Constraining the nature of the most extreme Galactic particle accelerator. H.E.S.S. observations of the microquasar V4641 Sgr

TL;DR

V4641 Sgr is a microquasar hosting a compact object with jet activity, interpreted as a potential site for multi-PeV Galactic particle acceleration. Using ~100 hours of H.E.S.S. data complemented by HI and CO gas measurements, the study resolves the gamma-ray emission around the system as highly elongated and spectrally hard, with a broad-band SED peaking near 100 TeV. The lack of dense ambient gas and X-ray constraints imply that a leptonic emission mechanism with rapid transport or a special environmental configuration best explains the observations, though a hadronic channel with extreme energetics cannot be completely ruled out. The work constrains acceleration sites and transport physics in microquasar environments and motivates future observations across X-ray, radio, and very-high-energy gamma-ray bands to pinpoint the origin of the emission and assess the Hadronic possibility in different environments.

Abstract

Microquasars have emerged as promising candidates to explain the cosmic-ray flux at petaelectronvolt energies. LHAASO observations revealed V4641~Sgr as the most extreme example so far. Using $\approx$100~h of H.E.S.S. data, we performed a spectro-morphological study of the gamma-ray emission around V4641~Sgr. We employed HI and dedicated CO observations of the region to infer the target material for cosmic-ray interactions. We detected multi-TeV emission around V4641~Sgr with a high significance. The emission region is elongated. We found a power-law spectrum with an index $\approx$1.8, and together with results from other gamma-ray instruments, this reveals a spectral energy distribution that peaks at energies of $\approx$100~TeV for the first time. We found indications (3$σ$) of a two-component morphology, with indistinguishable spectral properties. The position of V4641~Sgr is inconsistent with the best-fit position of the single-component model and with the dip between the two components. We found no significant evidence of an energy-dependent morphology. No dense gas was found at any distance towards V4641~Sgr. The peak of the SED at $\approx$100~TeV identifies V4641~Sgr as a candidate cosmic-ray accelerator beyond the so-called knee. The absence of dense target gas places stringent energetic constraints on hadronic interpretations, however. The H.E.S.S. measurement requires an unusually hard ($\approx 1.5$) spectral index for the protons. A leptonic scenario faces fewer obstacles if the particle transport is fast enough to avoid losses and to reproduce the observed energy-independent morphology. The absence of bright \xray emission across the gamma-ray emission region requires a magnetic field strength $\lesssim3$~$μ$G, however. Our findings favour a leptonic origin of the gamma-ray emission. This conclusion does not exclude hadron acceleration in the V4641~Sgr system.

Figures (13)

• Figure 1: Statistical significance of the H.E.S.S. excess counts with energies higher than 0.8 TeV above the background of nearly isotropic cosmic rays (indicated by the colour scale) before statistical trials were accounted for. The 68% containment region of the H.E.S.S. PSF is denoted with a white circle (left). The map was derived using a maximum likelihood test for a Gaussian kernel of radius 0.06$\degree$, which resulted in smoothing. The scale resulting from the combination of the PSF and this kernel is shown with another white circle (right).
• Figure 2: Left: Significance map. The 68% containment regions (corresponding to 1.5$\sigma$ for a 2D Gaussian) of the spatial models are overlaid. The best-fit extent is depicted with dashed lines in teal for model A (see Section \ref{['subsec:morphology']}), and in red and yellow for models B1 and B2 (see Section \ref{['subsec:morphology2']}), respectively. The best-fit position and its 95% confidence intervals are shown with crosses and solid lines using the same colour scheme. Middle: The measured spectral energy distribution (SED) for models A (teal circles) and B1 (red squares). The solid line depicts the spectral shape of the best-fit power law. The dark and light shaded areas depict the statistical and systematic error bands, respectively. The error bars represent the combined systematic and statistical errors. The upper limits are shown at the 95% confidence level. Right: The symbols follow those of the middle panel, but for models A and B2 (yellow diamonds).
• Figure 3: Broad-band spectral energy distribution. Flux points and upper limits measured by H.E.S.S. (red circles), HAWC (squares), LHAASO (diamonds), and Fermi-LAT. The Fermi-LAT points are shown both as published in Zhao2025 (light triangles) and re-scaled to the H.E.S.S. emission region size (dark triangles). The broad-band spectrum is observed to rise sharply until a peak forms at energies of $\approx$100 TeV, further extending up to energies of several hundreds of TeV.
• Figure 4: Profiles along and across the emission. Centre: Rotated and scaled version of the significance map is shown to help visualise the geometry of the different regions used to derive profiles. Top: The red circles represent the measured flux above 0.8 TeV along the major axis as a function of distance to V4641 Sgr. Distances are provided in degrees and parsecs adopting a distance of 6.2 kpc. The dashed and dotted lines depict the predictions of models A and B, respectively. Bottom: The gamma-ray spectral index measured along the major axis in the same regions. Square and round symbols indicate when emission is detected inside the region with significance below or above $2.5\sigma$, respectively. A dotted grey line and shaded region indicate the best-fit value and statistical uncertainty of the index parameter in model A. The best-fit index at distance $\approx 0.7\degree$ is negative and is thus not visible in the plot. The significance of the emission in that region is 0.3$\sigma$. Middle left: The symbols follow those of the bottom panel, but measured across the emission region. Middle right: The symbols follow those of the top panel, but measured across the emission region.
• Figure 5: Significance maps in the energy bands. The symbols and colours follow those of Figure \ref{['fig:significance']}, but for photon energies below 5 TeV (top left) and above (top middle). Significance maps derived with a larger smoothing kernel of 0.12$\degree$ are shown in the bottom row. The map for photon energies above 5 TeV is repeated to show the contours from the two HAWC significance maps provided in hawcv4641. Finally, the top right panel shows the size of the major axis in the "model A" fit for both H.E.S.S. (red points) and for the HAWC publicly available data (blue points). The dashed lines and shaded bands represent the best-fit size and uncertainty derived in larger energy ranges as presented in Table \ref{['tab:fit_params']} and hawcv4641. The error bars/bands represent the combined statistic and systematic uncertainty.