椭偏成像生物传感器是将光学椭偏技术与显微成像相结合而形成的一种新的检测方法，它具有免标记、高灵敏度、高通量等优点，在分子间相互作用研究、表面性质表征、生物标志物检测等领域具有广泛的应用。椭偏成像生物传感器是以一束偏振光为探测光，当其照射于样品表面，由于样品的调制作用改变了光波的偏振状态，其反射光将会携带样品的相关信息。通过测量光波复反射率的变化能够计算出样品的厚度和折射率等信息，进而推知样品的质量面密度的空间分布，所以，椭偏成像生物传感器的信号读出与待测物的质量面密度成正比关系。而对于低分子量分析物，由于其分子量比较小，质量面密度比较低，使其相应检测信号比较弱。因此，传统的椭偏成像生物传感器在低分子量分析物检测方面具有一定的局限性，特别是对复杂基体中痕量成分的检测，其灵敏度很难达到实际检测的要求。针对这一重要科学问题，本论文从以下几个方面开展了相关探索研究：1）从理论层面分析和探讨了影响传感器检测信号的相关因素；2）分别构建了基于双抗夹心、竞争吸附、高密度配基的信号放大体系，并建立了对临床诊断标志物（降钙素原、前列腺特异性抗原、糖化血红蛋白、糖化白蛋白）和环境污染物（赭曲霉毒素A、多氯联苯72、多氯联苯77）的免标记椭偏成像生物传感分析方法，且应用于实际样本的检测；3）建立了椭偏成像生物传感器在双抗夹心信号放大体系中非特异性信号的消除方法。研究结果有效地提升了椭偏成像生物传感器对低分子量分析物检测的灵敏度，拓宽了其应用范围，具有重要的理论意义和实际应用价值。 本论文具体研究结果如下： 1. 椭偏成像生物传感器信号放大理论研究：根据椭偏成像生物传感器在非零模式下的工作原理，通过理论推导得到了与其检测信号相关的主要因素（包括分子质量、分子浓度、配基密度、平衡解离常数），并探讨了各因素的变化对检测信号的影响。在此基础上，设计了适用于椭偏成像生物传感器的信号放大方法，并在理论层面对其可行性进行了讨论，为提高椭偏成像生物传感器对低分子量分析物检测的灵敏度奠定了理论基础。 2. 基于双抗夹心信号放大免标记椭偏成像生物传感器的构建及应用研究：针对低分子量分析物，使其捕获抗体和检测抗体与待测物形成双抗夹心复合物，利用检测抗体所带来的质量面密度的增加实现信号放大。同时，通过信号转换消除非特异性吸附，进一步提高方法的灵敏度。在此基础上，构建了免标记椭偏成像生物传感器对低分子量分析物（降钙素原和前列腺特异性抗原）的检测方法。该方法对降钙素原的检测范围为0.125-128 ng/mL，检测限为0.081 ng/mL，其灵敏度与信号放大前相比提高了近12倍；对总前列腺特异性抗原和游离前列腺特异性抗原的检测范围分别为0.5-256 ng/mL和0.1-36.5 ng/mL，检测限分别为0.126 ng/mL和0.092 ng/mL。该方法还应用于批量血清样品中降钙素原和前列腺特异性抗原的检测，其结果与临床检验方法具有很好的匹配度。检测结果还表明，非特异性吸附的消除有效提升了方法的准确度。另外，采用上述方法类似的原理，建立了糖化血红蛋白（%）与糖化白蛋白（%）的检测方法，总血红蛋白和总白蛋白的检测范围分别为0.05-0.2 μg/mL和0.2-1.5 μg/mL，检测限分别为0.645 ng/mL和17.496 ng/mL。该方法在检测时无需进行信号与浓度的转换，能够一步直接实现血液样品中糖化蛋白百分比的检测。 3. 基于竞争吸附置换信号放大免标记椭偏成像生物传感器的构建及应用研究：以免疫反应的竞争法为基础，当在基质表面加入待测样品与抗体时，基质表面修饰的待测物和样品中待测物与抗体竞争性结合，此过程有效增大了基质表面的质量面密度，实现了信号放大。在此基础上，建立了环境污染物赭曲霉毒素A的检测方法，其检测限可达0.489 ng/mL，检测范围为0.5-100 ng/mL，该方法对实际样品（小麦样品）中赭曲霉毒素A的加标检测结果与试剂盒方法具有良好的相关性。 4. 基于高密度配基信号放大免标记全内反射椭偏成像生物传感器的构建及应用研究：使用核酸适配体代替抗体作为配基，核酸适配体由于尺寸较小，其数量在单位面积上较抗体有明显增加，配基密度的增大能够有效提高体系质量面密度从而实现信号放大，它还能够与夹心方法相结合以达到二次信号放大的效果。基于上述原理，建立了环境污染物赭曲霉毒素A、多氯联苯72与多氯联苯77的检测方法。该方法对赭曲霉毒素A的检测限达66 pg/mL，检测范围为0.1-20 ng/mL；多氯联苯72与多氯联苯77的检测限分别为0.915 pg/mL和3.671 pg/mL，检测范围分别为1-1000 pg/mL和10-1000 pg/mL。目标分析物在实际样品（红酒样品，水样）中的加标实验也得到了满意结果。 综上，本文在理论研究的基础上，通过采用不同的信号放大策略，建立了椭偏成像生物传感器对复杂基体中低分子量分析物的系列检测方法，研究工作为生物传感器的构建提供了新的思路和方法。

The imaging ellipsometry (IE) biosensor is a detection method which combines optical ellipsometry and microscopic imaging, and has the advantages of free labelling, high sensitivity, and high throughput. It is widely used in the realms of intermolecular interaction, surface characterization, and biomarker detection. IE biosensor uses a beam of polarization light to irradiate the surface of the sample. Under its modulation, the polarization state of the light wave will change and reflected light would carry the information of the sample. By measuring the change of the complex reflectance, thickness and refractive index of the sample can be calculated, helping to further infer the spatial distribution of the mass density. In this way, the output signal of the IE biosensor has proportional relationship to the mass density. For low molecular weight analyte, its mass density is rather small, resulting in a low detection signal. Thus, traditional IE biosensor has difficulties in low molecular weight analytes detection for its sensitivity hardly meets the requirements of the practical applications, especially for those with trace amount in complex matrix. In response to the important scientific issue, this paper has carried out researches from the following aspects: 1) Analyzed and discussed the related factors influencing detection signals from the theoretical level. 2) Constructed label-free IE biosensors with signal amplification systems based on double-antibody sandwich, competitive adsorption, and high-density ligands for detections of clinical diagnostic markers (procalcitonin (PCT), prostate specific antigen (PSA), glycated hemoglobin (HbA1c), glycated albumin (GA)) and environmental pollutants (ochratoxin A (OTA), polychlorinated biphenyl 72 (PCB72), polychlorinated biphenyl 77 (PCB77)). And successfully applied the methods for actual sample detections. 3) Established the method for eliminating nonspecific signals in sandwich signal amplification system of the IE biosensor. The research results effectively improve the sensitivity of IE biosensor for the detection of low molecular weight analytes, broaden the scope of its application and have important significance both in theory and practical application. The research results of this paper are as follows: 1. Theoretical study on signal amplification of the IE biosensor: According to the working principle of the IE biosensor in off-null mode, the main factors related to its detection signal (including molecular weight, molecular concentration, ligand density, and equilibrium dissociation constant) were obtained through theoretical derivation. The influence of various factors on the detection signal has been discussed. Based on the results, we designed signal amplification systems for the IE biosensor and discussed the feasibility, laying a theoretical foundation for improving the sensitivity of the IE biosensors for detection of low molecular weight analytes. 2. Development and application of label-free IE biosensor based on sandwich signal amplification: For low molecular weight analytes, the capture antibody and the detection antibody formed a double-antibody sandwich complex with the analyte. The signal amplification has been realized by the increase of the surface mass density caused by the detection antibody. In addition, nonspecific adsorption has been eliminated by signal conversion, further improving the sensitivity of the method. A method for detecting low molecular weight analytes (PCT and PSA) was constructed using a label-free IE biosensor. The detection range of PCT was 0.125-128 ng/mL and the limit of detection (LOD) was 0.081 ng/mL. The sensitivity of the method was 12 times higher than the traditional IE biosensor. For total PSA and free PSA, the detection ranges were 0.5-256 ng/mL and 0.1-36.5 ng/mL, respectively. And the LODs were 0.126 ng/mL and 0.092 ng/mL, respectively. The method has also been applied to the detection of PCT and PSA in batches of serum samples, and the results had a good agreement with the clinical test method. The test results also showed that the elimination of nonspecific adsorption effectively improved the accuracy of the method. Additionally, using the similar principle, a method for detecting HbA1c (%) and GA (%) has been established. For total HbA1c and total HSA, the detection ranges were 0.05-0.2 μg/mL and 0.2-1.5 μg/mL, respectively. And the LODs were 0.645 ng/mL and 17.496 ng/mL, respectively. The method did not require conversion of signal and concentration during detection and can directly detect the percentage of glycated protein in the blood sample in one step. 3. Development and application of label-free IE biosensor based on competitive adsorption signal amplification: The detection principle is based on the competition method of immune reaction. When sample and antibody were added to the surface of the substrate, the analyte on the surface of the substrate and that in the sample would compete for the antibody, and the process effectively increased the surface mass density of the substrate, realizing the signal amplification. We have established a method for detecting environmental pollutants OTA and the LOD for OTA was 0.489 ng/mL with a detection range of 0.5-100 ng/mL. The method had a good correlation with the kit in the spiked detection of OTA in the wheat sample. 4. Development and application of label-free IE biosensor based on high-density ligands signal amplification: The nucleic acid aptamer has been used as the ligand instead of the antibody. Because the size of the aptamer is rather small, compared with the antibody, the amount of aptamer per unit area would increase significantly. The increase of the ligand density can effectively increase the mass density of the system and realize signal amplification. This system can also be combined with a sandwich method to achieve secondary signal amplification. Based on this principle, a method for detecting environmental pollutants OTA, PCB72, and PCB77 has been established. The LOD for OTA was 66 pg/mL with a detection range of 0.1-20 ng/mL. For PCB72 and PCB77, the LODs were 0.915 pg/mL and 3.671 pg/mL, respectively. The detection ranges were 1-1000 pg/mL and 10-1000 pg/mL, respectively. The spiked experiments on the analytes in actual samples (red wine samples, water samples) have got satisfactory results. In summary, based on the theoretical research, by using different signal amplification strategies, a series of detection methods of the IE biosensors for low molecular weight analytes in complex matrices have been established. The research work provides new ideas and methods for the construction of biosensors.