1594 GREENWOOD AND NIERENSTEIN : CLXXVI1,-Studies in the Chronzan Series. Part I. By ANNIE GREENWOOD and MAXIMILIAN NIERENSTEIN. IN order to obtain confirmatory evidence of the views expressed by one of us wibh regard to the action of metallic sodium and alcohol on catechin tetramethyl ether (this vol. pp. 971 1151) we thought it advisable to study the effect of these reagents on some simple derivatives of the chroman series. Our observations on 1-phenylchroman (I} and 3-phenylchroman (III) which are described in the present communication confirm in every respect the deductions made in the case of catechin tetramethyl ether. Thus we obtained 2-hydroxy-ay-diphenylpropane (11) and 2-hydr-oxy-aa-diphenylpropane (IV} by the action of metallic sodium on I-phenylchroman (I) and 3-phenylchroman (111) respectively.We also found that whereas the ay-derivative (11) is recovered unchanged on oxidation 2-hydroxy-aa-diphenylpropane (IV) yields under the same conditions 2-hydroxydiphenylacetic acid (V). 0 /\/><yHPh /\OH I ICH~-CH,.CH,P~ \/ 1 1 \ / \ P f l 2 CH2 (1). (11. ) (111 1 (1V. 1 w. 1 1-Phenylchroman (I) has been previously prepared by Harries and Busse ( B e r . 1896 29 380) and Feuerstein and Musculu STUDIES IN THE CXROMAN SERIES. PART I. 1695 (ibid. 1901 34 412). We have obtained 3-phenylchroman (111) according to Semmler’s scheme for the synthesis of chroman (Ber., 1906 39 2855). 3-Phenyldihydrocoumarin (VI) was reduced to 2-o-dihydroxy-aa-diphenylpropane (VII) and subsequently con-verted into 3-phmylchroman (111).E x P E R I M E N T A L. 2 -Aydroxy-a y -diphenylpropan.e ( T I ) . To a solution of 11.5 grams of l-phenylchroman prepared accord-ing to Harries and Busse’s method in 150 C.C. of alcohol are added 50 grams of metallic sodium and the solution is heated for eight hours on a water-bath. The alcohol is removed with steam and, at the same time 4.5 grams of unchanged l-phenylchroman are recovered. This unchanged material is again reduced 20 grams of sodium and 100 C.C. of alcohol being used. The combined aqueous solutions are extracted with ether in order to remove any unchanged l-phenylchroman (0- 72 gram of unchanged product was recovered on evaporation of the ether) the solution acidified with dilute sulphuric acid and extracted several times with ether.The dried ethereal extract leaves on evaporation an oil which solidifies on keeping over paraffin in a vacuum. It’ crystallises from light petroleum in slender glistening needles which melt sharply at 21.5O. The alcoholic solution turns violet with alcoholic ferric chloride. The yield is 69 per cent. of the theoretical (Found*: C=85-1; IX=7.6. C,,H,,O f requires C=84*9; H=7*5 per cent.). Diazomethane converts the ethereal solution into the methyl ether the yield being 86 per cent. of the theoretical. It is a viscous colourless oil which boils a t 143-147°/11-12 mm. (Found 1 C = 84.9 ; H =8.2. CI6Hl80 requires C = 85.0 ; H = 7.9 per cent.). * Dried over paraffin in a vacuum. -f Braun and Deut,sch (Bey. 1912 45 2187) have prepared 2-hydroxy ny-diphenylpropane in a different manner.They describe i t as a yellow, viscous odourlees oil which boils at 198-502°,/15 rnm. They calculate, however wrongly for C,,R1,O which requires C = 86.5 ; H = 5.8 and they find C = 86.63 ; H = 6-01 per cent. 5 Dried over-paraffin in zb vacuum 1596 GREENWOOD AND NIERENSTEIN : 3-PhenyZtZ.ihydrocoumarin (VI) . This substance has been prepared by Liebermann and Hartmann (Ber. 1891 24 2582) and Simonis (“Die Cumarine,” 1916, p. 132) but1 neithelr method was used by us. We have prepared it by the action of acetyl chloride on 2-methoxy-PO-diphenyl-propionic acid (compare Stoernier and Friderici Ber. 1908 41, 340). For the synthesis of this acid we used Fosse’s method (Ann. Chim. 1920 [ix] 13 105) which seemed t o us preferable to that employed by Stoernier and Friderici (7oc.c i t . p. 335). TwenLy-six grams of 2-methoxydiphenylcarbinol (prepared according to Stoermer and Friderici’s method Zoc. c i t . p. 332) are heated in an oil-bath for four hours at 120° with 50 grams of ethyl malonate. The excess of the ethyl malonate is distilled off and the solid crystallised from alcohol from which it separates in long, prismatic needles which melt a t 134-135O carbon dioxide being evolved (compare Stoermer and Friderici loc. c i t . who give 131O). The yield is 92 per cent. of the theoretical. 2-Methoxy-PP-diphenylpropionic acid (23 grams) dissolved in glacial acetic acid (100 c.c.) is converted into1 3-phenyldihydro-coumarin with acetyl chloride (15 grams) on keeping for forty-eight hours a t the ordinary temperature!.The precipitate formed on the addition of water contains some unchanged 2-methoxy-PP-diphenylpropionic acid which is remo’ved by dissolving the pro-cipitate in ether and washing the ethereal solution with a 10 per cent. sollution of sodium hydrogen carbonate saturated with carbon dioxide. The solid left’ on evaporation of the dried solution crystallises from alcohol in stout plates which melt at 8 2 O as given by Liebermann and Hartmann (Toc. c i f . ) . 2-o-Dih?/clron.?/-aa-c~i~?~,e?~yl~)~o~)(~ i t e (VII) . To a solution of 22 granis of 3-phenyldihydrocoumarin in 175 C.C. of alcohol 50 grams of metallic sodium are added and the solu-tion is kept a t the boiling point for eight hours when all the sodium disappears. The solution freed from alcohol by steam distillation is acidified extracted with ether and the drield ethereal extract leaves on evaporation an oil yielding a main fraction, which boils at 195-204°/11-12 mm.On redistillation of this fraction a colourless heavy oil is obtained which boils constantly a t 197-199O/11 mm. Jt is soluble in the usual organic solvents, and does not solidify even on prolonged keeping in a freezing mixture. The yield is 68 per cent. of the theoretical (Found % : C=79 1 ; H-7.3. C,,-,H,,O,? requires C-78.9; IT-’7.0 per cent.). * Driccl over parefhi in a vmxnnn STVTDIES I N THE CIIROMAN SERIES. PART I. 1597 3 1’h ciiylchromaii (111). A solution of 16 grams of 2-w-dihydroxy-act-diphenylpropane in 100 C.C. of absolute alcohol is heated on a boiling-water bath dry hydrogen chloride being passed through the solution a t intervals.The solution is reduced to 20 C.C. and diluted with 300 C.C. of water which causes a heavy oil to separate. The oil is extracted with ether and the etherelal extract washed first with dilute alkali and then with water. The dried ethereal extract gives on evapor-ation an oil which is dried in a vacuum over paraffin. The sub-stance obtained in this way is very readily soluble in the usual organic solvents but by dissolving about one part of the oil in about four parts of boiling light petroleum and keeping in a freezing mixture of ice and salt long prismatic needles are obtained which melt sharply a t 38-5O. The yield is 78 per cent. of the theoretical. The addition of ferric chloride to a suspension of 3-phenylchronian in concentrated sulphuric acid produces a reddish-violet coloration even i f the substance has been crystal-lised several times.Chroman prepared according t o Semmler’s method (Zoc. c i t . ) and 1-phenylchronian also1 give a faint violet coloration with ferric chloride and concentrated sulphuric acid. These observations are of great interest in connexion with the chemistry of catechin in view of the importance which has been attached to the fact that both catechin tetramethyl ether and the coumarans give a violet coloration with ferric chloride and con-centrated sulphuric acid (compare Kostanecki and Lamp Ber., 1906 39 4007 ; Freudenberg ibitl. 1920 53 [B] 1423) (Found * : C=86-1; H-6.8. Cl,H140 requires @-85.7; H = 6 * 6 per cent.).2-Hydro x y -aa-d iph c IL y 1 pro pane (IV) . Nine grams of 3-phenylchronian dissolved in 150 C.C. of alcohol, are heated with 40 grams of metallic sodium and the alcohol is removed with steam. From the fact that on steam distillation, no unchanged 3-phenylchroman was recovered as observed in the case of 1-phenylchroman it may bel deduced that the 3-phenyl-chroman nucleus undergoes fission more easily than the 1 -phenyl-chroman nucleus. The ethereal extract oE the acidified solution leaves on evapor-ation an oil which gives a main fraction boiling a t 226-231°/ 13-14 mm. The pure substance is a viscous colourless oil which boils at 214-216O/5-6 mni. and is soluble in all the usual organic solvents. The product does not solidify even on prolonged * Dried over paraffin in 8 vmuum 1598 STUDIESilIN TKE CHROMAN SERIES.PART I. keeping in a vacuum or in ay freezing mixture. The alcoholic solution of 2-hydroxy-aa-diphenylpropane turns violet with alcoholic ferric chloride. The yield is 71 per cent. of the theoretical (Found * C= 85.2 ; H = 7.8. C,,H,,O req,uires C = 84.9 ; H=7.5 per cent.). Diazomethane converts the ethereal solution into the methyl ether the yield being 74 per cent. of the theoretical. It is a viscous colourless oil which boils at 176-179O/4-5 mm. (Found* C=84-7; H=8*1. C,,H,,O requires C=85*0; H=7*9 per cent .). 2-~ydroxydiyhenylacetzc ,4 cid (V) . Five grams of 2-hydroxy-aa-diphenylpropane dissolved in 100 C.C. of a 20 per cent. solution of potassium hydroxide in water, are oxidised on a boiling-water bath for four hours witah 5 grams of potassium permanganate in 100 C.C.of water. The solution is filtered while hot and after cooling acidified with dilute sulphuric acid. The grey-coloured precipitate is not filtered but the solution extracted several times with ether in which the precipitate dis-solves. The dried ethereal extract leaves a solid which crystallises from dilute alcohol with the aid of animal charcoal in prismatic needles melting at 86-87O as given by Bistrzycki and Flatau (Ber. 1895 28 990). The yield is 64 per cent. of the theoretical. On heating 2 grams of the acid with acetic anhydride and anhydrous sodium acetate a quantitative yield of 2-phenyl-mumaran-1-one is obtained (compare this vd. p. 1155). It crystallises from alcohol in prismatic needles which melt a t 113-114O as given by Bistrzycki and Flatau (Eoc.c i t . p. 989). Neither the melting point of 2-hydroxydiphenylacetic acid nor of 2-phenylcoumaran-1 -one was found to be depressed when mixed with the corresponding substances which had been prepared accord-ing to Bistrzycki and Flatau’s method. Addendum. Marschalk (Ber. 1910 43 1700) a t the suggestion of Kostanecki has proposed the names depsan (from 6+0 to tan) for p-benzylcoamaran depsanoll foc leuco-p-beinzoylcounlaran and 3 5 3’ 4/-tetrahydroxydepsanol for catechin. 0 OH 0 Depsan. Catechin (Kostanecki’s formula). * Dried over paraffi in a vacuum DEPENDENCE OF OPTICAL ROTATORY POWER ETC. 1699 This nomenclature has been freely used by Kostanecki’s collaborators (compare for example Bielinki Diss.Bern 1911), which is to be regretted for the following reasons (1) From Emil Fischer’s well-known researches the name clepside which is also derived from 8 l 4 0 has become ciefinitely associated with a series of anhydrides. (2) The name tetrahydroxydepsanol suggested for catechin indicates that catechin possesses tanning properties which is not the case (compare Prcxter “ The Principles of Leather Manufacture,” 1903 ; Perkin and Everest “ The Natural Organic Colouring Matters,” 1918). (3) Most of tha recent researches point towards a chroman formula for catechin as originally suggested by A. G. Perkin (T. 1902 81 1172; 1905 87 404) and the possibility that catechin is a coumaran derivative has become very remote. For these reasons the authors would suggest that the names depsan depsanol etc. should no longer be used in connexion with catechin. The authors wish to express their indebtedness to the Collston Society of the University of Bristol for a grant which has covered the expenses of this research. CHEMICAL DEPARTMENT, BIOCHEMICAL LABORATORY, UNIVERSITY OF BRISTOL. [Recewed Noue&r Qth 1920.