Research On The Method And Device For Automated Monitoring Sedimentation Of Non-homogeneous Liquid Preparations

Sedimentation is a downward movement that particles deposited under gravity in the fluid. The speed of the sedimentation is connected with several factors such as the particle size, particle density, fluid density and viscosity, temperature, and fluid state of motion. Determination of sedimentation curve has different meanings in different industries. Considering of different requirements, the sedimentation process can be speeded up or slowed down by adding flocculent, coagulating agent, deflocculating agent and other additives.Sedimentation of non-homogeneous systems is a typical phenomenon, indicating that the physical instability is 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 is to accomplish automatic and quantitative monitoring 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 23 mm and 215 mm, respectively, with controlled temperature and light intensity. With equipped high performance camera, the sedimentation images of high spatial and temporal resolution were acquired, which can continuously capture sedimentation images with the resolution of 2048×2048 pixel at a maximum rate of 60 slides/second.This paper presents two image data processing methods. Method one:image segmentation algorithm mainly with Otsu method and morphological filtering algorithm. The whole target area of each picture can be collected and stored in its height. Even though the program was simple, the amount of calculation was large. The other method: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 when the picture was taken. Combined with 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 method was validated for great significance in determination of small volume samples, parallel control multiple batches, and long period of time automatic measurement.In conclusion, the dynamics of a non-homogeneous system with high precision, quantitative, analytical intelligence was achieved by use of sedimentation automatic monitoring method, which is of great significance in pharmaceutical formulations design and pharmaceutical process.