Automatic monitoring and quantitative characterization of sedimentation dynamics for non-homogenous systems based on image profile analysis

• Published in: Powder technology 2015-09, Vol.281, p.49-56
• Main Authors: Lu, Xiaolong, Liao, Zuhua, Li, Xue, Wang, Manli, Wu, Li, Li, Haiyan, York, Peter, Xu, Xu, Yin, Xianzhen, Zhang, Jiwen
• Format: Article
• Language: English
• Subjects: Automatic monitoring; Derivative mutation algorithm; Image processing; Non-homogeneous system; Sedimentation dynamics; Suspensions
• ISSN: 0032-5910; 1873-328X
• DOI: 10.1016/j.powtec.2015.04.080

Sedimentation of non-homogeneous systems is the typical phenomenon indicating the physical instability as a key measure to the quality control of the preparation products. Currently, the determination methods for the sedimentation of non-homogeneous preparations are based on manual measurement or semi-quantitative observation, lacking of either automation or quantitative dynamic analysis. The purpose of this research was to realize automatic and quantitative monitoring of the sedimentation dynamics for non-homogenous systems as suspension, emulsions at laboratory level. Non-contact measurement method has been established to determine the sedimentation behaviors in a standard quartz tube for sedimentation, with internal diameter and height 23mm and 215mm, respectively, with controlled temperature and light intensity. As high performance camera has been equipped, the sedimentation images with high spatial and temporal resolution could be acquired, which can continuously capture sedimentation images with the resolution of 2048×2048pixel at a maximum rate of 60slides/s. All the pictures were processed to extract the luminance matrix top-down along the fixed vertical midline of each picture, which implied sedimentation characteristics of the system at the moment the picture was taken. Combining all the luminance matrixes along vertical middle lines of the pictures, a time-luminance matrix profile was obtained. Digital image processing techniques were used to eliminate interference and establish a three-dimensional surface model of the sedimentation dynamics. Then, the derivative mutation algorithm has been developed for the intelligent identification of sedimentation interface with threshold optimization so as to quantitatively analyze the sedimentation dynamics with visualization. The sedimentation curve and sedimentation dynamic equation of the non-homogeneous system were finally outputted by numerical fitting. The methodology was validated for great significance in determinations of small volume samples, parallel control multiple batches, and long period of time automatic measurement. [Display omitted] •Identify sedimentation interfaces of non-homogeneous systems•Quantitatively analyze the sedimentation dynamic behavior•Recording the sedimentation profiles facilitates further analysis.•Great significance in determinations of small volume samples of automatic measurement