Methylcyclohexane-1 : 3-dione, previously imperfectly characterised, has been made by three methods. Monoenol ethers are prepared from several cyclohexane-1 : 3-diones and (a) ethanol in presence of toluene-psulphonic acid, (b) ethyl iodide and sodium ethoxide, or (c) ethyl orthoformate and an acid catalyst. Method (a) is not applicable to %methyl-diones, and (b) is useful only with 2-methyl-diones. Method (c) is generally applicable, but under certain conditions gives complex compounds formed by the coupling of three dione molecules through a CH group.DURING experiments directed towards the synthesis of steroids and other natural have examined the preparation of certain cyclohexane-1 : 3-diones (3-hydroxycyclohex-2-enones) and their enol ethers.cycZoHexane-1 : 3-dione was most easily made by hydrogenation of resorcinol in alkaline solution, but conditions milder than those hitherto recommended were necessary to prevent further reduction. With methyl iodide and methanolic sodium methoxide it afforded the 2-methyl compound, but the maximum yield was only 58% (cf. Friedman and Robinson, Chem. and I d . , 1951, 777) ; and this was not improved by methylation in aqueous sodium hydroxide with a large excess of methyl iodide, partial ring scission to 5-ketohexanoic acid also occurring. Further methylation of 2-methylcyclohexane-1 : 3dione (in methanol) gave the enol ether ; the 2 : 2-dimethyl compound could not be isolated. In contrast, 5 : 5-dimethylcyclohexane-1 : 3-dione always gave a mixture of the 2-methyl and the 2 : 2-dimethyl derivative, as reported by Desai (J., 1932Desai (J., , 1079 and HirsjaM (Chem. A h . , 1948,42,2936). Gilling (J., 1913Gilling (J., , 2029 described the reduction of 4-methylresorcinol ; the product was believed to be 4-methylcyclohexane-1 : 3-dione, but was not satisfactorily characterised. We confirmed the formation of the 4-methyl-dione, a syrup, which rapidly oxidised in air to (crystalline) a-methylglutaric acid. The dione was characterised as the methylenebisderivative, and was synthesised by two further, interdependent methods. Methyl 4-carbe t hoxy-5-keto-2-met hylhexanoat e (from ethyl ace t oacet at e, methyl a-me t hylacrylate, and sodium ethoxide) on successive treatment with sodium ethoxide and alcoholic potassium hydroxide yielded a separable mixture of a-methylgutaric acid and the 4-methyl-dione, in proportions which varied widely in different experiments (Table 1).

A wheat-straw hemicellulose, containing uronic acid residues (ca. 3%), gave on hydrolysis xylose and arabinose in the ratio of 11 : 1. Hydrolysis of the methylated polysaccahride yielded 2 : 3 : 5-tri-O-methyl-~-arabinose, 2 : 3 : 4-tri-O-methyl-~-xylose, 2 : 3-di-O-methyl-~-xylose, %O-rnethyl-~-xylose, and 2 : 3 : 4-tri-U-methyl-~-glucuronic acid, together with some 3-Umethyl-2-0-(2 : 3 : 4-tri-~-methy~-~-g~ucuronosy~)-~-xy~ose. The structure of the polysaccharide is discussed in the light of these and other results.PREVIOUS structural investigations have indicated that a variety of closely related polysaccharides constitute wheat-straw hemicellulose. Adams and Bishop showed that the hemicellulose contains a backbone of 1 : 4-linked p-D-xylopyranose residues, to which are attached through position 3 a number of L-arabofuranose and D-glucopyruronic acid residues as single-unit side-chains. Earlier studies in these laboratories showed that it was possible to fractionate the hemicellulose and to isolate a xylan, devoid of arabinose residues, but still containing uronic acid residues as an integral part of the molecular structure. These investigations were carried out on hemicellulose fractions isolated by alkaline extraction from straw which had previously been treated with acidified sodium chlorite solution to remove lignin. In view of the possibility of degradation of polysaccharides during chlorite delignifi~ation,~ we have now examined the hemicellulose extracted directly with alkali from wheat straw rather than from the holocellulose.The hemicellulose was extracted from straw, previously extracted with benzene and methanol, by the procedure used in the extraction of beechwood hemicellulose A,5 the structure of which we have previously studied.6 The polysaccharide contained uronic acid residues (ca. 3%) and yielded on hydrolysis xylose and arabinose in the ratio of 11 : 1. Graded hydrolysis liberated arabinose preferentially, suggesting that this sugar occurred in the furanose form, but selective removal of these labile residues to yield a xylan devoid of arabinose residues could not be achieved without extensive degradation of the molecule, as appreciable quantities of xylose and xylose-containing oligosaccharides were released before all the arabinose residues had been removed. On more vigorous hydrolysis, a resistant acidic fraction was isolated ; reduction of the derived methyl ester methyl glycoside with potassium borohydride , followed by hydrolysis, yielded 4-O-methylglucose, glucose, and xylose, indicating the presence in the polysaccharide of residues of glucuronic acid (partly as the 4-methyl ether).The xylan was converted into its fully methylated derivative, hydrolysis of which yielded the following methylated sugars, characterised by the formation of crystalline derivatives : 2 : 3 : 5-tri-O-methyl-~-arabinose, 2 : 3 : 4-tri-O-methyl-~-xylose, 2 : 3-di-O-methyl-~-xylose, and 2-O-methyl-~-xylose (in the approximate ratio of 3 : 1 : 41 : 4). Small quantities of 3-O-methyl-~-xylose were also present ;...

No abstract

No abstract