Open Collections

Gas absorption on wood pulp cellulose

University of British Columbia

Nitrogen, argon and oxygen adsorption isotherms at 78 °K have been determined on samples of solvent exchange dried, fully bleached, kraft pulp of western hemlock and Douglas fir. The solvent exchange drying sequence used was water-methanol-n-pentane with the n-pentane removed at room temperature. The pulp samples were in two groups: one group was beaten to various levels on a P.F.I. mill and solvent exchange dried from a water swollen state; the second group was air dried to different moisture contents and then solvent exchange dried. The presence of micropores (Dubinin definition) in solvent exchange dried cellulose has been shown. This finding was suggested by the results of Harris, who noted that adsorbents having an average pore radius of less than 18 Å indicated an average Kelvin pore radius of 18 Å. The finding of micropores in solvent exchange dried cellulose was substantiated by the pore volume distribution which indicated a very large volume of pores at 18 Å radius, and by the Dubinin plot technique for the measurement of micropore volume. Some solvent exchange dried wood pulp samples indicated that up to 70 percent of their total pore volume was present as micropores. The presence of micropores with the associated enhanced adsorption and restricted adsorption space complicates the analysis of isotherms so that the B. E. T. surface area and Kelvin type pore analysis cannot be considered reliable. An investigation into the effects of errors associated with assuming values for the physical properties of adsorbed molecules led to the conclusion that gas adsorption is useful for indicating trends or for use as a parameter for correlations of some pulp and paper properties but should not be used to postulate models of specified dimensions. The Kelvin type pore analysis and the Dubinin pore analysis have been shown to be very sensitive to the model assumed to describe the physical structure.of the wood pulp fibre. While gas adsorption techniques are of doubtful validity for solvent exchange dried pulp samples, the B.E.T. analysis, was found to be quite reliable on air dried paper sheets. The existence of the large volume of pores of approximately 18 Å radius (cylindrical pore model) or 25 Å wall separation (parallel sided fissure model) in solvent exchange dried cellulosic materials which have been found by other workers has been shown to be doubtful. This finding is substantiated by the accessibility data of Stone and Scallan which does not indicate a large volume of pores near the most common pore sizes found by gas adsorption. P. F. I. mill beating of the pulp was found to lower the surface area slightly, to shift the pore size distribution slightly to larger pore sizes and to substantially lower the volume of pores at the most common pore size. The Dubinin micropore volume analysis also indicated lowering of micropore volume with beating. These results led to the conclusion that beating affected the structure of the pulp fibre even in structural elements of the smallest size measured. Partial drying of the pulp handsheets prior to solvent exchange drying was found to lower the surface area by nearly two orders of magnitude and to shift the pore volume distribution strongly to the smaller pore sizes. The effects of .beating and drying were easily interpreted on the basis of the Stone and Scallan parallel sided fissure model of the structure of cellulose. Comparisons between isotherm shapes indicated solvent exchange dried wood pulp may have a structure similar to montmorillonite, which is known to have a flat plate structure. The Kaganer and "t"-plot analytical techniques were applied to the solvent exchange dried wood pulp as additional methods of obtaining estimates of the surface area. The surface areas determined by these methods were significantly larger than those found by the B. E. T. equation, however, the trends were found to be the same. Handsheets were made from the pulp samples and these sheets were subjected to physical tests. The surface areas of these sheets were determined and the bonded areas estimated. The bonded area was found to increase with the level of beating. Sheet density increased as bonded area increased, so did the breaking length and burst. Tear factor decreased as the bonded area increased.

Text

2011-06-03

For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.

http://hdl.handle.net/2429/35108

Chemical and Biological Engineering

University of British Columbia

Graduate