The variability behavior of NGC 925 ULX-3

TL;DR

This study monitors the ultraluminous X-ray source NGC 925 ULX-3 over two years, revealing a consistent 127-day periodicity, spectral characteristics, and implications for neutron star magnetic fields, without detecting pulsations.

Contribution

First detailed multi-year monitoring of NGC 925 ULX-3 revealing its periodicity, spectral properties, and constraints on neutron star magnetic fields, advancing understanding of ULX behavior.

Findings

Exhibits a 127-128 day periodicity with stable flux range.

Spectral analysis shows two thermal components with high temperature ratio.

No pulsations detected, upper limit on pulsed fraction is ~40%.

Abstract

We report the results of a 2019-2021 monitoring campaign with Swift and associated target-of-opportunity observations with XMM-Newton and NuSTAR, examining the spectral and timing behavior of the highly variable ultraluminous X-ray source (ULX) NGC 925 ULX-3. We find that the source exhibits a 127-128 day periodicity, with fluxes typically ranging from 1e-13 to 8e-13 ergs/s/cm2. We do not find strong evidence for a change in period over the time that NGC 925 ULX-3 has been observed, although the source may have been in a much lower flux state when first observed with Chandra in 2005. We do not detect pulsations, and we place an upper limit on the pulsed fraction of ~40% in the XMM-Newton band, consistent with some previous pulsation detections at low energies in other ULXs. The source exhibits a typical ULX spectrum that turns over in the NuSTAR band and can be fitted using two thermal components. These components have a high temperature ratio that may indicate the lack of extreme inner disk truncation by a magnetar-level magnetic field. We examine the implications for a number of different models for superorbital periods in ULXs, finding that a neutron star with a magnetic field of ~10^12 G may be plausible for this source. The future detection of pulsations from this source would allow for the further testing and constraining of such models.