Li Weiyong, Johnson Mark C, Bruce Rick, Rasmussen Henrik, Worosila Gregory D
Analytical Development, Johnson and Johnson Pharmaceutical Research & Development, LLC, Spring House, PA 19477, USA.
J Pharm Biomed Anal. 2007 Jan 17;43(2):711-7. doi: 10.1016/j.jpba.2006.07.015. Epub 2006 Aug 14.
Online blend uniformity study was conducted with a near infrared (NIR) sensor and a simulated formulation consisting of acetaminophen and four excipients. Quantitative calibration models were developed and validated for the sensor and assay results were obtained in real time for acetaminophen and three excipients. Mechanical thief samples were also collected during the study. The samples were analyzed offline by a bench-top near infrared spectrometer and used as reference. Comparison of the online and offline data shows a significant difference in standard deviation for acetaminophen and excipients. R.S.D. data calculated from the real-time assay values for acetaminophen was 3.5-13.2-fold lower than those from the offline results. The cause for the discrepancy is believed to be the large beam size of the online sensor. A simple complete-random-mixture model was used to explain the discrepancy. It is concluded that beam size is an important factor in quantitative online blending uniformity studies.
使用近红外(NIR)传感器和由对乙酰氨基酚及四种辅料组成的模拟制剂进行了在线混合均匀性研究。针对该传感器开发并验证了定量校准模型,并实时获得了对乙酰氨基酚和三种辅料的含量测定结果。在研究过程中还采集了机械取样器样品。这些样品通过台式近红外光谱仪进行离线分析,并用作参考。在线和离线数据的比较显示,对乙酰氨基酚和辅料的标准偏差存在显著差异。根据对乙酰氨基酚实时含量测定值计算的相对标准偏差(R.S.D.)数据比离线结果低3.5至13.2倍。据信,差异的原因是在线传感器的光束尺寸较大。使用简单的完全随机混合模型来解释这种差异。得出的结论是,光束尺寸是定量在线混合均匀性研究中的一个重要因素。