In 2014 two groups reported an unidentified X-ray emission line near 3.5 keV in stacked spectra of galaxy clusters and in Andromeda (Bulbul et al.; Boyarsky et al.), at energies matching no strong known atomic transition. The exciting interpretation was decaying dark matter: a 7 keV sterile neutrino decaying to a photon and an active neutrino would produce exactly such a line, scaling with dark matter column density. A decade of follow-up has been a slow deflation: deep observations of the Milky Way halo and dwarf spheroidals, where the dark matter signal should be cleanest, found nothing; Hitomi's high-resolution look at Perseus saw no line at the claimed strength; blank-sky analyses using XMM-Newton data excluded the sterile-neutrino parameter space at high significance (Dessert et al. 2020); and eROSITA's all-sky data have not confirmed the feature. Defenders contest the exclusion methodologies, and the original detections remain formally unretracted.
For ΛCDM the line was never a requirement, but its trajectory is symptomatic: the model needs some dark matter particle, the 3.5 keV line was among the most promising indirect candidates, and its collapse joins the WIMP direct-detection nulls, the axion non-detections, and the collider absences in an inventory of vanished candidates. Each null individually constrains parameter space; collectively they erode the expectation that the particle exists to be found.
The standing is a disputed residual: most of the community treats the line as atomic (potassium and chlorine blends, charge exchange) or instrumental, a minority maintains the signal with environment-dependent decay properties, and XRISM's high-resolution spectroscopy is delivering the decisive data on cluster after cluster.