Cooled-CCD and amorphous silicon-based neutron imaging systems for low-fluence neutron sources

摘要

Abstract: We have developed a neutron detection system for accelerator based neutron radiography and tomography based on a combination of scintillation screen and large aperture optics combined with a cooled CCD camera. The system is capable of detecting single neutron events and can therefore be considered as a possible detector for neutron scattering as well as conventional imaging. The system has a resolution of 0.1 mm or 1242 by 1152 pixels. The limit of image size is set by the light output of the scintillator, the light collection of the optical system, the size of the CCD and the desired signal to noise ratio. The lower limit on neutron flux is determined by the dark current of the chip. Equations for these limits have been derived and can be used to predict and optimize performance. The scintillation light output per incident neutron is large enough to permit the use of lens coupled systems with their increased flexibility and ease of implementation. The system can approach a quantum limited noise level, depending on the particular geometry used. For our current system, based on the use of NE 426 scintillator,$+3$/ a 1242 by 1152 pixel EEV CCD operating at $MIN@50C, and using a 100 mm focal length, F/0.9 lens, the maximum size for the imaging screen is 0.5 m, and the lower limit for flux is 1 n/pixel/s based on this size screen and a typical dark current of 10 e/pixel/s. We are now investigating a new type of imaging technology based on large amorphous silicon sensor arrays being developed by Xerox and others. A typical device is 200 by 250 mm with a pixel size of 127 $mu@m and the entire array with all electronics is in a 400 mm by 37 mm package. Major advantages of this device are the high light coupling between scintillator screen and the sensor as well as the more compact nature of such an array, since no lens systems is required and, potentially, a much lower cost. Currently, the noise performance is worse than that of CCDs, largely due to the current electronic readout but should be adequate to produce quantum limited images if electronics can be improved. !12