化学机械抛光过程跨尺度多场耦合问题的建模与协同求解方法研究

The requirement of chemical mechanical polishing (CMP) machining is gradually diversified in manufacturing of integrated circuit and micro/nano device. It is high cost and low efficiency to optimize CMP processing by traditional method relying on experimental data. Meanwhile, it is still difficult to obtain reliable information from the numerical simulation results of CMP processing to guide the design of CMP machine and the optimization of processing parameters, because quantitative simulation of CMP process is still a great challenge for the present CMP model and solution method. Accurate modeling and high-efficiency solution of trans-scale multi-physics coupling problem are the keys to realize quantitative analysis of whole-system model of CMP process. Aiming at solving this problem, accurate semi-empirical model of force, thermal, chemical concentration and material removal rate in micro-scale will be developed based on experimental observations and theoretical analysis of material removal process. Based on this model, micro/macro trans-scale multi-physics coupling model of CMP process will be built. Through the analysis of mapping relations of variables between different scale model and coupling relations in multi-factor problem, general solving strategy of trans-scale multi-factor problem will be developed, and then collaborative solving approach will be proposed. The research of this proposal will contribute to a new simulation platform for analysis of CMP process and system. Using this simulation platform, the action mechanism of various factors in CMP process will be revealed according to the research requirement under the premise of quantitative analysis of CMP process, and it is important to meet the diversification requirements of mechanism research and processing optimization. The research results of this proposal will offer theoretical basis for alternating the research approach of CMP process from experiment-based method to quantitative numerical simulation-based method.

集成电路和微纳器件制造领域对CMP工艺的需求逐渐多样化，而传统依靠试验数据进行工艺优化成本高、效率低，现有CMP数字化模型仍难以实现CMP过程全系统的定量模拟，无法有效指导CMP装备设计和工艺优化。跨尺度多场耦合问题的准确建模和高效求解是实现CMP全系统数字化模型定量分析的关键，针对这一难题，通过微观尺度材料去除试验和理论分析，建立微观尺度下的力、热、化学和材料去除的准确半经验数学模型，以该模型为基础，建立宏-微跨尺度多场耦合模型，通过跨尺度模型间变量映射关系和多场耦合问题的耦合关系分析，提出CMP过程跨尺度多场偶合问题的通用求解策略，并进行协同求解方法研究，建立CMP过程与系统的模拟平台，实现在CMP过程定量模拟的前提下，按需揭示抛光过程中各种因素的作用机制，满足CMP机理研究和工艺优化的多样化需求，为CMP工艺研究由依靠经验数据为主转变为以数字化模型定量分析为主提供理论依据。

随着微纳科学和技术的快速发展，对各种超平滑表面的加工需求不断增加。化学机械抛光（CMP） 技术作为超平滑表面加工的主流技术，在过去三十年中，伴随着集成电路的发展，工艺水平取得了很大的进步。然而，由于CMP加工过程的复杂性，技术的进步很大程度上取决于经验数据库的积累，面向未来在更广泛领域内的加工对象和需求，传统依赖于经验的工艺优化方法面临成本高和效率低的问题。为了突破CMP工程在建模和量化预测上的困难，本项目旨在提出一套可以实现不同类型材料CMP去除过程量化分析以及关键参数确定的系统方法，以实现CMP过程的量化预测，有效指导CMP工艺优化。.通过项目研究，提出了以微观尺度单粗糙峰CMP试验为基础，结合抛光垫-工件微观接触行为分析，最终实现宏观材料去除量化预测的思路和方法，取得初步成果。具体研究成果包括：1）提出了以单粗糙峰去除模型为基础建立宏观材料去除模型的方法，既可以描述化学机械协同作用下的原子尺度去除，同时又可以关联宏观尺度的材料去除过程。2）揭示了单晶硅和各类SiO2材料在CMP条件下的材料去除规律，提出了材料去除模型，准确描述了机械化学协同作用关系；3）揭示了抛光垫和工件间的微观作用关系，建立了多峰粘弹性接触模型，实现了微观接触面积的理论预测；4）建立了多粗糙峰作用条件下材料化学机械协同作用下的材料去除率预测模型，实现了高精度预测；5）搭建了CMP跨尺度多场耦合行为模拟软件平台架构，初步实现了材料去除率预测。.通过本项目研究，提出了CMP过程分析和仿真的新策略，开发了基础试验平台，提出了基本分析方法。研究成果对满足多样化的CMP机理研究需求和实际工艺优化具有指导价值，对实现CMP工艺的研究和工艺开发由经验为主转向量化仿真为主具有重要意义。