Evidence for optical pulsations from a redback millisecond pulsar

TL;DR

The paper reports a potential detection of optical pulsations at the spin period from the redback MSP PSR J2339-0533 using the SiFAP2 photometer. By combining SiFAP2 data (W1 and W2) with near-simultaneous Parkes radio timing, the authors fold the optical time series at the pulsar spin frequency and account for correlated SiFAP2 noise, finding a candidate signal in the longer W2 exposure with a significance of $2.9$–$3.5\,\sigma$ and a pulsed magnitude of $m_V^{\rm pulse}=26.4\pm0.6$ mag, corresponding to a pulsed flux of $F_V^{\rm pulse}=(3.1\pm1.5)\times10^{-16}$ erg s$^{-1}$ cm$^{-2}$. The derived pulsed efficiency is $\eta\sim3\times10^{-6}$ of the spin-down power, comparable to young isolated pulsars like the Crab but 50–100× lower than disk-accreting MSPs, implying that an accretion disk can boost optical emission. If confirmed, these pulsations would indicate that optical emission can arise close to the neutron star even in the absence of a disk, while highlighting disk interactions as a mechanism to enhance emission; future coordinated radio and larger-telescope optical observations are needed to confirm the signal and elucidate its origin.

Abstract

Recent detections of optical pulsations from both a transitional and an accreting millisecond pulsar have revealed unexpectedly bright signals, suggesting that the presence of an accretion disk enhances the efficiency of optical emission, possibly via synchrotron radiation from accelerated particles. In this work, we present optical observations of the redback millisecond pulsar PSR J2339-0533, obtained with the SiFAP2 photometer mounted on the Telescopio Nazionale Galileo. Data accumulated during the campaign with the longest exposure time (12 hr) suggest that its $\sim$18 mag optical counterpart exhibits pulsations at the neutron star's spin frequency. This candidate signal was identified by folding the optical time series using the pulsar ephemeris derived from nearly simultaneous observations with the 64-m Murriyang (Parkes) radio telescope. The detection significance of the candidate optical signal identified in those data lies between 2.9 and 3.5 $σ$, depending on the statistical test employed. The pulsed signal has a duty cycle of $\approx 1/32$, and the de-reddened pulsed magnitude in the V band is $(26.0 \pm 0.6)$ mag. At a distance of 1.7 kpc, this corresponds to a conversion efficiency of $\sim 3 \times 10^{-6}$ of the pulsar's spin-down power into pulsed optical luminosity, comparable to values observed in young, isolated pulsars like the Crab, but 50-100 times lower than in disk-accreting millisecond pulsars. If confirmed, these findings suggest that optical pulsations arise independently of an accretion disk and support the notion that such disks boost the optical emission efficiency.

Figures (6)

• Figure 1: SiFAP2 net light curves of J2339 during the different nights (labelled as N$i$) of the two observing campaigns W$i$ (see text for details on the background estimation). The orbital phase was evaluated considering the timing solution listed in Table \ref{['table:psr']}.
• Figure 2: Phase-aligned pulse profiles observed in W2 by SiFAP2 (black points), Parkes (green line) and Fermi LAT (magenta line; taken from 3PC catalogue, S23, and shifted assuming that the lag with the radio profile is the same as reported there). The normalization of the radio and gamma-ray profiles is arbitrary. The reference epoch was set to align the radio peak at phase 0.5. The blue line is a fit of the SiFAP2 pulse profile with four Gaussians added to a constant level.
• Figure 3: Ratio of the pulsed B ($\lambda_{\rm eff}=433$ nm, $W_{\rm eff}=90$ nm) luminosity to the spin down power for J2339 (orange dot) and the sample of detected optical pulsars. Data from Ambrosino2017, Papitto2019 and Ambrosino2021.
• Figure 4: Grey boxes show the observed density of $\chi_{\rm tr}^2$ values observed epoch folding SiFAP2 W2 data around 100,000 trial periods close but not overlapping with the pulsar period. Blue boxes show the distribution of $\chi^2$ values rescaled by $r=1.16$, and the black solid line the $\chi^2_{n-1}$ probability density function for $n=32$. The vertical dashed line marks the rescaled value of $\chi^2_{\rm W2}=64.4$.
• Figure 5: Grey boxes indicate the complementary cumulative distribution of the $\chi_{\rm tr}^2$ values obtained by epoch folding searching W2 data at 100,000 independent trial periods (see text for details). Blue boxes indicate the distribution obtained after rescaling the $\chi^2$ values by $r=1.16$. The expected $\chi^2$ tail distribution for $n=32$ is overplotted as a black line. The dashed horizontal line indicates the false alarm probability of the $\chi_{W2}^2=64.4$ value observed in SiFAP2 W2 data.