Determination of cefalexin in capsules using near infrared diffuse reflectance spectrophotometry

OBJECTIVE Near infrared diffuse reflectance spectrophotometry (NIRDRS) and chemometrics were applied to the quantitative analysis of cefalexin in capsules. METHODS MPLS (modify partial least squares) regression models were set up using a calibration set (25 samples).The validation set was set up by 15 samples. Applying this model to predict the validation set and detects the samples.RESULTS The determination coefficients R 2 was 1.00 as indicated from the cross-validation, the true mean predictive error RMSECV (root mean square error of cross validation ) was 0.40 .The mean prediction error RMSEP (root mean square error of prediction) obtained from the validation was 0.59 , the determination coefficients R 2 = 0.999 .The correlation coefficient of the true value and prediction value was 0.999 .The average recovery of the prediction set was 100.4% (RSD= 0.50% , n =15). CONCLUSION NIRDRS permits quantitative non-destructive determinations of the active constituents in pharmaceutical preparations with little or even no sample treatment. NIRDRS method is faster and easier than other methods.