56 PERKIN AND PHIPPS: NOTES ON SOME By ARTHUR GEORGE PERKIN, F.R.S., and SAMUEL Prrr~rs. TIIIS investigation contains an account of the yellow colouring matters which are present in the flowers of I’mnus spinosa and the Japanese dyestuff ‘; Fukugi,” 1,ogether with some results which have been obtained during the further study of morin, myricetin, hesperitin, and curcumin. T h f e C o Z o u ~ i ~ t g A f u t t e ? - of t h e FZoweys of 2jrunu.s S 21 i 12. 0 S C(. Some years ago, while examining t h e dyeing properties of certain natural products, it was noted t h a t some quantity of yellow colouring matter exists as a gluooside in the flowers of Yrzcnus spinoscc or common blackthorn, A cursory examination carried out at the time resulted in the isolation 05 a product which was apparently a mixture, because its acetyl derivative did not melt sharply, the meltingpoint beingNATURAL COLOURING MATTERS.57 very low for this class of substance. The attempts then made to effect a separation were not successful, but more recently, i n t h e hope t h a t a new colouriag matter might be present, the subject has been again investigated, and the difficulty overcome in an exceedingly simplo manner. I n order t o facilitate the work, an extract of these flowers prepared by Merck of Darmstadt was chiefly employed. Two hundred grams of the extract dissolved in 2 litres of water were treated with 50 C.C. of hydrochloric acid, boiled for three hours, and, after cooling, extracted with etber to remove the colouring matter, which was thus obtahed as a brownish-yellow, crystalline mass (3.5 grams).This was purified by two or three crystallisations from dilute alcohol, and a portion then converted into the acetyl compound in order t o determine whether its melting point coincided with that of the corresponding derivative of some known substance of this class. It mas found t o melt somewhat indefinitely a t 133--134", and gave on analysis C = 60.09 ; H = 4.28, these figures indicating that i t was most probably a mixture. The main bulk of the colouring matter was now dissolved in a small quantity of boiling acetic acid, arid the crystals which separated on cooling were collected and recrystallised from the same solvent until no alteration in the melting point could be observed ; the yield was 0.S gram.Found C = 62.93 ; IT = 3.89. C,,H,,O, require:; C = 62.94 ; K = 3.49 per cent. The substance, which coiisisted of pale yellow, glistening leaflets melting a t 276', dissolved in dilute aqueous alkalis with a pale yellow c 010 rat i o ti. The ucetp? compound crj stallised from methyl alcohol in colourless needles which melted at about 116", resolidified at a higher tempera- ture, and melted again at lS1-182'. Found C = 60.77 ; 3 = 3.97. C',,HG06(C1,H,0), reqiiires C: = 60.79 ; H = 3.96 per cent. On fusing the colouriug matter with caustic alkali, phloroglucinol (m. p. 210') and p-hydroxybenzoic acid (m. p. 208--210°) were obtained. These facts, together with an examination of i t s dyeing properties, indicated without doubt that this colouring matter was kampherol.As this dyestuff has recently been found in at least four other plants, it is likely t o occur in many other vegetable species. The acetic acid mother liquors from which the kampherol had separated mere allowed to absorb moisture by exposure to the58 PERKIN AND PHtPPS: NOTES ON SOME atmosphere, which caused the deposition of a small quantity of a mixture of a kampherol and a second substance. This precipitate was removed, the filtrate treated with a small quantity of hot water, and the yellow precipitate, which slowly separated, was collected and purified by conversion into its cccetyl derivative. Found C = 58.67 ; H = 4.26. C1,H,O7(C,H,O), requires C = 58-59 ; H = 3.90 per cent. This substance, which melted a t 191°, was decomposed with acid in the usual manner, and the regenerated colouring matter crystallised from dilute alcohol.Found C = 59-56 ; H = 3.61. U,,U1,O7 requires C = 59.60 ; H = 3.31 per cent,. It formed glistening yellow needles soluble in alkali solutions with a yellow coloration ; on fusion with caustic pchash, it gave phloroglucinol (m. p. 210”) and protocatechuic acid (m. p. 193-195’). It was evidently quercetin, and to this colouring matter and kampherol the dyeing property of these flowers is evidently due. A cursory examination of the flowers of the violet (VioZcc odoratn) and the white clover (Trzj’olium repens) by a method similar to that employed above indicated in each case the presence of puercetin in the form of a glucoside. This colouring matter was recognised by the melting point of its acetyl derivative, and its decomposition products with caustic alkali, and in consequence of these observations a fuller investigation appeared unnecessary.The Japanese D y e s t u f f “Pukugi.” We are indebted to the kindness of Professor E. Yoshitake, of Tokio, for this material, wbich consists of the wood of a tree, and was obtained in the form of an almost colourless, coarse powder; it appears, a t least until recently, to have been employed to a considerable extent as a yellow mordant dyestuff, principally in the form of extract. A preparation of this kind was also procured and consisted of brittle, rectangular cakes (4” x 2y x ly) of a yellowish-brown colour, which weighe 1 approximately 410 grams. The coarsely powdered extract, dissolved in ten times its weight of water, was boiled with 100 C.C.of hydrochloric acid for two hours i n order to decompose the glucoside. A somewhat viscous precipitate of the impure colouring matter thus separated, which, when cold, was washed by decantation, drained on a tile, and allowed- to dry a t the ordinaryNATURAI, COLOURING MATTERS. 59 temperature. The product was extracted with boiling alcohol, the extract evaporated and poured into a large bulk of ether, which caused the separation of a resinous impurity, and on evaporating the ethereal liquid the colouring matter now obtained was of a much lighter colour. For further purification, it was dissolved in boiling alcohol containing a trace of acetic acid, lead acetate solution added, the resulting yellow precipitate removed, the filtrate evaporated to a small bulk and poured into ether.The pale yellow, etherealliquid was well washed with water, evaporated to dryness, and the viscid residue left for some days. Minute crystals slowly separated, which were col- lected, washed with a small amount of ether, and crystallised first from dilute ethyl alcohol and then from methyl alcohol until the melt- ing point was constant. The product frequently contained a trace of the lead compound, which adhered somewhat tenaciously, and was best removed by means of ether, i n which it dissolved with difficulty. Found, (i) C = 66.03 ; H = 6.04. (ii) C = 65.17 ; H = 3.97. Cl7H,,O, requires C = 65.3’7 ; H = 3.84 per cent. It consisted of a mass of minute, prismatic, canary-yellow needles, which melted at 288 - 2 9 0 O .When crystallised from dilute alcohol, the air-dried product contains one and a half molecules of %water of crys talli sation. Found, H20 = 7.56. This new colouring matter, for which the name fuhgetin is proposed, is readily soluble in hot alcohol arid dissolves in aqueous alkalis or in cold sulphuric acid with a pale yellow coloration. The solution in t h e latter solvfent, on heating, becomes dull violet-red, and finally assumes an orange-brown tint, and,on dilution with water, now deposits a brown, amorphous precipitate soluble in aqueous alkalis to a dull red solution. With lead acetate in alcoholic solution, an orange-yellow precipitate is formed, whilst alcoholic ferric chloride develops a brownish-black coloration; on the other hand, alcoholic potassium acetate gives no insoluble salt, and mineral acids do not react to form the usual com- pounds.When examined by Zeisel’s method, it was found to contain no methoxyl groups. Fukugetin readily dyes mordanted fabrics, and it was at once ob- served that the shades produced were almost identical, except as regards the iron mordanted portion, with those given by luteolin. C,,H2,0,2,3H20 requires H,O = 7-35 per cent. Cr. 81. sn. Fe. Fukugetin. Dull orange-yellow. Orange-yellow. Bright yellow. Olive-brown. Luteolin. Brown orange-yellow. 9 ) > ? 0 live-black.60 PERKIN AND PHIPPS: NOTES ON SOME Crystalline acetyl and benzoyl derivatives of this new colouring matter could unfortunately not be obtained either by the usual pro- ceses or by the pyridine method, Methylation also led to the formation of a viscous product, although i t is possible that these difficulties might have been surmounted if a larger quantity of substance had been available for experiment.Bromine Compozcnd.-One gram of fukugetin was added to a solu- tion of one gram of bromine in a little glacial acetic acid. After twenty-four hours, the product was drained on porous tile, ground u p with a small amount of acetic acid, filtered a t the pump, washed once or twice with acetic acid, and purified by crystnllisation From nitro- benzene, two or three drops of acetic acid being added t o the solution when cooling. Two distinct preparations were made. C,7Hlo0,Br, requires C = 43.22 ; H = 2.54. Found, (i) C = 43.47 ; H = 2.43. Bibromofickugeti?a forms minute, flat needles melting at 280°, readily soluble in hot alcohol, more sparingly so in acetic acid.On fusion with alkaii in the usual manner, fukugetin gave protocatechuic acid (m. p. 194-196") and phloroglucinol (m. p. The above results are, unfortunately, too meagre to allow of the prediction of the constitution of this substance with any certainty, but the similarity of most of the general properties of the compound with those of luteolin, and the fact that it contains similar nuclei, point to the probable close relationship between these colouring matters. Interesting in this respect is the stability of alkaline solu- tions of fukugetin when exposed t o the air, for these do not undergo oxidation even after many days. This property, as showh i n former investigations, is somewhat characteristic of flitvone compoundd such a s apigenin and luteolia, but is not possessed by flavanol derivatives, of which fisetin and quercetin may be yuoted as examples.It is possible that the distinction between luteolin and fukugetin consists chiefly in the manner by which the catechol nucleus is connected with the pyrone ring, and one or two formul~e suggest themselves as very probable representatives of this colouring matter, It is intended, should a further supply of raw material be forthcoming, to continue these experiments in the hope of elucidating with certainty the con- stitution of this interesting substance. An examination of the dyeing properties of " fukugi " showed, as was to be expected, that it behaved in this respect in a n analogous manner t o weld (Reseda Zuteola).The similarity in shade was so marked that, except in point of strength-for fukugi is a stronger dye than weld- (ii) C = 43.53 ; H = 1.95. 2 10").NATURAT, C!OIiOUHIN G MATTERS. 61 i t is impossible to distingukh between them, and there can be little doubt that prior to the introduction of the synthetical colouring matters this dyestuff would have been a valuable addition to those already in use. E t h y l a t i o n 0 f *MoI.in. Although quercetin and other members of the flavanol class readily give, on alkylation, well-defined crystalline substances, morin has only yielded a methyl ether in a stBate of purity (Trans., 1896, 29, i92), for on ethylation in the usual manner, viscous products result which have hitherto refused to crystallise.As it seemed possible that the impurities might be more readily removed after acetylation, the residue from former experiments carried out some years ago was treated in the following manner. The resinous mass was digested with boiling acetic anhydride for some hours, the solution evaporated to a small bulk, and diluted with about twice its volume of methylated spirit. After several days, a small quantity of crystalline matter separated, and the mixture was now set aside for some weeks; the product was then collected, crystallised two or three times from methyl alcohol, and thus obtained in colourless needles, readily soluble in hot alcohol and melting a t 121-1 23". Found, C = 65.70 ; H = 6.23. C,,H,O,(OEt),~C,H,O requires C = 65.78 ; H = 6.06 per cent.To prepare the free tetraethyl ether, the acetyl compound was digested with boiling alcoholic potassium acetate, the solution poured into a small quantity of dilute hydrochloric acid and the product crystallised from methyl alcohol. Found, C = 66.50 ; H = 6-63. C,,H,O,(OEt), requires C = 66.65 ; H = 6.28 per cent. Morin teti-aethyl ether forms pale yellow, prismatic needles, sparingly soluble in cold methyl alcohol, and melting a t 126-128'. I n general properties, it closely resembles the tetramethyl ether previously described (Zoc. cit.).62 PERKIN AND PHIPPS: NOTES ON SOME The B r o m i n a t i o a of M y r i c e t i n in t h e presemce of A 2 ~ 0 7 ~ 0 2 . A most interesting property of morin, C,,H,,,O7, the colouring matter of old fustic, is that when brominated in the presence of alcohol i t yields tetrabromomorin ethyl ether, C1, H,0GBr4*0 Et..This peculiar behaviour, which is apparently not possessed by other flavanol derivatives, is considered by Herzig (Monutsh., 1897, 18, 700) to render doubtful the constitution assigned to it by one of the authors (Trans., 1896, 29, 792). Being in possession of a small quantity of myricetin, which is apparently a flavanol derivative capable of yielding a tetrabromo- compound (Trans., 1896, 69, 1287), the authors have studied its behaviour in these circumstances. Myricetin (1.9 grams) in 20 C.C. of alcohol was treated with 3.4 grams of bromine and the mixture left for forty-eight hours, when on cnutiously diluting with water a small quantity of crystalline precipitate gradually separated and was recognised as tetrabrorno- myricetin (loc.c i t . ) ; this was removed after some hours, the filtrate treated with a large volume of water, and a second deposit collected and crystallised two o r three times from dilute alcohol. Found, C = 31.29 ; H = 1.68 ; Et = 4.36. Cl,H,0,Br4Et requires C = 50.81 ; H = 1.51 ; Et = 4.38 per cent. Tetrcdwomornyricetin ethyl ether formed colourless needles very soluble in alcohol; it becomes red a t l l O o , commences t o sinter at 132O, and melt!, with decomposition at 146". This melting point is given with reserve, owing to the possibilit,y that a trace of tetrabrorno- myricetin itself may be associated w i t h the product. The lack of raw material did not permit of further experiment, but the result indicates t h a t in these circumstances myricetin behaves in an analogous manner t o morin, though somewhat less readily.The M o l e c u l a r W e i y h t s of l l e s p e i a i t i a a n d Curcumin. In a previous communication (Trans., 1898, 31, 1031), it was shown that the properties of hesperitin were in accord with the constitution OH *C,H,( OC H3) CH: C H CO *O C,H,( OH),, previously indicated by Hoffmann (Ber., 1876, 9, 685) and Tiemann and Will (ibid., 1881, 14, 848). The fact, however, that this substance gave well-defined crystalline salts having the f ormulze (C16H1406)2C2H302K andNATURAL COLOURINQ MATTERS. 63 C,,H270,,K suggested that its molecular weight was C32H28012, or twice that assigned to it by these authors.More recently it was shown (Trans., 1903, 88, 127) that certain substances, the molecular weights of which were well known, also gave somewhat peculiar salts by similar methods. Thus from gallaceto- phenone, C,H,O,, the salt C,,H,,O,,K, and from daphnetin, C,H,O,, the salt C,,HllO,K resulted, 2nd other cases might be cited. It was accordingly evident that these salts do not in all cases furnish trust- worthy indications of the molecular weight, and cryoscopic experiments were therefcre carried out, advantage being taken of the ready solubility of acetylhesperitin in naphthalene. 0.3351 in 13.90 naphthalene gave At - 0,375'. 0.4310 ,, 13.S7 naphthalene gave At - 0.490". Found, 15 = 449. Found, M= 443. C,,H,,O,(C,H,O), requires M = 428. It is consequently evident that the molecular weight of hesperitin is represented by the formula C,,H,,O,, and that the above-mentioned potassium and sodium salts belong t o the class of '' semi-substituted " compounds (Zoc.cit.). It has been shown by Ciamician and Silber (Ber., 1897, 30, 192) that the molecular weight of curcumin is most probably represented as C2,H2,,06 rather than C,,H,,O,, the formula originally assigned to i t by Jackson and Menke (Amer. Chem. J, 1882, 4, 77). Analyses of i t s mono-potassium salt (Zoc. cit.), although agreeing approximately with the formula CzlH,,O6E, were not completely satisfactory, owing to the difficulty of purifying this somewhat soluble compound. It therefore appeared interesting t o confirm these results, if possible, by the cryoscopic method, for which purpose a benzoylcurcumin would probably be suitable.This derivative was readily prepared by treating 2 grams of cur- cumin dissolved in 30 grams of pyridine with 23 grams of benzoyl chloride. The product was washed with water, the viscous residue dissolved in alcohol, and the crystals, which gradually separated, were purified by cry stallisation from a mixtnre of this solvent and benzene. Pound, C = 73-87 ; €1 = 4.48. C,,H,i0,(C7H,0)3 requires C: = 76.11 ; H = 4.70 per cent. BenzoyZcurctmin consists of fine, lemon-yellow needles melting at 0.3256 in 13.23 naphthalene gave At - 0.25c. These results therefore indicate that the molecular weight assigned 176-1 78". Found, M = 688. C2,Hlp0,(C'7H,0), requires M = 680.64 NOTES ON SOME NATURAL COLOUHING MATTERS.t o curcumin by Ciamician and Silber is correct, and poiut to the fact that this colouring matter contains three hydroxyl groups, although only two have hitherto been suspected. Curcumin itself is somewhat soluble in naphthalene, but apparently not sufficiently so for molecular weight determination ; the cryoscopic experiments gave 31 = 429, whereas the formula C,lH,,O, requires M = 368. The isolation of curcumin from turmeric is a t best a tedious opera- tion, and as the yields obtained by the published methods were not satisfactory, the following process was adopted. An alcoholic ( xtract of turmeric was treated with lead acetate solu- tion (Daube, Ber., 1870, 3, 709), and the precipitated lead compound of the colouring matter collected, thoroughly washed with alcohol, and then with water. The product suspended in warm water was decomposed with dilute sulphuric acid, and the resulting mixture of curcumin and lead sulphate well washed, drained on porous tile, and extracted with boiling alcohol. After concentration, the extract was poured into ether, the solution decanted from tarry matter, evaporated to a small bulk, and diluted with carbon disulphide. The mixture, when left exposed t o the a i r at the ordinary temperature, gradually deposited crystals, which were collected from time t o time, and i n this way a yield of approximately 0.56 per cent. of curcumin mas isolated from the sample of root employed. Little or no loss should occur by this process, and the mother liquors contain only a, trace of the so-called turmeric resin. The authors express their thanks t o the Research Fund Committee of the Chemical Society for a grant, which has been partly employed to cover the expenses of this research. CLOTH w ORKERS’ RESE AIKX L ABO ILATO ii Y , DYEING DEPARTMENT, THE YOILKSHIRE COLLEGE, LEEDS.