Pattern Defect Detection Research Articles

Repeat pattern segmentation of print fabric based on adaptive template matching

Patterned fabrics are generally constructed from the periodic repetition of a primitive pattern unit. Repeat pattern segmentation of printed fabrics has a very significant impact on the pattern retrieval and pattern defect detection. In this paper, we propose a new approach for repeat pattern segmentation by employing the adaptive template matching method. In contrast to the traditional method for template matching, the proposed algorithm first selects an adaptive size template image in the repeat pattern image based on the size of the original image and its local maximum edge density. Then it uses the sum of absolute differences as the matching features to identify the matched regions in the original image, and the minimum envelope border of the primitive pattern, typically as a parallelogram, can be determined from the results of the four adjacent matched templates. Finally, image traversal base on the obtained parallelogram is implemented over the original image using minimum information loss theory to produce a well-segmented primitive pattern with a complete edge structure. The results from the experiments conducted using an extensive database of real fabric images show that the proposed algorithm has the advantage of rotation invariance and scaling invariance and will not be affected when the background or foreground color is changed.

Data-driven approaches to optical patterned defect detection.

Computer vision and classification methods have become increasingly wide-spread in recent years due to ever-increasing access to computation power. Advances in semiconductor devices are the basis for this growth, but few publications have probed the benefits of data-driven methods for improving a critical component of semiconductor manufacturing, the detection and inspection of defects for such devices. As defects become smaller, intensity threshold-based approaches eventually fail to adequately discern differences between faulty and non-faulty structures. To overcome these challenges we present machine learning methods including convolutional neural networks (CNN) applied to image-based defect detection. These images are formed from the simulated scattering of realistic geometries with and without key defects while also taking into account line edge roughness (LER). LER is a known and challenging problem in fabrication as it yields additional scattering that further complicates defect inspection. Simulating images of an intentional defect array, a CNN approach is applied to extend detectability and enhance classification to these defects, even those that are more than 20 times smaller than the inspection wavelength.

Extreme ultraviolet lithography patterned mask defect detection performance evaluation toward 16- to 11-nm half-pitch generation

High-sensitivity and low-noise extreme ultraviolet (EUV) mask pattern defect detection is one of the major issues remaining to be addressed in device fabrication using extreme ultraviolet lithography (EUVL). We have designed a projection electron microscopy (PEM) system, which has proven to be quite promising for half-pitch (hp) 16-nm node to hp 11-nm node mask inspection. The PEM system was integrated into a pattern inspection system for defect detection sensitivity evaluation. To improve the performance of hp 16-nm patterned mask defect detection toward hp 11-nm EUVL patterned mask, defect detection signal characteristics, which depend on hp 64-nm pattern image intensity deviation on EUVL mask, were studied. Image adjustment effect of the captured images for die-to-die defect detection was evaluated before the start of the defect detection image-processing sequence. Image correction of intrafield intensity unevenness and L/S pattern image contrast deviation suppresses the generation of false defects. Captured images of extrusion and intrusion defects in hp 64-nm L/S patterns were used for detection. Applying the image correction for defect detection, 12-nm sized intrusion defect, which was smaller than our target size for hp 16-nm defect detection requirements, was identified without false defects.

Patterned mask inspection technology with projection electron microscope technique on extreme ultraviolet masks

High-sensitivity extreme ultraviolet (EUV) mask pattern defect detection is one of the major issues remaining to be addressed in device fabrication using extreme ultraviolet lithography. In order to achieve inspection sensitivity and suitability for the 1× nm node, a projection electron microscope (PEM) system is employed that enables high-speed/high-resolution inspection, which is not possible using conventional deep ultraviolet or electron beam inspection systems. By employing higher electron energy in the electron optics (EO) exposure system and by improving the PEM design, we have minimized the aberration that occurs when working with EO systems and we have improved the transmittance of the system. Experimental results showing the improved transmittance were obtained by making electron throughput measurements. To guarantee the tool’s aptness for 16-nm node EUV mask inspection, corresponding sized programmed defects on masks were designed, and the defect detection sensitivity of the EO system was evaluated. Improvements in image resolution and electron throughput have enabled us to detect 16-nm sized defects. The PEM system was integrated into a pattern inspection system for defect detection sensitivity evaluation.

Development of extreme ultraviolet mask pattern inspection technology using projection electron beam optics

Extreme ultraviolet (EUV) lithography with a 13.5 nm exposure wavelength is a leading candidate for the next-generation lithography because of its excellent resolution for 16 nm half pitch (hp) node devices and beyond. High sensitivity EUV mask pattern defect detection is one of the major issues to realize device fabrication with EUV lithography. First, to estimate targeted pattern defect detection size, a simulation for defect printability was carried out. In order to achieve the required inspection sensitivity applicable for 1X nm node, a projection electron microscopy (PEM) system was employed, which enabled us to do inspection in higher resolution and with higher speed in comparison to those of the conventional deep ultraviolet and electron beam inspection systems. By incorporating high electron energy and low optical aberration into the PEM, we designed a system for 16 nm hp node defect inspection. To guarantee the quality of the 16 nm node EUV mask, corresponding sized programmed defect masks were designed, and a PEM system defect detection was evaluated by using the current system for 2X nm generation.

EUV pattern defect detection sensitivity based on aerial image linewidth measurements

As the quality of EUV-wavelength mask inspection microscopes improves over time, the image properties and intensity profiles of reflected light can be evaluated in ever-greater detail. The SEMATECH Berkeley Actinic Inspection Tool (AIT) is one such microscope, featuring mask resolution values that match or exceed those available through lithographic printing in current photoresists. In order to evaluate the defect detection sensitivity of the AIT for dense line patterns on typical masks, the authors study the linewidth roughness (LWR) on two masks, as measured in the EUV images. They report the through-focus and pitch dependence of contrast, image log slope, linewidth, and LWR. The AIT currently reaches LWR 3σ values close to 9nm for 175nm half-pitch lines. This value is below 10% linewidth for nearly all lines routinely measured in the AIT. Evidence suggests that this lower level may arise from the mask’s inherent pattern roughness. While the sensitivity limit of the AIT has not yet been established, it is clear that the AIT has the required sensitivity to detect defects that cause 10% linewidth changes in line sizes of 125nm and larger.