Enhancing Measurement Accuracy of Nanolithography Alignment Using 2-D Hamming Self-Convolution Window

The Moiré fringe technique is an effective approach for achieving nanolithography alignment. However, the accuracy of fringe phase extraction is compromised by spectral leakage. Due to their limited capability to suppress spectral leakage, classical windows are inadequate for image processing in nanoscale alignment. To address this issue, we propose a novel 2-D Hamming self-convolution window (2D-HmSCW), developed through multiple time convolutions of the classical 2-D Hamming window. This technique improves the sidelobe performance of the window as the convolution order increases, thereby facilitating spectral leakage suppression. This approach addresses the inherent limitations of the 2-D Hamming window, particularly its edge oscillation. Simulation and experimental results are presented to verify the rationality and feasibility of this approach. Compared with classical windows, the 2D-HmSCW with a reasonable order can more effectively suppress spectral leakage. Consequently, the Moiré fringe image weighted 2D-HmSCW enables high-precision phase extraction using 2-D fast Fourier transform (2D-FFT), enhancing the alignment measurement accuracy to 1.93 nm. Additionally, we explore the feasibility of applying the 2D-HmSCW in other fringe-based high-precision measurement fields, such as Fourier transform (FT) profilometry and circular fringe projection profilometry, highlighting its potential for widespread applications.