Neutron Stars, Pulsars and Magnetars

At a glance

A neutron star is the collapsed core of a massive star, holding more mass than the Sun in a sphere the size of a city. Pulsars are neutron stars whose beams sweep past Earth as they rotate. Magnetars are neutron stars powered by extremely strong magnetic fields.

Sources checked October 2026. In October 2026 the ATNF home page showed version 2.9.0, but the history page gives counts only up to v2.8.1. The radii are half the NICER diameters quoted by NASA. The mean density and surface gravity are calculated for 1.4 M_☉ and 12 km, as orders of magnitude.

Sources: [1] [2] [3] [4] [5] [6]

Formation and internal structure

Neutron stars form in core-collapse supernovae. When a massive star exhausts its nuclear fuel, its iron core can no longer support itself and collapses in a fraction of a second. Electrons and protons merge into neutrons, releasing a flood of neutrinos. The collapse halts at nuclear densities, and the star’s outer layers are blown away. According to NASA, a collapsing core of roughly 1 to 3 solar masses becomes a neutron star, while heavier cores go on to form black holes. A star made of neutrons was first proposed in 1934; the discovery of pulsars in 1967 confirmed that such objects exist.

A neutron star is not a uniform ball of neutrons. Its thin crust contains atomic nuclei that grow richer in neutrons with depth, until free neutrons appear between them. The core is mostly neutrons, with smaller numbers of protons and electrons. Some models allow exotic phases at the centre, such as a “sea” of free quarks, but NICER results suggest that core matter is harder to compress than many such models predict.

NASA compares a neutron star to the size of Manhattan Island, with roughly half a million Earth masses inside. At such compactness, general relativity is needed to describe both the star’s structure and the light that leaves it.

Sources: [1] [2]

Mass, radius and the equation of state

The relation between pressure and density in neutron-star matter, the equation of state, sets both a star’s radius and its maximum possible mass. For PSR J0740+6620, radio timing of the Shapiro delay (the extra travel time of pulses passing close to the companion) gives 2.08 ± 0.07 M_☉ (68.3% credibility). Stars this heavy rule out the “softest” equations of state, which could not hold them up against collapse.

Radii are harder. NASA’s NICER X-ray telescope on the International Space Station times photons from hot spots on a pulsar’s surface to better than 100 nanoseconds. Gravitational light bending shapes the pulses, revealing the star’s compactness. In 2019 NICER found a radius of about 13 km for PSR J0030+0451, a star of about 1.4 M_☉. In 2021 two teams found about 12.4 km (Amsterdam-led) and 13.7 km (Maryland-led) for J0740+6620, with overlapping uncertainties. Although J0740+6620 is almost 50% more massive, it is essentially the same size. Pressure must therefore rise steeply in the cores of massive neutron stars.

The theoretical maximum mass of a non-rotating neutron star (the Tolman–Oppenheimer–Volkoff limit) is still not known exactly; observations show only that it is at least about 2 M_☉. The tidal deformation of the stars in the GW170817 merger, described below, adds an independent constraint.

Sources: [3] [2]

The discovery of pulsars and the lighthouse model

Pulsars were discovered at Cambridge in 1967. Jocelyn Bell, a PhD student supervised by Antony Hewish, was analysing chart recordings from a wire-array radio telescope at the Mullard Radio Astronomy Observatory when she noticed an odd signal. On 28 November 1967 the recordings showed a string of pulses about one and a third seconds apart. The source, CP1919, was jokingly labelled “LGM” (“Little Green Men”).

The 1974 Nobel Prize in Physics went to Antony Hewish and Martin Ryle; Bell was not included, which has long been debated. Bell has said her contemporaries were more upset than she was. In 2018 she received the Special Breakthrough Prize in Fundamental Physics and used the money to support under-represented groups in physics.

In the lighthouse model, a neutron star emits beams along its magnetic axis. If that axis is tilted from the rotation axis, the beams sweep through space and we see a pulse each time one crosses Earth. The energy comes from rotation, so the pulsar slowly spins down. Treated as a rotating magnetic dipole, the period P and its rate of change Ṗ give the surface field (B ∝ (PṖ)^1/2) and the characteristic age (τ = P/2Ṗ). A typical normal pulsar, with P ≈ 0.5 s and Ṗ ≈ 10⁻¹⁵ s/s, has a field of about 10⁸ T and an age of about 10 million years.

Sources: [7] [4]

Interactive lab: a pulsar’s age and magnetic field

The lab opens on the Crab pulsar: notice that the characteristic age worked out from its period and spin-down rate points back to the supernova seen in 1054, give or take a few hundred years. Then pick the first pulsar found (CP1919 in the text, now called B1919+21): its characteristic age is about 16 million years and it sits among the normal pulsars. The millisecond pulsar J0437−4715 and the magnetar candidate 1E 2259+586 land in opposite corners of the diagram; compare where they fall relative to the lines of constant field and constant age.

Millisecond pulsars, binary pulsars and tests of gravity

Millisecond pulsars have periods of about 1.4–30 ms and relatively weak fields, around 10⁴ T. The standard explanation is recycling: a neutron star in a binary first spins down as an ordinary pulsar, then, when its companion swells into a giant and overflows its Roche lobe, accretes matter and angular momentum and is spun up to millisecond periods. The companion is left as a white dwarf, which is why most millisecond pulsars have white-dwarf companions. The fastest known, PSR J1748−2446ad in the globular cluster Terzan 5, spins 716 times per second.

The first binary pulsar, PSR B1913+16, was found in 1974 by Russell Hulse and Joseph Taylor with the 300-m Arecibo telescope. The pulsar and its companion, a neutron star of similar mass, orbit each other in under eight hours. Timing showed the orbital period shrinking by about 75 millionths of a second per year, matching general relativity’s prediction for energy lost to gravitational waves to within about half a percent. This first, indirect evidence for gravitational waves earned the 1993 Nobel Prize in Physics.

In the double pulsar PSR J0737−3039A/B, both neutron stars have been detected as radio pulsars. Using 16 years of timing, Kramer and colleagues measured seven post-Keplerian parameters, more than for any other binary pulsar, and found that the orbital decay agrees with general relativity’s quadrupole formula at the 1.3 × 10⁻⁴ level (95% confidence), one of the most precise strong-field tests of gravity.

Sources: [4] [8] [9]

Pulsar timing arrays and the nanohertz gravitational-wave background

A pulsar timing array uses stable millisecond pulsars as a galaxy-sized gravitational-wave detector. It looks for tiny, correlated shifts in pulse arrival times. Gravitational waves with periods of years to decades leave a characteristic signature: the correlation between two pulsars depends on their angular separation on the sky, following the Hellings–Downs curve.

In late June 2023 the North American NANOGrav collaboration published its 15-year data set. In an analysis of 67 pulsars, the Hellings–Downs correlation appeared at about 3.5–4σ, and the background’s characteristic strain amplitude was about 2.4 × 10⁻¹⁵ at a reference frequency of 1/year. The European (EPTA, with the Indian InPTA), Australian (PPTA) and Chinese (CPTA, using the FAST telescope) arrays reported their own evidence at varying significance. A later comparison by the International Pulsar Timing Array (IPTA) found that their background parameters agree within 1σ.

The result counts as “evidence” rather than a definitive detection. Its amplitude and spectrum are consistent with a population of supermassive black-hole binaries, but more exotic cosmological and astrophysical sources cannot be excluded.

Sources: [10] [11]

Magnetars and fast radio bursts

Magnetars are neutron stars whose X-ray and gamma-ray activity is powered mainly by an ultra-strong magnetic field rather than by rotation. For most of the roughly 30 known magnetars, spin-down implies fields above 10¹⁰ T (10¹⁴ G); the first such measurement, for SGR 1806−20, gave about 8 × 10¹⁰ T. Their X-ray pulsations have periods of about 2–12 s. Magnetars emit short bursts and, rarely, giant flares: the one from SGR 1806−20 on 27 December 2004 released more than 10³⁹ J in its initial spike.

Fast radio bursts (FRBs) are millisecond radio flashes, mostly from other galaxies. On 28 April 2020 the CHIME telescope in Canada and the STARE2 array in the United States recorded an exceptionally bright radio burst from the Galactic magnetar SGR 1935+2154, while INTEGRAL, China’s Huiyan (Insight-HXMT) and the Konus instrument on Wind caught a simultaneous X-ray burst. It was the first FRB seen from within the Milky Way. The magnetar’s distance is poorly known (about 14,000–41,000 light years). The event strongly supports a magnetar origin for at least some FRBs.

Ersin Göğüş and Yuki Kaneko of Sabancı University in Türkiye were part of an international team that analysed 24 intense X-ray bursts from SGR 1935+2154 observed with Fermi/GBM and NICER in 2020. The study, published in Nature Astronomy in February 2021, found that X-ray bursts coinciding with FRBs differed spectrally from the others.

Sources: [5] [12] [13]

Neutron star mergers and X-ray binaries

On 17 August 2017 LIGO and Virgo recorded GW170817, the first gravitational-wave signal from two merging neutron stars, about 130 million light years away. The component masses were estimated at about 1.1–1.6 M_☉. Gamma rays arrived about two seconds after the gravitational waves, and some 70 observatories followed the event. The kilonova that followed showed freshly made heavy elements such as gold and platinum, supporting mergers as a major source of elements heavier than iron. For neutron-star physics, the tidal deformation encoded in the signal constrains the equation of state; the multi-messenger story is covered in a separate dossier.

Many neutron stars live in X-ray binaries, where gas pulled from a companion shines in X-rays as it falls in. The transferred matter also carries angular momentum and can spin the neutron star up: the recycling channel that produces millisecond pulsars.

Sources: [14] [4]

How to go further: data and archives

The standard reference for pulsar parameters is the ATNF Pulsar Catalogue, which lists all published rotation-powered pulsars, including those seen only at high energies. Its history page records the count for each version: 4,351 in v2.7.0 (October 2025) and 4,393 in v2.8.1 (June 2026). Users should cite Manchester et al. (2005, Astronomical Journal 129, 1993).

NASA’s HEASARC archive serves X-ray and gamma-ray data from NICER, Fermi, Swift, NuSTAR, XMM-Newton, IXPE and other missions, with software such as HEASoft and XSPEC. NANOGrav’s 15-year timing data and background products are on its data page. Gravitational-wave event catalogues (GWTC), open detector data and tutorials are available from the Gravitational Wave Open Science Center (GWOSC).

Sources: [6]

Known vs open questions

Established

• Neutron stars form in core-collapse supernovae; pulsars are rotating neutron stars.
• Neutron stars of at least about 2 M_☉ exist (PSR J0740+6620: 2.08 ± 0.07 M_☉).
• Typical radii are around 12–13 km, and massive neutron stars are not noticeably smaller than 1.4 M_☉ ones.
• Orbital decay in binary pulsars matches general relativity’s gravitational-wave prediction with high precision.
• Neutron-star mergers produce kilonovae and heavy elements.
• At least one Galactic magnetar has produced a fast radio burst.

Open questions

• What is the exact maximum mass, and what phase of matter (hyperons, quark matter) lies at the centre?
• How do magnetar fields form and decay, and do all FRBs come from magnetars?
• Is the nanohertz background due purely to supermassive black-hole binaries, or partly to new physics?
• Where exactly does pulsar radio emission originate, and what sets its beam geometry?