Search for X-ray emission from subdwarf B stars with compact companion candidates

Stellar evolutionary models predict that most of the early type subdwarf stars in close binary systems have white dwarf companions. More massive companions, such as neutron stars or black holes, are also expected in some cases. The presence of compact stars in these systems can be revealed by the detection of X-rays powered by accretion of the subdwarf’s stellar wind or by surface thermal emission. Using the Swift satellite, we carried out a systematic search for X-ray emission from a sample of twelve subdwarf B stars which, based on optical studies, have been suggested to have degenerate companions. None of our targets was detected, but the derived upper limits provide one of the few observational constraints on the stellar winds of early type subdwarfs. If the presence of neutron star companions is confirmed, our results constrain the mass loss rates of some of these subdwarf B stars to values <10^{-13}-10^{-12} Msun/yr.

💡 Research Summary

The authors performed a systematic X‑ray search for compact companions in subdwarf B (sdB) binaries using the Swift X‑Ray Telescope (XRT). Twelve sdB stars were selected on the basis of prior optical radial‑velocity studies (Geier et al. 2010) that suggested the presence of unseen massive companions—white dwarfs (WD), neutron stars (NS), or black holes (BH). The selection criteria emphasized low Galactic absorption (N_H < 5 × 10²⁰ cm⁻²) and short orbital periods (0.1–0.8 d) to maximise the expected accretion luminosity from wind capture.

Swift/XRT observations were carried out between May and August 2011, with total exposure times ranging from ~3 ks to ~5 ks per target. All data were obtained in photon‑counting mode. No X‑ray source was detected at the position of any of the twelve sdB stars. For each field a 3σ upper limit on the count rate was derived, assuming a power‑law spectrum with photon index 2 and the measured N_H values. These count‑rate limits were converted to unabsorbed 0.3–10 keV luminosity limits, typically 10³⁰–10³¹ erg s⁻¹.

To interpret the non‑detections, the authors assumed that any compact object would accrete from the sdB wind via Bondi‑Hoyle capture. The accretion radius R_a = 2GM/(v_o²+v_W²) depends on the compact object mass (M), the orbital relative velocity (v_o), and the wind velocity (v_W). They adopted v_W equal to the sdB escape velocity as a conservative estimate. For white dwarf companions they used M = 0.6 M⊙ and R = 10⁹ cm; for neutron stars M = 1.4 M⊙ and R = 10⁶ cm; for the black‑hole candidate they assumed M = 6 M⊙ with a 10 % accretion efficiency. Using the observed luminosity limits and L_X ≈ GMṀ/R, they derived upper limits on the wind mass‑loss rate (Ṁ_W).

The results show that for systems likely hosting white dwarfs the wind loss must be Ṁ_W ≲ 5 × 10⁻¹¹ M⊙ yr⁻¹, while for the four systems where a neutron star or black hole is suspected the limits are much tighter, Ṁ_W ≲ 3 × 10⁻¹³ M⊙ yr⁻¹. These values are well below the predictions of radiatively driven wind models (Vink & Cassisi 2002), which give Ṁ_W ≈ 10⁻¹²–10⁻¹¹ M⊙ yr⁻¹ for solar metallicity. The authors argue that only if the sdB stars have sub‑solar metallicities (Z ≲ 0.3 Z⊙) could the theoretical rates be reconciled with the observational limits.

The paper discusses several caveats. The wind velocity may be lower than the escape speed, which would increase the accretion radius and thus the expected X‑ray luminosity, making the limits even more stringent. The assumed power‑law spectrum is appropriate for accretion‑powered emission but would not capture thermal emission from a hot white dwarf or neutron star surface, which peaks below Swift’s most sensitive band. Moreover, Swift’s modest sensitivity means that very weak winds remain undetectable.

Despite these limitations, this work provides the first systematic X‑ray constraints on sdB stellar winds and on the presence of compact companions in such binaries. The authors suggest that deeper observations with more sensitive X‑ray observatories (e.g., XMM‑Newton, Chandra) targeting the closest and least absorbed candidates (such as PG 0101+039 or CPD −64 481) could detect accretion signatures if the wind mass‑loss rates are at the level of 10⁻¹² M⊙ yr⁻¹, as predicted by theory. Confirming or refuting the presence of neutron stars or black holes in these systems would have important implications for binary evolution pathways, including the formation of Type Ia supernova progenitors.