Empirical modeling of chlorine dioxide delignification of oxygen-delignified hardwood kraft pulp

Due to the complexity of delignification, it is still not accurate enough to describe the whole chlorine dioxide delignification system with mechanistic models such as kinetic studies. The simple aim of predicting and optimizing the process requires the use of economical empirical models, few of which are in the literature. In this study, principal component analysis was used to classify various bleaching response variables, including end pH, kappa number, viscosity, and optical properties such as brightness, into different categories. Statistical methods, including factorial design, multiple regression, and response surface methodology, were then used successfully to screen the importance of bleaching factors and their interactions. Models to control, predicate, and optimize the bleaching responses were established. In most cases, the results of empirical modeling coincided well with those from conventional research methods, but the empirical methodology was more economic or less laborious. The empirical models are robust enough to predicate and optimize the bleaching responses to chlorine dioxide delignification of oxygen-delignified hardwood kraft pulp.