Does the Fe L-shell blend bias abundance measurements in intermediate-temperature clusters?

TL;DR

This study assesses how modeling the Fe L-shell blend affects iron abundance measurements in intermediate-temperature galaxy clusters, finding a modest 5-8% systematic bias that supports current measurement reliability.

Contribution

It provides the first detailed quantification of Fe L-shell modeling systematics in galaxy clusters within 2.5-4.5 keV temperature range, using XMM-Newton data.

Findings

L-induced systematics are 5-6% across the temperature range.

Joint MOS and pn fits increase the bias slightly to 7-8%.

No significant dependence on X-ray flux was observed.

Abstract

In intermediate-temperature (T = 2.5 - 4.5 keV) galaxy clusters, abundance measurements are almost-equally driven by Fe K and L transitions, at $\sim$ 6.7 keV and 0.9 - 1.3 keV, respectively. While K-shell-derived measurements are considered reliable, the resolution of the currently available instrumentation, as well as our current knowledge of the atomic processes, makes the modelling of the L-line complex challenging, resulting in potential biases for abundance measurements. In this work, we study systematics related to the modelling of the Fe L-line complex that may influence iron-abundance measurements in the intermediate-temperature range. To this aim, we select a sample of three bright galaxy clusters, with long XMM-Newton observations available and temperature in the 2.5 - 4.5 keV range. We fit spectra extracted from concentric rings with APEC and APEC+APEC models, by alternatively excluding the L and K bands, and derive the fractional difference of the metal abundances, $\Delta Z/Z$, as indication of the consistency between K- and L-shell-derived measurements. The $\Delta Z/Z$ distribution is then studied as a function of the cluster radius, ring temperature and X-ray flux. The L-induced systematics, measured through an individual fit of each MOS and pn spectrum, remain constant at a 5 - 6% value in the whole 2.5 - 4.5 keV temperature range. Conversely, a joint fit of MOS and pn spectra leads to a slight excess of 1 - 2% in the above estimate. No significant dependence on the ring X-ray flux is highlighted. The measured 5 - 8% value indicates a modest contribution of the systematics to the derived iron abundances, giving confidence for future measurements. To date, these findings represent the best-achievable estimate of the systematics in analysis, while future microcalorimeters will significantly improve our understanding of the atomic processes underlying the Fe L emissions.