The use of total reflectance X-ray fluorescence (TXRF) for the determination of metals in the pharmaceutical industry.

The control of residual metals in active pharmaceutical ingredients (API's) and intermediates is critical because of their potential toxic effects. A variety of technologies are available to measure residual metals in pharmaceutical compounds including, AAS, ICP-AES, and ICP-MS. The newest technology is total reflectance X-ray fluorescence spectroscopy (TXRF) which uses primary X-rays to excite atoms which then emit secondary X-rays. The emitted X-rays are characteristic of the individual elements present, and the intensities of the emitted X-rays are proportional to the concentrations of the elements present in the sample. The benefits of TXRF are that it is essentially unaffected by matrix effects, is very sensitive (ppb's), requires small amounts of sample (5-10 mg), and requires very little sample preparation time. During this study, TXRF was used to quantitatively measure residual metals in API's and intermediates and such topics as sample preparation, sensitivity, linearity, reproducibility and accuracy are discussed. The results obtained by TXRF were compared with those obtained by ICP-MS for the same samples for Pd and Cu measurement, and statistical analysis indicated that the results obtained by the two technologies are equivalent at the 95% confidence level. A comparison is also made of the capabilities of the instruments using a tungsten (W) or a molybdenum (Mo) source for excitation. Both instruments could be used for the quantitative determination of residual metals in pharmaceuticals.

Development and characterization of "push-pull" sampling device with fast reaction quenching coupled to high-performance liquid chromatography for pharmaceutical process analytical technologies.

A push-pull sampling system interfaced on-line to high-performance liquid chromatography (HPLC) was developed for micro-volume real-time monitoring of reaction mixtures. The device consists of concentric tubes wherein sample was continuously withdrawn through the outer tube and reaction quenchant continuously delivered through a recessed inner tube. The device allowed sampling rates of 0.1-6.0 µL/min from a reaction vessel and stopped the reaction by passive mixing with quenchant to preserve the conditions observed in the reaction vessel. A finite element model of the system showed that reaction mixtures could be completely mixed with quenchant within 4.3s at a flow rate of 1.0 µL/min. The model also showed that an offset distance of 1mm between the push capillary and sample capillary tips is sufficient to avoid leakage of quenchant/diluent into the bulk sample for push flow rates up to 95% of the pull flow rate. The maximum relative push flow rate was determined to be 90% of the pull flow rate experimentally. Delay between sampling and delivery to the HPLC was from 111±3s to 317±9s for pull flow rates from 1.0 to 3.0 µL/min in agreement with expected delays based on tubing volume. Response times were from 27±1s to 52±6s over the same flow rate range. The sampler was tested to determine the effects of sample viscosity. The sampler was also used to demonstrate periodic sampling capabilities. As a test of the system, it was used to monitor the base-catalyzed hydrolysis of aspirin for 1.5h, demonstrating its utility for monitoring an ongoing reaction.