MIT/Lincoln Labs CCID20 W66C2 CCD Test Results

The first thinned, backside MITLincoln Labs 2Kx4K CCD, W66C2 was characterized in the Lick detector lab. This is a high-resistivity CCD and is 40 microns thick. A report is available.

QE curve: The QE measurement was made at two device temperatures.
Surface plot: This measurement was made at room temperature with the device in its chip carrier.
Brick wall: This is the pattern which results from the incomplete backside laser anneal.

This plot shows two QE measurements made at device temperatures of -130°C and -70°C. A new, QE calibrated diode was used in the measurements. The lack of response at 3500Å and shorter seems to indicate that the Lincoln ion-implant plus laser-anneal doesn't pin the first few hundred Angstroms of silicon at the back of the CCD. This is where all of the UV photons are absorbed, and it is clear we are not collecting the electrons from this layer. The blue QE (at 4000Å and longer) can be improved greatly with a two-layer AR coating. The single-lager AR coating on this device was optimized for about 6500Å. This results in about 30% reflection at 4000Å. The 40-micron thick CCD has better red response than the traditional 20-micron thick CCD, although a better AR coating would help here too. The only significant difference between the two curves is found at 9000Å and beyond, where the QE goes up with increased temperature, as expected.

QE plot of W66C2 at -130°C.

Here is a surface plot for the first thinned CCID20 device tested. The device number is W66C2. This device was especially difficult to scan because in some regions of the CCD the reflected laser beam was diffracted into a pencil-shaped beam several centimeters long. This made the central bright spot difficult to track. But refocus of the laser and modifications to the tracking software to allow for varying spot brightness allowed us to scan most of the surface. The reason for the odd behavior of the reflected laser beam isn't yet known. The end-on view shows that the peak-to-peak variation over the area scanned is about 12 microns.

Surface plot of W66C2 at room temperature.

End-on view of surface plotof W66C2 at room temperature.

The following image is a flat field at 400nm showing QE variations that result from incomplete annealing of the backside after boron implantation. The amplitude of the variations is about 12%.