Analytical Methods in Rubber Chemistry. (IV) The Determination of Peroxidic Oxygen

In the previous publication, of Farmer, Tristram and Bolland, it was stated that methods were required to estimate the various forms in which oxygen occurs united to the rubber molecule, and a method for the estimation of hydroxyl oxygen was discussed. The determination of peroxidic oxygen in oxidized rubber presents two distinct problems. It is necessary to employ a method which, first, is satisfactory in the presence of unsaturation and, secondly, ensures effective contact between the long rubber chains and the analytical reagent. These requirements are both satisfied by the modification described below of the colorimetric method of Young, Vogt and Nieuwland. Before describing the modified technique, some mention must be made of the earlier methods and of the errors which arise when they are applied to the determination of peroxidic oxygen in unsaturated compounds. All the earlier methods for the determination of peroxidic oxygen depend on the oxidation of potassium iodide and titration of the liberated iodine. This method is open to the objection that the liberated iodine immediately attacks the double bonds remaining in the oxidized material. This difficulty was recognized by Stephens, who, using a method in which the peroxidic oxygen was reduced with stannous chloride, succeeded in estimating about 80 per cent of the total peroxide present in cyclohexene peroxide. Yule and Wilson estimated the peroxide in cracked gasoline by shaking with an acid solution of ferrous thiocyanate and back-titrating the resulting ferric salt with titanous chloride. The results were later contested by Young, Vogt and Nieuwland, who described the colorimetric method which has been adopted, with some modification, in the present work.