Download Isotopic Fractionation of Silicon and Chlorine in Geological Systems and more Study notes Geochemistry in PDF only on Docsity!
Chapter 11
tween the core and silicate Earth. Observational differences of Si isotopic composition between metal and silicate phases in enstatite chondrites (Ziegler et al., 2010), experimental data (Shalar et al., 2011), and theoretical calculations (Schauble et al., 2007) indicate that the fractionation of Si isotopes between metal and silicate can be expressed as:
∆ 30 Sisilicate − metal ≅
7.5 × 10
6
T
This equation tells us that isotopically light Si preferentially partitions into the metal phase (i.e., the core) and the extent of fractionation decreases with temperature. We don’t know exactly the tempera- ture of equilibration, but assuming a difference in δ^30 Si between the bulk silicate Earth and carbona- ceous or chondrites of -0.15‰, Armytage et al. (2011) concluded that the core contains ~8.7 wt% Si while Ziegler et al. (2010) concluded the core contains at least 6% Si. This amount of Si would be consis- tent with geophysical observations, such as seismic velocity and moment of inertia, that constraint the density of the core are require that it contain a few percent of an element lighter than Fe or Ni. Fitoussi and Bourdon (2012) calculate that to account for the much greater difference in δ^30 Si between the Earth and Moon and enstatite chondrites, the Earth’s core would have to contain 28 wt% Si, an amount much greater than allowed by geophysical constraints. The similarity of δ^30 Si of the Moon and silicate Earth implies that the Earth’s core had mostly segregated before the Moon-forming impact and the δ^30 Si in silicates from proto-Earth and the impactor were homogenized in the aftermath of the impact (Army- tage et al., 2012). Considering Si, Cr, Ni, and O isotopic compositions, Fitoussi and Bourdon (2012) sug- gest that the Earth-Moon composition could be accounted for if the parental material were a mix of car- bonaceous and ordinary chondritic material with the addition of 15% of enstatite chondrite. The average δ3-^ Si of mantle-derived ultramafic xenoliths is -0.30±0.04‰, that of basalts is -0.28±0.03‰, and that of granitic rocks is -0.23±0.13‰. This implies that there is a slight fractionation associated with melting and fractional crystallization, with lighter silicon isotopes partitioning prefer- entially into SiO 2 -poor mafic minerals such as olivine. Savage et al. (2012) found that δ^30 Si in igneous rocks depended on SiO 2 composition approximately as:
δ 30
Si (‰) = 0.0056 × [ SiO 2 ] − 0.568 11.
Savage et al. (2012) also found a slight variation in δ^30 Si in granites that related to their petrogenesis: S- type granites (those derived by melting of sedimentary precursors and with Al 2 O 3 /(CaO+Na 2 O +K 2 O)>1) had slightly lighter δ^30 Si than either I-type (derived from melting of igneous precursors) or A- type (derived by fractional crystallization of a mafic parent magma) granites. Compared to other proc- esses, however, Si isotope fractionation in igneous processes is quite small; for example, in the course of evolution from basaltic to andesitic, the δ^30 Si of a magma would increase by only ~0.06‰, an amount only slightly greater than present analytical precision. This, of course, should not be surprising because (1) temperatures involved are high, (2) the oxidation state of Si does not vary, and (3) the atomic envi- ronments of Si in silicate minerals and melts do not differ greatly. At lower temperatures and where another form of Si is involved (silicic acid: H 4 SiO 4 ), fractionations are (not surprisingly) greater. Igneous minerals at the surface of the Earth undergo weathering reac- tions such as:
2NaAlSi 3 O 8 + 5H 2 O → Al 2 Si 2 O 5 (OH) 4 + 3SiO 2 + 2Na+^ + 2OH –^ + H 4 SiO 4
Thus when albite reacts with water, Si partitions between residual clay (kaolinite in this example), quartz, and silicic acid in aqueous solution. This is an overall reaction that involves a variety of inter- mediate steps and phases and furthermore occurs rather slowly. Thus determining the fractionation factors for such reactions experimentally is difficult. Méhuet et al. (2007) used a theoretical approach to calculate a ∆^30 Si kaolinite-quartz fractionation factor of -1.6‰ at 25˚C. Experimental estimates of the fractionation of between dissolved H 4 SiO 4 and biogenic opaline silica indicate that ∆^30 Si (^) opal-diss. Si ≈ -1.1‰
Chapter 11
(De la Rocha et al., 1997). Other studies indicate little fractionation in the transformation of opal to quartz, so we can infer that the fractionation between kaolinite and dissolved silica should be ~ -2.5‰. Consequently, we expect that weathering of silicate rocks to produce isotopically light clays and an iso- topically heavy solution. In addition, some of the silica released into solution by weathering can be ab- sorbed on the surface of oxides and hydroxides in soils such as ferrihydrite and goethite. Delstanche et al. (2009) experimentally determined fractionation factors for adsorption of Si on ferrihydrite as ∆^29 Si (^) ferri-
solu. ≈^ -0.81‰ and^ ∆ (^29) Si geoth.-solu. ≈^ -0.54‰. Assuming mass-dependent fractionation these imply^ (^30) Si frac-
tionation factors of ∆^30 Si (^) ferri-solu. ≈ -1.59‰ and ∆^30 Si (^) geoth.-solu. ≈ -1.06‰. This adsorption drives dissolved sil- ica further toward isotopically heavy compositions. The solubility of silica is quite limited; concentra- tions in rivers are typically less than 10 ppm. Consequently, isotopic fractionation between dissolved silica and solid phases has a smaller effect on the isotopic composition of solid phases than the solution. In addition to being the key ingredient in silicate rocks and minerals, silicon is also extensively bio- utilized. Terrestrial plants take-up silicic acid from the soil solution and incorporate it in various tis- sues, where it appears to increase rigidity, increase photosynthetic efficiency, limit loss of water by evapotranspiration, and increase the resistance to pathogens and grazing by herbivores, including in- sects. The ‘phytoliths’ formed in this process consist of hydrated opaline silica and vary widely in sili- con isotopic composition. Ding et al. (2009) suggest the latter is due to Rayleigh fractionation during precipitation of the phytoliths from plant fluids with an apparent fractionation factor of ∆^30 Si (^) phyto-solu. ≈ +2‰. This follows a fractionation during uptake of H 4 SiO 4 from the soil solution by roots of ∆^30 Si (^) roots-soil ≈ -1.2‰. Thus silica utilization by plants drives the soil solution towards heavier isotopic compositions. In view of the fractionations during weathering and uptake by the terrestrial biota, it is not surprising to find that rivers are isotopically heavy, with δ^30 Si ranging from -0.1‰ to +3.4‰ and averaging about +0.8‰ (De la Rocha et al., 1997; Georg et al., 2006; Ding et al., 2011). Isotopic compositions vary not only between different rivers, but also over time in individual rivers. Georg et al. (2006) found that δ^30 Si correlated positively with dissolved Si concentration and inversely with discharge and base cation flux. They suggest that the higher δ^30 Si are associated with active clay mineral formation during slow seepage of water through soils and low river discharge. Dissolved silica reaching the oceans is extensively bio-utilized and can be bio-limiting. Marine organ- isms utilizing silica include sponges and planktonic protists such as diatoms, silicoflagellates, and ra- diolarians. Of these, sponges appear to fractionate silica the most. De La Rocha (2003) found that δ^30 Si in modern sponge spicules range from -1.2‰ to -3.7‰ and fractionation factors ∆^30 Si (^) sponge-sw averaged -3.8±0.8‰. Diatoms, which account for some 75% of primary productivity in the oceans, are far more important in the marine silica cycle. These organisms live in the surface waters and build tests (or shells) of opaline silica. As a consequence, dissolved silica is depleted in the surface waters (Figure 11.23). As their remains sink, the tests tend to redissolve and consequently deep waters are enriched in dissolved silica. As we noted above, experiments suggest a fractionation factor, i.e., ∆^30 Si (^) opal-dissolv., of about -1.1‰ during formation of diatom tests, and as a consequence surface waters are typically en- riched in 30 Si. Curiously, Demerest et al. (2009) found that the fractionation during partial dissolution of diatoms tests was just the opposite: the solution became isotopically lighter and the residual tests iso- topically heavier with a ∆^30 Si (^) opal-dissolv. ≈ +0.55‰. Thus in both precipitation and dissolution of diatom tests, the lighter isotope reacts more readily (see Problem 11.4). Because δ^30 Si is oceans is tightly coupled with biological productivity and hence the carbon cycle, there is much interest in using silicon isotopes in understanding the modern silica cycle and using δ^30 Si in siliceous biogenic sediments to reconstruct productivity variations in the ancient oceans. To date, only a few such studies have been carried out, and interpretation of these variations remains somewhat equivocal. Improved understanding of the silica cycle in the modern ocean is needed before δ^30 Si can be used as a paleo-oceanographic proxy.
Chapter 11
δ^37 Cl (^) SMOC ), whose 37 Cl/ 35 is 0.319627 (Coplen et al., 2002). NIST-SRM 975 is the commonly used reference standard, whose isotopic composition is (^73) Cl/ 35 Cl = 0.31977 (δ (^37) Cl SMOC = +0.52‰)^ *^. Relatively few Cl isotopic analyses of me- teorites have been reported. Sharp et al. (2007) reported δ^37 Cl of +1.21‰ in Orgueil (CV1), +0.46‰ in Murchison (CM2) and - 0.38‰ in Allende (CV3); components of meteorites such as water-soluble extracts and sodalite (Na 8 Al 6 Si 6 O 24 C (^) l2 ) show greater variability. Because the oceans and evaporites, which both have δ^37 Cl of ~0‰, likely contain more than 80% and possibly more than 90% of the Earth’s chlorine in- ventory, the δ^37 Cl of the Earth is probably close to 0‰. Sharp et al. (2010) found that δ^37 Cl in lunar materials ranged from -0.7‰ to +24‰. They argued that both the positive value and the large spread resulted from fractionation during volatilization and loss of metal chlorides during eruption of the basalts. That in turn, they suggested, indicated the Moon is hydrogen poor because formation of magmatic HCl lessens the fractionation during degassing in ter- restrial basalts. Schauble et al. (2003) theoretically calcu- lated factors for chlorine. They found (not surprisingly) that the largest fractionation occurred between chlorine of different oxi- dation states. Fractionation between chlo- rides, chlorate (ClO 2 ), and perchlorate (ClO 4 – ) are as great as 27‰ and 73‰ at 298K, with (^37) Cl concentrating in the oxidized forms; however, naturally occurring chlorine in these oxidized forms is extremely rare and plays essentially no role in the natural chlo- rine cycle. Among chlorides, they predicted that 37 Cl will preferentially partition into organic molecules (by 5 to 9‰ at 295K) and into salts and silicates where it is bound to divalent metals (e.g., FeCl 2 ) in preference to monovalent ones (such as NaCl) by 2 to 3‰ at 298K. 37 Cl should also concentrate in sili- cates relative to brines by 2 to 3‰ at surface temperature. They also calculated ∆Cl 2 -HCl =
- (^) The supply of this standard has been exhausted and it has been superseded by NIST-SRM 975a, whose 37 Cl/ 35 Cl =
0.31970.
Figure 11.24. Chlorine isotope variations in terrestrial materials and meteorites.
Chapter 11
+2.7‰ and ∆HCl-NaCl = +1.45‰. Experiments by Eggenkamp et al. (1995) produced the following frac- tionation factors: ∆NaCl-soltn = +0.26‰, ∆KCl-soltn = -0.09‰, and ∆MgCI 2 .6H 2 O-soltn = -0.6‰. These fractionations are consistent with the observed narrow range and mean value of δ^37 Cl in halite evaporites (-0.9 to +0.9‰ and +0.06‰, respectively) and slightly lighter composition of potash (KCl) facies evaporites (average -0.3‰) (Eastoe et al., 2007). Liebscher et al. (2004) experimentally determined fractionations between NaCl solutions and aqueous vapor at 450˚C at pressures close to the critical point and found that ∆^37 Cl (^) vapor=liquid values were within ±0.2‰ of 0. Sharp et al. (2010) experimentally determined the fractionation between HCl vapor and gas and found ∆^37 Cl (^) HClvapor=HClliquid ≈ +1.7‰, consistent with theo- retical prediction. Figure 11.24 summarizes δ^37 Cl variations in natural materials. As noted above, seawater is isotopi- cally homogeneous within analytical error or nearly so †^. Phanerozoic marine evaporites are fairly ho- mogeneous, with a mean and standard deviation of -0.08‰ and 0.33‰, respectively. Eastoe et al., (2007) found small but significant variations between basins, which they attributed to contribution from non-marine sources, but found no systematic variation with basin size or age, other than a potash (KCl) facies being slightly lighter than halite facies, consistent with theoretically predicted fractionation. This suggests that the Cl isotopic composition of seawater has been uniform over the Phanerozoic. Both groundwaters and sediments of various types, including metasediments, also exhibited only limited variation in δ^37 Cl. In the surficial environment, the largest variation in δ^37 Cl appears to be in fumarole gases and fluids and in marine pore waters, particularly pore waters expelled through accretionary prisms. Somewhat surprisingly, low temperature fumarole effluents (those with measured temperatures <100˚C) show less
† (^) The sole reported exception is water from the upper current of the Bosphorus, which has a δ^37 Cl (^) SMOC of +0.4‰ (Eas-
toe, et al., 2007).
Figure 11.24. Chlorine and lead isotope ratios in oceanic island basalts and MORB. Chlorine isotope data from John et al. (2011), Bonifacie et al. (2008) and Layne et al. (2010).
