Study on the QSPR of Aldehydes and Alkones Based on Their Molecular Structureparameters
|Keywords||aldehydes andketones the molecular structureparameter physicochemical property quantitative structure-property relationships chemometricsmethods|
Quantitative structure-property relationship （QSPR） of compoundshas become an importantbranch of chemometrics, by establishing the QSPR model not only to predict or estimate the natureof the unknown samples, but also to analyse the characteristics ofmolecular structure parameters,which have the largest impact on properties of compounds. Researchers may understand the effectofthe micro-structure of materials on their macro-properties at the molecular leve, which havecertain guidance effect on molecular design. QSPR research has an important theoretic andapplication valuein the field of chemistry, life science and environmental science.This thesis studies thatquantitative structure-property relationship of aldehydes and ketones,theyare important industrial materials、drug materials、synthesis intermediates, on the other hand,they are one of the most importantpollutions in houses and public places, prolonged exposure tothese compounds can lead to headache、dizziness、nauseaand other symptoms, which causes greatharms to the environment and human health.Butthese physicochemical properties is hard to getfrom experiment at present, it is necessary to seek methods to predict them. Therefore, we need tostudy some properties of them, mainly analyzes and investigates its physicochemical properties(boiling point Tbp、molar refractionMR、constant pressure heat capacity Cp、critical temperature Tcand critical volume Vc), this study adopts the method of QSPR modelling.This thesis starting from the molecular structure of the aldehydes and ketones,a customizedmolecular space structural parameters （Yh） was constructed, according to molecular spatialcoordinates （x,y,z） of aldehydes and alkones compounds that obtained by using MM2molecularmechanism to optimize the structure after got three-dimensional configuration using chemicalsoftware chemoffice2010. Multivariate linear regression analysis and BP neural network and RBFneural network analysis were carried out between Yh、other two-dimensional molecular structureparameters （the number of carbon atoms Cnin molecular、functional group site index Soxand Windex） and boiling point Tbp、molar refractionMR、constant pressure heat capacity Cp、criticaltemperature Tcand critical volume Vcand other physical and chemical properties of aldehydes andketones, respectivelyobtained powerful fitting ability and good stability QSPR models. Finally, themodel established is proved by the testing set samples, and the test results show that the establishedQSPR model is of powerful prediction ability. This thesisexperimental datas ofthe physicochemical properties of aldehydes and ketonesaretaken from some literature, which savesa lot of cost、equipment and time of experiment, it will behelpful to research compound relationship between the structure and properties. Finally, systemcluster analysis method based on the molecular structure parameters and its physicochemicalproperties were used to classify aldehydes and ketones.According to the results of classificationhave overall consistency, shows that molecular structure parameters of the selected can exactly、reliablyexplain and manifestthe macroscopic properties of aldehydes and ketones.