No abstract

The intense blue colour formed when ketones of the type Ar*CH(CH,*NO,)CH,COArl (I) or ArCH (CN). CH,COArl (11) or some of their simple derivatives and related compounds are heated with ammonium formate or with certain other nitrogen-containing compounds has been shown to be due t o the formation of 2 : 2' : 4 : 4'-tetra-arylazadipyrromethines (111), which constitute a new chromophoric system, having a formal relationship to the phthalocyanines. (111, Ar = Arl = Ph) has been shown t o have the structure assigned to it by degradation with hydriodic acid to 2 : 4-diphenylpyrrole (IV, Ar = Arl = Ph), which is capable of reconversion into (111, Ar = Arl = Ph) by nitrous acid, via 5-nitroso-2 : 4-diphenylpyrrole (V, Ar = Arl = Ph), which condenses with a further molecule of (IV, Ar = Arl = Ph) .(V, Ar = Arl = Ph) has been reduced catalytically to give 5-amino-2 : 4-diphenylpyrrole (VII, Ar = Arl = Ph), which differs from the known 3-amino-2 : 4-diphenylpyrrole (VIII).Certain metallic complexes of (111, Ar = Arl = Ph) and a few analogues of types (111) and (IV) are described.

The structure 7-chloro-4 : 6-dimethoxycoumaran-3-one-2-s~iro-l'-(2'methoxy-6'-methylcycZohex-2'-en-4'-one) (IX ; R = Me) accounts for the known physical and chemical properties of griseofulvin and for the degradation products described in Parts 1-111 (preceding papers). FROM the oxidative degradations described in Part I1 (J., 1952, 3958), it was concluded that griseofulvin possesses a benzenoid ring (A) and a hydroaromatic six-membered ring (c) thereby confirming the views of Oxford, Raistrick, and Simonart (Bioclzem. 3979and C1,Hl,O,C1, which were shown in Parts I1 and I11 to be (VIII; R = H and OH respectively). These acids, derived from griseofulvic acid by alkaline peroxide and by permanganate respectively, are substituted coumaran-3-ones in which the carbonyl and oxygm ether bridges from partial structure (VI) are linked to the same carbon atom, adjacent to a C-methyl group. Union of partial structures (VI) and (VII) in a like manner leads to the two spiran structures (IX) and (X) which are isomeric methyl ethers of the Me0 CO CO-CH CH,

The structures proposed for griseofulvin, C17H1706Cl, by Oxford, Raistrick, and Simonart (Biochem. J., 1939, 33, 240) and by Grove and McGowan (Nature, 1947, 160, 574) are inconsistent with the ultra-violet and infra-red absorption spectra. Grove and McGowan's suggestion (Chew. and Ind., 1949, 647), based on spectroscopic data, that griseofulvin is the methyl ether of a 1 : 3-diketone has been proved conclusively by catalytic reduction of griseofulvic acid, Cl,H ,O,Cl, to two neutral non-lactonic alcohols, C,,Hl,O,C1 and C,,Hl,O,Cl, and by other evidence. The formation of decarboxygriseofulvic acid by the action of alkali on griseofulvic acid is shown by ultra-violet and infra-red spectroscopic evidence to involve a rearrangement of the molecule. GRISEOFULVIN, C,,H,,O,Cl, m. p. 220°, [a]:& +354", a colourless neutral compound containing three methoxyl groups, was isolated from the mycelium of Penicilliuutz griseofuluum Dierckx by Oxford, Raistrick, and Simonart (Biochem. J., 1939, 33, 240). Subsequently, it was isolated from P. jnnczeuskii Zal. [= P. nigricans (Bainier) Thom] and its unique biological activity on moulds noted by Brian, Curtis, and Hemming (Trans. Brit. 2tlycoZ. SOC., 1946, 29, 173; see also Brian, A m . Bat., 1949, 13, 59) and McGowan (Trans. Brit. Mycol. SOC., 1946, 29, 188) who originally called it ' I curling factor " before the identity with griseofulvin was established (Grove and McGowan, Nature, 1947, 160, 574; Brian, Curtis, and Hemming, Trans. Brit. Mycol . SOC., 1949, 32, 30).Although the chemical and biological properties of ' I curling factor " and griseofulvin were identical, the analytical data and molecular-weight determinations quoted by McGowan (Zoc. cit.) in support of his original formula C,,H,,O, agreed more closely with C,,H,,O,Cl than with C,,H,,O,Cl. However, a careful reinvestigation has established the empirical formula as C1,H,,O,C1, in agreement with the earlier work of Oxford ei al. and this is supported by molecular-weight determinations by both the Rast and crystallographic methods (the latter by Dr. A. F. Wells). Nevertheless, the structure (I) tentatively proposed by Oxford et al. (loc. cit.) is not entirely satisfactory even on the basis of the facts which they reported, and we have accumulated considerable evidence which cannot be reconciled with it. We have repeated practically all the experimental workRecent work on the correlation of absorption frequency and structure for the C=O and -OH stretching vibrations has been summarised by Grove and Willis (J., 1951, 877). Normal carboxylic esters generally absorb between 1715 (conjugated) and 1745 cm.-l (unconjugated). These frequencies may be raised by the presence of powerful electronattracting substituents (e.g., NO,) or lowered by intramolecular hydrogen bonding (chelation) ; however, neither of these complications is present in the griseofulvin molecule. Griseofulvin (Fig. 3b) has no absorption bands in the carboxylic ester range, and the Derivatives. Griseofulvin (0.4 g.) and bromine (0.6 g.) were dissolved ...

By M. A. THOROLD ROGERS.:Reprint Order Yo. 5802.1Methods for the preparation of NN-dialkylhydroxylamines from the corresponding amine are reviewed. The pyrolysis of tertiary amine oxides bearing a t least one ethyl or propyl radical is a satisfactory, if occasionally ambiguous, method. N-Oxides of (3-dialkylaminopropionic esters or nitrile, or of Mannich bases, decompose very readily to give the desired compound unambiguously.IN this series of papers we shall be primarily concerned with NA7-dialkylhydroxylamines, (R*CH,),N*OH. No new methods for the preparation of the monoalkylhydroxylamines are reported, but in later papers some of the properties of the monoalkylhydroxylamines will be compared with those of the dialkyl series. The cyclic hydroxylamines, e.g.,