Characterizing Charge Diffusion in CCDs

Charge-coupled devices, or CCDs, are the light detectors used in almost all modern cameras. As it turns out, they are also the detector of choice for many telescopes. One important measurement for any CCD that will be used in a telescope is the amount of “charge diffusion” (informally, the amount of light from one pixel which bleeds through to an adjacent pixel due to electrical effects within the CCD). An empirical measurement of charge diffusion is essential before any new CCD can be qualified for scientific use.

The schematic representation below shows a photon entering a CCD. At some point within the CCD, the photon is absorbed, resulting in the generation of a cloud of energetic electrons. The electrons spread out as they are pulled to the bottom of the CCD by an applied electric field. At the bottom of the CCD, the electrons are captured into pixels and ultimately conveyed to an amplifier for measurement. This schematic was stolen from the excellent paper by Rodney and Tonry, which inspired my subsequent work on the measurement of charge diffusion.


I developed a relatively simple numerical technique for estimating the amount of charge diffusion in a given reference image. My technique does not depend on a specific model for electron mobility within the CCD, resulting in reduced complexity and improved robustness as compared to previous methods. For a detailed description of the algorithm, consult:

David Lawrence, Paul O’Connor, et al., “Model-independent Characterization of Charge Diffusion in Thick Fully Depleted CCDs”, Publications of the Astronomical Society of the Pacific (August, 2011).

This research was carried out at the Instrumentation Division of Brookhaven National Laboratory between 2008 and 2010, in order to qualify new CCDs under development for the Large Synoptic Survey Telescope. I would like to thank Peter Takacs and Paul O’Connor for their exceptional mentorship while I was working on this project. Source code is available on GitHub.