Fast pyrolysis behavior of different celluloses and lignocellulosic biopolymer interaction during fast pyrolysis

摘要

Cellulose, hemicellulose and lignin are the three major components of biomass, constituting about 85 wt. %. In the current study, primary fast pyrolysis of biomass has been studied by using a system with an online micropyrolyzer-GC-MS/FID and gas analyzer for product analysis while a combination of analytical technologies including HPLC, UV-VIS and ICP was used for biomass characterization.In the first part, pyrolysis behavior of cellulose with different crystallinities and degree of polymerization has been studied. The results showed that primary pyrolysis behavior of cellulose was not affected by its crystalline structure or its molecular weight. In the second part, the interaction effects within cellulose-hemicellulose and cellulose-lignin under the condition of primary fast pyrolysis were studied, by comparing the pyrolysis fashion of their native mixture, physical mixture and superposition of single components. For cornstover, it has been found that no apparent interaction exists for either of its physical binary mixture. However, when pyrolyzing native cornstover cellulose-lignin, levoglucosan yield was significantly decreased accompanied by higher yield for low molecular compounds and furans. Further, native cellulose-lignin from red oak, pine and switchgrass had also been pyrolyzed. Similar cellulose-lignin interaction effects were observed for switchgrass, which were however, not shown for pine and red oak. For the native mixture of cornstover cellulose-hemicellulose, the yield of levoglucosan is similar to the one from either physical mixture or superposition, while different yield were observed for other products, which might be caused by the chemical difference between native hemicellulose and extracted hemicellulose.This study provides an insight to the chemistry involved during primary fast pyrolysis of different types of cellulose and lignocellulosic biopolymer, which facilitates building up a model to predict bio-oil composition.