Common Envelope Evolution

Common envelope evolution (CEE) is a brief, transformative phase in the life of an interacting binary star system in which an expanding donor star—typically a red giant, asymptotic giant branch (AGB) star, or red supergiant—undergoes dynamically unstable mass transfer and engulfs its companion star within a single, shared, non-corotating gaseous envelope. First proposed symmetrically by Polish astrophysicist Bohdan Paczyński in 1976 (building on earlier qualitative ideas by Jeremiah Ostriker in 1975 and William Sparks & Warren Stecher in 1974) to explain the origin of the short-period white dwarf + main-sequence binary V471 Tauri, common envelope evolution solves the orbital shrinkage problem of compact binary astrophysics. Because the dense core of the giant star (M_c) and the engulfed companion (M₂) orbit inside the extended gaseous envelope at supersonic speeds without corotating with the gas, hydrodynamic and gravitational drag (dynamical friction) transfers orbital energy and angular momentum from the binary pair to the surrounding envelope over a dynamical to thermal timescale of only 1 to 1,000 years. This inspiral shrinks the binary orbital separation a by factors of 100 to 1,000—either unbinding and ejecting the shared envelope to leave behind a tight post-common-envelope binary (the primary channel for cataclysmic variables, bipolar planetary nebulae, double white dwarfs, X-ray binaries, Type Ia supernovae, and gravitational-wave mergers of binary neutron stars and black holes) or terminating in a complete stellar merger (forming luminous red novae or hybrid Thorne-Żytkow objects).

Across the Milky Way and extragalactic surveys, astronomers observe numerous classes of close binary systems containing at least one evolved compact remnant—a white dwarf, neutron star, or stellar-mass black hole—with orbital periods of minutes to hours and orbital separations of a_f ≈ 0.5 to 5 R_☉ (0.002 to 0.02 au). Examples include cataclysmic variables (accreting white dwarfs with P_orb ≈ 1.3–10 hours), double white dwarf binaries (such as HM Cancri with P_orb = 5.4 minutes), low-mass X-ray binaries, and double neutron star or black hole binaries that coalesce within a Hubble time via gravitational-wave radiation.

However, for the progenitor star of the compact remnant to have synthesized its degenerate core via post-main-sequence nuclear burning, it must previously have evolved through a red giant or asymptotic giant branch phase during which its photospheric radius expanded to R_giant ≈ 100 to 1,000 R_☉ (0.5 to 5 au). If the binary had possessed its present-day tight separation (a_f ≈ 1–3 R_☉) on the main sequence, the two stars would have merged long before the primary could develop a degenerate white dwarf or neutron star core. Conversely, for the primary to grow undisturbed into a giant of radius 200 R_☉, the initial binary orbital separation must have been a_i ≳ 200–1,000 R_☉. Common envelope evolution provides the dissipative hydrodynamic bridge that shrinks a wide binary orbit (a_i ~ hundreds of R_☉) down to a compact orbit (a_f ~ a few R_☉) in less than a millennium.

A binary enters a common envelope phase when the expanding donor star (mass M₁ = M_c + M_env) fills its Roche lobe of effective Eggleton (1983) radius R_L,1(q) ≈ a · 0.49 q^2/3 / [0.6 q^2/3 + ln(1 + q^1/3)] (where q = M₁ / M₂ is the mass ratio) and the resulting mass transfer through the inner Lagrangian point L₁ becomes dynamically unstable.

The Mass-Radius Exponent Criterion (ζ_ad < ζ_L)

Stability of Roche lobe overflow is governed by comparing the donor star's adiabatic mass–radius exponent ζ_ad = (∂ ln R₁ / ∂ ln M₁)_S against the Roche-lobe mass–radius exponent ζ_L = (∂ ln R_L,1 / ∂ ln M₁):

• Radiative Envelope Donors (ζ_ad > 0): Stars with radiative outer envelopes contract when they lose mass on an adiabatic (dynamical) timescale, allowing stable mass transfer up to moderate mass ratios.
• Deep Convective Envelope Donors (ζ_ad ≈ −1/3 to 0, Hjellming & Webbink 1987): Red giants, AGB stars, and red supergiants possess deep convective envelopes whose specific entropy is nearly flat. When such a donor loses mass, its convective envelope expands adiabatically (R₁ ∝ M₁^−1/3, so ζ_ad < 0), while for q = M₁ / M₂ > q_crit ≈ 1.2–1.5 (when the donor is more massive than the accretor), mass transfer from M₁ to M₂ causes the orbit and the donor's Roche lobe R_L,1 to shrink (ζ_L > 0). Because the giant expands while its Roche lobe contracts (ζ_ad < ζ_L), a runaway feedback loop ensues on the donor's dynamical (sound-crossing) timescale (~weeks to years). Mass-transfer rates surge past 10⁻² to 1 M_☉ yr⁻¹, far exceeding the companion's Eddington luminosity accretion limit, and the overflowing gas engulfs both stars beyond the outer Lagrangian point L₂.

Four Hydrodynamic Phases of Common Envelope Evolution (Ivanova et al., 2013)

1. Phase I — Loss of Corotation and Plunge-In Prelude: Prior to or during runaway Roche-lobe overflow (or when the Darwin tidal instability occurs because the orbital moment of inertia drops below three times the giant's spin moment of inertia, I_orb < 3 I_spin), tidal torques can no longer keep the giant's envelope synchronized with the tightening orbit, and the companion plunges beneath the giant's photosphere.
2. Phase II — Rapid Dynamical Plunge-In (t ~ 1 to 100 orbital periods, months to years): Moving supersonically through the dense non-corotating envelope, the companion M₂ and the giant's compact core M_c excite trailing spiral density wakes. Gravitational drag (dynamical friction, F_DF ∝ G² M₂² ρ / (v_rel² + c_s²)) extracts orbital energy and angular momentum, driving an inspiral that reduces the separation a by 1 to 2 orders of magnitude and inflates the outer envelope into an expanding shock-heated cocoon.
3. Phase III — Self-Regulated Slow Spiral-In (t ~ 10 to 1,000 years): Once the expanded envelope density around the inner binary drops and the local gas approaches corotation with the inspiraling core–companion pair, dynamical friction weakens. Further orbital decay proceeds on the kelvin-Helmholtz thermal timescale of the envelope, assisted by hydrogen and helium recombination energy (~13.6 eV per H atom and 24.6 + 54.4 eV per He atom released as expanding ionized gas cools below 10⁴ K) and dust-driven radiation pressure winds.
4. Phase IV — Envelope Ejection or Stellar Merger: Either the cumulative orbital energy plus recombination energy exceeds the gravitational binding energy of the envelope—ejecting it as a bipolar planetary nebula with an equatorial torus and exposing a tight post-CE binary—or the binary coalesces before the envelope is unbound, producing a single rapidly rotating star or compact merger transient.

Because three-dimensional hydrodynamic simulations cannot easily resolve nine orders of magnitude in spatial scale (from a 10-km neutron star or 10⁴-km white dwarf to a 10⁹-km AGB envelope) across thousands of orbits, binary population synthesis codes parameterize common envelope outcomes using two conservation formalisms:

1. The Standard Energy Formalism (α_CE–λ Prescription; Webbink 1984; Livio & Soker 1988)

The α_CE formalism equates the gravitational binding energy E_bind of the donor's envelope to a fraction α_CE (the common-envelope ejection efficiency, typically 0 < α_CE ≤ 1, or >1 if non-orbital energy sources such as H/He recombination or accretion jets are absorbed into α_CE) of the orbital energy ΔE_orb released as the giant's core (M_c) and the companion (M₂) spiral inward from initial separation a_i to final separation a_f:

E_bind = α_CE ΔE_orb

Using the de Kool (1990) structural binding-energy parameter λ (where λ ≈ 0.05–1.5 depends on the giant's central concentration and whether internal thermal and ionization/recombination energy U_int is included) to write E_bind = −(G M₁ M_env) / (λ R₁), the energy balance becomes:

( G M₁ M_env ) / ( λ R₁ ) = α_CE [ ( G M_c M₂ ) / ( 2 a_f ) − ( G M₁ M₂ ) / ( 2 a_i ) ]

Because a_f ≪ a_i, the initial orbital energy term is small compared to the final term, yielding the direct scaling for the post-common-envelope orbital separation a_f:

a_f / a_i ≈ ( α_CE λ / 2 ) ( M_c M₂ / M₁ M_env ) ( R₁ / a_i )

If the calculated final Roche-lobe radius around either M_c or M₂ at a_f is smaller than its physical radius, envelope ejection fails and the two stars merge.

2. The Angular-Momentum Formalism (γ Prescription; Nelemans et al. 2000)

To explain observed double white dwarf binaries whose first mass-transfer episode appeared to occur without dramatic orbital shrinkage, Gijs Nelemans and colleagues (2000) introduced the γ formalism, which assumes that the specific angular momentum ΔJ / ΔM_env carried away by the ejected envelope equals a constant multiple γ (typically γ ≈ 1.5–1.75) of the initial specific orbital angular momentum J_i / (M₁ + M₂):

( J_i − J_f ) / J_i = γ [ M_env / ( M₁ + M₂ ) ]

For decades after Paczyński (1976), evidence for common envelope evolution was indirect, resting on the large population of short-period post-CE binaries and the discovery that at least 20% of all planetary nebulae host close binary central stars with orbital periods P_orb < 3 days whose orbital axes are aligned with the symmetry axes of their bipolar nebular lobes (Miszalski et al., 2009).

In 2008, direct real-time observational proof of common-envelope inspiral and merger was captured by the OGLE (Optical Gravitational Lensing Experiment) survey in the galactic transient V1309 Scorpii (Tylenda et al., 2011). Archival photometry covering six years prior to V1309 Sco's September 2008 Luminous Red Nova outburst (L_peak ≈ 3 × 10⁴ L_☉) revealed an eclipsing contact binary with an initial orbital period of P_orb = 1.44 days whose period decayed rapidly according to P(t) ∝ exp(−t / τ) as the stars lost corotation via the Darwin instability, shed an equatorial outflow through the L₂ point, and plunged into a shared common envelope to merge before our eyes.

Why is common envelope evolution necessary to form gravitational-wave mergers detected by LIGO and Virgo?

For two stellar-mass black holes (each ~10–30 M_☉) or two neutron stars (~1.4 M_☉) to merge via gravitational-wave emission within the 13.8-billion-year age of the universe (according to the Peters 1964 quadrupole decay formula t_GW ∝ a⁴), their orbital separation at the birth of the second compact remnant must be smaller than a_f ≲ 5 to 40 R_☉. Yet their high-mass red-supergiant progenitor stars expanded to radii of 500 to 1,000 R_☉. In the standard isolated-binary evolution channel, after the first star collapses into a black hole or neutron star, the second star expands as a supergiant and engulfs the first compact object in a common envelope, shrinking the orbit by a factor of >100 and stripping the hydrogen envelope to leave a tight Wolf-Rayet + BH/NS binary that subsequently collapses into a merging double compact object.

What is the role of hydrogen and helium recombination energy in ejecting a common envelope?

As the inspiraling binary injects orbital energy into the deep layers of the giant's envelope, the gas expands and cools below ~10⁵ K and ~10⁴ K. At these temperatures, doubly ionized helium (He²⁺ → He⁺ → He⁰) and ionized hydrogen (H⁺ → H⁰, followed by H + H → H₂) recombine with free electrons, releasing latent ionization energy of ~15–20 eV per nucleon—comparable in magnitude to the gravitational binding energy of the outer envelope of an AGB star, helping unbind the envelope during Phase III.

What happens if the inspiraling companion fails to eject the common envelope?

If the orbital energy released during inspiral is insufficient to unbind the donor's envelope (a_f drops below the tidal disruption or Roche limit of the core or companion), the two stars coalesce in a common-envelope stellar merger. Observationally, the ejection of the shocked outer layers during coalescence produces a multi-month optical/infrared transient known as a Luminous Red Nova (such as V1309 Scorpii or V838 Monocerotis), leaving behind a rapidly rotating single giant (an FK Comae star), a rejuvenated blue straggler, or—if the engulfed companion is a neutron star—a Thorne-Żytkow object.

1. Paczyński, B. (1976). "Common envelope binaries." In P. Eggleton, S. Mitton, & J. Whelan (Eds.), Structure and Evolution of Close Binary Systems (IAU Symposium No. 73, pp. 75–80). D. Reidel Publishing. https://doi.org/10.1007/978-94-010-1483-0_8
2. Webbink, R. F. (1984). "Double white dwarfs as progenitors of R Coronae Borealis stars and Type I supernovae." The Astrophysical Journal, 277, 355–360. https://doi.org/10.1086/161701
3. Tylenda, R., Hajduk, M., Kamiński, T., Udalski, A., Soszyński, I., Szymański, M. K., et al. (2011). "V1309 Scorpii: Merger of a contact binary." Astronomy & Astrophysics, 528, A114. https://doi.org/10.1051/0004-6361/201016221
4. Ivanova, N., Justham, S., Chen, X., De Marco, O., Fryer, C. L., Gaburov, E., et al. (2013). "Common envelope evolution: Where we stand and how we can move forward." The Astronomy and Astrophysics Review, 21(1), 59. https://doi.org/10.1007/s00159-013-0059-2
5. Röpke, F. K., & De Marco, O. (2023). "Simulations of common-envelope evolution in binary stellar systems: Physical models and numerical techniques." Living Reviews in Computational Astrophysics, 9(1), 2. https://doi.org/10.1007/s41115-023-00017-x