QD323 : Application of Chemometircs methods to Predict the Retention index of some organic Compounds
Thesis > Central Library of Shahrood University > Chemistry > MSc > 2018
Authors:
Elahe Salamati [Author], Naser Goudarzi[Supervisor], Davood Shahsavani[Supervisor], Mansour Arab Chamjangali[Advisor]
Abstarct: Generally, the amount of sulfur in crude oil, depending on its source, varies from 0.25 to about 7.89 of Weight percent. Generally sulfur is observed both organic and inorganic forms in crude oil. In addition to elemental sulfur more than 200 sulfur organic compounds are identified in the oil, which contains sulfides, Thiols or mercaptans, thiophene, which theophany, and di-benzoate molecules are much more complex. These compounds are known as persistent contaminants. In addition, these compounds are carcinogenic. Various methods, such as gas chromatography, have been proposed to segregate and identify these compounds, which inhibition index compares with standards. Therefore, the inhibition index of a soluble is an important parameter that can be used to identify sulfur heterocyclic aromatic compounds. But on the other hand, these methods are costly and time-consuming. By applying chemo metrics methods, prediction of inhibitory index of compounds is plausible (PASHs) and by using random forest methods (RF) and MARS methods and the artificial neural network SR-ANN and RF-ANN we predicted the inhibitory effect of these compounds. The results indicate that the use of the mentioned statistical method has a good potential to predict the empirical inhibitory index of these compounds, which R^2 results from RF and MARS and SR-ANN and RF-ANN methods are 0/944 and 0/984, 0/971and 0/932 respectively.
Keywords:
#Chemometrics #Random forest #MARS #Artificial Neural Network #Inhibition Indices (index) #Aromatic heterocyclic sulfur compounds Link
Keeping place: Central Library of Shahrood University
Visitor: