A binary pulsar in a 53-minute orbit

The study probes whether millisecond pulsars with moderate-mass companions can exist in extremely compact orbits, bridging the evolutionary gap between redback and black widow systems. Spider pulsars evolve via mass transfer and ablation, potentially transitioning from redbacks (0.1–0.4 M⊙ companions, P_orb < 1 day) to black widows (M_c ≪ 0.1 M⊙, ultracompact orbits). Prior to this work, no system combined a very short orbital period with a moderate-mass companion. The authors report the discovery and characterization of PSR J1953+1844 (M71E), a binary millisecond pulsar with a 53.3-minute orbit. They detail its timing solution, environment relative to the globular cluster M71, lack of eclipses, and constraints on companion type and orbital inclination, aiming to illuminate formation pathways and validate evolutionary models of spider pulsars.

The paper situates M71E within the context of spider pulsar evolution, citing foundational work on companion ablation and evolution of millisecond pulsars (e.g., Van den Heuvel & Van Paradijs; Roberts). It references observational demographics of redbacks and black widows, and recent extremes such as a 62-minute orbital-period black widow in a hierarchical triple and a record-mass neutron star in a spider system. Catalogues of globular cluster pulsars and ATNF data are used to place M71E among short-period binaries. Theoretical studies on formation channels include binary evolution toward helium/CO white dwarfs, stability of mass transfer, magnetic braking with ablated winds, and potential dynamical exchange interactions in clusters. Observational limits on detached white dwarf masses (≥0.14–0.16 M⊙) and brown dwarf/low-mass star mass–radius relations inform constraints on the nature of M71E’s companion.

Discovery and observational campaign: M71E (PSR J1953+1844) was first found by the FAST Galactic Plane Pulsar Snapshot (GPPS) survey and independently detected in archival FAST globular cluster data (2019-12-12). Follow-up timing observations were conducted from MJD 59293 to 59781 across 21 epochs (∼2 h per epoch), with the FAST 19-beam L-band receiver (1,050–1,450 MHz). Data specifics included PSRFITS search-mode recordings with 4,096 channels at 49.152 µs sampling or 2,048 channels at 98.304 µs sampling. Integrated profiles (5 min) with 128 phase bins were generated. Data reduction and timing: PRESTO (prepfold, get_TOAs.py) was used to produce TOAs (typically 32–64 per observation). Timing analysis employed TEMPO/TEMPO2 with the ELL1 binary model (appropriate for low-eccentricity systems). Initial fits used JUMPs between observations to derive orbital parameters; DRACULA was used to remove JUMPs and establish rotation count. Iterative folding with refined solutions produced the final timing solution. The timing yielded a mass function of 2.3×10⁻¹⁰ M⊙ and residual RMS of 35.138 µs over 635 TOAs spanning 488 days. Measured parameters included: RA 19:53:37.9464(1), Dec +18:44:54.310(2), spin frequency 225.01840471114(7) Hz, spin-down −1.11(1)×10⁻⁵ s⁻¹, orbital period Pb = 0.0370398638(5) d (53.3 min), TASC 58829.26006(1) MJD, projected semimajor axis x = 0.006668(2) lt-s, eccentricity components κ = 0.0003(6), η = 0.0005(6). Set quantities: reference epoch 59474.630459 MJD, DM = 113.1 pc cm⁻³, Solar System ephemeris DE440, model ELL1. Polarimetry and flux calibration: A 3,600 s observation on 2021-12-04 provided polarization profiles across 1,050–1,450 MHz, calibrated using a 10 K noise diode and FAST aperture efficiency of 0.63 via LAPUDA. Rotation measure RM = −475 ± 2 rad m⁻². Peak and mean flux densities were 0.632 ± 0.002 mJy and 0.092 ± 0.002 mJy, respectively. Multiwavelength counterpart search: Archival SDSS and 2MASS imaging showed no optical/IR counterpart. Chandra data (ObsID 5434) revealed 8 photons within 0.5″ of the timing position (offset 7.34 mas), energies 0.759–1.327 keV, phase clustering consistent with the pulsar with ∼3% chance probability. Estimated X-ray flux 3 ± 1 × 10⁻⁶ erg cm⁻² s⁻¹ and luminosity 6 ± 2 × 10²⁹ erg s⁻¹ at 4 kpc, suggesting rotation-powered emission. Companion constraints and modeling: Using the timing-derived mass function, the authors mapped inclination i versus companion mass M_c for assumed neutron star masses of 1.0, 1.4, 2.0 M⊙. White dwarf companions (≥0.14–0.16 M⊙) would require extremely face-on inclinations (i < 3.6°), a priori probability <0.3%, and lack an obvious formation channel at such tight orbits, disfavouring a WD. For stripped low-mass/brown dwarf companions, Roche-lobe constraints combined with mass–radius relations (solar metallicity LMX and zero metallicity LMZ models) yield allowed M_c ranges. Table 2 summarizes modeled inclination angles and companion masses across neutron star masses (1.0–2.0 M⊙), metallicities (Z = 0.02, 0), and evolutionary ages (0.6–10 Gyr). Evolutionary scenarios: The system is compared against binary evolution tracks (donor metallicities Z = 0.002 and 0.02; initial M_c = 0.4 M⊙, P_b = 0.70 d, evaporation efficiency f = 0.02), and an alternative track with an evolved main-sequence donor leading to a He-rich companion. Effects of evaporation by pulsar emission were not included in some models and may widen the orbit.

• Discovery and timing of the binary millisecond pulsar PSR J1953+1844 (M71E) with an orbital period of 53.3 minutes, the shortest known for a radio-detected binary pulsar at the time.
• Measured parameters: spin frequency 225.01840471114(7) Hz; spin-down −1.11(1)×10⁻⁵ s⁻¹; orbital period Pb = 0.0370398638(5) d; projected semimajor axis x = 0.006668(2) lt-s; DM = 113.1 pc cm⁻³; mass function f = 2.3×10⁻¹⁰ M⊙; timing RMS 35.138 µs over 635 TOAs across 488 days.
• Sky position J2000 19:53:37.95 +18:44:54.3, about 2.5 arcmin from the center of globular cluster M71; DM close to other M71 pulsars, but membership remains debated.
• No radio eclipses observed across the orbit, indicating a non-edge-on geometry.
• Chandra counterpart consistent with a faint X-ray source at the pulsar position: 8 photons, 0.76–1.33 keV; inferred flux 3 ± 1 × 10⁻⁶ erg cm⁻² s⁻¹ and luminosity 6 ± 2 × 10²⁹ erg s⁻¹ at 4 kpc; suggests rotation-powered emission and little/no accretion.
• Polarization measurement yields RM = −475 ± 2 rad m⁻²; mean flux density ∼0.092 mJy at 1.05–1.45 GHz.
• Companion type constraints: Detached white dwarf companions (≥0.14–0.16 M⊙) would require extremely face-on inclinations (i < 3.6°, probability <0.3%) and lack a straightforward formation path at this period, making a WD companion unlikely.
• Roche-lobe and mass–radius constraints for a stripped low-mass/brown dwarf donor imply M_c ≈ 0.047–0.097 M⊙ (typical ∼0.07 M⊙) and i ≈ 3.8°–12.1° (typical ∼8°) for neutron star masses 1.0–2.0 M⊙, consistent with the absence of eclipses.
• Placement in the companion mass–orbital period plane shows M71E as a bridging object between redbacks and black widows, extending to a tighter orbital regime but with moderate companion mass.

The detection of M71E fills a predicted but previously unobserved region of spider pulsar parameter space: very short orbital periods with moderate-mass companions. Its 53.3-minute orbit and inferred companion mass around 0.07 M⊙ support evolutionary paths where redbacks evolve into black widows through mass transfer, magnetic braking, gravitational radiation, and irradiation-driven evaporation. The lack of radio eclipses and the requirement of a low inclination are consistent with its compact, likely non-eclipsing geometry. Detached white dwarf companions are disfavoured due to the improbably small inclination angles required and the absence of clear formation channels that produce a detached NS–WD binary at such tight orbital periods. Stripped low-mass or brown dwarf donors confined within their Roche lobe more naturally explain the constraints and X-ray faintness indicative of a non-accreting state. Comparisons with binary evolution models indicate that standard tracks with donor metallicities Z = 0.002–0.02 can approach, but sometimes struggle to reproduce, the observed shortest orbital period, suggesting that donor composition (e.g., He-rich) and evaporation may be important. An evolutionary track with an evolved main-sequence donor and standard magnetic braking can reproduce both the mass and orbital period of M71E, though the onset and impact of pulsar-driven evaporation remain to be fully modeled. If associated with M71, dynamical exchange cannot be excluded, but current models are insufficient to assess this channel quantitatively. Overall, M71E offers stringent constraints on spider pulsar formation and the transition from redbacks to black widows, motivating refined models that include irradiation-driven mass loss and donor composition effects.

This work reports the FAST discovery and detailed timing of PSR J1953+1844 (M71E), a binary millisecond pulsar with a record-short 53.3-minute orbit and a likely stripped low-mass/brown dwarf companion of ∼0.07 M⊙. White dwarf companions are disfavoured by inclination constraints and formation considerations. The system’s properties bridge redbacks and black widows and provide empirical constraints for binary evolution models, highlighting the roles of magnetic braking, gravitational radiation, and pulsar-driven evaporation. The X-ray faintness indicates a rotation-powered, non-accreting state. Future work should include multi-band observations to detect and characterize the companion (e.g., HST/JWST spectroscopy for elemental abundances), continued radio timing to measure orbital derivatives, parallax, and proper motion, and broader radio/X-ray campaigns to probe pulsar–companion interactions. Improved evolutionary modeling incorporating evaporation and potential He-rich donors will help reconcile models with the observed ultrashort period.

• Cluster membership remains uncertain: the pulsar lies ∼2.5′ from the M71 center and has a DM similar but not identical to known M71 pulsars.
• Inclination and true companion mass are not directly measured; absence of eclipses only constrains geometry indirectly.
• No optical/IR counterpart is detected; the X-ray counterpart is based on 8 photons with a ∼3% chance clustering probability, limiting spectral and timing diagnostics.
• Evolutionary tracks used for comparison do not fully include evaporation effects in all cases and struggle to reproduce the exact 53.3-minute period for some metallicities, indicating model incompleteness.
• Dynamical formation via exchange interactions is not evaluated with self-consistent calculations, leaving that channel unconstrained.