Abstract: Near infrared diffuse reflectance spectroscopy with partial least square fitting (NIRSPLS) was applied for nondestructive quantification of the content of Pyrazinamide in Pyrazinamide tablets. A mathematical calibration model of NIRS was proposed using the first derivative spectrum in the wavelength range from 800 to 1200 nm. For this model, the regression coefficient (R) between predicted values and actual values in calibration set is as high as 0.9989 and the root meansquare error of prediction (RMSEP) is 0.563. Results show that this model is reliable with high prediction accuracy and is convenient for use, that can be used as general means in pharmaceutical analysis.

Key words: pharmacochemistry, near infrared spectroscopy, partial least squares, nondestructive analysis, pyrazinamide