1534 HIGGINBOTHAM AND STEPHEN CLX1X.-Studies zn the Coumaranone Series. Pal$ I. The Preparation of 4- 5- and 6-Methgl-coumaran-%ones and some Derivatives of o- m-, and p- Tolyloxyacetic Acids. By LUCY HIGGINBOTHAM and HENRY STEPHEN. STOERMER and Bartsch (Ber. 1900 33 3181) were the first to prepare the above three methylcoumaranones by condensing the corresponding p - m- and o-tolyloxyacetic acids by means of phosphoric oxide. The reaction which takes place in the formation, for example of 4-methylco;umaran-Z-one (I) from p-tolyloxyacetio acid may be represented as follows: 0 (1- ) Poor yields of the coumaranones were obtained by the above method and the compounds isolated and certain derivatives of them were impure as subsequent investigation has shown. The preparation of 4-methylcoumaranone (I) was carried out success-fully by Auwers and Muller (Ber.1908 41 4233) and Fries and Finck (ibid. p. 4271) almost simultaneously and the latter authors also describe the preparation of 5-methylcoumaranone.* The method in both cases depends on the withdrawal of the elements of hydrogen haloid from o-halogen-o-hydroxyacetophenones but, owing to the 'difficulty in obtaining the required acetophenones the method is not suitable for more general application. The present investigation provides a method for the preparation of all three coumaranones and is in principle analogous to the method given by Shermer and Bartsch indicated above but with * Auwers (Ber. 1916,49 809) has recentIy prepared 6-methylcoumaran-2-one by a similar method STUDIES IN THE COUMARANONE SERIES.PART I. 1535 the difference that the coumaranones are derived from the acid chlorides of the three tolyloxyacetic acids. Methods for the pre-paration of these acids have been known for some time but modifi-cations of them are described in the present paper which give improved yields and shorten the time required for reaction. The conversion of the acids into the acid chlorides was found to take place more readily using thionyl chloride rather than phosphorus pentachloride. Michael (Amer. Chem. J. 1889 9 216) has shown that phenoxyacetic acid on treatment with phosphorus penta-chloride yields the acid chloride but a t the same time a con-siderable amount of a mixture of o- and p-chlorophenoxyacetyl chlorides is formed.Using thionyl chloride almost quantitative yields of the three tolyloxyacetyl chlorides are obtained. Phenoxy-acetyl chloride is also readily prepared in this way. Stoermer and Atenstadt (Ber. 1902 35 3569) have shown that when phenoxyacetyl chloride is dissolved in benzene and treated with aluminium chloride two products result from bhe reaction, namely coumaran-2-one (m. p. 101O) and w-phenoxyacetophenone (m. p. 72O). These were separated according to the above authors, by distillation in a current of steam coumaran-2-one being volatile and o-phenoxyacetophenone being isolated from the residue after distillation. This procedure is not however in agreement with the statement of Mohlau (Ber. 1882 15 2497> who prepared w-phenoxyacetophenone (m. p. 72O) from w-bromoacetophenone and phenol in alkaline solution and describes the ketone as being volatile in steam.It is thus difficult to understand how Stoermer and Atenstadt could have isolated the ketone in the manner described. A repetition of their experiment has been made and, whilst no ketone was isodated a small amount of coumaran-2-0110 was obtained (m. p 101O). The latter substance was also isolated from the reaction which took place when a solution of phenoxyacetyl chloride in carbon disulphide was treated with aluminium chloride. In both cases much residue was left after distillation with steam. The above reaction was then applied to the three tolyloxyacetyl chlorides mentioned above and in the first series of experiments the reaction was carried out using an excess of benzene as solvent.None of the three w -tolyloxyacetophenones which might have been formed in the usual way by the condensation of the tolyloxyacetyl chlorides with benzene was obtained the only crystalline products isolated from the reactions being the 4- 5- and 6-methylcoiumaran-ones. The same compounds were formed by treating solutions of the acid chlorides in carbon disulphide with aluminium chloride. The yields of coumaranones obtained by either method of pro-cedure are approximately the same usually about 35 to 40 pe 1536 HIGGINBOTHAM AND STEPHEN : cent. of the theoretical and although not quite so good as those obtained by Auwers and Miiller and by Fries and Finck (Zoc. cit.), the method offers several advantages over that given by these investigators .For purposes of comparison attempts have been made to pre-pare the three o-tolyloxyacetophenones according to the method described by Kunckell (Her. 1897 30 577) who claims to have prepared o -m-to1 yloxyacet op henone by treating o - bromoacet o-phenone and nt-cresol dissolved in alcohol with sodium hydroxide, and in a similar way w-p-tolyloxyacetophenone was prepared from o-bromoacetophenone and p-cresol. Both experiments were repeated with a slight modification using o-chloroacetophenone instead of the corresponding bromo-compound and from each experiment two substances were obtained both of which contained chlorine and melted at' 118O and 1 4 8 O respectively. These com-pounds were isolated under different conditions the former being obtained by warming a mixture of w-chloroacetophenone with either m- or p-cresol in the presence of alcoholic sodium hydroxide on the water-bath for an hour.After filtering off the sodium chloride formed during the reaction from the alcoholic solution, which no longer possessed the pungent odour of the chloro-ketone, although the odour of cresol was evident fine needles were deposited which melted as above after recrystallisation from methyl alcohol. The compound of higher melting point was formed when the heat-ing of the alcoholic solution of the mixture was continued for several hours. After similar treatment the above-mentioned com-pound melting a t 1 4 8 O and crystallising from methyl alcohol in small prkrns was obtained. Subsequent investigation has shown that both substanem could be obtained in several ways; thus on treating an alcoholic solution of o-chloroacetophenone with the dry sodium compounds s f phenol 0- m- and p-cresolls and &naphthol, respectively each experiment gave one or other of the above chloro-compounds according to the duration of the reaction.The action of these sodium compounds and also thaC of sodium hydroxide on o-chlo'roacetophenone is the same as that of ammonia and sodium ethoxide on the same substance as was investigated by Staedel and Riigheimer (Ber. 1876 9 1769) and by Paal and Stern (Ber. 1899 32 531) respectively and both investigations have shewn that two chloro-compounds are formed namely the so-called u- and P-chlorodiphenacyla melting a t 1 1 7 O and 1 5 4 O respectively.Widman and Almstrom (AnnaZen 1913 400 86) have investigated the constitutions of both compounds and shown that a-chlorodiphenacyl is cis-2-chloro-3 4-oxido-3 5-diphenyltetra-hydrofuran (11) (m. p. 117-118O) and that the &compound i STUDIES IN THE COUXARAHONE SEEXES. PART I. 1537 t'he corresponding trans-isomeride (111) (m. p. 149O and not 154O, as given above). c1 O*CHPh*CH I,O*OHPh*CH / " I (3' / I I \-&h.O i3 ' L-CPh*" Gl (11.1 (111.) It is thus difficult to understand how various investigators in the past have succeeded in preparing a-alkylated or o-arylated acetophenones by the action of sodium compounds of alcohols or phenols on o-halogenacetophenones . An investigation of this problem is being carried out by one of us. The preparation of other derivatives of 0- m- and p-tolyloxy-acetic acids is described below.EXPERIMENTAL. 0- m- and p-Tolyloxyacetic Acids. The preparation of these acids by treating a mixture of chloro-acetic acid and the corresponding cresol with an aqueous solution of sodium hydroxide has been known for some time the ortho-acid having been first prepared by OglialorocTodarol and Cannone (Gazzetta 1889 18 511) and the meta- and para-acids by OglialorocTodaro and Forts (Gazzetta 1891 20 508) and Gabriel (Ber. 1882 14 923). Details of a modified method for t8he preparation of the three acids by the same reaction are given in D.R.-P. 79514 and 85490. Numerous experiments carried out according to the details given in the above investigations have shown that the yields of the acids nelver exceed 60 per cent.of the theoreltical loss of chloroacetic acid by hydrolysis probably taking place. The present investigation has shown that the effect of hydro'lysis can be mini-mised by adding ths cresol dissolved in alkali hydroxide to the chloroacetic acid. I n the particular case of the meta-acid the sodium hydroxide was replaced by an equivalent amount of potassium hydroxide which gave an increased yield of the product; in other respects the process was the same. Ninety-six grams (I mol.) of chloroacetic acid were heaked in an oil-bath a t 110-120° and a solution of 108 grams (1 mol.) of the cresol and 100 grams (2.5 mols.) of sodium hydroxide in 400 C.C. of water was slowly added with frequent shaking. The addition of the solution was complete in two hours when the product after being cooled to about 40° was acidified with approximately 2N-sulphuric acid.The tolyloxyacetic acid was collected drained, VOL. c3xVI.I. 3 1538 HICIQINBOTHAM AND STEPHEN: and finally washed with 20 C.C. of light petroleum to remove traces of adhering cresol. The weight of crude acid was 140 grams and a further 10 grams were extracted with ether from the filtrate the combined yield being 90 per cent. of the theoretical. The acid obtained in this way is pure enough for most purposes without further treatment. The temperature stated above a t which the acidification was carried out is of importance since the meta-acid undergoes decom-position to some extent by mineral acid' a t higher temperatures.On the other hand the sodium salt of the para-acid is only moderately soluble in water a t 40° and the free acid was best obtained by warming the solution to 70° when the sodium salt dissolved and slowly adding 2N-sulphuric acid at the same temperature until acid. The acids may be purified by crystallisation from dilute alcohol, and are then obtained in thin glistening plates the melting points of which are 151-152O 102O and 1 3 5 O for the ortho- meta- and para-acids respectively agreeing with those given in the literature. The acids are decomposed into the respective cresol and glymllic acid on distillation in a current of steam or by boiling with dilute mineral acids the meta-acid more readily than the para- the ortho-acid being the most resistant to such treatment.Preparation of the Ammonium Salts of the above Acids. In connexion with the preparation of the nitriles of the tolyl-oxyacetic acids the ammonium salts were obtained most readily in the following way. Twenty-five grams of the acid were dissolved in 100 C.C. of dry ether and the solution was saturated with dry ammonia a t the ordinary temperature. The anhydrous ammonium salt separated from the ethereal solution as a. fine crystalline powder. The ammonium salts of the 0- m- and p-acids decompose at 126O 18S0 and 1 7 7 O respectively. 0- m- and p-Tolyloxyacetyl Chlorides C9H,0,C1. Twenty grams (1 mol.) of the acid previously dried by fusion were finely powdered and placed in a small flask fitted with a reflux apparatus on a water-bath a t 60°; 18' grams (1-5 mols.) of thionyl chloride were added slowly and the reaction was moderated if necessary, by removal of the flask from the water-bath.The reaction was complete at the end of fifteen minutes and a. pale yellow liquid remained ; longer heating caused the product to darken in colour, These acid chlorides have hitherto not been described STUDIES IN THE COUMARANONE SERJES. PART I. 1539 resulting in a poor yield of the acid chloride. The excess of thionyl chloride was removed by distillation on the water-bath a t 30° under about 15 mm. pressure and the residue finally distilled a t 10 mm. Twenty grams of each acid chloride were obtained as colourless liquids the yields being about 90 per cent. of the theoretical and in ail three cases the products were obtained solid.B. p.110 mm. 35. p. m- ...... 126 19.3 o-Tolyloxyacetyl chloride ....... 120" 29-30' P- 9 9 39 ...... 124 17.9 They are al! soluble in carbon disulphide benzene or ether and are decomposed by water yielding the respective acids. 9 9 9 9 01- m- and p-Tolylozyacetarnides C,H,,O,N. (1) Twenty grams of the ammonium salt of o-tolyloxyacetic acid were heated under 10 mm. pressure in a distillation flask in an oil-bath a t 135O water passed over and the reaction was com-plete when the solid in the flask had liquefied the amide formed being then a t a temperature above its melting point. In the cases of the ammonium salts ob the meta- and para-acids, the temperature of the oil-bath was maintained a t 190-200° in consequence of the higher decomposition points.The liquid residue solidified on cooling and the product crystallised from alcohol. (2) On warming the acid chlorides with an excess of ammonium carbonate on the water-bath for half an hour and removing the ammonium chloride formed by triturating the solid mass with cold water and filtering the amides were obtained pure after crystal-lisation from alcohol and identical with those prepared by the first method. The three compounds were purified for analysis by recrystal-lisation from benzene being obtained in stout rhombic prisms and the purity of each substance was tested by boiling a known weight with aqueous sodium hydroxide and collecting the ammonia liberated in standard acid. The results were in close agreement with the theoretical values.M. p. M. p. o-Tolyloxyacetamide.. .......... 127" 128' 713- 9 9 ............ 118 111-112 P- 7 3 ............ 119 126 The figures in the second column are those given by Forte (Gazzetta 1893 22 ii 526). The amides are sparingly soluble in cdd methyl and ethyl alcohols or benzene but dissolve readily on warming and are slowly hydrolysed by boiling with water. 3 ~ 1540 HIMINBOTHAM AND STEPHEN : The yields of the amides obtained by both met,hods are about 90 per cent. of the theoretical. o- m- and p-Tolyloxyacetonitriles C,H,ON. The meta- and para-compounds were obtained by Stoermer and Schmidt (Ber. 1899 30 1705) from the oximes of the correspond-ing tolyloxyacetaldehydes the former being a yellow liquid (b. p. 254O) and the latter crystallising in needles (m.p . 40O). The ortho-compound has not hitherto been described. The three sub-stances are more readily prepared by treating the .amides with phosphoric oxide. Ten grams of the amide were intimately mixed with 10 grams of phosphoric oxide in a distillation flask and the mixture was warmed in an oil-bath a t 1 2 0 O . When the reaction had begun, the sidetube of the receiver attached to the flask was connected to the pump and a yellow oil distilled a t 130°/10 mm. After fifteen minutes the temperature of the oil-bath was raised to 150° and the distillation was complete after a further ten minutes. Six grams of nitrile were obtained in this way. B. p./lO mm. M. p. o-Tolyloxyacetonitrile.. ........... 133" yellow oil. m- 7 9 141 Y 9 P- ...............136 38-39' ............... 7 9 The nitriles are easily hydrolysed on boiling with aqueous sodium hydroxide yielding the corresponding acid and are soluble in the usual organic solvents. o- m- aitd p-Tolyloxyace tanilides C15H150,N. Them anilides may be readily prepared by the following method. Ten grams (1 md.) of the acid chloride were dissolved in 100 C.C. of dry benzene in a flask fitted with a reflux apparatus and 10 grams (2 mols.) of freshly distilled aniline slowly added with vigolrous shaking. After cooling the solid which had separated was filtered and the benzene solutioa containing some aniline was evaporated and the residue added to the main bulk. The whole was then triturated with dilute hydrolchloric acid to remove aniline hydrochloride and finally treated with cold dilute sodium carbonate solution to remolve traces of acid.Tho substance was then crystallised from ethyl alcohol in which the ortho-compound was the most and the para-compound the least soluble the meta-compound being moderately soluble. They are readily hydrolysed on boiling with dilute hydrochloric acid STUDIES IN TRE COUMARBNONE SERIES. PART I. 1541 M. p. o-Tolyloxyacetanilide (cubes) .................. 108.5" m- 9 9 (needles) 96 7 9 ( , ) .................. 109 .................. 23-These values are in closta agreement8 with those given by Forte (ibid.). 4-Methylcoamaran-2-one (I). Ten grams (1 mol.) of p-tolyloxyacetyl chloride were dissolved in 15 grams (3 mols.) of dry benzene and 8 grams of finely powdered aluminium chloride slowly added with vigorous shaking, the mixture being cooled in ice.The reaction proceeded slowly, and the solution became dark red but if the temperature was allowed to rise the mixture rapidly turned into a dark-coloured, tarry mass and the yield of cumaranone was considerably diminished in consequence of decomposition. The red solution, after remaining in the cold for two holurs was poured on polwdered ice mixed with hydrochloric acid and the oll suspended in the water submitted to distillatio'n in a current of steam for several hojurs until the distillate no longer showed a reducing action with Fehling's solution and the characteristic odour of the coumaranone was imperceptible. The upper layer of benzene containing some coumaranone in solution was separated from the aqueous layer, and the latter was extracted with ether several times to remove coumaranone dissolved in the water.The combined benzene and ethereal extracts were dried over anhydrous sodium sulph ate the solvents removed on the water-bath and on cooling the residue solidified to a mass of fine needles. The substance crystallised from ethyl alcohol in white needles melting a t 5 1 O . The yield was 3 grams. The semicarbazone (yellow needles from alcohol) melted a t 228O (heated slowly) and the oxime a t 143-144O (Stoermer and Bartsch, Zoc. cit. give 1 8 1 O and 144O respectively). The p-nitrophenylhydrazone was obtained by adding a solution of p-nitrophenylhydrazine in dilute acetic acid to an aqueous solution of the coumaranone.The hydrazone soon separated as a red crystalline powder and after crystallisation from dilute acetic acid it melted and decomposed a t 230-232O (Found N=14*4. C,,H,,O,N requires N = 14.85 per cent.). 6 -Methylcoumaran-2-cne was prepared in a manner similar to that described above from m-tolyloxyacetyl chloride and was obtained in needles melting a t 8 5 O . The semicarbazone and mime melted a t 208O and 1 5 6 O respectively. For the latter Fries and Finck (Zoc. cit.) give 1 6 5 O probably a misprint. The p-nitro 1542 COPISAROW CARBAZOLE-BLUE AND CARBAZOLE-VIOLET. phenylhydrazone crystallises from dilute acetic acid in red needles melting and decomposing a t 214-216O. 6 -Bet hylco umaran-2- one prepared f rom o-tolyloxy acetyl chloride, crystallised in needles melting a t 8 8 O . The semicarbazone and oxime melted a t 2 2 7 O and 148O respectively and the p-nitroplzenyl-hydrazone (red needles) a t 195O with decomposition. The above coumaranones give purple colorations with ferric chloride they reduce Fehling’s solution on warming and give an immediate precipitate of silver with Tollens’ reagent. On treat-ment with concentrated sulphuric acid they dissolve with the production of an intense red colour and subsequent formation of tarry matter. On remaining in the air for some time the crystals of the cournaranones become coated with a bright red-coloured substance probably an oxidation product. They have a characteristic odour resembling that of hyacinth. The authors desire to thank Professor 1,apworl;h for facilities placed at their disposal in connexion with this investigation and also Dr. W. \V. Adamssn for analysing several of the substancM described above. CHEMICAL DEPARTMENT, THE UNIVERSITY, MANCIIESTER. [Received October 14th 1920.