M. V. Guterres, C. C. Felício, M. M. C. Ferreira, J. C. Menezes, "Chemometric Methods For Monitoring Batch Chemical Reactions With Spectroscopy". Lisbon, Portugal, 20-23/09/2004: Ninth International Conference on Chemometrics in Analytical Chemistry (CAC IX, CAC-2002). Program and Abstracts Book, (2004). Poster 184. Section: Poster Session 5 - Chemometrics & Process / Industrial Applications.
CHEMOMETRIC METHODS
FOR MONITORING BATCH CHEMICAL REACTIONS WITH SPECTROSCOPY
GUTERRES, M.V; FELÍCIO, C.C; FERREIRA, M.M.C; MENEZES, J.C
State Univ. of Campinas, Instituto de Química, CAMPINAS, 13083-970, BRAZIL, Phone: 55 19 3788-3102, Fax: 55 19 3788-3023
ABSTRACT
CHEMOMETRIC METHODS FOR MONITORING BATCH CHEMICAL REACTIONS WITH SPECTROSCOPY. Guterres, M.V1; Felício, C.C2; Ferreira, M.M.C3; Menezes, J.C4 . 1UNICAMP, Instituto de Química, Caixa Postal 6154, CEP: 13084 – 971, Campinas - SP, Brasil. guterres@iqm.unicamp.br 2Instituto Superior Técnico, Departamento de Engenharia Química, Av. Rovisco Pais, P – 1049-001, Lisboa, Portugal. candida.felicio@ist.utl.pt 3UNICAMP, Instituto de Química, Caixa Postal 6154, CEP: 13084 – 971, Campinas - SP, Brasil. marcia@iqm.unicamp.br 4Instituto Superior Técnico, Departamento de Engenharia Química, Av. Rovisco Pais, P – 1049-001, Lisboa Portugal. Cardoso.menezes@ist.utl.pt Chemometric methods based on Multivariate Curve Resolution (MCR) and Analysis of Principal Components (PCA) are proposed for the diagnostic and statistical monitoring of chemical reactions followed by Near and Mid Infra-Red spectroscopy techniques. MCR method was used for characterization and diagnostic of the chemical reactions at the end of the batch. This method, allows detecting whether a batch of reaction is in agreement or not with the reference batch. PCA models and multivariate control charts are proposed for statistical on line monitoring of batch chemical reactions. Both chemometric methods were tested for the development of a system for experimental simulation of chemical reactions. For the cases investigated, it was observed that analysis with MCR method yields better results when applied to MIR data than NIR due to the better selectivity of MIR absorption bands for the substances used in the simulation. Online Statistical monitoring with PCA models was successful for both spectroscopic methods. Acknowledgments: FAPESP
Keywords: mcr, pca, batch reactions
Requested contribution format: poster
Corresponding author: MÁRCIA M. C. FERREIRA (marcia@iqm.unicamp.br)