186 PERKIN : DERIVATIVES OF MACLURIN. x 1 v. -Be I iva t itw 0 f lia ClU? * i? 1. Part 11. By A. G. PERKIN, F.R.S.E. I:: a previous communication (Eedford and Perkio, Trans., 1895, S7, 933), it was shown that maclurin, one of the colouring matters of old fustic (Mows tincto~icc), reacts with 2 molecules of diazobenzene and analogous substances, yielding crystalline compounds. Based upon the fact that maclurin forms a pentabenzoyl derivative, C,,R,0,(C7H,0), (Konig and Kostanecki, Bw., 1893, 26, 1994), and when fused with alkali yields protocatechuic acid and phloroglucinol, i t appears to be iL pentahydrosybenzophenone ; conse Iueiitly, the constitution of maclu- rine-azobenzene was represented as follows. OH OH/\-CO--/\N:N- C,H, 0110 OH\,/OH I I N:N*C,H, Ciamician and Silber (L’er., 1894, 27, 162s) found that maclurin, when digested with acetic anhydride and excess of anhydrous sodium acetate, behaves peculiarly, yielding, not a pentacetyl compound, but a substance having this composition less 1 molecule of water.Further, by the activn of hydriodic acid, this is converted, riot into rnaclurin, but into a compound, C,,IT,oO,, isomeric with fisetin. Phlore- tin, C,,H,,O,, also, a product obtained by the decomposition of the glucoside phloridzin, which exists in the root bark of the apple tree, behaved similarly. These authors point out that in this respect both phloretin and macliirin behave like cotoin, which had been previously examined by them (L’el.., 1894, 27, 400), and found to be a mono- methyl ether of benzoylphloroglucinol. This, when treated with acetic anhydride and sodium acetate, is converted into a substame having the coiistitution of a diucetyl derivative less 1 molecule of water, shown ultimately to be the acetyl compound of a inonomethyl 3ther of lvetadionyplienylcoumnrin.PERKIN : DERIVATIVES O F MACLURIN.187 Certain natural colouring matters and allied substances, a t present undergoing investigation, which possess the property of reacting with diazobenzene, yield by the ordinary method of acetylisation sticky pro- ducts difficult to purify, resembling maclurin in this respect. Their behaviour, when treated according to the process of Ciamician and Silber, has not at present been studied, for the derivatives formed not being normal acetyl compounds, cannot be used as a sure indication of the number of hydroxyl groups present in them.Having observed that the diazobenzene derivatives of maclurin, cyanomaclurin, &c., are considerably more stable towards reagents than the substances themselves, it appeared possible that the acetyli- sation might proceed in the normal manner, and further that by Lieber- mann’s method, the number of acetyl groups present could be deter- mined with accuracy and rapidity. The difficulty experienced in preparing a quantity of the substances under examination made it ad- visable t o first test the accuracy of such a method with maclurin itself, for not only is this colouring matter more readily procurable, but from Kostanecki’s results, the number of its hydroxyl groups is no longer open to doubt.It was at first intended to embody the results of this paper in an investigation of cyanomaclurin and catechin which is now proceeding, but having noticed that Herzig (Monatsh., 1896, 17, 421) has expressed his intention t o further study the acetylisation products of maclurin obtained by Ciamician and Silber, I thought it desirable t o submit this communication to the Society, so as to conclude for the present my study of maclurin. EXPERIMENTAL. A mixture of 1 gram of finely-powdered maclurinazobenzene, 1 gram of anhydrous sodium acetate, and 30 grams of acetic anhydride was digested at the boiling heat ; the maclurinazobenzene, without appearing to dissolve entirely, became slightly yellower, and after a few minutes the whole was transformed into a semi-solid mass, owing to the separation of yellow crystals.Excess of acetic anhydride was now added, and the mixture digested until complete solution was attained, the heating being continued for an hour longer. After diluting with acetic acid, the product was poured into a large volume of water, and the red precipitate collected and washed. Before drying at looo, it was found preferable to strongly press the product between bibulous paper, otherwise, when much water was present, the whole aggregated to a dark-coloured, resinous mass, whilst, if this was avoided, a friable red powder was obtained, which could be more readily purified. Treatment with alcohol removed a dark-coloured impurity, and the residue was then crystallised from boiling cumene ; this was greatly facilitated by188 PERKIN : DERIVATIYES OF MACLURIN.the addition of a trace of alcohol t o the hot cumene solution, the mass then separating entirely in the crystalline form, and the tendency of a portion t o gelatinise was avoided. 0.1420 gave 11 C.C. nitrogen a t 18' and 772 mm. 0.1230 gave O.ZS17 CO, and 0.0495 H,O. per cent. CI,H,O,(C,H,O),(N,* C,H,), requires N = 8-77 per cent'. Although the production of a triacetyl derivative was readily to be accounted for, it appeared advisable t o study the action of acetic anhy- dride when continued for a longer period, and for this purpose the heating was continued for 7 hours. The chief product of the acetylisa- tion was again found to be the above acetyl compound, accompanied, however, in this instance, by a larger quantity of the resinous impurity previously referred to.N = 9-09. C = 62.45 ; H = 4.47. C,3H,0,(C,H,0),(N,* C,H,), requires C = 62.42. H = 4.02 ; X = 9 40 An analysis gave the following result. 0*1708 gave 13.6 C.C. nitrogen a t 15' and 770 mm. Theory requires N = 9.40 per cent. I n order to determine the number of acetyl groups contained in this substance, it was suspended in a very little boiling acetic acid, a few drops of sulphuric acid added, and the clear solution heated to boiling for a short time; the cautious addition of boiling water caused the separation of crystals, and after the mixture had stood for 12 hours, they were collected and washed with water. The filtrate was pale yellow, evidently containing a trace of the substance still in solution ; it was therefore concentrated to a small bulk, treated with boiling water, and again evaporated, this process being repeated until but a faint odour of acetic acid was perceptible.After standing 1 2 hours, the minute quantity of crystals which had then separated was added t o the main bulk, which was dried first a t 100' and then a t 160". 0.7716 gave 0.6109 C,,H,O,(N,C,H,),. C1,HjOa(C2H30),(N,*C,H,), requiresC,,H,O,(N,C,H,), = 7S%6 per cent. To be sure that the decomposition product thus obtained was in reality maclurinazobenzene, it was analy sed, with the following result. 0.1222 gave 13 C.C. moist nitrogen a t 19" and 743 mm. N = 11.94. 0.1193 ,, 0.2786 CO, and 0.0430 H,O. C = 63.69 ; H = 4 00 C,,H,O,(N,* C,H,), requires C = 63.83 ; H = 3.S3 ; N = 11 -91 per cent.Its melting point also was found t o be identical with that of maclurinazobenzene, for when heated, both showed signs of decompo- sition a t 265", and melted at 276-277" with evolution of gas. It WRS observed that the melting point of maclurinazobenzene varied within a few degrees, according to the rapidity with which it was heated; if heated slowly it melted even as low as 270". The crystals obtained N = 9-45, Found, 79.17 per cent.PERRIN : DERIVATIVES O F RIBCLURIN. IS9 on decomposing the acetyl compound in this manner were remark- able for their size and beauty, as compared with those produced by the methods described in the previous paper (Eoc. cit.), a fact which a t first caused a doubt as to their identity. It mas, however, found subsequently that, on cautiously adding water to the unacetylised substance dissolved in zt hot dilute solution of sulphuric acid in acetic acid, crystals separate, identical with those obtained from the acetyl compound, ~twicccet~E.lncccl.uri~z,tcLxo6eizxene forms a mass of glistening, hair-like, orange yellow needles, melting at 240-243" with decomposition, sparingly soluble in boiling acetic acid and cumene.It is not aLtt;Lcked by cold, dilute alkalis, but on heating gradually dissolves, yielding an orange-red liquid from which acids precipitate red flocks of the free diazobenzene compound. It has been stated above that when maclurinazobenzene i d digested with acetic anhydride and sodium acetate, the mixtnre quickly solidifies unless a very large excess of the former reagent is present.I n order t o examine this product, the mass thus produced was stirred up with a small quantity of acetic acid, and the crystals collected, mashed .repeatedly with alcohol, then with water and dried. The glistening mass of yellow needles thus obtained was found t o consist entirely of the above triacetylmaclurin- azobenzene, it being thus evident that the process of acetylisation takes place with extreme rapidity, prolonged digestion being unneces- sary. I n this way, nioreovei-, the yield is considerably increased. Maclurin, as is well known, contains a phloroglucinol nucleus, and it therefore appeared of interest to study the acetylisation of phloro- glucinolazobenzene ; this was prepared according to the directions of Weselsky and Benedikt (Uer., 12, 2%)) and purified by crystallisation from alcohol.A nitrogen estimation gave N = 16.43 per cent., theory requiring 16-76 per cent. I n acetylising this substance, the method employed was identical with that described above with maclurinszobenzene. As the digestion proceeded, the bright orange-red solution a t first obtained gradually became browner and finally very dark, as if decomposition mas taking place. The addition of water caused the separatioii of a chocolate- coloured precipitate, which, after standing for 1 2 hours, was collected, washed, and dried. Treatment with slightly warm acetic acid removed from this product a dark-brownish, resinous substance, a small quantity of a red powder remaining undissolved; this was purified by crystalli- sation from acetic acid.0'0882 gave 11.4 C.C. nitrogen a t 16', aild 754 111111. I? = l.l*O-i. CGH,O,(N,* C61i15)2* C,H,O i,ecluires N = 14.89 per cent. The yield of the ncetyl compound olhained in this may being small,190 PERKIN : DERIVATIVES O F MACLURIN. it seemediprobable that, if digested for a shorter time, a larger quantity might be produced ; this was found to be the case. As soon as it was completely dissolved, the whole was allowed to cool, and the crystals, after treatment with acetic acid, were collected, washed with alcohol, then with water, and finally purified by crystallisation from acetic acid. 0.1357 gave 17.7 C.C. nitrogen a t 17", and 740 mm. H = 14.73 ; 0.1178 ,, C,H,O,(C,H30)(N,CGH,), requires C = 63-82 ; H= 4.25; N = 14.89'p.c. To determine the number of acetyl groups present, a similar method was employed as with the maclurin compound.0.6664 gave 0.5887 C,H,O;(N,* C,H,),. CGK303(C.LHSO)(COH5N,)~ requires CoH,O,(N,* C,H,), = 88.83 per cent. It was evidently a monacetyl compound. The regenerated phloro- glucinolazobenzene was analysod. 0.1342 gave 19.7 C.C. nitrogen a t 16', and 748 mm. N = 16 83. Phloroglucinolazobenzene melts and decomposes at 228-230". I~~oiLacet~Z~)~lo~o~lzcciizoZa,-obeil~e?ie was obtained as an orange-red mass of glistening needles meltring a t 222-223" wit'h decomposition, readily soluble in boiling glacial acetic acicl. Cold alkaline solutions do not dissolve it, but on warming, orange-red liquids are obtained, from which acids precipitate red flocks of the azo-compound.The above results show that, although in maclurin five hydroxyls exist, three only are present in its diazobenzene derivative. This is readily accounted f o r on examining the constitution shown above, two hydroxyls in the ortho-position relatively to the diazobenzene groups being in the ketonic condition. The same also is the case with phloroglucinolazobenzenc, which consequently contains, as found above, but one free hydroxyl group, and, if further proof were needed, there is here conclusive evidence of the existence in maclurin of five hydroxyl groups. It is interesting to notice that the peculiar reaction of maclurin with acetic anhydride and sodium acetate investigated by Ciamician and Silber (Zoc. cit.) does not, under the conditions here employed, take place with its diazobenzene derivative, as this behaves in the normal manner with these re-agents, I t s behaviour in this respect, and its stability when compared with maclurin, would appear, therefore, t o be due to the ketonic condition of the two hydroxylsabove referred to.Moreover, it appears probable that the acetylisation of the diazo- benzene derivatives of substances analogous to maclurin will yield a ready means of estimating their hydroxyl groups, in case these cannot! 0.2745 GO,, and 0.0495 H,O. C: = 63.55 ; H= 4.67. Found, 88.19 per cent. Theory requires N = 16.76 per cent.FORSTER : CONVERSION OF CAMPHOROXTME, ETC. 191 be determined in the ordinary way; and further experiments that are being carried out on the azobenzene derivative of catechin seem likely to corroborate this supposition, I n such substances, but little difficulty presents itself in determining the group or groups present to which this reactive capacity with diazo-compounds is due, and should ortho- azo derivatives be formed, the ketonic condition of such hydroxyl groups must thus be allowed for.LuteoZim.-Since the publication of my last communication on this colouring matter, a paper has appeared by Herzig (Nonatsh., Zoc. cit.) which confirms those results excepting as regards a slight difference in the melting point of luteolin triethyl ether. He, however, mentions a second product formed during the ethylation of luteolin, although its nature and composition were not determined. Although no account of this substance appears in my paper, I was fully aware of its existence, but did not consider it necessary t o mention i t at that time, believing that the prior publication of my results was sufficient to reserve f o r me the further study of this reaction. Vnderstanding, however, from a private communication from him, that he is investigating this product, I have withdrawn my claim to it, This, howcver, renders it necessary for me t o point out that, during the methylation of luteolin, the results of which I have already partly desc~ibed, rz second substance is similarly produced in conjunction with luteolin trimethyl ether. An account of this, now partly investiguted, I hope shortly to lay before the Society. For reasons similar to those given above, it has not been before men- tioned, although I have been aware of its existence for many months. I reserve this entirely for myself. CLOTH\VORKERS’ RESESRCII LABORATORY, DYEING DEPARTMENT. YORKSHIEE COLLEGE.