煤化工废水脱酚萃取剂计算机辅助设计

Computer-aided molecular design of extraction agents for coal chemical industry phenol-bearing wastewater treatment

低损高效萃取剂筛选设计是煤化工废水萃取脱酚技术研究的关键问题之一。针对当前常用萃取剂普遍存在的选择性单一、多元酚萃取能力差的问题,本文采用计算机辅助分子设计方法开展了针对煤化工废水典型酚类污染物萃取剂的筛选设计,主要工作包括:1)考虑到经验性初始基团筛选方法存在不确定性大的问题,建立了新的基团选择策略,即通过将已知或者潜在的具有较好萃取效果的化合物进行基团拆解和分析,得到用于分子设计的基团组,以减少因设计者个人偏好造成的偏差;2)采用分步求解方法对萃取剂设计MINLP问题进行求解,针对求解过程中搜索全部可行域耗时长的问题,提出可行域分块求解策略,大幅提高求解速度;3)对煤化工废水中含量较高的苯酚、苯二酚和苯三酚3种典型酚类污染物进行了萃取剂设计,分别得到128种、55种和3种萃取剂,并分析总结了萃取剂分子结构特点和规律。研究结果可为进一步进行对总酚脱除具有良好效果的萃取剂实验筛选提供支持。

Though some commonly used extraction agents are efficient for extracting monohydric phenols from wastewater, most of them cannot effectively remove polyphenols. Searching alternative efficient and low loss extraction agents for phenol-bearing wastewater treatment is one of the key issues needed to be conquered for coal chemical industry. This paper was focused on searching novel extraction agents for typical phenolic pollutants contained in wastewater through computer-aided molecular design （CAMD）. First, a new group selection strategy was formulated as the altemative of the experience based strategy. This new strategy identifies groups via analyzing the structural information about commonly used solvents in industrial practice and possible extraction agents found in the databases that meet the specified criteria. Second, the decomposed method was adopted to solve the formulated CAMD problem. To solve this problem efficiently, a new feasible domain decomposition strategy was set up. Third, extraction agents were designed for phenol, hydroquinone and pyrogallol separately, the number of candidate molecules generated for them are 128, 55 and 3, respectively. The obtained molecules can be used as potential candidates for further mixture design.