Insight-HXMT observations of jet-like corona in a black hole X-ray binary MAXI J1820+070

Black hole X-ray binaries (BHXRBs) exhibit transitions between hard and soft spectral states during outbursts. The hard state is dominated by Comptonization of photons in a hot corona, while the soft state shows a strong blackbody component from the accretion disk and a weaker high-energy tail. MAXI J1820+070 is a BHXRB discovered in 2018, observed across various wavelengths. Its distance has been measured through radio parallax. Insight-HXMT provided extensive broadband X-ray observations (1-250 keV) of MAXI J1820+070 during its outburst, offering a unique opportunity to study the evolution of its accretion flow. Previous studies using NICER data suggested coronal contraction during the outburst, while radio observations revealed a twisted compact X-ray jet. This paper focuses on analyzing the Insight-HXMT spectral data to understand the evolution of the accretion flow, particularly the corona and its interaction with the disk, during the high-energy-dominated hard state.

Several studies have explored the spectral and timing properties of BHXRBs. Fender and Belloni (2004) reviewed disc-jet coupling, while Done et al. (2007) modeled accretion flows. Yuan and Narayan (2014) focused on hot accretion flows, and Yan et al. (2020) examined coronal properties using NuSTAR and Swift data. The MAXI mission (Matsuoka et al., 2009) provided initial detection of MAXI J1820+070, with subsequent optical and radio observations (Baglio et al., 2018; Bright et al., 2018). The discovery of low-frequency quasi-periodic oscillations (LFQPOs) in both X-ray and optical bands (Gandhi et al., 2018; Yu et al., 2018) added further complexity. Previous X-ray observations by Swift (Kennea et al., 2018), NuSTAR (Buisson et al., 2018), and NICER (Homan et al., 2018) provided additional data. Specifically, Kara et al. (2019), using NICER data, suggested that the corona contracts during the outburst. The present work leverages the unique broad-band capabilities of Insight-HXMT to further investigate this.

The study utilized data from Insight-HXMT's High-Energy (HE), Medium-Energy (ME), and Low-Energy (LE) X-ray telescopes. Data reduction involved filtering based on criteria like pointing offset, Earth elevation, geomagnetic cut-off rigidity, and proximity to the South Atlantic Anomaly. Spectral analysis was performed using XSPEC, focusing on the first outburst of MAXI J1820+070. The spectra were modeled using a combination of components: tbabs (to account for galactic absorption), diskbb (for accretion disk radiation), relxillCp (relativistic reflection model from the corona), and xillverCp (non-relativistic reflection model for a narrow-line component). A Markov Chain Monte Carlo (MCMC) algorithm was employed to obtain best-fitting parameters. The reflection fraction (Re), defined as the ratio of coronal intensity illuminating the disk to the coronal intensity reaching the observer, was a key parameter of interest. The time evolution of parameters such as electron temperature, reflection fraction, emissivity profile index, and photon index were studied throughout the outburst decay. To investigate the effect of coronal bulk motion on the reflection fraction, the authors utilized the relxilllpionCp package, which incorporates relativistic outflow in the reflection model. Further analyses included evaluating the equivalent width (EW) of the iron line and exploring potential degeneracies between fitted parameters, such as iron abundance and disk density.

The spectral analysis revealed that the reflection fraction (Rf) increased during the rising phase and decreased during the decay phase of the outburst. The electron temperature showed a corresponding decrease during the rise and a slow increase during the decay. The emissivity profiles were found to be relatively flat, suggesting a spatially extended corona. The decrease in Rf during the decay, despite evidence of coronal contraction (from previous NICER data), was attributed to the outflowing motion of the corona. Simulations using relxilllpionCp showed that increased outflow velocity reduces the reflection fraction. The authors conclude that the corona contracts towards the black hole but simultaneously accelerates its outflow, leading to a net decrease in reflection fraction during the decay. The broad iron Kα line remained relatively stable, while the strength of the narrow component decreased over time. Analyses of other spectral parameters, such as the ionization parameter and iron abundance, were also conducted and discussed considering potential model limitations.

The observed decrease in reflection fraction during the decay phase, despite coronal contraction, strongly supports the model of an outflowing corona. The combination of spectral and timing data (contracting corona from NICER and decreasing reflection fraction from Insight-HXMT) provides a consistent picture of a dynamic corona. The model of a jet-like corona, interpreted as a standing shock, provides a physical explanation linking the X-ray emission region to jet formation. The study emphasizes the importance of considering both the position and bulk velocity of the corona to understand particle acceleration near the black hole. The finding of a faster outflow at closer proximity to the black hole has broader implications for understanding the dynamics of accretion flows in BHXRBs and potentially AGNs.

This study demonstrates the importance of broad-band X-ray observations in understanding the dynamic nature of accretion flows in BHXRBs. The observation of a decreasing reflection fraction despite coronal contraction strongly suggests a faster outflowing corona as it moves closer to the black hole. This finding supports a jet-like corona model and highlights the need for future studies incorporating both spectral and timing information to constrain the physical parameters of accretion flows in similar systems. Future research could focus on refining models to account for high-density effects and improve the simultaneous constraint of outflow velocity and coronal height.

The study acknowledges limitations in uniquely determining the outflow velocity and coronal height using spectral fitting alone. The constant density assumption in the relxillCp model may lead to artificially high iron abundances. The simplified model for the continuum may also influence the accuracy of equivalent width measurements. Future work with improved models that incorporate more realistic density profiles and more sophisticated modeling techniques will help to overcome these limitations.