摘要:

SODIUM HYPOCHLORITE AND AROMATIC SULPHONAMIDES. 371 XLL-The Action of Sodium Hypochlorite on the hornatic Sulphonamides. By HENRY STANLEY RAPER, JOHN THOMAS THOMPSON, and JULIUS BEREND COHEN. ALTHOUGH the general nature of the chemical change * resulting from the interaction of sodium hypochlorite and the aromatic sulphon- amides resembles that described by Chattaway and Orton in the case of the aromatic acylamines (Trans., 1899, '75, 1046 ; 1900, 77, 134, 789 ; Ber., 1899, 32, 3573), yet there are certain well-marked differ- ences in the character of the products which we venture to think are not without interest. These authors have shown that by the action of sodium hypo- chlorite on the acyl-anilides and -toluidides, the halogen in the first * A preliminary notice of this reaction has already appeared (Proc., 1901, 17, We have now supplemented these experiments by extending the reaction to 262).a number of other sulphonsmides.372 RAPER, THOMPSON, AND COHEN: THE ACTION OF Aniliiie ........................... o-Toluicline ..................... 112-Toluidine.. ................... p-Toluidine .................... m-4-Xylidine .................. a-Nnphthylamine ............ instance enters the para-position with respect to the amino-group if this position is free, but otherwise becomes attached t o an ortho-carbon atom. The orientating effect of the sulphonic group is different ; in the first instance, the halogen seeks the ortho-position with respect t o the amino-group, thus the benzenesulphonamides of aniline, p-toluidine, m-4-xylidine, and P-naphthylamine form ortho-compounds almost exclusively.If, however, a methyl group is present and the para- position to the amino-group is free, the conditions are modified and the halogen enters either the ortho-position to the methyl group or the para-position to the amino-group. Thus, the sulphonamide of m-toluidine gives a derivative of 6-chloro-m-toluidine. There is a considerable difference also in the readiness with which the different sulphonamides react. This is very marked in the case of those bases which are substituted in the ortho-position to the amino-group. The sulphonamides of o-toluidine and m-4-xy lidine are only acted on slowly, whereas the sulphonamide of a-naphthylamine, which may be regarded as an o-substituted compound, is quite unchanged by the hypochlorite.The following table gives a list of the compounds which have been studied : ArNHoS0,*C'6H, ' ArNH,. Melting p i n t of t h e sulph onainide. 110" 123-124 95 120 124-125 168 97-9s ~ ~ I'osition o f t h e 1 ' ~ o g e n relative to, 1 the smino-group. I I ~ (I- 2.1 - P- , l op- dichl oro- ' compound ' i 0- I 0- no nctioii I 0- I Melting point of the chloro- sul phonamide. 129-130" 124-125 130 114 110 148-149 - 130-131 The Action of Sodium Hypochlorite on Benzenesulphonunilide. Twenty grams of benzenesulphonanilide were dissolved in 200 C.C. of a solution of sodium hypochlorite (1 C.C. = 0.03 gram Cl) in the cold and lef t for 12 hours ; the brown solution was acidified with acetic acid and the semi-solid precipitate was filtered and washed.I t was then dissolved in about twice its weight of glacial acetic acid, to which a few drops of strong sulphuric acid were added, and digested for a n hour on the water-bath. The product was poured into water, from which n crys- talline compound separated, and, when purified by recryst:illisation from dilute acetic acid, melted a t 129-130°; 18 grama of the crude product or 11 grams of purified substance were obtained,SODIUM HYPOCHLORITE ON THE AROMATIC SULPHONAMIDES. 373 The mother liquor contained a small quantity of a semi-solid sub- stance which slowly became crystalline, and after recrystallisation from benzene and petroleum melted at 116O. This substance is probably benzenesulphon-p-chloroanilide (m. p. 12 lo), but it is not readily freed from oily impurity. On analysis, the substance melting at 129-130' gave the following numbers : 0992 gave 14.15 C.C.moist nitrogen a t 1 5 . 5 O and 758 mm. N = 5.66. 0,2097 ,, 0.1820 BaSO,. S = 11 -93. C,,H,,O,NClS requires N= 5.23 ; C1= 13.27 ; S = 11.90 per cent. The constitution of the substance was determined by hydrolysis ; two grams were heated with about 8 C.C. of concentrated hydrochloric acid in a sealed tube a t 190' for 4-5 hours. On opening the tube there was a strong odour of benzene, the presence of which was con- firmed by conversion into aniline, with which sodium hypochlorite gave the usual reaction. The appearance of benzene in tbe decomposition of the sulphonanilide is readily accounted for, seeing that Armstrong and Field (Ber., 18i4, 7, 406) and Jacobsen (Ber., 1876, 9, 258) have shown that sulphonic acids are converted into hydrocarbons by strong hydrochloric acid under pressure.The acid solution, after extract- ing with light petroleum, was made alkaline with sodium carbonate and distilled in steam ; the distillate was extracted with ether, the ether removed, and the residue dehydrated in a vacuum desiccator, when a yellow oil weighing 0.8 gram remained. The substance did not solidify on introducing a crystal of p-chloroaniline, even when cooled in ice; i t yielded an acetyl derivative melting a t 85-86' and a benzoyl derivative melting at 102'. This corresponds with o-chloroaniline, which was prepared for comparison and converted into the acetyl, benzoyl, and benzenesulphonyl derivatives.The reaction probably occurs in two stages, as in the chlorination of the acetyl derivatives studied by Chattaway and Orton, although the intermediate chloroamide was too unstable to be isolated in a pure state. 0.3285 ,, 0.1764 AgC1. C1= 13-57. N H *SO,Ph NCl *SO,Ph NH*SO,Ph /\ + f)C1 -+ I I A. I 1 \/ \/ \/ The main product is therefore the o-chloroanilide, together with a very small quantity of the para-compound. Wallach (Bey., 1877, 0, 424) found that benzenesulphonanilide, when heated at 100' with phosphorus pentachloride, yields benzenesulphon-p-chlorosnilide (m. p. 121'1, but no reference is made to an ortho-compound. We YOL. LXXXV. c c374 RAPER, THOMPSON, AND COHEN: THE ACTION OF have repeated and confirmed Wallach's work. The product is difficult t o purify, and only a small quantity of pure substance could be obtained, but we failed t o detect any ortho-compound.We have also synthesised Wallach's compound from p-chloroaniline and benzene- sulphonic chloride, and to complete the series we have also prepared benzenesulphon-m-chloroanilide from m-chloroaniline. This anilide, which has not been previously described, melts a t 122' like the para-compound. 0,2576 gave 12.85 C.C. moist nitrogen at 13'and 761 mm. ; N=5-'76. C,,H,,O,NClS requires N = 5.23 per cent. T!L~ Action of Sodium Hypochtorite onl Benxenesu~i~on-o-tolzcidide. Benzenesulphon-o-toluidide was described by Beckmann and Fellrath (Annabn, 1893, 273, 13) ; it crystallises from alcohol in rectangular tablets which melt a t 123-124'. When treated with sodium hypochlorite in the manner already described, i t gave an un- crystallisable product, but in acetic acid solution a colourless crystal- line substance was obtained.If warmed with alcohol for any length of time, this compound decomposed, a portion of the alcohol being oxidised t o aldehyde. From a mixture of benzene and light petroleum, the substance crystallised without any decomposition, and a product was obtained which melted a t 99-100'. 0.1951 gave 0.1000 AgC1; C1= 12.92. On crystallising the foregoing compound from acetic acid solution, 0.1014 gave 0-0530 AgC1; C1= 12.93. 0.1014 ,, 0.0843 BaSO, ; S = 11.41. C,,H,,O,NCIS requires C1= 12-61 ; S = 11.39 per cent. This substance was identified as benzenesulphon-5-chloro-o-toluidide, also obtained from benzenesulphonic chloride and 5-chloro-o-toluidine ; The chlorine therefore enters the para-position to the amino-group.The isomeric substance melting at 99-100' is probably the chloro- amide, C7H7*NC1*S0,*C,H,, for on heating with alcohol it loses chlorine and reverts t o the original substance. This explains both the conversion effected by acetic acid and the decomposition produced by alcohol, in which the latter reagent is oxidised to aldehyde in eliminating chlorine. C,,H,,O,NClS requires C1= 12.61 per cent, it becomes transformed into an isomeride melting a t 124-1259SODIUM HYPOCHLORITE ON THE AROJIATIC SULPRONAMIDES. 375 me Action of Sodium Hypochlorite on Bsnxenesulphon- m-$oluidide. Benzenesulphon-m-toluidide has not been previously described ; it crystallises from alcohol in colourless plates melting at 95".The action of the hypochlorite solution on this substanue is much more vigorous than in the preceding cases, both mono- and di-chloro9 derivatives being formed. The monochloro-compound was obtained by using the theoretical quantity of sodium hypochlorite and keeping the mixture cold ; after 12 hours, a large quantity of crystalline substance separated, which melted a t 275-280'. This is probably a sodium derivative having the formula C,H,Cl(CH,).NNa*SO,.C,KS ; it crystallises from ethyl acetate in fine needles. 0°8808 required 28.87 C.C. N/10 H,SO, ; Na = 7.54. C,,H,,O,NClSNa requires Na = 7-58 per cent. When digested for n few minutes with glacial acetic acid, the sodium compound is easily decomposed, yielding the mono-chloro-derivative, which crystallises in long, colourless prisms and melts a t 130'.Five grams of the toluidide yielded 3.2 grams of the pure mono-chloro- compound. 0,3790 gave 0.1970 AgCl ; C1= 12-86, 0.3790 ,, 0,3144 BclSO, ; S = 11.39. C1,H,,O,NGIS requires C1= 12.60 ; S = 11.37 per cent. The constitution of the substance was determined by hydrolysis as in the previous cases. A crystalline base was obtained which melted at 83' and gave an acetyl derivative melting at 89'. These numbers correspond with the melting points of 6-chloro-m-toluidine and its acetyl derivative. The formula of the original compound is therefore Me c1/\ (/NR*SO,Ph. No other mono-chloro-compound could be detected. The Dichloro-compound.-Five grams of the sulphon-m-toluidide were dissolved in excess of the hypochlorite solution ; the whole mass subsequently solidified and was collected at the pump.After crystal- lisation from acetic acid, 3 grams of pure substance (m. p. 114') and 1.7 grams of an impure product melting at about 80' were obtained. The latter is probably a mixture of the mono- and di-chloro-compounds, which are separated with difficulty. A better yield of the dichloro- compound was obtained either by warming the mixture of hypochlorite c c 2376 RAPER, THOMPSON, AND COHEN: THE ACTION OF and m-toluidide at about 40Ofor a short time or by adding sodium hypochlorite solution to the monochloro-compound. The substance melting at 114O was analysed with the following result : 0.2640 gave 0.2303 AgC1. Cl= 21.60.0,2640 ,, 0.2030 BaSO,. S= 10.53. On hydrolysis, acrystalline base was obtained, which was recrystallised from alcohol and melted at 58O, its acetyl derivative m e l t i q at 121'. The base was further identified by successive conversion into 2 : 3 : 6-tri- chlorotoluene (m. p. 45-46'}, 2 : 3 : 6-trichlorobenzoic acid (m. p. 165O), and dinitro-2 : 3 : 6-trichlorotoluene (m. p. 141-142') (compare Cohen and Dakin, Trans., 1902, 81, 1333). The structure of the sulphondichloro-m-toluidide is accordingly represented as follows ; C,,H,,O,NCl,S requires C1= 22.4 ; S = 10.10 per cent. Me CI/\Cl I INH*SO,Ph It is interesting to compare the action of the hypochlorite on the benzenesulphonyl and acetyl derivatives of m-toluidine. In the acetyl derivative, the hypochlorite produces substitution in the ortho- and para-positions to the methyl group (Cohen and Dakin, Trans., 1902, 81, 1332); in the sulphonamide, the chlorine atoms enter the two ortho-posi tions.\/ Action of Sodium H ypochlorits on Benxeneaubhon-p-toluidide. Benzenesulyhony-toluidide was described by Wallach (Be?-., 1876, 9, 424), who states that it does not yield a halogen derivative with phosphorus pentschloride, and attributes the fact to the presence of the methyl group in the para-position to the amino-group. With sodium hypochlorite, the reaction goes quite smoothly, and benzene- sulphon-3-chloro-p-toluidide (m. p. 110') is obtained, 5 grams of the toluidide giving 3.5 grams of pure product. 0-4020 gave 0.2068 AgCl. C1= 13-73, 0.4020 ,, 0.3383 BaSO,. S = 11.54.C,,H,,O,NClS requires C1= 12.60 ; S = 11 ~ 3 7 por cent. On hydrolysis with hydrochloric acid, 2 grams of the compound gave 0.9 gram of a yellow, oily base, which solidified on coolingin ice, and yielded an acetyl derivative melting at 1 1 8 O . These properties agree with those of the 3-chloro-p-toluidine. The chlorine, therefore, enters the ortho-position to the amino-group. An excess of hypo- chlorite produced no further substitution.SODIUM HYPOCHLORITE ON THE AROMATIC SULPHONAMIDES. 377 The Action of Sodium Hypochlorite on Benxenesulphon-m-4-xylidide. Benzenesulphon-m- 4-xylidide, which has not been previously de- scribed, is prepared in the usual way, and crystallises in colourless prisms meltiug at 124-1 25'. When treated with sodium hypochlorite solution, it remained unchanged, but in acetic acid a small amount of chlorination product was obtained, the yield being augmented by carrying out the reaction at about 50'.The white solid which separ- ated was recrystallised from alcohol and melted at 148-149'. Six grams of the m-xylidide gave 4.2 grams of pure product. 0.1293 gave 0.0619 AgC1. C1= 11-84. 0.1293 ,, 0.1004 BaSO,. S = 10866. On hydrolysis, 2 grams of the chlorination product gave 0.8 gram of a colourlees, crystalline base melting at 39-40', which formed an acetyl derivative melting a t 200-201°. The base obtained on hydro- lysis is identical with 5-chloro-m-4-xylidine.* The sulphon-5-chloro- m-Cxylidide has therefore the following structure : C,,H,,O,NClS requires C1= 11-89 ; S = 10.72 per cent. Me /\ Cl!, )Me ' An excess of sodium hypochlorite had no further action.Sodium Hypochlorite and Benzenesulp~on-a-nap~thalide. The sulphon-a-naphthalide obtained by the action of benzenesulph- onic chloride on a-naphthylamine crystallises from alcohol in colour- less needles melting at 168' ; it dissolves in aqueous sodium hypo- chlorite forming a deep brown solution, which, when poured into acetic acid, yielded a precipitate of the original substance. No change mas effected by the hypochlorite in the presence of acetic acid. The Action of Sodium Hppochlorite on Benxenesulphrm-P-nnthalide. Benzenesulphon-/I-naphthalide crystallises from alcohol in colourlees plates, which melt at 97-98'. On adding sodium hypochlorite, a yellow oil separated, which quickly solidified, the product being a For purposes of identification, the preparation of the three isomeric mono- chloro-derivatives of m-4-xylidine was undertaken, and an account of these sub- stances will appear separately.378 McKENZIE : THE ESTERIPICATION OF sodi urn derivative of ben zenesuI phonchloro-P-naph t halide, w hie h crys- tallises in glistening needles from benzene containing about 5 per cent.of alcohol. 0,8837 required 2 *45 C.C. NH,SO,. Na = 6.38. C,,H,,O,NCISNa requires Na = 6.76 per cent. The sodium compound was decomposed by boiling for a few minutes with dilute mineral acid ; the product dissolved in alcohol, from which it crystallised in colourless needles melting a t 1YO-13l0; 5 grams of p-naphthalide gave 4 grams of pore substance. 0.1843 gave 0.0840 AgCl. C1= 11.28. 0,1843 ,, 0.1355 BaSO,. S= 10.10. C,,H,,O,NClS requires C1= 11.18 ; S = 10.08 per cent. The compound is more easily hydrolysed than the benzenesulphon- toluidides, a temperature of 140' for one hour being suEcient for the purpose; 1.8 grams gave 0.7 gram of solid base, which melted a t 60' and formed an acetyl derivative melting a t 14'7-148'. These data correspond with the melting points of l-chloro-/3-naphthylamine and its acetyl derivative. The sulphonamide, therefore, has the following structure : c1 /\/\NH*SO,Ph. I l l I n this case, the chlorine atom enters the a-position contiguous to the amidic nitrogen. \/\/ THE YORKSHIliE COLLEQE, LEEDS.

ISSN:0368-1645    

DOI:10.1039/CT9048500371

出版商:RSC    

年代:1904

数据来源: RSC

摘要:

378 McKENZIE : THE ESTERIPICATION OF XLII.--T'e Esterijication of r-Mxndelic Acid by Menthol and Borneol. Ey ALEXANDER MCKENZIE. IT has been shown by Marckwald and Chwolles (Ber., 1898, 31, 783) that the resolution of an i-acid into its optically active components by means of alkaloids is due solely to the difference in solubility between the alkaloidal salts of the d- and 2-acids respectively, and not, as Pasteur supposed, to any difference in affinity exhibited by the two enantiomorphous acids towards one and the same base. This conclu- sion was based on the following observations : (1) that correspondingR-MANDELIC ACID BY MENTHOL AND BORNEOL. 379 amounts of the acid cinchonine salts of d- and I-tartaric acids respect- ively exhibited in solution the same rise of boiling point, (2) that the rate of hydrolysis of methyl d-tartrate by nicotine was the same as that of methyl I-tartrate, and (3) that, when methylethylacetic acid in aqueous solution was shaken with a quantity of brucine insufficient for complete neutralisation, the uncombined acid was optically inactive. The formation of a salt, therefore, by union of a n active acid with an active base takes place practically instantaneously, and may be considered as being essentially an ionic reaction.Whilst, then, a d-acid combines with an active base with the same velocity as its I-isomeride, the case might conceivably be different when the action depends on the spacial arrangement of the atoms in the molecule. That ester formation is a case in point is sufficiently indicated by the experiments of Menschutkin on the dependence of the rate of ester formation on the structure of the carbon chain, and by those of V.Meyer on diortho-substituted aromatic acids. The basis of Pasteur’s two cry stallisation methods for resolving i-com- pounds is undoubtedly physical, but his third method, the use of micro-organisms, is probably biochemical (compare E. Fischer, Zeit. physiol. Chem., 1898, 26, 60). On those considerations, Marckmald aDd McKenzie (Bey., 1899, 32, 2120) studied the esterification of r-mandelic acid by I-menthol and isolated I-mandeljc acid by a chemical method, distinct from Pasteur’s biological method in so far that the substances concerned are of known composition. When r-mandelic acid was heated with I-menthol a t 155O, the un- esterified acid was lavorotatory, and from i t a specimen of I-mandelic acid was prepared.The rate of ester formation of the enantiomor- phous acids towards one and the same active alcohol was accordingly different, since I-menthyl d-mandelate was formed more quickly than Fmenthyl E-mandelate. The esterified product, on the other hand, which was expected to contain an excess of I-menthyl d-mandelate, did not,, however, yield a dextrorotatory acid on hydrolysis, but was either inactive or feebly Iaevorotatory. It appeared, in fact, as if the ?--acid mas being converted into the I-isomeride without a corresponding pro- duction of the d-isomeride. I n order to account for this confusing result, experimental evidence was subsequently (Ber., 1901, 34, 469) submitted to show t h a t I-menthyl d-mandelate suffers partial race- misation during its formation more quickly than I-menthyl I-mandelate.Similar results were obtained (Zoc. cit.) with i-phenylethoxyacetic acid, but with i-ethoxypropionic acid there was no abnormality, since the rotation of the acid recovered from the esterified product was of the opposite sign t o that of the unesterified acid. Now Victor Meyer has found t h a t in the case of two isomeric aromatic esters, the particular one which is formed with the greater380 McKENZIE : THE ESTERIFICATIOK OF difficulty is also the less readily hydrolysable. His conclusion is based on experiments made with chloronaphthoic, bromobenzoic, and other acids (Ber., 1895, 28, 1262), and has been confirmed by Weg- scheider (Be?.., 1895, 28, 1468) and by Briihl (Ber., 1895, 28, 1913, 2868).Accordingly, I-menthyl d-mandelate and I-menthyl I-mandelate, which are not mirror-images, undergo hydrolysis by the same inactive base a t different rates. Of the two, the former is the more quickly produced and is the more quickly hydrolysable. The mixture of un- equal amounts of the two esters, which, on complete hydrolysis, yielded a feebly laevorotatory acid, gave, on partial hydrolysis, a dextrorotatory acid. The applicability of this method for resolving inactive compounds was further exemplified (Ber., 1901, 34, 469) in the case of sec-octyl alcohol. Whilst the difference in the rate of Formation of the tar- trates of d- and I-octyl alcohols was slight, there was a marked differ- ence in their rate of hydrolysis, so that, by the method of fractional hydrolysis, a partial revolution of the i-alcohol was effected.I n this connection, attention has already been directed (Ber., 1900, 33, 210) to the fact that Frankland and Price, in the esterification of racemised amgl alcohol by d-glyceric acid, observed that the unesterified alcohol was feebly laevorotatory (Trans., 1897, 71, 253). I n the light of the results obtained by Marckwald and the author, Walden has interpreted observations made by him during the inves- tigation, firstly, of the action of active amyl alcohol on inactive a-bromo- propionic bromide, and, secondly, of the action of active amyl iodide on inactive silver methylsuccinate (Ber., 1899,32,2703). Scholtz has studied the action of active amyl iodide on inactive 1-methyl-2-pipecoline in the hope that the formation of the quaternary iodide of the d-base would take place with a velocity different from that of the formation of the iodide of the I-base (Bey., 1901, 34, 3015); he obtained, however, only negative results.The following experiments include an examination of the action of borneol on Ir-mandelic acid, whilst the interaction of I-menthol and 1.-mandelic acid has also been further investigated. E x P E R I M E N T A L. Este?v&ation of r-MundeIic Acid by Borneol. Expt. 1.-The borneol employed bad the rotation [a];'- 37.4O (C = 17.374) in ethyl-alcoholic solution, a value in agreement with that of Haller (Anfi. Chim. PhYS., 1892, [vi], 2'7, 395). Twenty grams of r-mandelic acid were heated with 20 grams of 2-borne01 in an oil-bath at 16OOfor la hours.The product was dissolved in ether and theR-MANDELIC ACID BY MENTHOL AND BORNEOL. 381 unesterified acid separated from the esters by shaking the ethereal solution with a dilute aqueous solution of sodium carbonate. I n order t o separate effectively any traces of borneol or borneol esters present in the solution of sodium mandelate, the latter was concen- trated by evaporation on the water-bath and then extracted with ether. It was next acidified and extracted with ether, when 9.3 grams of mandelic acid were obtained with the following specific rotation in aqueous solution : I = 2 ; c = 7.48 ; a? - 0.58' ; [a]?- 3.9'. After distilling the ether from the ethereal solution containing the esters, the latter were hydrolysed by heating with an excess of alcoholic potassium hydroxide.When the alcohol had been removed, water was added and the precipitated borneol drained off. The filtrate wars completely freed from borneol as before; the mandelic acid isolated from it by acidification and extraction with ether was slightly Isvo- rotatory, the following data being obtained in aqueous solution : This experiment shows that d-mandelic acid is esterified by I-borneol more quickly than is I-mandelic acid. The unesterified acid is lsvo- rotatory, but the mixture of esters formed did not yield a dextro- rotatory acid, but a feebly lzevorotatory one. The latter result is to be attributed to the unequal rates of racemisation of I-bornyl d-man- delate and I-bornyl I-mandelate, the former being racemised during ifs formation more quickly than the latter.Expt. 2.-In order that the esterification might be more complete than in the previous case, 20 grams of r-mandelic acid were heated with 20 grams of I-borneol for 6 hours at 150'. The product was manipulated as before ; 3.7 grams of unesterified acid were recovered, and a determination of its specific rotation gave the following result : I = 2 ; c = 7.4 ; a r - 1.07" ; [ a ] r - 7.2' (in aqueous solution). The acid (15.5 grams) obtained by hydrolysing the mixed esters by an excess of alcoholic potassium hydroxide was feebly lzevorotatory, thus, I = 2 ; c = 10 ; a r - 0.27' ; [ a ] r - 1.3' (in aqueous solution). Expt. 3.-The effect of esterifying for a longer time a t a Iower temperature was next examined.Ten grams of r-acid were heated with 30 grams of I-borneol for 28 hours in a water-bath heated to boiling. The unesterified acid (2.6 grams) had the following rotation : I = 2 ; c= 10.4 ; ago - 2-18' ; [a]:'- 10.5' (in aqueous solution). The acid (6.6 grams) obtained from the esters was again feebly lsvorotatory, I = 4 ; c = 10.4 ; a F - 0.17'. This sign of rotation was confirmed by crystallising the 6.6 grams from water, withdrawing a crop of the r-acid and fiuding that the filtrate mas laevorotatory. The 3.7 grams of acid from Expt. 2 and the 2.6 grams from Expt. 3 were united and crystallised from water. After the removal of 3.1 grame of r-acid, the filtrate was converted into magnesium 2=3; ~ ~ : 8 . 5 4 ; aF-0.25'; \a-J$-1*5'.382 McKENZIE : THE ESTERIFICATLON OF s a l t ; a small crop of magnesium r-salt was withdrawn, and the filtrate from this acidified and extracted with ether.The resulting acid WBS next converted into cadmium salt, from the aqueous solu- tion of which a crop of cadmium r-mandelate was withdrawn, and the filtrate then decomposed by hydrogen sulphide. The resulting acid gave the following result : I = 2 ; c P 1.586 ; az - 3.53' ; [u]r - 111.3' (in aqueous solution). The isolation of the pure I-acid was not, however, attempted i n this case, the object of the experiment being simply t o show how the r-acid may be gradually eliminated from a mixture containing a large excess of it, together with a little of the I-isomeride. Expt. 4.-When r-mandelic acid (1 mol.) was heated with Kahl- baum's dextrorotatory borneol (1 mol.) for 1 hour at 155', the un- esterified acid was dextrorotatory, I = 4 ; c = 10 ; UY + 1 *SOo ; [a]T + 4.5' ; whilst the esterified product yielded a slightly dextro- rotatory acid.The rotation of the borneol used, [.IF+ 23' for c = 18,144 in ethyl alcoholic solution, indicated that it was R mixture of d- and iso- borneols. Expt. 5.--s-Mandelic acid was next so completely esterified by I-borneol in such a manner that no fractional esterification took place. The I-bornyl dEmandelate was then submitted t o frac- tional hydrolysis. Ten grams of r-acid were heated in a boiling water- bath for seven hours with 30 grams of I-borneol, a current of dry hydrogen chloride being passed into the mixture from time t o time, The product was dissolved in ether and shaken with dilute aqueous sodium carbonate until all the free acid had been removed.The esteri- fication was practically complete, since only 0.2 gram of mandelic acid was recovered. After expulsion of the ether from the ethereal solu- tion, the resulting oil was dissolved in ethyl alcohol and heated on cz boiling water-bath for 10 hours with 1 gram of potassium hydroxide, an amount insufficient for the complete hydrolysis of the esters. The potassium salt, when freed completely from the residual borneol and unesterified oil, yielded 1.7 grams of a Ievorotatory acid : I = 2 ; c = 6.8 ; The residual esters, on the other hand, after being heated with a large excess of alcoholic potassium hydroxide, gave 6-8 grams of an acid, a saturated aqueous solution of which, when examined with a polarimeter, reading to O.Olo, was inactive, as also was the filtrate obtained after removing a crop of 5 grams.A similar result was obtained in a subsequent experiment, where a larger amount of alkali was used in the initial hydrolysis. The mixed esters from 10 grams of acid were hydrolysed by 3.1 grams of potassium hydroxide, when 5.5 grams of acid were obtained, giving - 0.90' ; [*IF - 6.6' (in aqueous solution).R-MANDELIC ACID BY MENTHOL AND BORNEOL. 383 the following data : I = 2 ; c = 14 ; UF - 1 * 0 8 O ; [ u]y - 3.9" (in aqueous solution). The second hydrolysis, with excess of alkali, again gave an inactive product. The fractional hycirolysis of I-bornyl dl-mandelate thus proceeded in an unexpected manner.Although I-bornyl d-mandelate is more readily formed than I-bornyl I-mandelate, the latter appeared to be the more readily hydrolysable of the two. The formation of r-mandelic acid as the product of the final hydrolysis is discussed in the sequel. Esterajication of r-MandeIic Acid 6y l-MenthoZ. Marckwald and the author (Zoc. cit.) had previously isolated Lmandelic acid by means of the fractional esterification of the r-acid. I n the following experiment, the isolation of the d-acid was effected by the method of fractional hydrolysis. The esterification of r-mandelic acid with I-menthol according to the Fischer-Speier method was practically complete. After removal of the mineral acid and a trace of mandelic acid, the product was hydrolysed by a calculated amount of alcoholic potassium hydroxide solution sufficient for the hydrolysis of about half of the ester mixture, The resulting solution of potassium salt was separated from the residual esters and menthol and then yielded a mixture of mandelic acids, with [ u]" + 4.5' (c = 10.075) in aqueous solution.From the latter product, a specimen of d-mandelic acid, with the correct melting point and specific rotation, was isolated by the method of separation for the I-acid indicated by Marckwald and the author. The difference in the rate of hydrolysis of the two esters under con- sideration is more marked towards the beginning of the hydrolysis than later, since a more active product than in the case just described may be obtained by lessening the amount of alkali in the initial hydrolysis.Thus, when 10 grams of r-acid were heated with 30 grams of menthol in the presence of hydrochloric acid, and the product, after the removal of the free acid, was partially hydrolysed by an amount of alkali considerably less than half that required for complete hydrolysis, 1.2 grams of acid were obtained having [u]: + 1 1 . 3 O (c = 4.8) in aqueous solution. The residual esters yielded 6-5 grams of an acid with [a], - 2 . 5 O (c = 13) in aqueous solution. It should be noted, however, that the latter hydrolysis was quickly conducted by heating on tbe water-bath for half an hour with an amount of alcoholic potassium hydroxide just sufficient for the purpose. I-Menthyl dl-manndelate was prepared by heating r-mandelic acid on a boiling water-bath for 10 hours with 3 times its weight of kmenthol in presence of hydrogen chloride.From the ethereal solution of the384 McKENZIE : THE ESTERIFICATION OF product, the Free acid was removed by dilute sodium carbonate, when i t was found that only a trace of mandelic acid was obtained, from which it was concluded that the esterification was practically complete. After the ethereal solution had been dried, the ether was expelled and the oil distilled under diminished pressure. The menthol was completely removed and finally an oil, boiling at 225' under 30 mm. pressure, distilled over; this quickly solidified to a hard cake, which on analysis gave C = 74.4 ; H = 9.0 ; C,,H,,O, requires C = 74.5 ; H= 9.0 per cent. The product, which dissolves only sparingly in light petroleum, from which it separates in feathery, asbestos-like crystals, is readily soluble in benzene, chloroform, ethyl acetate, acetone, or pyridine, and moderately so in ethyl alcohol, from which it crystallises in needles.It is practically insoluble in water, with which it may be boiled for several hours without undergoing any appreciable hydrolysis. It melts a t 85-86'. A determination of its rotation in ethyl alcoholic solution gave the following result : I = 2 ; c = 10.890 ; aF - 16-17'; [a]F - 74.2'. Frankland first suggested the possibility of the resolution of esters of this type by submitting them to fractional crystallisation (Pasteur iMernos.iaZ Lecture, Trans., 1897, 71, 696. Compare also Wohl, Ber., 1898,31,2394), and Frankland and Price have shown (Trans., 1897,71, 253) that, although the ester formed from active amyl alcohol and i-di- benzoylglyceric acid is distinctly crystalline, no separation took place on crystallieation. A similar experience to the latter was encountered when the mandelate just described was crystallised from light petroleum and when the individual crops were allowed to separate at the temperature of the laboratory.The crystalline form of successive crops appeared to be perfectly homogeneous. After 10 crystallisations, the melting point of the product was the same as that of the original substance, whilst [a]y in ethyl alcoholic solution was -72.7' (c= 4.5624), a value agreeing with that previously quoted ; the result was the same when the ester was repeatedly crystallised from ethyl alcohol, no resolution having been effected.I-Menthyl dZ-mandelate is accordingly a partially racemic compound, and further attempts are being made to resolve it by varying the temperature conditions. Wheu I-menthyl dl-mandelate is hydrolysed with an excess of alcoholic potassium hydroxide, it yields r-mandalic acid, When hjdrolysed with an amount oE alkali insufficient for complete hydrolysis, but in considerable excess of that necessary for the hydrolysis of half the ester, the resulting potassium salt yields a lsevorotatory acid : 14.5 grams of the ester (corresponding with 7.6 grams of acid) were hydrolysed in the manner indicated and, fromR-MANDELIC ACID BY MENTHOL AND BORNEOL. 385 the potassium salt, 6 grams of acid were recovered giving the following data: I = 4; c = 12; uD - 2*08O; [a], - 4.3' in aqueous solution.The residual esters, when hydrolysed with an excess of alkali, yielded 1.1 grams of inactive acid. It is therefore possible by the fractional hydrolysis of I-menthyi dl-mandelate to obtain either a dextrorotatory or hvorotatory acid from the initial hydrolysis, according to the amount of alkali used. When the amount of alkali used is much less than that required for the hydrolysis of half the mixed esters, the acid resulting from the potassium salt formed is dextrorotatory. When the concentration of the alkali is increased, a larger quantity of dextrorotatory acid is obtained, the specific rotation of which is smaller than in the former case. Finally, by increasing still further the concentration OF the alkali, a lzevorotatory acid is obtained.In the experiments on fractional hydrolysis described in this paper, it has been repeatedly observed that the final hydrolysis of the mixed esters with alcoholic potassium hydroxide in excess led to the forma- tion of a totally racemised product, and the following experiments were accordingly conducted. Five grams of I-mandelic acid, melting at 133' and having [a], - 15S3 in aqueous solution, were esterified by heating on a boiling water-bath for 7 hours with 15 grains of 2-menthol in presence of hydrogen chloride. Alcoholic potassium hydroxide was then added to the product and the mixture boiled for 2 hours. After removal of the alcohols, the acid, recovered from the potassium salt, melted at 116-118' and gave [a], - 3.3' ( c = 707436) in aqueous solution. I n an experiment conducted with I-borneol in an analogous manner, the recovered acid gave [a], -9.7' (c = 6.432) in aqueous solution.Although it has been shown that I-mandelic acid yields i-phenyl- chloroacetic acid when heated with fuming hydrochloric acid at 95-100' in a Ljealed tube (Easterfield, Trans., 1891, 59, 72), and that d-mandelic acid by the action of phosphorus pentachloride somet,imes yields i-phenylchloroacetyl chloride (Kipping, Trans., 1903, 83, 1005), it was not considered probable that the partial racemisation observed in the experiments quoted took place during the actual esterification of the acid, but rather during the subsequent hydrolysis. That excess of alkali does actually racemise potassium I-mandelate was shown by experiment.One gram of I-mandelic acid was added to a solution of 4 grams of potassium hydroxide in 30 C.C. of water. The solution, examined in a 2-dcm. tube, gave about ago - 7*6', and did not appear to racemise after 12 hours at the temperature of the laboratory. Aftor being heated for several hours on the water-bath, the solution was quite in- active and yielded r-mandelic acid melting at 1 1 9 O .386 CHATTAWAY: ISOMERIC CHANQE OF Similarly, when I-mandelic acid was heated with a conbiderable excess of alcoholic potassium hydroxide, tho solution became quite in- active and yielded r-mandelic acid. The action of alkali on optically active esters and on the salts of optically active acids is being further examined. Summary. (1) When r-mandelic acid was heated with I-borneol, the unesterified acid was lavorotatory. The mixture of esters yielded a laevorotatory acid. (2) When E-bornyl dE-mandelate was submitted to fractional hydro- lysis, a laevorotatory acid was obtained from the initial hydrolysis, and an inactive acid from the final hydrolysis. (3) d-Mandelic acid was isolated from the dextrorotatory acid obtained from the initial hydrolysis of I-menthyl dZ-mandelate. (4) Z-Menshyl dE-mandelate yields either a dextrorotatory or a IEVO- rotatory acid from the initial fractional hydrolysis, according to the amount of alkali used. (5) Z-Menthyl dEmandelrtte is a partially racemic compound, which is not resolved into I-menthyl d-mandelate and I-menthyl Lmandelate when repeatedly crystallised from light petroleum at the temperature of the laboratory. (6) Potassium I-mandelate undergoes complete racemiaation when heated with a large excess of potassium hydroxide in aqueous or in alcoholic solution. The author desires to express his thanks to the Research Fund Committee of the Chemical Society for a grant, which defrayed the cost of this research. THE UNIVERSITY, BIRMINGHAM.

ISSN:0368-1645    

DOI:10.1039/CT9048500378

出版商:RSC    

年代:1904

数据来源: RSC

摘要:

386 CHATTAWAY: ISOMERIC CHANQE OF XLIII.-Isomeric Change of Uiacylanilides into Acyl- aminoke t ones. By FREDERICK DANIEL CHATTAWAY. A LARGE number of derivatives of aromatic amines in which aminic hydrogen is replaced by various atoms or groups readily undergo an intramolecular rearrangement, the substituting atom or group chang- ing place with a hydrogen atom of the nucleus. Such exchange invariably occurs between the substituting atom or group and aDIACYLANLLIDES INTO ACYLAMINOKETONES. 387 hydrogen attached to the nucleus in an ortho- or a para-position with respect to the amino-group, interchange between a group attached to the nikrogen and a hydrogen atom in either meta-position never having been observed. The transformations which have been most thoroughly studied are those in which alkyl, nitroxyl, or halogen radicles are con- cerned ; the ready substitution of aromatic amines by agents such as nitric acid, chlorine, or bromine is due to changes of this type which are brought about with great ease.Although the acyl groups very readily replace aminic hydrogen, these radicles have up to the present not been observed tomigrateinto the nucleus, although the diacylanilides are well known. Such intra- molecular changes, however, can be effected readily, and the reactions, in so far as they have been studied, follow a course exactly similar to that of other analogous transformations, thus : These intramolecular rearrangements take place a t somewhat high temperatures and under the influence of catalytic agents, hydrogen chloride being always effective, whilst in some cases zinc chlorids gives a better yield. Transference of more than one acyl group into the nucleus has, however, not yet been effected.Owing to the reactive character of the resulting ketones, the yields are never quantitative, and resinous or tarry by-products are always produced. When acetylation or benzoylation of an aromatic amine is effected by heating the base with the acid chloride, the product is in- variably coloured and small quantities of tarry matters may be formed. This is without doubt due to the occurrence, to a slight extent, of the isomeric change under consideration, and to the foramation of resinous products from the resulting ketone. The transformation of dibenzanilide into benzoylaminobenzophenone is the only one in which both the ortho- and the para-isomerides have been isolated in quantity, and it is this case which brings these changes into complete accord with other analogous intramolecular rearracge- ments.This transformation is the only simple method by which o-aminobenzophenone has yet been obtained. The change of diacetanilide into acetylaminoacetophenone and of dipropiooanilide into propionylaminopropiophenone where only the para-derivatives have been satisfactorily isolated have also been studied.388 CHATTAWAY : ISOMERIC CHANGE OF Tr a ns f o r m a t i o n of D iace tan i I id e i n t o A c e t y I-p-a m i n 0- a c e t o p h e n o n e, N( CO*CH,), /\ 1 1 \/ NH*CO*CH, /\ -+ I I \/ CO*CH, Acetanilide, when heated with an equivalent quantity of acetic anhydride or acetyl chloride for some time, yields diacetanilide, which can be obtained pure by distillation.If diacetanilide is mixed with about 10 per cent. of dry powdered zinc chloride and heated at 150-160° for several hours, or if at this temperature, it is treated with a slow current of dry hydrogen chloride, transformation takes place, and a considerable amount of the isomeric acetyl-paminoacetophenone is produced, There are clear indications of the simultaneous formation of the isomeric acetyl-o-aminoacetophenone, but the compound cannot be isolated with certainty, very probably because during the heating it becomes hydrolysed to some extent, and then, under the influence ol the zinc chloride, reacts with the p-amino- acet ophenone simultaneously formed to produce a-p-aminophen y 1- y-lepidine.This substance is always found as a product of the reaction and, as 0. Fischer has shown (Bsr., 1886, 19, 1038), it is formed almost quantitatively when a mixture of 0- and p-aminoacetophenones is heated together with zinc chloride. Some difficulty is experienced in isolating the different products, as much tarry or resinous matter is produced which is probably formed either by some complicated condensation of the aminoketones or from the unstable aminophenylacetylenes derived from these compounds by the action of the condensing agent (compare Baeyer, Ber., 1882, 15, 2174). The following process is found to give the best results : diacet- anilide is heated at about 140-150° for 8-10 hours and a very slow current of hydrogen chloride is passed through the melted mass ; a viscid, brown liquid is produced which solidifies on cooling, and from which acetyl-p-aminoacetophenone can be obtained by long-continued crystallisation, but as the process is very tedious the base is best extracted in the following manner.The product is completely dis- solved in excess of alcohol containing about one-third of its bulk of concentrated hydrochloric acid and hydrolysed by boiling for 6-8 hours; the resulting liquid is made slightly alkaline with caustic soda and distilled in a current of steam, which removes the aniline obtained from the untransformed diacetanilide, about 50 per cent. of the aniline used being thus recovered. The alkaline liquid in the distillingDIACPLANlLIDES INTO ACYLAMINOKETONES.389 flask is cooled and the tarry residue collected, the paminoacetophen- one being obtained from this by repeated extractions with boiling water. These solutions can be decolourired by heating with animal charcoal and concentrated, if necesrary, when the para-base, Khich separates on cooling, can be extracted with ether. If the stearn distillation is continued for some time after the aniline has passed over, a pale yellow distillate is obtained having a n intensely penetrating odour and reddening a pine splinter previously moistened with hydrochloric acid. This distillate undoubtedly contains a little of the ortho-base, and a yellow oil can be extracted from it by ether, but the amount produced is so extremely small that it cannot be identified with certainty.The yield OF p-aminoacetophenone is not good and seIdom exceeds 5 per cent, of the weight of diacetanilide used. With zinc chIoride as catalyst, the yield is better, being generally about 25 per cent. Diacetanilide is intimately mixed with about a n eighth of its weight of zinc chloride and heated at 150" for 12 hours. The product is poured into boiling water, washed several times to remove the zinc chloride, and then treated in the foregoing manner. This method is similar t o that employed by Klingel (Bey., 1885, 18, 2687) in preparing p-aminoacetophenone, except that diacetanilide is used in place of acetanilide and acetic anhydride in order to demon- strate the transformation, Klingel regarded the action as consisting of a direct introduction of the acetyl group into the ring without the formation of any intermediate product, There can, however, be no doubt that diacetanilide is formed i n his method, the p-aminoaceto- phenone being produced by the transformation of the diacetyl compound.A number of derivatives of p-aminoacetophenone have been pre- pared; these are formed by the ordinary processes, but in the preparation of the nitrogen-halogen derivatives great care must be taken t o avoid the liberation OF free halogen. The chloroamino- derivatives are prepared by suspending or dissolving the acylamino- ketone in chloroform and shaking it for some time with a well-cooled solution of potassium hypochlorite containing a n excess of potassium hydrogen carbonate. On separating the chloroform sulution, drying with calcium chloride, and evaporating off the solvent in a current of dry air, a yellow, viscid liquid is left, which, if pure, at once solidifies t o a colourless, crystalline mass, on cooling and stirring with a little light petroleum.The brornoamino-derivatives are made and treated in a similar way, using a solution of hypobromous acid to which a little freshly pre- cipitated mercuric oxide has been added to prevent any liberation of free bromine. VOL. LXXXV. D D390 CHATTAWAY : ISOMERIC CHANGE OF The compounds, unless otherwise stated, were crystallised from a Acetyl-p chloroaminoacetophenone, CH,* CO*C,H,*NCl*CO * CH,, crys- 0,2056 liberated I from 19.5 C.C. N/10 solution. C1 (as NC1) = 16.81. CloHl,O,NC1 requires C1 (as NCl) = 16.75 per cent.AcetyZ-p-bromoaminoacetophenone, CH,*CYO*C,H,*NBr*CO*CH,, crys- tallises from petroleum (b. p. 60-80°) in yellow, glistening, rhombic plates which melt at 83'. 0,2298 liberated I from 17.9 C.C. #/lo solution. Br (as NBr) = 31.14. C,,H,,O,NBr requires Br (as NBr) = 31-22 per cent. P~opionyZ-p-aminoacetophenone,CH3*CO*C6H4*NH~C0 *CH;CH,,crys- tallises from alcohol in glistening, transparent, colourless prisms and melts at 136'; its composition was established by estimating the chlorine in the corresponding chloroamino-derivative. mixture of chloroform and petroleum (b. p. 60--80°). tallises in clusters of glistening, pearly plates and melts a t 92O. P!ropio~~yE- p-chloroaminoc~cetophenone, CH8*CO*C6H,*&TCi*CO-CH2*CH3, 0.3020 liberated I from 26*6 C.C. N/10 solution.Cl(.as NCl) = 15.61. C11H,202NCl requires C1 (as NC1) = 15.71 per cent. Benxoyl-p-aminoacetophenone, CH,* CO* C,H,*NH* CO*C,H,, crystal- lises from alcohol or chloroform in colourless, pearly plates me1 ting at 205'. forms colourless, transparent plates which melt at 42'. Benxoyl-p-chloroccminoacetophenone, CH,*CO*C,H,*NC1*CO*C6H,. Benzoyl-p-aminoacetophenone is so sparingly soluble in cold chloro- form that it is better to add its alcoholic solution to potassium hypochlorite solution containing excess of hydrogen carbonate, the mixture being then shaken with a little chloroform until all the solid has disappeared. To remove the alcohol and to ensure the conversion of all the benzoyl derivative, the chloroform solution is repeatedly shaken with fresh hypochlorous acid.Benxoyl-p-chloroaminoacetophenone crys- tallises in clusters of colourless plates which melt a t 77'. 0-2511 liberated I from 18-1 C.C. #/lo solution. C1 (as NCl)= 12.77. Benxenesdphon-p-aminoacetophenone, CH,*CO*C,H,*NH*XO,*C,H,, crystallises from alcohol in clusters of, slender, colourless needles melting at 128'. Cl,H1202NCl requires C1 (as NC1) = 12.96 per cent. Benxertesulp~on- p-chloroaminoace~p~no~, CH,*CO* C6H,4NC1*S0, C6H5,DIACYLANILIDES INTO ACYLAMINOKETONES. 391 forms small, colourless, transparent plates, which are apparently flattened rhombs melting at 91'. 0.1804 liberated I from 11.6 C.C. N/lO solution. C1 (as NC1) = 11.39. C14H1203NClS requires C1 (as NCl) = 11.45 per cent. Toluene- p-sec~hon- p-ccminoacetophenone, CH3*CO*C6H,*NH*S0,*C6H;CH3, crystallises from alcohol in colourless, transparent, flattened prisms which melt at 203'.Toluene- p-sulphon- p-chloroaminoacetophenone, CH,. CO~C,H,*NC1*SO2*CBH;CH3, 0.2909 liberated I from 18.1 C.C. N/10 solution. C1 (as NCl)= 11.03. C,,Hl,O,NCIS requires C1 (as NC1) = 10.95 per cent. forms clusters of irregular, colourless plates melting at 93'. Tyana fovrnation of Dipropionani lide into Propiony I- p-aminopro- piophenone, N( CO*C2H,)2 NH* CO *C,H, /\ /\ I I \/ Under comparable conditions, dipropionanilide is transformed into the corresponding aminoketone exactly like diacetanilide, but only propionyl-p-aminopropiophenone could be isolated from the product in any quantity, although in this case also the ortho-compound is probably formed in small amount.When hydrogen chloride is used as a catalyst, the yield is small, but it is much increased by the employment of zinc chloride. The yield is better when the propionic acid formed is not separated from the dipropionanilide, about 50 per cent, of the theoretical amount being obtained in the following manner. Propionanilide (1 mol.) is heated at 140-150' for some hours with the equivalent amount (1 mol.) of propionic anhydride, then about a third of its weight of dry powdered zinc chloride is added and the heating continued for about twelve hours. The pale brown liquid produced, which solidifies on cooling, is washed with warm water to remove zinc chloride and hydrolysed by boiling for 8 hours with an excess of alcohol containing about one-sixth of its bulk of concentrated hydrochloric acid.A.fter distilling off the alcohol in a current of steam, the residue is made slightly alkaline with caustic soda and again distilled in steam t o remove any anilide obtained from unchanged dipropionanilide. On filtering the hot alkaline liquid from a very little tarry matter, the base separates from it on cooling in pale reddish-brown plates. A further D D 2392 CEATTAWAY : ISOMERIC CHANGE OF quantity can be extracted from the mother liquor by ether together with a very small amount of a pungent, oily substance which is probably the ortho-isomeride. 1D-~~iinopropiophenone as it first separates from the alkaline liquid has a somewhat reddish-broxn colour, which is, however, due to slight impurity and can be completely removed by recrystallisation.The substance is moderately soluble in alcohol o r chloroform, and sparingly 60 in ether; it crystallises from alcohol or a mixture of alcohol and ether in thick, colourless, transparent plates which are highly refractive, acd separates from chloroform in colourless, transparent, four-sided, flattened rhombs melting at 142'. It readily dissolves in hydrochloric acid forming a very soluble hydrochloride which crystal- lises in colourlees, flattened prisms. On mixing alcoholic solutions of p-aminopropiophenone hydro- chloride and platinic chloride, the platinichloride, 2(C,H,CO*C,H,-NH,),H2Pt CI,, crystallises out in small, flattened, bright orange-coloured prisms with domed ends, The double salt is sparingly soluble in hot water and in alcohol and blackens just above 200°, but does not melt.Pt = 27.19. C,sH,,02N2C16Pt requires P.4 = 27-52 per cent, 0.3813 yielded 0.1037 Pt. p-Aminopropiophenone easily dissolves in a hot dilute solution of sulphuric acid, and, on cooling, the sulphate, 2 (C,H,* CO*C,H,*NH,),H,SO,, crystallises out in glistening, colourless, pearly plates which are trans- parent when thin. On heating rapidly, it melts with decomposition a t about 223O, turning a bright red colour ; if heated slowly, it melts a few degrees lower with similar decomposition. SO4= 24.18. 0.4716 yielded 0.2771 BaSO,. ClsH2406N2S requires SO, = 24.24 per cent. Ace t y 1-p-aminopr opiopb enone, C,H,* COO C,H,*N H*CO* CH,, is easily prepared by adding the calculated quantity of acetic anhydride to the base, when considerable heat is developed and a crystalline mass separates.On completing the action by warming for a short time on the water-bath, freeing from acid by a dilute solution of potassium hydrogen carbonate, and repeatedly crystallising from alcohol, the com- pound is obtained in perfectly colourless, long, glistening, six-sided prisms terminated by pyramids which melt a t 175'. Kunckell (Ber., 1900, 33, 2641), who first prepared acetyl-p-amino- propiophenone by the action of propionyl chloride on acetanilide in tho presence of a large excess of aluminium chloride, states that it forms pale yellow needles rrielting a t 161°, and that the p-"minoin-opio-DTACYLANILIDES INTO XCYLAMINOKETONES. 393 phenone obtained from this formed long, yellow needles melting at 1409 The colour noted in both cases is without doubt due to a slight impurity.I n this method of preparation also, it is probable that the aminoketone is formed by the transformation of propionylacet- anilide under the influence of the aluminium chloride, and the very low melting point given by Kunckell for the acetyl derivative (14. below that obtained by the author) seems to support this view, as in the transformation of the mixed anilide i t is almost certain that the acetyl group as well as the prapionyl group would be transferred t o the ring, and t h a t the product, however carefully purified, would retain a Emall quantity of admixed propionyl-p-aminoacetophenone. From the base obtained in the above transformation, a number of derivatives have been made.The nitrogen chloride and bromide derivatives of the acylamino- propiophenones were all prepared by adding an alcobolic solution of the acylaminoketone t o a cooled solution of potassium hypochlorite containing hydrogen carbonate or to a solution of hypobromous acid containing suspended mercuric oxide. I n the case of the benzoyl derivatives, a large quantity of warm alcohol was required. Acet~Z-p-chloroaminopropiophenone,C2H5*CO* C,N,*NCl*CO-CH,, crys- stallises in colourless plates melting at 75". 0.2262 liberated I from 20 C.C. N/10 solution. C1 (as NCI) = 15.67. Acetyl-p-bromoaminopropiophenone, C,H,*C0.C,H,gNBr*CO*CH3, 0.2704 liberated I from 19.9 C.C. N/10 solution. Br (as NBr) = 29-42. CllH,,02NC1 requires C1 (as NCI) = 15.71 per cent. forms brilliant, short, yellow prisms which melt at 115'.Cl1H,,O2NBr requires Br (as NBr) = 29.60 per cent. PropionyEp-aminopropiophenone, C,H,*CO*C,H,-NH*CO*C, H,, crys- tallises from alcohol in slender, colourless prisms melting at 1 5 3 O ; it is easily soluble in alcohol and chloroform, sparingly so in ligroin. Prop'onyl -p-chloroaminopsopioplmone, C2H,*C0 * C,H,*NCl co c,H,, forms colourless, transparent plates and melts a t 80'. 0.1450 liberated I from 12.1 C.C. #/lo solution. C1 (as NC1) = 14.79. C,,H,,O,NCI requires C1 (as NC1) = 14.79 per cent. Pro~ionyZ-p-bi.omoaminopropio~~e~one, C2H,*CO*C,H,*NBr*CO*C2H5, separates in clusters of pale yellow, elongated plates which melt at 1 20°. Br (as NBr) = 27-96. 0.1258 liberated I from 8.8 C.C. N/10 solution. C,,H,,O,NBr requires Br (as NBr) = 28.14 per cent.Benxo yl-paminopropiophenmce, C2H,*CO*C,H,*NH*C0 C6H5, cry&394 CHATTAWAY: ISOMERIC CHANGE OF callises from alcohol, in which i t is sparingly soluble, in thin, colourless, elongated plates melting at 190'. Benxoyl-p-chloroaminolrropiopherro~, C,H,*CO*C,H,*NCZ*CO*C,H,, crystallises in clusters of colourless, transparent plates melting at 70". 0.1323 liberated I from 9.15 C.C. .iV/lO solution. C1 (as NC1) = 12.25. C,,H,,O,NCl requires C1 (as NC1) = 12.32 per cent. Benxoyl-p-bromoaminopropiophenone, C,H5~CO*C,~,oNBroCO*~6~0.1372 liborated I from 8.2 C.C. N/10 solution. Br (as NBr) = 23.S9. When heated quickly, all the nitrogen halogen derivatives of the acyl p-aminopropiophenones undergo transformation a t about ZOO-230°, this change being accompanied by much decomposition.Eenzenesdphon-p-aminopropiophenone, C,H,*CO~C,H,*NH*SO,*C,H,, separates in colourless, glistening plates melting at 165'. Be~zxeneszc~~o~~-p-chlorou~~~noprop~o~7~enon~, forms bright yellow, rhombic plates and melts a t 11 19 c,,~~,,o2NBr requires Br (as NBr) = 24.07 per cent. C2H5* CO*C6H4* NC1*S0,*C6H,, crystallises in colourless, transparent plates which melt at 81O. 0.1556 liberated T from 9.2 C.C. N/10 solution. C1 (as NC1) = 10.48. C,,H,,O,NClS requires C1 (as NC1) = 10.95 per cent. Transformation of DibenzccniEide i n t o Benxoyl-o- a no! -p-a m i n o b e n x o p hen o n e 8, NH*CO-C,H, NH*CO*C,H, "C6H5 h TT It was not thought necessary in this case to isolate the dibenzanilide, as this is known to be the first product of the action of benzoyl chloride on benzanilide (Compt rend., 1853, 37, 9 0 ; Ber., 1893, 26, 2853).The transformation and isolation of the products are best effected as follows. The equivalent quantity of benzoyl chloride (2 mols.) is added t o aniline (1 mol.) and the whole heated in an oil-bath ; the temperature is gradually raised as the interaction takes place, until it reaches 220-230°, at which point it is maintained for about 20 hours, A brown liquid is thus produced, which, on cooling, sets to a viscid, tarry mass, As it is difficult to isolate the isomeric benzoylaminoketones from this, it is best to hydrolyse the whole by dissolving it in a large excess of alcohol mixed with about one-third of its volume of con-DIACYLANILIDES INTO ACYLAMINOKETONES. 395 centrated hydrochloric acid.Complete hydrolysis is very difficult to effect, and 14 to 16 hours’ boiling is required. After the hydrolysis is complete, steam is blown rapidly through the liquid, when the alcohol first distils over and is followed by a considerable amount of ethyl benzoate. When this has all been removed, the acid liquid is rendered distinctly alkaline with sodiiim hydroxide, and steam is again passed in ; aniline is first expelled and then o-aminobenzophenone slowly distils over and either collects as a cryst.alline solid or separates in bright yellow needles from the cooled distillate. This product is perfectly pure after one crystallisation from alcohol. The pars-compound is not volatile in steam and remains behind in the distilling flask; a con- siderable amount crystallises out on filtering and cooling the hot liquid, and a further quantity can be obtained by repeatedly extracting the tarry residue with very dilute sulphuric acid, from which, on cooling, the base crystallises in a nearly pure state.It can be obtained pure by crystallisation from water containing a few drops of sulphuric acid, in which it dissolves only sparingly, or from a.lcoho1, in which it is readily soluble. The yields of 0- and p-aminobenzophenones are respectively about 15 per cent. and 45 per cent. of the weight of the aniline used. As in the analogous transformations of diacetanilide and dipropion- anilide, the presence of a little zinc chloride (compare Higgin, Trans., 1882, 41, 132) facilitates the transformation, and a slightly better yield of both isomerides is obtained.The process is carried out in the manner indicated above, about 10 per cent. G f finely powdered, dry zinc chloride being added. This process of intramole- cular rearrangement, which follows a precisely similar course t o the other well-known migrations of groups from the aminic nitrogen atom into the ortho- and para-positions of the ring, could in all probability be carried to the further stage, in which two acyl groups would be intro- duced into the nucleus in these positions. Although the use of a larger amount of benzoyl chloride (3 or 4 mols. to 1 mol. of aniline) leads to a somewhat diminished yield of the benzoyl-o- and -p-aminobenzophenones, which may be due to the occurrence of this additional transformation, yet no disubstituted product has been isolated.The isomeric change of dibenzanilide is of further interest since it affords the only simple method yet described of preparing o-amino- benzophenone, the process described being much less laborious than that employed by Ullmann and Bleier (Bey., 1903, 35, 4273), who converted toluene-p-sulphonanthranilic acid into its chloride, and treated benzene with this compound in the presence of aluminium chloride, finally liberating the base by hydrolysis with dilute sulphuric acid.396 CHATTAWAY : ISOMEfiIC CHANGE OF The following compounds have been prepared from the 0- and p-aminobenzophenones obtained in the foregoing transformation. The chloroamino- and bromoamino-derivatives of the acylamino- benzophenones were prepared respectively by adding an alcoholic solution of the acylaminoketone to a well-cooled solution of potassium hypochlorite containing an excess of potassium hydrogen carbonate, or t o a solution of hypobromoiis acid containing a little suspended mercuric oxide ; the products mere then extracted with chloroform and t h e solution treated as previously described.AcetyE-o-chZos.ounz.irzo benxophenone, C6H5* CO*C,H,* NCl COO CH,, separ- ates from a mixture of chloroform and petroleum i n brilliant, colour- less, flattened prisms with domed ends ; it melts at 102'. 0,2022 liberated I from 14.7 C.C. N/10 solution. C1 (as NCI) = 12.88. C,,H,,O,NCl requires C1 (as NCl) = 12.96 per cent. A cet yZ-0- bromouminobenxophenone, C,H5 'Coo C6H,*NBrm CO CH,, crys- tallises from a mixture of chloroform and petroleum in short, trans- parent, yellow prisms melting at 121'.0.1660 liberated I from 10.5 C.C. X / l O solution. Br (as NBr) = 25.28. C,,H,,O,NBr requires Br (as NBr) = 25.13 per cent. C6H,*C0 C6H,*2JC1 CO*C,H,, crystallises from light petroleum in clusters of colourless, flattened prisms which melt at 107". 0.1788 liberated I from 12.6 C.C. N/10 solution. C1 (as NC1) = 12.49. C,,H,,O,NCl requires Cl (as NC1) = 12.32 per cent. Propionyl-o-bromoaminobensophenone, C,,H,*CO*C,H,*NBr-CO.C,H,, crystallises from light petroleum in bright yellow, transparent plates melting at 90". 0.1375 liberated I from 8-15 C.C. N/10 solution. Br (as NBr) = 23.69. Propionyl-o-chloroaminoberzxop7enone, C,,H,,O,NBr requires Br (as NBr) = 24-07 per cent.BenzoyZ-o-chZoroamioi5enzophenone, C6H5* CO* C,H,*NCl*CO*C,H,, separates from light petroleum in colourless, glistening, rhombic plates which melt at 98". 0.1611 liberated I from 9.7 C.C. N/10 solution. C1 (as NC1) = 10.6'7. C,,K,,O,NC1 requires C1 (as NCl) = 10.56 per cent. AcetyZ-p-chZoroarnino6enxophefione, C6H5*CO=C,H,*NCl*CO*CH3, crys- tallises from petroleum (b. p. 60--80°) in groups of colourless, flattened, elongated plates melting at 124'. 0,2396 liberated I from 17.3 C.C. N,lO solution. C1 (as NCl) = 12.79. C,,H1,O,NC1 requires C1 (as NCl) = 12.96 per cent.DTACYLANILIDES INTO ACYLAMINOKETONES. 397 AcetyZ-p-bromoaminohenzophenone, C,H,9CO*C6H4*NBr*CO*CR,, is less soluble than i t s analogues in chloroform; it crystallises from a mixture of about 1 part of chloroform and 3 parts of petroleum (b.p. 6O-SO0) in groups of pale yellow, six-sided plates melting at 15 1'. 0.1 624 liberated I from 10.2 C.C. N/10 solution. Br (as NBr) = 25-11. C15Hl,0,NBr requires Br (as NBr) 2 2 5 . 1 3 per cent. Psopiony 2- p.nm$no benxophenone, c6w, - CO*C,H, NH co C,&, crys- tallises from alcohol in clusters of thin, colourlees, elongated plates which melt a t 139'. The composition of this and other acyl deriva- tives is found by the analysis of their chloroamino-derivatives. PropionyEp-chEoroarninobenzophenone, C,H,*CO*C6H,*NC1*CO*C2H5, separates from light petroleum in glistening, colourless, flattened prisms melting at 129'. 0.1730 liberated I horn 12 C.C. N/lO solution. C1 (as NC1) = 12.29.Propion yl-p-bromoaminobenzoph enone, C,H,-CO*C,H,*N Br *CO*C,H5, crystallises from light petroleum in pale yellow, transparent plates which melt at 123'. CI6H,,O2NC1 requires Cl (as NCl) = 12.32 per cent. 0.1828 liberated I from 11 C.C. N]10 solution. Br (as NBr) =24*06. C,,H,,O,NBr requires Br (as NBr) =24*08 per cent. Benxoyl-p-chloroaminobenxophenone, C,H,* CO*C,H,*NCl* CO-C,H,, separates from petroleum (b. p. 60--80°) in clusters of small, colour- less, rhombic plates melting at 107'. 0,2382 liberated I from 14 C.C. N/10 solution. C1 (as NCl) =10*42. C,oH,,02NCl requires C1 (as NCl) = 10.56 per cent. Benxoyz-p-bromoarninobenxophenone, C,H,*CO*C,H,.N.Br*CO*C,H,, crystallises from a mixture of chloroform and petroleum in clusters of bright yellow, transparent, rhombic plates and melts a t 9 3 O . 0.2288 liberated I from 12 C.C. X/10 solution. Br (as NBr) =20*97. C2,Hl,0,NBr requires Br (as NBr) = 21-03 per cent. Benxenasulphon-p.aminobenzophenone, C,H,*CO*C,H,*NH*SO,*C,H,, crystallises from alcohol in colourless, transparent plates and melts at 156'. Benxenesulp~on- p-chloroamino Benxophenone, CGH5* CO*C,H,*NCl* SO,*C,H,, separates from a mixture of chloroform and petroleum in coIourless, short, flattened prisms which melt a t 114'.398 GREEN AND PERKIN: 0.2969 liberated I from 16 C.C. N/10 solution. C1 (as NC1) =9*55. C,,H,,O,NClS requires C1 (as NC1) = 9.54 per cent. !To Zuens- p-szclp~on- p-aminobenxop~enone, C,H,*CO* C,H,*NH *SO,* C,H,-CH,, crystallises from alcohol in colourless, thin, flattened prisms melting at 184'. ~oluene-p-sulpl~on-p-cl~loi.oaminobenxo~T~enone, C,H,*CO*C,H,*NCt.SO,* C,H,*CH,, crystallises from a mixture of chloroform and petroleum in clusters of somewhat irregular, colourless plates melting a t 11 6'. 0.3264 liberated I from 16.9 C.C. N/10 solution. C1 (as NCl) =8*97, C20H,g0,NC1S requires C1 (as NC1) = 9.19 per cent, ST, BARTHOLOMEW'S HOSPITAL AND COLLEGE, E. C.

ISSN:0368-1645    

DOI:10.1039/CT9048500386

出版商:RSC    

年代:1904

数据来源: RSC

摘要:

398 GREEN AND PERKIN: XLI V. - The Constitution of Pheno lp h t h alein. By ARTHUR GEORGE GREEN and ARTHUR GEORGE PERKIN. THE well-known behaviour of phenolphthalein with alkalis-the pro- duction of a deeply-coloured salt from a colourless acid and the dis- appearance of the colour on the addition of an excess of alkali or of alcohol-is generally explained on the theory of electrolytic dissociation. I t is supposed that whilst phenolphthalein itself and its undissociated salts are colourless, its ions are red, and that the disappearance of the colour on adding an excess of alkali or alcohol denotes the suppression or retrogression of ionisation. This view, first put forward by Ostwald ( Wissenschafttl. Grundkugsn der Amulyt. Chemie, IIh Aufl., p. l l S ) , has been adopted by Herzig and Meyer (Herzig, Ber., 1895, 28, 3258; 1896, 29, 138; Monatsh., 1896, 17, 429; 1902, 23, 709; R.Meyer, Jahrhzcch d. C‘hem., 1899, 9, 404) and by 0. Fischer (Zeit. EZcrb. Text. Chern., 1902, 1, 281) ; whilst quite recently R. Meyer and 0. Spendler (Ber., 1903, 36, 2949), after a careful re-examination of the facts, have come to the conclusion that it still offers the best ex- planation, and that the quinonoid theory does not give a satisfactory solution. On the other hand, the quinonoid theory of colour has received its strongest experimental support from the work of Nietzki on the ethers of fluorescein, a compound itself closely allied to phenolphthalein, whilst the constantly increasing weight of evidence in general favourTHE CONSTITUTION OF PHENOLPHTHALEIN. 399 of this theory should lead us t o question its lack of application in this case.With the view of testing the validity of the ionic explanation, we have carried out the following experiments. If the ionisation of the phenolphthalein alkali salt is suppressed by the presence of an excess of free alkali, it appears to follow that the red colour should return if this excess be neutralised. We have found, however, that if an aqueous solution of phenolphthalein containing a n excess of caustic potash sufficient to decolorise it (or nearly so> is carefully cooled and gradually neutralised by a weak organic acid, such as acetic or car- bonic acid, the solution may be rendered quite neutral without any return of the red coloration. Nor does tho colour return on again rendering the solution alkaline.On the other hand, if the solution is acidified and heated or left for some time a t the ordinary temper- ature, then on again rendering alkaline, the colour returns in its full intensity. In order to ascertain the point at which neutrality occurs, and thereby gain some insight into the nature of the colourless compound present, a weighed quantity of phenolphthalein was dissolved in a measured volume of standard caustic potash sufficient to decolorise it all but a slight pink colour. The solution was cooled in ice and care- fully titrated with dilute acetic acid of known strength until neutral to litmus paper. This point exactly coincides with the disappearance of the residua1 pink colour. The experiment was repeated several times, and i t was always found that the quantity of acetic acid required corresponded with the presence in the neutral solution of a potassium salt, c' 2oH,,0,K. For example, 0.5 gram of Phenolphthalein was dissolved in 10 C.C.of caustic potash solution, of which 1 C.C. = 0.194 gram of KOH and titrated to neutrality with dilute acetic acid, of which 1 c.c.= 0.146 gram H4C202, namely, 14.3 C.C. corresponded with 10 C.C. of the caustic potash solution used. The volume of acetic acid required was 13-65 C.C. Percentage of KOH neutralised by phenolphthalein. Fo&d 17.6 Calculated for'C,,H,,O,K 17.6 The freshly prepared cold solution of the potassium salt was a clear, colourless liquid which soon became turbid through separation of free phenolphthalein.When rendered a1 kaline, it only gave a slight pink coloration, but if first allowed to remain until turbidity had set in a more or less intense red was produced. Further, if the colourless neutral solution was simply heated t o boiling, it assumed the deep red400 GREEN AND PERKIN: colour of alkaline phenolphthalein, whilst a colourless precipitate of free phenolphthalein separated out. At the same time, the solution is found to have become strongly alkaline to litmus paper. The red colour remains on cooling. All these facts seem to be quite inconsistent with the electrolytic hypothesis. On the other hand, they are capable of simple explanation by means of the quinonoid theory, according to which the phenomena must be regarded as hydration and dehydration changes accompanied by it transition from the benzenoid to the quinonoid type and vice versb.If we assign to free phenolphthalein the usual Iactonic constitution, we may assume that the first action of an alkali is to neutralise one of the phenolic hydroxyl groups. The phenolic salt thus formed would be unstable, and may be considered to undergo an immediate rearrange- ment by the transference of the metal from the phenolic to the carb- oxylic group. I n this way, the red quinonoid salt, would be formed without intermediate production of the hydrol. On the addition of a further quantity of an alkali, the elements of potass- ium (or sodium) hydroxide are taken up, giving rise t o the colourless compound If, now, the cooled alkaline solution be neutralised with acetic acid, the metal attached to the phenolic oxygen will be removed and the colourless solution will contain the potassium salt of the carbinol- csrboxylic acid,THE CONSTITUTION OF PHENOLPHTHALEIN, 401 This compound is unstable, and on boiling or when left at the or- dinary temperature it undergoes dehydration.This change seems to take place in two directions, partly by formation of the red quinonoid salt, and partly by formation of the colourless lactone. The latter reaction is accompanied by the liberation of potassium hydroxide, which causes the solution to become alkaline, It is, in fact, very probable that the dehydration occurs solely in the second sense, and t h a t the quinonoid compound is only formed from the intermediate lactone by the action of the free alkali simultaneously produced. We have made many attempts to isolate the colourless carbinol- carboxylic acid (or its salt) in a solid state, but owing to its instability and great solubility in water we have not succeeded in doing so.Nevertheless, there can scarcely be a doubt as to its presence in the neutral colourless solution. The view of the quinonoid structure of the red salts of phenol- phthalein does not, in our opinion, run counter to any of the observed facts. Thus the formation of colourless lactone derivatives by alkyla- tion (Herzig and Meyer) or benzoylation (Bistrzycki and Nenki), which have been brought forward as arguments against it, are readily explained if we assume the intermediate formation of the carbinol salt. The difference in the behaviour towards alkalis of phenolphthalein as compared with other triphenylmet hane derivatives is only one of degree, and is readily explicable as being due to the presence of the carboxyl group and its tendency to form a lactone with the carbinol hydroxyl group.All the acid colouring matters of the triphenyl- methane series (acid violets, acid greens, soluble blues, &c.) are similarly decolorised by a n excess of alkali, although with the individual colouring matter& there is a large variation in the time required to destroy the colour completely, some requiring only 2 or 3 minutes, whilst others need several days, Even with phenol- phthalein, an appreciable time is necessary, which in itself renders the ionisation explanation improbable.The decolorised solutions of the various dyes can be rendered quite neutral by careful addition of acetic acid in the cold without the colour returning, but as soon as the neutral point is passed the colour returns immediately. I n the case of erioglaucine, however, the colour does not a t once return on acidifying, but only when the solution is heated. This exceptional behaviour is very analogous to that of phenolphthalein, to which erioglaucine may be considered to be similarly constituted, since i t contains an SO,H group instead of a CO,H group in the ortho-position to the methane carbon atom. The onlydifficultyarising from the above view of the constitution of the coloured phenolphthalein salts lies in the behaviour of quinolphthalein, which, according to the researches of R.Meyer and others (R. and H,402 THE CONSTITUTION OF PHENOLPHTHALEIN. Meyer, Ber., 1895, 28, 2959; R. Meyer and L. Friedland, Ber., 1898, 31, 1739; R. Meyer and 0. Spendler, Ber., 1903, 36, 2949), has, in the free state, the constitution of a dioxyfluoran : This compound behaves in a very similar manner to phenolphthalein, although the possibility of a change toa para-quinonoid form is excluded. I n order to ascertain whether quinolphthalein exhibits in all respects a completely similar behaviour to phenolphthalein, we have carried out with it the same experiments as recorded above. Thus 0.5 gram of quinolphthalein was dissolved in 10 C.C. of caustic potash solution (1 C.C. =Om194 gram KOH). A violet-red solution was first formed, which quickly became nearly colourless, It was then immersed in a freezing mixture and carefully titrated t o neutrality with dilute acetic acid, of which 14.3 C.C.corresponded with 10 C.C. of the caustic potash used ; the volume of acetic acid thus required was 13.7 C.C. Percentage of KO€€ neutralised by quinolphthalein. Found 16.94 Calculated for C'20H1306K 16.86 The solution therefore contained the potassium salt of the carbinol- carboxylic acid. The cold neutral solution was quite colourless, but, on heating, the violet-red colour returned, whilst simultaneously the solution became alkaline. The behaviour is therefore exactly similar t o that of phenolphthalein. I n order to bring these facts into harmony, we should suggest for the coloured salts of quinolphthalein the ortho-quinonoid structure : \-/ OH The changes which take place with alkalis would then be represented in a similar manner to those suggested above for phenolphthalein.FREEZING POINT CURVES OF DYNAMIC ISOMERIDES. 403 Such a view is exactly analogous to that put forward by Kehrmann with regard to the constitution of the rosindones and their analogues of the oxazine series. THE CLOTHWORKERS’ RESEARCH LABORATORY, THE YORKSIIIRE COLLEGE, LEEDS.

ISSN:0368-1645    

DOI:10.1039/CT9048500398

出版商:RSC    

年代:1904

数据来源: RSC

摘要:

FREEZING POINT CURVES OF DYNAMIC ISOMERIDES. 403 XLV. -Freezing Point Curves of Dynamic Isomerides : -4 mmonium Thiocyanate and Thiocadmmide. By ALEXANDER FINDLAY. NOTWITHSTANDING the importance of many of the substances, corn- paratively few investigations of the mutual relationships of dynamic isomerides have been carried out from the point of view of the phase rule, and the following short study is therefore offered as a con- tribution t o our knowledge of these (compare Carveth, J. Physical Chem., 1898, 2, 159 ; 1899, 3, 437; Soch, ibid., 1898, 2, 364; Cameron, ibid., 1898, 2, 409 ; Hollmann, Zeit. phy8ikal. Chem., 1903, 43, 129). In the case of all pairs of dynamic isomerides, the different solid forms correspond with a single definite constitution, but in the liquid state, a condition of dynamic equilibrium is established between the two modifications. When, therefore, the fusion and solidification of such substances are studied, the phenomena which are observed will vary according as reversible isomeric transformation takes place with measurable velocity at temperatures in the neighbourhood of the melting points, or only at some higher temperature. The relative velocity of transformation, as compared with the time required to carry out the determination of the melting or freezing point, will also have an influence on the character of the system.If isomeric trans- formation is relatively rapid, the system will behave like a one-corn- ponent system, but if the velocity of change is comparatively slow, the behaviour will be that of a two-component system (see Bancroft, J.Physical Chem., 1898,2,143 ; Roozeboom, Zeit. physikal. Chem., 1899, 28, 494). Since it is probable that in all cases the velocity of change can be accelerated by means of catalytic agents, it may be possible to cause a two-component system to behave like a system of one com- ponent. This has been effected, for example, in the case of acetaldehyde and paraldehyde (Hollmann, Zeit. physikal. Chem,, 1903, 43, 129) by means of sulphuric acid. If no isomeric transformation takes place in the neighbourhood of404 FINDLAY: FREEZING POINT CURVES OF the freezing points, the freezing-point curve obtained will be of the ordinary form exhibited by non-isomeric substances. If, however, isomeric transformation takes place with appreciable velocity, the extent to which these curves will be realised wili depend on the velocity with which the change occurs as compared with the rapidity with which the determination of the fleezing point can be effected.The general behaviour may be explained with the help of Fig. 1. If one of the modifications is heated, a temperature will be reached at which it mill melt, but this melting poirrt (represented by A ) mill be sharp only if the velocity of isomeric transformation is relatively slow, If the substance is maintained in the fused condition for some time, a certain amount of the isomeric form will be produced, and on lowering the temperature the pure A-form will be deposited, not at the temperature of its melting point, but at some lower temperature, depending on the concentration of the B-form in the liquid phase.If isomeric transformation takes place slowly in comparison with the rate a t which deposition of the solid occurs, the liquid will become increasingly rich in the B-modification, and the freezing point will therefore sink continuously until the eutectic point is reached. A similar behaviour mill be found if we start with the second isomeric form. The curve DE represents the variation of the equilibrium in the liquid phase with the temperature. This curve may be perpendicular, or may slope either t o the right or left, its direction depending on whether or not a heat effect accompanies transformation of the one isomeride into the other (Le Chatelier’s Theorem). The point D at which this equilibrium curve cuts the freezing-point curve is called the 6‘ natural freezing point.” This is the only point at which a solid phase (in this case the A-modification) can exist in stable equilibrium with the fused mixture (solution), and if isomeric transformation takes place with sufficient rapidity, this is the only freezing point which will be realisable, no matter from which form we start.Evidently, therefore, the degree of completeness with which the freezing point cilrve can be obtained will depend on the rapidity with which the determinations can be carried out. Finally, since the natural freezing point is the only temperature a t which the system solid-liquid is in stable equilibrium, a little consid- eration will show that on starting with the pure B-form and heating i t for varying periods of time so as to cause the formation of the A-form, the freezing point which is obtained will first fall until the eutectio point is reached, and then rise until the natural freezing point is attained.The recent work of Emerson Reynolds and Werner (Trans., 1903, 83, 1) on the equilibrium between ammonium thiocganste and thiocarb-DYNAMIC ISOMERIDES. 405 amide showed that the velocity of isomeric transformation is compara- tively slow at temperatures below about 140'. As these investigators did not study the freezing points of nmmonium thiocyanate and thio- carbamide, and as they informed me of their intention of not proceed- ing further with the investigation, the matter seemed t o be of suffi- cient interest and importance to warrant a short study of these isomerides on the lines just laid down, and I now lay the results OF this research before the Society.EXPERIMENTAL. The ammonium thiocganste and the thiocarbarnide employed were obtained from Kahlbaum, and were recrystallised from water, the thiocarbarnide being well washed with absolute alcohol. The materials were carefully dried, and the attraction of moisture by the ammonium thiocyanate during the preparation of the mixtures was guarded against by carrying out the operations in a large hot-air chamber. For the determination of the freezing points, about 20 grams of the mixture of thiocarbamide and ammonium thiocyanate were placed in a test-tuhe fitted with a cork, through which passed a glass stirrer and a thermometer graduated in tenths of a degree.The tube WBS heated by means of a glycerin bath, also furnished with a stirrer. After the complete liquefaction of the mixture had been effected, the temperature of the bath war allowed t o fall slowly, and the determin- ation of the freezing point carried out in the ordinary manner. This determinstion was made several times with the same quantity of substance, the temperature of the bath being maintained at not more than about lo bolow the freezing point of the mixture. If the mix- ture was stirred, surfusion was notfound to occur to any great extent (not more than about 0.5'). I n the first set of experiments which were carried out, varying quantities of ammonium thiocyanate and thiocarbamide were intimately mixed, and a portion of the mixture was used for the freez- ing point determination, whilst the other portion was analysed.The method of analysis WAS that used by Emerson Reynolds and Werner, I n these cases, it was necessary to make sure that incorrect values of the freezing point were not obtained, owing t o change occur- ring in the composition of the mixture through isomeric transforma- tion, A series of freezing point determinations were therefore made with each mixture extending over a period of 30-40 minutes, the times at which the different freezing points were determined being noted. I t was found, however, that at temperatures below about 130°, the change of tbe freezing point was very small. I n later experiments, however, the composition of the mixture was VOL. LXXXV.E E406 FINDLAY: FREEZING POINT CURVES OF ascertained from samples withdrawn in small tubes from the fused mass, after the freezing point had been determined. I n this way, greater certainty was obtained that the composition thus determined corresponded with the part,icular freezing point. More especially was this method of procedure necessary with mixtiires containing as much as 50 per cent. of thiocarbamide, for, in this case, very appreciable alterat.ion of the composition took place at the temperature of the freezing point. By noting the time a t which the last freezing point was determined and the time a t which the sample was withdrawn, a fairly accurate estimate of the freezing point could be made corre- sponding with the ascertained composition. The values of the freezing point obtained are given in the following table, and represented graphically in Fig.1. Freezing Poiizt of illixtwes qf Ammonium Thiocyurzate and l'hiocarbamide. Percentage of thiocarbarnide. 9.1 12.s 16.6 17-4 22.3 25.9 Percentage of Freezing point. thiocarbamide. Freezing point. 135.5' 30-4 106.2" 131.0 3 2 9 4 105.9 127.4 35.7 110.5 124.0 409 117.5 11 8.2 51.3 131.3 113.4 Apart from the numbers given in the table, a series of other deter- minations were made a t different times, which also confirm the course of the curve. Although the values of the freezing point are given in tenths of a degree, no claim is made that these figures are accurate t o less than about 0*5", and in the case of the higher temperatures the error may be greater. As is evident from Fig.1 (p. 407), the freezing point curve of ammon- ium thiocyanate and thiocarbamide is of the simplest form, indicating the non-formation of a compound between the two components. From all solutions of the composition represented by AC, ammonium thio- cyanate separates out in the form of crystalline scales. From solu- tions of the composition represented by BC, thiocarbamide separates in fine, crystalline needles, which are quite different in appearance from those of ammonium thiocyanate. By allowing the mixture con- taining 30 per cent. of thiocarbamide to cool, the eutectic point was found to be 104.5'. This was confirmed in a later experiment.DYNAMIC ISOMERIDES, 407 Melting Point of Ammonium Thiocyanals and of Tlhcccrbccmide. Even at the melting points of these two substances, isomeric trans- formation can take place with quite appreciable velocity in the case of ammonium thiocyanate, and with very considerable velocity in the case of thiocarbamide.The determination of the melting point, therefore, canuot be carried out with any great degree of accuracy, since the temperature found will depend on the rapidity with which the operation is carried out. By carrying out the determination rapidly., the melting point of ammonium thiocyanate was found to be FIG. 1. ,,4 Percentage thiocarbamide. 1 4 9 O , as also found by other investigators. With regard to thio- carbamide, I believe that the true melting point, if it could be ob- tained, would be found to lie much higher than 1 5 9 O , the value given by Emerson Reynolds and Werner, and higher even than 170°, the value given by Pratorius-Seidler (J.pr. Chem., 1880, [ii], 21, 141). This conclusion has been arrived a t from determinations of the melting point carried out rapidly in the ordinary apparatus, which yielded values between 172' and 178'. To lessen as far as possible the inaccuracy introduced by too rapid heating, the melting points of the thiocarbamide and of camphor were taken a t the same timein the same bath. A determination of the melting point of camphor carried out more slowlyin the ordinary The camphor was found to melt first. E E 2408 FINDLAY: FREEZING POINT CURVES OF manner gave a value 177". The melting point of thiocarbamide is, therefore, probably higher than this. This conclusion is also supported by the position of the eutectic point; for it is well known t h a t t h e more nearly coincident are the melting pointsof two substances, the closer d:>es t h e eutectic point, a s a rule, correspond with the 50 per cent.mixture. Although this rule is only approximately true, one would nevertheless expect that if t h e melting points of ammonium thiocyanate and thiocarbamide were 149' and 159O respectively, the eutectic point would correspond with 50 rather than with 30 per cent. of thiocarbamide. Nutural lireezing Point. The natural freezing point is the point at which the curve of equi- librium in the liquid phase cuts the freezing-point curve. This point could be readily determined if transformation took place with sufficient velocity, since the natural freezing point would then be the temperature at which all mixtures of the two isomerides would freeze, But in the present case, transformation does not occur sufficiently rapidly.If, however, the equilibrium composition is determined at two temperatures, the natural freezing point can be obtained with sufficient accuracy by joining the equilibrium points so determined and prolonging the line downwards to meet the freezing-point curve. I n the case of ammonium thiocyanate and thiocarbamide, t h e equi- librium composition in the fused state has been found both by Waddell (J. Phyicak Chem., 1898, 2, 525) and by Emerson Neynolds and Werner (Zoc. cit.) t o be independent of the temperature, although the value of the equilibrium found t)y the different investigators wus not the same. From the fact that the latter investigators carried out their determination with larger quantities of su hstance, preference may be given to their value. Moreover, a determiuation which I also made gave a value agreeing with their result.Assuming, then, t h a t the equilibrium in the liquid phase at the temperature 170-180° occurs when the mixture contains 25 per cent. of thiocarbamide, and also that the equilibrium is not altered by change of temperature, i t is evident that the natural freezing point is about 11 4-1 1 5 O , the stable solid form at this temperature being ammonium thiocyanate. This fact is of interest on account of the difference in behaviour as compared with polymorphous forms of a substance. I n the latter case, the more stable form in the neighbour- hood of the melting point is that with the higher melting point.Here, however, it is the more fusible form. A similar behaviour has been found in the case of the isomeric benzaldoximes (Cameron, J. PIzysicccl Chem,, 1898, 2, 409), and, wherever found, it is a sure signDYNAMIC ISOMERIDES, 409 that isomeric or polymeric, and not polymorphous forms of a substance are being studied. Finally, from the fact that the equilibrium in the liquid phase is independent of the temperature, the conclusion can be drawn that, in accordance with Le Chatelier's theorem, reciprocal transformation of ammonium thiocyanate and thiocarbamide is not accompanied by any heat effect. This is apparently not unusual in the case of dynamic isomerides, and has been met with in the case of the isomeric benzaldoximes (Cameron, Zoc.c i t . ) and the isomeric anisaldoximes (Carveth, J. Pl~ysiccd Chew.. 1899, 3, 437). Passuge from the Curve BC to AC. As has already been pointed out, i t should be possible in the case of dynamic isomerides by starting with the leas stable form to obtain, on heating, first a fall and then a rise of the freezing point. Hitherto, this had been experimentally verified only in the case of the isomeric bend-o-carboxylic acids (Soch, J. Pfhysical Cherrh., 1898, 2, 364). It is, however, also found in the case of ammonium thiocyanate and thiocarbamide. Thiocarbamide was fused and the liquid maintained at a tempera- ture of 160-1'i0° for a certain time, and the freezing point then determined.The temperature was again raised and the freezing point again determined. I n this vay, the following freezing points were obtained: 1OSo, 103", 1 0 4 O , 105.5O, 106.5". i n the case of the first two determinations, the solid sepaiated out in needle-shaped crystals like thiocarbamide, but in the case of the others, apparently in scales, like ammonium thiocyanate. Cooling Curve. From the form of the cooling curve of a mixture, valuable con- clusions can be drawn with regard t o the changes which take place in the system. Indeed, as has recently been emphasised by Tammann (Zeit. anorg. Chem., 2903, 37, 303), it is possible, from a study of the cooling curve and from the (' halts " which such a curve exhibits, to get a complete insight into the mutual relations of mixtures of two substances, and to determine the composition of any compounds formed, without the necessity of making a single analysis.I have therefore supplemented the determinations of the freezing points by ascertaining the form of the cooling curve, starting with a mixture containing about 22 per cent. of thiocarbamide. The curve obtained is shown in Fig. 2 (p. 410). The mixture was first of all fused completely, the temperature thereby rising t o about 126'. The tempera-410 FINDLAY: FREEZING POINT CURVES OE' ture was then allowed to fall slowly, and the rate of cooling noted. So long as only liquid was present, the cooling curve remained continuous as shown by the curve AB. The portion BC, which ascends slightly, indicates the separation of solid at the freezing point.This is then followed by a rapid fall owing to the temperature of the bath having meanwhile sunk considerably below the tempera. ture of the mixture, whereby rapid separation of the solid phase ie caused and a consequent fall in the temperature owing to the increase in the concentration of thiocarbamide in the liquid phase. Cooling then goes on regularly until the eiitectic point is reached a t which the temperature remained constant ( 104*3°) until complete solidification had occurred. This curve, therefore, agrees with the freezing point FIG. 2. I D Time. curve in giving no indication of the formation of a compound (NH,*CNS)3,CS(NH,),, since the appearance OF such a compound would be indicated by a I' break " in the cooling curve between C and D.Anaclgsis of the Solid Phase. For the purpose of confirming the conclusions which can be drawn from the freezing-point curve as to the non-formation of a compound between ammonium thiocyanate and thiocarbamide, a mixture contain- ing about 27-28 per cent. of the latter isomeride was prepared, com- pletely fused, and allowed t o cool down until solid began to separate out. The bath temperature was kept a t 107-10S0, and the solid which separated was filtered off through platinum gauze. On analysis, only about 10 per cent. of thiocarbamide was found. The presence ofbPNAMlC ISOMERIDES. 411 this fairly large amount of thiocarbamide is t o be accounted for by the difficulty of filtration with the apparatus employed. It is evident, however, that the solid which separated from the fused mixture could not be the compound (NH,*CNS),,CS(NH2)2. Stability Limits.The term I' stability limit " was introduced by Knorr (Annalen, 1896, 293, 88) to indicabe that temperature above which liquefaction and isomeric change take place. As employed by Knorr and others, however, the term does not have a precise meaning, since it is used t o denote not the temperature at which these changes cnn occur, but the temperature at which the change is appreciable, and the introduction of an indefinite velocity of change renders the temperature of the stability limit also indefinite. This indefiniteness has also been recognised by Knorr (ibid., 1899, 306, 334). Bince, however, the stable modification can always undergo isomeric change and liquefy at temperatures above the natural freezing-point, but not below it, and, moreover, since the less stable modification can undergo isomeric transformation and liquefy at temperatures above the eutectic point, but will not liquefy below it, it would appear better, as the author has suggested, to identify the natural freezing point and the eutectic point with the stability limits of the stable and unstable modifications respectively.The opinion expressed by Knorr (Zoc. cit.) that '' theoretically " the stability limits of most desmotropic forms, with the exception of enantiomorphous substances, will probably coin- cide with the melting points, does not appear to me to be justified, since, as stated above, liquefaction and transformation can '' theoretically " take place at temperatures below the melting points of either isomeride. Sumrnai.y. 1. The freezing-point curve of ammonium thiocyanate (m. p. 149') and thiocarbamide (m. p. above 177") is of the simplest form, consist- ing of two branches meeting a t a eutectic point (104.3'). 2. The natural freezing point is 114-115", the stable solid form being ammonium thiocyanate. 3. Neither the freezing-point curve, nor the cooling curve, nor the analysis of the solid phase gives any indication of the formation of a compound of ammonium thiocyanate and thiocarbamide stable at temperatures in the neighbourhood of the freezing-point curve. 4. The transformation of ammonium thiocyanate into thiocarb- amide and vice versd does not appear t o be accompanied by any heat effect.412 HARVEY: A NOTE ON 5. The I‘ stability limits ” of dynamic isomerides (including desmo- tropic forms) are defined as the natural freezing point and the eutectic point for the stable and unstable forms respectively. CHEMICAL DEPARTMEXT, U X I V E ILS I ’I’ Y OF 13 I 1’. 11 I NGH A M.

ISSN:0368-1645    

DOI:10.1039/CT9048500403

出版商:RSC    

年代:1904

数据来源: RSC

摘要:

412 HARVEY: A NOTE ON XLVI. -A Note 07% Pheny 1 dirnet h y 1 a1 1 y larnrnonium Compounds. By ALFRED WILLIAM HARVEY. The work which forms the subject of this note was commenced some eight months ago, but its completion was unavoidably delayed, and in t h e meanwhile a paper was published on the subject by H. 0. Jones (Trans., 1903, 83, 1400) which contained results differing in some respects from those obtained by the author. I n this investigation, an attempt has been made to prepare and resolve into their optically active components, substituted quaternary compounds containing two similar groups. Phenyldimethylallylammonium iodide was prepared by mixing molecular proportions of dimethylaniline (48 grams) and ally1 iodide (67 grams). After a few days, the semi-solid product was finely ground and filtered, 67 grams of solid phenyldimethylallylammonium iodide being obtained.The impure product, which rapidly deliquesced on exposure to air, was further purified by washing with acetone, the final purification being effected by solution in absolute alcohol and precipitation by the cautious addition of ether. Phenyldimethyl- allylammonium iodide, when pure, melts at 86--S8O, and when crystallised from acetone forms large, pink tables, which were never obtained colourless ; the pure substance was not deliqiiescent. Phen y 1 dime tb y I ally lammonium d-cam phorsul phona te was prepared by mixing molecular proportions of the iodide and silver d-camphor- sulphonate ; both constituents were finely powdered before mixing, and the mixture was boiled for one hour in a mixture of equal parts of ethyl acetate and acetone.The precipitated silver iodide was filtered off and the salt, which crystallised very readily, was obtained in six fractions by successive crystallisations. No alcohol was used in the preparation and the product was never gummy, The salt crystallised in fine, colourless needles melting at 1 5 6 O . The rotatory power, for sodium light, of each of these fractions, in a two per cent'. solution, was found t o be the same, and the calculated [MI, was identical, within t h e limits of experimental error, with thatPHENYLDIMETHYLALLYLAMMONlUM COMPOUNDS. 413 of the activity due to the active acid ion, namely, [M.], + 51.7'. It was repeatedly observed that an alteration in the concentration of the solution caused a different value for the specific rotation to be obtained at the same temperature, thus, a 1 per cent).solution gave [ a ] D + 1 lago, whilst a 2 per cent. solution had [a]= + 13.3'. As it was possible that a separation of the dextro- and IZVO- isomerides might be effected by crystallisation a t a higher temperature, search was made for a suitable solvent having a high boiling point. Pgridine was chosen, as phenyldimethylallylammonium d-aarnphor- sulphonate is fairly soluble in this solvent in the cold and readily SO a t the boiling point. A crop of crystals was thus obtained a t about the temperature of boiling water. This fraction was not, however, found to exhibit any optical activity due to the basic ion. As this substance was found t o be very sparingly soluble in ethyl acetate, and as Jones described it as readily soluble in ethyl acetate, i t s solubility in this solvent was determined and found to be less than 8 parts in 5000 parts of a boiling solution of the salt in ethyl acetate.Four separations of the d-camphorsulphonute were ob- tained by the successive addition of quantities of ethyl acetate to a cold solution of the salt in pyridine, and these four fractions also showed no activity due t o the basic ion. 0.2006 gave 0,4697 CO, and 0,1423 H,O. N(C,H,)(CH,),*C,H,*C,,H,,*O*SO, requires C = 64.1 ; H = 7.8 per cent. Phenyldimethylallyla.mmonium platinichloride was obtained as a pale yellow crystalline precipitate on adding the requisite quantity of platinic chloride to an aqueous solution of the final fraction from ethyl acetate and acetone of phenyldimethylallylammon~urn d-camphor- sulphonate, acidified with hydrogen chloride.The precipitated salt, which was collected and thoroughly washed with water, when dried in the air, melted at 162-165' with decomposition. C = 63.8 ; H = 7.8. 0.6130 gave 0.1439 Pt. Pt = 26.9. [N(C,H,)(CH,),*C,H,],PtCI, requires Pt = 86.6 per cent. A cold aqueous solution of the platinichloride, which is slightly soluble in water, was examined in the polarimeter, but was found t o be quite inactive. A further examination was made by suspending about half a gram of the salt in water, slightly acidified with hydrogen chloride ; the platinichloride was then decomposed by passing a current of hydrogen sulphide through the cold solution ; the precipitated platinum sulphide was filtered off and the solution examined in the polarimeter. This solution, also, was found to be inactive. As Jones has succeeded in obtaining d- and I-isomerides of phenyl- benzylmet h yle t hylammonium salts having optical rotations of similar414 BURGESS AND PAGE: A NOTE ON THE COMPOSITION OF arithmetical value (Trans., 1904, 85, 223), the author proposes to revise the work of Pope, Peachey, and himself, who failed to find such a close agreement in the case of the d- and Z-phenylbenzylmethyl- allylammonium salts. The author’s thankg are due to the Research Fund Committee of t h e Chemical Society for a grant which defrayed the expenses of this research. CHEMICAL DEPARTMENT, THE GOLDSMITHS’ INSTITUTE, NEW CROSS, S.E.

ISSN:0368-1645    

DOI:10.1039/CT9048500412

出版商:RSC    

年代:1904

数据来源: RSC

摘要:

414 BURGESS AND PAGE: A NOTE ON THE COMPOSITION OF XLVI1.--A Note on the Composition of Distilled. Oil o,f Limes and a New Sesquiteypene. By HERBERT EDWARD BURGESS and THEODORE HENRY PAGE. An investigation of terpeneless or concentrated oil of distilled limes * showed that this product contained about 40 per cent. of free alcohols, The oil was separated by distillation in VQCUO into two fractions. (1) The lower fraction, which boiled a t about 100-105"/17 mm. and had a sp. gr. 0.930 at 15", was viscous and retained the distinctive odour of the oil. By repeated fractionation, it was obtained in quantities approximating to those found for the alcohols by direct estimation, and had the following constants: sp. gr. 0.934 a t 1 5 O , a-19'30' (100 mm. tube), and nD 1.4806 at 20'.When cooled in a freezing mixture and sown with a crystal of terpineol, this fraction became nearly solid. After separation and crystallisation from dilute alcohol, the product melted at 35'. I n a supercooled state, the substance had the following constants : sp. gr. 0.941 a t 15', [.ID - 20°, and n, 1.4829 a t 20'; it boiled a t 214'1762 mm. The terpineol was further characterised by the preparation of certain of its derivatives. Several of these were found to melt a little low, but this is stated to be the case for the derivatives of the active terpineols as compared with those of the inactive variety. The nitrosochloride was obtained in practically quantitative yield by adding hydrochloric acid diluted with glacial acetic acid to a well- cooled solution of the terpineol and amyl nitrite in glacial acetic acid, the product being precipitated with ice-water.After recrystallisation from ethyl acetate, it melted at 105-106'. * The oil obtained after distilling off the terpcnes.DISTILLED OIL OF LIMES AND A NEW SESQUITERPENE. 415 The nitroleanilide,'prepared from the preceding compound by warming with an alcoholic solution of aniline, crystallised from alcohol and melted at 150-151'. The phenylurethane slowly separated from a mixture of equal volumes of terpineol and phenylcarbimide. After crystallisation from alcohol it melted a t 111'. The terpineol was successively oxidised with 2.5 per cent. potassium permanganate solution and chromic acid mixture (sp. gr. 1-25), the latter reagent being carefully added to the first oxidation product dis- solved in a little water.The oil, which separated, solidified after some days, and when recrystallised from water melted at 62". This product corresponds with the ketolactone from terpineol (m. p. 35O), and was further characterised by conversion into the oxime (m. p 77'). These results indicate t h a t lsvorotatory terpineol (m. p. 35") forms a considerable proportion of the oxygenated compounds in distilled oil of limes. Although the peculiar odour of distilled oil of limes is attached to the terpineol fraction, the terpineol isolated is either practically odourless or has a faint odour of lilac. The aroma is probably due t o an isomeric liquid terpineol having a slightly lower boiling point, but a t present this supposition is not supported by conclusive evidence.(2) The second fraction boiled a t 130-140°/17 mm., and had a lower sp. gr. than the first, the value of this constant being under 0.900. This fraction was found t o give only a slight action with sodium, and hence consisted almost entirely of hydrocarbons, which from their boiling point seemed to be sesquiterpenes. It was diluted with two volumes of dry ether, saturated with hydrogen chloride, and left for several days. On evaporating off the diluent, crystals separated which were collected, and, after crystallisation from ethyl acetate, melted constantly a t 79-80'. This product forms colourless crystals readily soluble in ethyl acetate, acetone, or ether, but less so in alcohol, acetic acid, or chloroform; it is characterised by a remarkable power of crystallisation, and is never obtained in an oily state, although its melting point is comparatively low, The hydrochloride was mixed with 18 parts by weight of sodium acetate and 6 parts of acetic acid, and heated nearly t o the boiling point for an hour.The crystals first dissolved, but after a time sodium chloride separated. The cooled mixture was then poured into water, mashed, and distilled in steam. The dried distillate was fractionated, the last time over sodium, and obtained as a colouxless oil with a slight, but peculiar odour; it resinified with great readiness, even when left in a partially filled flask. It boiled at 131' under 9 mm. and at 262-263' (uncorr.) under 756 mm. pressure, i n the latter case with slight decomposition ; the following constants were416 PERKIN : 8-KETOHEXAHYDROBENZOIC ACID.also determined : sp. gr. 0.873 a t 15', [a],, +, O', nD 1.4935 at 1 5 O , an(\ 1.4910 at 19.5' ; mol. refract. 68.2, whereas CI,H,, (with three double linkings) requires 67-76, 0,2768 gave 0.8970 CO, and 0.2986 H,O. In dilute solution in acetic acid, the hydrocarbon combined with C = 88.38 ; H = 11.99. C,,H,, requires C = 88.24 ; H = 11 *76 per cent;. approximately 6 atoms of bromine. The hydrochloride gave C1= 34.2. C,,HZ4,3HC1 requires 34.0 per cent. This was the only derivative successfully prepared. The nitroso- chloride, nitrite, bromide, and acetate were unsatisfactory, From the above data, it seems apparent that distilled oil of lime3 con- tains a sesquiterpene, not hitherto described, and differing materially from the other members of the series, especially in specific gravity, which is much lower than the usual value for sesquiterpenes, but ap- proximates t o those found for olefinic or partially oletinic sesquiterpenes. The name " Zirnene " is suggested for this sesquiterpene, as having first been found in oil of limes. Limene is also present in hand-pressed lime oil and in lemon oil i n the fraction boiling between 120' and 140' under 9 mm. pres- sure. The characteristic hydrochloride (m. p. 79-80') was readily obtained by saturating this fraction with hydrogen chloride. Other oils of this series are being examined with the view of ascertaining whether limene is present or not.

ISSN:0368-1645    

DOI:10.1039/CT9048500414

出版商:RSC    

年代:1904

数据来源: RSC

摘要:

416 PERKIN : 8-KETOHEXAHYDROBENZOIC ACID. XLVlII.--6-Ketohexahydrobenxoic Acid. By WILLIAM HENRY PERKIN, jun. OF the three possible ($I-, y-, and &)-ketohexahydrobenzoic (cyclohex- anonecarboxylic) acidp, I CO,H CH" I C0,H co bO,H only the first two are known. The ester of the P-acid was synthesisedPERKIN : ~-KETOHEXABYDROBESZOIC ACID. 417 by Dieckmann (Bey., 1894, 27, 103; 1900, 33, 2683) from ethyl pimelate by the action of sodium : CH, *CH,*C02E t CH,-CH,>co ,<C H,-CH,*CO2E t --+- H2<CH2---CH*C0,Et y-Ketohexahydrobenzoic acid was obtained by Baeyer and Tut ein (Ber., 1889, 22, 2182) by first reducing hydroxyterephthalic acid to ketohexahydroterephthalic acid and then eliminating carbon dioxide by boiling with water : C;O,H C y32H CH / \ / \ / \- 7% yo ?I€ G-OH --3 yH2 yo \ / \ / \/ YH --f CH, CH, CH CH CH, CH, C0,H FH CO,H 7 CO,H Since the unknown 6-keto-acid might be expected to exhibit un- usually interesting properties, the author has been engaged for some time in devising a method for its synthesis, and has ultimately succeeded in preparing it in considerable quantities by the following series of reactions.When ethyl P-iodopropionate and ethyl sodio- cyanoacetate are allowed to interact i n molecular proportions at the ordinary temperature, a curious reaction takes place, which results in the formation of ethyl y-cyanopentane-ayc-tricarboxylate and regenera- tion of half of the ethyl cyauoacetate : 2C02Et*CHNa*CN + ZICH,*CH,*CO,Et = CO,Et*C(CN)(CH,*CH,*CO,Et), + CO,Et*CH,.CN + 2NaI. Under the conditions employed, it was to be expected that the mono- substitution product, e t h y 1 a-cyanopropane-a y-dicarboxy late, C0,Et-C H(CN) CH2*CH2*C0,Et, would be formed, but no trace of this substance could be isolated, the reaction proceeding apparently quantitatively in the direction shown above." Ethyl cyanopentanetricarboxylate is hydrolysed by boiling with concentrated hydrochloric acid with the formation of pentane-ayc-tricarboxylic acid, CO,K*CH(CH2*CH2~C0,H)~, which * In the case of the very similar action of ethyl 8-bromopropionate on ethyl sodiomalonate, Emery (L'cr., 1891, 24, 282) showed that the principal product of the reaction was the iioriiial mono-substitution product, namely, ethyl propauc- aay-tricarboxylate, C02Et'CH,'CH,'CH(C0,Et)z, but he also noticed that small quantities of ethyl pentane-ayyc-tetracarboxylate, CO,Et'CH;C H,*C( C0,Et)2*CH,*CH2*C0,Et, were produced at the same time.418 PEHKIN : 8-KETOHEXAHYDROBENZOIC ACID.melts a t 116-118°, and has already been described by Emery (Ber., 1891, 24, 384) and by Bottomley and Perkin (Trans., 1900, 77, 299). When the sodium salt of this acid is heated with acetic anhydride, or when the acid itself is digested with acetic anhydride and then distilled, a remarkable decomposition takes place with elimination of carbon dioxide and water and formation of 6-ketohexulqdrobenzoic acid (m. p. 68") : Similar cases of ring formation from dibasic acids have already been observed on one or two occasions, thus Lapworth and Chapman (Trans., 1900, '77, 464) showed that homocamphoronic acid, when slowly distilled, is converted into camyhononic acid, and Perkin and Thorpe (Trans., 1904, 85, 138) recently obtained ketodimethylpentamethylenecarboxylic acid by heating t,he sodium salt of dimethyl butanetricarboxylic acid with acetic anhydride : CH,-VH, ChTe2*C0 -+ CO,H*CH< CH,*CH,* C0,H Co2H'CH<CMe,* CO,H Experiments are in progressIwhich, it is hoped, will indicate to what extent this valuable method may be employed in the formation of closed chain compounds containing the keto-group.8-Ketohexabydrobenzoic acid crystallises with 1 molecule of water and shows all the properties of a keto-acid ; it yields the following derivatives, the preparation and properties of which are described in the experimental part of this paper : methyl 8-?cetohexuJ~ydrobmzoate, CO(CH,*CH,),CH*CO,Me (b.p. 140°/20 mm.) ; ethy2 8-ketohexahydro- benxoccte, CO(CH2-CH,),CH-C02Et (b. p. 158'/40 mm.) ; 8-ketoximehexa- hydrobenxoic acid, (OH)N:C(CH,*CH,),CH'CO,H (m. p. 147") ; the semicurbaxone, NH;CO'NH* N:C( CH,*CH,),CH*CO,H (m. p. 1 9 4'). When treated with phenylhydrazine, 6-ketohexahydrobenzoic acid yields a yellow oil, which is doubtless the hydrazone, and this, when boiled with hydrochloric acid, loses ammonia and is converted into a sparingly soluble crystalline substance having the formula C13H1202N. This decomposition is evidently analogous to the well-known conver- sion of phenylhydrazones into indole derivatives discovered by EmilPERKIN : &KETOHEXAHYDROBENZOIC ACID. 419 Fischer, and the substance CI3Hl3O2N is obviously tetrahydroca~baxole- p-carboxylic acid, formed according to the scheme : C H CH2 CH / '\ / \ \ / YH2 s*NH*Q gF CO,H*CH CH, C H CH CO,H*CH C--C CH / \ \/ '\ / CH CH2 YH, Q:N*NH-E QH -~ CH \ / CH2 It is interesting to observe that Baeyer and Tutein (Ber., lSS9, 22, 21 84) had previously described the exactly similar conversion of y-keto- hexahydrobenzoic acid into tetrahydrocarbazole-m-carboxylic acid by the action of phenylhydrazine and subsequent treatment with hydro- chloric acid.~arboxyhexccmethenyl-6-l%etohexali?/dro~enzo~c Acid.-During a series of experiments on the preparation of the methyl and ethyl esters of 8-ketohexahydrobenzoic acid (see above) i t was noticed that, when the acid was boiled with methyl alcohol and sulphuric acid, a large quantity of an oil was formed which, on account of its high boiling point (255O under 20 mm.pressure), was evidently a condensation product. The examination of this oil proved that it was methyl carboxyhexamethen yl- 8-ketohexahydrobenxoute, formed by the elimination of water from two molecules of methyl 8-ketohexahydrobenzoate, and therefore its con- stitution is obviously that represented by the formula : On hydrolysis, this methyl ester yields the free cayboxyhexa- methenyl-6-ketohexahydrobenzoic acid, a colourless, crystalline substance which melts a t 170". Although it is well known that ketopolymethyl- ene ring compounds readily condense with substances containing the >CO group, no case, like the above, seems to have been observed in which condensation of two molecules of the substance is brought about simply by boiling with methyl alcohol and sulphuric acid.Reduction of 6-Keto~exahydrobenxoic Acid. _Tiormation of 8-Hydroxy- hexahydrobenxoic Acid and the Conversion of this Acid into A3-Tetra- hydrobenxoic Acid. 8-Ketohexahydrobenzoic acid is readily reduced by sodium amalgam with formation of 8-hydroxyhexalr ydrobenxois (cyclohexanol-4-carboxyEic) acid,420 PERKIN : 8-RETOHEXAHYDROBENZOIC ACID. (m. p. 121"), and, since this acid shows no tendency t o form a lactone, it is evidently the trans-modificatioa. It was to be expected that both the cis- and trans-forms of this acid would be produced by the reduc- tion of the keto-acid, but, although a careful search was made, no trace of the cis-acid could be isolated.When the hydroxy-acid is heated with hjdrobromic acid, it is con- verted into 6-bronzo~exal~yclrobenxoic (browzocy clohesane- 4-carboxylic) mid (m. p. 167"), and this, when digested with sodium carbonate, readily loses hydrogen bromide with formation of A3-tati*al~ydrobelzxoic (A3-cy~lohexenecarboxyZic) acid, This new acid melts at about 13Oand distils at 237O (748 mm.) ; i t shows all the properties of a n unsaturated acid, since i t reduces per- mamganate, combines with hydrogen bromide yielding y-bromoheza- hpdrobenzoic acid (m. p. l22"), and at once decolorises bromine with formation of y8-dibromohexahydrobengoic ( 3 : 4 dibromocyclohexanecarb- oxylic) acid (m. p. 86"), cH2<CH2- CHBr'CH2>CH*COzH C H, and C H B r < ~ ~ ~ ~ ~ ~ > C H * C O z H The above synthesis of A3-tetrahydrobenzoic acid is a welcome addition to our knowledge of the chemistry of the tetrahydrobenzoic acids, Pince all the possible isomerided have now been prepared and carefully investigated.The Al-acid (m. p. 29' and b. p. 240-243') was obtained by Aschan (Annalen, 1892, 271, 267) from bromohexa- hydrobenzoic acid by elimination of hydrogen bromide, whereas the A2-acid (b. p. 235'), is formed by the direct reduction of benzoic acid with sodium amalgam (Hermann, Annalen, 1864, 132, 75; Aschan, loc. cit., p. 236). Action OJ Hpdrogelz Cyanide on 6- Ketohexa~~ydrobenxoic Acid. Formation, of the cis- and trans-Hod(fications of a-Hgdroxyhexu- hydroterephthcclic Acid und of AI- ~etrc~hydrotere~hthulic Acid. When 6-ketohexabydrobenzoic acid is treated with hydrogen cyanide i n the presence of potassium cyanide (compare Lapmortb, Trans., 1903,PERKIN : 8-KETOHEXAHYDROBENZOIC ACID.421 83, 995), addition takes place readily with formation of an oil, which consists of the mixed nitriles of the cis- and trams-modifications of a-hydroxyhexahydroterephthalic acid, and from which the trans-nitrile (m. p. 140') was isolated in a crystalline condition. By hydrolysing the mixed nitriles under the conditions described in the experimental part of this paper (p. 435), the corresponding hydr- oxy -acids, OH CO,H \ / C C0,H OH \ / c: / \ 7% y 3 2 \ / CH2 CH,' C /\ H CO,H Tras~s (m. 1). 228"). were easily separated in a state of purity. I n the absence of definite proof of their configurations, the more sparingly soluble and less fusible isomeride was taken as being the ts.a?zs-modikation.It is remarkable that when 6-ketohexshydrobenzoic acid is reduced, that is, when the keto-group, >CO, is converted into >CH*OH, only one modification of hydroxyhexnhydrobenzoic acid should result, whereas, in the very similar addition of hydrogen cyanide converting the keto-group into >C(CN)*OH, both the cis- and trans-modificationsof the nitrile of a-hydroxyhexahydroterephthalic acid should be formed. It is also interesting that in the latter case very much more of the &-modification is produced than of the trans-modification. Both t h e cis- and trans-modifications of hydroxyhexahydrotere- phthalic acid are decomposed on distillation, with loss of water and formation of h~-tetiiahydroterepht~cc~~c acid, / \ H CO,H Cis (m.p. 168-170"). QO2H C CO,H and the acid thus synthesised is identical with the A*-tetrahydrotere- phthalic acid which Baeyer (Amalen, 1888,245,160) obtained directly from terephthalic acid by reduction with sodium amalgam. The cis- and trccns-modifications of a-hydroxyhexahydroterephthalic VOL. LXXXV, F F422 PERKIN : 8-EETOHEXAHYDROBENZOIC ACID. acid are readily acted on by concentrated sulphuric acid, decomposition taking place in both cases in two distinct directions. Part of the acid is decomposed with elimination of carbon monoxide and water and formation of 8-ketohexahydrobenzoic acid, pP C*OH I CO,H co bO,H whereas the remainder is converted into A*-tetrahydroterephthalic acid by the simple elimination of water.The investigation of 8-ketohexahydrobenzoic acid and its derivatives i s being continued in various directions. E x P E R I MENTAL. Eth y l y-Cy anopelztane-a yets.icarbox y late, C0,Et (CN)C( CH,* CH, CO,Et),. I n preparing this cyano-ester, the quantities usually employed were sodium (25 grams), alcohol (400 c.c.), ethyl cyanoacetate (125 grams), and ethyl P-iodopropionate* (250 grams). The sodium is dissolved in the alcohol in a two-litre flask, and, after cooling, the ethyl cyanoacetate added all at once ; the flask is then cooled again in running water and the ethyl P-iodopropionate added in three portions, care being taken that the temperature never rises above 25", as otherwise some of the ethyl P-iodopropionate is liable to be decomposed with formation of ethyl acrylste (compare Trans., 1904, 85, 120).After being left overnight, the mixture is heated for half an hour on the water-bath, diluted with three times its volume of water, and extracted three times with ether. The ethereal solution is washed until free from alcohol, dried over * Owing t o the costliness of 8-iodopropionic acid, a number of comparative ex- periments were made with the object of discovering the best conditions for con- verting it into its ethyl ester, arid the following process was found to give excellent results. The acid (500 grams) is dissolved in alcohol (400 c.c.) and niixed with a cold saltition of sulphuric acid (150 grams) in alcohol (200 c.c.) and the whole allowed to remain for four days a t the ordinary temperature.The product, which will have separated into two layers, is mixed with three volumes of water, extracted twice with ether, the ethereal solution washed with water containing a little sulphurous acid (to remove iodine), and fractionated under reduced pressure. The boiling point is 136" under 100 mm. pressure, and the yield 625-535 grams, or 94 per cent, of that theoretically possible.PERKIN : &KETOHEXAHYDROBF,NZOIC ACID. 423 calcium chloride, and evaporated, and the residual oil fractiooated under 20 mm. pressure. At first, unchanged ethyl cyanoacetate passes over, then the temperature rises rapidly to 220°, almost the whole of the residue passing over between this and 240' ; on repeating the fractionation, the pure substance is then easily obtained, distilling almost constantly at 228' (20 mm.).0.1888 gave 0.3881 CO, and 0.1245 I3,O. 0.2880 ,, 12.3 C.C. nitrogen at 15' and 756 mm. N = 4.9. C = 57.0 ; H = 7.4. C,,H,,O,N requires C = 57.5 ; H = 7.3 ; N = 4.5 per cent. Ethyl cyanopentanetricarboxylate is a viscid, colourless oil, which, when placed in iz freezing mixture, showed no signs of solidifying. It is formed apparently quantitatively, according to the equation, given in the introduction (p. 417), since on one occasion, when four of the above quantities were worked up together, the yields obtained were ethyl cyanoacetate (190 grams), and ethyl cyanopentanetricarb- oxylate (680 grams), whereas the theoretical quantities are 240 and 686 grams respectively. Pentane-aye-tricarboxyEic Acid, CO,H*CH( CH2*CH,* CO,H),.Ethyl cyanopentanetricarboxylate is with difficulty completely hydrolysed by boiling with mineral acids, especially when the opera- tion is carried out in a reflux apparatus, and the alcohol formed cannot escape. This behaviour caused great difficulties a t first, and a number of experiments had to be made before the best conditions for hydro- lysis were discovered; now that these have been determined, the yield of acid is almost quantitnt,ive. The pure cyano-ester (100 grams) is mixed with concentrated hydro- chloric acid (200 c.c.) in a litre flask fitted with a ground-in air con- denser, and carefully heated just to boiling on a sand-bath. I n the course of about half an hour, the oil will have completely dissolved, and care is necessary at this stage to avoid loss owing t o frothing due to the elimination of carbon dioxide.After 1 hour, the air-tube is removed and the alcohol allowed to escape, a further quantity of hydrochloric acid (50-100 c.c.) is added, and the boiling continued in the open flask for 8 hours. The crystals of pentanetri- carboxylic acid mixed with ammonium chloride, which will have separ- ated overnight, are collected at the pump, washed with concentrated hydrochloric acid, and the filtrate again boiled in an open flask, when a further crop of crystals will be obtained. These are treated in the same way as the first crop, and the boiling of the filtrate and the whole cycle of operations continued as long as crystals continue to separate, The last mother liquors, which have a brown colour, are evaporated on the watwbath and the viscid syrup set aside to crystal- F F 2424 PERKIN : ~-RETOHEXAHYDROBENZOIC ACID.lise ; the crude acid is then placed in contact with porous porcelain, when i n the course of a few days it wiIl have become almost white. The crops of crystals are combined, thoroughly dried on the water- bath, powdered, and extracted with pure ether," preferably in a Soxhlet apparatus, by which means the acid is dissolved, and separated from the insoluble ammonium chloride. The concentrated ethereal solution deposits the pure acid in crysialline crusts, which are collected and washed with ether, and from the mother liquors the remainder of the acid is obtained by concentrating and recrystallising either from ether or from hydrochloric acid.0,1627 gave 0.2801 CO, and 0.0874 H,O. C = 47.0 ; H = 6.0. C,H,,O, requires C = 47.1 ; H = 5.9 per cent, Pentane-ay€-tricarbOxyZic acid melts at 116-1 18", and has already been prepared by Emery (Bey., 1891, 24, 284) from ethyl pentane- tetracarboxylate, (CO,Et),C(CH;CH,*CO,Et),, and by Bottomley and Perkin (Trans,, 1900, '7'7, 399) from ethyl pentanehexacarboxylate, (C02Et)2CH*CH,*C(C0,Et)2*CH2*CH(C02Et)2~ by hydrolysis, but the previously observed melting points were rather too low. The process given in this paper is by far the most suitable for the preparation of large qutntities of this acid. &Keto?Lexn?&ydrobenzoic Acid CO<CH,. C H 2 * C H 2 > ~ ~ * ~ ~ , ~ . CH, This acid may be obtained in a variety of ways from pentanetri- carboxylic acid, as, for example, by the distillation of the ammonium salt or, better, by heating the sodium salt with acetic anhydride under exactly the same conditions as those described in the preparation of ketodimethylpentamethylenecarboxylic acid from dimethylbutane- dicarboxylic acid (Trans., 1904, 85, 138), but, as the result of a large number of experiments, the following process, which should be followed exactly, was found to give by far the most satisfactory results.The pure tribasic acid, in quantities of not more than 15 grams, is heated with twice its weight of acetic anhydride for 6 hours, the acetic acid and excess of acetic anhydride are then distilled off, the heating being continued until a thermometer, immersed in the liquid, registers 180".The residue is then transferred to a distilling flask connected with the pump and heated in a metal-bath under 15 mm. pressure, the thermometer being in the liquid during the first stages of the opera- * I t is important that the acid should be heated until every trace of hydrogen chloride has been removed and the ether quite free from alcohol ; otherwise, partial etherification takes place and the acid does not crystallise.PERKIN : 8-KETOHEXAHYDROBENZOIC ACID. 425 tion. Some acetic anhydride passes over, and when the temperature reaches 200-210°, at which it is kept for 10 minutes, carbon dioxide is evolved and the pressure rises. The operation is sus- pended until the pressure is again reduced t o 15 mm., when the heating a t 220' is continued and, as soon as the evolution of carbon dioxide is seen to slacken considerably, the thermometer is drawn up to the usual position and the contents of the flask rapidly distilled until the temperature rises to 350°, the considerable brown residue remaining in the flask being neglected.The distillate, which from 15 grams of tribasic acid weighs about 7 grams, is dissolved in a little water, mixed with semicarbazide hydrochloride (5 grams) and sodium acetate (5 grams), and heated just to boiling, After remaining overnight, the sparingly soluble semi- carbazone (p. 437) is collected, washed with water, dissolved in hot dilute hydrochloric acid, and repeatedly extracted with ether. On evaporating the ethereal solution, a pale yellow oil is obtained which solidifies completely on cooling and consists of practically pure &kcto- hexahydrobenzoic acid.This is the most direct may, if small quantities only of the keto-acid are required, but in preparing the large quantities employed in this research the following method OF purification was found to be more convenient and economical, The distillate from the decomposition of 50 grams of the tribasic acid is fractionated under reduced pressure (15 mm.), when a large fraction passes over a t 200-225', leaving a viscid, brown residue in the flask. On refractionating, practically the whole distils at 205-215° under t h e same pressure and solidifies almost completely on cooling, and the acid thus obtained, after remaining in contact with porous porcelain for a few days, is pure enough for most purposes, and in this form" it was, unless otherwise stated, used in the experiments described in this paper.For the analysis, the acid was purified by conversion into the semicarbazone and then recrystallised from a mixture of benzene and light petroleum. O m 1 gave 0.3689 CO, and 0.1092 H,O. C = 58% ; H = 7.1. 0.1523 ,, 0.3288 CO, ,, 0.0960 H,O. C = 58.8; .H = 7.0. C,HIo0, requires C = 59.1 ; H = 7.0 per cent. s-Ketohexachydrobenxoic acid melts at about 67-68' with a slight previous softening, and is readily soluble in alcohol, ether, or benzene, but sparingly so in light petroleum; it dissolves readily in cold water, and is soluble in almost all proportions in hot water, from which it separates, on cooling, in magnificent groups of needles. propionate varied, in different experiments, from 130-160 grams.* When purified in this way, the yield of keto-acid from 1 kilo. of ethyl 8-ioilo-426 PERKIN : ~-KETOHEXAHYDROB~NZOIC ACID. The crystals thus obtained were dried at 40' and found t o contain one molecule of water of crystallisation. 0*1782 gave 0.3433 CO, and 0.1176 H,O. C =52.5 ; H= 7.3. C7Hlo0,,H20 requires C = 52.5 ; H = 7.5 per cent. The direct determination of the water of crystallisation was difficult, owing to the fact that the acid is slightly volatile a t loo", but, after heating at 80' until practically constant, 0.5096 gram lost 0.0592 gram, or 11.5 per cent., whereas the formula C7Hl,0,,H,0 requires 11.3 per cent. of water. On titration, 0.2177 of the crystals neutralised 0.0539 NaOH, whereas this amount of an acid, C7HIOOY,H20, should neutralise 0.0544 of the alkali, Formation of 6-Ketohexahydrobenzoic Acid frona the #odium Scdt qj Pentane-ayr-tricarboxylic Acid 6y the Action o f Acetic A n h y d d e , and from the Anmonium Salt by Distillation.I n the first experiment, the pure tribasic acid (41 grams) was dis- solved in water, mixed with sodium carbonate (33 grams), and evaporated t o dryness ; the dry salt was powdered and passed through a fine sieve and then heated with acetic anhydride at 165' for 4 hours, when large quantities of carbon dioxide were eliminated. The product was dissolved in water, mixed with semicarbazide hydrochloride (20 grams), when, after two days, a cake of the semi- carbazone of 6-ketohexahydrobenzoic acid had separated, and this was found to weigh 7 grams, I n the second experiment, 14 grams of the pure acid were dissolved i n excess of ammonia and evaporated to dryness.The viscid semi- solid mass was first heated under the ordinary pressure, when a quantity of ammonia and steam was given off and, as soon as the temperature of the liquid had risen to 150°, the whole was transferred to the vacuum distillation apparatus and the heating continued under 20 mm. pressure. The mass soon solidified completely, then melted again and gave off a quantity of gas, and much oil distilled over. The distillate, on treatment with semicarbazide hydrochloride and sodium acetate, yielded 3 grams of the semicarbazone of 6-ketohexahydrobenzoic acid. The Methyl and Ethgt Esters, tJ~e Oxinae, and tibe Xemicarbaxone of 6-KetoJbexahydrobenaoic Acid.Jfethyl 8-.KetohexaIqdrobenxoate, CO(CH,*CH,),CH*CO,Me.-A quantity of this ester was prepared by digesting the keto-acid with methyl alcohol and sulphuric acid in the usual way, but the yield wasPERKIN : 8-KETOHEXAHYDROBENZOIC ACID. 427 very poor owing to the formation of considerable quantities of methyl carboxyhexamethenylketohexahydrobenzoate (see p. 429). It is a colourless oil which boils at 140' under 20 mm. pressure. 0.1824 gave 0.4089 CO, and 0.1295 H,O. C = 61.2 ; H = 7.9. C,H,,O, requires C = 61-6 ; H = 7.7 per cent, Ethyl &Ketohexahydrobenxoate, CO(CH2*CH,),CH*C02Et.-This ester may be obtained, in an almost quantitative yield, by digesting the keto-acid (10 grams) with 40 grams of a 3 per cent.solution of hydrogen chloride in absolute alcohol in a reflux apparatus for 6 hours. The product is mixed with 3 times its volume of ether, water is then added, and the ethereal solution separated and washed 3 times with water and once with dilute sodium carbonate. After drying over calcium chloride and distilling off the ether, an oil is obtained which distils almost constantly at 158' (40 mm.), and consists of pure ethyl 8-ketohexahydrobenzoate. 0.1417 gave 0,3279 CO, and 0,1061 H,O. C=63.1; H=8*3. 6-Ketoximehexahydrobenxoic Acid, (0H)N : C( CH2*CH,),CH *C02H.- In preparing this oxime, the pure acid (2 grams) was dissolved in a little water, mixed with a strong solution of hydroxylamine hydrochloride (2 grams) and caustic potash (4 grams), and after remaining for 2 days, the product was acidified and extracted at least 16 times with ether.The ethereal solution was dried over calcium chloride and evaporated, and the viscid, oily residue, which gradually became semi- solid, left in contact with porous porcelain until quite dry, The substance was then dissolved in ether and the ethereal solution evaporated to a small bulk, when, on cooling, the oxime gradually separated in crystalline crusts. C9H,,0, requires C - 63.6 ; H = 8.2 per cent. 0.1704 gave 0.3336 CO, and 0.1080 H20. C=53.4 ; H = 7.0. 0-2014 ,, 14.8 C.C. nitrogen at 11' and 748 mm. N=8% C7H1,N0, requires C = 53.5 ; H = 7.0 ; N = 8.9 per cent. S-Hetoximehexahyd~.obenxoic acid, melts a t about 147' and is sparingly soluble in dry ether, chloroform, light petroleum, or benzene, but readily so in hot water or methyl alcohol. The.Xemicarbnxme of 6- Ketohemhp-hobenxoic acid, NH,*CO*NH*N: C(CH,*CH,),CH* C0,H. When ketohexahydrobenzoic acid is dissolved in a little water and mixed with a solution of semicarbaxide hydrochloride and sodium428 PERKIN : ~-EETOHEXAHPDROBENZOIC ACID. acetate, the semicarbazone separates at once as a sandy, crystalline powder. 0.1605 gave 28.7 C.C. nitrogen at 17' and 761 mm. This semicarbazone softens a t about 194' and decomposes a t 200O; it is very sparingly soluble ir, water and in alcohol, It dissolves readily in hydrochloric acid with liberation of the keto-acid, which may then be extracted with ether (see p. 435). N= 20.8. C, H130,N, requires N - 2 1 *2 per cent. Aci5on of Phenplhydruxine on 6-Ketohexahydrobenxoic Acid.Tormation of Tetrahydrocarbaxole-p-carboxylic Acid." A solution of the pure keto-acid (1 gram) in water was mixed with freshly distilled phenylhydrazine (2 grams) dissolved in dilute acetic acid, when, in a short time, pale yellow drops separated which, doubtless, consisted of hydrazone of the keto-acid. The substance showed, however, nosigns of crystallising, and therefore no attempts were made to analyse it ; it is also evidently an unstable substance, since i t rapidly became brown and gave ~ f f gas. The freshly prepared oily hydrazone dissolves in warm concentrated hydrochloric acid, and, if the solution is heated just to boiling and then poured into cold water, an ochre- coloured precipitate of crude tetrahydrocarbazolecarboxylic acid separates.The precipit ate was collected a t the pump, washed well with water, and left in contact with porous porcelain until quite dry ; it was then digested with benzene, filtered from a brown, amorphous impurity, and the benzene solution evaporated to a small bulk. The concentrated solution gradually deposited a, crust of warty crystals which was collected, washed with benzene, dried a t lOO', and analysed, 0.1635 gave 0.2184 ,, 11.4 C.C. nitrogen a t 15' and 749 mm. N=6.1. C,,H130,N requires C = 72.6 ; €I = 6.0 ; N = 6.5 per cent. Tetrahydrocarbacxole-p-carbozylic acid melts at about 195' and is readily soluble in alcohol, but sparingly so in chloroform, light petroleum, or benzene. The solution in sodium carbonate decolorises permanganate instantaneously. The finely powdered substance dissolves readily in concentrated hydrochloric acid and is repre- cipitated in flocks on dilution with water.When the acid is heated in a test-tube, it gives off gas and yields a crystalliue sublimate which is only partially soluble in sodium carbonate and has a pronounced odour of indole. * For the constitution of this acid, see p. 419. 0.4364 CO, and 0*0890 H,O. C = 72.8 ; H = 6.0.PERKIN : &KETOHEXAHYDROBENZOIC ACID. 429 Cus.boxghexacmethen y Z-8-htohexctl~ y ds. obenxoic Acid. As explained in the introduction (p. 419), the ester of this acid was accidentally obtained in the course of a n experiment made with the object of preparing methyl 8-ketohexahydrobenzoate from the acid by treatment with methyl alcohol and sulphuric acid in the usual way ; the details of the experiment are as follows.The acid (70 grams) was dissolved in methyl alcohol (200 c.c.), and concentrated sulphuric acid (50 c.c.) gradually added, the whole being well cooled during the addition. The mixture was digested for 4 hours in a reflux appsratus, and, when cold, mixed with 3 volumes of ether; water was then added and the ethereal solution separated and washed successively with water and dilute sodium carbonate. After drying over calcium chloride and evaporating, a viscid oil remained which, on fractionating under 20 mm. pressure, was readily separated into methyl 8-ketohexahydrobenzoate boiling at 140' (see p. 426) and a yellow, viscid liquid boiling at 255'. The latter was analysed with the following results : 0.266 gave 0,6356 CO, and 0.178 H20.C = 65.2 ; H = 7-4. C16H2205 requires C = 65.3 ; II: = 7.5 per cent. In order to obtain the free acid, this meth yE carboxyhexamethenyl-8-keto- hexuhydrobenxoate was dissolved in methyl alcohol mixed with a methyl alcoholic solution of caustic potash (8 grams) and heated to boiling for half a n hour. Water was then added and the solution evaporated until free fram methyl alcohol; i t was then acidified, when a viscid syrup separated which gradually solidified. After remaining in contact with porous porcelain until quite free from oily impurity, the acid was dissolved in dry ether, the ethereal solution concentrated and allowed t o remain, when the acid gradually separated as a hard, crystalline crust. 0.1508 gave 0,3474 CO, and 0.0958 H,O.C = 62% ; H = 7.0. C,,H,,05 requires C = 63.1 ; H = 6.8 per cent. CarboxyhexamethenyZ-8-ketohexahydrobenxoic acid melts at about 170°, but not sharply, since it softens a good deal below that temperature. It is readily soluble in alcohol, but sparingly so in ether and cold water; it dissolves, however, in hot water, and separates on cooling, in colourless, microscopic needles. The solution in dilute sodium carbonate instantly decolorises permanganate, and wben the aqueous solution of the acid is boiled with semicarbazide hydrochloride and sodium acetate, a very sparingly soluble semicarbazone is formed which was not further examined.430 PERKIN : 8-KETOHEXAHYDROEENZOIC ACID. Beduction of 8-Ketohexahydrobenxoic Acid. Pownation of trans-6-13ydroxyhexchydrobenxoic Acid, Sodium amalgam readily reduces 6-ketohexahydrobenzoic acid to the above hydroxy-acid, and, in studying this reduction, the keto-acid, in quantities of 5 grams, was dissolved in a slight excess of dilute sodium carbonate solution and treated with 150 grams of 4 per cent.sodium amalgam. The reduction was carried out in a large porcelain beaker, fitted with a mechanical stirrer and cooled externally by ice, and the sodium amalgam was added in quantities of 30 grams, a stream of carbon dioxide being passed during the whole operation. The product from 5 of these operations was separated from the mercury, acidified and extracted 10 times with ether, and since, even then, 8 grams of the reduced acid remained in the mother liquor, the extraction was continued by shaking on the machine with large quantities of ether.The combined ethereal extracts were dried over calcium chloride and evaporated t o a small bulk, when the hydroxy-acid (16 grams) gradu- ally separated in colourless nodules, and was further purified by a second recrystallisation from ether. 0.1860 gave 0.3932 CO, and 0.1391 H20. C = 58.4 ; H = 8.3. 0.1676 ,, 0.35'74 CO, ,, 0,1254 H,O. C=58.2 ; H=8*3 C7Hl2O3 requires C = 58-3 ; H = 8.3 per cent. trans-&Hydv*oxyhexahydrobenzoic acid melts at 1 20-121°, and is readily soluble in water or alcohol, but sparingly so in cold dry ether, chloroform, or benzene. This acid shows no tendency to yield a lactone ; when heated, it distils, and the oily distillate, which solidifies on rubbing, is completely soluble in cold sodium carbonate solution, and when the acid is boiled with 25 per cent.sulphuric acid, no signs of lactone formation could be detected. The ethereal mother liquors of the tran8-hydroxy-acid were evapor- ated to dryness and left for some days until the whole had become semi-solid. The mass was then spread on porous porcelain, and, when quite free from oil, recrystallised from ether, and in this may 5 grams of the pure trams-hydroxy-acid were obtained. I n order t o see whether any cis-hydroxy-acid had been formed dur- ing the reduction of the keto-acid, the porous porcelain was broken up and extracted in a Soxhlet apparatus with ether, but only a small quantity of oil was obtained, and after distilling this under 30 mm. pressure the distillate was found to be completely solublePERKIN : &RETOHEXAHYDROBENZOIC ACID.431 in cold dilute sodium carbonate solution, showing that no trace of lactone was present, and thus proving that no cis-hydroxy-acid had been formed. trans-8-Brornohexahydrobenxoic Acid and its Convmion into A3-Tetra- hydrobenxoic Acid, - - and h t o y-~romohexcchydrobenxo~c Acid and y8-Dibrornohexahydro- benxoic Acid. When finely-powdered 6-hydroxyhexahydrobenzoic acid is mixed with hydrobromic acid (saturated at OO), it dissolves, and even after several hours no separation takes place. If, however, the solution is heated in a sealed tube at 100' for 30 minutes, a viscid, oily layer forms on the surface of the hydrobromic acid, and this gradually becomes semi-solid.After washing with water, contact with porous porcelain readily removes the oily impurity, and the residue crystallises readily from light petroleum (b. p. 70-80') in colourless, glistening plates. 0*1388 gave 0.1254 AgBr. Br = 38.4. C7H,,0,Br requires Br = 38.6 per cent. trans-6-BromohexaiLyd.robertxoic acid softens a t 160' and melts at about 167O, and is readily soluble in chloroform, benzene, or alcohol, but less so in cold light petroleum. When the bromo-acid is boiled with water, hydrogen bromide is only slowly eliminated, and, indeed, the acid crystallises well from water if the operation is carried out moderately rapidly, and is then obtained in plates having a satiny lustre and exactly resembling benzoic acid, and, like this acid, the bromo-acid is also volatile in steam.A3-Tetrc~hydrobenxoic Acid.-In preparing this acid, 8-bromohexa- hydrobenzoic acid was dissolved in a considerable excess of moderately concentrated sodium carbonate and the solution heated to boiling for half an hour, On acidifying, As-tetrahydrobenzoic acid was pre- cipitated as a colourless oil, and, after extracting with ether, the ethereal solution was dried over calcium chloride, evaporated, and the residue distilled under 25 mm. pressure. With the exception of a small quantity of a higher boiling sub- stance (probably 6-hydroxyhexahydrobenzoic acid, produced from the bromo-acid by hydrolysis), the whole passed over at 145-150°, and, on432 PERKIN : 8-KETOHEXAHYDROBENZOIC ACID. subsequently distilling under the ordinary pressure, the acid was readily obtained pure as a colourless oil boiling constantly a t 237' (748 mm.).0.2348 gave 0.5700 CO, and 0.1 686 H,O. C = 66.2 ; H = 8.0. 0.1490 ,, 0.3626 CO, ,, 0.1064 H,O. C = 66.4 ; H= 7.9. C7H,,0, requires C = 66.6 ; H = 7.9 per cent. When placed in a freezing mixture, the distilled acid solidified com- pletely, and the melting point, determined by placing a thermometer in the melting acid, was about 13'. Possibly, if the acid were left in contact with porous porcelain in order to remove any trace of oily impurity, a higher melting point than this might be observed, but the amount of material at my disposal did not allow of this being done. A"-Tetrahydrobenzoic acid has a most unpleasant odour, closely resem- bling that of allylacetic acid ; it shows the properties of an unsaturated acid, since its solution in sodium carbonate decolorises permanganate instantly and the acid itself combines directly with both bromine and hydrogen bromide (see below).When the acid is exposed on a watch glass to the action of the air, very little change is noticed until after some days, and then a thin film of crystals forms on the surface, and these crystals appear to consist of benzoic acid, although this could not be established with certainty. I n any case, oxidation takes place only very slowly, and in this respect the A3-acid resembles the Al-acid and differs from the A2-acid, as the latter is rapidly oxidised to benzoic acid in contact with air. A3-Tetrahydrobenzoic acid shows no signs of being converted into a lactone when it is digested with dilute sulphuric acid.y-Brorno7~e~:al~yd~oben,-oic Acid.-A3-Tetrahydrobenzoic acid dissolves readily and completely in four times its volume of aqueous hydrobromic acid (saturated at 0') with evolution of heat, but the solution soon becomes cloudy, and in a short time a viscid oil separates on the surface of the hydrobromic acid. The oil gradually solidifies, and the aqueous layer becomes filled with colourless crystals. After 12 hours, water is added, the pasty, crystalline mass collected a t the pump, washed well, and left in contact with porous porcelain until quite dry. The residue crystallises from its warm dilute solution in light, petroleum (b. p. 60-80') in groups of flat needles, but from a hot concentrated solution the substance separates in glistening plates.Br= 38.6. C7H,,02Br requires Br = 38.6 per cent. y-Bv-omohexahydrobenxoic acid shrinks together a t 110' and melts at about 123,O, and repeated crystallisation did not alter this behaviour or make the melting point any sharper; it is practically insoluble in cold 0.1712 gave 0.1558 AgRr.PERKJN : &KETOI-IEXAHYDROBENZOIC ACID. 433 light petroleum or cold water, but readily soluble in benzene, alcohol, or ether. It dissolves t o a considerable extent in boiling water, and separates, on cooling, as a voluminous mass of glistening plates which closely resemble benzoic acid. I n its behaviour towards sodium carbonate, this y-bromo-acid differs in a marked manner from the 6-bromo-acid (see above), since, after boiling for half a n hour and acidifying, no sparingly soluble tetra- hydrobenzoic acid separates.On extracting with ether, a very soluble oily acid is obtained, which rapidly becomes insoluble in sodium carbonate and develops the odour of a lactone. It is probable t h a t the neutral oil produced in this way is the lactone of y-hydroxyhexa- hydrobenzoic acid. ys-Di6romo~exnl~~drobenxoic Acid.-When the solution of tetrahydro- benzoic acid in chloroform is cooled to - 10" and treated with bromine, the colour of the halogen disappears instantly, and only traces of hydrogen bromide are formed. As soon as the solution had acquired a permanent pale yellow tint, it was transferred t o a large watch glass and left until the chloroform had evaporated. A viscid, colourless oil resulted which did not crystallise for some days, but uitimately solidi- fied almost completely.The mass was rapidly washed with formic acid (sp. gr. 1.22) and recrystallised from this solvent, from which the dibromo-acid separates in small, glistening crystals. 0,1444 gave 0.1888 AgBr. Br=55*6. C7Hlo0,Br, requires Br = 55.9 per cent. ys-Dibro~rol~exalydrobenxoic acid melts at 84-86", and is sparingly soluble i n cold formic acid and light petroleum, but readily so in alcohol, ether, or chloroform, I t dissolves readily in sodium citrbonate, and, if t,he solution is boiled for 5 minutes and then acidified, nothing separates, but ether exkacts a very readily soluble syrupy acid which, when kept, shows signs of crystallising. Since this acid reduces ammoniacal silver solution, but does not decolorise permanganate, it is probably y6-dihydroxyhexahydrobenzoic acid produced from the dibromo-acid by hydrolysis.439 PERKIN : &KETOHEXAHYDROBENZOIC ACID. Action of Hydrogen, Cyanide on 6-lietohexahydrobenxoic Acid.Forma- tion of the cis- ccnd trans-Modifcations of a-Hydroxyhexahydro- terephthalic Acid, CO,H= (OH)C<:2:;2>CH*CO2H, and of Al-Tetra~~ydroterephth~lic Acid, I n the first experiments on the action of hydrogen cyanide on Gketohexahyd robenzoic acid, the conditions recommended by Baeyer and Tutein (Ber., 1889, 22, 2186) in the similar case of y-ketohexa- hydrobenzoic acid were employed, that is t o say, the acid was mixed with potassium cyanide and the mixture treated with hydrochloric acid. The addition did not take place satisfactorily under these con- ditions, and, after several experiments, the following process, based on the observations of Lapworth (Trans., 1903, 83, 995), was employed with excellent results.The pure acid (4 grams) was dissolved in a little warm water, cooled until crystallisation just commenced, and then mixed with pure potassium cyanide (5 grams) and 10 C.C. of a 15 per cent. solution of hydrocyanic acid. The liquid gradually darkened, and after 24 hours had become nearly black; i t was mixed with a little hydrochloric acid and extracted 6 times with ether. After drying over calcium chloride and evaporating, a light brown syrup (5 grams) was obtained, which, on rubbing with a glass rod, rapidly became semi-solid. The mass was left in contact with porous porce- lain until quite dry; the residue was dissolved in ether, the solution digested with animal charcoal, filtered, and concentrated.An equal volume of chloroform was then added and the ether distilled off, when a sandy, crystalline precipitate was soon deposited, which melted at 1 SOo, and was easily identified as aminomalononitrile, NH,*CH(CN),, produced by the polymerisation of some of the hydrogen cyanide em- ployed. The filtrate from this substance deposited, on spontaneous evaporation, a mass of crystals, which were collected and purified by re-dissolving in ether, adding chloroform, and distilling off the ether as before. The glistening, pale yellow plates thus obtained consist of the n i t d e of trans-a-hydroxyhexahydrotere~~~thalic acid.0.1805 gave 12.6 C.C. nitrogen a t 14' and 774 mm. C',H,,O,N requires N = 8-3 per cent. The melting point of this nitrile is not sharp; i t begins to shrink together a t 125' and is completely melted at 140'. It is readily N=S.4.PERKIN : 8-KETOHEXAHYDROBENZOIC ACID. 435 soluble in water or methyl alcohol, but sparingly so in chloro- form, benzene, or light petroleum. When heated in a test-tube, water is eliminated and an oil distils which does not solidify on cooling, and, since the solution of this oil in sodium carbonate does not decolorise permangnnate, it may very probably be the correspond- ing cyano-lactone. When the nitrile is hydrolysed under the con- ditions stated below, it yields ti.ans-a-hydroxyhexahydroterephthalic acid. The chloroform mother liquors of this trans-nitrile contain the &modification, but, as this did not appear to crystallise readily, no attempt was made t o isolate i t in a condition suitable for analysis.I n preparing the cis- and trans-modifications of hydroxyhexahydro- terephthalic acid in quantity, 20 grams of ketohexahydrobenzoic acid were converted into the mixed nitriles, and after extracting with ether, the oil was dissolved in 100 C.C. of concentrated hydrochloric acid and heated, first on the water-bath for three hours and then t o boiling for one hour. The whole was evaporated to dryness, the crystalline cake was then ground up and digested with absolute alcohol, and after filtering from the ammonium chloride the alcoholic solution was evaporated almost t o dryness. The residue, which still contained ammonium chloride, was dissolved in excess of dilute barium hydroxide and boiled until the evolution of ammonia had com- pletely ceased, the barium was then exactly removed by sulphuric acid, arid the filtrate from the barium sulphate evaporated almost to dryness.After remaining for 24 hours, the crystalline crust was left in con- tact with porous porcelain until quite dry, dissolved in water, digested with animal charcoal, and evaporated considerably, when a mass of white, nodular crystals of the twms-hydroxy-acid gradually separated. Thesa were collected and purified by crystallising twice from water, when the substance was obtained in hard crystals containing one molecule of water of crystallisation. After being exposed to the air for 6 days, the following results were obtained on analysis : 0.1654 gave 0.2822 CO, and 0.1023 H,O.C=46-5; H=6-9. C8H1,0,,H20 requires C = 46.6 ; H = 6-8 per cent. 0-2416 gram, heated a t 90" until constant, lost 0.0221 gram or 9-1 The per cent., whereas C,H,,O,,H,O contains 8-8 per cent. of water. anhydrous acid was also analysed. C = 51.3 ; H = 6.4. 0.1852 gave 0-3488 GO, and 0,1064 H,O. C,H,,O, requires C = 51.1 ; H = 6.4 per cent. On titration with decinormal caustic soda, 0.0956 gram required for436 PERKIN : &KETOHEXABYDIIOBENZOIC ACID. neutralisation 0.0408 NaOH, whereas this amount of a dibasic acid, C,H,,O,, should neutralise 0.0407 NaOH. trans-a-Hydroxyhexahydroterep~~t~~lic acid melts a t 228-230' with decomposition, and when heated in small quantities in a test-tube, water is eliminated and a crystalline sublimate of A1-tetrahydro- terephthalic acid is obtained (p.437). I t is readily soluble in water or alcohol, but sparingly so in cold ether, chloroform, benzene, or concentrated hydrochloric acid. The neutral solution of the ammonium salt gives no precipitate with barium or calcium chloride, but a heavy CaSeous precipitate with lead acetate ; with copper sulphate, a turbidity is produced, and, on warming, a pale blue, amorphous precipitate separates. cis-a-Hydroxyhexaillydroterepl~t?~alic Acid. - The aqueous mother liquors of the tmrzs-acid yielded, on further concontrntion, first an additional small crop of the same acid, and afterwards the cis-hydroxy- acid separated in masses of irregularly shaped crystals.These were collected and purified by repeated crystallisation from water, the substance being dried at 100' before analysis. 0.1594 gave 0,2958 CO, and 0.0912 H,O. The substance was then recrystallised from ether and again C = 50.6 ; H = 6.4. analysed. 0.1832 gave 0,3428 CO, and 0.1052 H,O. C8H120, requires C = 51.1 ; H = 6-4 per cent. On titration, 0.1220 gram neutralised 0,051 gram NaOH, whereas this amount of a dibasic acid, C,H,,O,, should neutralise 0.052 NaOH. cis-a- Hydroxyhexahydrote~ephthalic acid melts at 168-1 70°, and is readily soluble in water and alcohol, but sparingly so in chloroform, benzene, or ether. If, however, the finely-powdered substance is digested in a reflux apparatus with a large quantity of ether, i t dis- solves, and, after concentrating considerably, the pure acid gradually separates in hard, crystalline crusts.When crystallised from water, the hard nodules obtained do not, like tbe trms-acid, contain water of crystallisation, since the air-dried substance loses only very slightly in weight on drying a t 90'. If a small quantity of the dry acid is heated in a test-tube, it first melts and then gives off water, and a sublimate of Al-tetrahydro- terephthalic acid forms on the cooler portions of the tube. When the cis-hydroxy-acid is mixed with concentrated sulphuric acid and warmed to about 60°, carbon monoxide is evolved in quantity, and if, when the evolution has ceased, the pale yellow liquid is poured into water, a white, crystalline precipitate separates. This was collected and found to consist of pure A'-fetrahydro- C=51*0; H=6*4.PERKIN : 8-KETOHEXAHYDROBENZOIC ACID.437 terephthalic acid, and from the filtrate a quantity of 6-ketohexahydro- benzoic acid was obtained by extraction with ether and identified by conversion into the semicarbazone. The cis-hydroxy-acid shows a curious behaviour when heated with fuming hydrobromic acid (saturated a t 0') in a sealed tube in a boiling water-bath. The acid dissolves, and then a gas (presumably carbon dioxide) is slowly given off and, on diluting the yellow solution with water, a quantity of a very voluminous, white, amorphous precipitate separates. This substance, which contains bromine, seems to be formed quantitatively ; it is practically insoluble in water, but dis- solves in sodium carbonate. It separates from dilute formic acid apparently as a microcrystalline precipitate ; this shrinks together a t 160-165O and has no definite melting point. This substance was not further investigated, but it was noticed that the trans-hydroxy-acid also appeared to show the same behaviour on heating with hydro- bromic acid. The amount of the mixed hydroxyhexahydroterephthalic acids obtained by the addition of hydrogen cyanide to 6-ketohexahydro- benzoic acid and subsequent hydrolysis was upwards of 70 per cent. of the theoretical, and it is interesting to note that in this way a very much larger amount of the cis-acid is produced than of the trans- acid. h1-TetrahydrotwephthaZic Acid.-As explained above, this acid is produced when either the cis- or trans-modification of a-hydroxyhexa- hydroterephthalic acid is heated. In investigating this decomposition, 5 grams of the pure cis-acid was distilled under 10 mm. pressure, the sublimate was dissolved in hot sodium carbonate, the solution decolorised with animal charcoal, concentrated, and acidified, when the tetrahydro-acid separated as a microcrystalline deposit. This was collected and recrystallised from much water. 0.1406 gave 0.2894 CO, and 0.0764 H,O. C,H,,O, requires C = 56.4 ; H = 5.9 per cent, This acid melted above 300' and reduced permanganate, and was evidently identical with the Al-tetrahydroterephthalic acid described by Baeyer (Aniden, 1888, 245, 160). In order t h a t there should be no doubt as to its identity, the acid was converted into its methyl ester by digesting with a five per cent. solution of hydrogen chloride in methyl alcohol. The methyl ester, thus obtained, crystallised from dilute methyl alcohol in needles and melted at 38--39O, whereas Baeyer gives 39' as the melting point of methyl A'-tetrahydro- terepht halat e. C=56.2; H= 6.0. The author wishes to express his thanks to Mr. 5. S. Pickles, B.Sc., VOL. LXXXV. G Q438 RIPPING AND SALWAP: ARRANGEMENT IN SPACE OF THE for valuable assistance in carrgin'g out these experiments, and to state that much of the heavy expense entailed has been met by a grant from the Government Grant Commitee of the Royal Society. THE OWENS COLLEGE, MANCHESTER.

ISSN:0368-1645    

DOI:10.1039/CT9048500416

出版商:RSC    

年代:1904

数据来源: RSC

摘要:

438 RIPPING AND SALWAP: ARRANGEMENT IN SPACE OF THE XLIX.-The Arrarhgement in Space of the Groups Com- bined with the Tervalent Nitrogen Atom. By FREDERIC STANLEY KIPPINGI and ARTHUR HENRY SALWAY. THE discussion as to the cause of the existence of isomeric oximes and of that of certain other derivatives of tervalent nitrogen, which for so many years occupied a prominent place in current chemical litera- ture, apparently came to an end some time ago ; at any rzte, the occur- rence of such isomerides now excites but little interest, and is, in fact, taken as a matter of course, and the usual " ayn )) and " anti " config- urations having been respectively assigned to the two isomerides, the matter is dismissed from further consideration. It is far from our intention to attempt to make any review of the literature of this subject, as it would occupy far too much space even were we merely to refer in the briefest manner to the views of the principal workers in this field, or to recapitulate quite cursorily the various arguments which have led to the adoption of the '' syn " and (' anti ') configurational formuls for isomeric derivatives of ter- valent nitrogen ; it is necessary, however, in order to explain one of the objects of this investigation, to briefly recall the meaning and consequences of such stereochemical representations.According to Hantzsch and Werner (Bev., 1890, 23, 11), the exist- ence of this type of isomerides is to be explained by assuming that the groups around a tervalent nitrogen atom a r O situated at the corners of a regular tetrahedron, of which the nitrogen atom itself occupies the fourth corner ; the " syn " and " anti " configurations are thus respectively represented by the following symbols : N OHGROUPS COMBINED WITQ TEE TERVALENT NITROGEN ATOM.439 Now this hypothesis, which seems to be generally accepted by chemists as a means of explaining the phenomenon in question, involves, or seems to involve, certain consequences which are by no means confirmed by experimental evidence. In the first place, this conception premises the existence of isomerides in the case of a great number of compounds which hitherto have been obtained i n only one form (compare Hantzsch and Werner, Zoc. cit.). Azobenzene, for example, as far as is known, does not exhibit stereoisomerism, and yet the existence of ayn- and anti-forms of this and of axo- and aaoxy-compounds generally would be in accord- ance with theory ; as a matter of fact, the only cases of isomerism in such compounds which have yet been described are those of p-azoxy- toluene and trinitrazotoluene (Janovski and Reimann, Ber., 1889, 22, 40, and Janovski, Ber., 1889,254, 1172), and it would appear that even these may be capable of some other explanation.Many other examples of a similar absence of conceivable isomerides might be quoted, but it is unnecessary to do so; Hantesch, himself, has tried to obtain isomeric compounds containing dou bly-linked nitrogen in many cases where their existence seemed a probable consequence of his and Werner's theory (compare Ber., 1893, 26, 926).The question which more immediately concerns our work, however, is that of the possibility of existence of isomerism in compounds of the type NR,R,R,; if the three atoms or groups combined with ter- valent nitrogen are situated at the angles of a regular tetrahedron, secondary and tertiary nitrogen bases which are of the above type, that is to say, in which the three atoms or groups are all differenti from one another, should exist in enantiomorphously related forms, and the latter in all probability would be optically active ; such forms would be represented respectively by the configurational formulae : N N The possibility of the existence of such isomerides has been pre- viously considered, and, in fact, experiments have already been made in order to try and obtain such optically active compounds.Kraft (Ber., 1890, 23, 2780), at the suggestion of Hantzsch, attempted to resolve bmxylethylamine and p-tolylhydrszine by fractionally crystal- lising their hydrogen tartrates, whereas Behrend and Kiinig (Amlnakm, 1891,263,175 ; Ber., 1891,!24, 447c) carriedput similar and additional Q G 2440 KIPPING AND SALWAY: ARRANGEMENT IN SPACE OF THE experiments with P-benzylhydroxylamine, whilst Ladenburg (Bey., 1893, 26, 864) investigated methylaniline, tetrahydroquinoline, and tetrahydropyridine ; all these investigations, and likewise some experi- ments of Le Be1 (Compt. rend., 1891, 112, ll), were, however, uni- formly unsuccessful. Now it may be observed i n connection with the investigations just referred t o t h a t the point actually put t o the test of experiment was not so much the existence of isomeric tervalent as that of quinque- valent derivatives of nitrogen containing two identical groups or atoms.Even had isomeric salts been obtained, the latter might not have afforded enantiomorphously related bases when decortposed with alkalis, because in passing from the quinquevalent to the tervalent state a complete rearrangement in space of the nitrogen ‘‘ valencies ” might occur, and both the isomeric salts might give the same compound or the same mixture of isomerides. These considerations led us to undertake the present work in the hope of adding t o our knowledge of the configuration of tervalent nitrogen. Our first experiments were directed towards an investiga- tion of the products obtained by the action of an externally compen- sated acid chloride on certain primary and secondary amines.A dl-acid chloride, such as dl-phenylchloroacetyl chloride, gives with a dl-base, such as dZ-hydrindamine, a mixture of two isomeric phenyl- chloroacetohydrindamides, which can be separated by fractional crystal- lisation (Kipping and Hall, Trans., 1901, 79, 444) ; this is due, of course, to the formation of the four compounds, dBdA, IBlA, dEBZA, lBdA, which are enantiomorphously related in pairs, and which crystallise together, forming two different dl-isomerides. An externally compen- sated acid chloride may therefore be employed €or the detection of asymmetry in bases, a fact which in the first place we confirmed by proving that dl-benzylmethylacetyl chloride gives with dl-hydrind- arnine a mixture of isomeric dl-substituted amides from which the two compounds may be isolated without much difficulty.Now if instead of a base such as dl-hydrindamine, which owes i t s asymmetry to a carbon group, some primary or secondary base, such as p-toluidine or methylaniline, were treated with a dl-acid cbloride, it seems reasonable to suppose that two isomeric substituted amides should be obtained, provided that the three nitrogen valencies are loot arranged in one plane, but, aa surmised by Hantzsch and Werner in the case of the oximes ; the d-component of the acid chloride mould give equal or approximately equal quantities of the two compounds,GROUPS COMBINED WITH THE TERVALENT NITROGEN ATOM. 441 N which are not enantiomorphously related, whilst the I-component would give rise to two analogous isomerides ; these would then crystal- h e in pairs, just as in the case of the hydrindamides, giving two different dI-substituted amides.The experiments were tried with metbylaniline, p-tohidine, phenyl- hydrazine, and ben zylaniline, the acid chloride being dl-benzylmethyl- acetyl chloride; in no case, however, could we observe the least indication of the formation of isomerides. I n all these examples, if non-enantiomorphously related isomerides were produced at all, each of the four compounds which might con- ceivably have been formed would contain one group enantiomorphously related to t h a t in one of its isomerides ; thus, although dBdA and dBIA are not enantiomorphously related, a portion of the molecule of the one does bear such a relationship to a portion of that of the other; similarly with ZBZA and ZBdA.This might condition the formation of a double racemic salt { fzf2 such as is possibly produced from dI-hydrindamine and dl-mandelic acid (Kipping and Hall, Trans., 1901, 79, 442), and thus account for the failure of our attempts t o separate isomerides. I n order t o limit such a possibility t o some extent, we next made experiments on the behaviour of some primary and secondary bases towards optically active benzylmethylacetyl chloride, and examined the substituted amides derived from p-toluidine and benzylaniline ; no evidence of the formation of isomerides dBdA and ZBdA, however, could be obtained. Our last and most conclusive experiments on these lines were made with optically active bases and optically active benzglmethylacetyl chloride.If, in tho case of asymmetric carbon groups, the two enantiomorphously related compounds are generally separable as soon as an asymmetric radicle is substituted for one of the atoms or radicles in each of the groups, it would seem highly probable that the in- troduction of a second centre of asymmetry would greatly enhance the difference between the two isomerides aB regards their solubility, &c., and thus greatly facilitate their separation by fractional crys- tallisa tion. Now when a n optically active base such as I-menthylamine is treated with d-benzylmethylacetyl chloride, the resulting substituted amide contains two centres of asymmetry due t o carbon; if, in addition,442 RIPPING AND SALWAY: ARRANGEMENT IN SPACE OF THE the three nitrogen valencies are so arranged as to condition a third centre of asymmetry, two very dieerent isomerides, represented respectively by the following configurational formulte, should be obtained : p, As- Rzf N The bases examined in this way werev d-hydrindamine, E-methyl- hydrindamine, Lmenthylamine, and E-phenylethylamine ; in all cases, however, the product seemed to consist of one substance only, and when crystallised fractionally from various solvents retained its uniform character, W e conclude from these results that the three radicles, together with the tervslent nitrogen atom itself, are situated in one plane; that two of the radicles are symmetrically grouped with respect t o the third, and that in all probability this is true of any two, that is t o say, the whole arrangement is the most symmetrically possible one.&%amination of 0ximes.-Having failed to detect any indication of isomerism in the case of saturated tervalent nitrogen compounds which would be in accordance with the theory of Hantzsch and Werner, we next turned our attention t o the oximes themselves. Assuming that the isomerism of the syn- and anti-forms is structural and that the two compounds are represented by the formula $>C:N*OH and s > O < x H respectively, the latter, which contains an asymmetric carbon group, would exist in enantiomorphously related forms. If, then, an optically active acyl group mere substituted for the hydrogen atom, two non-enantiomorphously related optically active isomerides should be obtained, whereas by employing a dl-acy 1 chloride two different GLcompounds should be produced.I n tending, subsequently, t o investigate the benzaldoximes, we first studied the action of the dl-acid chloride on benzoinoxime, a substance which crystallises well, and the derivatives of which, it was thought, would show a similar behaviour. The reaction, however, proceeded in a manner altogether abnormal ; instead of obtaining a n acyl derivative by the simple displacement of the oximic hydrogen atom, we got a number of other products, namely, benzil, benzoin, benzaldehyde, benzonitrile, ammonium chloride, hydrogen chloride, benzylmethyl- acetic acid, and a compound melting at 126'; this decomposition of the oxime made i t impossible, of course, to accomplish the objects in view, but the result is not without interest as showing how easily theGROUPS COMBINED WITH THE TERVALENT NITROGEN ATOM, 443 benzoin molecule-which is generally so stable-may undergo decom- position.I n order to t r y and throw some light on this extraordinary reaction, we next studied the action of benzoyl chloride on benzoinoxime, but found that interaction proceeded in a normal manner, giving a benzoin- oxime benzoate in which the benzoyl group had displaced the oximic hydrogen atom; the constitution of the product was established by preparing the oxime of the benzoylbenzoin described by Zinin (Anna- Zen, 1857, 104, 117); the latter, which would necessarily have the constitution CoH5*CH(OBz)*CPh:N-OH, is not identical with the benzoyl derivative of benzoinoxime, so that our compound must be represented by the formula C,H5* CH( OH) CPh :N -0Bz.Failing to obtain the desired results with benzoinoxime, we made a few experiments with a-benzaldoxime, but were unabls to prepare the benzylmethylacetyl derivative owing to the decomposition of the aldoxime into benzonitrile. The Resolution of dl-Bases with tibe aid of d-Benzyl.metT~yl~ccety1 Chloyide. The resolution of a dl-base by fractionally crystallising the salt obtained by combining it with an optically active acid is very often unsuccessful owing to the fact that the two salts dBdA and ZBdA or dBZA and ZBZA may differ so slightly in solubility that no appreciable separation occurs, or, again, they may unite to form a partially racemic compound.Judging from the many cases which have been examined in this laboratory, it would seem that it is rather the exception than the rule for a resolution to take place, and were all the negative results of investigators on record, this conclusion would probably be amply substantiated; contrary to the views sometimes held, the use, in such experiments, of a relatively weak acid also seems to be just as likely to give satisfactory results as that of a relatively strong one, since hydrolytic dissociation seldom plays any important part, the difference in solubility of the salts being practically the only factor of importance. The investigations carried out during the last few years on the behaviour of dZ-hydrindamine have shown that the resolution of this base presents more than the usual difficulties, and that ordinary fractional crystallisation of its salts with d-bromocamphorsulphonic, d-chlorocamphorsulphonic, Reychler’s d-camphorsulphonic, Lmandelic, and d-tartaric acids fails to effect a separation of the two isomerides. In these circumstances, it occurred to one of us that a resolution might be accomplished by using an optically active acid chloride which would probably give isomeric substituted amides differing considerably in solubility j this expectation proved to be well-founded, inasmuch a s444 KIPPING AND SALWAY: ARRANGEMENT IN SPACE OF THE the product of interaction of dl-hydrindamine and d-benzylmethyl- acetyl chloride was easily separated into two isomerides having differ- ent specific rotations.d-Benzylmethylacetyl chloride was also employed for the resolution of dl-phenylethylarnine, and in this case also the two isomeric amides corresponding with the d- and E-bases respectively are separated with- out much difficulty. Judging from these examples, an optically active acid chloride may be employed with success when the use of an acid gives unsatisfactory results, but it will be necessary to wait until a greater number of cases have been tried before concluding that such a method is uni- versally applicable. One great drawback to the use of the acid chloride is that it may be impossible to obtain the optically active bases themselves, either because the substituted amides are so stable that they cannot be hydrolysed, or because so high a temperature is required to bring about hydrolysis that the bases undergo racernisa- tion.Thus, in the case of the benzylmethylacetohydrindamide, all ctttempts to regenerate the base were unsuccessful ; the amides are exceedingly stable towards ordinary hydrolysing agents at moderate temperatures, and although change occurs at high temperatures, it results in the formation of indene and ammonia. The application of an optically active acid chloride for the resolution of dl-alcohols or other dl-hydroxy-compounds has not yet been tried. EXPERIMENTAL. dl Hydrindamine and dl-Benzylmetlzylacetgl Chloride. I n order to satisfy ourselves that dl-benzylmethylacetyl chloride would prove serviceable in detecting asymmetry, we first tried its action on dl-hydrindamine, a base which is very prone to form partially racemic salts, and which is resolved into its enantiomorphously related components only with great difficulty. On adding the dl-benzylmethylacetyl chloride to the dl-base, both in ethereal solution, a development of heat occurred, and the hydro- chloride of the base separated in crystals ; after filtering, washing with water, and evaporating the ethereal solution, the substituted amide was obtained as an oil which rapidly solidified.When crystallised fractionally from methyl alcohol, ether, and other solvents, this product, which at first melts indefinitely at about 70°, is finally resolved into two isomerides. One of these crystallises in thick, transparent prisms melting at 110--21lo, but when the partially melted snbstance is allowed to cool it immediately solidifies, and when heated again melts at about 126O ; it is therefore dimorphous.Isolation Two Isomerides.GROUPS COMBINED WITH THE TERVALENT NITROGEN ATOM. 445 This amide is very readily soluble in cold methyl alcohol, moderately so in ether, but dissolves very sparingly in light petroleum (b. p. 40-60O). 0.1615 gave 0,4857 CO, and 0.1090 H,O. C,,H,,ON requires C = 81 -7 ; H = 7.5 per cent. The second isomeride crystallised from aqueous methyl alcohol in felted masses of needles melting sharply at 119.5' ; it seems to be far less soluble in cold ether than the preceding compound. C=82*0; H=75. 0.1497 gave 0.4484 CO, and 0-0996 H,O. C = 81.7 ; H = 7.4.C,,H,,ON requires C = 81.7 ; H = 7.5 per cent. Products from dl-Benzylmethylacetyl Chloride and Primary OT Secondary Bases. Having proved that evidence of asymmetry due to a carbon group is easily obtained with the aid of dl-benzylmethylacetyl chloride, we stcdied the products of the action of this compound on some primary and secondary bases. Methylaniline and the dl-chloride interact readily in ethereal solution. The resulting substituted amide remained in an oily condition for a, long time, but ultimately solidified completely; it was fractionally crystallised from warm light petroleum (b. p. 30-40') in which it was readily soluble, and was thus obtained in lustrous, trans- parent prisms melting sharply a t 54-55O; all the fractions had the same appearance, and except the last, which melted a few degrees below 54O, liquefied at the same temperature.Benxylmethylaceto- methylanilicle exists, therefore, in only one form ; it is readily soluble in all ordinary organic solvents. p-Toluidine and dl-benzylrnethylacetyl chloride interact vigorously in ethereal solution ; the benxylmthylaceto-p-tohidide, which is thus formed, solidified readily, and was fractionally crystallised from aqueous alcohol ; it separated in concentrically grouped needles melting at 130'. The very last mother liquors gave traces of a product melting about 2' lower than the rest, but there was no evidence of the existence of isomerides. Benzylmethylaceto-p-toluidide is readily soluble i n all ordinary organic solvents with the exception of light petroleum (b. p.30-40°), in which it dissolves somewhat sparingly. Benzylaniline and dl-benzglmethylacetyl chloride give a product which is at first rather difficult to obtain in a crystalline condition, but which afterwards crystallises easily from aqueous alcohol in ill-defined prisms; the compound was separated into a number of fractions, but the latter were all identical in appearance and melting point.446 ICTPPING AND SALWAY: ARRANGEMENT IN SPACE OF THE Benzylmethylaceto6en~~~an~lid~ is readily soluble in most organio 0.1452 gave 0.4457 CO, and 0.0921 H,O. solvents and melts, a t 69-70'. C = 83.7 ; H = 7.05. C,H,*CH,*CH(C~,)*CO.N(C~H~)*C,H, requires C = 83.9 ; H = 7.0 per Phenylhydrazine, when treated with the dl-chloride in ethereal solution, readily yielded ben~ylmethylacetophnylhydraxide, which crystallised from methyl alcohol in large, transparent, lustrous prisms melting at 116-117' and readily soluble in ethyl acetate ; fractional crystallisation failed to indicate the presence of an isomeride. cent.Products from d-Benaylmethylacetyl Chloride alzd Primary OT Secondwy Bases. For the reasons already stated, we next examined the products obtained with the aid of the optically active chloride ; in doing so, it was found that even when the acid chloride had the highest specific rotation previously observed, namely, [a], + 18' in ethereal solution (Kipping and Hunter, Trans., 1903, 83, lOOS), it still contained a small quantity of the I-isomeride, owing, doubtless, to the occurrence uf racemisation during its preparation, and gave rise to some of the dl-substituted amide; the presence of the latter was a serious dis- advantage because, although the main portions of the active and inactive amides could be separated without much difficulty, there still re- mained a considerable quantity of mixed products in which the required isomeric active compound might conceivably be concealed.I n the later and more important experiments, therefore, the active acid chloride was prepared from very carefully purified acid by a method which should exclude the possibility of racemisation (p. 447). p-Toluidine and d-benzylmet hylacetyl chloride gave a product which, when repeatedly fractionally crystallised from aqueous alcohol, was resolved into two compounds, one of which, however, was present in relatively small proportions; this was the amide of the dl-acid and melted at 130' (p.445). The main component, namely, d-benxyl- metiiylc6ceto-p-tol~idide, crystallised in large, flat, transparent, well- defined prisms melting a t 115-116' ; i t was readily soluble in cold alcohol and in ethyl acetate. 0.2364, dissolved in alcohol and the solution diluted to 25 c.c., gave, in a 200 mm. tube, a + 3.17', whence [a], + 167.6'. 0.5150, dissolved in chloroform and the solution diluted to 20 c.c., gave, in a 200 mm. tube, a + 5*67O, whence [a], + 110.1'. During the fractional crystallisation, no compound other than theGROUPS COMBINED WITH THE TERVALENT NITROGEN ATOM. 447 two already mentioned was obtained, and we concluded that optically active isomerides were not produced.The experiment was subse- quently repeated, using a sample of d-chloride free from the Z-iso- meride, and in t h i s case only one compound, namely, that melting at 115-1 1 6 O , was obtained. Benzylaniline and d-benzylmethyIacety1 chloride gave a product which was very difficult to crystallise even when the acid chloride had been prepared by the special method to be described later. This wae due to the fact that the product retained benzylaniline very tenaciously, and repeated washing with warm hydrochloric acid was required before the whole of this base was removed. Finally, however, the amide solidified to a crystalline mass, which was fractionally crystallised from light petroleum and from aqueous alcohol. The several fractions of d- benxyZmethyZacetobenzyZaniZide, c,H,.cH,~cH(cH,)~co~NH(c~H?,.~,H,, thus obtained melted at the same temperature (69-70°), which is also the melting point of the inactive amide already described (p. 446); moreover, a mixture of the crystals of the active and inactive amides also melted at 69-70', The specific rotation of the d-anilide was determined in chloroform solution : 0.3079, dissolved in chloroform, the solution diluted to 20 C.C.and examined in a 200 mm. tube, gave ~ + 0 * 2 7 ~ whence [uID+ 8 . P ; in methyl-alcoholic solution the specific rotation was [a], + 16'. Although only one active compound was isolated, and although there was no evidence of the formation of an isomeride in the above experiment, we are unable to state positively that only one substance was produced, as a portion of the product from the last mother liquors crystallised badly and had a brown colour, doubtless owing to atmo- spheric oxidation.Preparation of d-BenxyZmethyZacetyZ Chloride. Owing to the readiness with which the d-acid chloride is racemised under the influence of heat (Kipping and Hunter, Trans., Zoc. cit.), the compound prepared in the usual manner is almost invariably mixed with more or less of the Z-isomeride, the presence of which is very objectionable in such experiments as those already described, where it is a question of the existence of isomeric amides. In order to prepare the active compound in an optically pure condition, it is necessary to keep the temperature as low as possible and then to remove the phos- phorus oxychloride from the product without the application of heat. This can be accomplished by passing a current of dry air through the mixture under greatly reduced pressure, but the operation requires a considerable time before the last portions of phosphorus oxychloride448 KIPPING AND SALWAY: ARRANGEMENT IN SPACE OF THE are removed.The d-chloride thus prepared has a specific rotation in light petroleum [a], + 26.2, a value which is a trifle higher than that previously obtained (Zoc. cit.). This method of preparation does not remove every trace of phos- phorus compounds, for the acid chloride still gives a slight reaction with ammonium molybdate, probably due to the presence of phosphoric acid; this impurity, however, is of little consequence, as it is easily eliminated by washing with water, when the acid chloride is used for the preparation of a substituted nmide.Products from d-Benxylnzethy Zacetyl Chloride and Optically Active Buses. 1. l-PhenyZethylamine.-The active base, prepared by the method recently described (Kipping and Hunter, Trans., 1903, 83, 1147), was dissolved in dry ether and the active acid chloride added gradually, when a reaction immediately occurred, and the hydro- chloride separated in crystals. This salt was removed by filtra- tion, and the ethereal solution washed successively with dilute hydro- chloric acid, water, sodium carbonate, and water, by which treatment any remainicg hydrochloride, free base, or free acid is removed. After evaporating off the ether, the crude product,which immediately solidified, was then fractionally crystallised from benzene and thus separated into at least nine fractions, but one and the same compound melting a t 122' was obtained throughout, and no indication whatever of the presence of isomerides was observed.d-BenxyZmethylaceto-l-phen~Z~~hyZamide, C,H,*CH2*CH(CH,)*CO~NH.CK(CH{CH~)*C~H~, crystallises from benzene in beautiful, long, silky needles melting a t 122.5". It is readily soluble in ethyl acetate, chloroform, methyl alcohol, ether, or alcohol, but only sparingly so in light petroleum. 0,4315, dissolved in ether and diluted to 20 c.c., gave, in a 200 mm. tube, a + 0*37", whence [a], + 8.6". 2. l-MethyliLydrindc~rnine.-The reaction was conducted in dry ethereal solution, and since the product WRS not readily soluble in ether the solvent was allowed to evaporate and the residue washed at the pump with dilute hydrochloric acid, hot water, dilute sodium car- bonate, and hot water successively.The substance, which then melted at 150-152O, was now subjected to fractional crystallisation from benzene and separated into at least eight portions, but in this case, as before, only one compound was found to be present, although the last fractions had their melting point lowered slightly through the presence of a trace of unavoidable impurity. d-Benzylmetiiylace~o-1-Pnethylhydrindamide crystallises from benzene and alcohol in long, silky needles very similar to those of the acidGROUPS COMBINED WITH THE TERVALENT NITROGEN ATOM. 449 amide of I-phenylethylarnine and melting at 152"; i t is readily soluble in hot benzene, alcohol, chloroform, or ether.0.4071, dissolved in chloroform and diluted to 20 c.c., gave, in a 200 mm. tube, a - 1.08", whence [a], - 26.5". 3. 1-Menthylamine. -The optically pure base, prepared from its d-bromocamphorsulphonate (Tutin and Kipping, Trans., 1904, 85, 65) dissolved in ether, was added to an ethereal solution of d-benzylmethyl- acetyl chloride, when a vigorous reaction occurred ; the deposited menthylamine hydrochloride was separated by filtration and the ethereal solution washed with dilute hydrochloric acid, sodium carbon- ate, and water successively. The product was fractionally crystallised from benzene and separated into at least six fractions, but the first and last fractions were identical in crystalline form and melting point.d-BeIzxyZ~ethyZc~ceto-l-merzthyZamide cry stallises from benzene in long needles melting at 140'; it is readily soluble in all the ordinary organic solvents. tube, gave a + 0*58", whence [a], + 75". 4. d-€€$drindamine.-The reaction was conducted as in the case of the former active bases and the product fractionated from benzene. Six fractions were thus obtained, but no difference i n appearance or melting point could be observed on comparing the first and last deposits. d-BenxyZmethyZaceto-d-Aydrindumide cry stall ises from benzene in long, colourless needles melting a t 148-149"; it is readily soluble in methyl alcohol, chloroform, acetone or ethyl acetate ; moderately so in ether, and dissolves sparingly in light petroleum. 0,8580, dissolved in methyl alcohol and the solution diluted to 20 c.c., gave, in a 200 mm.tube, a - 1-33', whence [ aID - 15.5'. gave, in a 200 mm. tube, a + 0*76", whence [ u ] ~ + 6.1'. 0.7704, dissolved in chloroform and diluted to 20 C.C. in a 200 mm. 1.2482, dissolved in chloroform and the solution diluted to 20 c.c., Action Q#* dl-BcnzyZmethyZacetyZ Chloride on Benzoinoxime. Benzoinoxime was covered with dry ether or light petroleum and slowly treated with one molecular proportion of dZ-benzylmetbylacetyl chloride. No appreciable development of heat was observed, but the gradual progress of the reaction was indicated both by the conglomera- tion of the benzoinoxime into a united whole, and also by the slow evolution of hydrogen chloride. If the mixture is warmed to the boil- ing point of the solvent, the reaction is considerably accelerated, the450 KIPPINU AND SALWAP: ARRANGEMENT IN SPACE OF THE whole of the oxime dissolves, and a crystalline product is deposited in small quantities.The crystalline product was fractionally crystallised from alcohol ; the first deposit consisted of beautiful, long, lemon-yellow crystals which were evidently pure and meIted a t 96", being thus identified as benzil. The second deposit of crystals consisted of a mixture of small, colourless needles and of a white, fluffy, crystalline substance. The latter, which was immediately dissolved on washing with water, consisted of ammonium chloride. The residue, when crystallised, yielded colourless needles identified as benzoin (m. p. 134").The alcoholic mother liquors yielded further quantities of these three pro- ducts only. The ethereal solution gave, on evaporation, a yellow oil, which was washed with dilute aqueous sodium carbonate until the extracts were found to be free from acid. The substance thus removed was precipi- tated from the alkaline solution by the addition of dilute sulphuric acid, extracted with ether, and distilled under reduced pressure (1s mm.). It consisted almost entirely of benzylmethylacetic acid, boiling a t 164'. The ethereal solution, which had been extracted with sodium carbonate, gave, on evaporation, an oil, which had a strong odour of bitter almonds and which was distilled under atmospheric pressure, the fractions being collected at 179-190O and 190-200". The residue was then distilled under reduced pressure (17 mm.) and the distillate solidified to a yellow mass, which, when crystallised from alcohol, melted a t 95', and proved to be bend, the presence of which among the products of the reaction had been previously observed.The two fractions, boiling at 179-190' and 190--200' respectively, were too small for further fractional distillation, but as the former had the strong odour and boiling point of benzaldehyde, it was treated with phenyl- hydrazine in acetic acid solution, when benzaldehydephenylhydrazone (m. p. 154O) was obtained. The second fraction, boiling a t 190--200°, was found to be hydrolysed by prolonged boiling with an aqueous solution of sodium hydroxide, a copious evolution of ammonia taking place ; the alkaline solution, when neutralised, yielded a crystalline pre- cipitate, which was identified as benzoic acid.This fraction therefore consisted chiefly of benzonitrile. We have thus shown that the products of the interaction of benzoin- oxime and benzylmethylacetyl chloride are : (1 ) benzylmethylacetic acid ; (2) benzaldehyde ; (3) benzonitrile ; (4) benzoin ; (5) benzil ; (6) ammonium chloride ; and (7) hydrogen chloride. This decomposition of the molecule of benzoinoxime by the acid chloride is somewhat obscure, but apparently decomposition occurs in two different directions simultaneously. I n the first instance, the molecule is severed between the two central carbon atoms with theGROUPS COMBINED WITH THE TERVALENT NlTROGEN ATOM. 451 formation of benzaldehyde and its oxime, the latter of which by dehydration is converted into benzonitrila.This decomposition may be indicated thus : C,H,*cH(OrH)'i* $ *C,H, !OH; N .____..._ In the second place, the decomposition of the benzoinoxime takes place by the elimination of ammonia and the formation of benzil according to the equation : HON C6H5'cH(OH)*fl'C6H5 = C,H,'CO'CO*C,H, + NH, I n any case, the above decomposition of benzoinoxime at the ordinary temperature is a reaction of some interest, inasmuch as it affords one of the few known examples whereby the benzoincondensation of benzaldehyde is reversed. Zinin (Ber., 6, 1207), by repeated dis- tillation of benzoin, obtained a decomposition i n to benzaldehyde, benzil, deoxybenzoin, and water according to the equation : 3C6H,*CH(OH)*CO*C6H, = 2C6H5.CH0 + C,H,*CO*CO*C,H, + C6H;CH;CO*C6H5 + H,O ; whilst the same products were obtained by passing benzoin vapour through a red-hot tube (compare also Knoevenagel and Tomasczewski (Ber., 1903,36, 2829).Thinking that the decomposition of the benzoinoxime might be due t o the action of the hydrogen chloride which is liberated, experiments were made in which an excess of dry quinoline was previously mixed with the oxime; the presence of t h i s base, however, seemed to have little effect on the course of the reaction. It may be added that in many of the experiments the various products already enumerated were accompanied by very small quantities of a crystalline compound melting at 126O; this substance was not identified, but from its com- position and reactions it seemed to be the substituted amide, C6H,'CH(OH)'~H*CO*C6H5, produced by intramolecular change.C6H5*CH(OH)* 5 *(2,H5. N-OBz Benxoinoxime Benzoate, I n order to ascertain whether other acid chlorides would react normally with benzoinoxime giving acyl derivatives, or whether, like benzylmethylacetyl chloride, they would bring about a decomposition of the molecule, molecular proportions of benzoyl chloride and benzoin- oxime were allowed to interact in ethereal solution, A gradual452 KIPPING AND SALWAY: ARRANGEMENT IN SPACE OF THE evolution of hydrogen chloride occurred, but the reaction was not vigorous. After keeping for several hours a t the ordinary tempera- ture, the ethereal liquid was separated from the solid by filtration, The filtrate contained unchanged benzoyl chloride together with approximately equal proportions of benzoin and benzil.The solid product consisted chiefly of benzoinoxime benzoate, but it also con- tained a moderate quantity of benzoin and hydroxylamine hydrochloride. The presence of the latter was shown, after the removal of the benzoate by crystallisation, by evaporating the alcoholic mother liquors and washing the residue with water. The aqueous extract immediately reduced Fehling’s solution and also gave the reactions of a chloride. Benzoinoxime benzoate is easily purified by crystallisation from alcohol, in which it is only slightly soluble(approximate1y 1 part in 30 parts of boiling alcohol); it is thus obtained in leaf-like plates melting at 165-166’. 0-1540 gave 0.4301 CO, and 0.0729 H,O.0.4795 ,, 19.4 C.C. moist nitrogen a t 21° and 7603 mm. N = 4.6. Benzoinoxime benzoate is hydrolysed by acids and alkalis giving benzoin, benzoic acid, and hydroxylamine. It is moderately soluble in chloroform and ethyl acetate ; only slightly so in benzene. The foregoing method of preparing the benzoyl derivative leaves open several possibilities with regard to its constitution. Since benzoin- oxime contains two hydroxyl groups, it is uncertain whether the benzoyl group has entered the alcoholic or the oximic hydroxyl, or whether the benzoinoxime has undergone the Beckmann transforma- tion prior to benzoylation. The latter possibility, however, is shown to be untenable by a study of the reactions of the product. Since, on hydrolysis, it yields benzoin and hydroxylamine, it cannot have the constitution C,H,~CH(OH)*CO*NBz*C,H, or C,H,-CH( OH) * NBz* CO*C,H,, but may be either C=76.17 ; H=5*26. C,,H,,O,N requires C = 76.13 ; H = 5.14 ; N = 4.2 per cent.A decision between these two formulae has been rendered possible by the fact that we have prepared the compound having the latter constitution by benzoylating benzoin prior t o the introduction of the oxime group, a process which ensures the substitution of the hydrogen of the alcoholic hydroxyl group by the benzoyl radicle. I n order t o distinguish between the two isomerides, we have described them as benzoinoxime benzoate and benzoylbenzoinoxime respectively.GROUPS COMBINED WITH THE TERVALENT NITROGEN ATOM. 453 C,R,*CH(OBz)G *C,H, N*OH ' Benxo ylbenxoinoxirne, Benzoylbenzoin was prepared by heating molecular proportions of benzoin and benzoyl chloride at 150' for a n hour (Zinin, Eoc.cit.) ; the reaction was always accompanied by oxidation with the formation of benzil (about 10 per cent.). The benzoylbenzoin was boiled in aqueous alcoholic solution with twice the theoretical quantities of sodium acetate and hydroxylamine hydrochloride, the heating being continued during 6-12 hours; when the greater portion of the alcohol was then distilled off and the residue poured into water, the semi-solid product, on crystallisation from alcohol, deposited globular clusters of crystals melting at 148'. 0.2400 gave 0.1122 H,O and 0.6'704 CO,. 0.5295 ,, 20.3 C.C. moist nitrogen at 17.5" and 763.5 mm.N = 4.5. Benzoylbenzoinoxime is readily soluble in the ordinary organic C = 76.18 ; H=5*.19. C,,H1703N requires C = 76.1 3 ; H = 5.14 ; N = 4.2 per cent. solvents. Action of Benxylmethylcwetyl Chlorids on a-BenxaZdoxinae. When a-benzaldoxime, covered with light petroleum, is treated with benzylmethylacetyl chloride, the mixture becomes warm and hydrogen chloride is slowly evolved, while the transparent crystals of the oxime are converted into a fine, white powder. After leaving the mixture a t the ordinary temperature for several hours, the solid if separated by filtration and identified as benzamide. On evaporation, the light petroleum solution gave an oil which had a strong odour of bitter almonds; on distilling in steam, a portion passed over and was identified as benzonitrile.The residue consisted of benzylmethylacetic acid. I n one experiment, 10.7 grams of acid chloride and 7-1 grams of benzaldoxime were used, from which 5 grams of benzonitrile, 2 grams of benzamide, and 9 grams of benzylmethylacetic acid were obtained. I n a second experiment, only 0.5 gram of benzamide was produced from 8-3 grams of benzaldoxime, with a corresponding increase in the quantity of benzooitrile. Yh8 Resolution of dl-Bases with the Aid of an Active Acid Chloride. Action of d-Benxylmethykc~cetyl Chloride on dl-Eydrindamine. On slowly adding a dry ethereal solution of hydrindamine (2 mols.) to an ethereal solution of benzylmet hglacetyl chloride (1 mol.), pre- VOL. LXXXV. H H454 KIPI'ING AND SALWAY: ARRANGEMENT IN SPACE OF THE pared from the d-acid in the ordinary way, a vigorous reaction ensues with development of heat and precipitation of hydrindamine hydro- chloride, but after about half the base has been run in, the reaction moderates considerably.The solution, left overnight to ensure com- plete interaction, is then washed with water, the ether allowed to evaporate, and the crystalline product separated by filtration and washed with water until free from hydrochloride. On repeatedly crystallising the crude hydrindamide from methyl alcohol, the most sparingly soluble isomeride is obtained without much trouble ; this compound cry stallises from methyl alcohol in colourless needles or prisms melting at 148-149'; it is readily soluble in cold methyl alcohol, ethyl acetate, ether, and acetone, but practically insoluble, even in boiling water.This compound is lavorotatory in methyl- alcoholic solution, and is a derivative of d-hydrindamine, a s is proved by the fact that i t is identical with the d-benzylmethylaceto-d-hydrind- amide already described (p. 449) ; its specific rotation was determined with the following results : 0.5766, in 25 C.C. methyl alcohol and in a 200 mm. tube, gave a - 0*74*, whence [a],, - 1 6 O . 0.1411 gave 0,4232 CO, and 0.0952 H,O. C = 81.79 ; H = 7.5. 0.1419 ',, 0.4266 CO, ,, 0.1016 H,O. C=82*0 ; H-7.9. C,,H,,ON requires C = 81.7 ; H = 7.5 per cent. The various mother liquors remaining after the iso1:ition of the d-benzylmethylaceto-d-hydrindamide gave crystalline deposits, which, when submitted to fractional crystallisation, gave long, ill-defined needles melting a t 119-120' ; this substance is one of the externally compensated compounds identical with that prepared from dl-hydrind- amine and the dl-acid chloride (p. 444).The later crystalline deposits yielded a compound which crystallised from methyl aicohol iu well- defined prisms melting a t 11O-11lo; this is the isomeric externally compensated compound previously described. The mother liquors which then remained gave crystalline deposits of very indefinite melt- ing point, and from which, owing t o their very great solubility, it seemed impossible to isolate the last component of the original product, namely, the derivative of the other active base. In these circum- stances, the several most soluble deposits were collected, recrystallised until colourless, and then examined pol:irimetbrically in methyl alcohol solution, when it was proved that these fractions contain a compound which was dexti*o-rotatory, and which therefore represents the deri- vative of the other active base, namely, the I-isomeride. The acid chloride used in the above experiments had been distilled under reduced pressure, and contained a considerable quantity of the &compound ; the experiment was repeated, using the d-chloride pre-GROUPS COMBINED WITH THE TERVALENT NITROGEN ATOM.455 pared as described in this paper (p. 447). I n this case, the active amide melting at 148-149" was very easily isolated by two or three crystallisations from benzene, but again i t was found almost imposeible t o obtain any quantity of a pure preparation of the isomeride from the mother liquors, although a few crystals were obtained melting a t 113". Attempts t o hydrolyss the amide melting at 148-149" were unsuccessful ; it seems to be quite stable towards boiling concentrated aqueous or alcoholic potash or aqueous baryta, and is only very slowly hydrolysed by boiling hydrochloric or dilute sulphuric acids, giving mere traces of ammonia and of indene; more concentrated sulphuric acid at 150' is much more rapid in its action, but ammonia is the only basic product ; the liberated d-benzylmethylacetic acid is still optically active, thus indicating the stability of the asymmetric carbon group in this compound. The active hydrindamide, when heated in sealed tubes with alcoholic potash, hydrochloric acid, or sulphuric acid, or other hydrolysing agents, was decomposed into indene and ammonia, Act ion of d-Benxylmethy Zucety Z Chloride on dl-a-P?LelzyZetlzyZam~ne. The optically pure chloride of the d-acid was treated in dry ethereal solution with excess of freshly distilled base in the same solvent ; a vigorous reaction took place, heat being developed and a bulky pre- cipitate of the hydrochloride being formed. The ether was allowed t o evaporate, and the residue heated with a little water t o dissolve the salt formed, when a white, slightly viscid solid remained, which melted a t 95-100". The product was first crystallised from methyl alcohol ; the sparingly soluble fraction, which separated in thick prisms, melted at 118-1 19' ; this portion was first crystallised from light petroleum (b. p. 60-SO"), in which it was very sparingly soluble, and then from benzene; the fine, long needles thus obtained melted a t 119', and mere identical with those of the d- benzylmethylace to-Z-phenyle thyl- amide previously described (p. 448). 0.2925, dissolved in 25 C.C. ether gave, in a 200 mm. tube, a + 0*24', whence [a], + 10.3". The isomeric derivative of d-phenylethylamine was not isolated from the mother liquors, since it was found impossible to regain the active bases by the hydrolysis of their amides. A part of the expense incurred in this investigation has been met by a grant kindly awarded by the Government Grant Committee the Royal Society. UNIVERSITY COLLEGE, NOTTINGHAM. of

ISSN:0368-1645    

DOI:10.1039/CT9048500438

出版商:RSC    

年代:1904

数据来源: RSC

摘要:

456 RUHEMANN AND WATSON: CONTRIBUTIONS TO THE L.-Contributions to the KrLowled.qe of the @-Diketones. By SIEGFRIED RUHEMANN and EDWIN ROY WATSON. SOME years ago, J. Wislicenus (Annulen, 1899, 308, 219) published the results of the work undertaken by him in conjunction with Liiwen- heim, Schmidt, and Wells on the interaction of alcoholic potash with benzylideneacetophenone dibromide. They found that the compound thus formed had the formula CI6Hl2O2, and they regarded i t as the tautomeride of dibenzoylmethane, which had been first obtained by v. Baeyer and Perkin (Bey., 1883, 16, 2134) by boiling ethyl dibenzoylacetate with water, and afterwards by Claisen (Annalen, 1896, 291, 52) by the action of sodium ethoxide or metallic sodium on a mixture of ethyl benzoate and acetophenone. The two tauto- rneric forms of dibenzoylmethane could be represented by the following formulae : C,H,*C( OH):CH*CO*CGH, and CGH,*CO-CH,oCO*C,H5. Wislicenus is inclined to regard his compound as the diketonic form, and the old dibenzoylmethane as the enol-ketonic form, because, whilst the latter dissolves in alkali with the greatest ease and yields with ferric chloride a violet-red coloration and a greyish-green pre- cipitate with alcoholic copper acetate solution, the former is insoluble in alkali and neither gives at once a coloration with ferric chloride, nor is precipitated by copper acetate.That Wislicenus’s compound was a tautomeride of the long-known dibenzoylmethane seemed to follow from the fact that it gave the same compounds on treatment with bromine, hydroxylamine, phenylhydr- azine, hydrazine, or semicarbazide.W islicenus further showed that, although his compound remained unchanged on distillation or on boil- ing with acetic acid, it was transformed into the old dibenzoyl- methane under the influence of mineral acids. He therefore succeeded in effecting a transformation of his compound into the other, although the reverse change could not be accomplished. That the transformation of the one form into the other is not brought about either by boiling or under the influence of alkali seems t o be incompatible with our present knowledge of tautomeric com- pounds. These difficulties seem to have been recognised by Wislicenus himself, for he suggested the possibility of representing his compound as benzoylphenylethylene oxide, C,H,*CO*CH~CH*C,H,.‘0’ These discrepancies and the desire to prepare ketones of the acetyl- ene series in ordsr to examine wbether they are transformed by the action of bases into heterocyclic compounds, similar to those formedKNOWLEDGE OF THE P-DIKETONES. 457 from the esters of the acetylene seiies (Ruhemann and Stapleton, Trans., 1900, 67, 239), have induced us to repeat Wislicenus's experiments. Such ketones of the acetylene series have indeed been obtained by Nef (Annalen, 1899, 308, 264) from the sodium compounds of the hydrocarbons of the acetylene series under the influence of acid chlorides; yet it seemed to us possible to prepare them from the olefinic ketones, just as Claisen has prepared acetylenic aldehydes from their olefinic analogues (Ber., 1903, 36, 3664).I n following Wislicenus's directions, we obtained by the action of alcoholic potash on benzylideneacetophenone dibromide a compound with precisely the same properties as those recorded by this author for his dibenzoylmethane, but our analytical results differed considerably from the numbers given by Wislicenus, and agree, not with the formula Cl,H1202 for dibenzoylmethane, but with C,7H1602. This formula, which was, moreover, verified by a molecular weight determin- ation, would correspond with the following constitution, C6H5*C( O*C,H,) :CH*CO*C,H,, according to which the substance appears to be the ethyl ether of dibenxoylmethane, and would therefore be analogous to the corre- sponding esters which have lately been obtained by Lees (Trans., 1903, 83, 145).This constitution is supported by the fact that the compound CI7Hl6O2, on heating with hydrochloric acid, is decomposed, giving acetophenone, benzoic acid, and ethyl chloride. This reaction, therefore, takes place in accordance with the following equation : C,H,*C(O*C,H5):CH*C'O*C,H5 + HCI + H20 = C6H,*C02H + C,H,CI + CH,*CO*C6H5. Lowenheim, who effected the decomposition of the compound by means of concentrated caustic potash and obtained acetophenone and benzoic acid, did not observe the formation of alcohol. The correctness of the formula, C17H1602, for this compound is, moreover, supported by the fact that p-nitrobenzylideneacetophenone dibromide, on treatment with alcoholic potash, yields the correspond- ing nitro-derivative, Cl7Hl,(NO2)O,, which, therefore, is to be regarded as the ethyl ether of p-nitrodibenzoylmethane, ~0,°C6H~*~(O*C2H5):CH~C~*~6H,. As regards the formation of the ethyl ether of dibenzoylmethane from benzylideneacetophenone dibromide, there cannot be any doubt that the first action of the alcoholic potash is to remove 1 mol.of hydro- gen bromide and to yield the bromobenzylideneacetophenone, C6H5* CBr : CH'Co *C6H5,458 RUHEMANN AND WATSON: CONTRIBUTIONS TO THE a compound obtained by Schmidt (Wislicenus, Zoc. cit.) by the action of potassium acetate on the dibromide. We have obtained a similar compound, namely, bromobenzylideneacetone, C6H, *C Br: CH*CO*CH,, by the action of alcoholic potash in the cold on the benzylidene- acetone dibromide. The bromobenzylideneacetophenone is subsequently transformed into the ethyl ether of dibenzoylmethane, either by the further removal of hydrogen bromide and subsequent union of alcohol with the first formed benzoylphenylacetglene, or by the direct replace- ment of the bromine atom by the ethoxyl group.This constitution shows why the compound is insoluble in aqueous caustic potash and does not give a coloration with ferric chloride or a precipitate with copper acetate. It further explains why the cow- pound can be readily transformed into dibenzoylmethaue, and why the reverse change cannot be directly effected. Moreover, all the reactions of this compound, as described by Wislicenus, can be readily explained on this view of its constitution. For example, an ay-diphenylisooxazole is formed with hydroxylamine, C,H,.C(OC,H,):~H*CO*C,H, + NH,*OH = C, H,* OH + H,O + The formation of ay-diphenylisooxazole, as shown by Wislicenus (toc.cit.), is preceded by that of a compound which he regarded as an oxime of dibenzoylmethane. The fact, however, that this substance, which is produced also from v. Biteyer and Perkin's dibenxoylmethane, does not yield a coloration with ferric chloride, is not in harmony with the formula C,H,*C(OH):CH*C(C,H,):NOH. After having proved t h a t Wislicenus's dibenzoylmethane is ethoxybenzylideneacetophenone, its reaction with hydroxylamine may be expressed as follows : C,H,*C(O*C,H5):CH*CO*C,H5 + NH,OH = C,H5* C( NH*OH) :CK*CO* C,H, + C,H,O. The compound would thus appear to be a hydroxylamine derivative and not an oxime.This constitution agrees with the fact that the compound does not give a coloration with ferric chloride ; it explains, also, why the substance, although unchanged by neutral solvents, is readily transformed into ay-diphenylisooxazole under the influence of alkalis. It is thus isomerised to C,H,*C( INOH) 'CH,*CO*C,H,, which subsequently loses water to form the isooxazole derivative. A ready explanation is also afforded of the action of bromine in form- ing the monobromo- and dibromo-benzoylmethane compounds, previouslyKXOWLEDGE OF THE P-DIKETONES 459 prepared by Xeufville and v. Pechmann (Ber., 1890, 23, 3377) from dibenzoylmethane, C,H,*C(O*C,H,):CH*CO*C,H, + Br, = C,H,*CO*CHBr*CO*C,H5 + C,H,Br. In connection with this investigation, we record the result of our experiments on the behaviour of bases towards olefinic diketones, an account of the action of benzamidine on these diketones having been lately given (Trans., 1903, 83, 1371).The results obtained induced us to examine also the action of other bases on these diketones. We find that ammonia readily reacts with benzylidenencetylacetone, yield- ing a compound having the formula C,,H,,ON,, the formation of which may be thus represented : 2C,H5*CH:C(CO*CH3), + ZNH, = BC,H,*CH(NH,)*CH(CO*CH,), = CH,(CO*CH,), + C,,H,,ON,. The constitution of the product should probably be expressed as %allows : according to which it is to be regarded as acetyldiphenylmethyltetra- hydropyrimidine. This constitution is in agreement with the fact that the compound, although readily decomposed by dilute mineral acids with the formation of benzsldehyde, is yet comparatively stable towards alkalis.On heating with alkali, no ammonia is evolved, but on removing the acetyl group a compound is formed which is probably the corresponding diphenylmethyltetrahydropyrimidine. One of us (Ruhemann, Trans., 1903, 83, 378) has shown that by the action of ammonia on ethyl benzylideneacetoacetate, a compound was formed which was regarded as benzylideneaminoacetone, C,,H,,ON. On account of the ease with which this substance is decomposed, the above formula could not be verified by a molecular weight determina- tion. The behaviour of this compound is analogous to that of the substance obtained from benzylideneacetylacetone, and this makes it necessary to assign to the two compounds similar constitutions.Such considerations lead us to double the molecular weight of the compound obtained from ethyl benzylideneacetoacetate, and to regard it as ethyl diphenylmethyltetrahydropyrimidinecarboxylate. Its formation may then be expressed by the following equation : 2C,H,*CH:C(CO*CH3)C0,*C,H, + ZKH, = H,O + CH,*CO*CH2*C0,*C2H, + C,K,*Cy* CH( CO,*C,H,)* 8 *CH, NH-CH(C,H,) *K460 RUHEMANN AND WATSON: COXTRIBUTIONS TO THE These compounds are still under examination. The assumption that the first step in the reaction of ammonia on benzylideneacety Iacetone leads to the formation of an additive pro- duct, as indicated in the above equation, is rendered probable by the action of aniline on benzylideneacety lacetone. These compounds interact at the ordinary temperature to form the additive product, a-anilino-PP-diacetyl-u-phenylethane, C,H,-CH( NH*C,H,)*CH( COCH,),.This compound, however, on heating a t looo, decomposes with the formation of benzylideneaniline and acetylacetone according to the equation : C,H5*CH(NH*C,H,)*CH(C0*CHJ2 = C,H,*CH:N*C,H, + CH,(CO*CH,),. Phenylhydrazine reacts with benzylideneacetylacetone as well as with benzylidenebenzoylacetone with development of heat, but the additive products, which may be assumed to be first formed, cannot be isolated, since, in both cases, they at once decompose forming benz- aldeh ydephenylhydrazone. The action of semicarbazide on benzylideneacetylacetone and benzyl- idenebenzoylacetone is of especial interest ; in both cases, the reaction takes place between 1 mol.of the olefinic diketone and 1 mol. of the semicarbazide ; water (1 mol.) is eliminated, and compounds are obtained having the formulae C,,H,,O,N, and C,,Hl7O2N, respectively. These compounds might be regarded as semicarbazones, C,H,~CH*C(CO~CH,)~C(CH,):NeNHoCO.NH, and C6H,*CH:C(CO*C6H,)*C(CH,):N*NH*CO*NH2, but taking into consideration the behaviour of other bases towards the olefinic diketones, there remains the alternative view that, in these cases also, additive products are first formed, which subsequently lose water to form cyclic compounds, for example : C,H,*CH:C(CO*CH,), + NH,*NH*CO*NH, = C,H,*CH (NH*NH*CO*NH,)*CH( CO*CH,), = One of us has recently shown (Zoc. cit.) that benzamidine also reacts with benzylidenebenzoylacetone forming an additive product.The fact that this compound, on heating with hydrochloric acid, decomposes and yields dibenzamide led to the view that the constitution of this compound should be represented by the formula : NH:C(C,H,)*NH*C(C,H,)iOH).C(CO*CH,):CH.C,H,.KNOWLEDGE OF THE P-DIKETONES. 461 This constitution, however, would not be in harmony with the results recorded in this paper, according to which the bases react additively with the olefinic diketones in such a way as t o destroy the double linking. I n the light of these researches, the above compound must be repre- sented thus : C,H,*CH[NH*C(C,H,):NH]*CH( CO*C6H,)*CO~CH,, and this formula would also readily explain the formation of dibenz- amide. Under the influence of hydrochloric acid, decomposition takes place, as indicated by the equation : C,H,*CH[NH* C( C,H,): N H ]*CH( CO* C6H,)*CO*CH, = C,H,*CHO + NH,-C(C,H,):NH + CH,*CO*CH3 + C,H,*CO,H.The benzamidine and benzoic acid in turn react in the presence of mineral acid to yield dibenzamide, thus : C,H,*CO,H + NH,*C(C,H,):NH + HC1= (C6H,*C0)2NH + NH,CI. The reaction of benzamidine on benzylideneacetylacetone, which, as has been shown lately (Zoc. cit.), yields diphenylmethp Idihydropyrimidine, may well be explained in a similar manner by assuming that first the additive product, C,H5*CH[NH*C(C,H,):NH]CH(C0.CH,),, is formed, which then condenses to the hydropyrimidine. The constitution of the products formed by the interaction of benzylideneacetylacetone and benzylidenebenzoylacetone with semi- carbazide has not yet been definitely fixed, but without considering t h i s problem the general hypothesis may be deduced that, in the first place, ammonia and organic bases, on reacting with the olefinic diketones, destroy the ethylene linking to form additive compounds.This assumption may be extended to explain the behaviour of bases towards ethyl dicarboxyglutaconate and its derivatives. As shown by one of us and his pupils, ethyl benzyldicarboxyglutaconate, on treatment with ammonia, yields ethyl benzylmnlonate and ethyl aminomethylenemalonate. I n these cases, also, it may be supposed that the decomposition is preceded by the formation of additive products, which subsequently break up in the manner indicated by the following equation : WO,Et), CH(C0 Et) C H < q c ~ , p t ) , + NH,R’ = CH(NHR’)<CqCo;EL); = C HR(CO,Et)? + NHR*CH:C( CO,Et),462 RUHEMANN AND WATSON: CONTRIBUTIONS TO THE E X P E RIME N T A L.Action of Alcoholic Potcd on Benxylideneacetophenone Dibromide. Benzy lideneacetophenone is readily propared by Claisen's method (Ber., 1887, 20, 657) by the action of sodium methoxide in methyl- alcoholic solution on a mixture of benzaldehyde and acetophenone. Its transformation into the dibromide, CgH~*CHBr*CHBr*CO0C,HS, is best effected by dissolving the olefinic kecone in carbon disulphide instead of chloroform, as recommended by Wislicenus (Zoc. cit.), when almost the total quantity of the dibromide separates out ; i t dissolt-es in boiling alcohol with difficulty and, on cooling, crystallises in colourless needles which melt a t 158" (according to Wislicenus a t 156*5-157*5c).0.2078 gave 0.371 1 GO, and 0.0642 H,O. C = 48.70 ; H = 3.43. 0.2017 ,, 0.3618 CO, ,, 0.0624 H,O. C = 48-93 ; H = 3.43. 0,2371 ,, 0.2406 AgBr. Br=43*18. C,,H,,OBr, requires C = 48*91 ; H = 3.26 ; Br = 43.47 per cent. We have pointed out in the introduction that the dibromide reacts with alcoholic potash to yield the ethyl ether of dibenzoylmethane, c,H,*c( 0 C, H,): CH-CO*C,H,, instead of diben zoy lmet hane, C,H,*CO*CH,*CO*C,H,, as stated by Wislicenus. Our mode of procedure is practically the same as employed by this author : caustic potash (2 mols.) dissolved i n absolute alcohol is gradually added to the hot solution of the dibromide (1 mol.) in absolute alcohol, and the reaction is completed by digesting the mixture at 100" for about an hour.The alcohol is then distilled from the water-bath as far as possible, the residue treated with water and extracted with ether. After drying the ethereal solution with calcium chloride and evaporating off the ether, an oil is left which distils a t 212' under 10 mm. pressure. The yellowish, oily distillate, when left for a short time, solidifies almost completely. The solid is only sparingly soluble in cold light petroleum (b. p. 50--80"), but readily dissolves in the hot solvent. It is readily soluble in hot alcohol and crystallises from either solvent in colourless prisms which melt at 77-78" and have all the properties recorded by Wislicenus for his '' dibenzoylmethane." 0.2037 gave 0.6020 CO, and 0*1190 H,O.C = 80.60 ; K = 6.49. 0.1925 ,, 0.5695 CO, ,, 0.1120 H,O. C = 80.68 ; H= 6.46. 4.1818 ,, 0.5397 CO, ,, 0.1052 H,O. C= 80.96 ; H = 6.4%. C17Hl6O2 requires C =r 80.95 ; H = 6.35 per cent.KNOWLEDGE OF THE 6-DIKETONES. 463 These numbers differ considerably from those given by J. Wislicenixs, which show a close agreement with the percentage results required for dibenzoylmethane, C15H1202, namely, C = 80.36 ; H = 5.36 per cent. The formula C17H,,02 agrees also with the result of the molecular weight determination of the compound by the freezing point method. 0.405 in 22.49 benzene gave At - 0.105°. C17Hl,0, requires 11. W. = 252. The constitution C,H,.C(O*C,H,):CH*CO*C,HS follows from the decomposition which the compound undergoes when heated with con- centrated hydrochloric acid in a closed tube at 150-160° for two hours.It thus furnishes ethyl chloride, acetophenone, and benzoic acid. The formation of ethyl chloride is sufficiently proved by the fact that oa opening the cold tube hardly any pressure is noticed, but on warming, a gas escapes which burns with a green flame. The other products of decomposition have been isolated and identified as acetophenone and benzoic acid by their properties and by analysis. M. W. = 256. Action of AkohoZic Potash on p-NitrobenxyZide~ecccetop~~none Bibromide. The action of alcoholic potash on benzylideneacetophenone dibroxnide is analogous to the behaviour of the same reagent towards thep-nitro- benzylideneacetophenone dibromide, the latter reaction giving rise to the ethyl ether of p-nitrodibenzoylmethane : 1 4 N02*C,H,*C( O*C,H,):CH*CO* C,H,.p-Nitrobenzylideneacetophenone is formed by adding dilute caustic soda to an alcoholic solution of pnitrobenzaldehyde and acetophenone. It was thus prepared by Sorge (BeF., 1902, 35, lOSS), who also trans- formed it into the dibromide. W e find that the union of bromine with the olefinic ketone takes place more readily than stated by this investigator, and t h a t it is completed when the solution of both reagents in chloroform has been left overnight. The dibromide, which separates out on concentrating the solution, has been once re- crystallised from chloroform, when it is obtained in yellow needles. On gradually adding the exact quantity of alcoholic potash (2 mols.) dissolved in absolute alcohol t o the hot alcoholic solution of p-nitro- benzylideneacetophenone dibromide (1 mol.), potassium bromide separates out and the solution turns deep red.After digesting the mixture for an hour at looo, the alcohol is distilled from the water- bath and the residue extracted with ether. On evaporating off the ether, a dark oil is left behind which partially solidifies in the course of 2-3 days; the solid is collected at the pump and dissolved in hot dilute alcohol, from which solvent i t separates in yellow needles which,464 RUHEMANN AND WATSON: CONTRIBUTIONS TO TEE after two crystallisations from the same solvent, melt constantly a$ 89-90". 0,1993 gave 0.5018 CO, and 0.0923 H,O. 0.1965 ,, 8.4 C.C. moist nitrogen a t 17" and 744.5 mm. N = 4.85.C= 68.62 ; H= 5.14. 0.1959 ,, 8.4 C.C. 9 , at 19" and 755.5 mm. N=4*89. CI7H,,O,N requires C = 68.68 ; H = 5.05 ; N = 4.71 per cent. p-Nitrodibenzoylmethane is readily soluble in alcohol, chloroform, or ether and insoluble in cold potash, but on warming with the alkali, however, decomposition takes place, and the odour of acetophenone is perceptible. Action of Alcoholic Potash on the Benxylideneacetone Dibromide. In preparing benzylideneacetone, we have followed the directions given by Claisen and Ponder (*4nnalen, 1884, 223, 139). This compound readily takes up bromine and yields the dibromide, C6H,*CHBr*CHBr*CO*CH,, but for this purpose we find it advisable to employ carbon disulphide rather than chloroform as a solvent (compare Claisen and ClaparBde, Ber., 1881, 14, 2461), since the yield is better and the product is at once obtained in a pure state. On adding caustic potash (2 mols.), dissolved in absolute alcohol, to the warm alcoholic solution of the dibromide (1 mol.), potassium bromide separates, and the solution turns red without further heating.The alcohol was not distilled off, as in the previous cases, but the solution was diluted with water and extracted with ether. After removing the ether, a yellow oil was obtained boiling at 150-151O and 169--170' under 10 and 20 mm. pressure respectively. This compound was bromobenzylideneacetone, C6H5*CBr:CH*CO* CH,t which, however, as indicated by the analysis, was not quite pure, but probably contained a little of the ethyl ether of benzoylacetyl- methane. 0.2063 gave 0.4023 CO, and 0.0839 H,O.0.2264 ,, 0.1775 AgBr. Br=33.39. C,,H,OBr requires C = 53.33 ; H = 4.0 ; Br = 35.55 per cent. On adding phenylhydrazine to the alcoholic solution of bromo- benzylideneacetone and slightly warming, a crystalline precipitate slowly separates, consisting of the phenylhydrazone, C6H,*CBr:CH*C( CH,):N=NH*C,H,. This substance crystallises from alcohol in yellow, glistening plates which melt and decompose at 97". N=8*90. C =53% ; H= 4.51. 0-1831 gave 14.2 C.C. moist nitrogen at 20" and 763 mm. C,,H,,N,Br requires N = 8-89 per cent,KNOWLEDGE OF THE P-DIKETONES. 465; Bromobenzylideneacetone is analogous t o the bromobenzylidene- acetophenone, C6H CBr: CH* co C6H5, which Schmidt (loc. c k ) obtained by the action of potassium acetate on benzylideneacetophenone dibromide. We have as yet not heated the bromobenzylideneacetone with alcoholic patash, but we propose to carry out this experiment with the view of examining whether in this case the corresponding ethyl ether, C,H,- C( ODC2H,) :CH*CO*C H,, of ace tyl benzoylmethane is produced.Action of Ammonia and Ogganic .Bccses on OleJinic Biketones. One of us has shown lately (Trans., 1903, 83, 1371) that, in the first place additive products are formed by the action of benzamidine on these diketones, and such a product has been isolated on treat- ing benzamidine with benzylidenebenzoylacetone. We have already stated (p. 459) t h a t other bases form with these diketones similar additive products, which, however, are unstable and readily give rise to other compounds.Action of Ammonia on Benz~lideneacetylacetone and Bensylidenebenxoyl acetone. The solution of benzylideneacetylacetone in excess of alcoholic ammonia, when left overnight, deposits colourless plates. These are insoluble in ether, but readily dissolve in hot alcohol and melt somewhat indefinitely at 147'. 0.2010 gave 0.5739 CO, and 0.1248 H,O. 0.1844 ,, 15.4 C.C. moist nitrogen at 12' and 749 mm. N = 9.71. 0.1865 ,, 15.4c.c. ,, ,, ,, 1 5 O and 755 mm. N = 9.66. C,,H,,ON, requireg C = 78-08 ; H = 6.85 ; N = 9.59 per cent. With regard to the constitution of this compound, we have already (p. 459) expressed the view that it is acetyldiphenyl- methyltetrahydropyrimidine. It dissolves in cold dilute mineral acids, but on warming the solution, decomposition takes place, which is indicated by the formation of benzaldehyde.The com- pound is mor0 stable towards alkalis for, on boiling with alcoholic potash, no odour of benzaldehyde is perceptible, but it loses the acetyl group and yields a substance which is most probably diphenyl- methyltetrahydropyrimidine. This product is isolated by distilling off the alcohol from the water-bath and adding water to the residue; a yellow oil is thus precipitated, whilst the aqueouc solution contains acetic acid, which has been identified by its general reactions. The oil, which rapidly solidifies and crystallises from dilute alcohol in yellow leaflets, is still under examination. C = 77.87 ; H = 6.89.466 CONTRIBCTTIOKS TO THE KNOWLEDGE OF THE B-DIKETONES. The action of alcoholic ammonia on benzylidenebenzoylacetone differs from the preceding condensation.I n this case, the solution turned yellow, but did not deposit a solid, even after a long time. On adding water to the solution, a viscous product was precipitated which could not be induced to crystallise. Action of Aniline on Benx?/lidenecccetylc~cetone. A mixture of equal quantities of benzylideneacetylacetone and aniline, when left for a short time, develops heat and becomes solid. After washing with a little alcohol, the product is dissolved in hot dilute alcohol ; the solution, on cooling, deposits colourless needles which melt a t 113'. 0.2010 gave 0.5646 (20, and 0.1248 H,O. C=76*60. H=6*89. 0,2367 ,, 11 C.C. moist nitrogen a t 22' and 732 mm. N = 5-04, 0.2384 ,, 11 C.C.,, ,, ,, 21' and 733 mm. N=5*03. C18H190,N requires C = 76.86 ; H = 6.76 ; N = 4.98 per cent. This additive compound, C,H,*CH(NH*C,H,)*CH(CO*CH,),, is somewhat unstable, and when heated on the water-bath it readily decomposes into acetylacetone and benzylideneaniline. This mixture, on cooling, partially solidifies ; the adhering oil is pressed out from the solid, and the latter dissolved in hot dilute alcohol. On cooling, a white emulsion is produced in the solution, which gradually deposits a compound crystallising in colourless plates melting at 50-51". This compound was identified as benzylideneaniline by the foregoing properties and by analysis. The additive compound is also formed on heating the mixture of aniline and benzylideneacetylacetone a t 100".Action of Phenylhydraxine on Benxylideneacetplacetone and Benxylidenc- benxoylacetone. On mixing benzylideneacetylacetone (3 grams) and phenylhydrazine (2 grams), heat was developed and a semi-solid product results, which crystallised from dilute alcohol in almost colourless needles melting at 155-156', and was identified by these properties and by analysis as benzaldehgdephenylhydrazone. Benzylidenebanzoylacetone reacts with phenylhydrazine in a similar manner ; an alcoholic solution of these reagents, when boiled and subsequently cooled, deposits needles melting at 155-156". 0.2069 gave 0.6010 CO, and 0.1165 H,O. 0.2032 ,, 25.5 C.C. moist nitrogen a t 16Oand 743 mm. N= 14.29. C = 79.26 ; H = 6-35. CI3Hl2K2 requires C = 79.59 ; H = 6.12 ; K = 14.28 per cent.THE FORMATION OF PERIODIDES IN ORGANIC SOLVENTS. 467 Action of Xesnicarbaxide on Renxylid~necccetylc~cetone and on Benxylidelzc- benxoy kcmetone. Semicarbazide hydrochloride (1 mol.), dissolved in water, is mixed with a solution of sodium (1 atom) in alcohol, and then treated with benzylideneacetylacetone (1 mol.) ; heat is developed, and after a short time a precipitate is formed which dissolves in alcohol with difficulty and, on cooling, crystallises in yellow plates, which melt at 210' with evolution of gas. 0.2030 gave 0.4728 CO, and 0.1147 H,O. C = 63.52 ; H = 6*2i. 0.1604 ,, 24-5 C.C. moist nitrogen at 18Oand746.5 mm. N = 17-31. C,,H,,O,N, requires C = 63.67 ; H = 6-12 ; N = 17-14 per cent. Benzylidenebenzoylacetone reacts with semicarbazide under the same conditions. The white, crystalline precipitate which separates is insoluble in water, sparingly soluble in boiling alcohol, and melts at 230' with evolution of gas. For analysis, the compound was washed successively with water and alcohol. 0.2013 gave 0.5174 GO, and 0,1034 H,O. 0.1679 ,, 19.5 C.C. moist nitrogen at 9'and 744.5 mm. N = 13.68. Further experiments are required in order to decide whether these substances are cyclic compounds, as indicated in the introduction, or whether they are mono-semicarbazones of benzylideneacety lacetone and benzylidenebenzoylacetone, but at present we are inclined to hold the latter view. C = 70.10 ; H = 5-70. C,,H170,S, requires C = 70.35 ; H = 5-53 ; N = 13.68 per cent. GOWILLE AND CAIUS COLLEGE, CAMBRIDGE.

ISSN:0368-1645    

DOI:10.1039/CT9048500456

出版商:RSC    

年代:1904

数据来源: RSC