Protoplanetary Disk

Anatomy of a disk

A protoplanetary disk is built by physics older than any star: angular-momentum conservation during gravitational collapse. The parent molecular cloud core rotates slowly — perhaps once every million years — but as it shrinks under gravity, conservation of angular momentum forces material in the equatorial plane to spin faster and faster. The result is a rotating, flared, geometrically thin disk: vertical scale height H much less than radius R, but H/R ~ 0.05-0.1 at typical radii because thermal pressure puffs the disk a few percent in the vertical direction.

Within the disk, conditions vary radically with radius:

• Inner edge (< 0.1 AU). Stellar magnetic field truncates the disk; gas falls onto the star along field lines (magnetospheric accretion). Temperatures ~1500 K destroy refractory grains.
• Terrestrial-planet zone (0.1-3 AU). 500-1500 K, mostly silicate dust and iron, no ices. Birthplace of rocky planets.
• Water snow line (~3-5 AU for Sun-like). Temperature 170 K. Beyond it water ice condenses, tripling solid mass — the inflection point for forming Jupiter-class cores.
• Giant-planet zone (5-30 AU). 50-150 K, abundant ices, optimum for ~10 M_⊕ cores that capture H/He envelopes.
• CO snow line (~30 AU for Sun-like). Temperature 30 K. Beyond it CO ice condenses, locking carbon in solids.
• Outer disk (30-500 AU). 10-50 K, ice-coated grains, low-density gas. Birthplace of Kuiper-belt-like icy bodies and possibly the occasional super-Neptune.

Worked example — ALMA's resolution on HL Tau

The HL Tau image is the field's signature plate. ALMA in long-baseline configuration achieves angular resolution θ ~ 30 milliarcseconds at the relevant wavelength (~1.3 mm), where 1 mas = 4.85 × 10^-9 rad. At HL Tau's distance d = 140 pc = 4.32 × 10¹⁸ m:

Linear resolution Δx = d · θ
                        = 140 pc · (30 × 10⁻³ arcsec / 206265 arcsec/rad)
                        = 4.32 × 10¹⁸ m · (1.45 × 10⁻⁷)
                        ≈ 6.27 × 10¹¹ m
                        ≈ 4.2 AU  (~5 AU per beam)

That is comparable to the orbital separation of Jupiter and Saturn (5.2 and 9.5 AU). For the first time, an observatory had resolved structure inside a protoplanetary disk on the scale on which planets actually form. Bright rings were measured at ~14, 32, 67 AU and gaps at ~14, 33, 50 AU. To produce gaps that sharp, hydrodynamic simulations require a planet of at least Saturn mass at each gap radius — though alternatives (dust trap induced by gas-pressure bumps from non-ideal magnetohydrodynamics, dead-zone edges) remain viable for some of the features.

The minimum-mass solar nebula and disk mass

Hayashi (1981) reconstructed a minimum-mass solar nebula (MMSN) by augmenting each terrestrial planet's solid mass back to solar composition and spreading the gas over annuli. The result: a disk surface density

Σ(r) ≈ 1700 g/cm² × (r / 1 AU)⁻³⃗²

Integrated from 0.4 to 36 AU this gives ~0.01 M_⊙ — the smallest amount of material from which our solar system could have been assembled. ALMA has now measured similar surface density profiles in dozens of disks, finding masses 0.001-0.1 M_⊙ — a wide range with the median sitting at a few times the MMSN. The implication: most disks contain enough material to build several Jupiter-mass planets, and the bottleneck on planet formation is not mass but efficiency of growth.

From dust to planets — the growth chain

Stage	Size scale	Mechanism	Timescale	Bottleneck
Interstellar grains	0.01-1 μm	Inherited from molecular cloud	—	None — given
Sticky aggregates	1 μm - 1 mm	Brownian motion, turbulent encounters	10³ - 10⁴ yr	Bouncing barrier (mm)
Pebbles	1 mm - 10 cm	Coagulation in pressure bumps	10³ - 10⁴ yr	Radial drift "meter barrier"
Planetesimals	1 - 100 km	Streaming instability + gravitational collapse	10⁴ - 10⁵ yr	Need overdense clump
Protoplanets	100 - 6000 km	Runaway and oligarchic accretion	10⁵ - 10⁶ yr	Isolation mass
Gas giants	10 - 30 R⊕	Core accretion of H/He envelope above 10 M⊕	10⁶ yr	Disk dispersal clock
Final architecture	Solar-system scale	Late-stage migration and giant impacts	10⁶ - 10⁸ yr	Disk gone — gravity only

The most stubborn problem in the chain is the "meter barrier": meter-sized rocks drift radially inward at ~30 m/s, falling onto the star before they can grow further. Streaming instability — the spontaneous collective collapse of pebble-laden gas streams — is the leading resolution, jumping growth directly from millimeter pebbles to kilometer planetesimals in a single dynamical instability.

A timeline of the field

• 1796. Laplace formalises the nebular hypothesis: the solar system formed from a rotating gas cloud.
• 1945. Carl von Weizsäcker proposes turbulent eddies as the mechanism for planet formation.
• 1970. Safronov publishes Evolution of the Protoplanetary Cloud and Formation of the Earth and Planets, establishing modern planetesimal theory.
• 1981. Hayashi minimum-mass solar nebula model defines the reference disk.
• 1984. First direct imaging of disk material (around Beta Pictoris, debris disk) by Smith and Terrile.
• 1995. 51 Peg b discovery proves planets form efficiently — and migrate inward.
• 2003. HST images dozens of "proplyds" in the Orion Nebula — protoplanetary disks photoionised by O-star UV.
• 2014. ALMA HL Tau image — the moment that made the field visual.
• 2018. PDS 70 b detected accreting in Hα by VLT/SPHERE — a planet caught in the act of forming.
• 2020. Disk Substructures at High Angular Resolution Project (DSHARP) ALMA survey resolves 20 nearby disks; ring/gap structure is the norm, not the exception.
• 2024. JWST resolves the inner disk geometry and chemistry; complex organic molecules (e.g. CH₃CN, methanol) detected in TW Hya inner disk.

How disks die

• Accretion onto the star. Typical Ṁ ~ 10^-8-10^-7 M_⊙/yr; over 5 Myr accretes ~0.005-0.05 M_⊙ — comparable to the total disk mass.
• Photoevaporation. Stellar EUV/FUV/X-ray photons heat the disk surface; gas above the local escape speed flows away. Disk lifetime is short for stars with high X-ray luminosity.
• Magnetorotational disk winds. Magnetic-field-launched winds remove mass and angular momentum from the disk surface independent of viscous accretion.
• Locked into planets. Once material is incorporated into kilometer-plus bodies it no longer counts as disk mass.
• External photoevaporation. In OB associations, neighboring massive stars can dramatically shorten lifetimes — some Orion proplyds disperse in ~1 Myr instead of 10.

Why protoplanetary disks matter

• Planet origin. Every known exoplanet condensed from a disk like these. Disk statistics constrain planet-formation theory.
• Disk-planet coevolution. Forming planets shape the disk (gaps, spirals); the disk shapes the planets (migration, atmospheric capture). Both must be modelled jointly.
• Solar system origins. Meteorite ages and isotopes anchor our own disk's history; ALMA images other disks give an outside view of the same process.
• Astrochemistry. Complex organic molecules — precursors to amino acids and the building blocks of life — form in cold outer disk regions and are inherited by comets and planetesimals.
• Initial mass functions. Disk fragmentation in massive disks may produce binary companions and brown dwarfs without going through stellar collapse channels.

Common misconceptions

• "Disks are flat." Geometrically yes (H/R ~ 5%), but optically and chemically they have layered structure: a hot ionised surface, a warm intermediate layer where CO and HCN line emission dominates, a cold dense midplane where dust settles and planets form.
• "All disks have visible rings." Only with ALMA resolution. Many disks look smooth at lower resolution; the rings emerge when you achieve few-AU beam sizes.
• "Planets form quickly throughout the disk." No — growth is fast in some bands (terrestrial zone, gas-giant zone) and slow in others. The dichotomy in our solar system between rocky inner planets and gas-giant outer planets reflects this.
• "The disk and the star are the same age." The disk forms within ~10⁵ yr after the protostar; both age together. But disks dissipate before the star reaches the main sequence, so disk age < stellar pre-main-sequence age.
• "Every disk forms planets." Probably most do — the Kepler mission found that the average star hosts ~1 planet per stellar mass — but planet formation efficiency is not 100%, and some disks disperse before producing visible planets.

Open questions

• Are the HL Tau gaps caused by planets? Most plausibly yes, but alternative explanations (gas-pressure bumps from MHD effects, dust drift) remain viable for some gaps. Resolution requires direct planet detection in each gap.
• Why is the disk mass distribution so wide? Two orders of magnitude spread in 1-Myr disks at fixed stellar mass — what sets the initial reservoir?
• How efficient is gas accretion? Once a core reaches ~10 M_⊕, gas runaway is supposed to take ~10⁵ yr; observational evidence for this short phase is sparse.
• What is the role of episodic accretion? FU Orionis outbursts deliver ~0.01 M_⊙ in decades. How much of the final stellar mass comes from these episodes versus steady accretion?