A percolation-theory model of lignin degradation

Abstract

Lignin removal during pulping is the result of the chemical degradation of two distinct cross- linked lignin gel fractions (middle lamella and secondary wall) of different polydispersities, the values of which can be predicted by gelation models. Values of the cluster-size (polydispersity) exponent, T, were calculated using molecular-weight and gel-fraction scaling exponents obtained from the degradation data of five lignin-type models and seven pulping experiments. Results for the middle lamella agree with percolation- theory predictions (T = 2.17) for three-dimensional condensate gels. Also, low values of the cluster-size exponent obtained during secondary-wall degradation in whole wood were in agreement with predictions given for either three-dimensional diffusion-limited cluster-cluster aggregation processes (T = 2.00), or with percolation-theory predictions (T = 2.05) for networks arranged in two-dimensional monolayers. However, the degradation behavior of secondary-wall lignin during pulping of wood meal and sawdust was better described by the Flory-Stockmayer model of gelation (T = 2.501, a model which is applicable to networks having a low cross-linking density. Results for lignin-type model compounds were also interpreted with percolation and cluster-cluster aggregation models. 1.1. Delignification and Lignin Degradation. Lig- nin may be considered to be a cross-linked amorphous polymer network, or gel, of trifunctional monomers spanning two different regions of the fiber wall, which are the middle lamella and the secondary wall. The physical structure of these regions is different, and it is generally assumed that the local lignin structure reflects this difference. The respective rates of delignification of these two regions are indeed quite variable, and those variations are a function of pulping chemistry. In addition, the mac- romolecular properties of soluble (sol) lignin fragments resulting from the degradation of each network are also dissimilar and may be a function of the morphology of the network. Degradation corresponds to the breakdown of the lig- nin network per se in both the middle lamella and the secondary wall regions of the fiber; delignification, on the other hand, is the subsequent removal of soluble lignin fragments from both regions. Lignin degradation char- acteristics have been described by Szabo and Goring' with some success using the Flory theory of gelation,2 in combination with the topochemical model of delignifica- tion,l which empirically predicts the selective removal of treelike lignin fragments from the two regions. This selective removal was first observed by Bixlere3 The Flory theory assumes that intramolecular linkages are absent from the network. However, current knowledge about the nature of native lignin indicates that many lignin fragments do not have a treelike structure and have a significant amount of intramolecular linkages. In the present work, lignin degradation will be analyzed in terms of modern percolation concepts4* and compared to other suitable ge- lation models such as diffusion-limited cluster-cluster aggregation (CCA),4b*c the theory of which is well developed for rigid4blc cross-linked networks. It will be shown that there is good quantitative agreement between the prop- erties of flexible lignin fragments and those predicted by percolation theory for cross-linked networks of different polydispersities. Qualitative agreement with results pre- dicted by cluster-cluster aggregation for rigid networks will also be discussed.