Dead Zones: Where the MRI Switches Off in Protoplanetary Disks

What a Dead Zone Is: The Physical Basis

A protoplanetary disk accretes onto its young star because something transports angular momentum outward. In well-ionized disks that something is the magnetorotational instability (MRI), discovered as a disk process by Balbus & Hawley (1991): a weak magnetic field threading differentially rotating gas becomes unstable, amplifying into vigorous turbulence that behaves like a viscosity.

The catch is that the MRI needs the magnetic field and the gas to be coupled — the field can only stir the gas if the gas is ionized enough to feel it. In the cold, dense interior of a disk beyond ~1 AU, temperatures fall below ~1000 K, thermal ionization shuts off, and the ionization fraction plummets to as low as 10⁻¹⁴. There, the field slips freely through the neutral gas (high magnetic resistivity), the MRI cannot grow, and turbulence dies.

• The surface layers stay ionized by external cosmic rays and stellar X-rays and remain MRI-active.
• The shielded midplane — the dead zone — is magnetically decoupled and laminar.

This vertically stratified picture is the layered accretion model proposed by Charles Gammie in 1996.

The Mechanism: Ionization, Shielding, and Non-Ideal MHD

Whether the MRI operates at a given location is a competition between ionization (which couples field to gas) and recombination + magnetic diffusion (which decouples them). Non-thermal ionizing agents each penetrate only so far:

• Stellar X-rays penetrate a column of ~1–10 g/cm².
• Galactic cosmic rays penetrate ~100 g/cm² before being absorbed.
• Radioactive ²⁶Al decay provides a weak floor of ionization everywhere.

Because a minimum-mass disk has surface densities of hundreds to thousands of g/cm² near 1 AU, only the outer ~100 g/cm² on each face is ionized — the rest is dead. The coupling is quantified by non-ideal MHD terms in the generalized Ohm's law: Ohmic resistivity (dominant at high density, low ionization — the midplane), ambipolar diffusion (dominant in the tenuous outer/upper disk), and the Hall effect (important in between). MRI activity roughly requires the Elsasser number Λ = v_A²/(Ω·η) ≳ 1 and a magnetic Reynolds number Rm = Ω·H²/η ≳ 10². Where η (the diffusivity) is large, Λ and Rm fall below unity and the MRI is quenched.

Key Quantities and a Worked Estimate

Consider a solar-mass star with a disk at r = 1 AU. The orbital angular frequency is Ω = √(GM/r³) ≈ 2×10⁻⁷ s⁻¹ (a 1-year period). With midplane temperature T ≈ 280 K, the sound speed c_s ≈ 1 km/s gives a scale height H = c_s/Ω ≈ 0.05 AU.

• Surface density (MMSN): Σ ≈ 1700 g/cm² at 1 AU.
• Active column each face: ~100 g/cm² (cosmic-ray depth), so the dead-zone column ≈ 1700 − 200 ≈ 1500 g/cm² — roughly 90% of the mass is dead.
• Ionization fraction in the dead zone: n_e/n ~ 10⁻¹³ to 10⁻¹⁴.
• Effective viscosity: active α ~ 10⁻² versus dead α ≲ 10⁻⁴ — two or more orders of magnitude quieter.

The outer edge of the dead zone sits where the whole column becomes ionizable — typically ~10–30 AU, depending on how much small dust is present, since grains sweep up free electrons and shrink the active region. The inner edge sits near ~0.1–1 AU, where T rises past ~1000 K and thermal ionization of alkali metals (potassium, sodium) reactivates the MRI.

How It's Detected and Where It Appears

Dead zones are not seen directly — they are inferred from their consequences. The strongest modern evidence comes from ALMA, whose high-resolution dust images (starting with HL Tau in 2014) reveal concentric rings and gaps. A sharp jump in turbulence at a dead-zone edge produces a pressure bump that halts the inward drift of mm-sized dust, piling it into a bright ring — exactly the kind of dust trap needed to beat the notorious radial-drift barrier.

• Low turbulent line broadening: ALMA molecular-line measurements in disks like TW Hya and HD 163296 find turbulent velocities well below the sound speed (α ≲ 10⁻³), consistent with weak midplane turbulence.
• Accretion variability: episodic outbursts (see below) trace material piling up in a dead zone.

Dead zones are expected in essentially all T Tauri and Herbig Ae/Be disks during the planet-forming epoch (the first few million years), which is why they sit at the center of modern planet-formation theory.

How It Differs from Related Regimes

It's easy to confuse a dead zone with other quiet-disk ideas — but they are distinct:

• Dead zone vs. fully MRI-active disk: An active disk is turbulent top to bottom; a dead zone has an inert middle sandwiched by active skins (layered accretion).
• Dead zone vs. gravitationally stable laminar disk: A disk can be laminar simply because it's low-mass; a dead zone is specifically magnetically laminar despite having field and shear available.
• Dead zone vs. magnetized disk winds: Recent work argues much of a disk's accretion may actually be driven by magnetocentrifugal winds launched from the surface rather than MRI turbulence — the wind can carry angular momentum even over a dead midplane. This has shifted the debate but not eliminated the dead zone.
• Dead zone vs. the habitable-zone: Despite the shared word 'zone,' the habitable zone is a temperature band for liquid water on planets — unrelated to disk magnetism.

The Hall effect adds a twist: whether the field is parallel or antiparallel to the rotation axis changes the dead zone's structure, a genuine and unresolved subtlety.

Significance, Famous Cases, and Open Questions

Dead zones matter because they solve problems that plague planet formation. Their quiet, low-turbulence midplanes let dust settle into a thin layer, concentrate at pressure bumps, and reach densities where the streaming instability and gravitational collapse can build kilometer-scale planetesimals on ~10⁵-year timescales — turning micron dust into planet embryos.

The most dramatic observational signature is episodic accretion. In the classic FU Orionis picture, gas delivered by the active layers piles up in the dead zone until it becomes massive enough to go gravitationally unstable or thermally reignite the MRI; the trapped material then dumps onto the star, brightening it 100-fold for decades before the cycle resets. FU Ori itself (which brightened in 1936 and stayed bright) is the prototype.

Open questions:

• How large is the dead zone really, given uncertain dust abundance and cosmic-ray shielding by stellar/disk winds?
• Do MRI turbulence or magnetized winds dominate accretion — and does the dead zone accrete at all?
• How does the poorly-constrained Hall effect reshape the zone's boundaries?

Active surface layers vs. the dead midplane in a layered protoplanetary disk

Property	Active surface layer	Dead zone (midplane)
Ionization source	Cosmic rays, stellar X-rays, UV, radionuclides	Only radioactive decay (26Al) reaches deepest gas
Ionization fraction (n_e/n)	~10⁻¹² and up	as low as ~10⁻¹⁴ to 10⁻¹³
MRI status	Active — sustains turbulence	Suppressed — laminar flow
Effective viscosity α	~10⁻² to 10⁻³	~10⁻⁵ to 10⁻⁴ (near-zero)
Column shielded	outer ~100 g/cm² per face	everything below the active columns
Role in planet formation	drives accretion onto the star	traps dust, grows planetesimals