论文部分内容阅读
全芯片多参数联合优化是光刻分辨率增强技术的重要发展方向。提出了一种基于粒子群优化(PSO)算法的光源掩模投影物镜联合优化(SMPO)方法。将由像素表征的光源、由离散余弦变换基表征的掩模及由泽尼克系数表征的投影物镜编码为粒子,以图形误差作为评价函数,通过不断迭代更新粒子,实现光源掩模投影物镜联合优化。在标称条件和工艺条件下,采用含有交叉门的复杂掩模图形对所提方法的仿真验证表明,图形误差分别降低了94.2%和93.8%,有效提高了光刻成像质量。与基于遗传算法的SMPO方法相比,该方法具有更快的收敛速度。此外,该方法具有优化自由度高和优化后掩模可制造性强的优点。
Full chip multi-parameter joint optimization is an important development direction of lithography resolution enhancement technology. A combined projection lens optimization (SMPO) method based on particle swarm optimization (PSO) is proposed. The light source represented by the pixel, the mask characterized by the discrete cosine transform base and the projection objective represented by the Zernike coefficient are coded as particles, and the image error is taken as the evaluation function to optimize the joint of the light source mask projection objective by iteratively updating the particles. Simulation and verification of the proposed method using complex mask patterns with cross-gates under nominal conditions and process conditions show that the pattern errors are reduced by 94.2% and 93.8%, respectively, effectively improving lithographic imaging quality. Compared with the SMPO method based on genetic algorithm, this method has a faster convergence rate. In addition, this method has the advantages of high degree of freedom of optimization and high manufacturability of the optimized mask.