CCD Modelling and Verification for ESA's Euclid CCD Architecture

Clarke, Andrew Stephen (2015). CCD Modelling and Verification for ESA's Euclid CCD Architecture. PhD thesis The Open University.

DOI: https://doi.org/10.21954/ou.ro.0000efce

Abstract

This thesis discusses semiconductor modelling carried out to investigate the charge packets which are present in Charge Coupled Devices (CCDs). The details of charge distribution are of interest for scientific CCDs, in particular those which may become damaged due to the effects of harsh radiation environments such as those experienced in space. Therefore this thesis focuses on the CCDs used in the main focal plane instrument (VIS, a visible imager) on the European Space Agency's Euclid space telescope due for launch in 2018.

The thesis is spilt into two parts, with the early chapters concentrating on the development of device models and the interpretation of modelling results. The modelling work is verified in latter chapters to ensure that the model predictions concur with lab based measurements on test devices.

Model verification makes up a large part of this thesis, ensuring the accuracy of the models. In some cases initial modelling work was shown to predict parameters different to those measured in test devices, however, investigations are made which show manufacturing errors are partly to blame. During this work images were taken using a Focussed Ion Beam Scanning Electron Microscope (FIBSEM), which allows cross-sectional images of the device to be taken, these showed deviations from the designed ploy silicon electrode geometry, resulting in some undersized and some oversized gates, these reduced the accuracy of some models, which were later modified. Through this investigation a method for predicting gate geometry was developed based on the Silvaco device models.

Further testing was carried out to verify the charge packet relationship with CTE in test devices. CTE is measured using the Extended Pixel Edge Response (EPER) technique, which allows CTE to be measured over a range of signal sizes. The CTE data can be used to calculate the charge loss per pixel as an average across the device column and these values can be compared against the charge packet volume simulations.