Research On Size Distribution Estimation Of Magnetic Nanoparticles

In this article, scientific problems related to Magnetic Nanoparitcle (MNP, also known as Ferrofluid) are discussed thoroughly, including the characterization of MNP's particle size and its variation during the MNP's thermal dynamic process. The research on MNP's particle size and the changes of of MNP polymers'magnetic parameters during the seperation caused by heating, are fundamental to in vivo or in vitro magnetic biomarker analysis technology.For the particle size information characterization, this work combines the ill-conditioned equation's analytic solving methods and the quantitative optimization techniques, to decrease the ill-posed feature's influence on particle size distribution function's solution and increase the accuracy of the characterization. One typical feature of the nano-scale measurements technology is that the result is usually a distribution function, instead of a figure. The magnetic dynamic model of the MNP is a functional equation of the distribution function of its particle size, while the discrete model is a matrix equation of the particle size distribution function. In preprocessing of the magnetization curve measurement data, the removal of water-based background noise was employed to cancel the anti-magnetism of the MNP colloid caused by water solvent to achieve a better accuracy. Singular Value Decomposition (SVD), Tikhonov canonical SVD and linear constrained quadratic programming method were applied in the solution of inverse problem. The prior information of the magnetization process is utilized to improve the stability of solution. It is found that the Tikhonov canonical SVD method solved the problem of artificial oscillation caused by ill-posed feature. Based on the signal discretization and quantization, this work analyzes the minimum sampling points for the characteristic magnetization curve. The sampling points are deployed according to the optimal quantization theory. The experimental results show that, the condition number of the Langevin super-paramagnetic magnetization numerical equation can be reduced remarkably which consequently suppresses the artificial oscillation signal caused by ill-posed feature. With accurate particle size information, two particles'agglomeration is discovered for the first time, from the perspective of size analyzing by the particle size characterization technology based on magnetization curve. To study the impact of particle size on the physical properties of MNP colloidal solution, the relationship curve between inverse susceptibility versus temperature is used to explore the temperature changing process of MNPs and to analyze the agglomeration of MNP particles (mainly the dimers). For the first time, temperature controlled segregation of dimers is discovered in the perspective of magnetic susceptibility. Such a phenomenon can be described as an unstable and invertible state in MNP colloid, which is controlled by temperature. Above a critical temperature T*, only monomers exist; below T*, monomers and polymers coexist. The existence of dimers'temperature controlled segregation phenomenon indicates that, in the biomedical temperature range of 300-370 K, the magnetic susceptibility of MNP colloid may not be a constant, but changed with temperature and density. The experimental data matches well with Prof. Morais'theoretical model of dimer. This conclusion also is supported by the discovery of dimer by measurement technology introduced in Chapter four.