Jet Cocoon: The Overpressured Bubble Inflated by a Relativistic Jet

What a jet cocoon is and why it forms

A relativistic jet is a narrow, ultra-fast (Lorentz factor Γ up to hundreds) stream of plasma launched by a compact central engine — an accreting black hole or a rapidly spinning magnetar. When that jet must plow through dense surrounding matter, a fundamental mismatch arises: the jet beam moves at nearly the speed of light, but the jet head — where it rams the ambient gas — is slowed dramatically because it must sweep up and accelerate that gas.

Two shocks bracket the head: a forward shock driven into the ambient medium and a reverse shock that decelerates the incoming jet. Between them sits shocked, hot plasma at enormous pressure. Because the head advances slowly, this plasma cannot pile up indefinitely — it is squeezed sideways, escaping laterally and spreading backward around the jet. The accumulated shocked jet material plus shocked ambient gas forms the cocoon.

• Overpressured: its pressure vastly exceeds the surrounding medium.
• Confining: that pressure presses inward on the jet, collimating it.
• Two-component: an inner light shocked-jet part and an outer heavy shocked-ambient part, often separated by a contact discontinuity.

The mechanism: head velocity, collimation, and the L̃ parameter

The physics is set by a pressure balance at the jet head. The jet's momentum flux (its ram pressure) pushes forward; the ambient medium's inertia pushes back. Equating them across the head gives the classic result for the head velocity, most cleanly framed by Bromberg, Nakar, Piran & Sari (2011) through a single dimensionless parameter:

L̃ = L_j / (Σ_j ρ_a c^3) ≈ ρ_j h_j Γ_j^2 / ρ_a,

the ratio of the jet's energy density to the ambient rest-mass energy density (Σ_j is the head's cross-section, ρ_a the ambient density). The dimensionless head velocity follows:

β_h ≈ β_j / (1 + L̃^(-1/2)).

• When L̃ ≪ 1 the head is non-relativistic and the jet is strongly collimated by its cocoon.
• When 1 ≪ L̃ ≪ θ_0^(-4/3) the head is relativistic but still collimated.
• When L̃ ≫ θ_0^(-4/3) the cocoon can no longer confine the jet and it stays uncollimated.

The cocoon pressure that does the collimating is fixed by the sideways-flowing energy: p_c ≈ [L_j /(Σ_j c)]^(1/2) (ρ_a c^2 / β_h)^(1/2) (Matzner 2003), and the cocoon drives its own bow shock into the star at velocity β_c set by p_c ≈ ρ_a c^2 β_c^2.

Characteristic numbers and a worked example

Take a canonical long-GRB collapsar: a jet of luminosity L_j ≈ 10^50 erg/s boring through a Wolf–Rayet star of radius R ≈ 10^11 cm and mass ~10 M_sun, so the mean envelope density is ρ_a ≈ 3M/(4πR^3) ≈ 10 g/cm^3.

• For a jet opening angle θ_0 ≈ 0.1 rad and head radius r ≈ θ_0 z, one finds L̃ ~ 0.01–1 deep inside the star — squarely the collimated, sub-relativistic-head regime.
• The head crawls at β_h ~ 0.1–0.3, so crossing the star takes t ≈ R / (β_h c) ≈ few seconds — comparable to observed long-GRB durations, which is not a coincidence.
• Cocoon pressure reaches p_c ~ 10^18–10^20 erg/cm^3, and the shocked plasma is heated to keV–MeV temperatures.
• The energy stored in the cocoon is a sizeable fraction of the jet energy injected before breakout, ~10^49–10^50 erg.

For a neutron-star merger, swap in ejecta of ~0.01 M_sun expanding at ~0.2 c over ~10^10 cm. The jet must catch and pierce this moving screen within ~1 s or it is choked; the cocoon it inflates then carries the energy outward and breaks out as a quasi-thermal flash.

How cocoons are observed: shock breakout, thermal flashes, and radio lobes

Cocoons announce themselves in several ways. When the cocoon's leading shock reaches the surface of the star or ejecta where the optical depth drops to ~c/v, the trapped photons escape in a cocoon shock breakout — a brief, quasi-thermal burst of X-rays/gamma-rays. After breakout the hot cocoon expands and cools, radiating a cooling-envelope / cocoon emission in UV–optical over hours to days, plus later non-thermal radio and X-rays as it decelerates against the circumstellar medium.

The landmark case is GW170817 (2017), the first binary neutron-star merger with an electromagnetic counterpart. The associated short GRB 170817A was ~1000× underluminous compared with normal short GRBs. Gottlieb, Nakar, Piran & Hotokezaka and colleagues showed that a cocoon shock breakout from the merger ejecta — not the narrow jet pointed at Earth — naturally produced those weak gamma rays. Later VLBI showed superluminal motion, confirming a successful structured jet emerged from within the cocoon.

• FR II radio galaxies: the cocoon is the luminous radio lobe, fed by backflow from the hotspots.
• Microquasars & TDE jets: smaller cocoons inflate bubbles in the host gas.

Successful jets, choked jets, and related structures

Whether a cocoon becomes the whole show or just a sheath depends on whether the jet breaks out before the engine shuts off:

• Successful jet: the jet pierces the envelope, emerges with a fast core surrounded by the slower, wide-angle cocoon. Off-axis observers see mostly cocoon/structured-jet emission; on-axis observers see a classic GRB. This is the favored picture for GW170817.
• Choked (failed) jet: the jet stalls inside the star or ejecta and dumps all its energy into the cocoon. No collimated GRB escapes, but the cocoon still breaks out — producing low-luminosity GRBs, X-ray flashes, and fast blue optical transients.

The cocoon is a cousin of, but distinct from, several structures. It is not the accretion disk or the jet itself; it is the waste-heat bubble around the jet. It differs from a supernova shell (spherical, thermal-pressure driven) in being aspherical and jet-driven. In the structured-jet picture, the angular energy profile E(θ) — a bright core plus power-law wings — is essentially the fossil imprint of jet–cocoon mixing. Radio-galaxy cocoons (Scheuer 1974; Begelman & Cioffi 1989) are the same physics scaled up by ~10 orders of magnitude and dominated by synchrotron rather than thermal emission.

Significance, open questions, and famous cases

The jet cocoon reframed how we read high-energy transients. It explains why long-GRB durations (~seconds) match the jet-crossing time of a Wolf–Rayet star, why some GRBs are underluminous, and why merger counterparts can be seen even when the ultra-narrow jet misses Earth. It links AGN feedback (radio lobes heating cluster gas) to stellar-death physics through one shared mechanism.

Key milestones: the collapsar cocoon framework of MacFadyen & Woosley (1999) and Matzner (2003); the unified analytic theory of Bromberg et al. (2011); and the GW170817 cocoon interpretation of Gottlieb, Nakar & Piran (2018), tested against the multi-messenger afterglow.

Open questions remain:

• Magnetization: real jets are Poynting-flux dominated; how magnetic fields alter cocoon formation, mixing, and stability (kink instabilities) is actively debated.
• Mixing: the degree of baryon loading from jet–cocoon mixing sets the emergent Lorentz factor and the emission — hard to resolve even in 3D simulations.
• Neutrinos: choked jets in cocoons are candidate sources of high-energy IceCube neutrinos, still unconfirmed.
• Fast optical transients: how much of the AT2018cow-like fast-blue-transient population is cocoon breakout.

Jet cocoons across scales: collapsar (long GRB), NS-merger (short GRB), and FR II radio galaxy regimes.

Property	Collapsar / long GRB	NS-merger / short GRB	FR II radio galaxy
Central engine	Black hole / magnetar in a collapsed Wolf–Rayet star	Hypermassive NS or BH after binary NS merger	Supermassive black hole (10^8–10^9 M_sun)
Ambient medium	Stellar envelope, ρ_a ~ 1–100 g/cm^3	Merger ejecta, ~0.01 M_sun, v ~ 0.1–0.3 c	Intracluster/IGM gas, ρ_a ~ 10^-27 g/cm^3
Jet luminosity L_j	~10^50–10^51 erg/s	~10^49–10^51 erg/s	~10^44–10^47 erg/s (kinetic)
Cocoon size at breakout	~10^10–10^11 cm (stellar radius)	~10^10–10^11 cm (ejecta radius)	~10–100 kpc (radio lobes)
Cocoon temperature/energy	keV–MeV thermal photons	keV thermal shock-breakout photons	Relativistic e-, synchrotron GHz radio
Breakout / lifetime	~1–10 s to cross the star	~1–2 s after merger	~10^7–10^8 yr lobe lifetime