Detections of millisecond pulsars with the Fermi Large Area Telescope

The Fermi observatory was launched on June 11, 2008. It hosts the \emph{Large Area Telescope} (LAT), sensitive to $\gamma$-ray photons from 20 MeV to over 300 GeV. When the LAT began its activity, nine young and energetic pulsars were known in $\gamma$ rays. At least several tens of pulsar detections by the LAT were predicted before launch. The LAT also allowed the study of millisecond pulsars (MSPs), never firmly detected in $\gamma$ rays before Fermi. This thesis first presents the pulsar timing campaign for the LAT, in collaboration with large radiotelescopes and X-ray telescopes, allowing for high sensitivity pulsed searches. Furthermore, it lead to quasi-homogeneous coverage of the galactic MSPs, so that the search for pulsations in LAT data for this population of stars was not affected by an \emph{a priori} bias. We present a search for pulsations from these objects in LAT data. For the first time, eight galactic MSPs have been detected as sources of pulsed $\gamma$-ray emission over 100 MeV. In addition, a couple of good candidates for future detection are seen. A similar search for globular cluster MSPs has not succeeded so far. Comparison of the phase-aligned $\gamma$-ray and radio light curves, as well as the spectral shapes, leads to the conclusion that their $\gamma$-ray emission is similar to that of normal pulsars, and is probably produced in the outer-magnetosphere. This discovery suggests that many unresolved $\gamma$-ray sources are unknown MSPs.

💡 Research Summary

The paper reports the first systematic detection of millisecond pulsars (MSPs) as pulsed γ‑ray sources using the Fermi Large Area Telescope (LAT). When LAT began operations in 2008, only nine young, energetic pulsars were known to emit γ‑rays, and MSPs—despite theoretical predictions that dozens should be detectable—had never been firmly identified in this band. To overcome the sensitivity and timing challenges, the authors organized a coordinated pulsar timing campaign involving major radio observatories (Arecibo, Green Bank, Parkes, etc.) and X‑ray facilities (Chandra, XMM‑Newton). Precise timing models were derived for essentially all known Galactic MSPs, providing nanosecond‑level ephemerides that could be folded onto LAT photon arrival times. This campaign ensured quasi‑homogeneous sky coverage, eliminating a priori selection biases that could otherwise skew detection statistics.

Using more than two years of LAT data, the authors performed phase‑folded searches for each MSP, applying statistical tests such as the H‑test and Z²_n to assess the significance of any pulsed signal. Eight MSPs—among them PSR J0030+0451, J0218+4232, J0613‑0200, J0751+1807, J1231‑1411, J1614‑2230, J1744‑1134, and J2124‑3358—exhibited pulsed γ‑ray emission above 100 MeV with >5σ confidence. Their spectra are well described by power‑law models with exponential cutoffs (photon indices Γ≈1.5–2.0, cutoff energies E_cut≈1–3 GeV), essentially identical to those of normal (young) γ‑ray pulsars. Phase alignment between the γ‑ray and radio light curves shows modest offsets (a few milliseconds), but the overall morphology (single or double peaks) matches predictions of outer‑magnetosphere emission models, such as the Outer Gap or Slot Gap scenarios. This similarity strongly suggests that MSP γ‑ray production occurs in the same high‑altitude regions as in younger pulsars, rather than near the stellar surface.

The authors also attempted to detect γ‑ray pulsations from MSPs residing in globular clusters. Despite the large number of cluster MSPs, no significant pulsations were found, likely because individual cluster MSPs are fainter and the dense stellar environments produce higher background levels that mask weak signals. Nonetheless, the null result emphasizes the need for longer LAT exposure and refined timing solutions for cluster sources.

A key implication of the study is that many of the unidentified LAT sources (the “UNID” catalog) may be previously unknown MSPs. Since the LAT’s positional uncertainties are on the order of a few arcminutes, targeted radio searches guided by LAT source locations could uncover a substantial hidden MSP population, thereby accounting for a significant fraction of the Galactic diffuse γ‑ray background. The paper therefore not only expands the known γ‑ray pulsar population by ~10 % but also reshapes our understanding of the Galactic γ‑ray source demographics.

In summary, the work demonstrates that the combination of LAT’s high sensitivity and a comprehensive, high‑precision timing campaign can reveal MSPs as γ‑ray pulsars. The detected MSPs exhibit spectral and temporal characteristics indistinguishable from those of normal pulsars, supporting outer‑magnetosphere emission mechanisms. The findings open a pathway to identify many more MSPs among the LAT’s unidentified sources, refine models of pulsar magnetospheres, and improve estimates of the contribution of pulsars to the Galactic γ‑ray budget.