How did they form?
The HED meteorite clan may represent a group of meteorites all related by differentiation and igneous events early in the solar system. Given their potentially simple and straight forward conditions of formation - low pressure, dry and restricted fO2 (Hewins and Newsom, 1988) - they offer a chance to understand this early and possible widespread kind of activity in the early solar system. Despite this possibility, there have been many ideas proposed to explain the variation in lithologies and compositions found within the clan. This section attempts to summarize these ideas and highlight the outstanding issues for resolution.
Experimental petrology:
The HED meteorites have been a testing ground for experimental petrologists since the early 1970s. The partial melting model of Stolper (1977) was based on experiments performed at 1 bar on several eucritic starting materials. The idea that eucrites represented partial melts of chondrites came from the observation that many eucrites plotted near the eutectic of chondritic systems, as expected of a partial melt in a multi-component system (Figure 22). Studies by Walker et al. (1975) at the same time illustrated that the textures found in basaltic eucrites could be formed by slow cooling of basaltic liquids as expected in a shallow intrusion or surficial lava flow.
Figure 22: Changes in the phase equilibria of melting at pressures between 0.5 and 1 kb, relevant to the origin of eucrites and diogenites. Figure is from Grove and Bartels (1992).
Later work by Jurewicz et al. (1993) demonstrated that angritic basaltic meteorites could be formed by melting of a carbonaceous chondrite at oxidizing conditions, while eucritic melts could be formed by melting under reduced conditions. Additional work placed constraints on the dependence upon chondritic starting materials - ordinary versus carbonaceous (Jurewicz et al., 1995).
Attempts to relate eucrites and diogenites by a single crystallization sequence, as suggested by Delaney et al. (1983) and Hewins and Newsom (1988), were supported by melting experiments on eucrites and diogenitic parent liquid at elevated pressures by Bartels and Grove (1991) and Bartels and Grove (1992). These experiments showed that elevated pressure is required for a simple crystallization sequence, but of course the central pressure of Vesta is ~1.5 kb so there is a limit to the leverage pressure can have on such a small asteroid sized body (Figure 23). Nonetheless, this pressure range makes olivine and pyroxene a cotectic rather than a peritectic (reaction) boundary in a relevant depth and has become an important concept in unraveling the history of these igneous rock types.
Figure 23: Changes in the phase equilibria of melting at pressures between 0.5 and 1 kb, relevant to the origin of eucrites and diogenites. Figure is from Grove and Bartels (1992).
Another early model advanced for explaining the eucrites or relations between the HED meteorites is an origin by fractional crystallization (Mason, 1962; Ruzicka et al., 1997; Figure 24). Although this could not connect the diogenites and eucrites, it was a way of producing a range of eucritic liquids from a more magnesian (or chondritic) parent. Later models focused on the process of partial melting of chondritic material (Stolper, 1977; Consolmagno and Drake, 1977). Again, no clear connection was made between eucrites and diogenites in partial melting models and usually they require separate formation and origin (e.g., Shearer et al., 1997). A possible connection between eucrites and diogenites was proposed by Bartels and Grove (1991) by appealing to the shift from a peritectic to cotectic boundary between olivine and pyroxene at slightly elevated pressure. At 1 kb, pressure at the center of Vesta, the reaction boundary disappears making it possible to relate the eucrites and diogenites by fractional crystallization.
Figure 24: cartoon illustrating the concept of fractional crystallization for the generation of diogenites followed by eucrites (from Shearer et al., 1997).
An unresolved issue with many of these prior formation scenarios involved adequate heat sources for the multiple melting episodes required (e.g., Hewins and Newsom, 1988). For example, core formation, eucrites genesis, and diogenite genesis could all potentially require their own heat source since they may have occurred at three different times. Righter and Drake (1997) proposed a solution to this problem by having the chondritic HED parent body start completely molten to form a metallic core, crystallize to form the diogenites and eucrites during the cooling sequence (and taking advantage of the phase equilibria changes at elevated pressure), and finally allowing later fractionation in the crust to form more evolved (Nuevo Laredo trend) eucrites (Figure 25). This solution formed the eucrites and diogenites in the same sequence, but diogenites are not directly formed from eucritic parent melts, but rather from a diogenitic parent liquid that was consumed during the equilibrium crystallization process. The modeling also took into account dynamic constraints such as convective lock-up in the magma ocean that may become important in the later stages of crystallization, and also satisfied basic geochemical constraints such as Fe/Mn and O isotopes for the origin of HED meteorites (Righter and Drake, 1997; Boesenberg and Delaney, 1997).
Figure 25: Stages in the evolution of asteroid 4 Vesta (HED parent body) from model of Righter and Drake (1997). All rock types are produced during a single cooling trend (from initially molten state) and later crustal metamorphism and impact gardening to form breccias).
The general model of Righter and Drake (1997) was not able to explain all compositional variation represented in HED meteorites in our collections, and some notable exceptions led Barrat et al. (2007) and (2010) to propose that assimilation of crustal materials was important to both eucrites formation and diogenite parent liquids (Figure 26). Assimilation of crustal material seems energetically unavoidable in large magmatic systems that are undergoing fractionation as some point, even if it is in the final stages (e.g., Carmichael et al., 1974; p. 68).
Figure 26: Cartoon of model proposed for origin of some eucrites by assimilation of crustal material, in order to exaplin REE data (Barrat et al., 2007).
Crust formation
The igneous scenarios described in the previous section can explain many primary features of eucrites and diogenites, but many of these samples have been affected by later metamorphism due either to burial and thermal conditions, or to shock. The layered crust model for the HED meteorites, proposed by Takeda (1979), explains many complicated features recorded in the meteorites such as the homogeneous Mg# of the diogenites, pyroxene exsolution features in diogenites and eucrites, and some of the younger ages obtained on eucrites by isotopic studies (Figure 27). Shock resetting has become an accepted feature of eucrite Ar-Ar age spectra (Bogard and Garrison, 2003).
Figure 27: Young ages obtained for some HED meteorites by Ar-Ar and Rb-Sr dating approaches. Young ages are thought to be due to impact and/or thermal resetting (from Bogard and Garrison, 2003).
Evolution of thought based on growth of collections, and outstanding problems
The classic falls of the HED clan had an influence on early models for the origin of these meteorites. Meteorite such as Sioux County, Juvinas, Stannern, and Johnstown were readily available and helped define many of the key issues surrounding ideas for the origin of HEDs. Subsequently, the discovery of many new HEDs in Antarctica led to detailed studies of new sample suites and a realization that simple models may not suffice. After some time it was realized also that there were fundamental differences between the collection of falls and Antarctic meteorites (Takeda, 1991). For example, the standard Stannern and Nuevo Laredo eucrite trends had been verified to some extent, but there are recently collected samples defining new trends or falling off the main trends (e.g. Ruzicka study). With the numerous new samples coming from the hot desert localities, the field is again going through a transformation because the new samples are providing additional new insights into the compositional and chemical diversity among the HED meteorites.
Issues still requiring attention are many and few will be highlighted here. First, it has been recognized for some time that the HED clan is volatile element depleted. They cannot be formed from chondritic material that has its full complement of volatile elements (e.g., Righter and Drake, 1997). This was a feature of that model that was not addressed, and remains unsolved. A second example is the unclear connection between the mesosiderites and HED meteorites. We have close to 1000 samples in our world collections and statistically it seems like there should be more be more evidence of a connection between these two groups. Perhaps additional samples or DAWN observations will shed light on this issue. Third, connections between the diogenites and eucrites could be made with detailed chronologic data. Because diogenites have been difficult to date due to their dominantly orthopyroxene mineralogy, there are sparse diogenite ages that are of comparable quality to those measured from eucrites (e.g., Schiller et al., 2010). Fourth, new data from HED meteorites has shown that the simple trends identified based on bulk composition (Stolper, 1977) or lithologic-type (Ikeda and Takeda, 1985) do not include all variation observed among the HED groups (also Mittlefehldt et al., 2009; Figure 28). The significance of this variation should be explored and resolved. Finally, the inner solar system is thought to have experienced a heavy and late bombardment causing cratering and shock effects in many samples and the HED meteorites may record some of this history. Continued studies may help to integrate the bombardment history of the Earth-Moon system and Mars with that of a region of the asteroid belt.
Figure 28: Models for the origin of diogenites based on Sm vs. Sc . Mixture tracks show the effects of adding 5%, 15% and 30% trapped melt. The plotted Sm content of MET 00424 is a 2σ upper limit. Additional labeled meteorites: M2 - MET 00422; M4 - MET 00424; M6 - MET 00436 (Mittlefehldt et al., 2009).