XRISM Spectroscopy of the Stellar-mass Black Hole GRS 1915+105

TL;DR

The XRISM/Resolve spectrum of the stellar-mass black hole GRS 1915+105 reveals a warm, Compton-thick outer-disk obscuring the central engine in a sub-Eddington state. The emission is dominated by narrow, photoionized lines from He-like and H-like ions, modeled with three ionization zones plus resonant scattering and a distant neutral Fe K component, indicating an outer-disk atmosphere or slow wind at $r\sim10^{6}\,GM/c^{2}$. A warped, precessing outer disk provides a coherent geometric picture consistent with jet-angle changes and JWST IR data, though several model simplifications and the Fe XXVI RRC oscillations invite caution and further study. Overall, the work strengthens the AGN-like analogy for the obscured state of GRS 1915+105 and highlights the diagnostic power of high-resolution X-ray spectroscopy for accretion-disk atmospheres in X-ray binaries.

Abstract

GRS 1915$+$105 was the stellar-mass black hole that best reproduced key phenomena that are also observed in Type-1 active galactic nuclei. In recent years, however, it has evolved to resemble a Type-2 or Compton-thick AGN. Herein, we report on the first XRISM observation of GRS 1915$+$105. The high-resolution Resolve calorimeter spectrum reveals that a sub-Eddington central engine is covered by a layer of warm, Compton-thick gas. With the obscuration acting as a coronagraph, numerous strong, narrow emission lines from He-like and H-like charge states of Si, S, Ar, Ca, Cr, Mn, Fe, and Ni dominate the spectrum. Radiative recombination continuum (RRC) features are also observed, signaling that much of the emitting gas is photoionized. The line spectrum can be fit by three photoionized emission zones, with broadening and bulk velocities suggestive of an origin in the outer disk atmosphere and/or a slow wind at $r \simeq 10^{6}~GM/c^{2}$. The Fe XXV He-$α$ and Fe XXVI Ly-$α$ lines have a broad base that may indicate some emission from $r \sim 3\times 10^{3}~GM/c^{2}$. These results broadly support a picture wherein the current state in GRS 1915$+$105 is due to obscuration by the irradiated outer disk. This could arise through disk thickening if the Eddington fraction is higher than inferred, but it is more likely due to a warped, precessing disk that has brought the outer disk into the line of sight. We discuss the strengths and weaknesses of this interpretation and our modeling, and possible explanations of some potentially novel spectral features.

Figures (6)

• Figure 1: The XRISM/Resolve spectrum of GRS 1915$+$105 in its obscured state. The model in red is the best-fit model detailed in Table 1. It consists of a highly obscured disk plus power-law continuum and three photoionized emission zones in the outer accretion disk, weakly shaped by resonant scattering. The strongest lines are from He-like and H-like charge states of Si, S, Ar, Ca, Cr, Mn, Fe, and Ni. Solid vertical lines mark the lab energies of the observed He-like (top row) and H-like transitions (bottom row). A weak, narrow, neutral Fe K$_{\alpha}$ emission line is clearly detected at 6.4 keV, signaling the presence of cold gas at intermediate disk radii. Several RRC features are marked with triangles, including from He-like and H-like Fe at 8.8 and 9.3 keV, respectively. The data are binned using the "optimal" binning algorithm.
• Figure 2: The XRISM/Resolve spectrum of GRS 1915$+$105, shown in 2 keV segments. The model in red is the best-fit model detailed in Table 1. Solid vertical lines mark the lab energies of the observed He-like and H-like lines, as well as neutral Fe. The red-shift of the bulk of the emitting gas relative to lab wavelengths is evident, especially in the strongest Fe XXV and Fe XXV lines. RRC features are marked with triangles.
• Figure 3: The XRISM/Resolve spectrum of GRS 1915$+$105, in the band containing He-like and H-like lines from Cr, Mn, Fe, and Ni. The model in red is the best-fit model detailed in Table 1. Solid vertical lines mark the lab energies of the observed He-like (top row) and H-like transitions (bottom row). RRC features are marked with triangles.
• Figure 4: The XRISM/Resolve spectrum of GRS 1915$+$105 in the vicinity of the strongest emission line complexes (He-like Fe XXV, centered at 6.7 keV, and H-like Fe XXVI centered at 6.97 keV). The total best-fit model is shown in red, with individual photoionized components shown in blue (pion$_{1}$), cyan (pion$_{2}$) and orange (pion$_{3}$). In these and other lines from adjacent charge states, it is clear that multiple components are needed to describe He-like and H-like line fluxes, and detailed line shapes. It is evident that the lines are red-shifted with respect to their laboratory energy values, marked in gray at the top of the frame. Note that the component colors adopted in this plot are not connected to the elemental color scheme in other plots.
• Figure 5: The XRISM/Resolve spectrum of GRS 1915$+$105 in the band containing Fe XXV and Fe XXVI RRC features (marked with triangles). The best-fit model in other figures is now shown in blue. The Fe XXVI RRC is clearly not smooth, as expected if recombining electrons are thermally distributed; instead, oscillatory structure with a spacing of 30 eV is evident. This structure is nominally consistent with recombination from Landau levels in a highly magnetized corona ($B = 2.5\times 10^{9}$ Gauss; this model is shown in red). Alternatively, the lines could arise in a $v=0.34c$ flow wind, wherein iron retains intermediate charge states. Both explanations are unsatisfactory.