A mineral sputter model in agreement with solar wind ion irradiation experiments

The sputtering of material is mostly modeled using Binary Collision Approximation programs. Several advances were made in the last few years focusing on modeling mineral sputtering by ion impacts relevant for rocky planets exposed to solar wind. The most recent contribution, from Biber et al. [1], includes not only sputter yields, but also angular distribution data for the mineral enstatite. The existing data, although scarce, are important to validate mineral sputter simulations. A widely applicable model is integral for obtaining and interpreting information of particles ejected from exposed rocky bodies such as Mercury and the Moon. Moreover, ease of use is crucial whenever a new approach is proposed, which is to compete with the default model found in the user-friendly, but inaccurate TRIM code [2]. 

To best recreate experimental data from mineral sputtering, previously suggested approaches rely on increased surface binding energies as well as increased sample densities [3,4]. We review the capabilities and limitations of these and propose a new model to best approximate experimental results. In contrast to the earlier models, our approach achieves unprecedented agreement with available experimental data under normal incidence (Fig. 1). It thereby does not require any manual adjustments of simulation parameters to achieve realistic mineral densities and does not depend on computationally intensive determination of species-specific surface binding energies [4].

The new model considers a surface binding energy for species leaving the sample as well as a bulk binding energy within the sample based on the enthalpy of formation. The latter prevents long collision cascades due to energy loss in the sample whenever a bond of a mineral-forming compound (i.e., an oxide or sulfide) is broken. Favoring short collision cascades leads to a more prominent forward-tilt of the ejecta distribution as it is seen in experiments. The increased energy loss within the sample also causes a peak broadening in the energy distribution of ejected particles whilst shifting the peak positions slightly towards larger energies. We expect to see this behavior on oxygen-bearing minerals as the same tendencies were observed in energy distributions of irradiated oxidized metals [5,6,7]. While we wait for further experimental data our improved quantitative formulation of the mineral sputter process is a valuable contribution for achieving state of the art exosphere models for the Moon and Mercury.

Fig. 1: Sputter yield of various models in SDTrimSP compared to TRIM and experimental data [1]. Short forms: SB — surface binding energies (default); BB — bulk binding energies; SBB-C — combined SB and BB model, differentiating bound and free atoms within predefined compounds. 

 

[1] Biber, H., et al. (2022). Planet. Sci. J., 3, 271.

[2] Hobler, G. (2013). Nucl. Instrum. Methods Phys. Res. B, 303, 165–169.

[3] Szabo, P.S., et al. (2020). Astrophys. J., 891(1), 100.

[4] Morrissey, L. S., et al. (2022). Astrophys J. Lett., 925(1), L6.

[5] Dullni, E. (1984). Nucl. Instrum. Methods Phys. Res. B, 2(1–3), 610–613.

[6] Wucher, A., & Oechsner, H. (1986). Nucl. Instrum. Methods Phys. Res. B, 18(1–6), 458–463.

[7] Wucher, A., & Oechsner, H. (1988). Surf. Sci., 199(3), 567–578.