Inspection of complex objects utilizing energy-sensitive X-ray imaging

AbstraktAbstract

X-ray computed tomography (CT) is extensively used for non-destructive industrial analysis to study the internal structure of materials and complex objects. In the microelectronics industry, X-ray computed tomography has proven useful for quality analysis and defect detection. However, inspecting electronic devices with standard CT is challenging due to the presence of many metals neighboring light materials causing various types of artifacts that affect the quality of computed tomography. This thesis presents new techniques and data processing approaches in the field of energy-sensitive computed tomography, which are used to characterize materials and to study defects in electronic devices. Energy-sensitive computed tomography utilizes energy information from the projection data obtained by energy-sensitive methods. Several standard CT techniques were also tested and compared with the newly presented techniques. The results showed that the standard techniques are inefficient when the scanned samples, electronic components, have high metal content. The work also addresses the spectroscopic performance of photon counting detectors in computed tomography after correcting the threshold mismatch between pixels. This new correction methods have helped to increase the energy resolution and improve the ability of the detector to discriminate materials with similar attenuation properties. Such correction is useful for applications that require high resolution at high energy spectra where the mismatch between pixels increases due to the invalidity of standard calibration methods. The thesis mainly focuses on the inspection of epoxy underfill process for electronic devices by implementing data fusion of multiple computed tomography scans as an alternative method to reduce artifacts. To this end, the work presents a methodology of performing successful computed tomography that enable the inspection of light materials (epoxy) for samples with high metal that cannot be achieved using standard CT techniques. In addition, the work presents the first results of the epoxy underfill inspection for real samples from the industrial field, placing the X-ray inspection as an option for defect detection in the microelectronics industry.

X-ray computed tomography (CT) is extensively used for non-destructive industrial analysis to study the internal structure of materials and complex objects. In the microelectronics industry, X-ray computed tomography has proven useful for quality analysis and defect detection. However, inspecting electronic devices with standard CT is challenging due to the presence of many metals neighboring light materials causing various types of artifacts that affect the quality of computed tomography. This thesis presents new techniques and data processing approaches in the field of energy-sensitive computed tomography, which are used to characterize materials and to study defects in electronic devices. Energy-sensitive computed tomography utilizes energy information from the projection data obtained by energy-sensitive methods. Several standard CT techniques were also tested and compared with the newly presented techniques. The results showed that the standard techniques are inefficient when the scanned samples, electronic components, have high metal content. The work also addresses the spectroscopic performance of photon counting detectors in computed tomography after correcting the threshold mismatch between pixels. This new correction methods have helped to increase the energy resolution and improve the ability of the detector to discriminate materials with similar attenuation properties. Such correction is useful for applications that require high resolution at high energy spectra where the mismatch between pixels increases due to the invalidity of standard calibration methods. The thesis mainly focuses on the inspection of epoxy underfill process for electronic devices by implementing data fusion of multiple computed tomography scans as an alternative method to reduce artifacts. To this end, the work presents a methodology of performing successful computed tomography that enable the inspection of light materials (epoxy) for samples with high metal that cannot be achieved using standard CT techniques. In addition, the work presents the first results of the epoxy underfill inspection for real samples from the industrial field, placing the X-ray inspection as an option for defect detection in the microelectronics industry.