The possibility of orthor nuclear oxidation in the flavone series has been tested using 7-hydroxy-flavone and 3-methoxy-7-hydroxyl-flavone. By means of alkaline persulphate 7: 8-dihydroxy-flavone (10% yield) and 3-methoxy-7: 8-dihydroxy-flavone (20% yield) could be obtained. This seems to offer an easier method of preparing the flavonol, 3: 7: 8-trihydroxy flavone. The intermediate stage of the sulphate could be isolated pure from 7-hydroxyflavone.

The toxic properties of a number of simple hydroxy and methoxy flavanones and chalkones have been studied using fresh-water fish. The flavanones resemble the flavones in a general way. However the methyl ethers of the flavanones appear to be less toxic than the corresponding flavone derivatives, whereas with the hydroxy compounds the reverse seems to be the case.

With the chalkones the toxic symptoms set in more slowly but they are more persistent. The methoxy chalkones are less toxic than the isomeric flavanones whereas when a number of hydroxyl groups are present the reverse is the case. As soon as all the hydroxyl groups in chalkones are methylated the toxicity increases considerably.

A marked difference in the behaviour of naringin and of butrin towards hydrolysing agents has been noticed previously. Thus naringin yields the flavanone, naringenin, while butrin yields a mixture of the flavanone, butin, and the chalkone, butein. Similar difference in behaviour is also exhibited by the flavanones, naringenin and butin, when treated with mineral acids or alkalis.

Methylation has now been conducted of the glycosides, naringin and butrin and their aglucones naringenin and butin using varying molar proportions of dimethyl sulphate in the presence of anhydrous potassium carbonate and in acetone solution. From the study of the products and also of a number of simpler flavanone derivatives it is clear that the presence of a free 5-hydroxyl gives stability to the pyranone ring. This stability is not found when the 5-hydroxyl is non-existent as in butrin and butin or gets methylated as in some of the methyl ethers of naringin and naringenin and other flavanones; the products are chalkones in these cases.

The special influence of the 5-hydroxyl group in stabilising flavanone structure is attributed to the existence of chelation between this hydroxyl and the carbonyl group of the pyranone ring. The mechanism of the chalkone-flavanone conversion is discussed.

Partial demethylation of 2-hydroxy-4 ∶ 6-dimethoxy chalkone with aluminium chloride yields 2 ∶ 4-dihydroxy-6-methoxychalkone which could be cyclised to 5-methoxy-7-hydroxy-flavanone. The use of hydrobromic acid for the demethylation produces directly 5-hydroxy-7-methoxy flavanone due to initial ring closure and subsequent demethylation in the 5-position of the flavanone. When this reagent acts on 2-hydroxy-4 ∶ 6 ∶ 4′-trimethoxy chalkone sakuranetin, the naturally occurring monomethyl ether of naringenin is produced.

The method of demethylation using aluminium chloride in benzene medium is shown to be applicable to polymethoxy flavanones. 5∶7∶4′-, 5∶7∶8- and 5∶6∶7-trimethoxy flavanones yield the corresponding trihydroxy compounds. Similarly carthamidin and isocarthamidin are obtained from their tetramethyl ethers whose synthesis is described. This work provides synthetic confirmation of the constitution of these tetrahydroxy flavanones proposed by Kuroda. That no isomeric change takes place in the preparation of carthamidin is established by its methylation to its tetramethyl ether.

Iodine, in the presence of hot alcoholic sodium acetate, is shown to be a convenient reagent for the conversion of hydroxy flavanones into flavones. Naringenin, its 4′ and 4′: 7-dimethyl ethers, hesperetin and its dimethyl ether are thus oxidised smoothly into apigenin and its ethers and diosmetin respectively. The method is also suitable for glycosides; examples chosen are naringin, its monomethyl ether and hesperidin. The constitution of apiin is discussed and confirmed by correlation with that of naringin.

The method works smoothly in all cases where a free hydroxyl is present in the 5-position. In its absence a mixture is formed; by working in the cold the flavones can be obtained, whereas in the hot benzalcoumaranones could be isolated. In such cases the suitability of the phosphorus pentachloride method has been tested using 7-methoxy flavanone.

The reaction is considered to involve (1) iodination of the 3-position and (2) elimination of hydriodic acid and these are brought about smoothly with the help of the active and unstable acetate ions. If the second stage involves iodinated flavanone, flavone is obtained; on the other hand if the corresponding iodinated chalkone is present, benzal-coumaranone results.

A convenient method of providing confirmation of the constitution of carthamidin and isocarthamidin is to convert them into herbacetin and tangeretin. The earlier claim of Bargellini to have prepared herbacetin tetramethyl ether by the oxidation of 2-hydroxy-3:4:6:4′-tetramethoxy chalkone with alkaline hydrogen peroxide could not be supported. This reaction yields the corresponding benzalcoumaranone and not flavonol in agreement with the behaviour of analogous and simpler chalkones having a methoxyl group in the 6-position. An example is described where even a methyl group in that position brings about the same result. However by the action of amyl nitrite and hydrochloric acid carthamidin and isocarthamidin tetramethyl ethers yield the corresponding flavonol derivatives.

The stem bark of Prunus puddum yields sakuranetin (Ia), sakuranin (Ib), genkwanin (II) and prunetin (III). The identification of sakuranetin is confirmed by the preparation of its diacetate and conversion to genkwanin. Prunusetin is shown to be only a slightly impure sample of prunetin. Though the components belong to different groups of anthoxanthins, they have the common feature that they are all 7-methyl ethers of 5∶ 7∶ 4′-trihydroxy compounds.

A general method for the synthesis of the 7-methyl ethers of hydroxy flavones is described using as examples (1) tectochrysin, (2) genkwanin and (3) 7-O-methyl luteolin. It consists in preparing the corresponding 7-O-methyl flavanone and oxidising it with iodine. 7-O-Methyl eriodictyol is now made from (a) 2∶3′-dihydroxy-4∶6∶4′-trimethoxy chalkone and (b) 2-hydroxy-4∶6-dimethoxy-3′∶4′-dibenzoyloxy chalkone by treatment with hydrobromic acid. Iodine oxidation of it yields (3). The constitution of this product is confirmed by fully ethylating it and showing that the triethyl ether is identical with 7-methoxy-5∶3′∶4′-triethoxy flavone.

The constitution of prunetin as the 7-methyl ether of genistein is established by preparing its diethyl ether and showing that it is identical with 7-methoxy-5 ∶ 4′-diethoxy isoflavone obtained by independent synthesis. Prunetin itself has been synthesised by a method involving partial methylation of genistein using one mole of dimethyl sulphate. The theoretical considerations are discussed. The same procedure leads to a convenient preparation of sakuranetin.

Santal has been synthesised by methylating 5∶7∶3′∶4′-tetrahydroxy isoflavone with one mole of dimethyl sulphate. When three moles of dimethyl sulphate are employed santal dimethyl ether is obtained. An interesting observation has been made that demethylation of santal trimethyl ether under restricted conditions with hydriodic acid leads to the formation of a mixture of santal and nor-santal. They have been separated using aqueous sodium carbonate.

The study of daidzein derivatives provides further support to the conclusions already arrived at regarding the special features of partial methylation and demethylation in the isoflavone series. Partial methylation of daidzein gives a good yield of the 7-monomethyl ether (isoformononetin) and the formation of the isomeric formononetin could not be detected. The same 7-methyl ether is obtained most conveniently by the demethylation of daidzein dimethyl ether with hydrobromic acid and hydriodic acid. 5:7-Dihydroxy and di-methoxy isoflavones have also been examined with a view to compare them with corresponding flavones. 7-Methoxy-5-hydroxy isoflavone is obtained most readily by partial demethylation of dimethoxy isoflavone with hydriodic acid whereas chrysin dimethyl ether yields mostly chrysin under these conditions.

Using methods of nuclear oxidation 7:8-dihydroxy-2-methyl isoflavone, 5:7:8-trihydroxyisoflavone and 5:7:8-trihydroxy-2-methyl isoflavone and their derivatives have been prepared. Demethylation of the 5:7:8-trimethoxy isoflavones with or without a 2-methyl group does not produce isomeric change in the trihydroxy product. It could therefore be concluded that a phenyl group in the 3-position prevents this isomeric change just like a methoxyl (hydroxyl) in the same position and that substitution in the 2-position has no influence.

Two new substances, padmakastin, a glycoside and padmakastein, the corresponding aglucone have been isolated from the bark ofPrunus puddum. The properties and reactions of padmakastein indicate that it is dihydroprunetin. This has been confirmed by its oxidation to prunetin and by its preparation from prunetin by reduction. Padmakastein is the first isoflavanone to be isolated and it gives the well-known Durham test characteristic of rotenoids.

It is shown that the most convenient method of preparing prunetin is by the partial demethylation of genistein trimethyl ether with hydrobromic or hydriodic acid under controlled conditions. The 5-methyl ether of genistein (isoprunetin) has now been synthesised by the methylation of 7:4′-O-dibenzyl genistein and subsequent debenzylation. Its properties are very different from those of prunetin and it yields a diethyl ether which could be independently prepared from genistein. The present work conclusively proves that prunusetin cannot be 5-O-methyl genistein (isoprunetin) and should be the same as prunetin.

A new method of synthesis of flavonols is described. It involves the oxiation of α-methoxy chalkones with alkaline hydrogen peroxide. The special advantages are (1) it avoids the formation of benzal coumaranones and (2) it gives good yields owing to the stability to further oxidation conferred by the 3-methoxyl group. The same products are obtained using selenium dioxide as the oxidising agent. Complete methyl ethers of galangin and kæmpferol have been prepared. The method is also useful for the preparation of partial methyl ethers of flavonols. 3∶5∶7∶4′-Tetramethyl ether of quercetin and 3∶5∶7-trimethyl ether of quercetin are prepared. The former has been subjected to partial demethylation to yield the new 7∶4′-dimethyl ether of quercetin which is now named isorhamnazin.

Under ordinary conditions 2-hydroxy-6-methoxy chalkone yields with alkaline hydrogen-peroxide only 4-methoxy benzalcoumaranone and not 5-methoxy flavonol as claimed by Oliverio and Schiavello.

Heating 2-hydroxy-3∶4∶6-trimethoxy-chalkone with hydrobromic acid yields instead of the expected 7-methoxy-5∶8-dihydroxy flavanone, the isomeric 7-methoxy-5∶6-dihydroxy flavanone. The same product is obtained by the partial demethylation of 2-hydroxy-4∶5∶6-trimethoxy chalkone. Attempts to prepare 5∶8-dihydroxy-7-methoxy flavanone by the reduction of the corresponding quinone again resulted in the formation of the isomeric 7-methoxy-5∶6-dihydroxy flavanone. The 5∶8-dihydroxy compound has been obtained by the partial demethylation of 5∶7∶8-trimethoxy flavanone by means of aluminium chloride in nitrobenzene solution.

The partial demethylation of 5∶7∶3′∶4′-tetramethoxy isoflavone has now been effected under improved conditions and an almost quantitative yield of santal obtained. Partial methylation of nor-santal has been repeated using larger quantities and santal and its acetate prepared and studied.

Past records indicate that an 8-hydroxyl group in a flavone is involved in chelation but it is of a weak nature. 7-Methoxy-8-hydroxy flavanone can be easily obtained by the partial methylation of 7∶8-dihydroxy flavanone or the partial demethylation of 7∶8-dimethoxy flavanone, and can also be directly obtained from gallacetophenone-4-methyl ether. Its preparation and properties indicate the existence of stronger chelation involving the 8-hydroxyl group. Similar experiments have been done on the preparation of 7-methoxy-8-hydroxy chromone and isoflavone. Though the partial demethylation of 7∶8-dimethoxy compounds has been successful, partial methylation of the dihydroxy compounds has not been possible in these cases. The properties of the products also indicate weaker chelation in the chromone and isoflavone just as in the corresponding flavone. In the course of the synthesis of 7-methoxy-8-hydroxy flavanone and chromone independently for purpose of comparison the preparation of the intermediate gallacetophenone 4-methyl ether has been re-investigated and a convenient method has been worked out.

For the preparation of flavanones having a 2′-methoxy group the acid method of chalkone condensation employing benzoates of hydroxy ketones is convenient. 5-Hydroxy-2′-methoxy flavanone and 5∶7-dihydroxy-2′-methoxy flavanone have thus been prepared. By demethylating the former with aluminium chloride 5∶2′-dihydroxy flavanone is obtained which undergoes only partial methylation in the 5-position even with excess of the reagent confirming that the 2′-hydroxyl is highly resistant. Synthetic 5∶7-hydroxy-2′-methoxy flavanone and its derivatives made by us and by Shinoda and Sato agree, but they do not agree in their properties with natural citronetin and its derivatives.

5-Hydroxy-2′-methoxy flavone does not undergo isomeric change under ordinary conditions of demethylation. 5∶2′-dihydroxy flavone undergoes partial methylation smoothly in the 2′-position thus showing the existence of marked difference between flavanones and flavones.

In their behaviour towards alkaline hydrogen peroxide 2′-methoxy chalkones behave normally and yield flavonols if the 6-position is free and benzalcoumaranones if this position has a methoxyl group. On the other hand 2′-hydroxy flavanones, and 5-hydroxy flavanones remain unaffected. This result is explicable on the basis that these hydroxy flavanone structures are stable and do not undergo change into the corresponding chalkone structures which alone suffer oxidation.

Using the more easily accessible 5:7: 8-trihydroxy derivatives of isoflavones a convenient method has been worked out for the synthesis of 5: 6: 7-trihydroxy derivatives. It consists in the alkali fission of 5-hydroxy-7: 8-dimethoxy isoflavone and its 2-methyl derivative to yield 2: 6-dihydroxy-3: 4-dimethoxy-phenyl-benzyl ketone. This ketone on treatment with ethyl formate and sodium yields 5-hydroxy-6 : 7-dimethoxy isoflavone. Vigorous acetylation of the ketone using acetic anhydride and sodium acetate yields 2-methyl-5-hydroxy-6: 7-dimethoxy isoflavone which is compared with an authentic sample prepared by an independent method. 5-Hydroxy-7 : 8-dimethoxy isoflavone yields a small quantity of 5-hydroxy 6: 7-dimethoxy isoflavone during alkali fission.

The action of iodine and silver acetate varies with the type of flavanones employed, and yields the intermediate stages postulated in the new flavone synthesis using iodine and sodium acetate. The polyhydroxyflavanones, naringenin and hesperetin form directly the 3-acetates which can be hydrolysed to the 3-hydroxy compounds. On the other hand from naringenin 7:4′-dimethyl ether and 5:7-dimethoxy flavanone, the 3-iodo compounds can be isolated as the first intermediates. The behaviour of the 3-hydroxy flavanones in acid and alkaline conditions also varies; either dehydrogenation to flavonol or dehydration to flavone taking place.

Our thanks are due to Dr. Pew of the Forest Products Laboratory, Madison, U.S.A., for the generous supply of 3-hydroxy naringenin used in this work.