公开(公告)号：WO2015172027A1

公开(公告)日：2015-11-12

申请号：PCT/US2015/029896

申请日：2015-05-08

Applicant: KLA-TENCOR CORPORATION

Inventor： SHCHEGROV, Andrei V. ,  PANDEV, Stilian Ivanov ,  MADSEN, Jonathan M. ,  KUZNETSOV, Alexander ,  MIEHER, Walter Dean

IPC: H01L21/66

CPC classification number: G01N21/9501 ,  G01B11/27 ,  G06N99/005

Abstract: Methods and systems for creating a measurement model based only on measured training data are presented. The trained measurement model is then used to calculate overlay values directly from measured scatterometry data. The measurement models receive scatterometry signals directly as input and provide overlay values as output. In some embodiments, overlay error is determined from measurements of design rule structures. In some other embodiments, overlay error is determined from measurements of specialized target structures. In a further aspect, the measurement model is trained and employed to measure additional parameters of interest, in addition to overlay, based on the same or different metrology targets. In some embodiments, measurement data from multiple targets, measurement data collected by multiple metrologies, or both, is used for model building, training, and measurement. In some embodiments, an optimization algorithm automates the measurement model building and training process.

Abstract translation: 提出了仅基于测量训练数据创建测量模型的方法和系统。 然后使用经过训练的测量模型直接从测量的散射测量数据计算覆盖值。 测量模型直接接收散射信号作为输入，并提供重叠值作为输出。 在一些实施例中，根据设计规则结构的测量确定覆盖误差。 在一些其它实施例中，通过专门的目标结构的测量来确定覆盖误差。 在另一方面，测量模型被训练并用于基于相同或不同的度量目标来测量除叠加之外的附加参数。 在一些实施例中，来自多个目标的测量数据，由多个计量学收集的测量数据或两者都用于建模，训练和测量。 在一些实施例中，优化算法使测量模型构建和训练过程自动化。