To date, 55 stellar mass black hole candidates have been detected in transient X-ray binaries (BHTs). These systems are characterized by sporadic outburst episodes which increase their brightness by several orders of magnitude (allowing their detection by X-ray telescopes) and long periods of quiescence. In 46 years of X-ray astronomy, only 18 of the 55 black hole candidates have been dynamically confirmed. This is a sample too small for a statistical analysis of the population as a class. Population synthesis studies suggest that transient black hole binaries in the Galaxy could amount to several thousand systems but we are only detecting the tip of the population, those that have gone into outburst. Therefore, it is necessary to develop methods to discover a large fraction of the hidden population of quiescent black holes. Another important goal is to enlarge the sample of dynamically confirmed black holes through detailed optical studies of the brighter black hole candidates.
Taking advantage of the large aperture of the Gran Telescopio Canarias (GTC), we obtained spectra of XTE J1859+226, a historical black hole candidate discovered during an outburst in 1999. Several orbital periods have been published but with large uncertainties in the measurements. In addition, a mass function of 7.4 M for the compact object was reported, implying the existence of a black hole, but it was never published in a refereed
journal. Combining photometry taken with several telescopes installed at the Observatorio del Roque de los Muchachos on La Palma between 2000 and 2010 and the spectroscopy obtained with GTC, we derived a value of the orbital period of 6.6 h and a radial velocity amplitude of the secondary star of 541 km s¿1 . These values imply a mass function of the compact object of 4.5 M and therefore the dynamical confirmation of the presence of a black hole, since it exceeds the maximum allowed mass for a neutron star, being the first black hole confirmed with data taken with GTC.
Swift J1357.2¿0933 is another black hole candidate discovered during an outburst in 2010. Its high Galactic latitude (b = + 50¿ ) and proximity (1.5 kpc) allowed its detection despite its low X-ray luminosity during the peak of the outburst (only Lx ~ 1035 erg s¿1 ). This probably makes Swift J1357.2¿0933 the first of a hitherto hidden population of transient systems with very low X-ray luminosity. Spectra taken during the decay reveal an orbital period of only 2.8 h which, combined with an estimate of the radial velocity of the inner disc, imply a mass function of 3.6 M and therefore the confirmation of a black hole. Furthermore, using high time resolution photometry, we traced the outward expansion of a wave in the accretion disc obscuring the compact object. This vertical structure, which had never been observed before, produces regular optical dips without an X-ray counterpart, which suggest that the binary is seen at a very high inclination. This makes the black hole
in Swift J1357.2¿0933 a scaled-down version of the supermassive black holes found in the centre of Seyfert 2 active galactic nuclei which are hidden by a torus of matter.
Once the sample of dynamically confirmed black holes is expanded with these two new systems, we have analysed the observational and dynamical properties of all the known BHTs. The distribution of Galactic black hole transient shows a concentration in the plane, with ~70% located between -5¿ and 5¿ in Galactic latitude. In addition, the visible quiescent spectra of these objects show emission lines produced in the accretion disc superimposed on the spectrum of the companion star. These lines are mainly generated by excitation of H and He, being especially strong the H¿ line.
Based on this, we decided to exploit the IPHAS and UVEX surveys to search for quiescent black holes in our Galaxy. These two surveys cover the Galactic plane visible from the Northern hemisphere in r' , i' , g' and RGO U along with the narrowband filters H¿ and He I ¿5875. We have computed a code to obtain the synthetic magnitudes in these bands through convolution of the spectra with the transmission curves of the filters above. Applying this code to our database of dynamically confirmed black holes, we have developed colour-colour diagrams combining filters from the two surveys. We have also performed a selection of other H¿ emitters that populate the Galactic plane and contaminate our search: cataclysmic variables, Be stars, symbiotic and T-Tauris. After analysing the position of all of them in the diagrams we have defined appropriate colour cuts that allow an intensive search of BHTs in the IPHAS and UVEX catalogues with minimum contamination from other H¿ emitters.
Applying these criteria to the diagrams (r' ¿ i' ) vs. (r' ¿H¿), (r' ¿H¿) vs. (HeI ¿ r' ) and (r' ¿ i' ) vs. (HeI ¿ r' ) diagrams, we have been able to reduce the number of candidates that fulfil all our colour cuts, from the initial 137 million to ~ 190, 000 objects. Finally, we also used infrared magnitudes from the 2MASS project. This allows us to define a new colour cut in the IR diagram (J ¿ H) vs. (H ¿ Ks) that reduces the sample to 11,752 objects. To confirm the nature of these objects, we have started observational campaigns using the WYFFOS/AF2 spectrograph and low resolution grisms. Of the ~ 1, 000 spectra obtained, we have analyzed ~400 and found only 6 with an H¿/He i ¿5875 flux ratio > 2. Only one of them has a extreme H¿/He i ¿5875 > 10, consistent with quiescent black hole candidates. Dedicated follow-up spectroscopy is required to confirm that this is a true quiescent black hole.