Quantitative real-time monitoring of dryer effluent using fiber optic near-infrared spectroscopy

This paper describes a method for real-time quantitation of the solvents ev aporating from a dryer. The vapor stream in the vacuum line of a dryer was monitored in real time using a fiber optic-coupled acousto-optic tunable fi lter near-infrared (AOTF-NIR) spectrometer. A balance was placed in the dry er, and mass readings were recorded for every scan of the AOTF-NIR. A parti al least-squares (PLS) calibration was subsequently built based on change i n mass over change in time for solvents typically used in a chemical manufa cturing plant. Controlling software for the AOTF-NIR was developed. The sof tware collects spectra, builds the PLS calibration model, and continuously fits subsequently collected spectra to the calibration, allowing the operat or to follow the mass loss of solvent from the dryer. The results indicate that solvent loss can be monitored and quantitated in real time using NIR f or the optimization of drying times. These time-based mass loss values have also been used to calculate "dynamic" vapor density values for the solvent s. The values calculated are in agreement with. values determined from the ideal gas law and could prove valuable as tools to measure temperature or p ressure indirectly. (C) 2000 Wiley-Liss, Inc. and the American Pharmaceutic al Association J Pharm Sci 89: 1180-1186, 2000.