TOBlS OIL BROMIDE FILMS ETC. 69 Oil Bromide Films and their use in Determining the Halogen Absorption of Oils. BY HAROLD TOMS M.Sc. A.I.C. (Work done finder the Analytical Investigation Scheme.) (Read at the Meeting December 7th 1927.) IT has already been shown that the “insoluble bromide” of linseed oil (AKALYST, 1024 49 77; Ann.ua2 Refiorts 1924 158) is identical with that obtained from other vegetable drying oils (ANALYST 1926 51 387). It seemed desirable, therefore to establish as far as possible the constitution of this compound for it is very probable that the unsaturated compound from which it is derived is very largely responsible for the drying properties of these oils. The first part of this paper is an account of the efforts to establish the con-stitution of this substance and the second is an account of an accurate micro-method for determining the unsaturation value of oils.This method which appears to be of general applicability was evolved to overcome certain difficulties that arose in the course of the work mentioned above. Incidentally new light has been thrown on the peculiar behaviour of tung oil. PART I. THE CONSTITUTION OF THE INSOLUBLE BRoMIDE.-Analysis shows that the insoluble bromide obtainable from most of the drying oils is probably dilinoleno-linolin-bromide-a conclusion supported by the independent work of Eibner and Schmidinger (Chew. Umsclzazt 1923 30 293). Hydrolysis ought therefore to give two molecules of hexabromostearic acid and one molecule of tetrabromo-stearic acid. Although the first acid has been found the latter for a long time escaped detection ; the probable reason for this has only recently been discovered.After many failures it was decided to debrominate the bromoglyceride and to attempt the isolation of the linolic acid by hydrolysis of the unsaturated product. This is somewhat remarkable in view of the ease with which a- and p-linolenic acids are obtained by the debromination of hexabromstearic acid (cf. Erdmann and Bedford Ber. 1909, 42 1324). The usual methods were tried and failed. Even the remarkable method involving the use of sodium iodide in ethyl aceto-acetate solution which was used with success by Grun (Z. angew. Chetn. 1916 29 37) for debrominating act-dilinolin bromide failed in this case. Metallic sodium and metallic calcium in alcohol rcmowd the bromine but produced a solid hydrogenated fat.The debromination proved extremely difficult 70 TOMS OIL BROMIDE FIL&lS AND THEIR USE Partial success was ultimately obtained by means of zinc dust and 95 per cent alcohol. The reaction is very slow indeed; only after 80 hours of continuous boiling was any product obtained and even then. the yield was prohibitively small Attempts at synthesis have been entirely abortive for it was found that the acid chloride or bromide of hexabromostearic acid cannot be obtained and the three methods available for preparing glycerides of known structure viz. that of Fischer (Ber. 1920,53 (B) 1589) that of Bergmann (Ber. 1921,54 (B) 936) and that of the present author ( J . Ckzem. SOC. Trans. 1921,119 1035-1040) all require acid chlorides or bromides of reasonable purity.It was found that phosphorus chlorides and phosphorus bromides under various conditions carbonise the bromo acids whilst thionyl chloride gave only a small yield of the acid chloride which, since it could not be separated from the unchanged acid was useless for the delicate selective esterification methods mentioned above. EXPERIMENTAL. THE DEBROMINATION OF DILINOLENO-LINOLIN BRoMIDE.-The bromo glyceride (5 grms.) was suspended together with zinc dust (10 grms.) in 95 per cent alcohol (100 c.c.) and boiled continuously for 80 hours. After cooling the mixture was filtered with the aid of the pump. The residue contained much unchanged substance. The filtrate which contained a few suspended droplets of oil was concentrated poured into water extracted with ether and the ethereal extract dried over anhydrous sodium sulphate.On removing the ether a few drops of a deep yellow oil resulted. This oil could not be purified by distillation as it re-sinified on heating even at low pressure. This experiment was repeated several times on separate preparations of dilinoleno-linolin bromide. In each case the resulting compound was a yellow oil which hardened rapidly in the air giving a tough and somewhat crystalline film. It had a specific gravity of 0.957 at 10” C., and was soluble in ether acetone benzene and chloroform. Unsaturation Value.-The smallness of the amount available and the con-sequent need for the conservation of the material suggested the idea of “drying” a small weighed amount of the oil in an atmosphere of bromine and from the increase in weight determining the unsaturation value.The brominated film thus produced would be still available for crystallisation and subsequent deter-mination of the melting poing. (Bromine content of film = 55.6 per cent.; theory for dilinoleno-linolin bromide = 59.4 per cent). The film was crystallised from tetralin and acetone and gave a small amount of product which according to the mixed melting-point test was identical with the original bromoglyceride. In a later experiment a few drops of the oil were brominated in ethereal solution and the precipitate after crystallisation proved to be identical with the original compound. The filtrate from the bromination liquor left on evaporation a small amount of a bromine-containing oil which would not crystallise.This behaviour is comparable with that recorded by various in-vestigators for hexabromo- and tetrabromostearic acids I N DETERMINING THE HALOGEN ABSORPTION OF OILS 71 Hydrolysis ofthe Oil.-One grm. of oil was hydrolysed with 10 C.C. of 20 per cent. alcoholic potassium hydroxide solution. After two hours the acids were isolated in the usual manner. The mixed acids were dissolved in ether and hromi-nated and the resulting precipitate was found to be hexabromostearic acid. On evaporating the solvent from the filtrate there resulted a very viscous oil which up to the present has resisted all attempts to crystallise it. It was purified by dissolving it several times in ether and cooling the solution whereby further small amounts of hexabromostearic acid were removed.Finally a liquid product resulted which contained 56 per cent. of bromine* (Stepanow’s method) showing that probably it was mainly a liquid form of tetrabromostearic acid. It may be identical with that recently described by Santiago and West (Phil. J . Sci. 1927, 32 41-52). PART 11. GRAVIMETRIC DETERMINATION OF THE BROMINE VALUE.-During the attempts t o remove bromine from dilinoleno-linolin bromide it became necessary to deter-mine the unsaturation value of drops of oil far too small in amount to give results by the usual Wijs method. It therefore occurred to me to allow thin weighed films of oil to “dry” in an atmosphere of bromine. From the increase in weight the bromine value can be obtained and this by multiplying by the factor-At.Wt. of Iodine At. Wt. of Bromine’ should give the ordinary iodine value. The apparent success of this method as indicated in Part I suggested the possibility of a simple gravimetric method for finding the unsaturated value of minute quantities of oils in general. It was, therefore tested on a large number of oils and it was found that not only do the values agree well with those obtained by the usual methods for the majority of unoxidised oils but also that the appearance of the film is in many cases character-istic of the particular oil. METHOD AND APPARATUS.-A single drop of oil (0.02 to 0.03 grm.) is spread in a thin film about 0.2 mm. thick on a weighed micrascope slide and placed in a wide tube closed at each end with a waxed cork and containing a boat in which are placed a few drops of bromine.After 20 to 30 minutes the slide is taken out and the excess of bromine is removed either by heating to 50-60” C. or by a current of warm air. From the increase in weight of the slide the bromine value can be calculated. For the less unsaturated oils larger amounts are needed and then it is advisable to use a larger plate with a similar one for a counterpoise. The method is characterised by extreme simplicity for it eliminates all errors due to graduated flasks and burettes; furthermore it reduces manipulation to a minimum and hence is capable of great accuracy. * Theory for tetrabromostearic acid is 63.3 per cent 72 TOMS OIL BROMIDE FILMS AND THEIR USE The characteristics of the brominated films are indicated below : OIL.Linseed Linolenic acid Boiled linseed Candle-nut Rubber seed Perilla Poppy seed Tung Cotton seed Soya bean Menhaden Sea elephant Whale Castor Apricot kernel CHARACTER OF FILM. . . Similar to above. . . . . . . . . Hard gritty film wrinkled surface not very glossy colourless. Smooth brown glossy film. Inclined to wrinkle but soft sticky and glossy. Smooth glossy very sticky and viscous. Wrinkled hard no gloss but tends to run into streaks and lakes rather Varnish-like very soft and sticky. Very high gloss but soft and sticky no tendency to wrinkle. Yellowish, Extremely soft and sticky. Slightly glossy but soft. Opaque coarsely wrinkled dull surface.Shiny transparent film soft and sticky. Transparent film covered with minute granules. than to remain even. . . . . . . . . . . . . . . Transparent droplets. . . Opaque droplets. EXPERIMENTAL. (A) DUPLICATES SHOWING THE DEGREE OF ACCURACY OBTAINABLE. (i) Olive Oil : (a) Oil taken 0-0344 grm.; bromine absorbed 0.0185 grm. Bromine (b) Oil taken 0.0428 grm.; bromine absorbed 0.0224 grm. Bromine value = 53.8. value = 53-6. Calculated iodine value = 85.3 Value found by independent investigator = 85.4 (ii) Menhaden Oil : (a) Oil = 0.0601 grm.; bromine absorbed 0*0665* grm. Bromine (b) Oil taken 0.0242 grm.; bromine absorbed 0.0268. Bromine value = 110.6. value = 110.8. (iii) Tea seed Oil : (a) Oil taken 0.0276 grrn.; bromine absorbed 0-0157 grm.Bromine (b) Oil taken 0-0170 grrn.; bromine absorbed 0.0096 grm. Bromine value = 56.9. value = 56.5. Calculated iodine value = 90.2. Experimental , ?? = 90.0. * The film showed no change in weight after 5 days 1 OIL BROMIDE FILMS. 2 1. Perilla Oil. 2. Tung Oil. 3. Linseed Oil. 4. Menhaden Oil. (Magnijed 2& dinmeters. IN DETERMINING THE HALOGEN ABSORPTION OF OILS '13 (B) RESULTS OBTAINED WITH VARIOUS OILS. Oil. Linseed I (Cah.xtta) . . Linseed TI (Baltic) . . Boiled linseed Linolenic acid Perilla Rubber seed I Rubber seed I1 Candlenu t Sova and linseed mixture Tung Pgppy seed Soya bean Sunflower seed Oil. Castor Colza Cotton seed Maize Oil. Tea seed I Tea seed I1 Apricot kernel Olive Arachis I Arachis I1 Coconut (a) (b) Oil.Menhaden Sperm Dugong Sea elephant Penguin Seal Sardine Sea leopard Weddell seal Seal Drying Oils. Bromine value. 115.2 126.2 76.6 133.9 130.0 88.75 88.52 104.6 142.2 108.5 Calculated iodine value. 182.9 20 1 -8 121.6 212.2 206.4 141.2 1406 166.1 225.5 172.3 88.7 141.0 82.94 131.7 81.8 129.8 Semi-drying Oils. Bromine Calculated value. iodine value. 62.6 99.7 63-4 100.6 62.4 99.5 66.4 105.3 Non-Drying Oils. Bromine value. 56.5 * ' ( 56.9 50-8 65.5 53.6 61.8 60.2 5.5 4.79 Calculated iodine value. 90.2 80.8 104.0 85-3 98.4 95.5 8.7 7.62 -Fish and Marine Animal Oils. Bromine value. . . 110.6 57.2 . . 32.4 67.4 60.4 96.4 .. 103.8 77.6 82.6 . . 91.8 Calculated iodine value. 184.0 90.8 51.5 107.0 96.0 153.0 164-8 123.j 131-2 146.0 Observed iodine value (Wijs). 182.6 202.5 115.2 210.0 205.4 140.7 140.3 165-6 162.0 172.6 136.4 130.3 129.2 Observed iodine value (Wijs). 105.2 10643 101-5 106.2 Observed iodine value (Wijs). 90.0 81.0 101.3 83.9 99.7 95.8 7.2 7 -2 -Observed iodine value (Wijs). 180.0 90.5 52.5 107.3 89.3 153.1 121.0 127.1 13743 150. 74 TOMS 'OIL BROMIDE FILMS AND THEIR USE It will be seen from these results that the iodine value calculated from the bromine absorption as determined by this micro method agrees closely with the iodine value determined by the Wijs method in the case of most unoxidised oils.Oils containing oxidised acids (8.g. boiled linseed oil) give as was to be expected, a higher absorption with bromine than with the Wijs reagent. Some of the fish and marine animal oils were old specimens in which some decomposition of the glycerides had probably taken place. On the other hand castor oil containing ricinoleic acid gives a lower halogen absorption with bromine than with iodine chloride. This may be a distinguishing characteristic between oleic acid and ricinoleic acid. The high result in the first determination with coconut oil is obviously due to experimental error; in the second determination a much larger amount (0-06 grm.) of the same fat was used on a plate 6 inches by 2 inches. THE SPECIAL CASE OF TUNG OIL.-On reference to the first table of results (vide szCp~a) it will be seen that in the case of tung oil there is a difference of over 63 units between the calculated and observed iodine values.This difference, being far too great to ascribe to experimental error rendered it desirable that the matter should be investigated in as many cases as possible. The results of this investigation are summarised in the table below. Ttmg Oil. Sample. I I1 I11 IV V VI VII Wijs value Br value, experimental. experiment a1 . 165-2 139.3 161.7 (137.2 ( 136.6 169.6 146.3 167.8 143.4 168.3 146.2 154.9 128.3 168.4 145.6 Wijs value, calculated. 221.3 217.9 2 17.0 232.4 227.8 232.1 203.8 231.3 Ratio of experimental Wijs value.Calc. Wijs value to 1.34 1.35 1 *34 1 -37 1-36 1.37 1.32 1-37 These results are probably due to the peculiar structure of the characteristic acid of tung oil. The unsaturated acids of the ordinary oils contain groupings of double bonds such as CH,.CH=CH.CH,.CH= CH.CH and CH,.CH =CH.CH,.CH = CH.CH,CH = CH.CH, i.e. non-conjugated systems and hence each double bond acts independently of the others. It seems that both bromine vapour and iodine monochloride in solution act in the same way on such structures that is they give direct addition products which are completely saturated. In the case of iodine monochloride the product contains iodine and chlorine in amounts proportional to the atomic weights of these elements thus-X. CH,.CH x.CH2.CHCl and X.CH,.CHI Y. CH,.CH Y. CH,.CHI Or Y.CH,.CHCl - 11 gives with ICl I I N DETERMIKISG THE HALOGEN ABSORPTION OF OILS 75 a-Elaeostearic acid which is present to the extent of 60 per cent. or more, in tung oil has an entirely different structure thus-Such a conjugated structure would normally react according to Thiele's theory-C4H9.CH = CH.CH = CH.CH = CH. (CH,),.COOH. (i) -CH=CH.CH=CH.CH=CH-++, (ii) -CHX.CH=CH.CH = CH.CHX-+X, gives -CHX.CH=CH.CH =CH,CHX-gives -CHX. CHX. CH = CH. CHX. CHX.-(iii) -CHX.CHX.CH=CH.CHX.CHX-+X, gives -CHX. CHX.CHX.CHX. CHX. CHX.-and hence the last two atoms to enter the system would experience a definite retardation. Thus the usual method of determining the iodine value by contact with Wijs solution for two hours may be insufficient for complete saturation.In the case of pure a-elaeostearic acid (Boeseken Rec. Trav. Ckim. Pays-Bas, 1927 46; ANALYST 1928 54) it was shown recently that the addition of iodine monochloride takes place in three stages; the first is complete in a few minutes the second in half-an-hour and the third in six days. Incidentally it is a matter of common observation that tung oil is the only one which during a Wijs determination is characterised by the precipitation of iodine and it is very probable that in this case the final product is a chloro-compound. If the results obtained on the free acid can be applied to the oil then the ordinary iodine values correspond to a two-thirds saturation of the glyceride of a-elaeostearic acid. From the results recorded in this paper it appears that bromine vapour acts very differently from iodine monochloride in solution or even bromine itself in solution, which appears to saturate only two of the three pairs of double bonds (cf.Ber., 1925 58 216; 1916 59 1319) whereas bromine vapour appears to saturate all bonds simultaneously regardless of whether the system is conjugated or other-wise. If this is so then the more nearly the composition of any sample of tung oil approaches 100 per cent. of the triglyceride of a-elaeostearic acid the more nearly will the ratio calculated iodine approach the limiting value of 1.5. experimentaliodine value Attempts were made to make the vapour of iodine monochloride act on tung oil. At room temperatures a greenish film resulted giving an increase in weight of 7 to 10 per cent.; at 100" C. a brown flaky film which did not stick t o the glass, resulted. The increase in weight for a single sample varied from 60 to 70 per cent. Obviously iodine monochloride vapour is useless in this connection. CALCULATION OF THE PERCENTAGE OF a-ELAEOSTEARIC TRIGLYCERIDE IN TUNG OIL.-In the following calculation it is assumed that tung oil contains two types of substances. Type A consists of those which contain non-conjugated systems of double bonds and therefore gives the same iodine value regardless of the method used. Type B is the glyceride of a-elaeostearic acid which gives (i) only two-thirds its normal or theoretical iodine value with Wijs's solution in two hours; but (ii) gives its full value by the bromine vapour method and henc 76 TOMS OIL BROMIDE FILMS AND THEIR USE the full calculated iodine value.The molecular formula for a-elaeostearic tri-glyceride is C5,HgzOs and its molecular weight is 872. The theoretical iodine value after two hours corresponding to the addition of 12 atoms of iodine is 174.8 per cent. Let this equal n per cent. Also the complete iodine value corresponds to the addition of 18 atoms of iodine and is 262.5 per cent. Let this be m per cent. Further let the substances of type A have an iodine value of x per cent. under both sets of conditions. Suppose in any experiment under condition (i) the mixture (A+B) absorbs 9 per cent. of iodine and under condition (ii) it absorbs q per cent. of iodine then we have and but nB = & (A+B) m-n = 2626-174.8 = 87.7 A - 87.7- (4-9) B - - q-9 - -hence for any sample A/B can be found.and percentage of B = (4-p)*100 87.7 ’ where q is the iodine value of the sample calculated from the bromine value, and9 is the iodine value found by the usual Wijs method. The amount of a-elaeostearic triglyceride in each of the tung oils used as calculated from the table on page 74 is: Per Cent. I 64 -0 I1 63.2 I11 71 *7 IV 68.5 Per Cent. V 7243 VI 85.8 VII 7143 SUMMARY AND CONCLUSIONS.-~. The insoluble bromide obtainable from most of the drying oils has with difficulty been debrominated and the presence of hexabromostearic acid confirmed whilst the presence of tetrabromostearic acid has been rendered more probable by the isolation of a thick oil which appears to be a liquid form of this acid.2. A rapid and accurate micro-method for determining the unsaturation value of non-volatile oils in general has been devised and tested in a large number of cases. The method involves the preparation of brominated oil films and these are found to be largely specific in appearance. The values for most unoxidised oils obtained agree well with the iodine values as usually determined. The behaviour of tung oil has been found to be unique and a method of determining the active constituent of this oil has been devised. 3 IN DETERMINING THE HALOGEN ABSORPTION OF o m 77 The author wishes to acknowledge his indebtedness to the Director of the Imperial Institute for gifts of oil; also to Mr. C. Ainsworth Mitchell and to Mr.A. Chaston Chapman for gifts of oil and for checking some of the results here recorded. He would also like to thank Mr. T. J. Ward for making photomicrographs slides and prints of some of the oil bromide films. BIRKBECK COLLEGE, UNIVERSITY OF LONDON. DISCUSSION. The PRESIDENT congratulated Mr. Toms on his further contribution to the chemistry of the bromine compounds of fatty oils. In the past he had often experienced difficulty in getting an iodine value for tung oil which agreed with that found by other chemists. The explanation of this difficulty was to be found in the work of Mr. Toms which would thus have a direct practical value. Mr. C. A. MITCHELL recalled the fact that Hehner many years ago had brought before the Society a gravimetric bromine method (ANALYST 1895 20 49).This had given fairly good results under strictly empirical conditions but it had the drawback that the bromine and oil were mixed in a solution and that the heat required for complete removal of the solvent caused the bromine to combine by substitution as well as addition and thus the weight of the oil bromide never became quite constant. In Mr. Toms’s micro-method constant weight was obtained without difficulty. The principle of the method might be tried with other reagents in the form of a vapour such as chlorine osmium tetroxide and especially sulphur trioxide. . The method might also give useful results in the examination of heavy lubricating oils and vaseline. Mr. M. S. SALAMON said that he too had had difficulties with regard to the iodine value of tung oil and he welcomed the new method since it promised to put an end to disputes.He suggested the application of the method to the examination of resins and especially of shellac. Mr. E. HINKS asked for information as to the thickness of the films. Mr. TOMS replying said that he would try the various reagents that had been suggested on fatty oils and eventually on resins. It was essential that the film should be as thin as possible; otherwise a skin formed on the surface and protected the oil beneath from the action of the bromine. His practice was to spread a single drop of the oil with the finger so as to cover about two-thirds of the area of the slide. ADDENDUM.-sinCe this paper was in print my attention has been drawn to a paper by Sabalitschka and Dietrich (Plzarm. Ztg. 1924 69 425) in which a method based on the same principle (absorption of bromine vapour) is described. As quantities of 0.1 to 0.4 grm. of the oil are used on plates 14 cm. by 5 cm. and from 2 to 3 hours’ heating are required to expel the excess of bromine it is not surprising that these authors have found that some substitution occurs and that their results are uniformly higher than those obtained by the Hub1 method even in the case of solid fats and non-drying oils. As is mentioned above the oil film must be sufficiently thin to prevent the formation of a pellicle on the surface, which impedes the removal of the excess of bromine; this is not readily practicable with quantities of 0.1 grm. and upwards.-H. TOMS