No abstract

Further understanding of the reaction in which 2,3-butanediol is dehydrated over thorium oxide to methyl vinyl carbinol and butadiene has been gained by measuring the adsorption on thoria of water, methyl ethyl ketone and the above-mentioned diol, carbinol, and diene, at temperatures up to 200 �C. In some instances satisfactory isotherms could not be obtained because sufficient chemisorption occurred, followed by polymerization or dehydration reactions, to reduce seriously the area of thoria surface available for adsorption. There was evidence to suggest that water, alone of the vapours concerned, was taken up in greater amount than could be accounted for by adsorption. The bearing of this on the dehydration and its relation to the structure of the catalyst is briefly discussed. A suggestion in a previous paper(1) that water is the product which retards the catalytic dehydration is confirmed. At temperatures little above 50 �C, methyl vinyl carbinol is dehydrated by thoria to butadiene, while methyl ethyl ketone at room temperature undergoes self-condensation on thoria to yield a methyl heptenone. Also at quite low temperatures dimerization of butadiene brings about appreciable reduction of available surface. Some loss of activity on this account must therefore be expected when thoria is used to dehydrate 2,3-butanediol, even though butadiene is adsorbed only weakly.

By spectrophotometric scanning of reacting solutions, we have studied the processes which can occur when Oz is admitted to aqueous solutions containing the Ion [CoII(CN)5]3- in the concentration range 0.0002-0.4M, at 0�. A scheme is given to explain the formation of and inter-relations between [(CN)5CoIII-O-O-CoIII(CN)5]6-; [(CN)5CoIII-O-O-CoIV(CN)5]5-; [(CN)5CoIIIOH2]2- ; [CoIII(CN)6]3-, and a complex believed to be [(CN)5CoIII-O-OH]3-, which absorbs at 272 mp Preparative methods are described for obtaining several of the oxidation products in the crystaline state. Hydrolysis of [(CN)5CoIII-O-O-CoIII(CN)5]6- to [(CN)5CoIIIOH2]2- takes place via the 272 mμ complex, which is fairly stable in alkaline solution and is resistant to oxidation. The same 272 mμ complex can be formed more directly by oxygenation of the hydrido complex [(CN)5CoIIIH]3-, which is present in aged or hydrogenated solutions of pentacyanocobaltate(II). All of the pentacyano oxidation products can be reduced by sodium borohydride to [(CN)5CoIIIH]3-, which is further reduced by excess borohydride to an insoluble green compound and finally to metallic cobalt. We have briefly examined the reactions of ammonia and CN- with [(CN)5CoIIIOH2]2-, and the reaction of [(CN)5CoIII-O-O-CoIV(CN)5]5- with reducing agents, including [CoII(CN)5]3-.

By means of a low-temperature ESR technique, several paramagnetic complexes have been detected during the reaction of aqueous [CoII(CN)5]3- with O2. Of these only [(CN)5CoIII-O-O-CoIV(CN)5]5-, the final oxidation product, is stable in solution. It has a 15-line spectrum resembling that of [(NH3)5CoIII-O-O-CoIV(NH3)5]5+ but more sharply defined and independent of pH. Under very alkaline conditions the first oxidation product detectable by ESR has an 8-line signal, indicating a mononuclear complex, possibly [(CN)5CoIII-O-O]3-. It is an intermediate in the oxidation of [(CN)5CoIII-O-O-CoIII(CN)5]6- to [(CN)5CoIII-O-O-CoIV(CN)5]5- and may reach a high concentration in the presence of 0.5M KOH. Although it may also prove to be the expected mononuclear inter- mediate in the oxidation of [CoII(CN)5]3- to [(CN)5CoIII-O-O-CoIII(CN)5]6-, no 8-line signal is detected until the cyanocobaltate(II) has been consumed. Present in small amount is a mononuclear complex with 16 lines, apparently a by-product. Larger amounts can be formed by reducing [(CN)5CoIII-O-O-CoIII(CN)5]6- with NaBH4 in strong alkali at -20�C.

No abstract