刻蚀演化仿真及关键参数优化

发布时间：2018-04-21 05:31

本文选题：刻蚀仿真 + 线算法　；参考：《清华大学》2015年硕士论文

【摘要】：集成电路技术的快速发展对生产工艺提出了越来越高的要求。而刻蚀作为集成电路生产过程中的一个重要环节,其工艺参数对刻蚀结果有着很大的影响。通过刻蚀试验获得工艺参数的方法,效率低,成本高。而通过仿真研究刻蚀演化过程,分析不同工艺参数对刻蚀结果的影响,可以节省大量的人力、物力和财力。针对等离子刻蚀过程,本文研究了几种不同的算法,分别设计实现了二维和三维刻蚀演化仿真工具,并通过与实验结果比较验证了方法的有效性。此外,由于仿真过程中的相关参数求取困难,通过引入实验数据,对比实验与仿真之间的差异,利用优化方法对这些参数进行优化,使得仿真结果更加准确。本文主要完成了如下的研究工作:1)提出一种新的基于线算法和分子动力学相结合的二维刻蚀演化方法,实现了跨尺度的刻蚀演化仿真。应用分子动力学方法计算到达模型表面的粒子的实际刻蚀效果,并将其折算成局部刻蚀率返回给线算法,从而更新下一步的剖面形貌。该方法简化了相关参数,扩大了仿真方法的适用范围,实现了对刻蚀过程的二维表面演化仿真。2)改进了一种基于元胞自动机的三维刻蚀演化仿真方法。通过对模型压缩表示优化模型,使得演化过程中空间需求降低;同时利用刻蚀产额等效粒子对基底材料表面的刻蚀效果,降低了仿真实施的难度。通过研究不同实验条件对刻蚀结果的影响,验证了算法在进行三维刻蚀演化中的有效性。3)提出了依据实验数据来优化刻蚀产额参数的方法。通过实际的刻蚀剖面和模拟剖面之间的对比,设定适应度函数将刻蚀产额参数求解问题转化为最优化问题。应用序优化方法和禁忌搜索算法相结合来解决这个优化问题,通过优化参数获得更加准确的仿真结果。
[Abstract]:The rapid development of integrated circuit technology puts forward higher and higher requirements for production process. As an important part of IC production, etching parameters have great influence on etching results. The method of obtaining process parameters by etching test has the advantages of low efficiency and high cost. By studying the process of etching evolution and analyzing the influence of different process parameters on etching results, a lot of manpower, material and financial resources can be saved. Aiming at the plasma etching process, this paper studies several different algorithms, designs and implements two-dimensional and three-dimensional etching evolution simulation tools, and verifies the effectiveness of the method by comparing with the experimental results. In addition, because of the difficulty of obtaining the relevant parameters in the simulation process, by introducing the experimental data and comparing the differences between the experiment and the simulation, the optimization method is used to optimize these parameters, which makes the simulation results more accurate. In this paper, the following research work is completed: 1) A new two-dimensional etching evolution method based on the combination of linear algorithm and molecular dynamics is proposed, and the cross-scale etching evolution simulation is realized. The molecular dynamics method is used to calculate the actual etching effect of particles reaching the surface of the model, and the local etching rate is converted into a feed line algorithm to update the profile of the next step. This method simplifies the relevant parameters and expands the scope of application of the simulation method. The two-dimensional surface evolution simulation of etching process is realized. 2) an improved three-dimensional etching evolution simulation method based on cellular automata is proposed. Through the compression of the model and the optimization of the model, the space demand in the evolution process is reduced, and the etching effect of the equivalent particle of etching yield on the substrate surface is reduced, which reduces the difficulty of simulation. By studying the effect of different experimental conditions on the etching results, the validity of the algorithm in the evolution of three-dimensional etching is verified. The method of optimizing the etching yield parameters based on the experimental data is proposed. Through the comparison between the actual etching section and the simulated section, the fitness function is set up to transform the solution of etching yield parameters into an optimization problem. Order optimization method and Tabu search algorithm are used to solve the optimization problem, and more accurate simulation results are obtained by optimizing the parameters.
【学位授予单位】：清华大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN405

【相似文献】