Scientific Premise

A handful of ultra-short-period planets (USPs) are known to be actively losing mass, producing asymmetric, variable-depth transits caused by a comet-like tail of sublimated rock dust (e.g., KIC 12557548 b from Kepler, and the recently discovered nearest disintegrating planet from TESS). These "disintegrating planets" are literally exposing their interior composition to spectroscopic study — the dust tail carries mineral grains whose wavelength-dependent scattering/absorption signatures encode whether the mantle is iron-rich, silicate-dominated, or enriched in refractory species like corundum or perovskite. Yet only a few individual systems have been modeled in detail, and no systematic survey has attempted to classify the interior mineralogy of all known disintegrating planets using a common forward-modeling framework.

Target Datasets

• TESS light curves (multi-sector, for transit depth variability and color dependence from FFI vs. 20-s cadence)
• Kepler long-cadence data for KIC 12557548 b (4 years of variable transits)
• JWST Cycle 1–4 public archive: any NIRSpec, MIRI, or NIRCam time-series observations of USPs
• Spitzer 3.6 and 4.5 μm archival secondary eclipse and transit photometry for USPs

Novelty

Most existing work on disintegrating planets focuses on dynamical modeling of the mass-loss rate or on reproducing the asymmetric transit shape. The mineralogical angle — using multi-wavelength transit depths to constrain the dust composition and hence mantle geochemistry of rocky exoplanets — is potentially underexplored at the population level. This project would build a Mie-scattering + radiative transfer forward model for the dust tail and apply it uniformly to every known (and candidate) disintegrating system, producing the first comparative mineralogical study of exoplanetary mantles.

Concrete Workflow

1. Compile all confirmed and candidate disintegrating-planet systems (currently ~5–8 objects) from the literature and from a targeted search of TESS data for variable-depth, asymmetric transits around USPs.
2. For each system, extract multi-band transit depths: optical (TESS/Kepler), near-IR (JWST if available), mid-IR (Spitzer archival).
3. Build a parametric dust-tail model: particle size distribution (power law), composition (mixtures of olivine, pyroxene, iron, corundum, etc.), and optical depth profile.
4. Use Mie theory to compute wavelength-dependent extinction and fit the multi-band transit depths simultaneously.
5. Derive posterior distributions on grain composition fractions; compare across systems and to Solar System meteorite mineralogies.

Expected Signal / Observable

Wavelength-dependent transit depth variations of ~0.01–0.1% between optical and mid-IR bands, depending on grain size and composition. Iron-rich grains produce a flatter extinction curve; silicate grains show a pronounced ~10 μm feature in MIRI data. Even with only broadband photometry, the optical-to-IR depth ratio constrains the mean grain size and Fe/Si ratio.

Possible False Positives

• Stellar limb-darkening variations between bands can mimic chromatic transit depth changes. Mitigation: use consistent limb-darkening models (e.g., from ExoTiC-LD) and fit for limb-darkening coefficients.
• Variability from starspots produces epoch-to-epoch depth changes that are not dust-related. Mitigation: use contemporaneous out-of-transit flux monitoring to detrend spot modulation.
• Grain size degeneracy with composition in broadband data. Mitigation: use JWST spectroscopy where available; otherwise, clearly report the degeneracy.

Why It Matters

We have no direct measurement of the mantle composition of any rocky exoplanet. Disintegrating planets offer a unique natural experiment — literally ablating their interiors into observable space. A comparative mineralogical survey could reveal whether rocky planet mantles are universally Earth-like or show significant diversity, with implications for planetary formation, differentiation, and ultimately habitability.