Polluted white dwarfs reveal exotic mantle rock types on exoplanets in our solar neighborhood

Polluted white dwarfs (PWDs)—white dwarfs whose otherwise H- or He-dominated atmospheres are contaminated by heavy elements from accreted planetary debris—offer a near-direct probe of exoplanetary compositions because heavier elements sink rapidly unless they are being replenished. As evolved stars disrupt close-orbiting planets or planetesimals that cross the Roche limit, their debris is accreted and detected spectroscopically. While astronomy often labels rocky bodies as Earth-like, the inner Solar System planets (Mercury, Earth, Moon, Mars) differ markedly in bulk and silicate composition. Recent claims suggested that some PWDs contain signatures of continental (granitic) crust—potentially a unique indicator of plate tectonics and global water cycles—based largely on limited elemental indicators (e.g., Ca, Al, Li/Na). This study asks whether the full set of rock-forming elements that define rock types (Mg, Si, Ca, Fe) in PWDs actually supports the presence of continental or other crustal compositions, and seeks to identify the rock types (mantle versus crust) represented by accreted materials.

Earlier work established that a substantial fraction of white dwarfs show pollution consistent with rocky bodies and that accretion may preferentially sample silicate-rich material due to tidal resistance differences between metallic and silicate phases. Several recent studies proposed continental crust signatures on dozens of PWDs based on enrichments in Ca, Al, and certain alkali ratios, but these did not include Si—critical for distinguishing felsic continental crust where SiO2 is typically 60–75 wt%. Catalogs of nearby FGKM stars (e.g., Hypatia) have been used to infer possible exoplanet compositions from stellar abundances, suggesting a range but not the extremes reported for some PWDs. Comparisons to meteorites (chondrites, achondrites including ureilites) and to crustal and mantle rocks from Earth, Moon, and Mars provide a context for assessing whether PWD bulk and silicate compositions match known Solar System materials.

The study analyzes 23 polluted white dwarfs for which Mg, Si, Ca, and Fe are measured with reported uncertainties. Bulk (core + mantle + crust) compositions inferred from the stellar photospheres are compared against Solar System inner planets (Earth, Moon, Mars, Mercury) by considering their bulk silicate compositions and adding back metallic cores where necessary to enable like-for-like comparisons. To isolate silicate fractions, the authors compute bulk silicate planet (BSP = mantle + crust) compositions for each PWD by correcting the bulk for core formation assuming Earth-like partitioning of Fe between silicate and metal reservoirs (fraction of Fe retained in the mantle aFe ≈ 0.27). Compositions are renormalized to Mg + Si + Ca + Fe = 100% (or as oxides: MgO + SiO2 + CaO + FeO = 100%), with total Fe expressed as FeO for silicate comparisons. PWD bulk and BSP compositions are compared to: (1) inner planet bulk and mantle compositions; (2) meteorite classes (chondrites, achondrites, irons, stony irons); (3) crustal and mantle rock types from Earth, Moon, and Mars; and (4) >4,000 rocky exoplanet silicate compositions inferred from nearby FGKM main-sequence stars in the Hypatia Catalog. Mineralogical interpretations use classic and newly proposed ternary diagrams based on proportions of olivine, orthopyroxene, clinopyroxene, quartz, and periclase to classify mantle rock types and evaluate whether any PWD compositions correspond to crustal (particularly granitic) rocks. Uncertainties are propagated and shown on comparative plots.

• Across 23 PWDs with precise Mg, Si, Ca, Fe, bulk compositions span a much wider range than those inferred from nearby FGKM stars and do not closely match Earth, Moon, or Mars bulk compositions.
• Several PWDs overlap partially with certain meteorite fields (e.g., some achondrites such as ureilites in Mg–Si space), but differences (notably higher CaO in PWDs) indicate imperfect matches; high-Ca PWDs have some overlap with elemental flood basalts but overall PWDs remain distinct.
• Elevated Fe in many PWDs suggests some accretion of metallic core material. After correcting for core formation to derive BSP (silicate) compositions, SiO2 remains too low and MgO too high for any PWD to represent crustal rocks at significant fractions, contradicting prior claims of granitic (continental) crust on PWDs.
• Only one PWD silicate composition approximates bulk silicate Earth; most PWD BSPs match mantle rock types, many of which are not dominant on any inner planet in our Solar System.
• PWD silicate compositions reveal mineralogies that include exotic assemblages (e.g., quartz + orthopyroxene and periclase + olivine) requiring new rock classification schemes; proposed names include quartz pyroxenite, orthopyroxenite variants, periclasite dunite, and others.
• Q-bearing PWD mantles contain too much orthopyroxene (too high MgO) to be continental crust, which would plot at much higher quartz contents.
• The compositional diversity of PWD-derived exoplanets exceeds that of >4,000 FGKM-star-inferred rocky exoplanets, implying their uniqueness is unlikely to be inherited directly from parent star compositions.

The findings directly test and refute previous interpretations that continental (granitic) crust is commonly recorded in polluted white dwarf atmospheres. Using rock-defining elements and core-corrected silicate fractions, the study shows that PWD compositions are inconsistent with granitic or other crustal rocks and instead primarily sample mantle materials. The dominance of exotic mantle mineralogies, rarely (or not) represented among inner Solar System planets, indicates that rocky exoplanets in the solar neighborhood experienced accretion and differentiation pathways distinct from Solar System analogs. Because the PWD compositional spread exceeds that of nearby FGKM stars, stellar composition alone is insufficient to explain the diversity; planetary processes (e.g., degrees of melting, differentiation, metal-silicate segregation, volatile histories) likely dominate the variance. The proposed new mineralogical classification schemes provide a framework to describe and compare these previously unrecognized mantle rock types, facilitating more accurate geological interpretation of exoplanetary debris.

This work presents the first rock-type estimates for exoplanetary material accreted by polluted white dwarfs using rock-defining elements (Mg, Si, Ca, Fe) and core-corrected silicate compositions. Contrary to prior claims, no evidence supports significant fractions of continental or other crustal rocks; instead, most PWDs reflect mantle materials, with only one case approximating bulk silicate Earth. The majority exhibit exotic mineralogies (e.g., quartz + orthopyroxene; periclase + olivine) that lack direct Solar System counterparts, necessitating new classification schemes introduced here. These results reveal a greater diversity of rocky exoplanet compositions and evolutionary histories in the solar neighborhood than previously recognized. Future work should expand samples with high-precision multi-element measurements, refine corrections for accretion and diffusion processes, and integrate additional petrologically diagnostic elements to further constrain exoplanetary rock types.

• Elemental coverage is limited mainly to Mg, Si, Ca, and Fe; lack of comprehensive datasets for Al, Na, K, Ti, and other petrologically important elements constrains rock-type specificity.
• BSP calculations assume Earth-like Fe partitioning between silicate and metal; deviations in exoplanet differentiation could bias core-correction results.
• Polluted photospheres may accrete mixtures of materials (e.g., multiple bodies, varying silicate/metal fractions), complicating straightforward rock-type attribution.
• Comparative frameworks rely on renormalization (Mg+Si+Ca+Fe) and expression of Fe as FeO, which may not capture all redox or volatile effects.
• Sample size (n=23) and selection based on availability of precise measurements may limit generalizability across all PWD systems.