通过集成在线近红外光谱仪、实时图像采集与处理系统和群体粒数衡算模型,开发了高剪切湿法造粒过程的决策支持系统(Decision support system,DSS).利用近红外光谱仪和图像系统实时测量多个过程变量和产品质量指标,包括粉体混合均匀度、颗粒粘合剂含量、粒径分布以及团聚、破裂行为等,能够快速确定过程操作空间.同时,由过程分析平台得到的信息经分析处理后输入工艺过程模型模块,用于估算和校准群体粒数衡算模型中的团聚和破裂速率常数,以此持续提高模型精度.另一方面,模型可以指导实验体系寻找最优操作空间.该决策支持系统成功应用到了以微晶纤维素和甘露醇为原料,3％聚乙烯吡咯烷酮水溶液为粘合剂的高剪切湿法造粒过程中,对两个粘合剂喷淋速率下的造粒过程进行监测.DSS认定粘合剂喷淋过程分为四个阶段:润湿期、成核期、快速生长期和慢速生长期.不同阶段之间的分界点与粘合剂喷淋速率有关.在较高喷淋速率下,颗粒进入成核期和快速生长期所需粘合剂较少,但是对颗粒最终粒径无明显影响.此外,通过近红外光谱测定混合均匀度,确定了粉体的混合终点.该DSS系统将基于过程分析技术的高效实验和过程模拟结合,可以快速确定操作空间以及颗粒的生长行为,实时提供大量数据用于持续提高模型精确度和稳健性,提高造粒过程的优化效率.%A decision support system (DSS)is developed for a high shear wet granulation process that has integrated near in-frared spectroscopy (NIR),on-line imaging and image analysis,and population balance (PB)models.The integrated PAT (process analytical technology)based on NIR and imaging allows multiple process variables and granule properties to be characterized in real-time,including mixing uniformity,binder content,granule size and size distribution,as well as granule growth,aggregation and breakage,and facilitates rapid identification of the operational spaces.The PAT measurements also provide data for estimation of the rate constants of kinetics kernels in the PB models and continuous improvement of the models.The models on the other hand can be used to guide PAT based experiments in searching for optimum operational space.A case study applying the DSS to the granulation of microcrystalline cellulose and mannitol powders using 3％ aqueous solution of polyvinylpyrrolidone as the binder is described.The spraying period has shown to experience four phases:wetting,nucleation,rapid growth and moderate growth,but the boundaries separating the phases were different depending on the spray rate.At a high spraying rate,the binder contents at which the growth stepped into nucleation regime and rapid growth regime were lower.However,it didn't show significant influence on the median size of the granules formed.In addition,the mixing uniformity was monitored using NIR,allowing identification of the end point of pow-der mixing.The DSS combined the effective experiments based on PAT with process modelling,which enabled rapid characterization of the operational spaces and granule growth behavior.It also provided sufficient data collected real-time during the process to im-prove the robust and accuracy of the model continuously,consequently increasing the efficiency of optimization in granulation.
展开▼
