秸秆纤维素在亚/超临界乙醇中液化反应路径及5集总反应动力学模型的建立  

Reaction pathway and 5-lump kinetic model of cornstalk cellulose liquefaction in sub-and supercritical ethanol

作　　者：战晓青 解新安[1] 李璐[1] 王娅莉[1] 陶红秀[1] 

机构地区：[1]华南农业大学食品学院,广东广州510641

出　　处：《化工进展》2015年第11期3892-3897,共6页Chemical Industry and Engineering Progress

基　　金：国家自然科学基金项目(21176097)

摘　　要：将玉米秸秆纤维素在亚/超临界乙醇中液化反应产物划分为残渣（RE）、轻油（WSO）、重油（HO）、挥发分（VO）和气体（GAS）5个集总。根据液化过程及其产物分布,建立集总反应路径,并以此建立了5集总反应网络。对反应网络建立了动力学模型,采用遗传算法结合非线性规划的求解策略简化和优化动力学模型,求取了动力学参数包括指前因子（A）、活化能（Ea）和反应级数（n）,并用所建立的动力学模型对液化过程进行了模拟计算,模拟结果表明WSO和VO收率随停留时间增加而减少,而HO和GAS则随停留时间增加而增加,这一规律与实验数据变化一致。以实验数据对动力学模型进行验证,各集总模拟计算值与实验值相关系数在0.9345～0.9995之间,吻合较好,表明所建立的动力学模型是可信的,所采取的遗传算法结合非线性规划求解策略是可行的。The liquefaction products of cornstalk liquefied in sub/supercritical ethanol were divided into 5 lumps：residue （RE）,water solute oil （WSO）, heavy oil （HO）,volatile （VO） and gas （GAS）. Reaction pathways and 5-lump network were established according to the liquefaction process and product distribution,followed by kinetic modeling. Simplification and optimization were achieved using genetic algorithm （GA） combined with nonlinear programming. Pre-exponential factor （A）, reaction activation energy （Ea） and reaction order （n） were determined. Then the model is used to simulate the liquefaction process,and the result shows that WSO and VO yields grow with increasing remaining time,while HO and GAS yields behave just the opposite. The model is validated by experiment results,and correlation coefficient is ranged 0.9345—0.9995,which indicated that the model is reliable and the method used to calculatekinetic parameters is practicable.

关 键 词：纤维素 液化 亚/超临界 乙醇 动力学 遗传算法 

分 类 号：TK6[动力工程及工程热物理—生物能]