Prospects for sub-pixel resolution

In addition to revealing the dimensions of sub-pixel structures, the mesh experiments also hint at the possibility of improving the spatial resolution of the CCD beyond that of a 24 $\mu m$ pixel device. Before continuing, a short digression on event grades is necessary. The CCD registers an event when the charge cloud created by a photoelectric absorption is pulled into the potential of one of the gates. Some fraction of the time, the charge is divided into multiple pixels. Theoretically, this fraction, the branching ratio, depends on what depth in the silicon the interaction took place and where it occurs with respect to the pixel boundaries. According to this scheme, single pixel events (ASCA grade 0) come from photons that land in the center of the pixel, vertically split events (ASCA grade 2) come from photons that land near the vertical boundary of the two pixels, horizontally split events (ASCA grades 3 and 4) come from photons that land near the horizontal boundary of two pixels, and three and four pixel events (ASCA grade 6) come from photons that land in the corner of a pixel. The mesh experiments, for the first time, have conclusively proven that this simple picture is in fact correct. Figure 4.43 shows the 3x3 RP arrays from the O $K_{\alpha}$ data for grade 0, grade 2, grade 3 and 4, and grade 6 events. It is obvious that the horizontal and vertical split events come from the boundary regions and that the multiple pixel events come from the corner regions. At the same time, the single pixel RP events occur only in the center of the pixel.

The confinement of certain event grades to a specific area of the CCD is effectively like having smaller pixels inside a 24 $\mu m$ pixel and is the key to obtaining sub-pixel resolution. The branching ratios are a strong function of energy and penetration of the photon into the device. As the percentage of multiple pixel events increase, these mini-pixels will increase in size. Figure 4.44 shows two 3x3 pixel grids. Both grids show a geometric area (computed from the branching ratios) for the different event grades discussed above, one for Si $K_{\alpha}$ photons (1.740 keV) and one for Cu $K_{\alpha}$ photons (8.040 keV). Superimposed on each of these grids are the 33 % and 66 % enclosed energy curves for the HRMA. A full, mathematical investigation has not been performed, but the hope is that by comparing the branching ratios from an astronomical observation with ground calibration data, the source location can be determined to better than one pixel.

  

Figure 4.43: O $K_{\alpha}$ Representative Pixels for grade 0, grade 2, grades 3 and 4, and grade 6events.

  

Figure 4.44: HRMA encircled energy surfaces projected onto a schematic of sub-pixel locations