Skin-deep difference between Ryugu and Bennu driven by the microphysical evolution of surface materials

Introduction
Samples returned from asteroids Ryugu and Bennu exhibit mineralogical and isotopic similarities to CI carbonaceous chondrites [1–3]. Although CIs are among the leading candidates for delivering the building blocks of life to terrestrial planets [4], their distribution in the solar system remains poorly understood. This is largely due to the featureless nature of primitive asteroids, as well as terrestrial weathering in meteorites. With the availability of pristine spectra from the returned samples, we are now better positioned to identify CI-like asteroids among the large population of primitive asteroids observed in the past decades.

However, remote sensing of Ryugu and Bennu have revealed that space weathering effects may be more complex on primitive asteroids than previously assumed, and its mechanism needs to be better understood to infer the bulk asteroid composition from telescopic data. Despite their compositional and physical similarities, Ryugu and Bennu exhibit differences in their visible spectra. More intriguingly, their visible spectra likely evolved in opposite directions. On Ryugu, fresher craters are brighter and bluer [5, 6], whereas on Bennu, fresher craters are darker and redder [7]. Notably, the spectra of the freshest materials on both asteroids are similar, suggesting a similar initial state. These observations imply that compositionally similar asteroids can undergo divergent spectral evolution, leading to differences in their global spectra [8].

In this study, we investigate the causes of the opposing spectral evolutions of Ryugu and Bennu by integrating spectral analyses of returned samples with remote-sensing measurements of crater spectrophotometry and thermal inertia.

Methods
We measured multi-band visible spectra of over 400 grains and aggregates from Ryugu samples and approximately 660 mg of Bennu aggregate samples at the JAXA curation facility.

We analyzed the visible spectra and phase function slopes of 119 craters on Ryugu based on ONC-T data and 568 craters on Bennu based on MapCam/PolyCam data. The phase function slope was determined by fitting all available photometric data using a Hapke model [9] and calculating the ratio of radiance factors observed under incidence angle (i), emission angle (e), and phase angle (α) of i = 5°, e = 0°, α = 5°, and i = 15°, e = 0°, α = 15°. Thermal inertia values for the craters were obtained from previous studies [10–12].

Results
Limited influence from solar wind and micrometeorites: Our spectral analyses of returned samples suggest that the opposing evolutions of visible spectra may not be fully explained by classical space weathering mechanisms involving solar wind irradiation and micrometeorite bombardment. Although morphological and compositional signatures of these processes—including microcraters, amorphization, and dehydration—have been identified in Ryugu samples [13, 14], the average visible spectra of weathered grains are not statistically distinguishable from those of less weathered grains. Furthermore, although previous studies suggested that solar wind irradiation can exert opposite spectral effects when the hydration state of the material differs [15, 16], the comparable phyllosilicate abundance and bulk hydrogen content of Ryugu and Bennu samples argue against this hypothesis.

Spectral evolution driven by the evolution of surface microphysical properties: We propose that changes in surface microphysical properties —grain size, porosity, and roughness— through thermal fatigue, impacts, and electrostatic levitation may be a more important mechanism. Remote-sensing data show that fresher craters on Ryugu have brighter radiance factor, bluer spectral slope, higher thermal inertia, and shallower phase function slope. All of these trends are inverted on Bennu. Such coevolution of visible spectra with phase function slope and thermal inertia, which are known indicators of microphysical properties, suggests that the opposite evolution of microphysical properties may have driven the opposite spectral evolutions on Ryugu and Bennu.

These results are further validated by the significant spectral change observed after minor powder adhesion on Ryugu grains. A millimeter-sized Ryugu grain (C0179) showed a bluer spectral slope than the remote-sensing spectra after removing the powders adhered to the surface by brushing and high-pressure gas. In contrast, the powders (<50 µm) produced from the same grain exhibited spectral slopes more than twice as red as the remote-sensing spectrum. Remarkably, covering just ~10% of the surface area of the powder-free grain with these fine powders reproduced the global spectrum. This result indicates a direct causal link between microphysical properties and spectra. Nevertheless, the powder-free grain remains bluer than the remote-sensing spectrum of Bennu. This suggests that additional factors—such as a potentially higher abundance of phosphates on Bennu [17]—may also have a non-negligible contribution.

Discussion
Model calculation of grain-size evolution on asteroid surfaces [18] suggests that the difference in asteroid size may be one influential factor causing the opposite microphysical evolutions. On small bodies, electrostatic lofting is the primary mechanism controlling fine grain abundance. Due to a twofold difference in escape velocity, the maximum loftable grain size is ~60 µm for Ryugu and ~32 µm for Bennu. Consequently, fine grains on the order of tens of microns tend to accumulate on Ryugu’s surface but are preferentially lost from Bennu.

Such a mechanism provides the first comprehensive model for explaining the spectral discrepancy revealed by remote sensing and the compositional similarity from sample analyses of Ryugu and Bennu. This further implies that asteroids composed of CI-like materials likely do not correspond to a single spectral type, but may instead be distributed across C, Cb, and B types, due to their diverse microphysical evolution.

References
[1] Yokoyama et al. (2022). Science, 379(6634).
[2] Greenwood et al. (2023). NatAstron, 7(1).
[3] Lauretta et al. (2024). MAPS, 59(9).
[4] Alenxander. (2017). Philos Trans A Math Phys Eng Sci, 375(2094).
[5] Sugita et al. (2019). Science, 364(6437).
[6] Morota et al. (2020). Science, 368(6491)
[7] DellaGiustina et al. (2020). Science, 370(6517).
[8] Yumoto et al. (2024). Icarus, 420.
[9]Hapke (2012). Cambridge university press.
[10] Shimaki et al. (2020). Icarus, 348.
[11] Rozitis et al. (2020). Science advances, 6(41).
[12] Rozitis et al. (2022). JGR: Planets, 127(6).
[13] Noguchi et al. (2023). NatAstron, 7(2).
[14] Matsumoto et al. (2024). Science advances, 10(3).
[15] Lantz et al. (2017). Icarus, 285.
[16] Clark et al. (2023). Icarus, 400.
[17] Milliken et al. (2025). 56^th LPSC, #1331.
[18] Hsu et al. (2022). NatAstron, 6(9).