The X-ray integral field unit (X-IFU) proposed for ESA’s Athena X-ray observatory will consist of 3840 transition edge sensor (TES) microcalorimeters optimized for the energy range of 0.2 to 12 keV. The instrument will provide unprecedented spectral resolution of ~ 2.5 eV at energies of up to 7 keV and will accommodate photon fluxes of 10’s mcrab (1000’s cps).
Over the past two years the baseline configuration has evolved from the original proposal. The current baseline consists of a uniform large pixel array (LPA) of 5” pixels, AC-biased within their superconducting-to-normal transition and read out using frequency domain multiplexing (FDM). The baseline pixel design is approximately a factor of two times slower than in the original concept. High count-rate accommodation, needed for bright point source observations, is now achieved by defocusing the telescope optic to spread the photons over a larger number of pixels. In this paper we report on Mo/Au TES designs that are being optimized to meet the baseline pixel parameters and performance goals. This includes detailed studies on the optimization of the thermal heat sink and the impact of different TES geometries (including TES size and normal metal feature geometries) on the DC-biased transition shape. We discuss how these geometric effects ultimately impact important performance metrics such as energy resolution, decay time, slew-rate and array scale uniformity.
Our Mo/Au TESs have historically been designed and optimized for DC-biased operation, however, the primary readout technology uses an AC drive to bias the TES. Depending upon the drive frequency, the AC bias affects the TES transition shape in two ways. Firstly, due to losses from the bias current coupling to metallic components in close proximity to the TES and secondly introducing fine structure in the transition due to Josephson effects. We present latest pixel design optimizations targeted at mitigating these frequency dependent effects in order to achieve commensurate performance with that obtained in the DC case.