PDF(853 KB)
PDF(853 KB)
PDF(853 KB)
基于人工神经网络的合成脱水长春碱工艺优化研究
Process optimization of the synthesis of anhydrovinblastine using artificial neural networks
Anhydrovinblastine, which is a crucial intermediate in the synthesis of vinorelbine and vinflunine, can be synthesized from vinblastine, but the reported yields are not sufficiently high for practical application. The synthetic conditions for this reaction have been optimized using orthogonal tests and artificial neural networks. A network model was established by improving backpropagation of an artificial neural network (BP-ANN) with 3 layers, and was shown to be correct. A wider range of factors was then introduced into the BP-ANN model in order to obtain the optimal conditions as follows: the vinblastine was dissolved in DMF with a concentration of 15mmol/L; the reagent was oxalyl chloride, C2O2Cl2, with the stoichiometric proportion between the reagent and vinblastine being 50∶1; the reaction temperature was 4℃ and the duration was 12 hours; the extractant phase was the mixture of ether and methylene chloride (volumn ratio 10). The maximum yield of anhydrovinblastine obtained under these conditions was 70.88%.
[1]Fahy J. Modifications in the upper or velbenamine part of th
e vinca alkaloids have major implications for tubulin interacting activities［J］. Current Pharmaceutical Design, 2001, 34(7): 1181-1197.
［2］Fahy J, Thillaye du Boullay V, Bigg D C H. New method of synthesis of vinca alkaloid derivatives［J］. Bioorganic & Medicinal Chemistry Letters,
2002, 12(3): 505-507.
［3］Gueritte F, Langlois N, Langlois Y， et al. 20'Deethylanhydrovinbla
stine［J］. Journal of Organic Chemistry, 1981, 46(1): 5393-5395.
［4］Hardouin C, Doris E, Rousseau B. Selective deoxygenation of leurosine: concise access to anhydrovinblastine［J］. Journal of Organic Chemistry, 2002,
67(2)： 6571-6574.
［5］Langlois N, Gueritte F, Langlois Y, et al. Application of a modificat
ion of the polonovski reaction to the synthesis of vinblastinetype alkaloids［
J］. Journal of the American Chemical Society, 1976, 98(1): 7017-7024.
［6］陈永江，陈洪明，李莉，等. 酒石酸长春瑞滨的合成［J］. 中国医药工业杂志，19
99, 30(1): 6-8.
［7］陈允平，王旭蕊，韩宝亮. 人工神经网络原理及其应用［M］. 北京：中国电力出版
社，2002: 1-3.
［8］党建武. 神经网络技术及应用［M］. 哈尔滨工程大学出版社，2000.［9］吴启迪，汪镭. 反馈式神经网络智能控制［M］. 上海：上海科技教育出版社, 2004: 6-7.〖ZK)〗
［10］Lacher R C. Backpropagation learning in expert networks［J］. IEEE Transaction on Neural Networks, 1992, 1(3): 62-72.
［11］Liu Jiangto, Chang Hongbin, Xu Zuyao, et al. Prediction of the flow
stress of highspeed steel during hot deformation using a BP artificial neural
network［J］. Journal 〖JP2〗of Materials Processing Technology, 2000, 103(2): 200-〖JP〗205.
［12］胡德文，王正志，王耀男，等. 神经网络自适应控制［M］. 长沙：国防科技大学
出版社，2006.
［13］包俊敏，夏德生，陈云华. 合成长春瑞滨的方法：CN 1552716A［P］. 20041208.
/
| 〈 |
|
〉 |
