Structures of Biopolymers in the Pulp and Paper Industry

Principal Investigator

Lawrie Dunn

Chemistry

University of Tasmania

Cellulose and lignin are the biopolymers of interest in this project. Both occur naturally in the hardwoods, softwoods and grasses(bamboo) used as raw materials by the pulp and paper industry. These three sources generate three different types of lignins. Gaussian 94 techniques have been used to establish optimised molecular geometries and atomic electron distributions for lignin dimer species with ß-0-4 structural linkages. Structural features found to be crucial for oxidative cleavage of lignin include the ß-0-4 aryl ether linkage, an aromatic 4-hydroxyl group and a -CH2OH group on Cb in the lignin sidechain. Additional base seems to initiate slow deprotonation at Ca which then competes with the Ca-Cb cleavage.

Because of the abundance of the lignin biopolymer in wood and because the fundamental chemical changes induced in the lignin biopolymer during pulping, bleaching and paper discoloration processes are of such considerable chemical complexity and industrial significance, our original aims and objectives for this project are theoretical studies of this biopolymer at the highest possible level of theoretical chemistry.

We are approximately half-way towards completion of our planned lignin ß-0-4 dimer studies. Our Gaussian-94 RHF 3-21g, 6-31g, 6-31g(d) results so far all seem to indicate that the planned MP2 6-31g(d) studies would indeed provide additional, relevant, fundamental molecular information for each of lignin, cellulose, phenolformaldehyde and polyethyleneoxide, all so important to the pulp and paper industry.

Co-Investigators

Karen Stack

Chemistry

University of Tasmania

Projects

g53 - VPP

What are the results to date and the future of the work?

Fully optimised geometries and atomic charge distributions of the p-hydroxyphenyl(php), guaiacyl(gua) and syringyl(syr) momomer units of lignin in grasses, softwoods and hardwoods respectively have now been completed at the Gaussian-94 RHF/3-21g, RHF/6-31g and RHF/6-31g(d) levels of theory. MP2/6-31G(d) optimisations will be initiated with these RHF/6-31g(d) minimum energy geometries. These calculations have been carried out in water (T=298K, e =78.30) to simulate the aqueous environment of pulp mills. For the monomer species studied, frequency calculations have confirmed these optimised results as genuine minima on the


- Appendix B


appropriate molecular potential energy surfaces. Similar calculations have been completed only at the Gaussian-94 RHF/3-21g level for the php-php, gua-gua and syr-syr lignin dimers with the ß-0-4 structural linkage. These results at this level already seem to confirm that the favoured configuration about the Ca-Cb linkage is indeed erythro rather than threo .

Lignin dimer geometry optimisation calculations at higher levels of theory, using the RHF/3-21g optimised geometries as input, have been successful but the high theory level minima characterisation calculations with Gaussian-94 have so far not been as successful. Various sorts of computational difficulties have arisen. Attempts to solve these with the assistance of both colleagues and ANUSF staff, by experimenting with program memory allocation e use of MRFS, have taken time and used our time allocation non-productively for our planned lignin 3-mer, 4-mer and 5-mer calculations.

With these present calculational difficulties for our high polymer lignin species, the emphasis of this project has shifted temporarily to dimers of cellulose, dimers and trimers of phenolformaldehyde resin and octamers of the linear polyethyleneoxide molecule. These last two species are retention aids used extensively in the pulp and paper industry. Geometry optimisations, charge distributions and potential energy minima identifications for these cellulose, phenolformaldehyde and polyethyleneoxide molecules at the RHF/3-21g, RHF/6-31g and RHF/6-31g(d) levels of theory are almost completed. MP2/6-31g(d) optimisation calculations will be initiated with these RHF/6-31g(d) minimum energy geometries.

What computational techniques are used?

Gaussian-94