Quantum Efficiency of ACIS CCD w215c2r

Release notes for preliminary quantum detection efficiency curve for ACIS front-illuminated detectors.

M. Bautz
29 August 1997




ASCII table suitable for use with qdp. There are two columns:

  1. Energy (keV)
  2. Quantum Efficiency (for ASCA grades 0,2,3,4,6)


This file provides a preliminary estimate of the quantum efficiency of the ACIS I3 CCD detector (MIT Lincoln Laboratory model ccid17, serial number w215c2r.) The data are for ASCA grades 0,2,3,4,6, with a split-event threshold of 14 electrons (13 adu; adu= analog-to-digital converter units) and an event threshold of 40 electrons (38 adu).

These estimates were derived from MIT CSR subassembly calibration measurements, and reflect the spatially averaged detection efficiency of the device.

A simple "slab and stop" model of the gate structure (see K. Gendreau, MIT PhD. Thesis, 1995) was fit to relative quantum efficiency measurements of w215c2r at 4 energies (525 eV, 677 eV, 1740 eV and 2015 eV), using reference detector w190c3 as a quantum efficiency standard. The spatially averaged relative efficiencies used in the model fit are 1.002, 1.018,1.000,and 0.993 at these four energies; formal statistical errors in these relative quantum efficiency measurements are about ±4x10-4. Systematic errors in these relative measurements are thought to be less than 3%.

The detection efficiency of the reference detector w190c3 was determined from fits to white-light data obtained at PTB/BESSY synchrotron storage ring. The assumed model parameters for the reference detector were:

Assumed Gate Structure Parameters for w190c3 Reference Detector
Si0.260 µmfit
Si020.351 µmfit
SiN30.026 µmfit
CSSiOx0.480000 µmfixed
CSWidth5.20000 µmfixed
CSSi1.00000 µmfixed

See ACIS Team preliminary calibration report (in preparation) for an explanation of the model parameters and other details. The error in the absolute efficiency of the reference detector is believed to be less than 5% in the energy range 0.3-4 keV.

The best-fit parameters for w215c2r used to derive this quantum efficiency curve are:

Si0.266 µmfit
Si020.339 µmfit
Si3N40.0200 µmfixed
CSSiOx0.480000 µmfixed
CSWidth5.20000 µmfixed
CSSi1.00000 µmfixed

The depletion depth, which determines the response at higher energies, was constrained using the 5.9 keV branching ratio method of Prigozhin; see the ACIS Team preliminary calibration report (in prepartion) for details. The best fit depletion depth is 65 ±2 µm. This estimate agrees, within 5%, with measurements made relative to the solid state beam normalization detector at the MSFC X-ray Calibration Facility during ACIS flat-field testing.

Intended use:

When combined with ancillary response functions (arf) representing the HRMA effective area and ACIS optical blocking filter transmission as functions of energy, and the normalized response function matrix previously release (w215c2r_norm.html), this quantum efficiency curve may be used for estimating counting rates and simulating spectra obtained from proposed ACIS/AXAF observations.

  1. Errors in this efficiency curve are believed to be no greater than 10%, except at energies in the immediate vicinity (within 20 eV) of characteristic absorption edges of N, O and Si, where near edge structure (which has been measured but not included in this prediction) is known to cause quantum efficiency variations of order 20-30%. XAFS will be included in subsequent releases of ACIS quantum efficiency curves.
  2. This quantum efficiency relation is expected to be fairly representative of all ACIS FI detectors. Spatial variations in quantum efficiency within one ACIS FI detector are generally less than 5% when averaged on 32x32 pixel scales. Detector to detector variations among ACIS FI devices can be as large as 15%.
  3. The gate structure models assumed here (slab and stop) are too simple and will be elaborated in subsequent releases.
  4. Pileup corrections have not been included in relative quantum efficiency measurements; these are known to be small (<1%). Pileup corrections in the absolute measurement of the reference detector efficiency are not based on the Jones model.
  5. This data in this file (w215c2r_eff_897.qdp) are replica of /suiko/d3/ti/Response/w215c2r/w215c2r.qdp of April 17 1997 14:58.

Mark W. Bautz
Thu Dec 4 18:33:45 EST 1997 Last modified: Thu Dec 4 19:12:09 EST