Isotope Geochemistry: The Anomalous Earth - W Isotopes and the Formation of the Moon, Study notes of Geochemistry

Download Isotope Geochemistry: The Anomalous Earth - W Isotopes and the Formation of the Moon and more Study notes Geochemistry in PDF only on Docsity! Isotope Geochemistry Chapter 5 127 3/2/11 At that point, the Earth appeared to be puzzlingly anomalous among differentiated planetary bodies in that silicate-metal differentiation appeared to have occurred quite late. Subsequently, Yin et al. (2002) reported W isotope measurements carried out in two laboratories, Harvard University and the Ecole Normale Supérieure de Lyon, which showed that the chondrites Allende and Murchison had W isotope ratios 1.9 to 2.6 epsilon units lower than the terrestrial standard (Figure 5.10). Yin et al. (2002) also analyzed separated metal and silicate fractions from two ordinary chondrites (Dhurmsala and Dal- gety Downs) that allowed them to estimate the initial 182Hf/180Hf of the solar system as 1 × 10-4. In the same issue of the journal Nature, Kleine et al., (2002) reported similarly low εW (i.e., -2) for the carbona- ceous chondrites Allende, Orgueil, Murchison, Cold Bokkeveld, Nogoya, Murray, and Karoonda measured in a third laboratory (University of Münster). Furthermore, Kleine et al. (2002) analyzed a variety of terres- trial materials and found they all had identical W isotopic composition (Figure 5.10). It thus appears that the original measurements of Lee and Halliday (1995) were wrong. The measurement error most likely relates to what was at the time an entirely new kind of instrument, namely the multi-collector ICP-MS. Yin et al. (2002) considered two scenarios for the formation of the core. In the first, which they call the two-stage model in which the Earth first accretes (stage 1) and then undergoes core formation (stage 2), induced by the giant impact that forms the moon. In this scenario, core formation occurs 29 million years after formation of the solar system. In the second scenario that they believed more likely, metal segregates continuously from a magma ocean. In this continuous model, the mean age of core forma- tion is 11 million years. In contrast, they concluded that the parent body of the eucrite class of achon- drites (suspected to be the large asteroid Vesta) underwent core forma- tion within 3 million years of formation of the solar system. Klein et al. (2002) reached similar conclusions.Yin et al. (2002) considered two scenarios for the formation of the core. In the first, which they call the two-stage model in which the Earth first accretes (stage 1) and then undergoes core formation (stage 2), induced by the giant impact that forms the moon. In this scenario, core formation is inferred from an “isochron” drawn through the ter- restrial, lunar, and chon- dritic W isotopic compo- sitions. That isochron has an age correspond- ing to 29 million years after formation of the so- lar system. In the second scenario, which they be- lieved more likely, metal segregates continuously from a magma ocean. In this continuous model, the mean age of core formation is 11 million years. In contrast, they concluded that the par- ent body of the eucrite class of achondrites (sus- pected to be the large asteroid Vesta) un- derwent core formation within 3 million years of Figure 5.10. W isotope ratios measured in chondrites, the iron meteorite Toluca, and terrestrial materials. Data from Kleine et al. (2002), Yin et al (2002), Touboul et al (2007, 2009). Docsity.com Isotope Geochemistry Chapter 5 128 3/2/11 formation of the solar system. Klein et al. (2002) reached a similar conclusion that the Moon and the Earth’s core formed early. The next twist in the story came when it was recognized that 182W could be produced by beta decay of 182Ta which in turn is produced cosmogenically through the reaction 181Ta(n,γ)182Ta. Because the Earth’s surface is somewhat protected from cosmic rays by the atmosphere, this reaction is entirely trivial at the surface of the Earth. However, the Moon has no atmosphere. Furthermore, the lack of lunar geologic activity means that materials are exposed at the surface for very long times (hundreds of millions of years and more). Consequently, the cosmogenic production of nuclides in lunar materials can be sig- nificant. Early studies failed to appreciate this or failed to fully correct for this effect. Touboul et al. (2007), working with W-rich, Ta-poor metal grains separated from lunar samples, showed that previous high εW values in lunar samples were due to cosmogenic effects and when these were eliminated the Moon has and isotopically homogeneous W isotopic composition that is identical to that of the silicate Earth – about 2 epsilon units higher than chondrites. The homogeneous isotopic composition of lunar materials Moon, including samples with highly variable Hf/W ratios such as anorthosites from the ancient lunar crust and mare basalts formed by later melting of the lunar mantle, implies that 182Hf had effectively decayed completely before the Moon formed (or at least before its magma ocean solidified). Given the analytical precision with which W iso- topes can be measured, this implies an age of the Moon of greater than 60 Ma after the beginning of the solar system (the age of Allende CAI’s). The minimum age of the Moon is constrained by Sm-Nd ages of the lunar anorthositic crust. This age is 4.456±0.040 Ga, or about 100 Ma after the beginning of the solar system. Touboul et al. (2007) thus estimate an age for the Moon of 62 +90/-10 Ma after the start of the solar system. There is a very broad consensus among scientists that the Moon formed as a result of a collision be- tween a nearly full-sized Earth and a second body (sometimes called Thera) about the size of Mars (the Giant Impact Hypothesis). Debris thrown into orbit about the Earth as a result of the impact subse- quently coalesced to form the Moon. A collision between bodies of these sizes releases enormous amounts of energy; depending on how efficiently the gravitational-kinetic energy is converted to heat, potentially enough energy to entirely melt the Earth. At the very least, one expects significant melting would have occurred on the Earth and one of the virtues of the hypothesis is that it explains the exten- sive melting experienced by the Moon. (Because it has remained geologically active, all evidence of a magma ocean on the Earth has been erased; because it quickly became geologically inactive, evidence of a magma ocean on the Moon has been preserved.) The significance of this is that such an impact would have provided a last opportunity for the metal to segregate from the mantle and sink into the core (a process that almost certainly requires the metal, if not also the silicate, to be liquid). We expect therefore, that the Moon-forming giant impact should also be the time of final separation of the core. If, however, the core formed in a single event 60-100 Ma after the start of the solar system, we would that the silicate Earth would have a W isotopic composition very close to chondritic; instead the terres- trial value is 2 epsilon units higher than chondritic. Furthermore, if, as silver isotopes clearly reveal, cores of asteroids formed within 10 Ma of the start of the solar system, why was core formation on the Earth so delayed? We can reconcile these observations if we assume that core formation was a more or less continuous process that began very early and only ended with the giant impact. In essence, the scenario goes like this. Once planetesimals exceed a few 10’s of km in radius, heating, either from 26Al or release of gravitational energy in collisions, caused melting that allowed and metal and silicate to separate, with the metal sinking to form cores of the planetesimals. As the bodies grew into larger ones through collision with other planetesimals, enough energy was released to allow the cores of colliding bodies to merge. This process, known as oligarchic growth, continued to build larger and larger plane- tary embryos. As the embryos merged through collision, fewer and fewer embryos remained and the collisions were less frequent. Slowing the planetary growth process, but the collisions, when they did occur, were far more energetic. The process, in the case of the Earth, culminates when the last two bod- ies in the Earth’s orbital neighborhood (or “feeding zone”) collide in the giant impact that forms the Docsity.com Isotope Geochemistry Chapter 5 Spring 2010 131 3/2/11 it would have to be 10% higher. If the increase occurred later, the Sm/Nd ratio would have to be even higher. This increase in Sm/Nd might seem small; after all, we have already stated that the assumption that the Earth has chondritic abundances of refractory elements is probably only valid to 10%. Yet this small difference is very important in interpretation of Nd isotope systematics. For the two scenarios above, 5 Ma and 30 Ma, the εNd of the accessible Earth would be +6.9 and +11 respectively. These val- ues fall within the range of values of mid-ocean ridge basalts. Recalling that the Isua samples have a 30 ppm excess in 142Nd relative to a terrestrial standard∗, this means that the Isua samples have a 50 ppm excess in 142Nd relative to chondrites. How might the increase in Sm/Nd come about? First, we need to recall that meteorites come from the asteroid belt and their compositions might not be representative of the composition of the inner so- lar nebula from which the Earth and the other terrestrial planets formed. Its possible the inner solar nebula had a higher Sm/Nd ratio. That said, it is very difficult to see why this should be so. The ob- servable fractionation in primitive meteorites relates to volatility and lithophile/siderophile tendency. As we have seen, Sm and Nd have quite high and very similar condensation temperatures and neither shows a significant siderophile tendency. Although the possibility cannot be excluded, there is simply no good reason to believe that the Sm/Nd ratio of the material from which the Earth should be differ- ent from chondrites. It is also possible that the early solar system was isotopically heterogeneous and that the material from which the Earth formed was richer in either 146Sm or 142Nd than the material that formed chondrite parent bodies. Andreasen and Sharma (2006) and Carlson et al. (2007) found that carbonaceous chondrites have a roughly 100 ppm deficit of 144Sm. This is significant because while 142Nd is primarily a s-process element produced in red giants, both 146Sm and 144Sm are p-process-only isotopes produced in supernova explosions. Incomplete
mixing
of
material
from
different
stars
thus
 could
result
in
variations
in 142Nd/144Nd in the early solar system. However, Andreasen and Sharma (2006) concluded that while the observed isotopic heterogeneity in 144Sm implied enough variation in 146Sm to explain 142Nd/144Nd variations observed among chondrites, it could not explain the difference between the Earth and chondrites. At present, there are two principal theories to explain the difference between terrestrial and chon- dritic 142Nd/144Nd. The first postulates that the bulk Earth does have a chondritic Sm/Nd ratio (and chondritic 142Nd/144Nd), but that low Sm/Nd and 142Nd/144Nd material is sequestered in a hidden res- ervoir deep in the mantle and that consequently the observable Earth has high Sm/Nd and high 142Nd/144Nd. The second postulates that the Earth has high, non-chondritic Sm/Nd and 142Nd/144Nd as a consequence of loss of low Sm/Nd material during its accretion as a consequence of collisional ero- sion. Let’s now explore these two possibilities. The first possibility was suggested by Boyet and Carlson (2005). They suggested that Earth under- went early differentiation forming an early enriched reservoir, such as a primordial crust that sank into the deep mantle and has not been sampled since. This differentiation might have occurred as a conse- quence of crystallization of a terrestrial magma ocean analogous to the lunar magma ocean. Alterna- tively, crystallization of the terrestrial magma ocean might have left a layer of residual melt, similar to the KREEP source on the Moon. Boyet and Carlson (2005) noted that if it were rich in Fe and Ti, as is the lunar KREEP reservoir is, once crystallized, the EER could have sunk into the deep mantle, where it remains because if its high density. As Boyet and Carlson (2005) point out, this early enriched reservoir must have formed in the upper mantle. Below the 660 km discontinuity, Mg- and Ca-perovskite would crystallize and fractionate incompatible elements in a manner much different than observed. The second possibility, “collisional erosion”, has been suggested by Caro et al. (2008), O’Neill and Palme (2008), and Caro and Bourdon (2010). As discussed earlier, planetary bodies are thought to form through the process of “oligarchic growth”. The initial stages of this process involve aggregations of dust-sized particles to form sand, which in turn aggregate to form pebble-sized particles, etc. The final ∗ The two standards commonly used in Nd isotope ratios measurements are the “La Jolla” standard and the “Ames” standard. Both are solutions created from industrially purified Nd. Docsity.com Isotope Geochemistry Chapter 5 Spring 2010 132 3/2/11 stages of this process involve infrequent, very energetic collisions between large bodies. Sufficient en- ergy is released in these collisions that the growing planet extensively melts. Between collisions, one might reasonably expect a primitive basaltic crust to form through crystallization at the surface. Caro et al. (2008), O’Neill and Palme (2008), and Caro and Bourdon (2010) suggest that a substantial fraction of this crust was blasted away in these collisions, leaving the Earth depleted in elements that were concen- trated in that crust: incompatible elements. The existence of meteorites from the Moon and Mars dem- onstrates that collisions can indeed accelerate fragments of a planet’s surface to escape velocity. Addi- tional evidence of the erosive capacity of these collisions is the large crater in the southern hemisphere of the large asteroid Vesta discovered by the Hubble Space Telescope. The volume excavated by this crater is roughly 1% of the total mass of Vesta, most of which appears to have been lost from the aster- oid. Another observation that supports this idea is the density of the planet Mercury, which is 5427 kg/m3, implying it has a much larger iron core than the other terrestrial planets, occupying 75% of its diameter (in contrast to a little of half for the Earth). An alternative way of looking at this is Mercury’s mantle is anomalously small. It has been proposed that Mercury lost much of its silicate mantle through as a consequence of a giant collision (Benz et al., 1988). Common to both hypotheses is the idea that planetary melting and consequent differentiation begins during, rather than after, planet accretion. Both hypotheses rely on the idea of formation, through melt- ing and fractional crystallization, of a primitive crust enriched in incompatible elements. Such a crust would have a low Sm/Nd ratio, leaving the remainder of the planet with a higher Sm/Nd ratio than the material from which it accreted. In the Boyet and Carlson (2005) hypothesis, this early crust sinks into the deep mantle where it remains as an isolated reservoir. In collisional erosion hypothesis, this early crust is lost from the Earth. Several observations suggest the latter hypothesis is preferable. The first of these observations is that both the Moon and Mars also appear to have 142Nd/144Nd ratios higher than chondritic. All lunar rocks are ancient, so not surprisingly, they have variable 142Nd/144Nd ratios. When these ratios are plotted against 147Sm/144Nd they display a correlation indicating the bulk Moon has 142Nd/144Nd about 17 ppm higher than the chondritic value, identical, within uncertainty, to the modern terrestrial value (Boyet and Carlson, 2008). These lunar data are particularly significant. First, the similarity of 142Nd/144Nd in the Earth and Moon imply that the Sm/Nd ratio of the Earth- Moon system was fixed before the giant impact that formed the Moon. One reaches the same conclu- sion when one considers the timing of the Moon forming event. As we saw in the previous section, tungsten isotope ratios of the lunar rocks are uniform and indistinguishable from those of the Earth, in- dicating the Moon-forming event occurred at least 50 and as much as 150 million years after the start of the solar system. Unreasonably high Sm/Nd ratios would be required to explain the 142Nd/144Nd of the Earth and Moon if the fractionation occurred that late. Thus the early enriched reservoir hypothesized by Boyet and Carlson must predate the giant impact. It is very unlikely this reservoir could have sur- vived this event without being remixed into the mantle (Caro and Bourdon, 2010). 142Nd/144Nd and 147Sm/144Nd ratios in shergottites (meteorites from Mars) suggest the Martian 142Nd/144Nd ratio is, like the Earth and Moon, 10 to 20 ppm higher than chondritic and Caro et al., (2008) argue that this suggests that all planetary objects in the inner solar system have Sm/Nd ratios higher than chondritic. However, the other Martian meteorites, the naklites and Chaussigny plot above this correlation. Caro et al. (2008) explain this by suggesting that the Martian mantle source of these meteorites underwent later differentiation. However, since there are multiple possible interpretations of the data, the Martian evidence must be considered equivocal at best on this point. Geophysical considerations also cast doubt on the existence of a highly incompatible element- enriched reservoir in the deep mantle. In order to explain the 6% increase in Sm/Nd ratio in the re- maining mantle, this reservoir would have to contain at least 40% of the Earth’s inventory of highly in- compatible lithophile elements. This group of elements includes the heat-producing elements K, U, and Th. Thus the early enriched reservoir at the base of the mantle would be responsible for some 40% of the heat production in the Earth, and 70% or more of the heat production in the mantle. If this were the case, the mantle should be heated mainly from below. As Davies (2009) points out, this would pro- Docsity.com Isotope Geochemistry Chapter 5 Spring 2010 133 3/2/11 duce a very different style of convention than actually occurs. Convective layers heated mainly from below are dominated by plumes that initiate as instabilities at the base of the layer. While some plumes do form at the base of the mantle and rise through it, the dominant form of convection in the Earth’s mantle is plate tectonics, which is the kind of convection expected in systems heated from within (lava lakes, for example, convect in a way similar to plate tectonics). At present, the collisional erosion hypothesis seems to provide the best explanation for the observa- tions we now have. However, regardless of which of these explanations is correct and whether an early-formed crust was eroded by collisions or sunk into the deep mantle, the implications of the non- chondritic Sm/Nd ratio of the Earth are profound. Both hypotheses imply that at least the observable Earth is depleted in the incompatible elements that would have been concentrated in that early crust. This would include the heat producing elements, K, U, and Th, so there are geodynamic implications as well. Origin of Short-lived Nuclides There is some debate over exactly how the short-lived radionuclides were synthesized. As we saw earlier in this chapter, heavy element nucleosynthesis occurs mainly in red giant stars and in superno- vae. Anomalous isotopic compositions of stable elements, which we discuss below, provides clear evi- dence that meteorites contain some material synthesized in both these environments. However, they provide no constraints on when this happened. Only with the unstable nuclides can we address the question of when. On galactic scales of time and space, red giants and supernovae continually inject newly synthesized elements into the interstellar medium. Those nuclides that are unstable will steadily decay away. These two competing processes will result in steady-state abundance of these nuclides in the interstellar medium. The abundances of 107Pd, 129I, and 182Hf listed in Table 16.1 roughly match the expected steady-state galactic abundances and hence do not require a specific synthesis event. How- ever, the abundances of shorter-lived 10Be, 26Al, 41Ca, 53Mn or 60Fe in the early Solar System requires syn- thesis of these nuclides at the time of, or just before, Solar System formation. The conventional view is that these nuclides were synthesized in a red giant and/or a supernova in the region where the solar system formed just shortly before its formation. Some of these elements, such as 26Al are most efficiently synthesized in red giants; others, such as 60Fe are most efficiently syn- thesized in supernovae. Thus most models invoke both environments, which may or may not have been the same star at different times. From an astronomical perspective, nucleosynthesis shortly before the solar system formed is not surprising: stars usually form not in isolation, but in clusters in large clouds of gas and dust known as nebulae. The Great Nebula in Orion is a good example. Some of the stars formed in these stellar nurseries will be quite large and have short lifetimes, ending their existence in supernova explosions. Thus stellar death, including the red giant and supernova phases, goes on simultaneously with star birth in these nebulae. Indeed, one popular hypothesis is that the formation of the solar system was actually triggered by a supernova shock wave. Boss and Vanhala (2001) pro- vide a good discussion of this view. Evidence of the existence of 10Be in some CAI’s has led to an alternative hypothesis, namely that many of the short-lived extinct radionuclides were produced by spallation within the solar system as it was forming. As we have seen, 10Be is not synthesized in stars, hence it presence in CAI’s and other primitive chondritic components is problematic for the red giant/supernova injection hypothesis. An- other key observation is that young protostars emit X-rays. X-rays are produced by accelerating charged particles. Hence some have suggested that near the surface of the accreting protosun, mag- netic reconnection events could produce flares that accelerate ions up to very high energies – essentially creating cosmic rays. Spallation would occur when the accelerated particles encounter dust grains – the CAI’s – that happen to be close to the forming Sun (within 0.1 AU). According to this theory, some of these irradiated CAI’s would have been carried back out to the vicinity of the asteroid belt by the ener- getic “X-winds” that are associated with these protostars. This theory can readily account for the abundances of 10Be, 26Al, 41Ca, and 53Mn observed. If it is correct, it solves the problem of the age gap between CAI’s and chondrules. Based on their apparent 26Al/27Al ratios, CAI’s appear to be several Docsity.com