Detection and clustering of mixed-type defect patterns in Wafer Bin maps

Abstract

A wafer bin map (WBM) is a map that consists of assigned bin values for each die based on wafer test results (e.g., value 1 for good dies, and value 0 for defective dies) in semiconductor manufacturing. Many times, the bin values of adjacent dies are spatially correlated, which form some systematic defect patterns. These non-random defect patterns occur by some assignable causes

therefore, identifying these systematic defect patterns is important to know root causes of failure and to take actions for quality management and yield enhancement. In particular, as wafer fabrication processes have become more complicated, mixed-type defect patterns (two or more different types of defect patterns occur simultaneously in a single wafer) occur more frequently than the past. For more effective classfication of wafers by their defect patterns, mixed-type defect patterns need to be detected and separated into several clusters of different patterns

subsequently, each cluster of a single pattern can be matched a well-known defect type (e.g., scratch, ring) or it can suggest the emergence of a new defect pattern. There are several challenges in detecting and clustering mixed-type defect patterns: 1) separation of random defects from systematic defect patterns

2) determination of the number of clusters

and 3) clustering of complex shapes of defect patterns. To address these challenges, this thesis proposes a new framework for detection and clustering of mixed-type defect patterns. First, a new preprocessing method, called the connected-path filtering algorithm, is proposed to denoise WBMs. Subsequently, the infinite warped mixture model is adopted for clustering of mixed-type defect patterns

this model is flexible to deal with complex shapes of defect patterns, and furthermore, the number of clusters does not need to be specified in advance but is automatically determined simultaneously during the clustering procedure. The proposed method is validated with real data from a semiconductor industry. The experimental results demonstrate the effectiveness of the proposed method in estimating the number of underlying clusters as well as in clustering of mixed-type defect patterns.