电子束曝光邻近效应校正 PECEBL proximity effect correction

从版图和剂量测试，到可核验的工艺交接。From layout and dose tests to a verifiable process handoff.

邻近效应校正解决什么问题What PEC addresses

电子在胶与基底中的散射会影响附近区域的曝光。Georgia Tech 的说明指出，点扩散函数（PSF）描述散射后的能量分布，PEC 可结合版图估算剂量分布。因此，同一工艺中的孤立线、密集线和阵列边缘值得分别检查。[1]Electron scattering in resist and substrate affects nearby regions. Georgia Tech describes PSFs as energy-distribution models used with layout data for PEC. Isolated lines, dense lines and array edges therefore merit separate checks. [1]

适合什么需求Suitable projects

适合超表面阵列、纳米光栅、精细电极等项目中，已观察到或需要验证图形密度相关尺寸偏差的需求。EPFL 将 PEC 列为 BEAMER 数据准备功能之一；是否启用、如何配置，应由实际加工平台结合样品确认。[2]Useful for metasurface arrays, nanogratings and fine electrodes when density-dependent dimensional variation is observed or needs evaluation. EPFL lists PEC among BEAMER data-preparation functions; the receiving facility must decide whether and how to apply it. [2]

询价与交接：先备齐六项输入Six inputs for a project review

以下为微纳Hub 的交接建议：①原始 GDS/OASIS、top cell、单位、图层与版本；②目标线宽、间距、阵列周期及关键区域；③基底、膜层顺序与厚度；④胶型、胶厚及已知显影条件；⑤拟用设备、加速电压与既有剂量测试记录，未知项标待确认；⑥显影后或图形转移后的验收要求。原始设计与校正后的写入数据应分别命名，保留处理记录。MN Fab Hub suggests providing: original layout, top cell, units, layers and revision; target widths, gaps, pitch and critical regions; substrate and film stack; resist and development information; intended tool, voltage and existing dose tests; and acceptance after development or pattern transfer. Mark unknowns for review. Keep original and corrected data separately named with processing records.

剂量测试怎样做得可比较Make dose tests comparable

Georgia Tech 提供用于特定胶/基底流程的线宽和间距剂量表征图形。[1] 微纳Hub 建议把器件中的孤立、密集、边缘结构放入测试片，记录位置和测试编号；先固定胶厚、显影与测量方法，再比较剂量或校正方案。剂量范围和步进由平台确定，不预填通用数值。Georgia Tech provides test patterns for dose characterization across widths and spacings in a given resist/substrate process. [1] We suggest representative isolated, dense and edge structures with location IDs. Hold resist, development and measurement conditions fixed while comparing dose or correction settings. Let the facility define the dose range and steps.

常见风险：参数移植与重复校正Risks in parameter reuse

PSF 与电子能量、材料堆栈有关，不应只因胶名相同就沿用另一套校正参数。[1] 交付时说明数据是否已做 PEC、偏置或剂量分配，避免双方重复处理。EPFL 建议先确认基础曝光显影的结果，再逐项增加校正并评估效果。[3]PSFs depend on beam energy and the material stack; matching a resist name alone does not establish a transferable model. [1] State any prior PEC, bias or dose assignment to avoid duplicate processing. EPFL recommends evaluating the baseline exposure/development before adding corrections one at a time. [3]

验收与下一步Acceptance and next step

建议约定关键尺寸测量位置、孤立/密集结构的采样数、允许偏差和报告方式；明确验收发生在显影后、刻蚀后还是 Lift-off 后。校正文件生成成功不能代替实物验收。把版图、材料膜层、已有 SEM 和目标尺寸发给微纳Hub，先梳理资料缺口与小样验证路线，再由实际加工资源确认可行性、周期及报价。Agree measurement locations, sampling of isolated/dense features, tolerances and reporting, with a clear stage: after development, etching or lift-off. Successful data generation is not sample acceptance. Share the layout, stack, existing SEM images and targets with MN Fab Hub to prepare a pilot evaluation; feasibility, schedule and pricing require review by the actual facility.

官方资料Official references

核查日期：2026-09-07。下列资料解释工艺与平台用法，不构成微纳Hub设备性能承诺。Checked September 7, 2026. References describe process concepts and facility practices, not MN Fab Hub performance guarantees.