The performanceofdifferentchemometricapproacheswasevaluatedinthespectrophotometricde- termination ofpharmaceuticalmixturescharacterizedbyhavingtheamountofcomponentswithavery high ratio.Principalcomponentregression(PCR),partialleastsquareswithonedependentvariable (PLS1)ormulti-dependentvariables(PLS2),andmultivariatecurveresolution(MCR)wereappliedtothe spectral dataofaternarymixturecontainingparacetamol,sodiumascorbateandchlorpheniramine (150:140:1,m/m/m),andaquaternarymixturecontainingparacetamol,caffeine,phenylephrineand chlorpheniramine (125:6.25:1.25:1,m/m/m/m).TheUVspectraofthecalibrationsamplesintherangeof 200–320 nmwerepre-treatedbyremovingnoiseanduselessdata,andthewavelengthregionshaving the mostusefulanalyticalinformationwereselectedusingtheregressioncoefficients calculatedinthe multivariatemodeling.Allthedefined chemometricmodelswerevalidatedonexternalsamplesetsand then appliedtocommercialpharmaceuticalformulations.Differentdataintervals, fixedat0.5,1.0,and 2.0 point/nm,weretestedtooptimizethepredictionabilityofthemodels.Thebestresultswereobtained using thePLS1calibrationmodelsandthequantification ofthespeciesofaloweramountwassig- nificantly improvedbyadopting0.5datainterval,whichshowedaccuracybetween94.24%and107.76%.
Optimization of wavelength range and data interval in chemometric analysis of complex pharmaceutical mixtures
De Luca M;IOELE, Giuseppina;RAGNO, Gaetano
2016
Abstract
The performanceofdifferentchemometricapproacheswasevaluatedinthespectrophotometricde- termination ofpharmaceuticalmixturescharacterizedbyhavingtheamountofcomponentswithavery high ratio.Principalcomponentregression(PCR),partialleastsquareswithonedependentvariable (PLS1)ormulti-dependentvariables(PLS2),andmultivariatecurveresolution(MCR)wereappliedtothe spectral dataofaternarymixturecontainingparacetamol,sodiumascorbateandchlorpheniramine (150:140:1,m/m/m),andaquaternarymixturecontainingparacetamol,caffeine,phenylephrineand chlorpheniramine (125:6.25:1.25:1,m/m/m/m).TheUVspectraofthecalibrationsamplesintherangeof 200–320 nmwerepre-treatedbyremovingnoiseanduselessdata,andthewavelengthregionshaving the mostusefulanalyticalinformationwereselectedusingtheregressioncoefficients calculatedinthe multivariatemodeling.Allthedefined chemometricmodelswerevalidatedonexternalsamplesetsand then appliedtocommercialpharmaceuticalformulations.Differentdataintervals, fixedat0.5,1.0,and 2.0 point/nm,weretestedtooptimizethepredictionabilityofthemodels.Thebestresultswereobtained using thePLS1calibrationmodelsandthequantification ofthespeciesofaloweramountwassig- nificantly improvedbyadopting0.5datainterval,whichshowedaccuracybetween94.24%and107.76%.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
