基于格点优化策略的化学机械研磨模型优化方法研究

Optimization of Chemical Mechanical Planarization Models Using Grid Window Refinement Strategies

  • 摘要: 在超大规模集成电路制造中,化学机械研磨工艺(chemical mechanical planarization, CMP)是实现多层金属互联结构全局平坦化的关键步骤,其效果直接影响后续光刻焦深与电路性能。本文聚焦于CMP模型的优化问题,特别是针对具有特殊几何结构的热点图形的预测能力。通过AFM实测数据与模拟结果对比发现,传统20 μm格点窗口无法准确解析由热点图形引起的Erosion缺陷。为此,提出采用3 μm小格点窗口配合1.5 μm偏移策略,可显著提升模型的预测精度。该优化策略不仅能有效识别版图中的潜在制造风险区域,还可追踪多层金属结构中下层Erosion对上层铜厚度的影响。研究表明,格点窗口尺寸与位置的精细调整对CMP模型的准确性具有决定性作用,为提高芯片良率与制造可靠性提供了重要支持。

     

    Abstract: In very large-scale integrated (VLSI) circuit manufacturing, CMP is essential for achieving global surface flatness across multilayer metal interconnects, significantly influencing lithographic precision and electrical performance. This study focuses on the optimization of CMP modeling, particularly in the accurate prediction of hotspot patterns with unique geometries. Experimental comparisons with atomic force microscopy (AFM) measurements indicate that conventional 20 μm grid windows are inadequate for capturing localized Erosion defects. A refined strategy employing 3 μm grid windows with a 1.5 μm offset is introduced, markedly enhancing prediction accuracy. This method not only identifies process-sensitive regions within the layout but also facilitates the tracking of Erosion-induced variations in metal thickness across higher metal layers. The findings underscore the critical role of grid resolution and positioning in CMP modeling and offer a reliable approach to improving IC yield and manufacturing robustness.

     

/

返回文章
返回