March, 1936
CHEMICAL
EXAMINATION OF THE SEED OF h X E S BALSAMEA (L) MILLER
[ CONTRIBUTrON FROM THE DEPARTMENT OF CHEMISTRY, UNIVERSITY
523
OF WISCONSIN]
The Chemical Examination of the Seed of Abies Baisamea (L) MillerlJ BY
s. R.BENSON AND H.N. CALDERWOOD
perature, approximately 25'. The extract yielded a volatile oil with the following characteristics: b. p. 165-180" Although the balsam fir, Abies balsamea (L) (740 mm.); sp. gr.2626, 0.8706; #"D, 1.4721; [(Y]*~D, Miller, grows over a large area of North America3 -25.32. In addition to the volatile oil, glucosides, tanand is rut extensively for pulpwood the only nins and resins were recovered, which will be investigated chemical examination of its seeds is the work of E. and reported in later publications. The husk and endosperm, separated readily in the dry von Tubeuf4 giving one photomicrograph of a oleoresin-free seed, represented 16 and SO%, respectively, cross-section of the seed showing the starch, pro- of the total seed. The endosperm after grinding was tein and fat content as determined by histological extracted with petroleum-ether, b. p. 60-65". and yielded methods. This paper is the first report in the after complete removal ot the solvent a clear golden literature upon the proximate composition of the yellow bland fatty oil. The oil yield based upon the endosperm material is 19.8%. seed and the characteristics of its fatty oil. Fatty Oil.-The fatty oil was examined by recognized The seed contains noticeable quantities of oleo- proceduress and these results are summarized in Table 11.
Introduction
resin, situated between the placenta scale tissue of the wings or husk and the endosperm, and raised a problem not encountered in securing the common fatty oils from seeds.
Experimental Collection of Material.-A total weight of 160 kg. of cones was collected which yielded 35 kg. of pure seed. Composition of Seed.-The proximate composition of the seed is shown in Table I.
TABLE I COMPOSITION OF SEEDSOF ABIESBALSAMEA Moisture, % Crude fiber, $& Crude protein, o/; Total ash, % (a) Soluble ash, % (b) Insoluble ash, Total water-ethanol extract, % (a) Non volatile fraction, % 1. Water soluble, % 2. Water insoluble, % (b) Volatile fraction, % Petroleum ether extract, %
As received
Moisturc-free basis
8.61 32.14 8 62 1.72 0 62 1.16 21.9 17.5 4.06 13.44 4 4 10.69
35.13 9.42 1 88 0.67 1.26 23 9 19.1 4 43 14.69 4.8 11 68
...
The best menstrum for obtaining the oleoresin free from fatty oil was found to be 75% alcohol. The air-dried unground seeds were extracted by percolation a t room tem(1) A portion of a dissertation submitted to the Graduate School of the University of Wisconsin by S. R Beeson in partial fuliiLment of the requirements for the Ph D.degree, June, 1935. ( 2 ) Presented in part before a joint meeting of the Wisconsin Sections of the American Chemical Sockty and the Wisconsin Academy of Sciences, Arts and Letters, Beloit, Wisconsin, April 27, 1935 (3) (a) Raphael 208, "Balsam Fir," U. S Dept Agr Bull, 66 (1914); (b) Geo. B Sudworth, "Check List of Forest Trees of North America, Their NafnRs and Ranges," U S Dept of Agr Miwl Circular, 99, 29 (1927). (4) E. von Tuheut, N u l w w i s s Z F o r 8 l - L u d w I 16, 231 (1917)
TABLIE I1 CONSTANTS OF FATTY OIL PROM BALSAMFIR SEED Sp. gr. 25"/25O 0.9279 Ref. index a t 25' 1.4783 Solidifying point -37.OOC. Titer test -17.5'C. Acid value 0.88 Iodine value (Nanus) 141.0 Saponification value 185.6 Acetyl value 15.4 Per cent. unsaponifiable 4.8 Reichert-Meissl number 0.53 Polenske number .21 SoiuMe acids trace Insoluble acids Per cent. 83.2 Thiocyanogen value 93.3 Saturated acids (lead salt-ether method) Per cent. 1.6 Melting poiat, "C. 58 Iodine value 14.0 Mean molecular weight 282.0 Unsaturated acids (lead salt-ether method) Per cent. 80.8 Iodine value 155.5 Mean molecu!ar weight 279.0 The unsaturated acids were examined using the bromination method of Eibner and Muggenthdlero and the thiocyahogen method of Kaufman.' These results are repoiTed in Table 111. A further study of the individual unsaturated acids was made by converting them into hydroxy acids by the method of Hazura.* The oxidation products upon separation yielded the hydfoxy acids reported in Table N. ( 5 ) A 0 A C Methods of Analysis, 3d ed , Washington, D C , 1930, pp 314-330 (6) Eibner and Muggenthaler, Farbcn Ztg , 18,131 (1912). (7) Kaufman, Z -hTahr Ccnussm , 81, 15 (1927). (8) Halure, Monatsk , 8 , 147, 156, ZBO (1887). 8, 180, 1'30 46'1 478, 944,947 (1888), 10, 190 (1880).
524
s. R. BENSON AND H. N.CALDERWOOD
Vol. 58
TABLEI11 OF THE UNSATURATED ACIDS EXAMINATION
with melting points identical with those obtained by Nicolet and Coxlo for the tetrahydroxy acids Bromine Thiocyanogen prepared from a single tetrabromostearic acid, the Acid method method structure of which has been determined by W. C. Linolenic acid, % 3.1 5.0 Snit." If it is assumed that the oxidation of lin179 Hexabromide, m. p., "C. Bromine in hexabromide, % 63.0 oleic acid with permanganate proceeds in a man54.6 59.7 Linoleic acid, 70 ner similar to that of oleic acid wherein a single Insoluble tetrabromide, m.p.. "C. 107-109 olefinic acid yields 'but one hydroxy derivative, Bromine in tetrabromide, % 54.0 the fatty oil of the balsam fir seed contains two 40.0 33.3 Oleic acid, % stereoisomeric linoleic acids. Further work upon TABLEIV the structure of these was in progress when the HYDROXYDERIVATIVES ISOLATEDBY PERMANGANATEpaper by Green and Hilditch12appeared wherein OXIDATION they state all seed fats contain but a single linoNeutral equivalent Acid M. P., 'C. Obsd. Calcd. leic acid. The results of these studies will appear Dihydroxys tearic 130 315 316 later. Tetrahydroxystearic 152 349 348 The single dihydroxystearic acid obtained Tetrahydroxystearic 173 348 348 showed, by its melting point and degradation Hexahydroxystearic 174 384 380 Hexahydroxystearic 200-203 . .. ... products, the oleic acid present in the oil to be the 9-octadecenoic acid. Discussion It is a pleasure to express our gratitude to D. C. Van Ostrand and Bruce G. Buell of Messrs. Although the fatty oils from seed of the botanithe Patten Timber Company, Amasa, Michigan, cal family Pimceae are repeatedly reported in the for their generosity and assistance in obtaining literature to have a terebinthic odor and taste, the material used in this investigation. only a few genera, including Abies, have oleoresin vesicles in the seed? In this paper a method is deSummary scribed for separating the oleoresin and fatty oil in Abies balsamea seed. The fatty oil was found 1. This paper is the first chemical examination to be glycerides of stearic, linolenic, oleic and of the seed of Abies baZsamea (L) Miller reported linoleic acids. The mean molecular weight and in the literature. melting point of the saturated acids demonstrated 2. The seed of Abies balsamea contains on the them to be essentially stearic. moisture free basis 19.8% oleoresin and 11.7% of The bromide and thiocyanometric reactions fatty oils. have enabled a quantitative study of the unsatu3. These substances as far as can be deterrated acids of the fatty oil. The values, shown mined by chemical methods are located in sepain Table 111, obtained by the two methods are rate parts of the seed; the oleoresin between the not identical but are however of the same order parchment-like tissue of the wings and the testa, of magnitude. The thiocyanometric method is the fatty oil inside the testa and distributed largely mathematical and serves but to give the throughout the endosperm. relative percentages of acids of the oleic, linoleic 4. The common physical and chemical conand linolenic type. The bromide studies in addi- stants of the fatty oil have been determined. tion to giving the relative amounts $f these differ5. The fatty oil has been shown to be essenent types enabled the identification of the linolenic tially glycerides of stearic, oleic, linoleic and linacid in the oil but did not permit a clear cut sepa- olenic acids. ration of the linoleic and oleic acids and therefore 6 . The relative amounts of acids of the satuthe hydroxy derivatives of these acids were pre- rated, oleic, linoleic and linolenic types have been pared. determined. The hexahydroxy- and hexabromostearic acids 7. The linoleic acid present was shown to be obtained proved the linoleic acid to be 9,12,15- 9,12,15-octadecatrienoic acid. The oil appears octadecatrienoic acid. to contain two isomeric linoleic acids which Two tetrahydroxystearic acids were obtained (lo) Nicolet and Cox, THISJOURNAL, 44, 144-52 (1922). (9) C. von Tubeuf, Nalurwiss. 2. Porsl-Landw., 14, 358 (1916), and 16, 18 (1917).
(11) W C. Smit, Hec. trav. chim., 49,539-51, 675-85 (1930). (12) Green and Hilditch, Biochem. J . , 19, 1552-63 (1935).
March, 1936
525
THECONSTITUTION O F DIBARBITURIC ACID
yielded tetrahydroxystearic acids with melting points of 152 and 173'. The oleic acid present
has been proven to be 9-octadecanoic acid, RECEIVED JANUARY 2, 1936
MADISON,WISCONSIN
[CONTRIBUTION FROM THE DEPARTMENT OF CHEMISTRY, YALE UNIVERSITY]
Researches on Pyrimidines. CLI. The Constitution of Dibarbituric Acid' BY ROLLIND. HOTCHKISS AND TREAT B.
JOHNSON
CHsN-CO HN-CO OC-NH When A. von Baeyer reported in 1864 the prepal i I 1 I / ration of barbituric acid, he described, among OC CBrz OC CH HC CO I 1 i l i I l l its other properties, its conversion by heat into an CHsN-CO HN-C-0-C-NH insoluble new acid, bibarbituric acid.2 He also I I1 reported the isolation of several salts and two formula I1 would not have the stability toward bromine derivatives of this new pyrimidine. In hydrolysis that dibarbituric acid does, nor would the seventy years which have elapsed since that it give enol reactions and form salts when all four time, barbituric acid and'its derivatives have been of the cyclic nitrogen positions have been methylstudied extensively. Dibarbituric acid, however, ated. The barbituric acid nuclei must, therefore, has received only passing mention by investigabe linked through carbon atoms, and, since the tors who were dealing with other substance^.^^^ linkage is easily cleaved by bromine water under The investigation reported in this paper is conthe conditions mentioned above, it is probably a cerned with the determination of the structure of double bond. It seems possible, therefore, to exthis difficultly soluble pyrimidine compound. clude all other formulas for dibarbituric acid exVon Baeyer postulated that dibarbituric acid cept I11 and IV, and their respective tautomeric can be represented by the formula CeHeNrOs, modifications. corresponding to the union of two molecules of HN-CO HN-CO HN-CO OC-NH barbituric acid with loss of one molecule of water. 1 1 I I OC C=C CHI OC C HzL LO This empirical formula is supported by the results I ' I 1 I now obtained. The literature does not record HN-CO HN-co HN-CHn C-NH any attempt to give a structural expression, exI11 IV cept that of Conrad and G u t h ~ e i t who , ~ ~ proDefinite experimental evidence rendering it posposed the formulation 111. sible to decide between formulas 111 and IV was On treatment with dimethyl sulfate and poobtained through the hydrogenolysis of N-tritassium hydroxide, dibarbituric acid yields a methyldibarbituric acid. Repeated attempts to trimethyl and a tetramethyl derivative. When reduce dibarbituric acid and its N-methyl homothe latter is warmed with bromine water it underlogs by a variety of methods failed to result in the goes bromination and cleavage into two moles of preparation of dihydro derivatives. At high tem1,3-dimethyl-5,5-dibromobarbituric acid (I). This peratures and pressures, however, catalytic reducfact indicates that the dibarbituric acid molecule tion of the N-trimethyl derivatives in dry dioxane contains two intact barbituric acid nuclei united solution brought about a smooth hydrogenolysis. through a linkage not involving the nitrogen On the basis of the results obtained from this reatoms. A di-enol anhydride, as represented by
; I\,
(1) From a thesis presented by Rollin D. Hotchkiss to the Graduate Faculty of Yale University in June, 1935, as partial fulfilment of the requirements for the degree of Doctor of Philosophy. (2) A. von Baeyer, Ann., 180, 145 (1864). Throughout this paper the name dibarbituric acid will be used in conformity with modern nomenclature. (3) (a) M. Conrad and M. Guthzeit, Bcr , 16, 2846 (1882); (b) C. Matignon, A n n . chim. phys., [ 6 ] 28, 292 (1893); (c) H. Biltz and H. Wittek, Ber., 64, 1035 (1921); (d) H. Biltz and T. Kohler, ibid., 66, 2482 (1923). (4) It seems possible that insoluble products observed by Wood and Anderson, J . Chcm. Soc., 96, 979 (1909), and Grimaux, Bull. sot. chim., 81, 146 (1879), might be impure dibarbituric acid, PIthough these authors did not recognize them as such.
(6) (1') CHaN-CO (2')
I O L (5')CI 1
(3') CHsN-CO
(6')
(5)
OC-NCHs
I t HzC CO I I
-=C-NH (4')
(1)
HZ
(2) (3)-
(4)
V CHaN-CO
I I OC CHz I I
CHaN-CO
VI
OC-NCHa
+ HnCI I
I t
CO
HzC-NH
VI1
