Amplitude and spatial frequency characterization of line edge roughness using AFM

To obtain a metrology method for line edge roughness (LER) detection and characterization using atomic force microscope (AFM), a LER quantificational method using image processing theory was presented, which is used to analyze AFM images of silicon lines and extract LER characteristics. Two types of analysis methods of LER were established: amplitude measurement and spatial frequency analysis. The amplitude parameters can reflect the uniformity of line edge topography, while the spatial frequency characterization method combining wavelet-based multiresolution analysis with power spectral density (PSD) analysis can give the spatial analysis of sidewall profile effectively. The measurement results of samples show that the main energy of LER is at the low frequency region. The dominant spatial frequency of LER is ∼0.04 nm ^-1, and the features with a characteristic wavelength of 500 nm have the largest contribution to the LER.