【作者】 陈怡；

【导师】 袁慎峰；

【作者基本信息】 浙江大学 ， 化学工程（专业学位）， 2019， 硕士

【摘要】 费托合成产物成分复杂,大部分为正构烷烃和伯醇,其中高碳醇的经济价值较高,被广泛应用于精细化工产品中。但因为正构烷烃和伯醇之间存在共沸现象,普通精馏难以对其进行分离,而萃取是一种有效又经济的分离方式。本文选取辛醇-壬醇-十二烷-十三烷混合物为研究对象,采用先萃取分离醇烃混合物后精馏精制的工艺方法,以年处理2000吨辛醇-壬醇-十二烷-十三烷混合物为例,通过实验和模拟,对其分离工艺进行设计和优化,为工业生产提供依据。本文以乙醇水溶液作为萃取剂,首先通过液液平衡实验获得常压下293.15K、298.15K和303.15K乙醇-辛醇-十二烷-十三烷-水、乙醇-辛醇-十三烷-水、乙醇-壬醇-十二烷-水、乙醇-壬醇-十三烷-水4个四元体系的液液相平衡数据,使用NRTL热力学模型对液液相平衡数据进行关联拟合,获得对应的二元交互作用参数,用于预测乙醇-辛醇-壬醇-十二烷-十三烷-水六元体系的液液相平衡数据,预测结果和实验结果有较高的一致性,表明4个四元体系液液平衡数据回归关联所得的二元交互作用参数可以用于预测六元体系液液平衡数据,并进一步用于萃取流程的模拟和优化。通过分析分配系数和选择性系数发现,75%的乙醇水溶液是较合适的萃取剂,298.15K是进行萃取操作较为合适的温度。其次,本文通过筛选,选定正十六烷作为反萃剂,测定了常压298.15K、303.15K和308.15K条件下的乙醇-辛醇-壬醇-十二烷-十三烷-水-十六烷七元液液平衡数据,并用NRTL模型进行了回归,和Aspen Plus的默认参数相比,回归得到的二元交互作用参数有更高的准确性,可用于后续萃取的模拟与优化中。综合考虑选择性系数、传质速率和粘度等,298.15K是进行反萃操作较为合适的温度。最后,本文利用Aspen Plus模拟设计并优化了分离辛醇-壬醇-十二烷-十三烷混合物的工艺流程,以年处理2000吨辛醇-壬醇-十二烷-十三烷混合物为例,用75%的乙醇水溶液作为萃取剂,通过多级逆流萃取,萃余相进行水洗获得99.8%纯度的十二烷-十三烷产品;萃取相以正十六烷作为反萃剂进行反萃,所得的萃余相经过一系列精馏处理,获得纯度为99.5%的辛醇-壬醇产品。萃取剂和反萃剂通过精馏循环使用。更多还原

【Abstract】 The components of Fischer-Tropsch synthesis products are complex,most of which are n-alkanes and primary alcohols.Primary alcohols especially higher alcohols have high commercial value,which are widely used as raw materials of fine chemical products.Due to the azeotropy between n-alkanes and alcohols,it is difficult to separate by normal distillation,while extraction is considerd as an effective and economical separate method.In this paper,octanol-nonanol-dedocane-tridecane mixture was seperated by combination of extraction and distrillation,and a process of treating 2000 t/a octanol-nonanol-dodecane-tridecane mixture was designed and optimized.In this paper,firstly,the LLE data of ethanol-octanol-dodecane-water,ethanol-octanol-tridecane-water,ethanol-nonanol-dodecane-water,ethanol-nonanol-tridecane-water at 293.15K,298.15K and 303.15K under atmospheric pressure were obtained experimentally.The NRTL equation was used to correlate the LLE data and the binary interaction parameters were acquired to predict the LLE data of ethanol-octanol-nonanol-dodecane-tridecane-water system.The predicted data agreed with the experimental data well,which indicated the binary interaction parameters acquired from LLE data of four quaternary systems could be used to predict LLE data of six-component system.The data of six-component system were evaluated by distribution coefficient and selectivity,which showed that 75%ethanol aqueous solution was suitable for separating octanol and nonanol from octanol-nonanol-dodecane-tridecane mixture.298.15K was the suitable temperature for extraction operation.Secondly,n-hexadacane was selected as reverse-extraction solvent.The LLE data of ethanol-octanol-nonanol-dodecane-tridecane-water-hexcadacane at 298.15K,303.15K and 308.15K under atomosphic pressure were obtained experimentally,and the LLE data was regressed by NRTL model.Comparing with default parameters in Aspen Plus database,the regressed parameters were more accurate.The results showed that the regressed parameters could be used in separation process simulation and optimization.By comparing distribution coefficient and selectivity at different temperatures,and considering higher temperature would give a higher mass transfer rate.Comprehensively,298.15 K was the suitable temperatureFinally,a separation process of treating 2000t/a octanol-nonanol-dodecane-tridecane mixture was designed and optimized by Aspen Plus.75%ethanol aqueous solution was selected as extractant,the raffinate phase was purified and 99.8%dodecane-tridecane product was obtained.Extract phase was reverse-extracted by hexcadacane,and the 99.5%octanol-nonanol product was obtained by a series of distillation operation.Extractant and reverse-extraction solvent were recycled.更多还原