Sunayana Bhargava
In recent studies, there have been several reports of a detection of an unexplained X-ray emission at 3.5 keV in various astrophysical systems. One interpretation of this excess is from the decay of sterile neutrino dark matter. The most influential study to date analysed 73 galaxy clusters observed by ESA's XMM-Newton satellite. This work supersedes previous studies by searching for evidence for a 3.5 keV excess in the spectra of 117 optically and X-ray confirmed galaxy clusters - the largest sample of its kind. In our analysis of individual spectra, we are able to identify three systems with an excess of flux at 3.5 keV (one of which might be due to a discrete emission line of unknown origin). We then bin the remaining 114 clusters according to their X-ray temperature in order to search for an increase in 3.5 keV flux with temperature (a reliable proxy for dark matter halo mass). However, we find no evidence for a positive trend, thereby placing the most stringent upper limit on a non-detection of a line from cluster studies thus far. We conclude that a highly debated 3.5 keV flux excess from multiple X-ray studies is a non-ubiquitous feature in galaxy clusters that does not scale according to the cluster's underlying halo properties, and is therefore unlikely to originate from sterile neutrino dark matter.
The work presented here supersedes previous studies by searching for evidence for a 3.5 keV excess in the spectra of 117 optically and X-ray confirmed galaxy clusters - the largest sample of its kind. The clusters are between redshifts of 0.1 and 0.6. A key aim of this study is to identify whether the cluster’s X-ray temperature scales with the appearance of a 3.5 keV excess, given that temperature is a reliable tracer of underlying dark matter mass. Only clusters with reliable temperaturs are used for this study. The final sample contains 117 clusters within the range 1 to 11 keV.
The methodology requires firstly to blueshift every cluster spectrum to the rest frame (redshift equal to zero), in order to smear out artefacts in the data, which may contaminate the line search. Next, three separate but related spectral tests are performed. The first is to search for the line in each cluster individually, then in four bins of different temperature ranges, and finally all the clusters in the sample. We use a well-motivated plasma model to fit the baseline cluster spectrum. We then iterate a Gaussian line between 3.5 and 3.6 keV, searching for a fit improvement relative to the baseline model. In departure from previous studies, we favour a simultaneous fitting approach for multiple clusters instead of stacking all spectra into one composite spectrum.
In our analysis of individual spectra, we are able to identify three systems with an excess of flux at 3.5 keV. These are described by three figures where the red line (the fit improvement) is above the 3-sigma threshold of a “significant detection” at approximately 3.5 keV. We suggest the feature in one of these clusters may be due to a discrete emission line of unknown origin. We then bin the remaining 114 clusters according to their X-ray temperature in order to search for an increase in 3.5 keV flux with temperature. However, we find no evidence for a positive trend, thereby placing the most stringent upper limit on a non-detection of a line from cluster studies thus far. We quantify such a non-detection based on the flat red line at 3.5 keV, highlighting an absence of any fit improvement when adding a Gaussian line to the baseline model. We conclude that a highly debated 3.5 keV flux excess from multiple X-ray studies, while present, is a non-universal feature in galaxy clusters that does not scale according to the cluster's dark matter properties, and is therefore unlikely to originate from sterile neutrino dark matter.