Advanced optical quality assurance of silicon micro-strip sensors for the CBM Silicon Tracking System

Abstract

The Compressed Baryonic Matter (CBM) experiment at the Facility for Anti-proton and Ion Research (FAIR) in Darmstadt, Germany, aims to study strongly interacting matter under extreme conditions. The Silicon Tracking System (STS) is the key detector to reconstruct the charged particle tracks produced in heavy-ion collisions. This paper describes a setup for optical quality assurance of silicon micro-strip sensors used in the STS. Machine Vision Algorithms (MVA) were used to analyze microscopic scans of the sensors for the presence of defects with a resolution of about 1 μ m and to perform metrology tasks. The software developed has a recognition and classification rate of 87% for defects like scratches, shorts, broken metal lines etc. The use of advanced image processing employing neural networks allows to further improve the identification rate to 96%. • Custom-built setup for sensor optical inspection, motorized XYZ, zoom, focus stages • Automated calibration, scanning, defect finding and analysis • Machine vision algorithms provide defect detection rate of 87%. • Advanced algorithms (deep neural networks) improve the detection rate to 96%. • μ m -precision 3D-metrology algorithms (thickness, parallelism, warp).