Swift J174610.4-290018: Galactic Center X-ray Transient

• Swift J174610.4-290018 is a transient X-ray source near Sgr A* with episodic outbursts detected by Swift/XRT and NuSTAR, offering insights into low-luminosity accretion phenomena.
• Spectral analyses reveal hard X-ray continua with strong iron-line features modeled using absorbed power-law and thermal plasma fits that illustrate complex emission processes.
• Competing interpretations as a VFXT neutron-star LMXB versus a symbiotic binary/recurrent nova underline the need for multi-wavelength follow-up to resolve its nature.

Searching arXiv for the specified source and cited papers to ground the article in current records. Tool call: arXiv search for Swift J174610.4-290018 and the provided arXiv IDs. Swift J174610.4–290018, often abbreviated as J1746–2900, is a transient X-ray source within $\sim 400''$ of Sgr A* and about 7 arcmin from Sgr A* that was first flagged on 2024 Feb 22 by Swift/XRT during the daily Galactic-Center monitoring and again on 2025 Apr 4 by NuSTAR. The source is also identified in archival catalogs as 4XMM J174610.7–290020 and 2CXO J174610.7–290019. Its astrophysical nature remains debated: one interpretation classifies it as a Very Faint X-ray Transient (VFXT) neutron-star low-mass X-ray binary, while another suggests a symbiotic binary/recurrent nova scenario. Both interpretations are motivated by the combination of low-luminosity outbursts, hard X-ray continua, and strong iron-line emission reported for the source (Stel et al., 2 Oct 2025, Hua et al., 30 Sep 2025).

1. Discovery, designation, and source context

Swift J174610.4–290018 entered the literature as a new Galactic-center transient after Swift/XRT detected it on 2024 Feb 22 during daily monitoring. In the 2025 outburst cycle, the source was again flagged, this time by NuSTAR on 2025 Apr 4. The same object had already appeared in the Chandra and XMM point-source catalogs between 2000 and 2010 under the designations 2CXO J174610.7–290019 and 4XMM J174610.7–290020 (Stel et al., 2 Oct 2025).

The designation history is important because the source was not entirely new in an archival sense. Earlier Chandra/ACIS observations intermittently detected a faint source at RA = 17h 46m 10.67s, Dec = –29° 00′ 19.44″. Those detections, together with later outbursts, established a long-term record spanning quiescent intervals, elevated states, and at least two clearly recognized modern outbursts, with one study additionally identifying a plausible outburst in 2005 (Hua et al., 30 Sep 2025).

In the VFXT framework, the source belongs to a class of X-ray binary systems that exhibit occasional outbursts with peak X-ray luminosities $L_X < 10^{36}\ {\rm erg\ s^{-1}}$, much lower than typical X-ray transients. In the competing recurrent-nova framework, the same observational record is interpreted in terms of thermonuclear runaway on a massive white dwarf embedded in a dense circumbinary environment (Stel et al., 2 Oct 2025, Hua et al., 30 Sep 2025).

2. Outburst phenomenology and luminosity evolution

The Swift-XRT count-rate light curve was constructed assuming a dust-scattered, absorbed power-law with $N_{\rm H}=2\times10^{23}\ {\rm cm^{-2}}$ and $\Gamma=2$, using daily XRT snapshots of $\sim 1$ ks each. In that reconstruction, the 2024 outburst had onset at MJD $\simeq 59900$, duration $\Delta t \simeq 50\ {\rm d}$, a slow rise to a peak absorbed 2–10 keV flux $$F_{2-10}\simeq 9\times10^{-12}\ {\rm erg\ cm^{-2}\ s^{-1}}$$, and a return to quiescence by MJD $\simeq 59950$. The 2025 outburst had onset at MJD $\simeq 60037$, $L_X < 10^{36}\ {\rm erg\ s^{-1}}$0, peak flux $L_X < 10^{36}\ {\rm erg\ s^{-1}}$1, and then a rapid fade (Stel et al., 2 Oct 2025).

Assuming a distance $L_X < 10^{36}\ {\rm erg\ s^{-1}}$2kpc, the corresponding 2–10 keV luminosities are

$L_X < 10^{36}\ {\rm erg\ s^{-1}}$3

and

$L_X < 10^{36}\ {\rm erg\ s^{-1}}$4

A second analysis, based on Swift, NuSTAR, Chandra, and archival Chandra data, reported a broader 2024 evolution. In that treatment, Swift/XRT first detected the rise on 2024 Feb 22 (MJD 59824), the pre-April 2024 flux was $L_X < 10^{36}\ {\rm erg\ s^{-1}}$5–$L_X < 10^{36}\ {\rm erg\ s^{-1}}$6, and the 2–8 keV flux decayed by $L_X < 10^{36}\ {\rm erg\ s^{-1}}$7 order of magnitude over $L_X < 10^{36}\ {\rm erg\ s^{-1}}$8 days, with hints of flattening or rebrightening around MJD 60460. The same study identified a 2005 outburst extending from 2005 Feb 8 to 2005 Jul 22, with a peak at $L_X < 10^{36}\ {\rm erg\ s^{-1}}$9 on 2005 Feb 27 and an overall duration of $N_{\rm H}=2\times10^{23}\ {\rm cm^{-2}}$0 days (Hua et al., 30 Sep 2025).

Interval	State	Reported properties
2001–2004	Quiescence	2–8 keV flux $N_{\rm H}=2\times10^{23}\ {\rm cm^{-2}}$1–$N_{\rm H}=2\times10^{23}\ {\rm cm^{-2}}$2; $N_{\rm H}=2\times10^{23}\ {\rm cm^{-2}}$3
2005	Outburst	Chandra plateau from 2005 Feb 8 to 2005 Jul 22; peak $N_{\rm H}=2\times10^{23}\ {\rm cm^{-2}}$4
2024	Outburst	Either $N_{\rm H}=2\times10^{23}\ {\rm cm^{-2}}$5d with $N_{\rm H}=2\times10^{23}\ {\rm cm^{-2}}$6, or $N_{\rm H}=2\times10^{23}\ {\rm cm^{-2}}$7d with decay over $N_{\rm H}=2\times10^{23}\ {\rm cm^{-2}}$8d
2025	Outburst	$N_{\rm H}=2\times10^{23}\ {\rm cm^{-2}}$9d; $\Gamma=2$0

Rapid day-to-day flux fluctuations around the 2024 peak were reported to be intrinsic, not due to dust scattering. No significant periodicity was found via Gregory–Loredo or Lomb–Scargle, in contrast to the 1537 s candidate from XRISM (Hua et al., 30 Sep 2025).

3. Spectral models and iron-line diagnostics

The X-ray spectroscopy of Swift J174610.4–290018 is dominated by hard continua and iron-line structure. Spectral fits to deep XMM-Newton/EPIC-pn data and to NuSTAR FPMA/B spectra converged on two phenomenological descriptions. The first is an absorbed power-law plus iron line(s), fgcdust*tbabs*(powerlaw+gauss+gauss), with $\Gamma=2$1–$\Gamma=2$2, photon index $\Gamma=2$3–$\Gamma=2$4, Fe XXV fixed at $\Gamma=2$5keV with flux $\Gamma=2$6, and Fe XXVI at $\Gamma=2$7keV with flux $\Gamma=2$8. The second is an optically thin thermal plasma, fgcdust*tbabs*apec, with $\Gamma=2$9–$\sim 1$0 and $\sim 1$1–$\sim 1$2keV (Stel et al., 2 Oct 2025).

In each case, a narrow Gaussian line at $\sim 1$3–$\sim 1$4keV with width $\sim 1$5eV is required. This feature can be viewed either as two unresolved ionized-iron transitions, Fe XXV and Fe XXVI, or as a single broad reflection feature. Its equivalent width is $\sim 1$6eV, reminiscent of standard disk-reflection models in which the line profile $\sim 1$7 depends on ionization parameter $\sim 1$8 and reflection fraction $\sim 1$9–0.5. The same analysis noted that the present spectra do not tightly constrain $\simeq 59900$0 or $\simeq 59900$1, but that the centroid and narrow width indicate an origin at $\simeq 59900$2 if Doppler-broadened, or in a photoionized corona (Stel et al., 2 Oct 2025).

A separate analysis fitted all spectra in XSPEC v12.15 using either an absorbed optically thin thermal plasma model

$\simeq 59900$3

or a bremsstrahlung continuum plus four Gaussians

$\simeq 59900$4

with line centroids fixed at $\simeq 59900$5keV, identified as Cr XXIII K$\simeq 59900$6, Fe I K$\simeq 59900$7, Fe XXV He$\simeq 59900$8, and Fe XXVI Ly$\simeq 59900$9 (Hua et al., 30 Sep 2025).

That study reported, for the quiescent Chandra state, $\Delta t \simeq 50\ {\rm d}$0, $\Delta t \simeq 50\ {\rm d}$1keV, $\Delta t \simeq 50\ {\rm d}$2eV, $\Delta t \simeq 50\ {\rm d}$3, $\Delta t \simeq 50\ {\rm d}$4, and $\Delta t \simeq 50\ {\rm d}$5. For the 2005 and 2024 outbursts observed by Chandra, the best-fit temperatures remained near $\Delta t \simeq 50\ {\rm d}$6–$\Delta t \simeq 50\ {\rm d}$7keV, while the iron-line ratios evolved and the 6.7 keV equivalent width remained large. Swift data from Feb–Mar 2024 yielded $\Delta t \simeq 50\ {\rm d}$8keV and an upper limit $\Delta t \simeq 50\ {\rm d}$9eV at $$F_{2-10}\simeq 9\times10^{-12}\ {\rm erg\ cm^{-2}\ s^{-1}}$$0, whereas NuSTAR data from Apr 2024 yielded $$F_{2-10}\simeq 9\times10^{-12}\ {\rm erg\ cm^{-2}\ s^{-1}}$$1keV and $$F_{2-10}\simeq 9\times10^{-12}\ {\rm erg\ cm^{-2}\ s^{-1}}$$2eV (Hua et al., 30 Sep 2025).

4. Archival detections, quiescence, and the 2004 flare

Archival Chandra and XMM observations establish that Swift J174610.4–290018 is not solely an eruptive 2024–2025 source. Between 2000 and 2010, the cataloged source displayed 2–10 keV luminosities from $$F_{2-10}\simeq 9\times10^{-12}\ {\rm erg\ cm^{-2}\ s^{-1}}$$3 to $$F_{2-10}\simeq 9\times10^{-12}\ {\rm erg\ cm^{-2}\ s^{-1}}$$4, with upper limits down to $$F_{2-10}\simeq 9\times10^{-12}\ {\rm erg\ cm^{-2}\ s^{-1}}$$5 (Stel et al., 2 Oct 2025).

One analysis divided the archival record into 2001–2004 quiescence, a 2005 outburst, and a subsequent $$F_{2-10}\simeq 9\times10^{-12}\ {\rm erg\ cm^{-2}\ s^{-1}}$$6 yr quiescent interval. In the 113 Chandra pointings from 2005 Jul to 2024 Feb, the source remained below a 2–8 keV flux of $$F_{2-10}\simeq 9\times10^{-12}\ {\rm erg\ cm^{-2}\ s^{-1}}$$7, corresponding to $$F_{2-10}\simeq 9\times10^{-12}\ {\rm erg\ cm^{-2}\ s^{-1}}$$8, with no long-term trend. This led to the suggestion of a recurrence timescale $$F_{2-10}\simeq 9\times10^{-12}\ {\rm erg\ cm^{-2}\ s^{-1}}$$9yr (Hua et al., 30 Sep 2025).

A particularly consequential archival event was identified in the 2004 XMM observation ObsID 0202670701. In that dataset, a flare with fast rise of $\simeq 59950$0s, slow decay of $\simeq 59950$1s, and total duration of 523 s was uncovered. Time-resolved blackbody fits using fgcdust*tbabs*bbodyrad yielded a peak $\simeq 59950$2keV rapidly softening to $\simeq 59950$3keV, a bolometric fluence $\simeq 59950$4, and an implied radius

$\simeq 59950$5

The reported softening and light-curve shape were taken to identify the event as an intermediate-duration type I X-ray burst, implying thermonuclear burning on a neutron-star surface (Stel et al., 2 Oct 2025).

This burst candidate is central to the classification problem. If the identification is correct, it strongly supports a neutron-star accretor. A plausible implication is that the archival light curve records both low-rate accretion episodes and a thermonuclear surface event, rather than only shock-powered emission.

5. Competing physical interpretations

One interpretation treats Swift J174610.4–290018 as a typical VFXT NS-LMXB seen at low accretion rates. In that picture, the low peak luminosity, the potential type I burst, the power-law slope $\simeq 59950$6, the high-temperature plasma with $\simeq 59950$7keV, and the ionized Fe lines all point to a neutron-star low-mass X-ray binary. Yoshimoto et al. (2025) further proposed an accretion-disk corona geometry in which the innermost emission is blocked by a thick, edge-on disk and the observed continuum and iron lines arise from scattering in an extended, hot ($\simeq 59950$8keV), optically thin corona of scale height $\simeq 59950$9. In that geometry,

$\simeq 60037$0

and resonance-scattered Fe XXV/XXVI emission naturally emerges from the photoionized plasma. This scenario was argued to explain both the faint outburst peaks and the stable, narrow iron-line complex (Stel et al., 2 Oct 2025).

A second interpretation explicitly examined the low-mass X-ray binary/accretion disk corona scenario and disfavored it. The principal objection was that the quiescent spectrum, with observed $\simeq 60037$1, would imply intrinsic $\simeq 60037$2, regarded as far too low to sustain an ADC, and would not produce the observed Fe lines in quiescence. The same study also noted that high-inclination LMXBs show X-ray eclipses, dips, and periodicities, none of which were seen here (Hua et al., 30 Sep 2025).

The alternative proposed in that work is a symbiotic binary/recurrent nova scenario. In this model, thermonuclear runaway on a massive white dwarf accreting from a red-giant wind produces ejecta with $\simeq 60037$3–$\simeq 60037$4 and $\simeq 60037$5 that shock the dense circumbinary environment, heating plasma to $\simeq 60037$6–$\simeq 60037$7keV and producing strong Fe lines. The model was presented as a natural explanation for the quiescent spectrum, the two discrete outbursts separated by $\simeq 60037$8yr, and the Fe K fluorescence and highly ionized Fe line evolution (Hua et al., 30 Sep 2025).

The recurrent-nova interpretation was supported by comparison with RS Oph, described there as the prototypical symbiotic recurrent nova with $\simeq 60037$9 and recurrence $L_X < 10^{36}\ {\rm erg\ s^{-1}}$00–$L_X < 10^{36}\ {\rm erg\ s^{-1}}$01yr. Under the standard recurrent-nova relation

$L_X < 10^{36}\ {\rm erg\ s^{-1}}$02

taking $L_X < 10^{36}\ {\rm erg\ s^{-1}}$03 and $L_X < 10^{36}\ {\rm erg\ s^{-1}}$04yr gives $L_X < 10^{36}\ {\rm erg\ s^{-1}}$05 and requires a massive white dwarf with $L_X < 10^{36}\ {\rm erg\ s^{-1}}$06 (Hua et al., 30 Sep 2025).

The present literature therefore does not present a settled classification. One line of evidence emphasizes the putative type I burst and the ADC-compatible iron-line phenomenology; the other emphasizes the quiescent plasma spectrum, the absence of LMXB-like periodic signatures, and the similarity to recurrent novae.

6. Astrophysical significance and observational outlook

Swift J174610.4–290018 is significant because its observational properties place it at the intersection of two different source classes in a heavily obscured Galactic-center environment. As a VFXT candidate, it exemplifies low-rate accretion with peak 2–10 keV luminosities in the $L_X < 10^{36}\ {\rm erg\ s^{-1}}$07–$L_X < 10^{36}\ {\rm erg\ s^{-1}}$08 range, hard spectra, and iron-line diagnostics. As a recurrent-nova candidate, it would represent the first nova detected in the Galactic center, implying the survival of wide symbiotic binaries in the nuclear star cluster and pointing to a hidden population of massive white dwarfs near Sgr A* (Stel et al., 2 Oct 2025, Hua et al., 30 Sep 2025).

The recommended follow-up program in the recurrent-nova study is explicitly multi-wavelength: infrared imaging and spectroscopy to identify a red-giant donor through dust-penetrating bands; high-resolution X-ray spectroscopy to map non-equilibrium ionization and element abundances; radio observations to image ejecta and probe the circumbinary medium; and continued Swift/XRT, Chandra, and NuSTAR monitoring to measure recurrence, shock evolution, and future outbursts (Hua et al., 30 Sep 2025).

For the NS-LMXB interpretation, the decisive observables are different. Additional detections of type I bursts, stronger constraints on the iron-line profile, and tighter measurements of reflection or coronal parameters would directly test the accretion-disk corona picture. This suggests that the source is likely to remain a diagnostic object for distinguishing low-luminosity neutron-star accretion from shock-powered white-dwarf eruptions under Galactic-center conditions.