EVALUATION OF DIFFERENT INJECTION TECHNIQUES IN THE GAS-CHROMATOGRAPHIC DETERMINATION OF THERMOLABILE TRACE IMPURITIES IN A DRUG SUBSTANCE

A gas chromatographic method for the determination of an epoxide in a drug substance was validated. Hot splitless injection of the sample ex tract was identified as a critical step with unsatisfactory repeatabil ity. Therefore, the gas chromatographic procedure was re-evaluated usi ng an instrument equipped with an on-column inlet and a split-splitles s inlet with a programmable column head pressure. Further, the re-eval uated method also included the determination of the corresponding chlo rohydrin that might occur as a reaction product of the epoxide. Three different sets of parameters were chosen: splitless injection using th e same parameters as were used in the validation experiments, splitles s injection with a high initial column head pressure and cool on-colum n injection. Generally, the precision of repeated injections of standa rd solutions and of the analysis of spiked sample extracts was improve d using the newer instrument. On-column injection gave the best result s in terms of both precision and absolute peak areas. The limits of de tection and quantification for the overall method were 0.09 and 0.29 m u g/g, respectively, for the epoxide and 0.09 and 0.31 mu g/g, respect ively, for the chlorohydrin. On hot splitless injection the chlorohydr in formed the epoxide, and also losses of the epoxide were observed. O wing to a shorter residence time in the hot injector block, a high ini tial column head pressure could successfully reduce the degradation of the analytes. However, the precision was not improved because of occa sional leakage of the septum immediately after injection, which result ed in uncontrolled losses and discrimination.