WFC3 TV3 Testing: Quantum Yield in the UV

The gain for the integrated WFC3 UVIS-1’ flight detector has been computed using the mean-variance method on flatfields taken in a small set of UV filters. These gains are compared to the gain computed in the visible (F606W) in order to estimate the quantum yield for the UVIS-1’ detector. The measured quantum yields are 1.07, 1.08, 1.03, and 1.00 e-/ photon, for F218W, F225W, F275W, and F336W, respectively, with errors of ~0.02 e-/photon, except for F218W, where the error is ~0.06 e-/photon due to some adverse effects attributed to the D2 lamp warm-up. The measured values are up to ~30% less than the predicted quantum yields (1.55, 1.45, 1.25, and 1.01e-/photon). Introduction At optical wavelengths, an incoming photon generates a single electron within the silicon of a CCD, i.e., the quantum yield is 1. In devices like WFC3, which are directly sensitive to the UV (unlike WFPC2, which had a lumogen coating to down-convert UV photons to visible wavelengths), there is a finite probability that a UV photon will generate more than one electron, or a quantum yield >1. This behavior can be a benefit, for example, in detecting faint targets in the UV. Quantum yield values are used in a variety of situations. Quantum efficiency (QE) curves are typically scaled downwards by the quantum yield, thereby avoiding QE values greater than 1 (Janesick 2001). Of course, applying a quantum yield correction that is too high can result in an underestimate of the actual QE of the device in the UV. An exposure time calculator (ETC) must account for quantum yields larger than 1 in order to provide the Association of Universities for Research in Astronomy, Inc., for the National Aeronautics and Space Administration. WFC3 Instrument Science Report 2008-47 correct predictions of the number of expected electrons in response to a particular input flux (e.g., when estimating exposure levels attainable without saturating the device), yet it must also provide correct S/N estimates (extra electrons cause the noise to deviate from a pure Poisson distribution based on the number of incoming photons). Both the WFC3 ETC (http://www.stsci.edu/hst/wfc3/tools/etc; see also Brown 2008) and the HST synthetic photometry package (http://www.stsci.edu/resources/software_hardware/stsdas/synphot; Laidler 2008) include the capability of specifying quantum yield terms. For WFC3, the quantum yield was initially expected to be 1.7 e-/photon at 200 nm, decreasing linearly to 1.0 at 340 nm; specifically, the factors were computed as λcrit / λ, where the critical wavelength is 339.4 nm which corresponds to the wavelength of the photon with an energy (3.65 eV) sufficient to generate an electon/hole pair in silicon (Janesick 2001). However, there were indications during TV3 that the quantum yield in WFC3 was not as large as expected, thus a program was included to measure the values in the UVIS channel. This report presents the results of that test. Observations and Analysis The flatfields used in the quantum yield analysis are summarized in Table 1. The UV data were taken with four different filters using the deuterium (D2) lamp of the internal calibration subsystem, while the F606W data were taken using the ground system optical stimulus (CASTLE). All but one set were read out in subarray (1024x1024) format; the subarray was positioned in the outer corner of the C amp, including the overscan regions so that a bias level subtraction could be performed. The full-frame set was part of the standard gain program and has been included here as a control sample: the same subarray area was extracted from the full-frames and processed in the same manner as the data taken as subarrays. Each set contained at least six pairs of images with exposure times chosen to provide exposure levels of ~500-30,000 DN at the nominal gain of 1.5 e-/DN. Table 1. List of flatfields. tvnum range filter observation date illumination data format 54142-54153 F218W 2008-03-27 D2 calsystem subarray 54154-54165 F225W 2008-03-27 D2 calsystem subarray 54166-54177 F275W 2008-03-27 D2 calsystem subarray 54178-54189 F336W 2008-03-27 D2 calsystem subarray 54190-54201 F606W 2008-03-27 CASTLE subarray 50946-50970, 50997 F606W 2008-03-12 CASTLE subsection of full-frame