An upper limit on the frequency of short-period black hole companions to Sun-like stars

TL;DR

This study searches for close black hole companions to Sun-like stars using TESS data and finds no evidence, setting a strict upper limit on their frequency and challenging some existing population models.

Contribution

The paper provides the first observational upper limit on the frequency of short-period black hole companions to Sun-like stars, based on a large TESS survey and spectroscopic follow-up.

Findings

No close black hole companions detected among 250 candidates.

Fewer than 1 in 100,000 Sun-like stars host such black holes.

Results challenge some optimistic black hole population models.

Abstract

Stellar-mass black holes descend from high-mass stars, most of which had stellar binary companions. However, the number of those binary systems that survive the binary evolution and black hole formation is uncertain by multiple orders of magnitude. The survival rate is particularly uncertain for massive stars with low-mass companions, which are thought to be the progenitors of most black hole X-ray binaries. We present a search for close black hole companions (separations less than 20 solar radii) to AFGK-type stars in TESS, i.e. the non-accreting counterparts to and progenitors of low-mass X-ray binaries. Such black holes can be detected by the tidally induced ellipsoidal deformation of the visible star, and the ensuing photometric light-curve variations. From an initial sample of 4.7 million TESS stars, we have selected 457 candidates for such variations. However, spectroscopic followup of 250 of them shows that none are consistent with a close black hole companion. On the basis of this non-detection, we determine (2 $\sigma$ confidence) that fewer than one in $10^5$ Solar-type stars in the Solar neighbourhood host a short-period black hole companion. This upper limit is in tension with a number of ``optimistic'' population models in the literature that predict short-period black hole companions around one in $10^{4-5}$ stars. Our limits are still consistent with other models that predict only a few in $10^{7-8}$.