LECTURA 2

As we have seen, radiogenic and trace element evidence strongly constrain observed mantle heterogeneity to have developed in the outer part of the Earth: above the 670 km discontinuity. These data are more equivocal as to whether material at the Earth’s surface has played a role. Stable isotopes can answer this question. The reason is, as noted earlier, that fractionation of isotopes between different species decreases with the square root of temperature and becomes nearly negligible in most cases at mantle temperatures. Oxygen isotope fraction- ations of greater than one per mille or so occur only when rock interacts with water at or near the surface of the Earth; hence oxygen isotopes are potential tracers of material that was once at the surface of the Earth. In addition, unique mass-independent fractionation of sulphur isotopes occurs only in the atmos- phere, so those too provide tracers of surficial material in the mantle. Lithium, carbon, boron, nitrogen, magnesium, sulphur, chlorine, and iron isotopes also suggest that surficial material has found its way into the mantle.

Oxygen isotopes significantly different from the mantle value were found in subduction-related volcanics, suggesting sediment is being subducted to depths of 100 km or more (e.g., Magaritz et al., 1978). Furthermore, oxygen isotope ratios that are both well above and well below the mantle value (~+5.5 ‰) have been found in eclogite xenoliths from kimberlites (e.g., MacGregor and Manton, 1986; Pernet-Fisher et al., 2014; Smart et al., 2014). Eclogitic diamonds also show highly variable carbon (Mattey, 1987) and nitrogen (Marty and Dauphas, 2003) isotopic compositions. These variations strongly suggest that the eclogitic precursors

originated at or near the surface, most likely as oceanic crust, and had been carried into the upper mantle in ancient subduction zones and stranded in the subcontinental lithosphere. Walter et al. (2011) reported d13_{C between –15 ‰ and}

–24 ‰ in several diamonds that contain inclusions matching the compositions of basaltic phases in the lower mantle. Such carbon isotopic compositions strongly suggest an origin as biologically produced organic matter, which had apparently been subducted into the deep mantle and subsequently returned to the shallow mantle in the Cretaceous by a mantle plume rising beneath Brazil.

Perhaps the most dramatic discovery linking diamonds to the Earth’s surface was mass-independent fractionation of sulphur in diamond sulphide inclusions (Farquhar et al., 2002). Most isotopic fractionation is mass dependent, which is to say that the extent of the fractionation depends on the difference in mass between the two isotopes; thus, for example, the variation of the 33_S/34_{S ratio}

is about half that of the 34_S/32_{S ratio. Mass independent fractionation (MIF) refers}

to fractionation where this relationship does not hold and ∆33_{S, the per mille devi-}

ation from predicted mass-dependent fractionation, is a measure of this effect. Mass-independent oxygen isotope fractionation is observed in the modern stratosphere and ultraviolet mass independent sulphur isotope fractionation of sulphur species has been demonstrated in laboratory experiments (see review of Thiemens, 2006). MIF sulphur is common in sedimentary and hydrothermal sulphides and sulphates in rocks older than 2.3 Ga, but disappears completely by 2.0 Ga (Farquhar et al., 2000; Farquhar and Wing, 2003). There is a host of other evidence suggesting that significant amount of atmospheric oxygen first developed just prior to 2.3 Ga (reviewed in Kasting and Catling, 2003). Farquhar and Wing reasoned that in the absence of oxygen and ozone, ultra- violet radiation could penetrate through the whole atmosphere and fractionate atmospheric sulphur, but the development of stratospheric ozone around 2.3 Ga severely limited the penetration of ultraviolet radiation into the troposphere, ending mass-independent fractionation of sulphur. Thus the diamonds analysed by Farquhar et al. (2002) apparently contain sulphur that had cycled through the atmosphere in the Archean or earliest Proterozoic and was subsequently subducted and stored in the subcontinental lithosphere.

Stable isotope ratios in diamonds and eclogites thus clearly confirm the subduction of crustal material into the shallow mantle, but is there evidence that it can be carried into the deep mantle? Eiler et al. (1997) found significant oxygen isotope variations in some oceanic island basalts. Unlike the Ito et al.

(1987) study of oxygen isotopes in MORB described earlier, Eiler et al. analysed oxygen in individual olivine phenocrysts, which has the advantage of avoiding weathering effects. There is about a 0.5 ‰ fractionation of d18_{O between olivine}

and basaltic liquid, so that mantle olivine has d18_{OSMOW ≈ +5.2 ‰ compared to}

the average MORB value of ≈ +5.7 ‰. Eiler et al. reported d18_O

SMOW values as

high as 6.1 ‰ in basalts from the Society Islands. Basalts from these islands have particularly high 87_Sr/86_{Sr and low e}

Nd and were the ones that White and

Sedimentary materials typically have high d18_O

SMOW, as high as +20 ‰ in shales

or +30 ‰ in limestones. Assuming a d18_O

SMOW of +15 ‰, Eiler et al. calculated

that the mantle source of Society Islands lavas contained up to 5 % of a sedi- mentary component. Elevated d18_{O also occurs in Samoan peridotite xenoliths}

and Eiler et al. suggested that they too could contain a recycled sedimentary component. Subsequently, Workman et al. (2008) reported variations of d18_{O in}

olivines in Samoan lavas that correlated positively with 87_Sr/86_{Sr and} 207_Pb/204_Pb

and incompatible element ratios, confirming the presence of recycled material in the source of Samoan lavas (Fig. 6.8).

Figure 6.8 87_Sr/86_{Sr in Samoan lavas plotted against oxygen isotope ratios measure in}

olivine phenocrysts. Line shows a mixing model between clastic sediment addition (d18_{O = 25 ‰, Sr = 250 ppm,} 87_Sr/86_{Sr = 0.75) to the least enriched}

Samoan mantle component in Ta’u lavas (data from Workman et al., 2008).

Eiler et al. (1997) also found that some basalts from other island chains, particularly those with the HIMU isotopic signature, had low d18_O

SMOW. Low

d18_{OSMOW can be produced by hydrothermal interaction between water and rock,}

and Eiler et al. suggested that the low values reflected assimilation of hydrother- mally altered material in the volcanic edifice. However, the remarkable discovery of MIF sulphur in olivine sulphide inclusions from one of these islands, Mangaia of the Cook-Austral chain (Cabral et al., 2013), suggests that these low d18_{O values}

reflect the presence of very ancient recycled hydrothermally altered oceanic crust rather than modern assimilation.

Cabral et al.’s discovery is one of the most remarkable ones in mantle geochemistry because it both unequivocally demonstrates the presence of recy- cled surficial material in the mantle and constrains the age of this material to be 2.3 billion years or older. As I explained earlier, such MIF sulphur appears to have been exclusively produced by photolysis when ultraviolet radiation could penetrate deeply into the atmosphere prior to the Great Oxidation Event. Unlike carbon, nitrogen, and sulphur evidence for surficial material in diamonds, the sulphur in Mangaia olivines could not have been stored in the subcontinental lithosphere and must have come from the convecting mantle, most likely the deep mantle.

Mass independent sulphur fractionation in the early Earth produced oxidised sulphur with negative ∆33_{S anomalies and reduced sulphur with posi-}

tive ∆33_{S anomalies. Cabral et al.’s data are shown in Figure 6.9. They (with one}

exception) exhibit negative ∆33_{S anomalies, which are characteristic of Archean}

barites and hydrothermal sulphides (Farquhar and Wing, 2003), in contrast to sulphides in diamonds, which exhibit positive ∆33_{S anomalies. This led Cabral}

et al. to argue that anciently subducted hydrothermally altered oceanic crust was the carrier of the mass independently fractionated sulphur.

Figure 6.9 Sulphur isotope ratios in sulphide inclusions in olivines (triangles) from Mangaia, Austral-Cook Islands (Cabral et al., 2013) and in diamonds (circles) from the Orapa kimberlite pipe in Botswana (Farquhar et al., 2002).

Labidi et al. (2013) reported new high precision sulphur isotope data for MORB and showed that d34_{SCDT in MORB is negative, in contrast to previous}

work that indicated that mantle d34_S

CDT was close to the chondritic value of

correlates positively with 87_Sr/86_{Sr and negatively with} 143_Nd/144_{Nd (Fig. 6.10).}

They interpreted these correlations as evidence of mixing between a DMM component with d34_S

CDT ≈ 1.5 ‰ and an ‘enriched’ component associated with

the Discovery and Shona mantle plumes. They argue the data are most consistent with the enriched component being recycled sediment with d34_S

CDT ≈ +10 ‰.

They found no evidence of mass-independent fractionation (∆33_{S and ∆}36_{S = 0}

within error), which implies that the recycled component is of Proterozoic or younger age. Subsequently, Labidi et al. (2015) reported d34_{S values ranging}

from +0.11 ‰ to +2.79 ‰ in the reduced sulphur fraction of Samoan lavas (the subordinate sulphate fraction had higher d34_{S). Furthermore, d}34_{S in the reduced}

sulphur correlated with 87_Sr/86_{Sr in glasses (Fig. 6.10). They argued the correlation}

“requires the EM-2 endmember to be relatively S-rich, and only sediments can account for these isotopic characteristics.”

Figure 6.10 d34_{SCDT vs.} 87_Sr/86_{Sr in lavas from Samoan volcanoes (Malumalu, Vailulu’u,}

Ta’u, and Muli and MORB from the South Atlantic. Also shown are mixing curves between a mantle component and potential sedimentary or upper crustal end-members with distinct d34_{S and S/Sr ratios (modified from Labidi} et al., 2015).