Multivariate optimization of the denitration reaction of nitrocelluloses for safer determination of their nitrogen content

The nitrogen content is of paramount importance to predict the explosive or non-explosive character of nitrocellulose (NC), and hence its applications. There are still quite a few really effective analytical methodologies allowing its determination, due to great NC molecular complexity. One of the approaches giving access to nitrogen content consists in releasing the nitrogenic moieties through alkaline hydrolysis. For the first time, this work reports on the optimization of the denitration yield by means of an experimental design within reasonable ranges of sodium hydroxide concentration, temperature, and time. The experiments were conducted with non-explosive and explosive NC standards. An original capillary electrophoresis (CE) method was used to monitor nitrite and nitrate ions released during hydrolysis. Because of their very different chemical properties, denitration conditions were optimized separately for both sub-classes of NCs to maximize their denitration yields, applying desirability analysis on modeled denitration yields. Mild, safe, and robust optimized conditions were drawn. The denitration yields (95% for non-explosive NCs, 92% for explosive NCs) experimentally obtained under these conditions were in good agreement with model predictions. For practical purposes, correction factors based on these maximal denitration yields are proposed for the first time to correct the determination of nitrogen content, based on preliminary alkaline denitration. This new strategy was successfully applied to determine nitrogen contents of NCs in real explosive samples (smokeless gunpowders). (C) 2015 Elsevier Ireland Ltd. All rights reserved.