ALBERT

Description: The different electron distributions in the hard and soft spectral states of black hole binaries could be caused by kinetic processes and changing because of varying physical conditions in the corona. In the presence of a magnetic field in the corona, the electron distribution can appear thermal, even when acceleration mechanisms would produce non-thermal distributions. This is due to fast and efficient thermalization through synchrotron self-absorption. In this paper, we have analysed data from 6 yr of observations of Cygnus X-1 with the INTEGRAL observatory and produced 12 high-quality, stacked broad-band hard X-ray spectra representative of the whole range of spectral shapes observed in this source. We then fit these spectra with hybrid thermal/non-thermal Comptonization models and study the evolution of the physical parameters of the accretion flow across the spectral transition. In particular, we use the belm model to constrain the magnetic field in the corona through its effects on the coronal emission. Indeed, the hot electrons of the X-ray corona produce soft (optical–UV) synchrotron radiation which is then Comptonized and may affect the temperature of the electrons (and thus the slope of the X-ray spectrum) through Compton cooling. We find that in the softer states, the emission is dominated by Comptonization of the disc photons and the magnetic field is at most of the order of 10 6 G. In the harder states, the data are consistent with a pure synchrotron self-Compton model, although a significant contribution of Comptonization of disc photons may not be excluded. If the non-thermal excess observed above a few hundred keV in the hard state is produced in the same region as the bulk of the thermal Comptonization, we obtain an upper limit on the coronal magnetic field of about 10 5 G. If, on the other hand, the non-thermal excess is produced in a different location (such as the jet or a different part of the corona), the constraints on the magnetic field in the hard state are somewhat relaxed and the upper limit rises to ~10 7 G. We discuss these constraints in the context of current accretion flow models.