화학공학소재연구정보센터
Chemical Engineering Journal, Vol.164, No.2-3, 268-274, 2010
Labscale fluidized bed granulator instrumented with non-invasive process monitoring devices
Fluidized bed granulation is a common size enlargement process in the pharmaceutical industry, in which fine powder is agglomerated using a liquid binder to obtain larger granules. However, fluidized bed granulation is a complex process by nature, being difficult to control due to the strong interactions between many process variables, such as moisture content and granule size. In order to control the process via correct pathway, one should be able to monitor the process, i.e. measure the process variables in-line. In this study, three Process Analytical Technology (PAT) devices, i.e. an acoustic emission sensor, a flash topography particle size analyzer and multi-point NIR probes, were developed and instrumented into a labscale fluidized bed granulator for simultaneous granulation process monitoring. Parallel techniques were used for characterizing the granule size distribution and moisture content of granules during fluidization. Cellets(R), protease granules and caffeine formulation were used as samples. Granulation was carried out in a custom made modular top spray granulation chamber. The granule size values obtained with the acoustic emission and flash topography particle size analyzer were in good agreement with the values measured with offline reference methods. Multi-point NIR (eight probes) and acoustic emission methods were able to detect the three granulation phases, mixing, agglomeration and drying. Single location monitoring does not reflect the heterogeneous sample during granulation, and thus, more information can be obtained by multi-point NIR. In general, the sensitivity of the multi-point NIR technique is susceptible to the fouling of probe windows whereas the acoustic emission technique is sensitive to background fluidizing air flows as well as external interference. The most informative data can be obtained when multiple PAT techniques are applied simultaneously for in-line process monitoring. (C) 2010 Elsevier B.V. All rights reserved.