engleski

Application of QSPR modeling for development of sustainability policy based ion chromatography methodologies for environmental monitoring

The concept of sustainability has generated much discussion and has been incrementally induced in environmental management thinking. However, implementation of these changes in industrial monitoring processes has progressed at a slow pace. One reason for this is the enduring myth that economic profitability must always be sacrificed to achieve environmental goals. However, innovative methodologies have been developed that contribute to improving both the environmental and economic corporate bottom line. This paper highlights examples of such methodologies in ion chromatography method development for environmental monitoring. This work focuses on molecular modeling and artificial intelligence in ion chromatography method development with pulse amperometic detection. The core of isocratic elution prediction is QSPR model developed using 22 different carbohydrates and all the structures were optimized using AM and PM1 algorithm. The genetic algorithms were optimized (number of mutations, crossover function) in terms of obtaining optimal dimensionality reduction. Feed forward and cascade forward back propagated artificial neural network methodology were applied and optimized to insure high accuracy of QSRR model prediction. Isocratic elution information was later transferred to gradient elution mode using numerical integration model. The results show that none of developed models includes systematic error, and the obtained predictions (RMSEP=11.66, 10.67, and 7.10 %, respectively) indicates good prediction ability. That should justify possible implementation of developed methodology in sustainably policy based routine environmental monitoring using ion chromatography.

environmental analysis; QSPR modeling; sustainability policy; ion chromatography

nije evidentirano