Fitting Methods: FI Device Model Parameters.

The spatially-averaged relative quantum efficiency measurements described in section 4.7.2 were used to constrain slab-and-stop model gate structure parameters for the flight FI devices. For each device, the MIT CSR relative efficiency measurements at 525 eV, 677 eV, 1740 eV and 2015 eV were used to fit for three model parameters: the mean silicon gate thickness, the mean silicon dioxide thickness, and the silicon nitride thickness. In each case, the reference detector quantum efficiency was modelled using the slab and stop model parameters derived from the undispersed synchrotron measurements, as described in section 4.6.1.

Estimates of the errors in the relative quantum efficiency measurements were derived from the comparison of relative efficiency measurements made at MIT CSR to those made in Phase I (ACIS flat-field) of the XRCF calibration (see section 4.7.3.) Thus, at these energies, for the front-illuminated detectors, we take the standard deviation of the error in each spatially-averaged relative quantum efficiency measurement to be 0.6%, independent of energy. Fitting was performed by minimizing the $\chi ^{2}$ statistic. Best-fit gate-structure model parameters, along with other adopted model parameters, are listed in Table 4.63. In all cases the fits (which had but one degree of freedom each!) were statistically acceptable given the adopted measurement errors.