基于SAFT的单晶硅内部缺陷时域检测成像方法

Imaging Method of Monocrystalline Silicon Internal Defects Based on SAFT Time-domain Detection

为了提高单晶硅内部缺陷的测量精确度和空间分辨率,针对单晶硅材料提出了合成孔径聚焦技术直接时域重建的方法,将多个A超扫描测量结果连续叠加,合并样本点的相位信息,根据TOF算法实现数据映射和同相求和。基于6dB-drop的概念,提出了可描述缺陷边界特征的测量方法,进而推测缺陷的范围、位置、形状及特征,为单晶硅切片工序的计算提供更多信息。流时域重建方法可用于处理实验数据,进而描绘缺陷的特征,简单有效地提高空间分辨率。实验结果表明,该方法可清楚辨别出两个间隔小且直径为0.8mm的孔洞,且天然缺陷的尺寸和方位的量化非常接近于实际测量出的切割试样。

A direct time-domain reconstruction and sizing method of SAFT was developed to ira-prove the spatial resolution and sizing accuracy of monocrystalline silicon internal defects. The basic idea of the reconstruction algorithm was to coherently superimpose multiple A-scan measurements, in- corporating the phase information of the sampling points. The algorithm involved data mapping and in -phase summation according to time-of-flight （TOF）. A sizing method was proposed to characterize the flaw boundary based on the concept of 6 dB-drop. The extents, orientation and the shape of the flaw might then be inferred to provide more informations for monocrystalline silicon slicing process technology. The developed method was used to process the experimental data to characterize the flaws. Using the developed method, the improvement of spatial resolution was observed. Results indi- cate that two closely spaced 0.8 mm-diameter holes are clearly identified, and the quantification of size and orientation of the natural flaw are very close to the actual measurements after cutting the testing piece apart.