5 times its weight of acetone, maintained for 1 hour at 5°C. The solid, stearine fraction from this second fractionation containing principally symmetrical and asymmetrical disaturated glycerides (SOS and SSO) was filtered and washed twice with acetone. The solvent was removed from the solid, stearine fraction of this second fractionation by evaporation in vacuo to give a mid-fraction m 20% yield based on the original tallow.

The tallow mid-fraction dissolved in 3 times its weight of petroleum ether of boiling range 100-120" was enzymatically interesterified as described in Example 3

using 20 g of lipase-celite catalyst prepared from Rhizopus japonicus lipase and a flow rate of 12.0 ml/hr.

After removing the solvent the interesterified fat was fractionated from a solution in 3 times its weight of acetone, maintained for 2 hour at 17°C. The solid, stearine fraction containing principally trisaturated glycerides, was filtered and washed with acetone. After combining the filtrate and wash the solvent was removed from the fluid, oleine fraction which was recovered in 82% yield based on the interesterified fat.

Following removal of the solvent the oleine fraction was further fractionated from a solution in 5 times its weight of acetone, maintained for 1 hour at 5°C. The solid, stearine fraction from this second fractionation containing principally symmetrical disaturated glycerides (SOS) was filtered and washed twice with acetone. The solvent was removed from the solid, stearine fraction to give a mid-fraction in 39% yield based on the starting tallow mid-fraction.

In Table 6 glyceride compositions are given for the tallow mid-fraction starting material and the mid-fraction derived from the interesterified tallow mid-fraction. From these data it will be seen that the mid-fraction from the interesterified fat contains substantially more of the desirable symmetrical disaturated cis- onounsaturated glycerides (SOS) and less of the undesirable asymmetrical disaturated cis-monounsaturated glycerides (SSO) than the starting tallow mid-fraction. The ratio of SOS to SSO in the tallow mid-fraction is 3.5, while m the mid-fraction from the interesterified fat it is 7.6.

TABLE 5 - Triglyceride Compositions

SSS SES SOS SSO Others

Tallow mid-fraction 7.0 4.6 51.4 13.7 23.3

Mid-fraction ex. interesterified 4.5 0.6 75.6 10.0 9.3 tallow mid-fraction

S = Saturated fatty acyl group

E = Trans monounsaturated fatty acyl group

O = Cis monounsaturated fatty acyl group.