Establishment of Universal Quantitative Models for Determination of Amoxicillin Sodium and Clavulanate Potassium for Injection Using Near-infrared Reflectance Spectroscopy

OBJECTIVE To develop a near-infrared (NIR) reflectance spectroscopic quantitative method for rapid and non-destructive determination of amoxicillin sodium and clavulanate potassium for injection. METHODS Using cluster analysis method, diffuse reflectance spectra of 26 batches of simulated samples and 40 batches of commercial samples from different manufactures were divided into calibration and validation set. The calibration model was developed using partial least squares (PLS) algorithm. RESULTS Three quantitative methods were explored at the concentration ranges of 21.28%-75.57% for amoxicillin, 2.67%-15.85% for clavulanate potassium, 0.46%-15.7% for water. The root mean square errors of cross validation (RMSECV) and the root mean square errors of prediction (RMSEP) were 1.42% and 1.54% for amoxicillin, 0.56% and 0.71% for clavulanate, 0.11% and 0.14% for water, respectively. CONCLUSION Three nondestructive quantitative NIR methods used in this study are applicable for rapid analysis of amoxicillin sodium and clavulanate potassium for injection from different manufactures.