Boson Star

What a boson star actually is

Every star you have ever heard of is held up against its own gravity by one of three things: thermal pressure from fusion (the Sun), electron degeneracy pressure (white dwarfs), or neutron degeneracy pressure (neutron stars). All three rely on matter built from fermions — particles that obey the Pauli exclusion principle and therefore refuse to be packed into the same quantum state. A boson star is what you get when you remove that crutch entirely and ask: can a cloud made of bosons — particles that are perfectly happy to pile into the same state — hold itself up?

The answer, remarkably, is yes. A boson star is a stationary, self-gravitating lump of a quantum field whose particles all occupy a single macroscopic quantum state — a gravitationally bound Bose-Einstein condensate. There is no fusion, no nuclear burning, no Pauli pressure. What keeps it from collapsing is the same effect that keeps the electron in a hydrogen atom from falling into the proton: confine a quantum wavefunction to a small region and Heisenberg's uncertainty principle forces it to carry large momentum, which shows up as an outward "quantum pressure." Balance that gradient pressure against gravity and you have a star with no atoms, no surface, and no light of its own.

Boson stars were first written down by David Kaup in 1968 (for a complex scalar field, originally called a "Klein-Gordon geon") and independently by Remo Ruffini and Silvano Bonazzola in 1969. They have never been observed. But because they are smooth, horizonless, and can be made extremely compact, they have become the workhorse model for testing whether the dark, massive objects we call black holes truly have event horizons.

The Einstein–Klein–Gordon system

A boson star is a self-consistent solution of general relativity coupled to a complex scalar field φ. Gravity curves spacetime; the field's energy density sources that curvature; the curvature in turn tells the field how to evolve. In the simplest case the field is a free, massive scalar with Lagrangian density

L = -gᵘᵛ ∂ᵤφ* ∂ᵥφ - m²|φ|²/ℏ²

Varying the action gives the curved-space Klein-Gordon equation for the field and Einstein's equations for the metric, sourced by the field's stress-energy tensor:

□φ = (m²/ℏ²) φ          (Klein-Gordon, in curved spacetime)
Gᵤᵥ = 8πG/c⁴ · Tᵤᵥ[φ]   (Einstein, sourced by the field)

The trick that makes a static star out of a field with a mass term is a harmonically oscillating phase:

φ(r, t) = σ(r) · e^(-i ω t)

The amplitude profile σ(r) is time-independent, so the metric and the energy density are static — but the field underneath is oscillating at frequency ω. This is the crucial subtlety: a boson star is not a frozen ball of stuff, it is a coherently breathing field whose observable (the energy density) is stationary even though the field itself never stops oscillating. The conserved U(1) charge of the complex field counts the number of bosons N, and it is this Noether charge — not baryon number — that is conserved and stabilises the configuration.

How quantum pressure beats gravity

You can recover the essential mass scale with a back-of-envelope uncertainty argument, no general relativity required. Pack N bosons of mass m into a ball of radius R. The condensate is a single wavefunction of extent R, so each quantum carries momentum p ~ ℏ/R and a kinetic energy ~ p²/2m ~ ℏ²/(2 m R²). The gravitational binding energy per particle is ~ G N m² / R. The total energy per particle is

E(R) ≈ ℏ²/(2 m R²)  −  G N m² / R

The first term (quantum pressure) blows up as R → 0; the second (gravity) wants R → 0. Minimising E(R) gives an equilibrium radius and, crucially, a maximum particle number above which no minimum exists and the star collapses. Setting dE/dR = 0 and pushing to the relativistic limit reproduces the Kaup result up to a numerical factor: the maximum mass scales as M_Pl²/m, where M_Pl = √(ℏc/G) ≈ 2.18 × 10⁻⁸ kg ≈ 1.22 × 10¹⁹ GeV/c² is the Planck mass.

The take-home physics: because bosons share one wavefunction, the only thing resisting collapse is the gradient energy of that wavefunction. That is far weaker than fermion degeneracy, so a non-interacting boson star is tiny in mass. To build a heavy one you either use an extraordinarily light boson, or you add a repulsive self-interaction.

The mass scale — why the particle mass is everything

The single most important number for a boson star is the mass m of the constituent boson. The maximum stellar mass depends inversely on it.

Configuration	Maximum mass M_max	For m = 1 GeV	For m = 10⁻¹⁰ eV
Non-interacting ("mini")	≈ 0.633 MPl²/m	~8 × 10⁻²⁰ M☉ (asteroid)	~1 M☉ (stellar)
Repulsive self-interaction (λ ~ 1)	≈ 0.06 √λ · MPl³/m²	~0.1 M☉ (sub-stellar)	≫ galaxy mass
"Massive" (large λ)	scales as √λ · MPl³/m²	tunable to any scale	tunable to any scale

Two clean numerical facts to anchor the scaling. For a free field the Kaup limit is

M_max ≈ 0.633 M_Pl²/m ≈ 8.5 × 10⁻²⁰ (GeV/m) M☉

so a GeV-mass boson makes only an asteroid-mass star, while an ultralight boson of m ≈ 2 × 10⁻¹⁷ eV reaches the 4 × 10⁶ M☉ scale of the Galactic Centre. Adding a quartic self-interaction λ|φ|⁴ changes the scaling from M_Pl²/m to M_Pl³/m², a result derived by Marco Colpi, Stuart Shapiro and Ira Wasserman in 1986:

M_max ≈ 0.06 √λ · M_Pl³/m² ≈ 0.06 √λ (GeV/m)² M☉

The extra factor of M_Pl/m (which is enormous — about 10¹⁹ for a GeV boson) is exactly why self-interacting boson stars can rival the masses of neutron stars and supermassive black holes while a free-field star cannot.

Compactness, light rings, and the absence of a horizon

What makes a boson star a mimicker rather than a curiosity is compactness — how close the radius approaches its own Schwarzschild radius r_s = 2GM/c². Define the compactness C = GM/(Rc²). A black hole sits at the horizon, C = 1/2. A neutron star reaches C ≈ 0.2–0.3. Boson stars span a wide range:

Object	Compactness C = GM/Rc²	Light ring?	Horizon?	Surface?
Sun	~2 × 10⁻⁶	No	No	Yes
White dwarf	~3 × 10⁻⁴	No	No	Yes
Neutron star	~0.2	Marginal	No	Yes (hard)
Dilute boson star	0.01 – 0.1	No	No	No (smooth taper)
Compact boson star	up to ~0.3+	Yes (two light rings)	No	No
Schwarzschild black hole	0.5 (at horizon)	Yes (one)	Yes	No

The decisive structural difference is that the field amplitude σ(r) never has a sharp edge — it falls off exponentially as e^(−mr) at large radius, so there is no surface where the density drops to zero discontinuously. And because there is no point of infinite redshift, there is no event horizon: a photon or test particle aimed at the centre passes through the dilute interior and out the other side. The most compact, strongly self-interacting models can nevertheless develop a pair of light rings (an outer unstable one and an inner stable one) and cast a dark central region in an image — close enough to a black hole's shadow that you need precision data to tell them apart.

Varieties: scalar, vector, charged, rotating, and oscillatons

• Scalar boson star (Kaup, 1968). The canonical case: a complex spin-0 field with a global U(1) symmetry whose Noether charge is conserved particle number. Spherically symmetric ground states are stable below the Kaup limit.
• Proca / vector star. Built from a massive complex spin-1 (Proca) field instead of a scalar. They support equilibrium configurations and, intriguingly, allow rotating solutions that connect smoothly to the spherical limit — unlike scalar stars, whose rotating branch is quantised in angular momentum J = N ℏ.
• Charged boson star. A complex field coupled to electromagnetism (Jetzer, 1989). Coulomb repulsion adds to quantum pressure, raising the maximum mass until the charge-to-mass ratio approaches the gravitational instability.
• Oscillaton. The real-field cousin. A real scalar field cannot form a truly static star, but it can form a long-lived quasi-stationary configuration whose metric oscillates in time — an "oscillaton." Axion dark matter, being a real field, forms oscillatons rather than true boson stars.
• Self-interacting / "massive" star. Add λ|φ|⁴ and the maximum mass jumps by the factor M_Pl/m, as Colpi, Shapiro and Wasserman showed; these are the models heavy enough to mimic astrophysical black holes.

Where boson stars show up: dark matter and mimicker tests

Boson stars are not merely a thought experiment, because the universe may well contain fundamental bosons in abundance.

• Fuzzy / ultralight dark matter. If dark matter is an ultralight scalar with m ~ 10⁻²² eV, its de Broglie wavelength is kiloparsec-scale and quantum pressure suppresses small-scale structure. Cosmological simulations (Schive, Chiueh & Broadhurst, 2014) show that every halo forms a dense solitonic core at its centre — a galaxy-scale boson star — surrounded by a granular, wave-like halo. This is one of the most actively studied alternatives to cold dark matter.
• Axion stars. The QCD axion (m ~ 10⁻⁵ eV) and axion-like particles can clump into gravitationally bound oscillatons. Dilute axion stars are stable up to a maximum mass; above a critical density they can collapse and emit a relativistic burst of axions or, in some models, radio photons (a "bosenova").
• Black-hole mimickers. This is the headline application. The Event Horizon Telescope images of M87* (2019) and Sagittarius A* (2022) show dark shadows; boson-star models have been fit to both to ask how horizonless an object could be and still match the data. The orbit of the star S2 around Sgr A* — tracked to a pericentre of ~120 AU with relativistic precession measured by the GRAVITY instrument — constrains any extended mass inside that orbit.
• Gravitational-wave echoes. When two compact objects merge, the remnant "rings down." A black hole's ringdown is absorbed at the horizon; a horizonless object can instead trap waves in a cavity between its light ring and centre, releasing them as a train of delayed echoes. LIGO–Virgo–KAGRA searches for such echoes are a direct, model-independent test for the absence of a horizon.

Concrete numbers to keep in your head

A worked sense of scale. Suppose dark matter is an ultralight scalar of mass m = 10⁻²² eV/c² = 1.8 × 10⁻⁵⁸ kg — the canonical "fuzzy" value. The non-interacting maximum (Kaup) mass for such a field is

M_max ≈ 0.633 M_Pl²/m = 8.5 × 10⁻²⁰ (GeV/m) M☉
      = 8.5 × 10⁻²⁰ × (10⁹ eV / 10⁻²² eV) M☉
      ≈ 8 × 10¹¹ M☉

That ceiling is enormous — of order the entire mass of a large galaxy, and thousands of times more than any soliton a halo actually builds — which is the real point: an ultralight field is so far below its own collapse threshold that the Kaup limit is irrelevant to it. The mass of a real fuzzy-dark-matter soliton is instead set by the halo it condenses in. Cosmological simulations give core masses of order 10⁷–10⁹ M☉ and core radii of order a kiloparsec (about 3 × 10¹⁹ m): enormously diffuse, ultra-low-compactness objects, nothing like a black hole. To make a compact star that could mimic the Galactic Centre instead, you need a much heavier boson: a non-interacting field reaches the 4 × 10⁶ M☉ of Sgr A* near m ~ 2 × 10⁻¹⁷ eV, and a self-interacting field can do it with a heavier boson still. The field's oscillation frequency, ω ≈ mc²/ℏ, is roughly 1.5 × 10⁻⁷ rad/s for m = 10⁻²² eV (a period of about a year) but climbs to ~3 × 10⁻² rad/s for the m ~ 2 × 10⁻¹⁷ eV mimicker — slow by particle standards, but the engine of everything the star does.

Common misconceptions and edge cases

• "A boson star is just a black hole made of bosons." No — the defining feature is the absence of a horizon. A boson star that became compact enough to form a horizon would simply collapse into an ordinary black hole, and there is a maximum mass beyond which that happens. The interesting regime is precisely the one where it stays horizonless.
• "The star is static, so the field must be static." The energy density is static, but the complex field oscillates as e^(−iωt) forever; that coherent oscillation is what supports the star. For a real field you cannot even make it perfectly static — you get a time-oscillating oscillaton instead.
• "Bosons can't form stars because they have no exclusion pressure." They have no Pauli pressure, true, but the gradient energy of a confined wavefunction (quantum pressure) and any self-interaction supply the needed support. The lack of Pauli pressure only means the maximum mass is small unless the boson is ultralight or self-interacting.
• "Excited states are stable stars too." Radially excited boson stars (with nodes in σ(r)) exist mathematically but are dynamically unstable: they either decay to the nodeless ground state or collapse. Only the ground-state branch below the maximum mass is stable.
• "If the EHT saw a shadow, the object can't be a boson star." A sufficiently compact boson star also casts a shadow via its light ring, so a shadow alone does not prove a horizon. Discriminating a horizon from a horizonless mimicker requires shadow size and sharpness, gravitational-wave echoes, or orbits that probe the deep interior — not just the existence of a dark patch.