GOALI: Measurement & Modeling of Heterogeneous Nanoscale Deformation of Copper-Polyimide High Density Interconnect Structures Subjected to Thermal Cycling

基于Cu-聚酰亚胺系统的总高密度互连（HDI）结构相对于Al-SiO2系统在信号传播速率方面表现出改进。 然而，面外热膨胀系数的大失配在处理期间产生高界面应力，这可能不利地影响结构的可靠性。 随着布线层的数量增加，这个问题将变得更加严重。 最近的实验证据表明，聚酰亚胺的弹性性能和热膨胀是1 μ m宽度及以下的结构尺寸的函数。 额外的实验证据表明，在这些结构中的Cu的非弹性变形是本地化的晶界和在Cu和聚酰亚胺之间的界面。 此外，位错滑移的塑性变形已被证明是可以忽略不计的相关温度范围。 由于特征尺寸是在Cu晶粒尺寸的量级上，基于非弹性变形的连续模型的有限元模型将不能准确地描述这些HDI结构中的应力。 建议的分析和实验计划的目标是开发和评估策略，在这些结构中的扩散运输的基础上建模异质，非弹性变形。 具体而言，非弹性变形将被限制在晶界和双材料界面。 界面和晶界本构模型将基于Coble蠕变和晶界滑动机制。 晶粒内部和聚酰亚胺中的变形将被建模为线性弹性（具有温度依赖的弹性常数和热膨胀系数）。 将使用热台扫描探针显微镜测量界面和晶界滑动的量和速率作为温度的函数。 测量晶界和界面处的材料堆积，将用于量化Coble蠕变机制的材料传输速率。 第二个相关的目标是开发策略来测量由用于限定导体线图案的蚀刻造成的损伤引起的聚酰亚胺的性质的程度和变化。 减少HDI结构中特征尺寸的趋势需要开发策略来模拟热机械变形，该变形考虑到变形的异质性，并认识到通过扩散传输的变形将对纳米结构具有一阶重要性。 了解这些影响并记录其大小将对整个微电子行业具有广泛的意义。 ***

9631319 Gross High density interconnect (HDI) structures based on the Cu-polyimide system have exhibited improvements in signal propagation rate relative to the Al-SiO2 system. However, the large mismatch in out-of-plane thermal expansion coefficients generates high interfacial stresses during processing that may adversely affect the reliability of the structure. This problem will become more severe as the number of wiring levels increases. Recent experimental evidence has shown that the elastic properties and thermal expansion of the polyimides are a function of structure size for 1 (m widths and below. Additional experimental evidence suggests that the inelastic deformation of the Cu in these structures is localized on grain boundaries and at the interface between the Cu and the polyimide. Furthermore, plastic deformation by dislocation glide has been shown to be negligible for the relevant temperature range. Since the feature size is on the order of the Cu grain size, finite element models based on continuum models of inelastic deformation will not accurately describe the stresses in these HDI structures. The goal of the proposed analytical and experimental program is to develop and evaluate strategies to model heterogeneous, inelastic deformation based on diffusive transport in these structures. Specifically, inelastic deformation will be confined to the grain boundaries and bimaterial interfaces. The interface and grain boundary constitutive models will be based on the mechanisms of Coble creep and grain boundary sliding. Deformation in the grain interiors and in the polyimide will be modeled as linearly elastic (with temperature dependent elastic constants and coefficients of thermal expansion). A hot stage scanning probe microscope will be used to measure the amount and rate of interface and grain boundary sliding as a function of temperature. Measurement of the buildup of material at grain boundaries and interfaces will be used to quantify rates of material transport for the Coble creep mechanism. A second, related goal is to develop strategies to measure the extent and change of properties of the polyimides caused by damage from the etch used to define the pattern for the conductor lines. The trend toward reducing the size of features in HDI structures requires the development of strategies to model thermomechanical deformation that take into account the heterogeneous nature of deformation and recognize that deformation by diffusive transport will be of first order significance for nanostructures. Understanding these effects and documenting their magnitude will have broad significance to the entire microelectronics industry. ***