40 RILEY: THE DETERMINATION OF ACETYL VALUES FOR USE IN [Vol. 76 The Determination of Acetyl Values for Use in Component Fatty Acid Analyses of Castor Oils BY J. P. RILEY A comparison has been made between the British Standards Institution (second method) and the Association of Official Agricultural Chemists' methods for the determination of acetyl values, and it has been concluded that both procedures give very similar results. Where results of high accuracy are required several determinations should be made and the mean calculated. A procedure has been developed for the complete analyses of castor oils and tested by application to mixtures of fatty acids of known composition. NUMEROUS methods have been proposed for the determination of the acetyl value of fats, most of which are based either on direct acetylation of the oil in the cold with a mixture of pyridine and acetic anhydride and subsequent determination of the amount of acetic anhydride consumed,l y 2 or on acetylation with boiling acetic anhydride and determination of the acetic acid obtained by hydrolysis of the resultant acetyl comp0und.39~ The former method suffers from the defect that the end-point of-the titration is rather vague and many modifications have been suggested for increasing the accura~y,~9~3~ but for the ultimate purpose of component acid determination direct acetylation with acetic anhydride alone has been found preferable.In the present study a comparison has been made between the British Standards Institution (second m e t h ~ d ) ~ and the Association of Official Agricultural Chemists4 methods for the determination of acetyl values, both of which depend on the determination of the saponification values of the oil and the acetylated oil.It has been found that both methods give very similar results, but it is suggested that where results of high accuracy are required the saponification values both of the oil and its acetyl compound should be the mean of at least five determinations. The accuracy of the B.S.I. method has been tested by using pure specimens of methyl ricinoleate and methyl 12-hydroxystearate ; and it is concluded that with care results having an accuracy of &I per cent. can be obtained. The acetyl value method, when applied to castor oils, does not give a true estimate of the amount of ricinoleo-glycerides present, since small amounts (about 0.6 per cent.) of 9 : 10-dihydroxystearo-glycerides that are present6 are also estimated.The dihydroxystearic acid may be determined with sufficient accuracy by crystallisation of the mixed fatty acids of the oil (after removal of unsaponifiable matter by the S.P.A. methods) from ethyl acetate at 0" C. The precipitated dihydroxystearic acid, which has only a low solubility under these conditions, is weighed after filtration and washing, and to this weight is added a correction for its slight solubility in the ethyl acetate used as solvent and wash liquor. Knowledge of the acetyl value and percentage of dihydroxystearic acid enables the percentage of ricinoleic acid in the oil to be calculated. Besides ricinoleic acid, the other principal constituent of castor oils is linoleic acid, which occurs to the extent of about 4 to 6 per cent, The linoleicJan., 19511 COMPONENT FATTY ACID ANALYSES OF CASTOR OILS 41 acid occurring in castor oil has been identified as the normal 9 : 12-linoleic acid of seed fats (J.P. Riley, unpublished observation). It may be determined by spectrophotometric examination of the mixed acids (after removal of unsaponifiable matter), after alkali isomerisa- tion at 180" C for 60 minutes under the conditions of Hilditch et al.1° I t is not necessary to make a correction for the absorption due to ricinoleic acid since it has been shown that ricinoleic acid treated under these conditions shows negligible absorption. Oleic and saturated acids, which occur in small amounts in the oil, are determined by iodine value and by difference respectively .As a test of the proposed method for the analysis of castor oils, mixtures have been made up from pure methyl ricinoleate, oleate and linoleate and analysed by the recommended procedure with results which agree with the calculated values. EXPERIMENTAL PREPARATION OF PURE ACIDS AND ESTERS- Methyl ricinoleate-Brown and Green11 first prepared methyl ricinoleate by low tempera- ture crystallisation, but attempts to repeat their work were unsuccessful. After a study of a number of solvents it was concluded that acetone was the most suitable, but that the solubility of methyl ricinoleate in it varied greatly with temperature. Crystallisations were therefore carried out by slowly cooling the solution to a temperature at which the desired amount of solid had separated and then maintaining it a t this temperature for 3 hours.A commercial cold drawn castor oil (600 g, I.V. 82-9) was hydrolysed and the un- saponifiable matter extracted. The recovered acids were methylated at room temperature. The neutral esters (566 g) were crystallised from acetone first in 11 per cent. solution a t -50" C and then- in 9 per cent. solution at -40" C. The soluble product (341 g) from the latter crystallisation was then recrystallised three times in 5 per cent. solution in acetone a t -65" C, -50" C and finally at -70" C, by which means there resulted 190.7 g of methyl ricinoleate S.E. 311.2 (S.E. (theory) 312.0), I.V. 80.0 (I.V. (theory) 81.2), n? 1.4657, nz 1.4603.Ricinoleic acid-Methyl ricinoleate (1.88 g) was hydrolysed by boiling for 10 minutes with 10 ml of 10 per cent. alcoholic potash. The soaps were acidified with dilute sulphuric acid, and the liberated acids extracted with ether. Most of the ether was removed by distilla- tion, and the last trace in vacuo at room temperature. Yield of ricinoleic acid 1.73 g, I.V. 83.8 (I.V. (theory) 85.2), EtZi at 234 mp after alkali isomerisation at 180" for 60 minutes. 1.1. MethyZ 12-hydroxystearate-Commercial cold drawn castor oil was alkali refined by washing with 10 per cent. potassium hydroxide solution and then hydrogenated with Raney nickel catalyst at 100" C. The product (I.V. 2.4) was saponified and the resultant acids (90 g) were ground up and extracted in a Soxhlet extractor with light petroleum (b.p.40" to 60" C) for 25 minutes. The residue (86 g) was recrystallised three times in 1 per cent. solution in petroleum (b.p. 60" to 80" C) yielding 69 g of 12-hydroxystearic acid, m.p. 82" C. The acid (20 g) was esterified at room temperature with methyl alcohol containing 0-5 per cent. of hydrogen chloride, yielding 19.7 g of methyl 12-hydroxystearate which after crystallisation from acetone a t -20" C had m.p. 57.5" to 58" C (Grummitt and Siedschlag12 record 56" to 9 : 10-Dihydroxystearic acid-The mixed acids (296 g) from 304 g of castor oil were dissolved in 1500 ml of ethyl acetate and cooled to 0" C. After standing overnight a t 0" C, the precipitated dihydroxystearic acid (1.36 g) was filtered off and twice recrystallised from 50 ml of ethyl acetate at 0" C, yielding 1.01 g of 9 : 10-dihydroxystearic acid, m.p.141" to 142" C. \ Determination of the solubility of dihydroxystearic acid-Dihydroxystearic acid (0.5 g) was stirred a t 0" C for 3 hours with 550 ml of redistilled ethyl acetate, and the solution was filtered through a cooled sintered glass filter. 500ml of the filtered solution was evaporated to dryness in a platinum basin, leaving 0.0925 g of dihydroxystearic acid; whence the solubility of dihydroxystearic acid is 0.0185 g per 100 ml of ethyl acetate at 0" C. Il.lethyZ oleate-Methyl oleate prepared as described by Hilditch and Pathak13 had the following characteristics-I.V. 84.8, E:?i at 234 mp after alkali isomerisation at 180" C for 60 minutes 4-3, whence its component acid composition was-saturated 1.6 per cent., oleic 97-9 per cent.and linoleic 0.5 per cent. by weight. Methyl Zinoleate-A concentrate of linoleic acid obtained by low temperature crystallisa- tion of the mixed fatty acids of a sunflower seed oil was methylated and fractionated. The 57" C).42 RILEY: THE DETERMINATION OF ACETYL VALUES FOR USE I N [Vol. 76 fraction having I.V. 165.3, Ei:& at 234 mp after allkali isomerisation at 180" C for 60 minutes 796, was used in the investigation. The component acids of the sample, calculated from the above data were: oleic 12.1 per cent., linoleic 8'7-9 per cent. by weight. DETERMINATION OF ACETYL VALUE BY B.S.I. AND A.O.A.C. METHODS- SaponiJication equivaled-Weigh out 3 g of tlne oil (or 1-5 g of the acetylated oil) into a 300-ml flat-bottomed flask, and add 25 ml of neutral alcohol, followed by 50 ml of approxi- mately 0.5 N alcoholic potash.Heat the mixture under reflux for 2 hours and then titrate with 0-5 N sulphuric acid using phenolphthalein as indicator. Run a blank determination under the same conditions but omitting the sample. Carry out saponification value determinations on both the oil and its acetyl compound in quintuplicate and calculate the acetyl value according to the formula: Acetyl value-Acetylate 15 g of the oil as described in B.S.S. 684 (1950). S' - s 1 - 0*001^)75s Ac.V. = where S is the mean sap. value of the oil S' is the mean sap. value of the acetylated oil. All the saponification tests in the text and tables of this paper are recorded as saponifica- tion equivalents (S.E.), and not as in the B.S.I. and A.O.A.C.specifications as saponification values. The formula for converting saponification equivalents to saponification values is- 56,100 S.E. COMPARISON OF THE B.S.I. AND A.O.A.C. MEiTHODS FOR THE DETERMINATION OF ACETYL VALUES A neutralised commercial castor oil and the rnethyl ,esters prepared from another castor oil were examined by both the B.S.I. and A.O.A.C. methods with the results shown in Table I. COMPARISON OF B.S.I. AND A.O.A.C. ACETYL VALUE METHODS Oil S.E. 309.3 310.8 309.7 310.2 310.0 310.6 Acetylated oil I--- B.S.I. A.O.A.C. S.E. S.E. 180.8 179.3 180.4 179.7 181.3 179.6 181.6 180.5 181.1 180.2 Methyl esters S.E. 314-3 312.9 312.0 313.4 312-7 3 12.5 Acetylated methyl esters -7 B.S.I.A.O.A.C. S.E. S.E. 189.1 189.3 188.0 189.1 188.7 188.7 187.5 189.4 187.8 188-3 188.7 189.3 Mean . . . . .. . . 310.2 Mean deviation . . .. . . 0.14y0 Standard deviation . . . . 0.566 Acetyl value . . .. .. - 181.0 179.9 0.17% 0.22% 0.469 0.707 149.3 152.4 313.0 0.20y0 0.787 - 188.3 189.0 0.27% 0.19% 0.624 0.436 137.1 135.8 From the above results it will be seen that both methods give very similar results and that the determined acetyl values do not differ from the mean by more than 1 per cent. For consistency in the later parts of the work, the B.S.I. method has been employed. EXAMINATION OF METHYL RICINOLEATE AN11 METHYL 12-HYDROXYSTEARATE BY THE B.S.I. METHOD The accuracy of the adopted method has been further tested by the examination of pure samples of methyl ricinoleate and methyl 12-hydroxystearate with the results shown in Table 11.The figures shown indicate that there is good agreement between the determined and calculated saponification equivalents and acetyl values. The determined acetyl values areJan., 19511 COMPONENT FATTY ACID ANALYSES OF CASTOR OILS 43 about 0.6 per cent. low, but it is impossible to say whether this is due to an error in the method, or to the presence of small amounts of impurities in the esters used as standards. COMPLETE ANALYSIS OF CASTOR OILS Dissolve about 60 to 80 g of the castor oil in 300 ml of ether and neutralise it by washing with 10 per cent. aqueous potassium hydroxide and then with water till free from alkali. Divide the neutral oil into two portions.(i) Use 30g for the determination of acetyl value. (ii) Saponify 30 to 40g of the oil by boiling for 1 hour with excess of 10 per cent. alcoholic potash, extract the unsaponifiable matter with ether and determine it according to the S.P.A. methodg (with volumes increased proportionately). Liberate the fatty acids from the soaps with dilute sulphuric acid and extract them with ether, remove the ether by distillation. Use the first portion (2 g) for Divide the resultant acids into two portions. TABLE I1 EXAMINATION OF PURE ESTERS BY B.S.I. PROCEDURE Methyl ricinoleate Methyl 1 2-hydroxystearate Acetylated Acetylated 7- Ester ester Ester ester Number of determinations . . .. 5 5 2 2 S.E. mean .. .. .. . . 311.2 177.3 313.8 178-4 S.E. theory . . .. .. . . 312.0 177.0 314.0 178-0 Mean deviation .. .. .. 0.12% 0.13% 0.06 yo 0.06% Standard deviation . . .. . . 0.510 0.374 - - Acetyl value (found). . .. .. 157.4 156-7 (theory) . . .. .. 158.5 157.6 the determination of linoleic acid by alkali isomerisation at 180" C for 60 minutes followed by spectrophotometric examination at 234 mp by the method of Hilditch et aZ.1° Dissolve the remaining portion of the acids (about 25 to 35 g) in 10 ml of ethyl acetate for each gram of acids and cool to 0" C overnight. Filter the precipitated dihydroxystearic acid through a cooled, weighed sintered glass filter (porosity 3), wash it with a known volume (about 30 ml) of cold ethyl acetate in small portions. Dry the washed precipitate in a vacuum desiccator under high vacuum and weigh. The product consists of substantially pure dihydroxystearic acid and should have a melting-point of not less than 139" C (pure 9 : 10- dihydroxystearic acid has m.p.141" to 142" C). In calculating the percentage of dihydroxy- stearic acid in the mixed fatty acids, a correction must be made for its solubility (0.0185 g per 100 ml at 0" C) in the known volume of ethyl acetate used as solvent and wash liquor. CaZcztZations-Percentage of triacetyltriricinolein in the acetylated oil- = P. Ac.V. x 100 159.1 The percentage of ricinoleic acid in the mixed fatty acids of the oil is different from the above because of the change in molecular weight on acetylation. It can be calculated with sufficient accuracy by means of the following approximation Let 298p + 0.96 (100 - P) = y 352-7 Then, if the proportion of ricinoleic acid in the mixed fatty acids is R per cent.298P x 100 352.7~ This figure for ricinoleic acid is too high, since it includes the contribution of the two hydroxyl groups in the dihydroxystearic acid also present. Let the percentage of the latter = D. Then the actual percentage of ricinoleic acid = R - 2D. Oleic acid is determined by its iodine value after making allowance for the ricinoleic and linoleic acids. R = Saturated acids are determined by difference.44 RILEY [vol. 76 The proposed method for the analysis of castor oils has been tested by the examination of two mixtures of methyl ricinoleate, oleate and linoleate in known amounts, with the results shown in Table 111. TABLE I11 EXAMINATION OF MIXTURES OF KNOWN COMPOSITION Esters .. .. .. .. .. S.E. .. .. .. .. .. S.E. mean deviation .. .. S.E. acetyl compound . . .. S.E. mean deviation .. .. Acetyl value . . .. .. .. Iodine value . . .. .. .. Mixed acids: Iodine value . . .. .. Et zi a t 234 mp (180°/60 min.) . . Mixture 1 Mixture 2 found 308.7 183.1 144.4 86.0 0.16% 0.23 %, 89.5 38.4 Whence the component esters of the mixtures are- found % w . 1 Methyl ricinoleate . . .. .. 90.1 Methyl oleate .. .. .. 5.7 Methyl linoleate . . .. .. 4.2 calc . 310.2 183.5 144.3 85.5 - - 89.5 39.9 calc. 90.0 5.6 4.4 % (wt.)* found 306.6 189.4 131.3 90.0 0.08 % 0.07 % 93.9 82.9 found 80.0 11.2 8.8 % ( W e ) calc. 308-7 190.9 129.6 90.1 - - 94.3 83.3 calc. % d(g)* 9.9 9.2 * Calculated on the assumption that the composition of methyl oleate and linoleate concentrates are: saturated ester 1-6%, methyl oleate 97.9% and meth,yl linoleate 0.5%; and methyl oleate 12.1% and methyl linoleate 87-9y0 (wt.) respectively.An example of the analysis of a commercial cold drawn castor oil of B.P. quality by the proposed procedure is given below- Characteristics of alkali re$ned oil-S.E. 309.7; I.V. 82.9; Ac.17. 150.2; n:6 1.4778; n;O 1-4741; [cc]zO + 4-42’. The refined oil (50-57 g) yielded 0.244 g of unsaponifiable matter and 46.32 g of mixed fatty acids, which gave 0-1681 g of dihydroxystearic acid, m.p. 139” C (corrected weight 0.261 g). Mixed acids after removal of unsafionijable matter-1.V. 87-0; Ei& at 234 mp (after alkali isomerisation at 180” for 60 minutes) 48.5. The component fatty acids of the oil calculated from these results are-saturated 0.9; oleic nil ; linoleic 5-4 ; ricinoleic 92-6 ; dihydroxystearic 0.6 ; unsaponifiable 0.5 per cent.by weight. Further application of the method to castor oils from various sources will be reported See page 41. Acetylated oil, S.E. 180.5. elsewhere. The author thanks Professor T. P. Hilditch, F.R.S., for his criticism and guidance during this work. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. REFERENCES British Standard Specification 684 (1950) (first method, p. 42). Verley, A,, and Bolsing, F., Ber., 1901, 34, 3354. Britis, Standard Specification 684 (1950) (second method, p. 44). “Methods of Analysis of the Association of Official Agricultural Chemists,” 6th Edition, 1945, West, E. S., Hoagland, C. L., and Curtis, G. H., J . Biol. Chem., 1934, 104, 627. Wilson, H. N., and Hughes, W. C., J . SOC. Chem. Ind., 1939, 58, 74. Ogg, C . L., Porter, W. L., and Willets, G. O., I n d . Eng. Chetn., Anal. Ed., 1945, 17, 394. Eibner, A., and Munzing, E., Chem. Umschau, 1925, 32, 159. “Report of Sub-committee on the Determination of Unsaponifiable Matter in Oils and Fats,” Hilditch, T. P., Morton, R. A., and Riley, J. F’., Ibid., 1945, 70, 67. Brown, J. B., and Green, N. D., J . Amer. Chem. SOC., 1940, 62, 738. Grummitt, O., and Siedschlag, K. G., J . Amer.. Oil Chem. SOL, 1949, 26, 690. Hilditch, T. P., and Pathak, S. P., Proc. Roy. :TOG., A , 1949, 198, 323. Washington, D.C., p. 501. Analyst, 1933, 58, 203. THE UNIVERSITY LIVERPOOL May, 1950