一氧化碳/苯乙烯共聚催化体系的优化  

OPTIMIZATION OF CARBON MONOXIDE/STYRENE COPOLYMERIZATION CATALYST SYSTEM

作　　者：姚芳莲[1] 陈锦言[1] 邓联东[1] 孟继红[1] 许涌深[1] 孙经武[1] 

机构地区：[1]天津大学化工学院,天津300072

出　　处：《催化学报》1999年第2期155-160,共6页

基　　金：韩国三星集团资助

摘　　要：用醋酸钯、１ ，１０邻菲咯啉、对甲苯磺酸、对苯二醌和甲醇组成的催化体系进行了一氧化碳／苯乙烯共聚反应，成功地制得了一氧化碳和苯乙烯的线性交替共聚物 聚酮（ＳＴＣＯ） ． 在催化体系组成的筛选过程中，对人工神经网络的应用进行了探索性研究． 首先，采用ＶｉｓｕａｌＣ＋ ＋ １－５２ 编制了ＢＰ神经网络，利用正交实验结果对该神经网络进行了训练，训练误差小于５％ ． 用训练后的网络预测了该催化体系用量及其交互作用与其催化活性之间的关系． 实验及预测结果表明，该网络系统生成的三维图可很好地体现不同催化体系组成对其催化活性的影响． 在此基础上，利用穷举法对较大范围内的催化体系组成进行了预测，确定出较好的催化体系组成范围． 通过进一步优化，该催化体系的最高催化活性达到４２－０ ｋｇ／（ｍｏｌ·ｈ）．Palladium acetate, 1,10 phenanthroline, p toluenesulfonic acid, p quinone and methanol were used as the catalyst in copolymerization of carbon monoxide and styrene, and the linear alternative copolymer—polyketone (STCO) was prepared. An artificial neural network (ANN) was used in the optimization of the catalyst system. First a BP back propagation neural network was designed by Visual C++ 1.52 and trained by orthogonal experimental results, the training error was less than 5%. Then the trained network system was used to anticipate the effect of the amounts of catalyst ingredients and their interaction on the catalytic activity. The network produced a three dimensional graph that can express the relationship between the catalytic activity and the composition of the catalyst. The catalytic activity in a large range of the catalyst composition can be anticipated. Finally, the composition ranges of the catalyst with higher catalytic activity were found. With the help of the network, the highest catalytic activity of 42 0 kg/(mol·h) was obtained.

关 键 词：一氧化碳 苯乙烯 共聚 催化剂 优化 聚酮 STCO 

分 类 号：O633.1[理学—高分子化学] O643.36[理学—化学]