The quantum efficiency and diffractive image artifacts of Si:As IBC mid-IR detector arrays at 5 $-$ 10 $\mu$m: Implications for the JWST/MIRI detectors

TL;DR

This paper examines the high quantum efficiency of Si:As IBC mid-IR detectors used in JWST/MIRI, models the origin of diffractive artifacts caused by internal reflections, and discusses implications for space-based infrared observations.

Contribution

It provides a detailed analysis of the factors leading to high QE in MIRI detectors and models the diffraction artifacts affecting their imaging performance.

Findings

Quantum efficiency reaches up to 60% between 5-10 μm.

The diffractive artifact results from internal reflections off metallic contacts.

Artifact properties depend on detector architecture and photon diffraction paths.

Abstract

Arsenic doped back illuminated blocked impurity band (BIBIB) silicon detectors have advanced near and mid-IR astronomy for over thirty years; they have high quantum efficiency (QE), especially at wavelengths longer than 10 $\mu$m, and a large spectral range. Their radiation hardness is also an asset for space based instruments. Three examples of Si:As BIBIB arrays are used in the Mid-InfraRed Instrument (MIRI) of the James Webb Space Telescope (JWST), observing between 5 and 28 $\mu$m. In this paper, we analyze the parameters leading to high quantum efficiency (up to $\sim$ 60\%) for the MIRI devices between 5 and 10 $\mu$m. We also model the cross-shaped artifact that was first noticed in the 5.7 and 7.8 $\mu$m Spitzer/IRAC images and has since also been imaged at shorter wavelength ($\le 10~\mu$m) laboratory tests of the MIRI detectors. The artifact is a result of internal reflective diffraction off the pixel-defining metallic contacts to the readout detector circuit. The low absorption in the arrays at the shorter wavelengths enables photons diffracted to wide angles to cross the detectors and substrates multiple times. This is related to similar behavior in other back illuminated solid-state detectors with poor absorption, such as conventional CCDs operating near 1 $\mu$m. We investigate the properties of the artifact and its dependence on the detector architecture with a quantum-electrodynamic (QED) model of the probabilities of various photon paths. Knowledge of the artifact properties will be especially important for observations with the MIRI LRS and MRS spectroscopic modes.