Observing gravitational waves from the first generation of black holes

The properties of the first generation of black-hole seeds trace and distinguish different models of formation of cosmic structure in the high-redshift universe. The observational challenge lies in identifying black holes in the mass range ~100-1000 solar masses at redshift z~10. The typical frequencies of gravitational waves produced by the coalescence of the first generation of light seed black-hole binaries fall in the gap between the spectral ranges of low-frequency space-borne detectors (e.g., LISA) and high-frequency ground-based detectors (e.g., LIGO, Virgo and GEO 600). As such, these sources are targets for proposed third-generation ground-based instruments, such as the Einstein Telescope which is currently in design study. Using galaxy merger trees and four different models of black hole accretion - which are meant to illustrate the potential of this new type of source rather than to yield precise event-rate predictions - we find that such detectors could observe a few to a few tens of seed black-hole merger events in three years and provide, possibly unique, information on the evolution of structure in the corresponding era. We show further that a network of detectors may be able to measure the luminosity distance to sources to a precision of ~40%, allowing us to be confident of the high-redshift nature of the sources.