Spider Pulsars (Black Widow & Redback)

The spider and its mate

Picture two stars locked in an orbit so tight that the whole dance fits inside the diameter of our Sun. One of them is a neutron star spinning hundreds of times a second — a city-sized ball of nuclear matter rotating faster than a kitchen blender. The other is a battered, bloated little star with one face permanently turned toward its companion, that face heated to a furnace glow while its back stays dim and cool. The neutron star is not merely orbiting its partner. It is killing it: a relativistic wind streaming from the pulsar boils gas off the companion's surface, gram by gram, year after year.

Astronomers named the phenomenon after spiders whose females eat the male after — or during — mating. The first one found, PSR B1957+20, became the black widow: a pulsar whose companion has been whittled down to a wisp of about 0.022 solar masses, well on its way to being consumed entirely. A second class, with heavier and more stubborn companions that survive the onslaught, took the name redback, after an Australian widow spider whose females are likewise cannibalistic but whose victims sometimes escape. Collectively they are the spider pulsars: the most intimate, most violent binaries in the pulsar zoo.

Where they come from: recycling a dead pulsar

To understand a spider you have to rewind a billion years. Neutron stars are born spinning fast but spin down quickly: a young pulsar like the Crab loses rotational energy to its magnetosphere and fades below the radio "death line" within tens of millions of years. A spider pulsar should be long dead. It is not, because it was recycled.

In a binary, when the companion star evolves and swells to fill its Roche lobe, it dumps gas onto the neutron star through an accretion disk. That infalling matter carries angular momentum, and over 10⁸–10⁹ years it spins the neutron star back up to millisecond periods — this is the millisecond pulsar recycling scenario, the engine that produces every spider. The accreting phase is observed directly as a low-mass X-ray binary (LMXB). When accretion finally ceases, the rejuvenated, rapidly spinning pulsar switches its radio beam back on — and now its wind is aimed point-blank at the very star that fed it. The prey becomes the victim.

The recycling also has a heavy consequence, literally. The accreted matter adds rest mass to the neutron star. That is why spiders host the most massive neutron stars ever weighed, pressing against the upper mass limit set by the dense-matter equation of state.

The mechanism: a relativistic blowtorch

A recycled pulsar's reservoir of energy is its rotation. The spin-down luminosity — the rate at which it bleeds rotational kinetic energy — is

Ė = 4π² I Ṗ / P³

where I ≈ 10⁴⁵ g·cm² is the neutron star's moment of inertia, P its spin period, and Ṗ the spin-down rate. For a typical spider with P ≈ 2–3 ms, Ė comes out around 10³⁴–10³⁵ erg/s. Most of this leaves the pulsar as a magnetised, relativistic wind of electron–positron pairs.

Now put a star a few light-seconds away. The fraction of the wind it intercepts is roughly its cross-sectional area divided by 4π a², where a is the orbital separation. For an orbital separation of about 1.5 R☉ and a companion radius of a few tenths of a solar radius, the companion catches on the order of 0.1–1 percent of the spin-down power. That sounds small, but it is delivered to a tiny, low-gravity star, and it concentrates entirely on the day side. The irradiating flux at the companion can exceed its own intrinsic luminosity by orders of magnitude.

The day side is heated to 5,000–10,000 K above the night side — directly measured from the photometric light curve, which shows a strong sinusoidal brightness variation as the heated face rotates into and out of view. The deposited energy ionises the surface and drives a thermal wind off the companion. Where this wind meets the oncoming pulsar wind, the two ram pressures balance and form an intra-binary shock, a curved sheet of shocked plasma wrapped around the companion. That shock is the source of the systems' orbitally modulated X-rays, and the ionised cloud behind it is what eclipses the radio pulses.

Quantifying the ablation

How fast does the prey actually shrink? The instantaneous mass-loss rate scales with the fraction of spin-down power that goes into unbinding surface material:

Ṁ_c ≈ ε · (Ė / 4π a²) · (π R_c² / v_esc²)

where R_c and v_esc are the companion's radius and surface escape speed and ε is an efficiency factor of order a few percent. Plugging in spider-typical numbers gives Ṁ_c in the range 10⁻¹² to 10⁻¹⁰ solar masses per year. Compare that to the companion's remaining mass: a 0.02 M☉ black-widow companion losing 10⁻¹¹ M☉/yr would, at a constant rate, evaporate in about 2 × 10⁹ years — comparable to the system's age. The rate is not constant, of course; it slows as the companion shrinks and its irradiated cross-section falls, which is precisely why finishing the job within a Hubble time is an open question.

Quantity	Black widow	Redback
Companion mass	0.01 – 0.05 M☉	0.1 – 0.7 M☉
Companion type	Semi-degenerate / He core	Non-degenerate, near main sequence
Orbital period	~1.5 – 10 h	~2 – 15 h
Radio eclipse fraction	~10 % of orbit	~30 – 60 % of orbit
Day-side temperature	up to ~10,000 K hotter	several thousand K hotter
State switching	No	Yes (transitional MSPs)
Prototype	PSR B1957+20	PSR J1023+0038

How we observe them: eclipses, X-rays, gamma rays

Three signatures together identify a spider, and the discovery pipeline that found most of them ties all three to one space telescope.

• Radio eclipses. As the pulsar swings behind the companion's ablated plasma cloud near superior conjunction, the radio pulses are absorbed, scattered, and dispersed. Crucially the eclipses are frequency-dependent — they last longer and the pulses arrive more delayed at lower observing frequencies, because free-free absorption and dispersion both scale strongly with wavelength. This is the fingerprint that distinguishes a plasma eclipse from a simple geometric one. The companion star itself is far too small to occult the pulsar.
• Optical light curves. The heated day side makes the companion vary by one to several magnitudes over an orbit. Modelling that light curve gives the companion temperature, the orbital inclination, and — combined with radial velocities of the companion's spectral lines — the neutron-star mass. This is how the record-breaking masses are weighed.
• X-rays from the shock. The intra-binary shock radiates synchrotron X-rays that are modulated on the orbital period, often peaking near the companion's superior or inferior conjunction depending on the shock geometry. Missions like Chandra, XMM-Newton, and NICER trace it.

The reason we now know of dozens of spiders, rather than the one or two known before 2009, is the Fermi Large Area Telescope. Millisecond pulsars are prolific gamma-ray emitters, and Fermi's all-sky survey produced a long list of unidentified gamma-ray point sources. Radio and optical follow-up of those sources turned up tightly orbiting, eclipsing binaries by the dozen — the modern spider population is largely a Fermi harvest. Many were first spotted as gamma-ray sources, then confirmed as pulsars by detecting their pulsations.

Real numbers and famous examples

• PSR B1957+20 — the original black widow. Discovered by Fruchter, Stinebring & Taylor in 1988, about 6,500 light-years away in Sagitta. Spin period 1.6 ms, orbital period 9.2 hours, companion ≈ 0.022 M☉. Its name gave the entire class its identity. Its companion shows a comet-like tail of ablated gas streaming away from the system.
• PSR J0952−0607 — the heaviest neutron star. A black widow with a 1.41 ms spin period (one of the fastest known), about 3,200–5,700 light-years away. Optical spectroscopy of its evaporating companion yields a neutron-star mass of ≈ 2.35 ± 0.17 M☉ (Romani et al. 2022), the current record and a stringent lower bound on the maximum neutron-star mass.
• PSR J1311−3430 — nearly nothing left. Discovered as a Fermi gamma-ray source, this black widow has a 2.5 ms spin, a brutally short 93-minute orbit, and a companion of only ≈ 0.01 M☉ — a hydrogen-depleted, helium-dominated remnant, a star ablated almost out of existence.
• PSR J1023+0038 — the prototype transitional redback. Famously caught switching between a radio-pulsar state and an accreting low-mass X-ray binary state. It had a visible accretion disk in 2000–2001, lost it, and then formed one again in 2013 — the clearest real-time demonstration of the LMXB↔MSP link.
• Globular-cluster spiders. Dense clusters like Terzan 5, 47 Tucanae, and Omega Centauri are spider factories, because frequent stellar encounters exchange partners and drive binaries into the tight configurations spiders require. The transitional source in M28 (IGR J18245−2452) was caught swinging between states in 2013.

What happens at the end

The long-term fate divides along the black-widow/redback line. Redbacks, with their heavier non-degenerate companions, can still drive episodic accretion — they are the transitional systems, flipping between radio-pulsar and X-ray-binary behaviour on timescales of years. They have not yet ablated their mate to insignificance.

Black widows are further along. Their companions are already semi-degenerate husks. The compelling open question is whether continued ablation can fully destroy the companion, leaving behind a solitary, fully recycled neutron star spinning hundreds of times a second with no partner at all. Such isolated millisecond pulsars exist — perhaps 10–20 percent of the field millisecond-pulsar population — and the black-widow ablation channel was historically the leading explanation. The difficulty is that ablation self-limits: as the companion shrinks, it intercepts less wind and loses mass more slowly, so the last sliver is the hardest to remove. Whether the spider truly finishes its meal, or whether isolated millisecond pulsars come from neutron-star mergers and other paths, remains genuinely unsettled.

Misconceptions and edge cases

• The eclipse is not the star blocking the pulsar. A 0.02 M☉ companion is roughly the size of a large planet — far too small to geometrically occult a beam from several light-seconds away at most orbital geometries. The eclipse is caused by ionised, ablated plasma absorbing and scattering the radio signal, which is why it is frequency-dependent.
• The pulsar is not "eating" the companion the way a black hole accretes. In the rotation-powered spider state, material is being blown off the companion by the wind and largely lost from the system — not funnelled onto the neutron star. Direct accretion happens only in the earlier LMXB phase and in transitional redbacks' accreting episodes.
• Black widow ≠ redback by spin. Both are millisecond pulsars with similar spin periods. The dividing line is companion mass and degeneracy, not the pulsar's rotation rate.
• "Spider" is not the same as a generic eclipsing binary pulsar. The defining feature is active ablation of the companion by the pulsar's high-energy output, producing an intra-binary shock and orbitally modulated emission — not merely a line of sight that happens to pass through a stellar atmosphere.
• The recycling is mostly finished, but mass can still creep up. The bulk of the spin-up and mass gain happened in the LMXB phase. In the present spider phase the neutron star is mostly shedding the companion, not growing — yet transitional redbacks can still accrete in bursts, which is why their masses are also among the largest known.