The radial drift paradox at small radii
For decades, theoretical models of planetary formation have wrestled with the speed at which dust should vanish from circumstellar disks. In a typical protoplanetary disk of gas and dust orbiting a young star, the gas experiences slight outward radial pressure support, causing it to orbit at marginally sub-Keplerian velocities. Millimetre- and centimetre-sized solid pebbles do not feel this pressure support directly and attempt to orbit at full Keplerian speeds. The resulting headwind saps orbital energy from the pebbles, causing them to migrate rapidly inward towards the central star. In an unperturbed, smooth disk, this process—known as radial drift—should drain solid matter into the central star on timescales far shorter than the typical multimillion-year lifetimes of young stellar systems.
Yet observations consistently show that older, Class II protoplanetary disks retain substantial reservoirs of millimetre-sized dust. To explain why this material has not long since plunged into the star, theorists invoke pressure traps: local pressure maxima within the gas disk where the outward pressure gradient reverses or vanishes. In these zones, the aerodynamic headwind drops to zero, collecting dust pebbles into long-lived rings or rims.
Over the past decade, high-resolution observations using the Atacama Large Millimeter/submillimeter Array (ALMA) confirmed this picture in spectacular fashion across large, bright protoplanetary disks spanning tens or hundreds of astronomical units (au). In these sprawling systems, concentric rings, gaps, and spirals are nearly universal. However, these giant systems represent an observational bias. The vast majority of protoplanetary disks in nearby star-forming regions are far smaller, with characteristic dust radii measuring less than 10 au.
Until recently, astronomers could not determine whether these compact disks harboured the same structural features or whether their small sizes reflected fundamentally different evolutionary pathways. If compact disks lacked dust traps, their persistent millimetre emission remained an unresolved paradox. If they did possess them, their substructures had simply been hidden by the limits of telescope resolution.
Pushing ALMA to two-astronomical-unit resolution
Resolving structural features inside a disk whose entire outer boundary spans less than 10 au pushes modern radio interferometry to its absolute limit. In a study published in The Astrophysical Journal Letters, an international team led by Pietro Curone targeted 11 of the smallest Class II protoplanetary disks in the nearby Lupus star-forming region. Each selected target possessed an effective dust radius (enclosing 68 percent of the millimetre flux) of less than 10 au.
The researchers used ALMA Band 7 to measure continuum emission at a wavelength of 0.9 millimetres. By exploiting ALMA’s most extended baseline configurations, the campaign reached an angular resolution of 0.02 arcseconds. At the distance of the Lupus molecular cloud, this corresponds to a spatial scale of just 2 au—the highest angular resolution currently achievable with ALMA at this submillimetre wavelength.
The observations revealed that compact disks are far from featureless clumps. Out of the 11 surveyed systems, four exhibited unambiguous substructures, while an additional three displayed tentative evidence of structural variations. Among the confirmed structures was a clearly defined central cavity with a radius of only 1.6 au in one disk, placing an evacuated inner hole directly within the terrestrial planet-forming zone of that star. In two other disks, the high-resolution images uncovered faint, delicate outer rings encircling bright, compact inner cores.
To contextualise these detections, the team integrated their sample into broader demographic data from the wider corpus of Class II disks observed with ALMA. To ensure fair statistical comparison, they restricted their analysis to disks whose physical diameters were resolved by at least three resolution elements across the beam.
The resulting demographic pattern is telling. Among disks with diameters greater than 30 au that meet this resolution threshold, virtually all of them—at or above 95 percent—show discernible substructures. In the intermediate-to-compact regime spanning diameters between 3 and 30 au, approximately 50 percent exhibit confirmed substructures. While that detection rate is lower than in the giants, it is substantial, particularly given the severe observational hurdles that remain when inspecting compact systems.
What gap opening implies for planet populations
The presence of narrow gaps and evacuated cavities at orbital radii between 1 and 5 au provides crucial constraints on the mechanisms sculpting young solar systems. While several hydrodynamical processes can theoretically form rings—including dead zones in magnetic turbulence or condensation fronts known as snowlines—the most compelling explanation for narrow, isolated gaps remains tidal clearing by newborn planets embedded within the gas.
Because the gravitational torque required to clear an annular gap scales with orbital radius and local disk properties, opening a gap very close to a star requires substantially less planetary mass than carving a wide trough tens of astronomical units out. In the distant, outer regions of extended disks, opening gaps typically demands massive gas giants comparable to Jupiter or Saturn. Within 5 au, however, the team’s dynamical calculations show that opening observable gaps requires planets with masses falling in the super-Earth to mini-Neptune regime.
This mass scale aligns closely with the known demographics of mature exoplanetary architectures. Exoplanet transit and radial velocity surveys have demonstrated that super-Earths and sub-Neptunes on compact orbits are the most common planets in the Milky Way, far outnumbering wide-orbit gas giants. The ALMA detections indicate that the architectural foundations for these ubiquitous small planets are laid down early, within the first few million years of a stellar system’s life.
Nevertheless, the findings must be interpreted with careful attention to physical limitations. Detecting substructure in roughly half of the resolved compact disks does not imply that the remaining half lack features. In compact disks, dust densities are exceptionally high, making the inner regions optically thick at submillimetre wavelengths. If dust emission is completely optically thick, radiation from interior structures cannot escape cleanly, washing out the brightness contrast between gaps and rings. Consequently, high optical depth almost certainly masks additional rings and gaps in the survey sample.
Furthermore, ALMA’s 2 au resolution element, while the current peak of millimetre astronomy, remains broad compared to the scale of inner terrestrial orbits. Features narrower than 1 to 2 au cannot be deblended from the background disk light. Probing the formation zones of Mercury- or Earth-analogue orbits will require even longer baselines or shorter observing wavelengths than currently available.
Sources
- Probing Planet Formation within 10 au: Dust Substructures in Extremely Compact Protoplanetary Disks, Pietro Curone et al., arXiv; published in ApJL 1010, L24 (2026), 2026-10-14
- Publisher record (DOI)
The R&D takeaway
For research planners and scientific funding consortia, these findings demonstrate that high-resolution instrumentation designed around large baseline extensions can overturn demographic assumptions driven by observational selection bias. Expanding our understanding of planet formation will require sustained investment in extreme-baseline submillimetre interferometry and next-generation facilities capable of penetrating high optical depths at sub-astronomical-unit scales. Research and development roadmaps for astronomical facilities should prioritise angular resolution over raw collecting area when the target science goals involve terrestrial and inner-disk architectures.
The R&D Innovate desk