Photo diodes for machine vision: device characteristics and a-Si:H deposition and analysis
2003 (English)Doctoral thesis, comprehensive summary (Other academic)
During the last years the area of digital cameras based on CMOS technology has grown rapidly. In CMOS, signal processing circuitry and sensor elements can be combined on the same chip, which has resulted in advanced machine vision chip designs. However, the potentially powerful combination of diodes and in-pixel signal processing circuitry, suffers from problems of area utilization trade-offs. By using hydrogenated amorphous silicon (a-Si:H) diodes that are post-processed on the CMOS chip we can achieve a number of advantages. They are better with respect to area utilization and in some sense avoid problems like bulk wafer defects. In addition the evolution of CMOS processes is moving in a direction that degrades the properties of sensor elements.
This thesis explores photo diodes made of a-Si:H with a special focus on fabrication methods and device characteristics for machine vision applications. The work includes characterization of state-of-the-art diodes describing their lag current, photo current rise and fall time, and stability. To further explain themeasured characteristics a device model based on the physical semiconductor material properties has been developed.
This thesis also explores the use of de magnetron sputter deposition as a fabrication method for the a-Si:H p-i-n diodes. The impurity incorporation for the p- and n-doped layers was accomplished using doped targets. This technique is unusual and required a particular research effort. Dielectric functions for the intrinsic as well as n- and p-doped films were extracted and used to predict the optical properties of a p-i-n diode.
The conclusion of this research project is that a-Si:H diodes are suitable for machine vision chips.
Place, publisher, year, edition, pages
Linköping: Linköping studies in science and technology , 2003. , 40 p.
Linköping Studies in Science and Technology. Dissertations, ISSN 0345-7524 ; 799
Engineering and Technology
IdentifiersURN: urn:nbn:se:liu:diva-34896Local ID: 23851ISBN: 91-7373-602-3OAI: oai:DiVA.org:liu-34896DiVA: diva2:255744
2003-03-21, Sal Visionen, Linköpings Universitet, Linköping, 10:15 (Swedish)
Böhm, Marcus, Prof.
List of papers