UBC Theses and Dissertations
Predictive modeling of high-pressure NAEM-catalyzed methanol pulping of spruce wood Lohrasebi Peydeh, Abdolhossein
Environmental, economic and raw material concerns encourage researchers to search for alternative pulping and bleaching methods. Organosolv pulping and bleaching are well-known alternative methods to traditional pulping and bleaching processes. The neutral alkali earth metal (NAEM) salts-catalyzed methanol pulping process produces high yield screened pulps from all lignocellulosic materials. However, when such refractory softwoods as spruce, are cooked by means of organosolv pulping at high alcohol concentrations (>70%), the penetration of liquor into the wood cell walls is partially blocked due to the frequent presence of aspirated pits in the air-dried fibers and air in the lumen. As a result of the decrease in liquor penetration, the diffusion of chemicals to the reaction sites is also decreased. Consequently, delignification and fiber liberation are delayed. To study whether increasing the cooking pressure changes the pulpability of air-dried wood chips (MC=7%), this study was set up in such a way that three major cooking variables (time, temperature, and particularly pressure) were varied in a split-plot experimental design with at least two replicates, and their individual and combined effects on delignification were assessed. The main objective of this research was to examine the effects of high pressure on spruce organosolv pulping behavior in terms of fiber liberation point (FLP), degree and selectivity of delignification, screened yield and residual lignin content of the resulting pulps at the FLP. The cooking conditions were optimized to achieve high-screened yield and low Kappa number pulp at the FLP. In this research, batch, isothermal delignification of spruce wood chips was carried out by means of high-pressure NAEM-catalyzed organosolv pulping. The cooking liquor contained methanol + water (80/20 by volume), and NAEM salts with liquor/wood ratio of 10 mL/g wood, all of which remained constant throughout the course of the cooks. The primary cooking variable was hydraulic reactor pressure (500-4000 psi) with cooking time (10-120 min) and temperature (190°-210°C) as secondary variables. Models were then developed to predict the responses of pulping, such as total yield, reject content, lignin-free yield, Kappa no., amounts of lignin and sugars dissolved and remaining. It was found that, in addition to cooking temperature and time, pressure had also a statistically significant effect on pulping behavior of spruce wood and on pulp chemical properties. Pulping was pressure sensitive particularly when cooking at low temperatures (190°-200°C) for short cooking times until the fiber liberation point was reached. Pressure was more effective in lowering reject content and consequently increasing screened-yield, than in reducing Kappa number. Thus, the issue of screen rejects could be resolved by applying high pressure (optimally 2000 psi) to the softwood organosolv pulping system while producing pulps of various qualities/grades that can be utilized for different applications. The high pressure did not affect residual mannose and xylose contents significantly; their dissolution depended on cooking temperature and time. Arabinose and galactose disappeared very early in the cook before the FLP. The topochemistry of delignification and fiber liberation (low reject) can be attributed to the removal of arabinose and galactose, in addition to the partial removal of pectins and lignins from the compound middle lamella. It was also found that there was a significant correlation between pH of spent liquor and chemistry of pulping (pulp composition). Dissolution of mannose and xylose were associated with the pH of the spent liquor during different stages of cooking. In all cases, the pH decreased from its initial value of 6.5 of the fresh cooking liquor and reached its optimum level of 4.2±0.2 at approximately the fiber liberation point, buffered by NAEM catalysts. Furthermore, dissolution of mannose and xylose were highly intercorrelated (r=0.99). Removal of xylose and mannose were also highly correlated with delignification, i.e. r=0.885 and 0.926, respectively. Moreover, the reject content of pulp was also correlated with xylose, mannose and lignin contents of the pulp, having a correlation coefficient (r) of 0.628, 0.677 and 0.838, respectively. Mathematical models were then developed, tested, validated and finalized. Using SAS software to run different subset selection procedures and taking several parameters into consideration, MAXR technique was found to be the most appropriate procedure to use when building the multiple regression models. All model-building procedures were followed and explained for the future modeling uses. The final models can predict many properties of the softwood NAEM pulp from the cooking variables with high accuracy (R²>92%) and thus can be applied for process control purposes in the pulp mill. The response variables modeled include total yield, reject content, Kappa number, residual and dissolved lignin, lignin-free yield, and mannose and xylose contents of the NAEM pulp. The general form of the models for most responses is as follows: Y = b₀ - b₁T + b₂Tt – b₃t – b₄ t[sup ½] – b₅P R²=92-99%. Optimum cooking conditions to obtain early fiber liberation, low reject content (53%. Increasing the cooking temperature above 200°C is not recommended since it extensively dissolved hemicelluloses and degraded cellulose, resulting in a lower pulp yield and viscosity. The effects of the pulping conditions on the response variables, as predicted by the models, are consistent with the known physical and chemical reactions occurring during the organosolv pulping process. The distinctly different behavior of high-pressure NAEM-catalyzed organosolv pulping, compared to uncatalyzed organosolv pulping, seems to be due to the unique topochemistry of delignification (selective removal of lignins, arabinose, galactose, and pectins from the compound middle lamella), different mechanisms of reactions, and non-condensed structure of the dissolved and residual lignins.
Item Citations and Data