**Origin of HED Meteorites from the Spalling of Mercury – Implications for the Formation and Composition of the Inner Planets**

Anne M. Hofmeister and Robert E. Criss *Department of Earth and Planetary Sciences, Washington University in St Louis, MO, USA* 

## **1. Introduction**

22 Will-be-set-by-IN-TECH

152 New Achievements in Geoscience

Gustafson Bas, (1994). Physics of zodiacal dust. *Annual Rev. Earth and Planet. Science* 22,

Luethi D. et al. (2008). High–resolution carbon dioxide concentration record 650 000–800 000

Mayewski Paul A. et al. (2004). Holocene climate variability. *Quaternary Research* 62, 243-255. Milankovitch M.M. (1941). Kanon der Erdbestrahlung und seine Anwendung auf das

The Microlensing Observations in Astrophysics (MOA) Collaboration & The Optical

Muller R.A. & MacDonald G.J. (1995). Glacial cycles and orbital inclination. *Nature* 377,

Muller R.A. & MacDonald G.J. (1997). Glacial cycles and astronomical forcing. *Science* 277,

Nufer R., Baltensperger W. & Woelfli W. (1999). Long term behaviour of a hypothetical planet in a highly eccentric orbit. URL http://xxx.lanl.gov/abs/astro-ph/9909464 Orlova L.A. et al. (2001). Chronology and environment of woolly mammoth (Mammuthus

Schirrmeister L. et al. (2002). Paleoenvironmental and paleoclimatic records from permafrost

Tiedemann R., Sarntheim M. & Shackleton N.J. (1994). Astronomic time scale for the Pliocene

Woelfli W.; Nufer R. & Baltensperger W. (2002). An additional planet as a model for the

Woelfli W. & Baltensperger W. (2007a). Arctic East Siberia had a lower latitude in the

Woelfli W. & Baltensperger W. (2007b). On the change of latitude of Arctic East Siberia at the

Pleistocene. *Anais da Academia Brasileira de Ciências,* 79(2): 183–193.

Seager R. (2006). The Source of Europe's Mild Climate. *American Scientist* 94, 334–341.

Snowball bibliography, URL: www.snowballearth.org/bibliography.html

Pleistocene Ice Age. URL: arxiv.org/abs/physics/0204004

end of the Pleistocene. URL: arxiv.org/abs/0704.2489

Eiszeitproblem. *Königliche Serbische Akademie,* Spez. Publikation 133, 1–633, Belgrad.

Gravitational Lensing Experiment (OGLE) Collaboration, (2011). Unbound or distant planetary mass population detected by gravitational microlensing. *Nature* 473,

primigenius Blumenbach) extinction in northern Asia. *The World of Elephants,*

deposits in the Arctic region of Northern Siberia. *Quaternary International* 1998,

Atlantic 18O and dust flux records of Ocean Drilling site 659. *Paleoceanography* 9,

Imbrie J.J., (1984). *Milankovitch and Climate,* Part 1, (eds. A. L. Berger et al.), pp. 269–305. Kashiwaya K, Ochiai S, Sakai H, Kawai T, (2001). Orbit-related long-term climate cycles revealed in a 12-Myr continental record from Lake Baikal. *Nature* 410, 71–74.

Klein F. & Sommerfeld A. (1910). Über die Theorie des Kreisels. *Teubner,* Leipzig.

years before present. *Nature* 453, 379–382.

International Congress, Rome, 718–721.

Petit J.R. et al. (1999). *Nature,* 399, 429–436.

552–554.

349–352.

107–108.

215–218.

97–118.

619–638.

The bulk chemical composition of the Earth is poorly constrained because samples are limited to only the outer 10% of its radius. Meteorite data are used to represent the huge zones that we cannot sample, but for this approach to provide a realistic model of the Earth, we need to know where various meteorites originated and what processing they underwent. Currently popular models of the Solar System presume that the Sun formed first, followed by assembly of the planets from mineral dust in a disk that condensed locally from hot gas (e.g., Boss, 1998). More recent findings that comets are mixtures of phases formed at high and low temperature (Zolensky et al., 2006) came as a great surprise to planetary scientists because this contraindicates the existence of a condensation gradient as proposed by Lewis (1974). The alternative to dust condensation around a pre-formed Sun is a nebula containing dust inherited from older generations of stars. Meteorites possess small amounts of isotopically distinctive pre-solar material, proving that the nebula included some dust that predates our Sun (e.g., Bernatowicz & Zinner, 1997). Recent detection of silicate dust in molecular clouds (van Breeman et al., 2011), which are loci of star formation, suggests that pre-solar dust was abundant. Disk models also have problems conserving angular momentum (Armitage, 2011) and fail to explain the first order characteristics of the Solar System (upright axial spin and nearly circular orbits). Thus, independent evidence points to formation of planets and Sun simultaneously from a 3-dimensional dusty nebula (Hofmeister & Criss, 2012), in which case, virtually all meteoritic material has been processed. What represents the bulk Earth?

It is generally accepted that chondritic meteorites are very primitive material, and bulk compositions of the silicate Earth have been derived from various averages. Models based on enstatite chondrites (Javoy, 1995; Lodders, 2000), which have Earth-like oxygen isotopes and Fe-metal content, agree with heat flux measurements (Hofmeister & Criss, 2005). Newly measured flux of neutrinos from within the Earth (Gando et al., 2011) support the enstatite chondrite model. However, because even the most primitive meteorites show some evidence of alteration and are comprised of non-equilibrium assemblages (e.g., Brearley & Jones, 1998), bulk compositions so derived are estimates, not exact representations of the Earth. Hence, it is worthwhile to consider alternative means of inferring bulk composition.

Origin of HED Meteorites from the Spalling of Mercury –

asteroids are cousin fragments of a much larger parent body.

**1.1.1 Is Vesta a mesosiderite? Was Vesta highly impacted?** 

strength. The lumpy shape resembles that of iron-rich meteorites.

**1.2 Is early Mercury the source of HEDs and related meteorites?** 

in an impact is evidenced by the Caloris basin, and its antipodal structure.

modeling are covered in Sections 1.2 and 3.

Implications for the Formation and Composition of the Inner Planets 155

Attribution of HED meteorites to Vesta and a few spectrally similar asteroids, despite weak evidence for orthopyroxenes or olivine (e.g., Burbine et al., 2001), sets the view that this tiny object is a major parent body. Spectral variations do not require differentiation: 1) dust on the surface could be non-uniform in grain-size, distribution, or mineral proportions, 2) Vesta could be a remnant from a larger body, or 3) Vesta could be a conglomerate, as are many meteorites. Section 2 shows that Vesta is too small to sort phases by density in its own

The Vesta hypothesis is counter-evidenced by the abundance of HEDs, which outnumber all other achondrites combined (Table I). Assignment of the HEDs to Vesta further requires a few asteroids provide IIIAB irons, and most mesosiderites and pallasites, which constitute a huge proportion of the collection (Table I). It is far more likely that Vesta and similar

Dynamical arguments that were set forth against proposals of Mars as the source of SNC meteorites (e.g., Vickery & Melosh, 1983) were refuted by subsequent isotopic and chemical evidence (summarized by Grady, 2006). Recognition of lunar and Martian meteorites, coupled with Mercury being of intermediate size to these parent bodies led Love & Kiel (1995) to suggest that some Mercurian samples may reside in available collections. Although impact calculations have been extensively revised to explain the existence of lunar and Martian meteorites, the view is nevertheless still held that dynamics forbid Mercury as a meteorite source (e.g., Melosh & Tonks, 1993). Clearly, dynamical arguments can neither prove nor preclude planetary ejection, let alone establish transport. Problems with such

Low orbital eccentricity of ~0.09 for Vesta and Ceres, compared to an average of 0.26 for 25 large asteroids which is similar to those of Pallas and Juno (http://nssc.gsfc.an.gov), indicates that Vesta was not more heavily impacted. We propose that Vesta is a mesosiderite with ~20% iron, based on its density of 3.9 g cm-3, and spectra (Section 4), and represents below of 1% of the mass ejected from Mercury. High iron content provides material

Mercury's unique characteristics are its high orbital eccentricity (0.2056 compared to 0.0067- 0.094 for the other seven planets), the large size of its Fe core compared to its mantle, and its similarity to the Moon, with craters upon craters, ejecta, impact basins, and smooth plains that are probably volcanic (e.g., Spudis & Guest, 1988). Delivery of immense kinetic energy

Various hypotheses attempt to explain Mercury's special characteristics. Attribution of Mercury's Fe rich composition to high-temperature nebular condensates being located near the Sun (Lewis, 1974) cannot be the cause because condensation temperatures of Fe0 and Mg2SiO4 (forsterite) only differ by a few degrees (Lodders, 2003) which precludes their sorting during condensation. Attribution of Fe-enrichment to vaporization of its surface (Cameron et al., 1988) does not address high orbital eccentricity and requires excessively

gravitational field. Section 4 shows that grain-size variations explain Vesta's spectra.

The present paper considers how more processed meteorites (the differentiated achondrites) may be related to the Earth. We demonstrate that the largest class of achondrites (the HEDs) has been wrongly assigned to an asteroidal parent body: a planet is required. The likely source is evolving Mercury, which suffered severe impacts that ejected the majority of its mantle to space (Section 1.2). This connection permits us to decipher information about the Earth during core formation, a very early and important step in its history which is not documented in terrestrial samples, and to estimate bulk planetary composition.

## **1.1 Problems in assigning HED meteorites to the asteroid Vesta**

Many different classes of achondrites are recognized, but most of these classes contain few meteorites (e.g., Mittlefehldt et al., 1998, summarized in Table 1). The largest single class of achondrites, the HEDs (= howardite, eucrite, and diogenite) is a distinctive petrographic family whose affinity has been confirmed by oxygen isotopes (e.g., Taylor et al., 1965). Eucrites are similar to pigeonite-plagioclase basalts and basaltic cumulates, while diogenites are similar to orthopyroxenites, and howardites are brecciated mixtures of these two types (e.g., Dymek et al., 1976; Shearer et al., 1997). It is further recognized that mesosiderites, pallasites, and IIIAB iron meteorites are geochemically similar to each other and to HEDs (e.g., Duke & Silver, 1967; Clayton & Mayeda, 1996). Section 5 provides further discussion.


\* Most contain chondrules (spheres of once melted material, typically olivine+pyroxene+feldspar rim or of forsterite+enstatite+Fe0), calcium-aluminum-rich inclusions, and/or pre-solar grains.

† Virtually all carbonaceous chondrite mass is in the Allende meteorite (3000 kg).

‡ Part of the HED family, sensu lato.

§ Of the remaining achondrites, 298 are C-rich ureilites but 67 aubrites (related to enstatite chondrites) have 4300 kg of the mass. Brachinites and angrites have chemical affinities to HEDs (Section 5).

Table 1. Summary of meteorite types with simple descriptions. Density from Britt and Consolmagno (2003). Number and mass from *The Meteoritical Bulletin* database (www.lpi.usra.edu/meteor/metbull.php, accessed 9/26/2011).

Sourcing the HEDs to the 3rd largest asteroid, Vesta, is based on spectral comparisons (e.g., Gaffey, 1997). Slight variations in spectra across its surface were interpreted as differing proportions of pigeonite to other phases. Detection of a basaltic mineral was taken as evidence that the asteroid had differentiated, and thus has a core (e.g., Ruzika et al., 1997).

The present paper considers how more processed meteorites (the differentiated achondrites) may be related to the Earth. We demonstrate that the largest class of achondrites (the HEDs) has been wrongly assigned to an asteroidal parent body: a planet is required. The likely source is evolving Mercury, which suffered severe impacts that ejected the majority of its mantle to space (Section 1.2). This connection permits us to decipher information about the Earth during core formation, a very early and important step in its history which is not

Many different classes of achondrites are recognized, but most of these classes contain few meteorites (e.g., Mittlefehldt et al., 1998, summarized in Table 1). The largest single class of achondrites, the HEDs (= howardite, eucrite, and diogenite) is a distinctive petrographic family whose affinity has been confirmed by oxygen isotopes (e.g., Taylor et al., 1965). Eucrites are similar to pigeonite-plagioclase basalts and basaltic cumulates, while diogenites are similar to orthopyroxenites, and howardites are brecciated mixtures of these two types (e.g., Dymek et al., 1976; Shearer et al., 1997). It is further recognized that mesosiderites, pallasites, and IIIAB iron meteorites are geochemically similar to each other and to HEDs (e.g., Duke & Silver, 1967; Clayton & Mayeda, 1996). Section 5 provides further discussion.

**Main type Description Type Mineralogy ρ, g cm-3 Number Total kg**  Chondrites unmelted stones\* ordinary pyroxene, olivine 3.3 36326 43000 carbonaceous hydrated, C 2.1-3.5 1385 3281† enstatite ~MgSiO3, albite 3.6 506 850 Achondrites melted stones HEDs‡ basaltic, see text 2.9-3.3 1033‡ 1250‡ Martian pigeonite, olivine 3.2 99 100

Stony- irons Metal+ silicates mesosiderites‡ HED minerals 4.2 170‡ 6300‡ pallasites‡ olivine 4.8 89‡ 13100‡ Irons Fe with<20% Ni non-magmatic most have low Ni 7.5 280 150000

\* Most contain chondrules (spheres of once melted material, typically olivine+pyroxene+feldspar rim or

§ Of the remaining achondrites, 298 are C-rich ureilites but 67 aubrites (related to enstatite chondrites) have 4300 kg of the mass. Brachinites and angrites have chemical affinities to HEDs (Section 5). Table 1. Summary of meteorite types with simple descriptions. Density from Britt and Consolmagno (2003). Number and mass from *The Meteoritical Bulletin* database

Sourcing the HEDs to the 3rd largest asteroid, Vesta, is based on spectral comparisons (e.g., Gaffey, 1997). Slight variations in spectra across its surface were interpreted as differing proportions of pigeonite to other phases. Detection of a basaltic mineral was taken as evidence that the asteroid had differentiated, and thus has a core (e.g., Ruzika et al., 1997).

of forsterite+enstatite+Fe0), calcium-aluminum-rich inclusions, and/or pre-solar grains. † Virtually all carbonaceous chondrite mass is in the Allende meteorite (3000 kg).

(www.lpi.usra.edu/meteor/metbull.php, accessed 9/26/2011).

‡ Part of the HED family, sensu lato.

Lunar anorthite, basalt 2.7, 3.8 146 60 other§ as above, fassaite ~3 576§ 4700§

magmatic IIIAB‡ most have low Ni 7.5 288‡ 136000‡ other magmatic variable Ni 7.5 485 164000

documented in terrestrial samples, and to estimate bulk planetary composition.

**1.1 Problems in assigning HED meteorites to the asteroid Vesta** 

Attribution of HED meteorites to Vesta and a few spectrally similar asteroids, despite weak evidence for orthopyroxenes or olivine (e.g., Burbine et al., 2001), sets the view that this tiny object is a major parent body. Spectral variations do not require differentiation: 1) dust on the surface could be non-uniform in grain-size, distribution, or mineral proportions, 2) Vesta could be a remnant from a larger body, or 3) Vesta could be a conglomerate, as are many meteorites. Section 2 shows that Vesta is too small to sort phases by density in its own gravitational field. Section 4 shows that grain-size variations explain Vesta's spectra.

The Vesta hypothesis is counter-evidenced by the abundance of HEDs, which outnumber all other achondrites combined (Table I). Assignment of the HEDs to Vesta further requires a few asteroids provide IIIAB irons, and most mesosiderites and pallasites, which constitute a huge proportion of the collection (Table I). It is far more likely that Vesta and similar asteroids are cousin fragments of a much larger parent body.

Dynamical arguments that were set forth against proposals of Mars as the source of SNC meteorites (e.g., Vickery & Melosh, 1983) were refuted by subsequent isotopic and chemical evidence (summarized by Grady, 2006). Recognition of lunar and Martian meteorites, coupled with Mercury being of intermediate size to these parent bodies led Love & Kiel (1995) to suggest that some Mercurian samples may reside in available collections. Although impact calculations have been extensively revised to explain the existence of lunar and Martian meteorites, the view is nevertheless still held that dynamics forbid Mercury as a meteorite source (e.g., Melosh & Tonks, 1993). Clearly, dynamical arguments can neither prove nor preclude planetary ejection, let alone establish transport. Problems with such modeling are covered in Sections 1.2 and 3.

## **1.1.1 Is Vesta a mesosiderite? Was Vesta highly impacted?**

Low orbital eccentricity of ~0.09 for Vesta and Ceres, compared to an average of 0.26 for 25 large asteroids which is similar to those of Pallas and Juno (http://nssc.gsfc.an.gov), indicates that Vesta was not more heavily impacted. We propose that Vesta is a mesosiderite with ~20% iron, based on its density of 3.9 g cm-3, and spectra (Section 4), and represents below of 1% of the mass ejected from Mercury. High iron content provides material strength. The lumpy shape resembles that of iron-rich meteorites.

## **1.2 Is early Mercury the source of HEDs and related meteorites?**

Mercury's unique characteristics are its high orbital eccentricity (0.2056 compared to 0.0067- 0.094 for the other seven planets), the large size of its Fe core compared to its mantle, and its similarity to the Moon, with craters upon craters, ejecta, impact basins, and smooth plains that are probably volcanic (e.g., Spudis & Guest, 1988). Delivery of immense kinetic energy in an impact is evidenced by the Caloris basin, and its antipodal structure.

Various hypotheses attempt to explain Mercury's special characteristics. Attribution of Mercury's Fe rich composition to high-temperature nebular condensates being located near the Sun (Lewis, 1974) cannot be the cause because condensation temperatures of Fe0 and Mg2SiO4 (forsterite) only differ by a few degrees (Lodders, 2003) which precludes their sorting during condensation. Attribution of Fe-enrichment to vaporization of its surface (Cameron et al., 1988) does not address high orbital eccentricity and requires excessively

Origin of HED Meteorites from the Spalling of Mercury –

2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4 Log (Radius, km)

Io

**r min,rocks ~1200 km**

> Moon J=0.394

> > icy Jovian moons

values needed for core formation. Data from Lodders & Fegley (1998).

**2.2 A minimum planetary size is required for gravitational settling** 

Core size correlates with both *g* and *R* (Fig. 2) because the latter are related through:

where G is the gravitational constant, and is the mean density. Forming a planetary core involves sorting of different density phases in a gravitational field. We therefore focus our discussion on gravitational acceleration. Because iron and silicate melts are immiscible (e.g., Goodrich and Bird, 1985) and density of liquids and crystals generally correlate, core formation can be viewed as sinking of Fe0 particles or blobs within a silicate "fluid" in the hot interior of the planet. Settling is described in terms of a Stokes (terminal) velocity:

where is dynamic viscosity of the "fluid," *d* is particle diameter, and is the density difference between the particle and the fluid. The percentage difference between silicate and ice being similar to that of Fe and silicate and being related to the density (compression) explains why one trend in surface g exists for bodies with different

J=0.366

J=0.378 Mars

0

5

large asteroids

compositions (Fig. 2b).

10

15

20

25

30

a

35

Implications for the Formation and Composition of the Inner Planets 157

~1.2 ms-2 (Fig. 2b). The trends show that an object must have a mass in excess of ~1022 kg to

g

Fig. 2. Dependence of core size, expressed as the ratio of core mass to total planetary mass, on (a) the logarithm of the surface radius and (b) the surface gravitational acceleration. Diamonds = asteroids. Dots (with error bars) = rocky moons. Open circles = icy moons with highly uncertain core sizes. Mercury was not included in the linear regression, although the x-intercept of the fits would change little if it were. Measured moments of inertia (*J*=*I*/*Mr*<sup>2</sup> where *r* is radius and *M* is total mass) are indicated. Bodies with cores have *J*<0.4, which is the value for a sphere of constant density. Dashed vertical arrows emphasize minimum

0

.

0 2 4 6 8 10 Gravitational accelleration, m s -2

g = 4G*R/3* (1)

*v* = *g*d2/(18), (2)

Icy moons b Ceres

Mars

Io

min~1.2 ms -2

Moon

5

10

15

20

25

30

35

Earth

Venus

have a metallic core, which is greater than the mass of the entire asteroid belt.

Earth J=0.3307

> Venus J=0.33

high surface temperatures. It is difficult to reconcile the topography with vaporization. Loss of Mercury's mantle through catastrophic fragmentation following many giant impacts, in which most of the material was reassembled while Mercury's orbit ranged dramatically, including beyond the Earth (Wetherill, 1988) has problems as regards orbital excursions, and incorrectly assumes that the impactor plus Mercury constituted a bound state, see Section 3.

Fig. 1. Mercury's internal structure and surface features. Left = comparison of the terrestrial planets to scale. Surfaces = heavy lines. Metal cores = shaded sections. Data from Lodders & Fegely (1998). Right = photo of a limb of Mercury taken by the Mariner spacecraft. The lower edge shows about 560 km across. From http://history.nasa.gov/EP-177/ch2-2-2.html.

Hofmeister and Criss (2012) recently proposed a different history for Mercury, whereby repeated impacts stripped away the mantle through the late heavy bombardment (LHB) that occurred up to 800 Ma after formation (Gomes et al., 2005). Given the physical characteristics of Mercury, its proximity to the gravitational focus of the Solar System (the Sun), and events like the LHB we propose that the spallation of Mercury before core formation was complete provided HEDs, pallasites, mesosiderites, irons, and others. Minimal addition of kinetic energy is needed to promote ejecta to distant orbits (Section 3). We conclude that the asteroid belt is a junkyard of debris that originated from diverse regions of the Solar System over its history, particularly small bodies (e.g., various moons, Mars, and Mercury as it evolved) and outer Solar System primitive material drawn in, e.g., during the LHB. Oxygen isotope data implicate early Mercury as the parent body of the HEDs (Section 5) and lead to bulk chemical compositions for Earth and its core (Section 6).

## **2. Did Vesta differentiate under its own gravity?**

This section compares characteristics of Vesta to those of bodies with known cores. The observed trends are explained using Stokes' law and stability arguments.

### **2.1 Limitation of known cores to round, large bodies**

Small planets are cold relative to large, due to their high surface area to volume ratio, which provides for greater radiative cooling to space, and shorter distances over which internal heat diffuses. Hence, their potential for differentiation via convection and gravitational segregation is limited. Mass fractions of planetary cores depend on planetary radius (Fig. 2a). Trends of the rocky bodies suggest that a minimum radius of ~1200 km is required to form a core. Considering icy and rocky bodies for which the moment of inertia is known, differentiation is also associated with a minimum surface gravitational acceleration (*g*) of

high surface temperatures. It is difficult to reconcile the topography with vaporization. Loss of Mercury's mantle through catastrophic fragmentation following many giant impacts, in which most of the material was reassembled while Mercury's orbit ranged dramatically, including beyond the Earth (Wetherill, 1988) has problems as regards orbital excursions, and incorrectly assumes that the impactor plus Mercury constituted a bound state, see Section 3.

**MARS**

Fig. 1. Mercury's internal structure and surface features. Left = comparison of the terrestrial planets to scale. Surfaces = heavy lines. Metal cores = shaded sections. Data from Lodders & Fegely (1998). Right = photo of a limb of Mercury taken by the Mariner spacecraft. The lower edge shows about 560 km across. From http://history.nasa.gov/EP-177/ch2-2-2.html.

Hofmeister and Criss (2012) recently proposed a different history for Mercury, whereby repeated impacts stripped away the mantle through the late heavy bombardment (LHB) that occurred up to 800 Ma after formation (Gomes et al., 2005). Given the physical characteristics of Mercury, its proximity to the gravitational focus of the Solar System (the Sun), and events like the LHB we propose that the spallation of Mercury before core formation was complete provided HEDs, pallasites, mesosiderites, irons, and others. Minimal addition of kinetic energy is needed to promote ejecta to distant orbits (Section 3). We conclude that the asteroid belt is a junkyard of debris that originated from diverse regions of the Solar System over its history, particularly small bodies (e.g., various moons, Mars, and Mercury as it evolved) and outer Solar System primitive material drawn in, e.g., during the LHB. Oxygen isotope data implicate early Mercury as the parent body of the HEDs (Section 5) and lead to bulk chemical compositions for Earth and its core (Section 6).

This section compares characteristics of Vesta to those of bodies with known cores. The

Small planets are cold relative to large, due to their high surface area to volume ratio, which provides for greater radiative cooling to space, and shorter distances over which internal heat diffuses. Hence, their potential for differentiation via convection and gravitational segregation is limited. Mass fractions of planetary cores depend on planetary radius (Fig. 2a). Trends of the rocky bodies suggest that a minimum radius of ~1200 km is required to form a core. Considering icy and rocky bodies for which the moment of inertia is known, differentiation is also associated with a minimum surface gravitational acceleration (*g*) of

**MERCURY**

6370 km 0 3390 km

**2. Did Vesta differentiate under its own gravity?** 

**2.1 Limitation of known cores to round, large bodies** 

observed trends are explained using Stokes' law and stability arguments.

**VENUS**

**EARTH**

~1.2 ms-2 (Fig. 2b). The trends show that an object must have a mass in excess of ~1022 kg to have a metallic core, which is greater than the mass of the entire asteroid belt.

Fig. 2. Dependence of core size, expressed as the ratio of core mass to total planetary mass, on (a) the logarithm of the surface radius and (b) the surface gravitational acceleration. Diamonds = asteroids. Dots (with error bars) = rocky moons. Open circles = icy moons with highly uncertain core sizes. Mercury was not included in the linear regression, although the x-intercept of the fits would change little if it were. Measured moments of inertia (*J*=*I*/*Mr*<sup>2</sup> where *r* is radius and *M* is total mass) are indicated. Bodies with cores have *J*<0.4, which is the value for a sphere of constant density. Dashed vertical arrows emphasize minimum values needed for core formation. Data from Lodders & Fegley (1998).

#### **2.2 A minimum planetary size is required for gravitational settling**

Core size correlates with both *g* and *R* (Fig. 2) because the latter are related through:

$$\mathbf{g} = 4\pi\wp\mathbf{G}\mathbf{R}\mathbf{\mathcal{B}}\tag{1}$$

where G is the gravitational constant, and is the mean density. Forming a planetary core involves sorting of different density phases in a gravitational field. We therefore focus our discussion on gravitational acceleration. Because iron and silicate melts are immiscible (e.g., Goodrich and Bird, 1985) and density of liquids and crystals generally correlate, core formation can be viewed as sinking of Fe0 particles or blobs within a silicate "fluid" in the hot interior of the planet. Settling is described in terms of a Stokes (terminal) velocity:

$$
v = \gcd \text{pd}^2/(18\eta). \tag{2}$$

where is dynamic viscosity of the "fluid," *d* is particle diameter, and is the density difference between the particle and the fluid. The percentage difference between silicate and ice being similar to that of Fe and silicate and being related to the density (compression) explains why one trend in surface g exists for bodies with different compositions (Fig. 2b).

.

Origin of HED Meteorites from the Spalling of Mercury –

without adding any kinetic energy (Fig. 3).


**3.2.1 Sling-shotting to distant reaches** 

orbital speed of Mercury is 48 km.

smaller, ~105 J.

minimum kinetic energy to be promoted far beyond Venus' orbit.

**3.2.2 Effect of kinetic energy beyond the minimum** 

The kinetic energy of Mercury in its orbit is ~106 J.


0

**A.U.**

0.2

0.4

Implications for the Formation and Composition of the Inner Planets 159

(e.g., Symon, 1971). Hence, an object in a circular orbit of radius *r* from the Sun has the same total energy as an object in extremely eccentric orbits whose distance at aphelion can approach 2*r*. Due to this property, ejecta can be transported to beyond the orbit of Venus

Mercury

Venus

e=0

s-1, a mere 1%

s-1. Because the

**0.206** 0.48 0.66


Fig. 3. Elliptical orbits for objects sharing the orbital energy of Mercury (solid curves, Mercury in bold), compared to the nearly circular orbit of Venus (dashed curve). Labels indicate the eccentricity of the orbits, which vary from a perfect circle (*e* = 0) to nearly unity.

For any ejecta, including those in elliptical orbits sharing the kinetic energy of Mercury's orbit, promotion to even more distant parts of the Solar System is possible. We base this remark on use of Jupiter's gravitational well to "slingshot" the famous Voyager spacecraft into the outer Solar System. If Venus were in a suitable position relative to the trajectories of ejected material in a highly eccentric orbit (*e*>0.86, Fig. 3), it is possible for that material with

If the perturbing collision adds energy to the object being ejected, far greater distances than

increase over existing orbital energy. For transport to infinity, the velocity needed for an

only doubling their kinetic energy. It should be clear that energies arising from impacts

the Sun. Hence, material ejected from an inner planet, if on an appropriate trajectory, has

g-1, because orbital velocities become progressively slower with distance from

s-1, escape of ejecta from the Sun's pull to infinity requires

g-1 whereas that of the asteroids is much

Venus' orbit can be realized. Escape from Mercury itself requires 4.3 km.

object to escape the gravitational well of the Sun at Mercury's orbit is 68 km.

were sufficient to move material from the inner to the outer Solar System.

**A.U.**

0.86 e=0.97

The core grows at a rate comparable to Stokes' velocity. Large planets with large *g* remain hot longer, forming proportionately larger cores (Fig. 2b). For bodies with *g* exceeding ~2 m. s-2, core size directly depends on *g*, as indicted by Eq. 2. Because large planets retain more heat than small, viscosity decreases as planet size increases which makes the dependence of *v* on g weaker than linear. However, Eq. 2 assumes that particles sink and reach a terminal velocity within the static fluid, but does not describe whether sinking is possible. Figure 2 demonstrates that a minimum g is needed for coherent, downward motions.

Asteroids are too small to possess cores. It is immaterial whether asteroidal bodies are mixtures of iron and silicate or of silicate and ice, given similar density contrasts. Because the density contrast among various silicate mineral pairs are an order of magnitude smaller, asteroids could not have sorted olivine from plagioclase. Our conclusion that Vesta and all other asteroids could not differentiate due to internal gravitational sorting does not preclude these being differentiated material. Rather, our analysis shows that the differentiation occurred within a larger body with sufficient mass to permit density sorting.

## **3. Derivation of achondrites from the inner Solar System**

Derivation of meteorites from Mars has been demonstrated (summarized by Grady, 2000). Mercury has g = 3.70 m. s-2, close to that of Mars (3.71 m. s-2) and an escape velocity of 4.3 km. s-1, low compared to Mars (5 km. s-1). Together these characteristics show that ejection of material from Mercury is not merely possible, but expected. Hence, objections to Mercury as a parent body rest on the notion that transport of material outward from the innermost Solar System is prohibited due to the huge gravitational influence of the Sun (e.g., Melosh & Tonks, 1993; summarized by Love & Kiel, 1995). Such difficulties have been overstated. As discussed below in general terms, classical mechanics indicates transport is feasible.

## **3.1 Dynamical constraints**

Calculations concerning ejection of objects from planetary bodies and the consequent orbits are a variant of the three-body problem of classical mechanics. A system of three or more masses, moving under their mutual gravitational forces, cannot be solved in any general way (e.g., Goldstein, 1950; Symon, 1971). For example, the three-body problem cannot be reduced to three one-body problems. Very few special solutions exist: 1) the restricted threebody problem can be solved, wherein one mass is small and does not perturb the motions of the other two large masses (e.g., the trajectory of a rocket from Earth to the Moon); 2) motions of the planets can be accurately calculated, which represent a numerical solution obtained from specified initial conditions that holds over some period of time; and 3) a particular solution exists for three-bodies orbiting about the center of mass (discovered by Lagrange). Dynamical studies of ejection of material from planets, and the resultant trajectories, are neither general nor particular solutions to the n-body problem, but are numerical results predicated on various assumptions, whether implicitly or explicitly stated. Hence, the fate of an impact can only be deduced in general terms from simple approaches.

## **3.2 Outward transport of material ejected during collisions**

The total kinetic and potential energy of a particle in an elliptical orbit depends on the semimajor radius but not on the eccentricity (*e*), semi-minor radius, or angular momentum

The core grows at a rate comparable to Stokes' velocity. Large planets with large *g* remain hot longer, forming proportionately larger cores (Fig. 2b). For bodies with *g* exceeding ~2

Asteroids are too small to possess cores. It is immaterial whether asteroidal bodies are mixtures of iron and silicate or of silicate and ice, given similar density contrasts. Because the density contrast among various silicate mineral pairs are an order of magnitude smaller, asteroids could not have sorted olivine from plagioclase. Our conclusion that Vesta and all other asteroids could not differentiate due to internal gravitational sorting does not preclude these being differentiated material. Rather, our analysis shows that the differentiation

Derivation of meteorites from Mars has been demonstrated (summarized by Grady, 2000).

material from Mercury is not merely possible, but expected. Hence, objections to Mercury as a parent body rest on the notion that transport of material outward from the innermost Solar System is prohibited due to the huge gravitational influence of the Sun (e.g., Melosh & Tonks, 1993; summarized by Love & Kiel, 1995). Such difficulties have been overstated. As

Calculations concerning ejection of objects from planetary bodies and the consequent orbits are a variant of the three-body problem of classical mechanics. A system of three or more masses, moving under their mutual gravitational forces, cannot be solved in any general way (e.g., Goldstein, 1950; Symon, 1971). For example, the three-body problem cannot be reduced to three one-body problems. Very few special solutions exist: 1) the restricted threebody problem can be solved, wherein one mass is small and does not perturb the motions of the other two large masses (e.g., the trajectory of a rocket from Earth to the Moon); 2) motions of the planets can be accurately calculated, which represent a numerical solution obtained from specified initial conditions that holds over some period of time; and 3) a particular solution exists for three-bodies orbiting about the center of mass (discovered by Lagrange). Dynamical studies of ejection of material from planets, and the resultant trajectories, are neither general nor particular solutions to the n-body problem, but are numerical results predicated on various assumptions, whether implicitly or explicitly stated. Hence, the fate of an impact can only be deduced in general terms from simple approaches.

The total kinetic and potential energy of a particle in an elliptical orbit depends on the semimajor radius but not on the eccentricity (*e*), semi-minor radius, or angular momentum

s-2) and an escape velocity of 4.3

s-1). Together these characteristics show that ejection of

s-2, close to that of Mars (3.71 m.

discussed below in general terms, classical mechanics indicates transport is feasible.

demonstrates that a minimum g is needed for coherent, downward motions.

occurred within a larger body with sufficient mass to permit density sorting.

**3. Derivation of achondrites from the inner Solar System** 

**3.2 Outward transport of material ejected during collisions** 

s-2, core size directly depends on *g*, as indicted by Eq. 2. Because large planets retain more heat than small, viscosity decreases as planet size increases which makes the dependence of *v* on g weaker than linear. However, Eq. 2 assumes that particles sink and reach a terminal velocity within the static fluid, but does not describe whether sinking is possible. Figure 2

m.

km.

Mercury has g = 3.70 m.

**3.1 Dynamical constraints** 

s-1, low compared to Mars (5 km.

(e.g., Symon, 1971). Hence, an object in a circular orbit of radius *r* from the Sun has the same total energy as an object in extremely eccentric orbits whose distance at aphelion can approach 2*r*. Due to this property, ejecta can be transported to beyond the orbit of Venus without adding any kinetic energy (Fig. 3).

Fig. 3. Elliptical orbits for objects sharing the orbital energy of Mercury (solid curves, Mercury in bold), compared to the nearly circular orbit of Venus (dashed curve). Labels indicate the eccentricity of the orbits, which vary from a perfect circle (*e* = 0) to nearly unity.

## **3.2.1 Sling-shotting to distant reaches**

For any ejecta, including those in elliptical orbits sharing the kinetic energy of Mercury's orbit, promotion to even more distant parts of the Solar System is possible. We base this remark on use of Jupiter's gravitational well to "slingshot" the famous Voyager spacecraft into the outer Solar System. If Venus were in a suitable position relative to the trajectories of ejected material in a highly eccentric orbit (*e*>0.86, Fig. 3), it is possible for that material with minimum kinetic energy to be promoted far beyond Venus' orbit.

## **3.2.2 Effect of kinetic energy beyond the minimum**

If the perturbing collision adds energy to the object being ejected, far greater distances than Venus' orbit can be realized. Escape from Mercury itself requires 4.3 km. s-1, a mere 1% increase over existing orbital energy. For transport to infinity, the velocity needed for an object to escape the gravitational well of the Sun at Mercury's orbit is 68 km. s-1. Because the orbital speed of Mercury is 48 km. s-1, escape of ejecta from the Sun's pull to infinity requires only doubling their kinetic energy. It should be clear that energies arising from impacts were sufficient to move material from the inner to the outer Solar System.

The kinetic energy of Mercury in its orbit is ~106 J. g-1 whereas that of the asteroids is much smaller, ~105 J. g-1, because orbital velocities become progressively slower with distance from the Sun. Hence, material ejected from an inner planet, if on an appropriate trajectory, has

Origin of HED Meteorites from the Spalling of Mercury –

1>e>0

a

52<v<68 km s-1

**e=0** 

Mercury is not isolated and the Sun pulls ejecta with *v* < 68 km.

restricted by geometry but are more likely to intersect Earth over time.

e=1 e=2 v>68 km s-1

Implications for the Formation and Composition of the Inner Planets 161

System. Given the great distances, parallel paths reasonably approximate these impacts. Consequently, Mercury's outer leading quadrant is struck preferentially, and most of the resulting recoil paths point away from the Sun or are tangential to Mercury's orbit (Fig. 4c).

A minimum energy escape trajectory from an isolated body follows a parabolic path (*e* = 1). With increasing energy, hyperbolae are expected (*e* = 2; e.g., Symon, 1971). However,

With more energy, the ellipse is larger, making excursion to the location of the asteroid belt possible for certain values of kinetic energy and initial paths. Aperiodic bounded orbits which exist for central forces (Symon, 1971) would be possible trajectory paths for intermediate energy particles, with initially tangential trajectories (Fig. 5b). This type of orbit can be open as shown, or closed, similar to the patterns of a drawing implement known as a "spirograph." Orbits of meteor strikes are presumed ellipses (e.g., Gounelle et al., 2006). To our knowledge, aperiodic bound orbits have not been considered, yet these are not only less

**impactor** 

ejecta escape the Solar system. (b) Aperiodic bound orbits are permissible for a central force

Regular and distinct trends exist for the mass of dusty and gassy bodies that orbit the Sun with distance (Fig. 6). These trends result from 3-d accretion of the pre-solar nebula (Hofmeister & Criss, 2012), such that the dusty bodies form first, and gas is augmented onto dusty protoplanets in accord with Eq. 3. The decrease in mass for the dusty bodies going as *r*-1.7 is consistent with conservation of angular momentum and energy during 3-collapse. Independent of our model, the trend for dusty primary satellites shows that objects in the asteroid belt formed in a different manner than did the planets and dwarf planets, especially insofar as the total mass in the asteroid belt falls two orders of magnitude below the trend.

The innermost planet seems to have lost practically all of its mantle, as its core extends out to nearly 80% of the planet's radius, and constitutes ~⅔ of its total mass (Fig. 1). A plausible interpretation is that proto-Mercury was once at least as large as Mars (Fig. 1) and that

Fig. 5. Schematic of possible ejecta orbits. (a) Simple orbits showing the connection of eccentricity with velocities needed to escape the Sun's control. Only for v>68 km.

(Symon, 1971) and could arise from initially tangential recoil paths.

**3.4 Type and mass of material excavated from evolving Mercury** 

**3.3 Mass arguments for the asteroid belt as debris** 

b

s-1 into bound orbits (Fig. 5a).

M

asteroid belt

s-1 can the

more than enough kinetic energy for placement in the asteroid belt. Once material reaches the belt, interactions and collisions with inward bound material could reduce their kinetic energy to an amount appropriate to the orbital energy in this region.

#### **3.2.3 Solar controls on the draw of Mercury and on impact and ejecta trajectory**

Geometry, properties of central forces, and relative sizes of Mercury and the Sun suggest possible paths. Importantly, gravitational competition with the Sun severely limits Mercury's draw. Due to the central nature of this force, for a particle within a dust cloud to be incorporated in a body orbiting a larger central mass (Fig. 4a),

$$\left(\begin{matrix}r\_{\text{cloud}}\\\\\end{matrix}\bigg|r\_{\text{orbit}}\right)^{2} \leq \bigvee\_{\text{central}} \mathbf{M}\_{\text{central}} \tag{3}$$

(Hofmeister and Criss, 2012). Although Mercury would actually draw in more distant material above and below its orbital plane and away from the Sun, our spherical formula suffices due to the large orbital radii and small planet mass: using the masses of the Sun and Mercury, provides a cloud size of 4.4x107 m, which is only 20 times Mercury's planetary radius (2.4x106 m) and an insignificant portion of its orbital radius (5.79x1010 m). Hence, a tiny spherical cloud reasonably describes the gravitational draw of Mercury. Impacts derived from cloud material are radial inward and help assemble the planet's mass. Our interest is rather collisions occurring after formation that would scour evolving Mercury.

Fig. 4. Geometrical control on impacts on a planet. (a) Gravitational competition. The size of a cloud collecting unto an orbiting body (dashed circle) is determined by where the gravitational force (black arrow) of the central object (Sun) balances that (white arrow) of the orbiting protoplanet (dot). (b) Input trajectories. The planet (circle) in its orbit, intercepts material that is stationary or moving towards the planet (dotted arrows). Material focused on the Sun strikes the planet at nearly right angles (heavy dashed arrows). Shaded area = the most affected quadrant. (c) Initial output trajectories are shown for direct hits, glancing blows and strikes in the middle of the preferred quadrant.

Once Mercury was formed, two paths for impact are relevant. One is a "windshield" effect, whereby Mercury collides with debris entering its orbit (Fig. 4b). Because of the great distance to the Sun, these trajectories are approximately tangential to the orbit. The other impact path is approximately radially inward to the Sun (Fig. 4b), which is the gravitational focus of the Solar

more than enough kinetic energy for placement in the asteroid belt. Once material reaches the belt, interactions and collisions with inward bound material could reduce their kinetic

Geometry, properties of central forces, and relative sizes of Mercury and the Sun suggest possible paths. Importantly, gravitational competition with the Sun severely limits Mercury's draw. Due to the central nature of this force, for a particle within a dust cloud to

2

(Hofmeister and Criss, 2012). Although Mercury would actually draw in more distant material above and below its orbital plane and away from the Sun, our spherical formula suffices due to the large orbital radii and small planet mass: using the masses of the Sun and Mercury, provides a cloud size of 4.4x107 m, which is only 20 times Mercury's planetary radius (2.4x106 m) and an insignificant portion of its orbital radius (5.79x1010 m). Hence, a tiny spherical cloud reasonably describes the gravitational draw of Mercury. Impacts derived from cloud material are radial inward and help assemble the planet's mass. Our interest is rather collisions occurring after formation that would scour evolving Mercury.

b to the Sun

Fig. 4. Geometrical control on impacts on a planet. (a) Gravitational competition. The size of a cloud collecting unto an orbiting body (dashed circle) is determined by where the gravitational force (black arrow) of the central object (Sun) balances that (white arrow) of the orbiting protoplanet (dot). (b) Input trajectories. The planet (circle) in its orbit, intercepts material that is stationary or moving towards the planet (dotted arrows). Material focused on the Sun strikes the planet at nearly right angles (heavy dashed arrows). Shaded area = the most affected quadrant. (c) Initial output trajectories are shown for direct hits, glancing

Once Mercury was formed, two paths for impact are relevant. One is a "windshield" effect, whereby Mercury collides with debris entering its orbit (Fig. 4b). Because of the great distance to the Sun, these trajectories are approximately tangential to the orbit. The other impact path is approximately radially inward to the Sun (Fig. 4b), which is the gravitational focus of the Solar

blows and strikes in the middle of the preferred quadrant.

v = 48 km s-1

a

*r M r M*

orbit central

(3)

c

to the Sun

**3.2.3 Solar controls on the draw of Mercury and on impact and ejecta trajectory** 

energy to an amount appropriate to the orbital energy in this region.

be incorporated in a body orbiting a larger central mass (Fig. 4a),

cloud

System. Given the great distances, parallel paths reasonably approximate these impacts. Consequently, Mercury's outer leading quadrant is struck preferentially, and most of the resulting recoil paths point away from the Sun or are tangential to Mercury's orbit (Fig. 4c).

A minimum energy escape trajectory from an isolated body follows a parabolic path (*e* = 1). With increasing energy, hyperbolae are expected (*e* = 2; e.g., Symon, 1971). However, Mercury is not isolated and the Sun pulls ejecta with *v* < 68 km. s-1 into bound orbits (Fig. 5a). With more energy, the ellipse is larger, making excursion to the location of the asteroid belt possible for certain values of kinetic energy and initial paths. Aperiodic bounded orbits which exist for central forces (Symon, 1971) would be possible trajectory paths for intermediate energy particles, with initially tangential trajectories (Fig. 5b). This type of orbit can be open as shown, or closed, similar to the patterns of a drawing implement known as a "spirograph." Orbits of meteor strikes are presumed ellipses (e.g., Gounelle et al., 2006). To our knowledge, aperiodic bound orbits have not been considered, yet these are not only less restricted by geometry but are more likely to intersect Earth over time.

Fig. 5. Schematic of possible ejecta orbits. (a) Simple orbits showing the connection of eccentricity with velocities needed to escape the Sun's control. Only for v>68 km. s-1 can the ejecta escape the Solar system. (b) Aperiodic bound orbits are permissible for a central force (Symon, 1971) and could arise from initially tangential recoil paths.

## **3.3 Mass arguments for the asteroid belt as debris**

Regular and distinct trends exist for the mass of dusty and gassy bodies that orbit the Sun with distance (Fig. 6). These trends result from 3-d accretion of the pre-solar nebula (Hofmeister & Criss, 2012), such that the dusty bodies form first, and gas is augmented onto dusty protoplanets in accord with Eq. 3. The decrease in mass for the dusty bodies going as *r*-1.7 is consistent with conservation of angular momentum and energy during 3-collapse. Independent of our model, the trend for dusty primary satellites shows that objects in the asteroid belt formed in a different manner than did the planets and dwarf planets, especially insofar as the total mass in the asteroid belt falls two orders of magnitude below the trend.

## **3.4 Type and mass of material excavated from evolving Mercury**

The innermost planet seems to have lost practically all of its mantle, as its core extends out to nearly 80% of the planet's radius, and constitutes ~⅔ of its total mass (Fig. 1). A plausible interpretation is that proto-Mercury was once at least as large as Mars (Fig. 1) and that

Origin of HED Meteorites from the Spalling of Mercury –

**3.5 Statistical arguments for planetary origins of meteorites** 

Also, spectra of asteroids preferentially represent larger bodies.

X-type

Vestoids All others

Asteroids (Bus and Binzel)

equivocal. Grain-size and dust coverage affect surface spectra (Section 4).

**3.6 The asteroid belt as the junkyard of the Solar System** 

S-type

C-type

Asteroids (Tholen)

C-type

Vesta Mars & Moon

S-type

M-type

Implications for the Formation and Composition of the Inner Planets 163

shock metamorphism, large volumes of melt, and small volumes of melt ejecta (tektites and small spherules) (French, 1998). Mercury has low gravity and lacks an impeding atmosphere. Impacts into pools of melt on Mercury splattered innumerable chondrules into space. Crystallization of the melt in space would provide round rather than the aerodynamic shapes of tektites, and would have formed bodies of diverse sizes. More massive impacts would have vaporized material, which would have escaped and then condensed. The internal structure of Mercury suggests that most ejecta would be from the mantle. However, impacts add their own matter to the surface, which would also be subsequently ejected. Hence, primitive material, differentiated material, both their melts, and mixtures of all such debris are possible. Vesta could be a conglomerate of ejecta from Mercury.

Asteroids are classified by remotely sensed data by several schemes. Spectra (Bus & Binzel, 2002) or spectra plus albedo (Tholen, 1989) are further used to link these types to meteorites. Broad categories of asteroids common among these schemes occur in percentages that are inconsistent with the numerical occurrence of meteorites (Fig. 7). Current collections have ~40,000 samples (Table 1) and should be statistically representative. Mass distribution gives us a different perspective and is skewed towards high iron content, due to material strength.

> Irons & Stony irons

Fig. 7. Large disparity between two different asteroidal classifications (left) and meteorite types (right). Dark "C-type" asteroids represent a majority, yet are considered as the

primary sources of the rare carbonaceous chondrites. Vesta is only one of >106 asteroids, but is presumed to be the source of the HEDs, which represent >½ of the achondrites, and are geochemically tied to ~ ⅓ of the irons, and most stony-iron meteorites. Data from Table 1.

The current linkage is statistically untenable, as is well known. Problems exist with attribution of C-type asteroids to carbonaceous chondrites. Spectra of C-types are featureless with low albedo, consistent with dusty surfaces; it may simply be that their mineralogy is

The above dynamical, mass, and statistical considerations lead to an alternative view of the asteroid belt. Classical mechanics requires that ejecta from early Mercury predominately

Meteorites by mass

Irons

Achondrites Chondrites Carbonaceous chondrites

HEDs

Meteorites by number

Ordinary & enstatite chondrites

AchondritesCarbonaceous chondrites HEDs

Stony irons

>3x1023 kg mostly of silicates was blasted off this battered planet early in its history. The total mass of the asteroid belt (~3x1021 kg) is only 1% of Mercury's loss, and half of the mass in the belt is sequestered in one object, Ceres. Much of Mercury's lost mantle would have been absorbed within the inner Solar System, but this huge quantity of material will offset the low probabilities of ejecting material and transferring it to Earth-crossing orbits. Transport of <1% of Mercury's loss to the asteroid belt is reasonable, and, more importantly, is much more than sufficient to generate the aggregate mass of achondrite and other classes of meteorites in available collections, including the ~1000 kg of HEDs (Table 1). Other bodies will provide achondritic material, but in much smaller amounts and ejecta originate at various times. Chondritic material would represent dust originating from this region.

Fig. 6. Mass of large bodies orbiting the Sun vs. distance from the origin. The Sun is plotted at its body radius to use a logarithm plot, needed to represent the vastness of the Solar System. Open circles and dotted line = gas giants. Dashed line at right side = radius of the pre-solar nebula, calculated from the Sun's mass, assuming an initial density like that of molecular clouds. Filled circle = heavy elements in the Sun (Basu & Antia, 2008), taken to represent the dust component of the pre-solar nebula (PSN). Squares = inner rocky planets, except Mercury (open square) which has lost significant mass. Diamonds = dwarf planets of the outer Solar System. Line = fit to the circles, squares and diamonds. Grey diamond = Sedna. Horizontal lines describe substantially elliptical orbits. Triangles = individual asteroids (grey for the total mass). Data from Lodders & Fegely (1998) and http://pds.jpl.nasa.gov/.

The heavily cratered surface of Mercury and its location as the innermost planet provide evidence for deep excavation. Saturation of the surfaces of Callisto and Ganymede with craters implies that the Jovian moons were also extensively bombarded, due to the gravitational focus of impactors by Jupiter. By analogy, early, evolving Mercury received an intense flux of impactors, due to its proximity to the Sun.

Our understanding of impacts is skewed toward late, low energy events which do not cause great loss of planetary matter. Such minor impacts provide dust (rayed craters), breccias,

>3x1023 kg mostly of silicates was blasted off this battered planet early in its history. The total mass of the asteroid belt (~3x1021 kg) is only 1% of Mercury's loss, and half of the mass in the belt is sequestered in one object, Ceres. Much of Mercury's lost mantle would have been absorbed within the inner Solar System, but this huge quantity of material will offset the low probabilities of ejecting material and transferring it to Earth-crossing orbits. Transport of <1% of Mercury's loss to the asteroid belt is reasonable, and, more importantly, is much more than sufficient to generate the aggregate mass of achondrite and other classes of meteorites in available collections, including the ~1000 kg of HEDs (Table 1). Other bodies will provide achondritic material, but in much smaller amounts and ejecta originate at various times. Chondritic material would represent dust originating from this region.

> 10<sup>9</sup> 10<sup>10</sup> 10<sup>11</sup> 10<sup>12</sup> 10<sup>13</sup> 10<sup>14</sup> 10<sup>15</sup> Distance from Sun, m

Vesta

Fig. 6. Mass of large bodies orbiting the Sun vs. distance from the origin. The Sun is plotted at its body radius to use a logarithm plot, needed to represent the vastness of the Solar System. Open circles and dotted line = gas giants. Dashed line at right side = radius of the pre-solar nebula, calculated from the Sun's mass, assuming an initial density like that of molecular clouds. Filled circle = heavy elements in the Sun (Basu & Antia, 2008), taken to represent the dust component of the pre-solar nebula (PSN). Squares = inner rocky planets, except Mercury (open square) which has lost significant mass. Diamonds = dwarf planets of the outer Solar System. Line = fit to the circles, squares and diamonds. Grey diamond = Sedna. Horizontal lines describe substantially elliptical orbits. Triangles = individual asteroids (grey for the total

The heavily cratered surface of Mercury and its location as the innermost planet provide evidence for deep excavation. Saturation of the surfaces of Callisto and Ganymede with craters implies that the Jovian moons were also extensively bombarded, due to the gravitational focus of impactors by Jupiter. By analogy, early, evolving Mercury received an

Our understanding of impacts is skewed toward late, low energy events which do not cause great loss of planetary matter. Such minor impacts provide dust (rayed craters), breccias,

Mars Earth

Individual asteroids

mass). Data from Lodders & Fegely (1998) and http://pds.jpl.nasa.gov/.

intense flux of impactors, due to its proximity to the Sun.

**Total asteroid belt**

Venus

**gas M~r 1.1**

Mercury (spalled)

Eris

**edge if** 

**PSN=1.6x10 -19 gcm-3**

Haumea **Sedna**

Jupiter

Saturn Uranus Neptune

Pluto

Makemake

10<sup>20</sup>

10<sup>22</sup>

10<sup>24</sup>

10<sup>26</sup>

10<sup>28</sup>

10<sup>30</sup>

Sun

heavy elements in Sun

> **dust M~r-1.7**

shock metamorphism, large volumes of melt, and small volumes of melt ejecta (tektites and small spherules) (French, 1998). Mercury has low gravity and lacks an impeding atmosphere. Impacts into pools of melt on Mercury splattered innumerable chondrules into space. Crystallization of the melt in space would provide round rather than the aerodynamic shapes of tektites, and would have formed bodies of diverse sizes. More massive impacts would have vaporized material, which would have escaped and then condensed. The internal structure of Mercury suggests that most ejecta would be from the mantle. However, impacts add their own matter to the surface, which would also be subsequently ejected. Hence, primitive material, differentiated material, both their melts, and mixtures of all such debris are possible. Vesta could be a conglomerate of ejecta from Mercury.

## **3.5 Statistical arguments for planetary origins of meteorites**

Asteroids are classified by remotely sensed data by several schemes. Spectra (Bus & Binzel, 2002) or spectra plus albedo (Tholen, 1989) are further used to link these types to meteorites. Broad categories of asteroids common among these schemes occur in percentages that are inconsistent with the numerical occurrence of meteorites (Fig. 7). Current collections have ~40,000 samples (Table 1) and should be statistically representative. Mass distribution gives us a different perspective and is skewed towards high iron content, due to material strength. Also, spectra of asteroids preferentially represent larger bodies.

Fig. 7. Large disparity between two different asteroidal classifications (left) and meteorite types (right). Dark "C-type" asteroids represent a majority, yet are considered as the primary sources of the rare carbonaceous chondrites. Vesta is only one of >106 asteroids, but is presumed to be the source of the HEDs, which represent >½ of the achondrites, and are geochemically tied to ~ ⅓ of the irons, and most stony-iron meteorites. Data from Table 1.

The current linkage is statistically untenable, as is well known. Problems exist with attribution of C-type asteroids to carbonaceous chondrites. Spectra of C-types are featureless with low albedo, consistent with dusty surfaces; it may simply be that their mineralogy is equivocal. Grain-size and dust coverage affect surface spectra (Section 4).

## **3.6 The asteroid belt as the junkyard of the Solar System**

The above dynamical, mass, and statistical considerations lead to an alternative view of the asteroid belt. Classical mechanics requires that ejecta from early Mercury predominately

Origin of HED Meteorites from the Spalling of Mercury –

0.5 1 1.5 2 2.5 Wavelength, μm

*d*. (c) Schematic of emissions, where *I*BB is blackbody intensity.

Mesosiderites

Estherville

Fig. 9. Effect of grain-size. (a) Comparison of orientational differences in spectra for Vesta to

Padvarminkai (Gaffey, 1976)] and powdered mesosiderites as labeled (Burbine et al., 2007). Vesta spectra (Gaffey, 1997) are shown for two rotational aspects (dots), with each compared to the average (lines) of 4 to 5 rotational aspects. The uncertainties are roughly 2-3 times the

reflections for the most intense ray paths at near-normal incidence (the angle is exaggerated and the white arrow is not labeled). The light received by a detector (*I*meas) relative to incident intensity (*I*0) is a combination of reflection and absorption, depending on thickness,

Measurements are affected by back reflections (e.g., Hofmeister et al., 2003). If the surface layer is very thick then the back reflections are not sensed. Reflectivity at high frequency is flat for a bare, large, single crystal due to Fresnel's law, (*r* = [(*n*-1)2+*k*2]/[(*n*+1)2+*k*2] where *n* is the index of refraction, k = *a*/4 where *a* is the absorption coefficient, and = frequency.

the grain-size dependence of reflectance-absorbance spectra from eucrites [solid = Padvarminkai (Hiroi et al., 1995); dashed = Macibini (Burbine et al. 2001); dotted =

symbol size, and the placement of the points near 2 m is not exact because spectra presented at different scales were merged. Spectra offset for clarity. (b) Schematic of

d

c

d

b

(1-α) 2

2

Imeas,emit = εIBB(1-r)

I0r(1-r)<sup>2</sup>

I0r(1-r)(1-α)

εIBBr

I0r(1-r)(1-α)

I = εIBB

I0

I0(1-r)

I0(1-r)(1-α)

I0(1-r)<sup>2</sup>

(1-α)

Imeas, refl

I0r

powder 298 K a

Lamont

Vaca Muerta

**4.2 Effect of grain-size on spectra** 

baseline, and strengths of the features.

Implications for the Formation and Composition of the Inner Planets 165

Grain-size controls whether peaks are superimposed positively or negatively on the

Light produced by a body at any frequency is the product of the blackbody function times the emissivity. Kirchhoff's law for a body in thermodynamic equilibrium requires that absorptivity (= *I*abs/*I*0, where *I*abs = the intensity of light actually absorbed) equals emissivity () (e.g., Bates, 1978; Brewster 1992). For an opaque material such as a metal, 1= +*r*, where *r* =*I*ref/*I*0 is reflectivity. Although asteroids are large and opaque, dust grains on their surface are partially transparent. For the part of the asteroid that is sampled, 1= +*r* +*t*, where *t* = *I*tran/*I*0 is transmittivity of the uppermost layer as derived by Bates (1978) for dielectrics (e.g. silicates).

occupy elliptical orbits and aperiodic bound orbits. Over time, the amount of material is non-circular orbits will diminish. The inner Solar System has been swept clean. The asteroid belt has a large number of objects because the distances are large, so the clearing rate is low, and a planet did not form in this area. On this basis, a likely repository for the ejecta and leftovers of accretion is the asteroid belt. Gravitational forces (e.g. Jupiter) have shepherded errant objects into this veritable junkyard of debris. Nothing requires this zone to be the locus of accretion for all of its material.

## **4. Limitations of spectroscopic assignments**

Spectra from remote objects are obtained under uncontrolled conditions and contain a mixture of absorption, emission, and reflection features. To relate these data to laboratory mineral spectra, we need to understand how spectra are affected by sampling conditions.

#### **4.1 Effect of temperature on spectra**

For a remotely probed surface, the most important effect of temperature (*T*), is the control it exerts on whether the object is emitting or reflecting light over any given frequency range.

Blackbody emission curves depend strongly on *T* and frequency (Fig. 8a). The 1st law of thermodynamics requires that the flux from Vesta match that received from the Sun (Fig. 8b); roughly speaking, the areas under the curves must be equal. Due to the properties of Planck curves, Vesta, which is cold (85<*T*<255 K; e.g., Lucey et al., 1998) outputs virtually all its light below 2200 cm-1, and therefore is emitting in the infrared but reflects light in the near-IR to visible.

Fig. 8. Factors governing whether light is emitted or reflected. (a) Dependence of blackbody curves on temperature, as labeled. Except for the Sun at ~6000 K (grey), the left y-axis pertains. Arrows indicate spectral ranges sampled by different instruments. (b) Schematic of the first law of thermodynamics. Black arrow = light from Sun. Grey arrows = light emitted, which integrated over the area and frequency must equal the flux received. (c) Cold objects emit at low frequency, and reflect sunlight at high frequency (speckled arrow). Both are modified by spectral properties, which depend largely on whether the surface is bare (d) or dusty (e), as sketched for the case of reflected sunlight.

### **4.2 Effect of grain-size on spectra**

164 New Achievements in Geoscience

occupy elliptical orbits and aperiodic bound orbits. Over time, the amount of material is non-circular orbits will diminish. The inner Solar System has been swept clean. The asteroid belt has a large number of objects because the distances are large, so the clearing rate is low, and a planet did not form in this area. On this basis, a likely repository for the ejecta and leftovers of accretion is the asteroid belt. Gravitational forces (e.g. Jupiter) have shepherded errant objects into this veritable junkyard of debris. Nothing requires this zone to be the

Spectra from remote objects are obtained under uncontrolled conditions and contain a mixture of absorption, emission, and reflection features. To relate these data to laboratory mineral spectra, we need to understand how spectra are affected by sampling conditions.

For a remotely probed surface, the most important effect of temperature (*T*), is the control it exerts on whether the object is emitting or reflecting light over any given frequency range. Blackbody emission curves depend strongly on *T* and frequency (Fig. 8a). The 1st law of thermodynamics requires that the flux from Vesta match that received from the Sun (Fig. 8b); roughly speaking, the areas under the curves must be equal. Due to the properties of Planck curves, Vesta, which is cold (85<*T*<255 K; e.g., Lucey et al., 1998) outputs virtually all its light below 2200 cm-1, and therefore is emitting in the infrared but reflects light in the

**0**

Fig. 8. Factors governing whether light is emitted or reflected. (a) Dependence of blackbody curves on temperature, as labeled. Except for the Sun at ~6000 K (grey), the left y-axis pertains. Arrows indicate spectral ranges sampled by different instruments. (b) Schematic of the first law of thermodynamics. Black arrow = light from Sun. Grey arrows = light emitted,

(c) Cold objects emit at low frequency, and reflect sunlight at high frequency (speckled arrow). Both are modified by spectral properties, which depend largely on whether the

**1 10<sup>14</sup>**

**2 10<sup>14</sup>**

**Sun 5850 K**

which integrated over the area and frequency must equal the flux received.

surface is bare (d) or dusty (e), as sketched for the case of reflected sunlight.

a

Near-IR to visible spectroscopy

**3 10<sup>14</sup>**

**4 10<sup>14</sup>**

b

c

d e

**5 10<sup>14</sup>**

locus of accretion for all of its material.

**4.1 Effect of temperature on spectra** 

near-IR to visible.

0

100 K

**200 K** 

**IR**

298 K

2 10<sup>7</sup>

4 10<sup>7</sup>

6 10<sup>7</sup>

8 10<sup>7</sup>

1 10<sup>8</sup>

**4. Limitations of spectroscopic assignments** 

0 2500 5000 7500 1 10<sup>4</sup>

Frequency, cm-1

500 K

Grain-size controls whether peaks are superimposed positively or negatively on the baseline, and strengths of the features.

Light produced by a body at any frequency is the product of the blackbody function times the emissivity. Kirchhoff's law for a body in thermodynamic equilibrium requires that absorptivity (= *I*abs/*I*0, where *I*abs = the intensity of light actually absorbed) equals emissivity () (e.g., Bates, 1978; Brewster 1992). For an opaque material such as a metal, 1= +*r*, where *r* =*I*ref/*I*0 is reflectivity. Although asteroids are large and opaque, dust grains on their surface are partially transparent. For the part of the asteroid that is sampled, 1= +*r* +*t*, where *t* = *I*tran/*I*0 is transmittivity of the uppermost layer as derived by Bates (1978) for dielectrics (e.g. silicates).

Fig. 9. Effect of grain-size. (a) Comparison of orientational differences in spectra for Vesta to the grain-size dependence of reflectance-absorbance spectra from eucrites [solid = Padvarminkai (Hiroi et al., 1995); dashed = Macibini (Burbine et al. 2001); dotted = Padvarminkai (Gaffey, 1976)] and powdered mesosiderites as labeled (Burbine et al., 2007). Vesta spectra (Gaffey, 1997) are shown for two rotational aspects (dots), with each compared to the average (lines) of 4 to 5 rotational aspects. The uncertainties are roughly 2-3 times the symbol size, and the placement of the points near 2 m is not exact because spectra presented at different scales were merged. Spectra offset for clarity. (b) Schematic of reflections for the most intense ray paths at near-normal incidence (the angle is exaggerated and the white arrow is not labeled). The light received by a detector (*I*meas) relative to incident intensity (*I*0) is a combination of reflection and absorption, depending on thickness, *d*. (c) Schematic of emissions, where *I*BB is blackbody intensity.

Measurements are affected by back reflections (e.g., Hofmeister et al., 2003). If the surface layer is very thick then the back reflections are not sensed. Reflectivity at high frequency is flat for a bare, large, single crystal due to Fresnel's law, (*r* = [(*n*-1)2+*k*2]/[(*n*+1)2+*k*2] where *n* is the index of refraction, k = *a*/4 where *a* is the absorption coefficient, and = frequency.

Origin of HED Meteorites from the Spalling of Mercury –

**5.1 Geochemical characteristics of meteorites** 

geochemical characteristics, discussed below.

Venus 0.72 7000\*

**5.2 Chemical trends in the Solar System** 

were established long ago by Wasserburg et al. (1964).

1 and 2 (Surkov et al., 1987).

distribution are well-understood.

Implications for the Formation and Composition of the Inner Planets 167

Solar system inherited from the pre-solar nebula. An important marker is the ratio of volatile to refractory elements, which is exemplified by K/U and Rb/Ba ratios. A ratio involving Al is useful because this refractory element has a short-lived isotope. We selected Ga for normalization as it is a more volatile Group IIIA element with a similar ionic radius. Stable isotope ratios (e.g. Si, O) have proven utility to petrogenesis. Oxygen isotopes are exceedingly important because this element is abundant and its fractionation and

Detailed descriptions of the geochemical and petrological characteristics of HED meteorites are available (see references in Section 1.1). Compared to other achondritic meteorites, the HED family includes the oldest known basalts in the Solar System, is the most volatile

Ordinary chondrites, by far the most abundant stony meteorites (>90%), are widely believed to originate from the asteroid belt. This attribution is consistent with chemical uniformity of this group of meteorites. Persuasive arguments for the origination of SNC meteorites from Mars have been advanced (see Grady, 2006). In contrast, assigning the HED family to Vesta or Vestoids (e.g., Drake, 2001) contradicts physical data (Section 1), abundances (Fig. 7), and

 **A.U. K/U Rb/Ba 104 Ga/Al 17O 18O 30Si** 

Ordinary chondrites 2-3 60000 0.6 5 0.7 to 1.3 3.7 to 5.6 -0.46 Mars, SNCs 1.52 16000 0.17 4 0.27 4.5 -0.48 Earth upper mantle 1.00 10000 0.10 2 0.00 5.7 -0.68 Moon, Lunar 1.00 3000 0.015 0.3 0.00 5.7 -0.45

HEDs, protoMercury 0.39 3700 0.007 0.2 -0.25 3.7 -0.45 Ureilites, CAI, K-ch ~0 <1000 0.002 0.005 <-0.8 <8.1 -0.47 Table 2. Geochemical and isotopic characteristics of meteorites and rocky planets. Data Sources: BVSP (1981), Lodders and Fegley (1998), Lodders (1998), Kitts and Lodders (1998), Molini-Velsko et al. (1986), McKeegan et al., (1998); MacPherson et al. (2005), Franchi et al. (2008), Armytage et al. (2011) and others. For A.U.~0, oxygen values were taken from a cluster of ureilites and a chondrule from the Kakangari chondrite (Prinz et al., 1989). For R chondrites, the average excludes hot desert finds. \*Gamma ray spectrometer results for Vega

The trends in Table 2 are coherent and regular. Similarities in the K/U ratio among disparate rock types on Earth, and the large difference of this ratio from that in chondrites,

Note that the K/U ratio of HED meteorites is much more similar to rocky materials from the inner Solar System than it is to materials presumed to have originated outside Earth's orbit.

R chondrite ~3 >40000 >0.35 8 4.3

depleted, the most strongly reduced, and distinct in terms of isotopes (Table 2).

For grain sizes of rocks, measured reflectance is reduced by back reflections (Fig. 9b). As spectra of eucrites demonstrate (Fig. 9a), larger grain size at the surface means deeper absorption features. Small grains are not detected: these contribute only to scattering.

Regarding emitted light, for a bare, thick surface or large grains, = , so emissions go as IBB(1-*r*). At frequencies where transitions occur less light is received, i.e., peaks point down as seen in the spectra of Christensen et al. (2000). (Note: 1-r has been incorrectly interpreted as emissivity for large grains). If the surface is covered with dust, this layer is heated by lattice conduction from below and emitted light comes only from this layer. Because reflection peaks are broader than absorption peaks, emissions from a dusty surface go as IBB = IBB and peaks point up as seen in spectra from small grains (Low & Coleman, 1966; Bates, 1978) and references therein.

## **4.3 Effect of Fe2+ contents and grain-size on near-IR reflectance spectra of Vesta**

High frequency reflection spectra (Fig. 9a) have the strong peaks of pigeonite (as in eucrites). Fe-bearing plagioclase may be present: its 8000 cm-1 band (Hofmeister & Rossman, 1984) is weak for Serra de Magé eucrite. No obvious peaks for orthopyroxene (dioginites) or olivine (pallasites) exists. Such may be extracted by peak fitting, but variations in grain-size of pigeonite suffice to explain the scant differences, as is clear from Fig. 9. Variation in grain-size were inferred from recent analysis of Vesta's mid-IR light-curves (Chamberlain et al., 2011).

Mesosiderite spectra are similar to eucrites, but the presence of orthopyroxene broadens the peak at 2000 nm and shifts it to 1800 nm (Fig. 9). The surface of Vesta was impacted. Rapid surface cooling would produce pigeonite as in terrestrial lavas. From Fig. 9, the surface of Vesta could be a mixture of pigeonite and metal. The proportion of pigeonite is high, based on peak depths, which is consistent with Vesta's density (Section 1.1.1).

Importantly, the depths of the peaks are connected with Fe2+ content of the minerals. Orthopyroxene in the HEDs has Fe/(Mg+Fe) of 0.25 which is half that of pigeonites. Therefore, pigeonite will dominate if in equal proportions. Orthopyroxene is also coarse grained: large crystals will not provide back reflections. The spectra of Vesta could contain all phases of mesosiderites, but spectra only prove that pigeonite is present and abundant.

## **4.4 Problems with interpretation of mid-IR emission spectra from Vesta**

Spectra obtained of Vesta using the Infrared Satellite Observatory (ISO) record its emitted light. A broad peak near 450 cm-1 (Heras et al., 2000) indicates a blackbody temperature of 130 K, consistent with previous inferences (e.g., Lucey et al., 1998). ISO spectra in this range and below are noisy, obscuring features of minerals. Mid-IR features are weak, consistent with a dust covering (Dotto et al., 2000; Lim et al., 2005). Hence, these features are in emission. Unfortunately, spectral analyses by these authors assume that emissions are that of an opaque (metallic) body [= IBB(1-r)] whereas presence of a dust cover requires that Imeas,emit = IBB for a dielectric (silicate) material (Fig. 9bc). The fits need redoing.

## **5. Geochemical arguments for HEDs as inner Solar System material**

Meteorites display a bewildering complexity of chemical and isotopic relationships. The following discussion focuses on characteristics that reveal radial chemical gradients in the

For grain sizes of rocks, measured reflectance is reduced by back reflections (Fig. 9b). As spectra of eucrites demonstrate (Fig. 9a), larger grain size at the surface means deeper absorption features. Small grains are not detected: these contribute only to scattering.

Regarding emitted light, for a bare, thick surface or large grains, = , so emissions go as IBB(1-*r*). At frequencies where transitions occur less light is received, i.e., peaks point down as seen in the spectra of Christensen et al. (2000). (Note: 1-r has been incorrectly interpreted as emissivity for large grains). If the surface is covered with dust, this layer is heated by lattice conduction from below and emitted light comes only from this layer. Because reflection peaks are broader than absorption peaks, emissions from a dusty surface go as IBB = IBB and peaks point up as seen in spectra from small grains (Low & Coleman, 1966;

**4.3 Effect of Fe2+ contents and grain-size on near-IR reflectance spectra of Vesta** 

High frequency reflection spectra (Fig. 9a) have the strong peaks of pigeonite (as in eucrites). Fe-bearing plagioclase may be present: its 8000 cm-1 band (Hofmeister & Rossman, 1984) is weak for Serra de Magé eucrite. No obvious peaks for orthopyroxene (dioginites) or olivine (pallasites) exists. Such may be extracted by peak fitting, but variations in grain-size of pigeonite suffice to explain the scant differences, as is clear from Fig. 9. Variation in grain-size were inferred from recent analysis of Vesta's mid-IR light-curves (Chamberlain et al., 2011).

Mesosiderite spectra are similar to eucrites, but the presence of orthopyroxene broadens the peak at 2000 nm and shifts it to 1800 nm (Fig. 9). The surface of Vesta was impacted. Rapid surface cooling would produce pigeonite as in terrestrial lavas. From Fig. 9, the surface of Vesta could be a mixture of pigeonite and metal. The proportion of pigeonite is high, based

Importantly, the depths of the peaks are connected with Fe2+ content of the minerals. Orthopyroxene in the HEDs has Fe/(Mg+Fe) of 0.25 which is half that of pigeonites. Therefore, pigeonite will dominate if in equal proportions. Orthopyroxene is also coarse grained: large crystals will not provide back reflections. The spectra of Vesta could contain all phases of mesosiderites, but spectra only prove that pigeonite is present and abundant.

Spectra obtained of Vesta using the Infrared Satellite Observatory (ISO) record its emitted light. A broad peak near 450 cm-1 (Heras et al., 2000) indicates a blackbody temperature of 130 K, consistent with previous inferences (e.g., Lucey et al., 1998). ISO spectra in this range and below are noisy, obscuring features of minerals. Mid-IR features are weak, consistent with a dust covering (Dotto et al., 2000; Lim et al., 2005). Hence, these features are in emission. Unfortunately, spectral analyses by these authors assume that emissions are that of an opaque (metallic) body [= IBB(1-r)] whereas presence of a dust cover requires that Imeas,emit = IBB for a dielectric (silicate) material (Fig. 9bc). The fits need redoing.

Meteorites display a bewildering complexity of chemical and isotopic relationships. The following discussion focuses on characteristics that reveal radial chemical gradients in the

on peak depths, which is consistent with Vesta's density (Section 1.1.1).

**4.4 Problems with interpretation of mid-IR emission spectra from Vesta** 

**5. Geochemical arguments for HEDs as inner Solar System material** 

Bates, 1978) and references therein.

Solar system inherited from the pre-solar nebula. An important marker is the ratio of volatile to refractory elements, which is exemplified by K/U and Rb/Ba ratios. A ratio involving Al is useful because this refractory element has a short-lived isotope. We selected Ga for normalization as it is a more volatile Group IIIA element with a similar ionic radius.

Stable isotope ratios (e.g. Si, O) have proven utility to petrogenesis. Oxygen isotopes are exceedingly important because this element is abundant and its fractionation and distribution are well-understood.

## **5.1 Geochemical characteristics of meteorites**

Detailed descriptions of the geochemical and petrological characteristics of HED meteorites are available (see references in Section 1.1). Compared to other achondritic meteorites, the HED family includes the oldest known basalts in the Solar System, is the most volatile depleted, the most strongly reduced, and distinct in terms of isotopes (Table 2).

Ordinary chondrites, by far the most abundant stony meteorites (>90%), are widely believed to originate from the asteroid belt. This attribution is consistent with chemical uniformity of this group of meteorites. Persuasive arguments for the origination of SNC meteorites from Mars have been advanced (see Grady, 2006). In contrast, assigning the HED family to Vesta or Vestoids (e.g., Drake, 2001) contradicts physical data (Section 1), abundances (Fig. 7), and geochemical characteristics, discussed below.


Table 2. Geochemical and isotopic characteristics of meteorites and rocky planets. Data Sources: BVSP (1981), Lodders and Fegley (1998), Lodders (1998), Kitts and Lodders (1998), Molini-Velsko et al. (1986), McKeegan et al., (1998); MacPherson et al. (2005), Franchi et al. (2008), Armytage et al. (2011) and others. For A.U.~0, oxygen values were taken from a cluster of ureilites and a chondrule from the Kakangari chondrite (Prinz et al., 1989). For R chondrites, the average excludes hot desert finds. \*Gamma ray spectrometer results for Vega 1 and 2 (Surkov et al., 1987).

## **5.2 Chemical trends in the Solar System**

The trends in Table 2 are coherent and regular. Similarities in the K/U ratio among disparate rock types on Earth, and the large difference of this ratio from that in chondrites, were established long ago by Wasserburg et al. (1964).

Note that the K/U ratio of HED meteorites is much more similar to rocky materials from the inner Solar System than it is to materials presumed to have originated outside Earth's orbit.

Origin of HED Meteorites from the Spalling of Mercury –

various processes and may onset early from impact heating

et al., 2005).

Implications for the Formation and Composition of the Inner Planets 169

fractionation trend and an ordinary mass-dependent fractionation trend. Practically all Solar system materials lie between the 1:1 lines known as EC (Equilibrated chondrites) and CCAM (carbonaceous chondrite anhydrous minerals). The CCAM line includes CAI inclusions from Allende which are very old, refractory, were rich in short lived isotopes and extend to incredibly low values. CAIs from enstatite and ordinary chondrites have similarly low values (McKeagen et al., 1998; Guan et al., 2000). Several lines of evidence point to CAI's being near the center of the Solar System and representing a distant reservoir (MacPherson

The 1:2 trends are superimposed on the 1:1 trends and are due to ordinary fractionation processes that continue today on Earth (the TFL). For any given planet, surface materials will lie to the upper right along these 1:2 lines (Fig. 11), while deep interior materials will lie to the lower left. Mars' meteorites lie above the TFL, whereas the HEDs lie below by a similar amount. These 1:2 trends develop during late stage heating of formed bodies by

Fig. 11. Oxygen isotopes. Tiny triangles = ureilites. Open squares = winonaites. Grey square = IIICD irons. Black squares = IAB irons. Circles = HEDS. Dark grey dots = mesosiderites. Light grey dots = main group pallasites. Black dots = IIIAB irons. Open cross = angrites. Open diamond = brachinites. Light X = aubrites. Heavy X = enstatite chondrites. Open triangles = Mars meteorites. Filled triangles =IIE irons. Filled diamonds = IVA irons. Small plus = H chondrites. Tiny black squares = L chondrites. Tiny grey squares = LL chondrites. Large plus = R-chondrites. Light lines = fractionation trends for various bodies, as labeled. Heavy lines = solid-gas interaction, as labeled. Double arrow indicates the Solar System gradient. Data from Clayton and Mayeda (1996); Franchi et al. (2008); and other sources.

The Rb/Ba ratio is also consistent and characteristic for each of the rock suites under consideration (Fig. 10). This and the Ga/Al ratio increase markedly in the order that the K/U ratio increases (Table 2). The remarkable coherency of Rb/Ba ratios in lunar materials, which are distinct from the chondritic ratio, has been demonstrated (BVSP, 1981, p. 251).

#### **5.2.1 Attribution of the HEDs to protoMercury: implications for elemental compositions**

Not all element ratios will vary in the same manner as those in Table 2, nor would such coherency be expected given the profound differences in planetary masses, timescales and extents of differentiation, and other factors. What is remarkable, however, is that so many compelling geochemical signatures vary in the same order, with the HED family near one extreme. These regular variations involving both major and minor elements, as well as the stable isotopic variations of the most abundant elements, are consistent with an origin from the inner Solar System, most logically, from early Mercury. The inventory of volatile and large ion lithophile elements in protoMercury would have been greatly reduced by spalling, leaving a remnant planet highly depleted in these elements, including K, U, and Th. Chemical data provided by Nittler et al (2011) and Peplowski et al. (2011) confirm our deduction.

Fig. 10. Zonation in the pre-solar nebula. (a) Rb vs. Ba concentrations for Solar System materials, suggesting HEDs are refractory inner Solar System material. Sources as in Table 2. (b) One hypothesis for origination of the zonation. From left to right: the pre-solar nebula of dust and gas (white cloud) is stable, but a nova injects material (grey) including refractory CAIs, while producing or inducing short-lived isotopes such as 26Al (grey). The injection diffuses, but the center of its mass provides a gravitational instability, precipitating nebula contraction, rotation around this locus, and planetary accretion. (c) Alternatively, two passing and interacting stars provide distinct reservoirs of dusty gas.

#### **5.3 Oxygen isotope variations in the Solar System**

Graphs of 17O vs. 18O values of achondrites and other meteorite types (Fig. 11) display a lattice-work of 1:1 and 1:2 lines, which respectively represent a mass-independent

The Rb/Ba ratio is also consistent and characteristic for each of the rock suites under consideration (Fig. 10). This and the Ga/Al ratio increase markedly in the order that the K/U ratio increases (Table 2). The remarkable coherency of Rb/Ba ratios in lunar materials, which are distinct from the chondritic ratio, has been demonstrated (BVSP, 1981, p. 251).

**5.2.1 Attribution of the HEDs to protoMercury: implications for elemental compositions**  Not all element ratios will vary in the same manner as those in Table 2, nor would such coherency be expected given the profound differences in planetary masses, timescales and extents of differentiation, and other factors. What is remarkable, however, is that so many compelling geochemical signatures vary in the same order, with the HED family near one extreme. These regular variations involving both major and minor elements, as well as the stable isotopic variations of the most abundant elements, are consistent with an origin from the inner Solar System, most logically, from early Mercury. The inventory of volatile and large ion lithophile elements in protoMercury would have been greatly reduced by spalling, leaving a remnant planet highly depleted in these elements, including K, U, and Th. Chemical data provided by Nittler et al (2011) and Peplowski et al. (2011) confirm our

Earth

b

c

Fig. 10. Zonation in the pre-solar nebula. (a) Rb vs. Ba concentrations for Solar System materials, suggesting HEDs are refractory inner Solar System material. Sources as in Table 2. (b) One hypothesis for origination of the zonation. From left to right: the pre-solar nebula of dust and gas (white cloud) is stable, but a nova injects material (grey) including refractory CAIs, while producing or inducing short-lived isotopes such as 26Al (grey). The injection diffuses, but the center of its mass provides a gravitational instability, precipitating nebula contraction, rotation around this locus, and planetary accretion. (c) Alternatively, two

Graphs of 17O vs. 18O values of achondrites and other meteorite types (Fig. 11) display a lattice-work of 1:1 and 1:2 lines, which respectively represent a mass-independent

time 0

Moon

passing and interacting stars provide distinct reservoirs of dusty gas.

deduction.


Chondrites

SNC/Mars

HED

0 0.5 1 1.5 2 2.5 3 3.5 Log Ba, ppm

**5.3 Oxygen isotope variations in the Solar System** 

fractionation trend and an ordinary mass-dependent fractionation trend. Practically all Solar system materials lie between the 1:1 lines known as EC (Equilibrated chondrites) and CCAM (carbonaceous chondrite anhydrous minerals). The CCAM line includes CAI inclusions from Allende which are very old, refractory, were rich in short lived isotopes and extend to incredibly low values. CAIs from enstatite and ordinary chondrites have similarly low values (McKeagen et al., 1998; Guan et al., 2000). Several lines of evidence point to CAI's being near the center of the Solar System and representing a distant reservoir (MacPherson et al., 2005).

The 1:2 trends are superimposed on the 1:1 trends and are due to ordinary fractionation processes that continue today on Earth (the TFL). For any given planet, surface materials will lie to the upper right along these 1:2 lines (Fig. 11), while deep interior materials will lie to the lower left. Mars' meteorites lie above the TFL, whereas the HEDs lie below by a similar amount. These 1:2 trends develop during late stage heating of formed bodies by various processes and may onset early from impact heating

Fig. 11. Oxygen isotopes. Tiny triangles = ureilites. Open squares = winonaites. Grey square = IIICD irons. Black squares = IAB irons. Circles = HEDS. Dark grey dots = mesosiderites. Light grey dots = main group pallasites. Black dots = IIIAB irons. Open cross = angrites. Open diamond = brachinites. Light X = aubrites. Heavy X = enstatite chondrites. Open triangles = Mars meteorites. Filled triangles =IIE irons. Filled diamonds = IVA irons. Small plus = H chondrites. Tiny black squares = L chondrites. Tiny grey squares = LL chondrites. Large plus = R-chondrites. Light lines = fractionation trends for various bodies, as labeled. Heavy lines = solid-gas interaction, as labeled. Double arrow indicates the Solar System gradient. Data from Clayton and Mayeda (1996); Franchi et al. (2008); and other sources.

Origin of HED Meteorites from the Spalling of Mercury –

trends, we assume that the gas phase is greatly enriched in 16O.

Fig. 12. Reservoirs and processes altering oxygen isotopes and mineral phases and compositions. Symbols for meteorites as in Fig. 11 with the exception of open cross = Rchondrites, and a few additions: Square with cross = acapulcoites. Square with X = Kchondrites. Right triangles = dark inclusions in CAIs. Light lines = late, ordinary fractionation within the various bodies, as labeled. Heavy lines = mixing or gas fractionation, as labeled. Grey double arrow indicates the Solar System gradient.

Implications for the Formation and Composition of the Inner Planets 171

consistent with the presence of pure, magnesium endmember silicates (forsterite, enstatite, and diopside), possibly calcium carbonate (Molster et al., 2001; Kemper et al., 2002), along with CAI minerals (Hofmeister et al., 2004). Hydrosilicates are possible (Hofmeister & Bowey, 2006). Dense dust clouds around young stellar objects contain ices (H2O, CO, CO2, CH4, hydrocarbons) thought to nucleate on the dust, summarized by Bowey and Hofmeister (2005) who identified melilite (a CAI phase) and Mg-rich amphiboles (hydrated pyroxenes). Both dust reservoirs condensed from hot gas expelled from pre-existing stars (Fig. 10bc). The large, and possibly pre-existing, dust reservoir involves O, Mg, Si, Fe, consistent with major elements in the Earth and with stellar nucleosynthesis. This reservoir is exemplified by the ordinary and R chondrites and dominates the outer reaches (Fig. 12). The large dust reservoir also contained lesser amounts of high-temperature phases, and lower temperature phases with volatile elements. The small, probably injected, dust reservoir provided the CAI inclusions, short-lived isotopes, refractory elements, and lesser amounts of iron and other materials which dominate the large dust reservoir. For a simple explanation of the 1:1

## **6. Geophysical implications**

To obtain bulk compositions of the silicate Earth and its core requires coherently explaining Fig. 11 in general terms that are consistent with fundamental principles and the data, and understanding the assembly, spalling, and resultant layers of Mercury.

## **6.1 Attribution of the HEDs to evolving Mercury: implications for layer compositions**

The large size of Mercury's core shows that it was deeply spalled both during and after core formation. Spalling was caused by late-stage impacts from the outer Solar System (Section 1.2), which added material from distance because impactors bury themselves while ejecting target material (French, 1988). Thus, the HED family trend lies closer to the TF line than proto-Mercury did. The upper layers of early Mercury must be represented in the meteorite collection. The trend of winonaites, IAB and IIICD irons lies below the HEDs (Fig. 11). This material is more pristine, consisting of non-magmatic irons and is similar to the enstatite chondrites associated with the Earth (e.g., Javoy, 1995) corroborating proto-Mercury as the source.

The HEDs represent early Mercury's present mantle, the mesosiderites are a combination of poorly sorted core-mantle boundary material and late-stage additions, and the IIIAB irons represent core material. Main-group pallasites represent material ejected from the zone of gravitational settling during core formation. Approximate compositions are in Table 3.


Table 3. Zones in Mercury and Earth and their connections with meteorite types.

#### **6.2 Explanation of solar system gradients and trends**

We dismiss the hypothesis that the material which formed the planets post-dates the Sun. The "condensation gradient" is invalid in view of recent data on comets (Zolensky et al., 2006). Simultaneous formation of the Sun and planets by 3-d nebular collapse is supported by the axial spins and orbital characteristics of the planets (Hofmeister & Criss, 2012) and indicates rapid accretion over ~1-3 Ma. What did the Solar System form from? Pre-solar grains indicate diverse sources (Bernatowicz & Zinner, 1997). For simplicity, given Table 2 and Fig. 11, we consider that two reservoirs of dust existed, which were embedded in an immense reservoir of gas. Gas constituents are mostly H2 with He and CO. Its mass is ~100 times larger than the dust mass, based on solar composition (Basu & Antia, 2008).

Our analysis is also based on the mineralogy of dust in astronomical environments. Infrared dust emissions of the exploded circumstellar dust of NGC 6302 contain many peaks,

To obtain bulk compositions of the silicate Earth and its core requires coherently explaining Fig. 11 in general terms that are consistent with fundamental principles and the data, and

**6.1 Attribution of the HEDs to evolving Mercury: implications for layer compositions**  The large size of Mercury's core shows that it was deeply spalled both during and after core formation. Spalling was caused by late-stage impacts from the outer Solar System (Section 1.2), which added material from distance because impactors bury themselves while ejecting target material (French, 1988). Thus, the HED family trend lies closer to the TF line than proto-Mercury did. The upper layers of early Mercury must be represented in the meteorite collection. The trend of winonaites, IAB and IIICD irons lies below the HEDs (Fig. 11). This material is more pristine, consisting of non-magmatic irons and is similar to the enstatite chondrites associated with the Earth (e.g., Javoy, 1995) corroborating proto-Mercury as the

The HEDs represent early Mercury's present mantle, the mesosiderites are a combination of poorly sorted core-mantle boundary material and late-stage additions, and the IIIAB irons represent core material. Main-group pallasites represent material ejected from the zone of gravitational settling during core formation. Approximate compositions are in Table 3.

Bulk silicate Earth ½ enstatite chondrites + ½ ordinary chondrites + ices

We dismiss the hypothesis that the material which formed the planets post-dates the Sun. The "condensation gradient" is invalid in view of recent data on comets (Zolensky et al., 2006). Simultaneous formation of the Sun and planets by 3-d nebular collapse is supported by the axial spins and orbital characteristics of the planets (Hofmeister & Criss, 2012) and indicates rapid accretion over ~1-3 Ma. What did the Solar System form from? Pre-solar grains indicate diverse sources (Bernatowicz & Zinner, 1997). For simplicity, given Table 2 and Fig. 11, we consider that two reservoirs of dust existed, which were embedded in an immense reservoir of gas. Gas constituents are mostly H2 with He and CO. Its mass is ~100

Our analysis is also based on the mineralogy of dust in astronomical environments. Infrared dust emissions of the exploded circumstellar dust of NGC 6302 contain many peaks,

Earth's core ~85%Fe + ~5%Ni + ~10%C with minor S, N, P Table 3. Zones in Mercury and Earth and their connections with meteorite types.

times larger than the dust mass, based on solar composition (Basu & Antia, 2008).

understanding the assembly, spalling, and resultant layers of Mercury.

**Object or layer Meteorite analogy or composition**  Proto-Mercury Winonaites + IAB + IIICD + ices

Core-mantle sorting Pallasites, main group

Proto-Earth enstatite chondrites + ices

**6.2 Explanation of solar system gradients and trends** 

Bulk silicate early Mercury HEDs

Mercury's evolving core IIIAB

**6. Geophysical implications** 

source.

consistent with the presence of pure, magnesium endmember silicates (forsterite, enstatite, and diopside), possibly calcium carbonate (Molster et al., 2001; Kemper et al., 2002), along with CAI minerals (Hofmeister et al., 2004). Hydrosilicates are possible (Hofmeister & Bowey, 2006). Dense dust clouds around young stellar objects contain ices (H2O, CO, CO2, CH4, hydrocarbons) thought to nucleate on the dust, summarized by Bowey and Hofmeister (2005) who identified melilite (a CAI phase) and Mg-rich amphiboles (hydrated pyroxenes).

Both dust reservoirs condensed from hot gas expelled from pre-existing stars (Fig. 10bc). The large, and possibly pre-existing, dust reservoir involves O, Mg, Si, Fe, consistent with major elements in the Earth and with stellar nucleosynthesis. This reservoir is exemplified by the ordinary and R chondrites and dominates the outer reaches (Fig. 12). The large dust reservoir also contained lesser amounts of high-temperature phases, and lower temperature phases with volatile elements. The small, probably injected, dust reservoir provided the CAI inclusions, short-lived isotopes, refractory elements, and lesser amounts of iron and other materials which dominate the large dust reservoir. For a simple explanation of the 1:1 trends, we assume that the gas phase is greatly enriched in 16O.

Fig. 12. Reservoirs and processes altering oxygen isotopes and mineral phases and compositions. Symbols for meteorites as in Fig. 11 with the exception of open cross = Rchondrites, and a few additions: Square with cross = acapulcoites. Square with X = Kchondrites. Right triangles = dark inclusions in CAIs. Light lines = late, ordinary fractionation within the various bodies, as labeled. Heavy lines = mixing or gas fractionation, as labeled. Grey double arrow indicates the Solar System gradient.

Origin of HED Meteorites from the Spalling of Mercury –

0123

**K**

A.U.

0

**CCAM**

may aid in deciphering parent bodies.

**7. Conclusions and future research** 

The light elements in the core are C, S, N and P.

than the upper mantle. Earth's bulk 18O value is close to +4.

The smaller the object's core, the more Fe-rich the basalts.

predictions to guide future work:

quenching due to impacts.

HEDs.

0.1

0.2

1/Range of

labeled.

δ18O values

0.3

0.4

0.5

Implications for the Formation and Composition of the Inner Planets 173

This trend is consistent with larger bodies being hotter and staying active longer. This trend

In contrast, the inverse of the length of the 1:1 lines depends on distance from the nebula center (Fig. 13). This important new finding corroborates that the small reservoir with shortlived isotopes was indeed at the center of the nebula, and suggests that our Solar System formed after a nova. Involvement of a second star is mandated by the existence of two dust

We provide an alternative view of the origins of meteorites and the asteroid belt, based on fundamental principles. The connection of HEDs and related meteorites with early Mercury provides a coherent view of the ~40,000 samples of meteorites, suggests new relationships with the planets, and provides a consistent picture of their formation. Our analysis removes guesswork in estimating chemical compositions for the planets and provides many testable

 Core formation involved redox reactions dominated by C, and released divalent iron into mantle minerals. This process increased the proportion of olivine to pyroxene. The 18O values of Earth are vertically zoned, with the lower mantle being lower in 18O

The oxygen isotope composition of Mercury will be shown to be similar to that of the

Asteroid 4Vesta is a mesosiderite that has undergone minor surface melting and rapid

Carbonaceous chondrites are too refractory to describe the Earth's composition.

reservoirs. The 1:1 line lengths could be derived from light-energy from the nova.

Fig. 13. Dependence of mass dependent fractionation on mass and of the inverse massindependent fractionation on distance from the center. Objects and fractionation trends as

1

0 2 10<sup>24</sup> 4 10<sup>24</sup> 6 10<sup>24</sup> **Object mass, kg**

individual chondrite types vs asteroid belt

Earth

2

Mars

Mercury

proto-Mercury

3

4

5

Spread in 18 O values

6

7

8

**EC**

Upon assembly of the planets, and heating via impacts or radioactivity, dust in the main reservoir reacted with trapped ices, exemplified by CO:

$$2\text{Fe}^{0} + \text{CO} \uparrow \text{MgSiO}\_{3} \text{ (enstatiite)} \rightarrow \text{FeC} + \text{(Mg}\_{2}\text{Fe)SiO}\_{4} \text{ (olivine)}.\tag{4}$$

This reaction converted end-member pyroxene to solid-solution olivine. The isotopic drive was towards the CO reservoir, which was rich in 16O. This reaction is germane to core formation, suggesting that carbon is the core's light element, consistent with solar abundances. Earth's core has ~10% light elements, mostly C, consistent with ubiquitous occurrence of olivine with Fe/(Mg+Fe) ~ 0.1 in meteorites, mantle samples, and terrestrial basalts. Equation 4 and similar reactions provide for enrichment of Ni accompanying incorporation of C, S, N and P in the core while lithophiles (e.g., Ge and Ga) are transferred to the silicates. Comparison of the composition of IAB to IIIAB irons associated with evolving Mercury (see Mittlefehldt et al., 1998) supports our proposal.

The small dust reservoir (CAIs, ureilites) has many phases and many reactions are possible. Because metals that condense at high *T* are rare (e.g., V, Mo, Nb, Pt group), reactions similar to Eq. 4 in the small reservoir could go to completion. Melilite could be converted to forsterite and fassitic pyroxene would be formed, for example. Without ample Fe0, the following reaction is suggested to be important with dust serving as nuclei:

$$\text{CO} + \text{H}\_2 \rightarrow \text{C} + \text{H}\_2\text{O}.\tag{5}$$

Metals such as V were incorporated in hibonite, ample graphite was produced, material was hydrated and/or carbonated, and the CAI phases were mass-independently fractionated.

Reactions like Eq. 4 and 5 proceed with time and this sequence is recorded in meteorities which were ejected from evolving Mercury and other planets as the late stages of accretion progressed. Nearly full-formed planets provided the ejecta, in accord with early assembly involving dust, with large impacts being associated with the late influx from great distance (Hofmeister & Criss, 2012). The earliest ages deduced from isotopic studies correspond to condensation of the dust reservoirs from their stellar sources. Early chondrule ages (Amelin et al., 2002) of 2.5 Ma later record the onset of large impacts, and the end of proto-planet assembly, consistent with Hofmeister and Criss' (2012) analysis using the time-dependent virial theorem.

Mixing of these two dust reservoirs created a graded solar system (Figs. 11 and 12). Because of contraction and events like the LHB, the proportion of the main (outer) reservoir increased with time. The EC line pertains to the asteroid belt, whereas the K line pertains to the dust in the inner Solar System. The current planets lie about half way between these 1:1 lines. To this, CO and H2O and other ices, which were frozen on dust drawn into protoplanets, were added in indeterminate amounts.

#### **6.3 Fractionation trends and evolutionary processes**

After assembly of the bodies and heating, mass-dependent fractionation occurs, superimposing lines parallel to TF for individual planets and large masses. All 1:2 lines for meteorite types are about the same length, consistent with similar thermal histories of the body after ejection to the asteroid belt. The longer line for the HEDs and related meteorites is consistent with origination in Mercury, given that the large Earth has an even longer 1:2 line. The degree of ordinary fractionation directly depends on the mass of the object (Fig. 13).

Upon assembly of the planets, and heating via impacts or radioactivity, dust in the main

 2Fe0 + CO +MgSiO3 (enstatite) FeC + (Mg,Fe)SiO4 (olivine). (4) This reaction converted end-member pyroxene to solid-solution olivine. The isotopic drive was towards the CO reservoir, which was rich in 16O. This reaction is germane to core formation, suggesting that carbon is the core's light element, consistent with solar abundances. Earth's core has ~10% light elements, mostly C, consistent with ubiquitous occurrence of olivine with Fe/(Mg+Fe) ~ 0.1 in meteorites, mantle samples, and terrestrial basalts. Equation 4 and similar reactions provide for enrichment of Ni accompanying incorporation of C, S, N and P in the core while lithophiles (e.g., Ge and Ga) are transferred to the silicates. Comparison of the composition of IAB to IIIAB irons associated with

The small dust reservoir (CAIs, ureilites) has many phases and many reactions are possible. Because metals that condense at high *T* are rare (e.g., V, Mo, Nb, Pt group), reactions similar to Eq. 4 in the small reservoir could go to completion. Melilite could be converted to forsterite and fassitic pyroxene would be formed, for example. Without ample Fe0, the

Metals such as V were incorporated in hibonite, ample graphite was produced, material was hydrated and/or carbonated, and the CAI phases were mass-independently fractionated. Reactions like Eq. 4 and 5 proceed with time and this sequence is recorded in meteorities which were ejected from evolving Mercury and other planets as the late stages of accretion progressed. Nearly full-formed planets provided the ejecta, in accord with early assembly involving dust, with large impacts being associated with the late influx from great distance (Hofmeister & Criss, 2012). The earliest ages deduced from isotopic studies correspond to condensation of the dust reservoirs from their stellar sources. Early chondrule ages (Amelin et al., 2002) of 2.5 Ma later record the onset of large impacts, and the end of proto-planet assembly, consistent with Hofmeister and Criss' (2012) analysis using the time-dependent

Mixing of these two dust reservoirs created a graded solar system (Figs. 11 and 12). Because of contraction and events like the LHB, the proportion of the main (outer) reservoir increased with time. The EC line pertains to the asteroid belt, whereas the K line pertains to the dust in the inner Solar System. The current planets lie about half way between these 1:1 lines. To this, CO and H2O and other ices, which were frozen on dust drawn into

After assembly of the bodies and heating, mass-dependent fractionation occurs, superimposing lines parallel to TF for individual planets and large masses. All 1:2 lines for meteorite types are about the same length, consistent with similar thermal histories of the body after ejection to the asteroid belt. The longer line for the HEDs and related meteorites is consistent with origination in Mercury, given that the large Earth has an even longer 1:2 line. The degree of ordinary fractionation directly depends on the mass of the object (Fig. 13).

CO + H2 C + H2O. (5)

reservoir reacted with trapped ices, exemplified by CO:

protoplanets, were added in indeterminate amounts.

**6.3 Fractionation trends and evolutionary processes** 

virial theorem.

evolving Mercury (see Mittlefehldt et al., 1998) supports our proposal.

following reaction is suggested to be important with dust serving as nuclei:

Fig. 13. Dependence of mass dependent fractionation on mass and of the inverse massindependent fractionation on distance from the center. Objects and fractionation trends as labeled.

This trend is consistent with larger bodies being hotter and staying active longer. This trend may aid in deciphering parent bodies.

In contrast, the inverse of the length of the 1:1 lines depends on distance from the nebula center (Fig. 13). This important new finding corroborates that the small reservoir with shortlived isotopes was indeed at the center of the nebula, and suggests that our Solar System formed after a nova. Involvement of a second star is mandated by the existence of two dust reservoirs. The 1:1 line lengths could be derived from light-energy from the nova.

## **7. Conclusions and future research**

We provide an alternative view of the origins of meteorites and the asteroid belt, based on fundamental principles. The connection of HEDs and related meteorites with early Mercury provides a coherent view of the ~40,000 samples of meteorites, suggests new relationships with the planets, and provides a consistent picture of their formation. Our analysis removes guesswork in estimating chemical compositions for the planets and provides many testable predictions to guide future work:


Origin of HED Meteorites from the Spalling of Mercury –

*Icarus*, Vol. 158, pp. 146-177, ISSN: 0019-1035

Pergamon, New York, ISBN: 0-08-028086-2

Vol. 215, (06-2011), pp. 57-61, ISSN: 0019-1035

1180, ISSN: 1086-9379

pp. 15-30, ISSN: 1529-6466

1133-1141, ISSN: 0004-6361

1086-9379

0016-7037

Houston.

Implications for the Formation and Composition of the Inner Planets 175

Burbine, T.H.; Greenwood, R.C.; Buchanan, P.C.; Franchi, I.A., Smith, C.L. (2007).

*and Planetary Science Conference*, League City, Texas, USA, 12-16 March 2007 Bus, S.J., Binzel, R.P. (2002). Phase II of the Small Main-Belt Asteroid Spectroscopic Survey.

Basaltic Volcanism Study Project (1981). Basaltic Volcanism on the Terrestrial Planets.

Cameron, A.G.W., Fegley, B., Benz, W., & Slattery, W. L. (1988). The strange density of

pp. 692-708. University of Arizona Press, ISBN 0-8165-1085-7, Tucson, AZ. Chamberlin, M.A., Sykes, M.V., Tedesco, E.F. (2011). Mid-Infared lightcurve of Vesta. *Icarus*,

Christensen, P.R., Bandfield J. L., Hamilton V.E., Howard D.A., Lane M.D., Piatek J.L., Ruff

Clayton, R.N. and Mayeda T.K. (1996) Oxygen isotope studies of achondrites. *Geochim.* 

Consolmagno, G.J., Brit, D.T. (2003). Stony meteorite porosities and densities: A review of

Criss, R.E., Farquhar, J. (2008). Abundance, Notation, and Fractionation of Light Stable

Dotto, E., Muller. T.G., Barucci M.A., Encrenaz Th. Knacke R.F. Lellouch E, Doressoundiram

Drake, M.J. (2001). The eucrite/Vesta story. *Meteor. Planet. Sci.*, Vol. 36, pp. 501-513, ISSN:

Duke, M.B., Silver L.T. (1967). Petrology of eucrites, howardites and mesosiderites. *Geochim.* 

Dymek, R.F., Albee A.L., Chodos A.A., and Wasserburg G.J. (1976). Petrography of

French, B.M. (1998) *Traces of Catastrophe: A Handbook of Shock-Metamorphic Effects in Terrestrial* 

Franchi, I.A. (2008). Oxygen Isotopes in Asteroidal Materials. *Reviews in Mineralogy & Geochemistry; Oxygen in the Solar System*, Vol. 68, pp. 345-397, ISSN: 1529-6466 Gaffey, M.J. (1976). Spectral reflectance characteristics of the meteorite classes. *J. Geophys.* 

Gaffey, M.J. (1997) Surface lithologic heterogeneity of asteroid 4 Vesta. *Icarus,* Vol. 127, pp.

Gando, A., Gando, Y., Ichimura, K. et al. (2011). Partial radiogenic heat model for Earth

revealed by geoneutrino measurements. *Nature Geoscience*, Vol. 4, (09-2011), pp.

minerals. *J. Geophys. Res*., Vol. 105, pp. 9735-9739, ISSN: 0885-3401

*Cosmochim. Acta*, Vol. 60, pp. 1999-2017, ISSN: 0016-7037

*Cosmochim. Acta,* Vol. 58, pp. 3921-3929, ISSN: 0016-7037

*Res.*, Vol. 81, pp. 905-920, ISSN: 0885-3401

130-157, ISSN: 0019-1035

647-651, ISSN: 1752-0894

Reflectance spectra of Mesosiderites: Implications for asteroid 4 Vesta. *38th Lunar* 

Mercury: theoretical considerations, In: *Mercury*, C. Chapman, C. & Vilas, F. (Eds),

S.W., and Stefanov W.L. (2000) A thermal emission spectral library of rock-forming

the data through 2001. *Meteoritics and Planetary Science*, Vol. 38, No. 8, pp. 1161-

Isotopes. *Reviews in Mineralogy & Geochemistry; Oxygen in the Solar System*, Vol. 68,

A., Crovisier J., Brucato J.R., Colangeli L., and Mennalla V. (2000). ISO results on bright Main Belt asteroids: PHT-S observations. *Astron. Astrophys.,* Vol. 358, pp.

isotopically-dated clasts in the Kapoeta howardite and petrologic constraints on the evolution of its parent body. *Geochim. Cosmochim. Acta,* Vol. 40, pp. 115-1130, ISSN:

*Meteorite Impact Structures.* LPI Contribution No. 954, Lunar and Planetary Institute,


## **8. Acknowledgment**

Special thanks to Janet E. Bowey (University College London) for help with ISO data. We thank our Washington U. colleagues (R. F. Dymek, B. Fegley, B. L. Jolliff, K. Lodders and R. Korotev) and W. Hamilton (Colorado School of Mines) for helpful discussions. Partial support was provided by NSF EAR-0757841 and AST-0908309 funds.

## **9. References**


Studies of Vesta's moment of inertia will disclose that it is internally homogeneous and

 Properly analyzed emission spectra will help correct the asteroid classification system. All differentiated meteorites will be eventually linked to planetary bodies and other

 The preSolar nebula was grossly homogeneous, but chemically and isotopically zoned in detail, with refractory elements more concentrated in the center, and volatile

Mineral dust in the preSolar nebula was dominated by pure enstatite and iron metal.

Special thanks to Janet E. Bowey (University College London) for help with ISO data. We thank our Washington U. colleagues (R. F. Dymek, B. Fegley, B. L. Jolliff, K. Lodders and R. Korotev) and W. Hamilton (Colorado School of Mines) for helpful discussions. Partial

Amelin, Y., Krot, A.N., Hutcheon, I.D., Ulyanov, A.A. (2002) Lead Isotopic Ages of

Armytage, R.M.G., Georg, R.B., Savage, P.S., Williams, H.M., Halliday, A.N. (2011) Silicon

Armitage, P.J. (2011). Dynamics of Protoplanetary Disks. *Annual Review of Astronomy and* 

Basu, S., Antia, H. M. (2008). Helioseismology and Solar Abundances. Physics Reports, Vol.

Bates, J.B. (1978) Infrared emission spectroscopy. *Fourier Transform IR Spectr.* Vol. 1, pp. 99-142 Benz, W., Slattery W.L., and Cameron A.G.W. (1988) Collisional stripping of Mercury's

Bernatowicz, T.J., Zinner, E.K. (1997). *Astrophysical implications of the laboratory study of* 

Boss, A.P. (1998). Temperatures in protoplanetary disks. *Ann. Rev. Earth Planet Sci.*, Vol. 26,

Bowey, J.E. and Hofmeister, A.M. (2005). Overtones and the 5-8 µm spectra of deeply

Brearly, A.J., Jones, R.H., 1998. Chondritic meteorites. *Rev. Mineral.*, Vol. 36, pp. 1-398, ISSN:

Brewster, M.Q. (1992) *Thermal Radiative Transfer and Properties.* John Wiley and Sons, Inc. Burbine, T.H., Buchanan P.C., Binzel, R.P., Bus S.J., Hiroi T., Hinrichs J.L., Meibom A., and

embedded objects. *Monthly Notices of the Royal Astronomical Soc.* Vol. 358, pp. 1383-

McCoy T.J. (2001) Vesta, vestoids, and the howardite, eucrite, diogenite group: Relationships and the origin of spectral differences. *Meteorites Planet. Sci.*, Vol. 36,

Chondrules and Calcium-Aluminum-Rich Inclusions. *Science*, Vol. 297 pp. 1678-

isotopes in meteorites and planetary core formation. Geochim. Cosmochim. Acta,

Many Earth-crossing orbits are not elliptical but aperiodic bounded orbits.

large objects such as the Moon that have a minimum radius of 1200 km.

support was provided by NSF EAR-0757841 and AST-0908309 funds.

*Astrophysics*, Vol. 49, pp. 195-236, ISSN: 0066-4146

mantle. *Icarus,* Vol. 74, pp. 516-528, ISSN: 0019-1035

457*,* No. 5–6, pp. 217-283, ISSN: 0370-1573

*presolar materials*. AIP, New York.

pp. 53-80, ISSN: 0084-6597

pp. 761-781, ISSN: 1086-9379

1393, ISSN: 0035-8711

1529-6466

lacks a core.

**8. Acknowledgment** 

**9. References** 

All meteorite materials are processed.

1683, ISSN: 0036-8075

vol. 75, 3662-3676

elements concentrated in outer zones.


Origin of HED Meteorites from the Spalling of Mercury –

ISSN: 0036-8075

278, ISSN: 1086-9379.

1453-1455, ISSN: 1559-128X

5865-5871, ISSN: 0885-3401

pp. 398, ISSN: 1086-9379

8075.

*Series*, Vol. 341, pp. 225-250, ISSN: 1-58381-208-3

*Astrophys.*, Vol. 372, pp. 165–172, ISSN: 0004-6361

333, pp. 1850-1852. ISSN: 0036-8075.

*Planet. Sci.,* Vol. 33, pp. 197-213, ISSN: 1086-9379

Survey." *Icarus*, Vol. 173, pp. 385-408, ISSN: 0019-1035

Implications for the Formation and Composition of the Inner Planets 177

Kemper, F., Jaeger, C., & Waters L.B.F.M. (2002). Detection of carbonates in dust shells around evolved stars. *Nature,* Vol. 415, pp. 295–297, ISSN: 0028-0836 Kitts K., Lodders, K. (1998). Survey and evaluation of eucrite bulk compositions. *Meteoritics* 

Lewis, J.S., (1974). The temperature gradient in the solar nebula. *Science*, Vol. 186, pp. 440-43,

Lim, L.F., McConnochie, T.H., Bell, J.F., Hayward, T.L. (2005) "Thermal infrared (8-13 µm)

Lodders, K. (1998). A survey of shergottite, nakhlite and chassigny meteorites whole-rock compositions. *Meteoritics Planet. Sci.,* Vol. 33, pp. 183-190, ISSN: 1086-9379. Lodders, K. (2000). An oxygen isotope mixing model for the accretion and composition of

Lodders, K. (2003). Solar system abundances and condensation temperatures of the elements. *The Astrophysical Journal*, Vol. 591, (06-2003), pp. 1220-1247, ISSN: 0004-637X. Lodders, K., Fegley, B. (1998). *The Planetary Scientist's Companion*. Oxford, New York. Love, S.G., Kiel, K. (1995). Recognizing Mercurina meteorites. Meteoritics, Vol. 30, pp269-

Low, M.J.D., Coleman, I. (1966). Measurement of the spectral emission of infrared radiation

Lucey, P.G., Keil K., Whitely, R. (1998). The influence of temperature on the spectra of the A-

MacPherson, G.J., Simon, S.B., Davis, A.M., Grossman, L., Krot, A.N. (2005). Calcium-

McKeegan, K.D., Leshin, L.A., Russell, S.S., MacPherson, G.J. (1998). Oxygen Isotopic

Melosh, H.J., Tonks, W.B. (1993). Swapping Rocks: Ejection and Exchange of Surface

Mittlefehldt, D.W., McCoy, T.J., Goodrich, C.A., Kracher, A. (1998). Non-chondritic meteorites from asteroidal bodies. *Rev. Mineral*, 36, pp. 1-195, 1529-6466 Molini-Velsko, C., Mayeda T.K. and Clayton R.N. (1986) Isotopic composition of silicon in meteorites. *Geochim. Cosmochim. Acta*, Vol. 50, pp. 2719-2726, ISSN: 0016-7037 Molster, F.J., Lim, T.L., Sylvester, R.J., Waters, L.B.F.M. Barlow, M.J., Beintema, D.A., Cohen,

Nittler, L.R. and 14 others (2011) The major-element composition of Mercury's surface from

Pepowski, P.N. and 16 others (2011) Radioactive elements on Mercury's surface from

of minerals and rocks using multiple-scan interferometry. *Appl. Opt*. Vol. 5, pp.

asteroids and implications for their silicate chemistry. *J. Geophys. Res.*, Vol. 103, pp.

Aluminum-rich Inclusions: Major Unanswered Questions. In: Krot, A.N., Scott, E.R.D., Reipurth, B. (2005). *Chondrites and the Protoplanetary Disk ASP Conference* 

Abundances in Calcium-Aluminium-Rich Inclusions from Ordinary Chondrites: Implications for Nebular Heterogeneity. *Science*, Vol. 280, pp. 414-418 ISSN: 0036-8075

Material Among the Terrestrial Planets. *Meteoritics*, Vol. 28, No. 3, vol. 28, (1993),

M., Cox, P., Schmitt, B., (2001). The complete ISO spectrum of NGC 6302. *Astron.* 

MESSENGER X-ray spectrometry. Science, Vol. 333, pp. 1847-1850. ISSN: 0036-

MESSENGER: implications for the Planet's formation and evolution. Science, Vol.

rocky planets. *Space Sci. Rev*., Vol. 92, pp. 341-354, ISSN: 0038-6308

spectra of 29 asteroids: the Cornell Mid-Infrared Asteroid Spectroscopy (MIDAS)


Goldstein, H. (1950). *Classical Mechanics.* Addison-Wesley Publishing Co., Inc., Reading,

Gomes, R., Levinson, H.F., Tsiganis, K., & Morbidelli, A., (2005). Origin of the cataclysmic

Goodrich, C.A.and Bird J.M. (1985). Formation of iron-carbon alloys in basaltic magma at

Gounelle, M., Spurny, P., Bland, P.A. (2006). The orbit and atmospheric trajectory of the

Grady, M.M. (2006). The history of research on meteorites from Mars. In: McCall, G.J.H.,

Guan, Y., McKeegan, K.D., MacPherson, G.J. (2000). Oxygen isotopes in calcium-aluminum-

Heras A.M., Morris P.W., Vandenbussche B., Müller T.G. (2000). Asteroid 4 Vesta as seen

Hiroi T., Binzel R.P., Shunshine, J.M., Pieters, C.M., Takeda, H. (1995). Grain sizes and

Hofmeister, A.M., Bowey, J.E. (2006). Quantitative IR spectra of hydrosilicates and related

Hofmeister, A.M and Criss, R.E. (2005) Earth's heat flux revisited and linked to chemistry.

Hofmeister, A.M and Criss, R.E. (2012). A thermodynamic model for formation of the Solar

Hofmeister, A.M., Rossman, G.R. (1984). Determination of Fe3+ and Fe2+ concentrations in

Hofmeister, A.M., Keppel E., Speck A.K. (2003). Absorption and reflection spectra of MgO

Hofmeister, A.M., Wopenka, B, Locock, A., (2004). Spectroscopy and structure of hibonite,

Javoy, M., 1995. The integral enstatite chondrite model of the earth. *Geophys. Res. Lett.*, Vol.

*Tectonophysics,* Vol. 395, pp. 159-177, ISSN: 0040-1951

*Cosmochim. Acta*, Vol. 68, pp. 4485-4503, ISSN: 0016-7037

late heavy bombardment period of the terrestrial planets. *Nature*, Vol. 435, pp. 466-

Uivfaq, Disko Island: the role of carbon in mafic magmas. *J. Geology*, Vol. 93, pp.

Orgueil meteorite from historical records. *Meteoritics & Planetary Science*, Vol. 41,

Bowden, A.J., Howarth, R.J. (Eds.). (2006). *The History of Meteoritics and Key Meteorite Collections: Fireballs, Falls and Finds.* Geological Society, Special Publication

rich inclusions from enstatite chondrites: new evidence for a single CAI source in the solar nebula. *Earth and Planetary Science Letters*, Vol. 181, pp. 271-277, ISSN 0012-

with the ISO short wavelength spectrometer. In: *Thermal Emission Spectroscopy and Analysis of Dust, Disks, and Regoliths.* M.L. Sitko, A.L. Sprague, and D.K. Lynch, pp.

mineral composition of Vesta-like asteroids. *Icarus,* Vol. 115, pp. 374-386, ISSN:

minerals. *Monthly Notices Royal Astronomical Society,* Vol. 367, pp. 577-591, ISSN:

System via 3-dimensional collapse of the dusty nebula. *Planetary and Space Science*,

feldspar by optical and EPR spectroscopy. *Phys. Chem. Minerals,* Vol. 11*,* pp. 213-

and other diatomic compounds. *Mon. Not. R. Astron. Soc.*, Vol. 345, No. 1, (03-2003),

grossite, and CaAl2O4: implications for astronomical environments. *Geochim.* 

Massachusetts, ISBN: 0-201-02510-8

No. 1, pp. 135-150, ISSN: 1086-9379

Grady, M.M. (2000). *Catalogue of Meteorites*. Cambridge University Press.

469, ISSN: 0028-0836

475-492, ISSN: 0022-1376

256. 13: 978-1-86239-194-9.

205-212, ISBN: 978-1-58381-532-8

ISSN 0032-0633 Vol. 62, pp. 111-131

22, pp. 2219-2222, ISSN: 0094-8276

224, ISSN: 0342-1791

pp. 16-38, ISSN: 0035-8711

821X

0019-1035

0035-8711


**8** 

*Russia* 

**The Eruptions of Sarychev Peak Volcano,** 

**Kurile Arc: Particularities of Activity** 

Alexander Rybin1, Nadezhda Razjigaeva2, Artem Degterev1,

More then 68 quaternary volcanic edifices are picked out in the Kurile Islands, among them 36 are active and potentially dangerous. During 45 thousand years not less then 12 large explosive eruptions, connected with the formation of the calderas, have occurred. During these eruptions a great amount of ash emitted and large pumice-pyroclastic covers formed,

During the historical time (about 250 years for the Kurile Islands) about 29 great and catastrophic eruptions were fixed (Gorshkov, 1967). In XXth century the volcanoes of the Central and Northern Kurils were the most active and productive by the volume of the erupted material. These tendencies have continued in current century. Explosive eruptions occurred many times on Chikurachki volcano (Girina et al., 2008). Phreatic and phreato-magmatic explosions were detected on the Ebeko, Berga, Chirinkotan, Severgin

Because of rare population of the Kurile Islands the damage at the eruptions was not so considerable as it usually was in densely populated island countries such as Indonesia, Japan and others. However, some cases were written in the historical documents, when the eruptions in the Kurile Islands caused the material damage and victims. In 1778 during the strong eruption of Raikoke volcano 15 Russian manufacturers died. There are the historical materials about the destruction of ainu settlements in Shiashkotan Island after the eruption of Sinarka volcano (Gorshkov, 1967). In 1933 at the eruption of Severgin volcano in Harimkotan Island the Japanese settlement was destroyed, tsunami waves formed by the eruption, caused the death of several people in the neighboring islands Onekotan and

During the eruptions of Sarychev Peak volcano in 1946 and 1973 the evacuation of military unit was conducted in Matua Island. At Tyatya volcano eruption in 1973 the military camp

that caused the climate and landscape changes (Melekestsev et al., 1988).

**1. Introduction** 

volcanoes.

Paramushir (Miyakate, 1934).

 **and Influence on the Environment** 

*1Institute of Marine Geology and Geophysics of Far East Branch,* 

Kirill Ganzey2 and Marina Chibisova1

*of Russian Academy of Sciences, Yuzhno-Sakhalinsk, 2Pacific Institute of Geography of Far East Branch, of Russian Academy of Sciences, Vladivostok,* 

