Chandra reveals 84 X-ray sources in a little-explored energy band

How many luminous objects escape a survey because they emit in the very band it does not examine? That question drove a new analysis of the Chandra space observatory’s archive. Rather than looking farther away, Mustafa Muhibullah and colleagues at the University of Alabama and the Center for Astrophysics | Harvard & Smithsonian revisited observations of six nearby galaxies and changed the energy range used in the search. The study was published September 9 in Nature Astronomy.

The researchers selected points that appeared between 0.15 and 0.3 kiloelectron volts, a unit describing the energy of each particle of light, but were not significantly detected above 0.3. That upper value commonly serves as the lower cutoff in X-ray surveys. The team compared images of the same region in different bands, combined archival observations taken through 2017, and visually inspected the candidates. Using older records took advantage of the instrument’s greater sensitivity to these energies early in the mission.

The selection yielded 84 sources, with seven to 21 in each galaxy. When the lower cutoff was raised from 0.15 to 0.2 kiloelectron volts, a more reliable range for the detector, 81 of the 84 remained detectable. In 16 comparison fields outside the galaxies, the authors found two similar candidates. Using that control, they estimated that objects unrelated to the studied galaxies contaminated the sample at a level below roughly 3%. The population therefore appears unlikely to be merely a product of selecting noise or foreground sources.

The chosen name, hypersoft X-ray sources, describes the exceptionally low energy of the detected radiation. The brightest objects approach 10³⁸ ergs per second—about 10³¹ watts—in the narrow band examined alone, according to estimates that correct for radiation absorbed on its way to us and assume equal emission in all directions. Emission models suggest that an even greater share of the energy may emerge in the extreme ultraviolet, which gas between stars readily absorbs. That total power is an extrapolation, not a direct ultraviolet measurement.

There are indications that part of the sample contains white dwarfs, compact stellar remnants that can receive matter from a companion. At least seven of the 18 sources selected in central Andromeda coincide with systems known to have undergone nova explosions. Nuclear burning at the surface during these events can keep the remnant hot. For other members of the sample, the authors consider systems containing black holes or neutron stars. The same radiation pattern may thus encompass different physical engines.

Identifying the nature of the sources will make it possible to investigate whether some are precursors of Type Ia supernovae, explosions that destroy white dwarfs. It will also allow estimates of how much their photons contribute to stripping electrons from gas in galaxies. The survey now provides a concrete sample for tackling those questions and shows that part of a galaxy’s inventory may depend on where the observational boundary is drawn.