In Gladman (2003), Gladman and Burns (1996), and Gladman et al. (1996), it was determined through numerical simulations that a small flux of Mercurian meteorites to Earth should be possible. They concluded that 30–65% of ejected material would be re-accreted to Mercury in 10 m.y. (75% in 50 m.y.). During that time frame, a portion of the ejected material would make the journey to a Venus-crossing orbit, where 5–10% of that material would be accreted. After many close encounters with Venus, some material would be efficiently delivered to an Earth-crossing orbit. When all is considered, ~0.1% of the material initially ejected from Mercury would accrete to Earth during a time frame of 10 m.y.
In similar studies by Wetherill (1984) and Love and Keil (1995), they calculated that the Earth intercept percentage is lower by a factor of ten; i.e., only 0.01% of the material initially ejected from Mercury would accrete to Earth during a time frame of 2–10 m.y. In contrast, Gladman (1996) determined that 0.5% of Mercurian material would accrete to Earth in a time frame of 50 m.y. Additional results were compiled by Melosh and Tonks (1993), who utilized an orbital evolution model to analyze a statistically significant number of particles. Their model demonstrated that almost all of the Mercurian ejecta was either re-accreted to Mercury in a median time frame of ~30 m.y., or that it was collisionally destroyed, with only a very small fraction of the ejecta impacting Venus. Surprisingly, their model also predicted that ~30% of ejecta from Venus should impact Earth in a median time frame of ~12 m.y., which suggests a higher likelihood of finding a venusian meteorite than a Mercurian meteorite.
The most recent study of the transfer of Mercurian impact ejecta to Venus and Earth or its re-accretion to Mercury has been conducted by Gladman and Coffey (2009). They utilized numerical calculations of particles to simulate impact-generated ejections from Mercury. In this new study it was determined that previous estimations of impacts onto Mercury were too low, and that higher impact speeds of 20–70 km/s are more realistic. Therefore, according to theory, the ejection velocities used for the simulations were in the range of 4–25 km/s, and the particles were followed over a period of 30 m.y. The transfer efficiency of the ejected meteoroids for various ejection velocities which had sizes of a decimeter or larger were derived. Overall, 50% were re-accreted to Mercury in 30 m.y., and 15% more over the suceeding 30 m.y. The cumulative accretion to Venus was 20%. The cumulative accretion to Earth over the same 30 m.y. for all ejection velocities studied was determined to be 2–5%. The lower ejection velocities (4 km/s) and shorter timeframes (10 m.y.) produce efficiencies that are in agreement with previous studies; the increased velocities (up to ~20 km/s) and extended timeframe (30 m.y.) account for the higher accretion efficiency, about half that of Mars efficiencies.
While a lunaite should have an FeO content and a crystallization age similar to a mercurian meteorite, it would also have a high content of implanted solar gases. Notably, many lunaites contain clasts composed of high FeO mare basalt. Optical and near-infrared spectra of Mercury indicate that the regolith is consistent with a composition having a 3:1 ratio of labradorite and enstatite, commensurate with ~1.2 wt% FeO and 0 wt% TiO2 (Warell and Blewett, 2004).
An intriguing theory has been set forth by investigators from the University of Washington in Seattle. They suggest that the angrite group of meteorites may represent material from Mercury that was collisionally stripped from the planet, thereby explaining the chemical and mineralogical differences between Mercury and the angrites; e.g., the higher FeO-abundance of angrites compared to that on the present surface of Mercury, and the reversal of the Fe/Mn ratios for olivine and pyroxene compared to those values measured for other planetary bodies. Nevertheless, even accepting that collisional-stripping of a hypothetical FeO-rich basaltic (angritic) crust on Mercury occurred, Hutson et al. (2007) find it implausible that Mercury initially differentiated under oxidizing conditions to form the angritic crust, and then subsequently differentiated under reducing conditions to form the now-exposed surface that we can observe today. Hutson et al. (2007) have also determined that other mineralogical features identified in angrites, which were attributed to rapid decompresion on a planetary-sized body such as Mercury, are more consistent with typical cooling processes during crystallization of a melt.
Ongoing studies of newly discovered angrites by investigators at UWS and other institutions has led to revised models describing the possible connection of angrites to Mercury, such as the occurrence of metasomatic processes to explain the disequilibrium textures observed. Metasomatic processes may also be responsible for the inferred oxidizing conditions under which angrites formed, as evidenced by the association of rhönite and ferric iron in NWA 4590, and by other Fe-metal–oxide associations found to exist in some other angrites. Furthermore, the UWS investigators have modeled the petrogenesis of the various angrite lithologies, the masses of which could now reside in the asteroid belt, or perhaps may still orbit around the original collisionally-stripped parent body (Mercury?).
Other studies which relate to a possible angrite–Mercury connection have been conducted. For example, in a study of siderophile element depletions (Ni, Co, and W) and their association to core segregation of the angrite parent body, K. Righter (2008) proposed a scenario which would lead to the observed siderophile element depletions observed in angrites. The model is consistent with a small differentiated asteroid having a mantle only a few hundred km in radius surrounding a small metallic core. His model employs the conditions of a reduced periotite mantle and FeNi-metal core which attained metal–silicate equilibrium under very low pressures. The oxidation state of material derived from this hypothetical body, consistent with that observed in some angrites, could have occurred later, possibly during the eruption phase, through C-assisted redox processes.
In a study of the shock-modified bencubbinite Fountain Hills, and the CB group in general, Weisberg and Ebel (2009) discussed the severe impact-related characteristics of this meteorite and other CB members as they show evidence for a major collision early in the history of the CB parent body. They revealed an abundance of evidence which indicates that Fountain Hills experienced impact shock forces greater than those observed for any other chondritic body, and they contrast this severe impact with the hypothesized collision on Mercury that is considered to have stripped away much of its original mantle. They argue that the general characteristics of bencubbinites, i.e., metal-rich, refractory-rich, and volatile-depleted, are consistent with its formation in the innermost solar system, possibly near the orbit of Mercury. Furthermore, they contend that the bulk composition of bencubbinites shows some similarities to Mercury as well. It may be speculated that bencubbinites formed from a vapor cloud that was produced by a massive collisional impact on Mercury. A possible link between this carbonaceous group and Mercury will be the subject of future investigations through data gathered by the MESSENGER (MErcury Surface, Space ENvironment, GEochemistry, and Ranging spacecraft) spacecraft.
Although the transfer dynamics of ejecta from Mercury to Earth are more complex than delivery from Mars, the probability of finding a Mercurian meteorite may be higher than was once thought. Perhaps several are already in the worldwide collections but have not yet been recognized as such. Indeed, it may even be possible to find venerian (Venus) meteorites and terrene/terran (Earth) meteorites, but their recovery is less likely because of other constraining factors such as a high escape velocity, a thick atmosphere, and a short (m.y.) time frame for re-accretion.