On the Immunoassay of Human Immunoglobulin using Surface Plasmon Resonance: Experiment and Numerical Simulation

• Main Authors: Cheng-Han Lin, 林承翰
• Other Authors: Jeng-Shian Chang
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/39507342242079298953

碩士 === 臺灣大學 === 應用力學研究所 === 96 === Surface plasmon resonance (SPR) biosensor has been widely used as the apparatus for the biomolecule detection in the last two decades, since it has several advantages in analyzing the interactions among biomolecules, such as label free, non-destructive, highly sensitive, high throughput, and capable of monitoring dynamic biomolecular interaction in real time. Most of the existing works focused merely on experiments. In this thesis we perform not only the experiments using a home-made angle-detection SPR system but also the numerical simulations based on the FEM software, Comsol Multiphysics v3.3 to study the behavior of the antibody-antigen interaction. In the immunoassay experiments we use human immunoglobulin Human IgG1 and Anti-Human IgG1 as the receptor-ligand pair in PBS buffer solution for the antibody-antigen interaction with the concentration of Anti-Human IgG1 being 50, 25 and 10 , respectively. The response curves of binding reaction are represented as the evolution of SPR angle versus time during the binding of Anti-IgG1 to the immobilized IgG1 on the gold film, under various concentrations of Anti-IgG1. By manipulation of measured data, the association constant and dissociation constant can be extracted in the usual fashion as done in the literature. In numerical simulation, we first perform the transporting analysis of the analyte in the solution along the tubes to obtain the concentration profile of the analyte in the solution at the inlet to the reaction chamber. Then we setup a 3-D model for the reaction chamber and perform the binding reaction simulation based on finite element calculation. The simulated results turn out to be inconsistent with the experimental curves, mainly due to incorrect affinity constants and . A more accurate way to calculate and from the measured data is thus raised. With our corrected and , the simulated curves match quite well with the experimental results.