本文采用双光束干涉曝光的方法制备了金纳米阵列， 在面积为20mm×20mm的玻璃衬底和硅衬底上得到了均匀的光栅阵列、纳米点阵和纳米孔阵结构，用该方法制备LSPR传感的衬底，不仅装置简单成本低，而且可以通过调节曝光时间和显影时间，实现对纳米结构的控制。选择形状均匀良好的金纳米孔阵结构用实验室搭建的LSPR传感测试系统测量了空气、水和浓度分别为5%、10%、15%的盐溶液的透射光谱，经过处理数据，得到灵敏度这一表征LSPR传感测试性能的参数高达800nm/RIU。并且按照电镜看到的尺寸进行了FDTD仿真模拟，得到灵敏度为271.7nm/RIU，远小于实验测得的，而且共振波长也有四五十纳米的差距，可能是由于制备的纳米孔阵结构不是规则的圆柱造成的。同时，还对三种纳米结构进行了FDTD仿真，改变参数，以得到性能最好的结构，指导实验进行。

We employed double two-beam interference lithography to pattern a periodic array of photoresist on Si or glass covered with a film of ITO with the size of 20mm×20mm, followed by thermal evaporation of gold to produce the Au arrays. By varying the time of exposure and development, we can get array of nanoholes, nanodots and gratings with different size. And we can change the angle of the incident light to get different period. So it’s a cheap and convenient way to get different kinds of Au arrays. After that we used the most uniform nanohole arrays with the period of 430nm and the damater of hole about 200nm to get the transmission spectra under the condition of air, water and different concentration (5%,10%,15% ) of NaCl. We find the sensitivity is as high as 800nm/RIU. However, we just get 271.7 nm/RIU by FDTD simulation of the same size of nanohole arrays, and the wavelength of LSPR is tens of nanometers shorter than that of experiment. Maybe it’s because the nanohole in experiment is not a cylinder as in FDTD. Also we employed FDTD to simulate the three kinds of arrays and change some parameters to get the structure with highest sensitivity to guide our experiment further.