Predictive Semiconductor Lithography based on Theoretically-Informed Reaction-Diffusion Models

基于理论反应扩散模型的预测半导体光刻

• 批准号: 1731185
• 负责人: Gila Stein
• 金额: $ 18.7万
• 依托单位: University of Tennessee Knoxville
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2016
• 资助国家: 美国
• 起止时间: 2016-09-01 至 2019-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1731185&HistoricalAwards=false
• 关键词: Predictive; Semiconductor; Lithography; based; Theoretically

1437878 (Stein)The most sophisticated integrated circuits, such as microprocessors and memory chips, are patterned by a process called projection lithography. In a typical lithographic process, a silicon wafer is coated with a radiation-sensitive polymer film ("resist") and exposed to a pattern of light. Radiation triggers a reaction that generates the latent chemical image, and patterns are developed by selectively washing away the exposed (or un-exposed) material. Industrial applications require high-throughput processes, so the resists must be highly sensitive to radiation. This is achieved with a process termed "chemical amplification". Chemically-amplified (CA) resists have two principal components: (i) a lipophilic polymer with acid-labile protecting groups; and (ii) a low concentration of photoacid generator (PAG). Exposing the resist to radiation generates a strong acid-counterion catalyst, and heating at moderate temperature promotes the acid-catalyzed decomposition of protecting groups along the polymer backbone. This deprotection reaction changes the polymer polarity for development in an aqueous base. CA systems are highly efficient because each photon absorbed by the resist generates approximately 0.3-2 acids (depending on the photon energy), and each catalyst cleaves hundreds of bonds so a low radiation dose is "amplified" through chemistry. However, the excellent sensitivity of CA resists comes at a price, because acid diffusion during reaction will limit the pattern resolution of CA resists. While CA resists have been studied for more than 40 years, there are no quantitative models that predict the spatial extent-of-reaction with nanoscale resolution. This poses an increasingly important roadblock for the semiconductor industry, as current research and development efforts are targeting sub-10 nm feature sizes.Intellectual Merit:The aim of this research program is to identify the fundamental physics and chemistry that produce anomalous catalyst transport and catalyst deactivation. The PI hypothesizes that (i) anomalous catalyst transport is induced by dynamic heterogeneities in the glassy polymer films; and (ii) catalyst loss is a signature of acid-catalyzed side reactions. A series of experiments are planned that will reveal these effects. The knowledge acquired through this research will be developed into a predictive lithography tool that can accelerate materials development and process optimization. The spatial extent-of-reaction in CA resists cannot be directly measured with experiments, so experimental investigation needs to be coupled to models that predict composition at nanometer length scales. The intellectual merit of this research is the development of an efficient, quantitative, and spatially-resolved reaction-diffusion model for CA resists that is applicable in a wide range of processing conditions with parameters determined by experimental investigation. To achieve this goal, the study will employ a minimal set of variables to be determined through optimization while incorporating the underlying physics and chemistry of the glassy polymer matrix and acid catalyst. The framework developed through this research will offer a means to investigate processing conditions and formulations that have yet to be realized in laboratory experimentation.Broader Impacts :The outcomes of this research program may immediately impact materials design and process development for next-generation lithography, as predictive models reduce or eliminate the need for iterative experimentation. After establishing the applicability of the anomalous transport model over a wide range of conditions, detailed descriptions of the method will be available to the general public via publications and presentations, furthering the development of lithography simulation that remains a bottleneck in industrial applications. The students engaged by this program will receive training in polymer physics and chemistry, computational methods, and designing experimental procedures. The PI has a record of engaging high school students and undergraduate students in laboratory research, and also participates in community outreach through NSF-funded K-12 camps at the University of Houston. Outcomes of this program will be incorporated into the PI's graduate course "Chemical Processing for Microelectronics."

1437878（Stein）最复杂的集成电路，如微处理器和存储器芯片，通过称为投影光刻的工艺形成图案。在典型的光刻工艺中，硅晶片涂覆有辐射敏感聚合物膜（“抗蚀剂”）并暴露于光的图案。辐射引发产生潜在化学图像的反应，并且通过选择性地洗掉暴露（或未暴露）的材料来显影图案。工业应用需要高通量工艺，因此抗蚀剂必须对辐射高度敏感。这是通过一种称为“化学放大”的过程来实现的。化学放大（CA）抗蚀剂具有两种主要组分：（i）具有酸不稳定保护基团的亲脂性聚合物;和（ii）低浓度的光酸产生剂（PAG）。将抗蚀剂暴露于辐射产生强的酸抗衡催化剂，并且在中等温度下加热促进沿着聚合物主链的保护基团的酸催化分解。这种脱保护反应改变了聚合物的极性，以便在碱水溶液中显影。CA系统是高效的，因为被抗蚀剂吸收的每个光子产生大约0.3-2个酸（取决于光子能量），并且每个催化剂裂解数百个键，因此低辐射剂量通过化学被“放大”。然而，CA抗蚀剂的优异灵敏度是有代价的，因为反应期间的酸扩散将限制CA抗蚀剂的图案分辨率。虽然CA抗蚀剂已经研究了40多年，但没有定量模型可以预测具有纳米级分辨率的反应空间范围。这对半导体工业构成了越来越重要的障碍，因为当前的研究和开发工作的目标是亚10 nm的特征尺寸。智力优点：本研究计划的目的是确定产生异常催化剂运输和催化剂失活的基本物理和化学。 PI假设（i）玻璃状聚合物膜中的动态不均匀性引起异常催化剂传输;和（ii）催化剂损失是酸催化副反应的特征。计划进行一系列实验来揭示这些影响。通过这项研究获得的知识将被开发成一种预测性的光刻工具，可以加速材料开发和工艺优化。CA抗蚀剂中反应的空间范围不能用实验直接测量，因此实验研究需要与预测纳米长度尺度组合物的模型相结合。这项研究的智力价值是开发一个有效的，定量的，和空间分辨的反应扩散模型CA抗蚀剂，是适用于在广泛的工艺条件与实验研究确定的参数。为了实现这一目标，该研究将采用一组最小的变量，通过优化来确定，同时结合玻璃状聚合物基质和酸催化剂的基本物理和化学。通过这项研究开发的框架将提供一种手段，以调查尚未实现在实验室experiments.Broader影响的加工条件和配方：这项研究计划的结果可能会立即影响下一代光刻的材料设计和工艺开发，预测模型减少或消除迭代实验的需要。在建立了异常传输模型在各种条件下的适用性之后，该方法的详细描述将通过出版物和演示文稿向公众提供，进一步发展光刻模拟，这仍然是工业应用中的瓶颈。参与该计划的学生将接受高分子物理和化学，计算方法和设计实验程序的培训。PI有让高中生和本科生参与实验室研究的记录，并通过NSF资助的休斯顿大学K-12营地参与社区外展。该计划的成果将纳入PI的研究生课程“微电子化学处理”。"