- The paper reports the use of the Einstein Probe (EP) to identify and characterize a new X-ray transient, EP J175257.3–351923, with properties suggestive of a black hole low-mass X-ray binary (BH LMXB).
- Using multi-wavelength observations, Huangl et al. determined the source to have a peak 2-10 keV flux of ≈ 4e-10 erg cm-2 s-1 and a hard stable slope between ≈ 1.6 - 2.2.
- Among other finds, the fast-rise exponential-decay (FRED) outburst morphology indicated BH LMXB like morphology, with the hard-state persistence hinting at a possible failed outburst, though the lack of pulses or QPO's make the BH classification a candidate status, not a confirmation.
Overview
Huang et al. report the discovery and multiwavelength characterization of EP J175257.3–351923 (EP250916a), a new X-ray transient detected by the Einstein Probe (EP) Wide-field X-ray Telescope on 2025 September 16 at Galactic coordinates l=355.3090∘, b=−4.6086∘ (2606.10566). The source underwent a ≳40-day outburst with a peak 2–10 keV flux of ∼4×10−10 erg cm−2 s−1, remained in the hard spectral state throughout, and lacks a detectable optical counterpart. The authors classify it as a candidate black hole low-mass X-ray binary (BH LMXB) in a faint or very faint outburst, at an assumed distance of 8 kpc implying LX​∼1036 erg s−1. The identification of the compact object is explicitly not definitive; the BH interpretation rests primarily on the high coronal cutoff energy.
Discovery and follow-up campaign
The transient was first detected by EP/WXT and localized by EP/FXT to within 10″, then refined by Swift/XRT to R.A. = 17h52m57.3s, Dec. = −35°19′22.9″ with a 2.2″ (90%) uncertainty. The follow-up program combined EP/FXT, Swift/XRT (21 observations over roughly one month), SVOM/MXT and VT, NuSTAR (~21 ks on September 24), and GROND optical/NIR imaging. This campaign is representative of a broader pattern: over its first two years, EP has uncovered more than a dozen faint XRB candidates, including EP240809a, EP240904a, EP250315b, EP J174942.2–384834, and EP250623a. The paper's central observational claim is that EP's sensitivity now routinely reaches the faint (LX​∼1036−37 erg s−1) and very faint (b=−4.6086∘0 erg sb=−4.6086∘1) regimes that earlier all-sky monitors such as RXTE/ASM, Swift/BAT, and MAXI largely missed.
Optical constraints
Five catalogued optical/IR sources lie within or near the Swift/XRT error circle, but small-aperture UVOT photometry shows their fluxes vary by less than a factor of two over ~40 days while the X-ray flux varies by more than an order of magnitude, arguing against any of them being the counterpart. The authors derive b=−4.6086∘2 upper limits of b=−4.6086∘3 mag (Swift/UVOT) and b=−4.6086∘4 mag (SVOM/VT). Correcting for b=−4.6086∘5 mag at an assumed distance of 8 kpc gives b=−4.6086∘6 mag and hence an X-ray-to-optical flux ratio b=−4.6086∘7, consistent with LMXBs hosting low-mass companions. Two caveats are stated plainly: photometry in this crowded field is unreliable (GROND photometry could not be performed at all), and the 2.2″ aperture used for the upper limit includes contaminating flux from neighbors, making it conservative — though this does not affect the companion-type inference.
Spectral evolution and broadband modeling
Joint Swift/XRT + NuSTAR spectra were fitted with constant*tbabs*(cutoffpl + relxill + diskbb), yielding b=−4.6086∘8. The key parameters are:
| Parameter |
Value |
| Photon index b=−4.6086∘9 |
≳400 |
| Cutoff energy ≳401 |
≳402 keV |
| Disk inner temperature ≳403 |
≳404 keV |
| Inclination |
≳405 deg |
| Ionization ≳406 |
≳407 |
| Iron abundance ≳408 |
≳409 |
| ∼4×10−100 |
∼4×10−101 cm∼4×10−102 |
Several relxill parameters could not be constrained and were frozen (∼4×10−103, ∼4×10−104, emissivity indices fixed at 3, reflection fraction ∼4×10−105), so the reflection constraints are weak by construction. A simpler absorbed cutoff power-law leaves soft residuals and a 20–40 keV hump, motivating the disk plus reflection components. The measured ∼4×10−106 exceeds the Galactic H i column of ∼4×10−107 cm∼4×10−108 from HI4PI, suggesting intrinsic absorption.
A notable methodological limitation concerns SVOM/MXT: because its bandpass cannot disentangle thermal and non-thermal components, fixing disk parameters to Swift/XRT values significantly softens the inferred photon index. The authors caution that MXT-derived ∼4×10−109 values should be interpreted cautiously due to this degeneracy.
Outburst morphology and state behavior
The long-term light curve follows a fast-rise exponential-decay (FRED) profile with rise timescale −20 days, decay timescale −21 days, and peak time ~2.4 days after discovery. The FRED fit has reduced −22, attributed to cross-calibration offsets between instruments and the simplicity of the model; EP/WXT points are systematically offset during decay due to its narrow bandpass. Throughout the outburst, −23 stays within ~1.6–2.2 with no softening, and the hardness–intensity diagram shows no transition away from the hard state — behavior characteristic of "failed" or hard-state-only outbursts. The diskbb-derived inner radius evolves consistently with a truncated disk at −24, which is why that value was adopted in the reflection fit.
Timing analysis
No Type-I bursts, coherent pulsations, QPOs, or red noise are detected in any dataset. The NuSTAR power spectrum yields a 99% upper limit of 10.3% integrated fractional rms over 0.1–64 Hz, and a PRESTO acceleration search constrains the pulsed fraction to below ~6–7% at 99% confidence. The absence of a QPO distinguishes this source from EP J182730.0−095633, which showed a stable ~40 mHz QPO with 20–30% rms amplitude, though featureless PSDs are documented in other VFXBs.
Nature of the compact object
The evidence assembled is circumstantial but mutually consistent: FRED morphology, hard-state persistence, truncated disk with weak reflection, no pulsations or bursts, high X-ray-to-optical flux ratio, and similarity to VFXBs such as IGR J17285−2922, IGR J17062−6143, and MAXI J1848−015. The strongest single discriminator invoked is the cutoff energy: −25 keV is argued to be more typical of BH systems, since NS binaries benefit from additional seed photons (disk, surface, boundary layer) that cool the corona to lower electron temperatures. However, the paper concedes that a NS in the propeller regime cannot be excluded on current data, and no dynamical mass measurement exists. The BH classification is therefore a candidate status, not a confirmed identification.
Limitations and open questions
Three assumptions materially affect the results. First, the luminosity classification as a faint/VFXB depends entirely on the assumed 8 kpc distance; no independent distance estimate exists. Second, the reflection modeling relies on frozen parameters chosen partly from the diskbb evolution, so the reported inclination and ionization carry systematic uncertainty beyond the statistical errors quoted. Third, the optical non-detection may reflect extinction, crowding, and intrinsically weak reprocessing rather than a genuinely faint companion. The paper leaves open whether deep optical spectroscopy can secure a counterpart and orbital period, and whether quiescent observations can establish the compact object mass.
Conclusion
This Letter presents a well-monitored faint X-ray transient whose temporal, spectral, and multiwavelength properties collectively favor a black hole LMXB in a hard-state-only outburst, while stopping short of a firm compact-object identification. Its principal contribution is demonstrative: EP detects and enables broadband characterization of transients at −26 erg s−27 near the Galactic plane that previous monitors would have missed, opening a population whose demographics remain poorly constrained.
