On the Formation of Sugars From Formaldehyde

This Thesis is concerned principally with the formose reaction, the conversion of formaldehyde to sugars: it begins with an historical account. The various catalysts, accelerators products are tabulated and the kinetic characteristics reviewed. A number of mechanisms that have been proposed at one time or another, are discussed in the context of whether they refer to the first part of the reaction (the initial formation of glycolaldehyde) or to the second (the creation of higher sugars). The aldol and benzoin condensation, and the Cannizzaro reactions, which have relevance to the formose reaction,are briefly mentioned. An examination of the alumina-catalysed formose reaction follows this. Such a system has been used to elucidate the behaviour of some of the accelerators. In place of the earlier suggestions that "active formaldehyde" was produced by such accelerators as benzoin and vitamin C, new mechanisms have been proposed. The minimum requirement for sugar accelerators has been shown to be the possession of the hydroxyacetyl group. In addition, a new type of accelerator has been found: this has the minimum requirement of a methylene alpha-diketone or a benzoyl methyl group, e.g. diacetyl and kojic acid. The two classes of accelerator coincide in benzoyl carbinol. While neutral and basic alumina catalysed the formose reaction, acidic alumina did not. As this was presumably a pH effect, investigations were undertaken with solutions of various pH to determine at what pH sugar formation became significant. For experimental reasons preliminary investigations were carried out in sodium hydroxide solution, which is homogeneous at these pHs. The change to -sugar was very sudden, so sudden that a "switch on" parameter could be assigned to it. It was in solutions of the highest pH that sugars were formed: immediately below this the final pHs after a prolonged reaction time were approximately constant, probably because of a bufferring effect caused by the Cannizzaro reaction. In solutions of the lowest initial pH no changes were observed. Highly charged ions formed sugars at lower pHs. The tetramethylammonium ion seemed not to favour the Cannizzaro reaction, instead permitting sugar formation at a much lower pH than would have been expected on comparison with other monovalent ions. This observation strongly supports the hypothesis that co-ordination to cations is not necessary for formaldehyde's conversion to sugars: the effect of cyanide ions on this system reinforces this. The reduction of carbon dioxide to formaldehyde photochemicalIy has also been reviewed. Much uncertainty is attached to the validity of earlier claims, although more recent work suggests that the aldehyde is actually formed. The possibility of the formose reaction on the primitive earth was entertained. Starting with the components of the primitive atmosphere, the formation of formaldehyde was reviewed, and the local formation of organic compounds under such circumstances considered (with special reference to formaldehyde). The formose reaction has been shown to be catalysed by rocks. Other factors affecting it have also been mentioned. The formose system comprises a complex, dynamic set of reactions. Speculations on such a system being a primitive phenotype under the influence of a clay genetic material are made