Discovery of the redback millisecond pulsar PSR J1728-4608 with ASKAP

We present the discovery of PSR J1728-4608, a new redback spider pulsar identified in images from the Australian SKA Pathfinder telescope. PSR J1728-4608 is a millisecond pulsar with a spin period of 2.86 ms, in a 5.05 hr orbit with a companion star. The pulsar exhibits a radio spectrum of the form $S_ν\propto ν^α$, with a measured spectral index of $α= -1.8(3)$. It is eclipsed for 42% of its orbit at 888 MHz, and multi–frequency image–domain observations show that the egress duration scales with frequency as a power law with index $n = -1.74$, where longer duration eclipses are seen at lower frequencies. An optical counterpart is detected in archival Gaia data within $0.5’’$ of the radio position. It has a mean G-band magnitude of 18.8 mag and its light curve displays characteristics consistent with a combination of ellipsoidal modulation and irradiation effects. We also report the nearest Fermi $γ$-ray source, located 2$’$ away from our source, as a possible association. A radio timing study constrains the intrinsic and orbital properties of the system, revealing orbital period variations that we attribute to changes in the gravitational quadrupole moment of the companion star. At the eclipse boundary, we measure a maximum dispersion measure excess of $2.0 \pm 1.2 \ \mathrm{pc\ cm^{-3}}$, corresponding to an electron column density of $5.9 \pm 3.6 \times10^{18} \ \mathrm{cm^{-2}}$. Modelling of the eclipse mechanism suggests that synchrotron absorption is the dominant cause of the eclipses observed at radio wavelengths. The discovery and characterisation of systems like \psr\ provide valuable insights into pulsar recycling, binary evolution, the nature of companion-driven eclipses, and the interplay between compact objects and their plasma environments.

💡 Research Summary

The authors report the discovery of a new redback millisecond pulsar, PSR J1728‑4608, using archival imaging data from the Australian Square Kilometre Array Pathfinder (ASKAP). The source first stood out in the VAST and EMU surveys as a periodic, eclipse‑like dip with a ∼5‑hour recurrence in a 15‑minute cadence light curve at 943.5 MHz. A second independent ASKAP observation reproduced the same pattern, confirming the variability as intrinsic to the source. Follow‑up observations with ATCA (2.1 GHz), Parkes/UWL (0.7–4 GHz), uGMRT (0.55–0.75 GHz), and MeerKAT (2.6–3.5 GHz) established a spin period of 2.86 ms, an orbital period of 5.05 h, and a steep radio spectrum $S_\nu\propto\nu^{-1.8\pm0.3}$. At 888 MHz the pulsar is eclipsed for 42 % of its orbit; the egress duration follows a power‑law scaling $\propto\nu^{-1.74}$, indicating longer eclipses at lower frequencies.

Timing analysis over several months yielded a dispersion measure (DM) of 65.6 pc cm⁻³ and a DM excess of $2.0\pm1.2$ pc cm⁻³ at the eclipse boundaries, corresponding to an electron column density of $5.9\pm3.6\times10^{18}$ cm⁻². The timing solution also revealed secular variations in the orbital period, which the authors attribute to changes in the gravitational quadrupole moment of the companion (the Applegate mechanism).

A Gaia DR3 source (Gaia DR3 5951944861092533248) lies within 0.5″ of the radio position, has a mean G‑band magnitude of 18.8 mag, and a parallax of $0.63\pm0.23$ mas, implying a distance of roughly 1.6 kpc. Its optical light curve, obtained from LCOGT observations in g, r, i, and z bands, shows a combination of ellipsoidal modulation and heating by the pulsar wind, consistent with a semi‑degenerate companion of 0.3–0.7 M⊙ typical for redbacks.

The nearest Fermi LAT source, 4FGL J1728.0‑4606, lies 2′ away, outside the 95 % error ellipse but still considered a possible γ‑ray counterpart given the high background in the Galactic plane. No definitive γ‑ray pulsations have yet been detected.

The authors model the eclipse mechanism using the measured DM excess and the frequency dependence of the egress. They argue that synchrotron absorption by a magnetised plasma surrounding the companion dominates, as cyclotron absorption would require unrealistically high magnetic fields or electron temperatures. The synchrotron model naturally explains the observed $\nu^{-1.74}$ scaling and the modest DM increase at eclipse ingress/egress.

Overall, the paper demonstrates that image‑domain searches with wide‑field, high‑cadence instruments like ASKAP can efficiently uncover short‑period binary MSPs that are difficult to find with traditional beam‑forming surveys. The detailed multi‑wavelength follow‑up provides a comprehensive picture of PSR J1728‑4608: a redback system with a steep radio spectrum, strong, frequency‑dependent eclipses driven by synchrotron absorption, a companion whose quadrupole moment variations modulate the orbital period, and a plausible—but not yet confirmed—γ‑ray counterpart. These results enrich the growing sample of redback pulsars, offering valuable constraints on binary evolution, pulsar recycling, and plasma physics in extreme binary environments, and set the stage for future discoveries with the full SKA.