Characterizing the influence of neutron fields in causing single-event effects using portable detectors

Abstract

The malfunction of semiconductor devices caused by cosmic rays is known as Single Event Effects(SEEs).
In the atmosphere, secondary neutrons are the dominant particles causing this effect. The neutron flux density in atmosphere is very low. For a good statistical certainty, millions of device hours are required to measure the event rate of a device in the natural environment. Event rates obtained in such testings are accurate.
To reduce the cost and time of getting the event rate, a device is normally taken to artificial accelerated neutron beams to measure its sensitivity to neutrons. Comparing the flux density of the beam and the flux density of a location in the atmosphere, the real time event rate can be predicted by the event rate obtained. This testing method was standardized as the neutron accelerated soft error rate (ASER) testing in JEDEC JESD89A standard.
However, several life testings indicated that the neutron flux density predictions given by the accelerated testings can have large errors. Up to a factor of 2 discrepancy was reported in the literature. One of the major error sources is the equivalence of the absolute neutron flux density in the atmosphere and in accelerated beam.

This thesis proposes an alternative accelerated method of predicting the real-time neutron error rate by using proxy devices. This method can avoid the error introduced by the uncertainty in the neutron flux density.
The Imaging Single Event Effect Monitor (ISEEM) is one of the proxy devices. It is the instrument originally developed by Z. Török and his co-workers in the University of Central Lancashire. A CCD was used as the sensitive element to detect neutrons. A large amount of data sets acquired by Török were used in this work. A re-engineered ISEEM has been developed in this work to improve ISEEM performance in life testings. Theoretical models have been developed to analyze the response of ISEEM in a wide range of neutron facilities and natural environment. The agreement of the measured and calculated cross-sections are within the error quoted by facilities. Because of the alpha contamination and primary proton direct ionization effects, performance of ISEEM in life testings appeared to be weak.