Download Subsequent - Geochemistry - Lecture Notes and more Study notes Geochemistry in PDF only on Docsity!
Chapter 9
subsequent to hydrothermal low temperature metamorphism), and contrasts with the homogeneity of peridotite xenoliths (Figure 9.06), many of which also come from kimber- lites. Since eclogites are the high pressure equivalent of basalts, this suggests the ecolgites come from subducted oceanic crust. Further evidence to this effect is the correlation observed between O isotope ratios, Sr isotope ratios, and K and Rb concentrations, as these will also be affected by metamor- phism of the oceanic crust. Geobarometry suggests these eclogites last equilibrated at depths of 165-190 km. Radiogenic isotope evidence suggests they are 2.47 Ga old. We examined the radiogenic isotope evidence for mantle heterogeneity in Chapter
- One interpretation is that this heterogeneity results from recycling of oceanic crust and sediment into the mantle. Since sediment, and, to a lesser degree, oceanic crust (as we shall see in the next section) differs in its oxygen isotopic compositions from normal mantle material, we might expect this recycling to produce variations in O isotope composition of basalts. As Figure 9.24 shows, this is indeed observed. EM2 basalts (defined by their radiogenic isotopic compositions as shown in Figures 6.11 and 6.12) show sys- tematically higher δ^18 O than MORB, while HIMU and low 3 He/ 4 He basalts (many HIMU basalts have low 3 He/ 4 He) have systematically low δ^18 O. It is in EM2 basalts, such as those from the Society Islands, where the best evidence for a recycled sedimentary component is seen (e.g. Figure 6.18). Even among MORB, there is a correlation between incompatible element enrichment and δ^18 O (Figure 9.24). Since δ^18 O should be effectively insensitive to the extent of partial melting and fractional crystallization, this correlation is likely to reflect a correlation between these properties in the MORB source. Since O isotope ratios can be changed only at low temperatures near the Earth’s surface, Eiler and Schiano (2000) interpret this as evidence of a recycled component in MORB.
S TABLE ISOTOPES IN H YDROTHERMAL S YSTEMS
Ridge Crest Hydrothermal Activity and Metamorphism of the Oceanic Crust
Early studies of “greenstones” dredged from mid-ocean ridges and fracture zones revealed they were depleted in 18 O relative to fresh basalts. Partitioning between of oxygen isotopes between various min- erals, such as carbonates, epidote, quartz and chlorite, in these greenstones suggested they had equili- brated at about 300° C (Muehlenbachs and Clayton, 1972). This was the first, but certainly not the only,
Figure 9.25. Relationship between δ^18 O and K 2 O and (La/Sm)N (the chondrite-normalized La/Sm ratio) in MORB glass. Solid line shows the ex- pected result if recycled altered oceanic crust is mixed with depleted peridotite and then melted 10%. Dashed line shows the expected result if a 10% melt of eclogitic recycled oceanic crust is mixed with a 10% melt of depleted peridotite. From Eiler and Schiano (2000).
Chapter 9
evidence that the oceanic crust underwent hydrothermal metamorphism at depth. Other clues in- cluded highly variable heat flow at ridges and an imbalance in the Mg fluxes in the ocean. Neverthe- less, the importance of hydrothermal processes was not generally recognized until the discovery of low temperature (~20° C) vents on the Galapagos Spreading Center in 1976 and high temperature (350° C) “black smokers” on the East Pacific Rise in 1979. Various pieces of the puzzle then began to fall rapidly into place and it was soon clear that hydrothermal activity was a very widespread and important phe- nomenon. Most of the oceanic crust is affected to some degree by this process, which also plays an im- portant role in controlling the composition of seawater. Hydrothermal metamorphism occurs because seawater readily penetrates the highly fractured and therefore permeable oceanic crust. A series of chemical reactions occurs as the seawater is heated, transforming it into the reduced, acidic, and metal-rich fluid. Eventually the fluid rises and escapes, forming the dramatic black smokers. Fluid in many “black smokers” vents at temperatures of about 350° C *^. This results from the density and viscosity minimum that occurs close to this temperature at pressures of 200-400 bars combined with a rapidly decreasing rock permeability above these tempera- tures. Of the reactions that occur in the process, only one, namely oxygen isotope exchange, concerns us here. Seawater entering the oceanic crust has a δ^18 O of 0, fresh igneous rock has a δ^18 O of +5.7. As sea- water is heated, it will exchange O with the surrounding rock until equilibrium is reached. At tempera- tures in the range of 300-400° C and for the mineral assemblage typical of greenschist facies basalt, the net water-rock fractionation is small†^ , 1 or 2‰. Thus isotopic exchange results in a decrease in the δ^18 O of the rock and an increase in the δ^18 O of the water. Surprisingly, there have only been a few oxygen isotope measurements of vent fluids; these indicate δ^18 O of about +2. At the same time hydrothermal metamorphism occurs deep in the crust, low-temperature weathering proceeds at the surface. This also involves isotopic exchange. However, for the temperatures (~2° C) and minerals produced by these reactions (smectites, zeolites, etc.), fractionations are quite large (some- thing like 20‰). The result of these reactions is to increase the δ^18 O of the shallow oceanic crust and decrease the δ^18 O of seawater. Thus the effects of low temperature and high temperature reactions are opposing. Muehlenbachs and Clayton (1976) suggested that these opposing reactions actually buffered the iso- topic composition of seawater at a δ^18 O of ~0. According to them, the net of low and high temperature fractionations was about +6, just the observed difference between the oceanic crust and the oceans. Thus, the oceanic crust ends up with an average δ^18 O value about the same as it started with, and the net effect on seawater must be close to zero also. Could this be coincidental? One should always be suspicious of apparent coincidences in science, and they were. Let’s think about this a little. Let’s assume the net fractionation is 6, but suppose the δ^18 O of the ocean was –10 rather than 0. What would happen? Assuming a sufficient amount of oceanic crust available and a simple batch reaction with a finite amount of water, the net of high and low temperature basalt- seawater reactions would leave the water with δ^18 O of –10 + 6 = –4. Each time a piece of oceanic crust is allowed to equilibrate with seawater, the δ^18 O of the ocean will increase a bit. If the process is repeated enough, the δ^18 O of the ocean will eventually reach a value of 6 – 6 = 0. Actually, what is required of seawater–oceanic crust interaction to maintain the δ^18 O of the ocean at 0‰ is a net increase in isotopic
- While this is typical, temperatures of 400° C or so have also been found. Most low-temperature vents waters, such as those on the GSC appear to be mixtures of 350° C hydrothermal fluid and ambient seawater, with mixing occur- ring at shallow depth beneath the seafloor. Although hydrothermal fluids with temperatures substantially above 400° C have not been found, there is abundance evidence from metamorphosed rocks that water-rock reactions occur at temperatures up to 700° C. † While the mineral-water fractionation factors for quartz and carbonate are in the range of +4 to +6 at these tempera- tures, the fractionation factor for anorthite and chlorite are close to zero, and that for magnetite is negative.
Chapter 9 Spring 2011
topic exchange with oceanic crust is much too slow to dampen these short-term variations. That water-rock interaction produces essentially no net change in the isotopic composition of the oce- anic crust, and therefore of seawater was apparently confirmed by the first thorough oxygen isotope study of an ophiolite by Gregory and Taylor (1981). Their results for the Samail Ophiolite in Oman are shown in Figure 9.26. As expected, they found the upper part of the crust had higher δ^18 O than fresh MORB and while the lower part of the section had δ^18 O lower than MORB. Their estimate for the δ^18 O of the entire section was +5.8, which is essentially identical to fresh MORB. Ocean Drilling Project (ODP) results show much the same pattern as the Samail ophiolite, though no complete section of the oceanic crust has yet been drilled. ODP results do show that hydrothermal alteration is not uniform. In Hole 504B, the deepest hole yet drilled, the transition to the hydrothermally altered zone was found to be quite sharp. If the Muehlenbachs and Clayton hypothesis is correct and assuming a steady-state tectonic envi- ronment, the δ^18 O of the oceans should remain constant over geologic time. Whether it has or not, is controversial. Based on analyses of marine carbonates, Jan Veizer and his colleagues have argued that it is not. Figure 9.27 shows the varia- variation in δ^18 O in marine carbonates over Phanerozoic time. The isotopic composition of carbonates reflect (1) the composition of water from which they precipitated and (2) the fractionation between water and carbonate. The latter is large (~30‰) and temperature dependent. However, over reasonable ranges of temperature, the fractionation factor will vary by only a few per mil. (As is convention, the data in Figure 9.27 are reported relative to PDB, which is about +30‰ relative to SMOW. This roughly matches the fractionation be- tween water and carbonate: water with δ^18 OSMOW= 0 should precipitate carbon- ate with δ^18 OPDB≈ 0.) The variations in Figure 9.27 are much larger than this, implying that seawater oxygen isotopic composition has indeed varied signifi- cantly. Subsequent reaction and equili- bration of carbonates with pore water could shift the δ^18 O of the carbonates, but these data are from carefully screen samples that are least likely to be al- tered in this way. Some of the short- term variations might be due to changes in ice volume. Increasing ice volume would shift δ^18 O positively, and indeed, there is some evidence that δ^18 O was higher when the climate was colder, such as during the late Ordovi- cian and Carboniferous glaciations. In addition to the short-term variations
Figure 9.28. Change in sealevel and the isotopic composi- tion of water in the model of Wallmann (2001) compared with actually sea level and the Veizer et al. (1999) carbon- ate data. Model is the run at two different recycling effi- ciencies (r), 0.3 and 0.5. From Wallmann (2001).
Chapter 9 Spring 2011
there is an overall increase from about - 8‰ in the Cambrian to 0‰ at present. This clearly contradicts the idea that ridge crest hydrothermal activity buffers the δ^18 O to a constant value. Wallmann (2001) has produced a box model of the isotopic composition of ocean water that matches, in a very qualitative way, the secular increase in δ^18 O observed by Veizer. He argues that the long-term isotopic composition of seawater depends not only on ridge crest hydrothermal activity, but also on a number of other inputs to and removals from the ocean, particularly those related to the deep subduc- tion cycle. He also argues that the isotopic composition of subducted water in oceanic crust and sedi- ment that is less that the isotopic composition of water degassed from the mantle, and further, that wa- ter has been subducted at a higher rate than it has been degassed from the mantle. Over the Phanero- zoic, this, he argues, has produced a decrease in the mass of the oceans and in sealevel. His model this shows an increase in δ^18 O and a decrease in sealevel (Figure 9.28). However, the controversy has not been resolved. Based on δ^18 O of ancient ophiolites (the Oman ophiolite is Cretaceous), Muehlenbachs continues to argue that the oxygen isotopic composition of seawater has not changed significantly throughout Phanerozoic and Proterozoic time (e.g., Muehlenbachs and Furnas, 2003).
Meteoric Geothermal Systems
Hydrothermal systems occur not only in the ocean, but just about everywhere that magma is in- truded into the crust. In the 1950’s a debate raged about the rate at which the ocean and atmosphere were created by degassing of the Earth’s interior. W. W. Rubey assumed that water in hydrothermal systems such as Yellowstone was magmatic and argued that the ocean and atmosphere were created quite gradually through magmatic de- gassing. Rubey turned out to be wrong. One of the first of many im- portant contributions of stable isotope geochemistry to understanding hydro- thermal systems was the dem- onstration by Craig (1963) that water in these systems was meteoric, not mag- matic. The argument is based upon the data shown in Figure 9.29. For each geothermal system, the δD of the “chloride” type geothermal waters is the same as the local precipitation and groundwater, but the δ^18 O is shifted to higher values. The shift in δ^18 O results from “high” temperature (~300°C) re- action of the local meteoric water with hot rock. However, because concentra- tion of hydrogen in rocks is nearly 0 (more precisely because ratio of the mass of hydrogen in the water to mass of hydrogen in the reacting rocks is ex- tremely high), there is essentially no change in the hydrogen isotopic com- position of the water. If the water in- volved in these systems was magmatic, it would not have the same δD as local
Figure 9.29. δD and δ^18 O in meteoric hydrothermal systems. Closed circles show the composition of meteoric water in the vicinity of Yellowstone, Steamboat Springs (Colorado), Mt. Lassen (California), Iceland, Larderello (Italy), and The Gey- sers (California), and open circles show the isotopic composi- tion of chloride-type geothermal waters at those locations. Open triangles show the location of acidic, sulfide-rich geo- thermal waters at those locations. Solid lines connect the me- teoric and chloride waters, dashed lines connect the meteoric and acidic waters. The “Meteoric Water Line” shows the cor- relation between δD and δ^18 O observed in precipitation shown in Figure 8.10.
Chapter 9 Spring 2011
parcel of water, dW , passes through the system and induces and incremental change in the isotopic composition of the rock, d δ r. In this case, we can write: Rcr d δ r =( δ (^) w^ i^ − (^) [ Δ + δ r ]) cwdW 9.1 4
This equation states that the mass of isotope exchanged by the rock is equal to the mass of isotope ex- changed by the water (we have substituted ∆ + δr for δ
f w ). Rearranging and integrating, we have:
W
R
= ln
δ r f
− δ r
i
− δ r
f
+ δ w
i
cr
cw
Thus it is possible to deduce the water rock ratio for an open system as well as a closed one. Using this kind of approach, Gregory and Taylor (1981) estimated water/rock ratios of ≤ 0.3 for the gabbros of the Oman ophiolite. It should be emphasized, however, that this can be done with other iso- tope systems as well. For example, McCulloch et al. (1981) used Sr isotope ratios to estimate wa- ter/rock ratios varying from 0.5 to 40 for different parts of the Oman ophiolite.
The Skaergaard Intrusion
A classic example of a meteoric hydro- thermal system is the Early Tertiary Skaer- gaard intrusion in East Greenland. The Skaergaard has been studied for nearly 75 years as a classic mafic layered intrusion. Perhaps ironically, the initial motivation for isotopic study of the Skaergaard was determination of primary oxygen and hy- drogen isotopic compositions of igneous rocks. The results, however, showed that the oxygen isotope composition of the Skaergaard has been pervasively altered by hydrothermal fluid flow. This was the first step in another important contribution of stable isotope geochemistry, namely the demonstration that most igneous intru- sions have reacted extensively with water subsequent to crystallization. Figure 9.30 shows a map of the Skaer- gaard with contours of δ^18 O superimposed on it. Figure 9.31 shows a restored cross section of the intrusion with contours of δ^18 O. There are several interesting features. First, it is clear that circulation of water was strongly controlled by permeability. The impermeable basement gneiss experi- enced little exchange, as did the part of the intrusion beneath the contact of the gneiss with the overlying basalt. The basalt is quite permeable and allowed water to flow freely through it and into the intrusion. Figures 9.30 and 9.31 define zones of low
Figure 9.30. Oxygen isotope variations in the Skaergaard Intrusion. LZ, MZ, and UZ refer to the ‘lower zone, ‘mid- dle zone’ and ‘upper zone’ of the intrusion, which dips 20 - 25° to the southeast. UBZ refer to the ‘upper border group’. The δ^18 O = +6 contour corresponds more or less to the trace of the gneiss-basalt contact through the intrusion (SW to NE). The gneiss is essentially im- permeable, while the basalt is highly fractured. Thus most water flow was above this contact, and the gabbro below it retained its original ‘mantle’ isotopic signature (+6). After Taylor (1974).
Chapter 9 Spring 2011
δ^18 O, which are the regions of hydrother- mal upwelling. Water was apparently drawn into the sides of the intrusion and then rose above. This is just the sort of pattern observed with finite element models of fluid flow through the intru- sion. Calculated water-rock ratios for the Skaergaard were 0.88 in the basalt, 0.52 in the upper part of the intrusion and 0. for the gneiss, demonstrating the impor- tance of the basalt in conduction the water into the intrusion and the inhibiting effect of the gneiss. Models of the cooling his- tory of the intrusion suggest that each cm^3 of rock was exposed to between 10^5 and 5 × 106 cm^3 of water over the 500,000 year cooling history of the intrusion. This would seem to conflict with the wa- ter/rock ratios estimated from oxygen iso- topes. The difference is a consequence of each cc of water flowing through many cc’s of rock, but not necessarily reacting with it. Once water had flowed through enough grams of rock to come to isotopic equilibrium, it would not react further with the rock through which it sub- sequently flowed (assuming constant tem- perature and mineralogy). Thus it is im- portant to distinguish between W/R ratios calculated from isotopes, which reveal only the mass (or molar) ratio of water and rock in the net reaction, to flow mod- els. Nevertheless, the flow models dem- onstrate that each gram of rock in such a system is exposed to an enormous amount of water. Figure 9.32 is a cartoon illustrat- ing the hydrothermal system deduced from the oxygen isotope study.
Oxygen Isotopes and Mineral
Exploration
Oxygen isotope studies can be a valuable tool in mineral exploration. Mineralization is very often (though not exclusively) associated with the region of greatest water flux, such as areas of upward moving hot water above intrusions. Such areas are likely to have the lowest values of δ^18 O. To under- stand this, let’s solve equ. 9.15, the final value of δ^18 O:
δ r^ f^ =( δ (^) r^ i^ − δ w^ i^ + Δ) e
− Wcw
Rcr + δ
w
i − Δ 9.1 6
Figure 9 .31. Restored cross-section of the Skaergaard in- trusion with contours of δ^18 O.
Figure 9.32. Cartoon illustrating the hydrothermal system in the Skaergaard intrusion. After Taylor (1968).
.
Layered Series
Border Group
present
Gneiss Topography
Basalt
Chapter 9 Spring 2011
sulfur isotopes are undoubtedly the most com- plex. This complexity arises in part because of there are five common valence states in which sulfur can occur in the Earth, +6 (e.g., BaSO 4 ), + (e.g., SO 2 ), 0 (e.g., S), – 1 (e.g., FeS 2 ) and – 2 (H 2 S). Significant equilibrium isotopic fractionations oc- cur between each of these valence states. Each of these valence states forms a variety of com- pounds, and fractionations can occur between these as well. Finally, sulfur is important in bio- logical processes and fractionations in bio- logically mediated oxidations and reductions are often different from fractionations in the abio- logical equivalents. There are two major reservoirs of sulfur on the Earth that have uniform sulfur isotopic composi- tions: the mantle, which has δ^34 S of ~0 and in which sulfur is primarily present in reduced form, and seawater, which has δ^34 S of +20 and in which sulfur is present as SO
2 - 4.^ Sulfur in sedimentary, meta- morphic, and igneous rocks of the continental crust may have δ^34 S that is both greater and smaller than these values (Figure 9.35). All of these can be sources of sulfide in ores, and further fractionation may occur during transport and deposition of sul- fides. Thus the sulfur isotope geochemistry of sul- fide ores is remarkably complex.
Sulfur Isotope Fractionations in Magmatic
Processes
Sulfur is present in peridotites as trace sulfides, and that is presumably its primary form in the mantle. At temperatures above about 400 °C, H 2 S and SO 2 are the stable forms of sulfur in fluids and melts. In basaltic melts, sulfur occurs predomi- nantly as dissociated H 2 S: HS–.^ It is unlikely that significant fractionation occurs between these forms during melting. Indeed, as we have seen, the mean δ^34 S in basalts (~+0.1) is close to the value in meteorites, which is presumably also the mantle value. The solubility of H 2 S in basalt appears to be only slightly less than that of water, so that under moderate pressure, essentially all sulfur will remain dissolved in basaltic liquids. The solubility of H 2 S in silicate melts is related to the Fe content, decreas- ing with decreasing Fe. As basalts rise into the crust, cool, and crystallize, several processes can affect the oxidation state and solubility of sulfur and the produce isotopic frac- tionations. First, the decreasing pressure results in
Figure 9.36. Fractionation of sulfur isotopes be- tween fluid and melt (shown by dashed curves) as a function of oxygen fugacity and temperature for PH 2 O = 1 kB. Solid lines show equal concen- tration boundaries for quartz + magnetite ® fay- alite (QM-F), H 2 S ® SO 2 , and Magnetite ® Hema- tite (M-H). After Ohmoto and Rye (1979).
Figure 9.35. δ^34 SCDT in various geologic materials (after Hoefs, 1987).
Chapter 9 Spring 2011
some of the sulfide partitioning into the gas (or fluid) phase. In addition, H 2 can be lost from the melt through diffusion. This increases the ƒO 2 of the melt, and as result, some of the sulfide will be oxidized to SO 2 , which is very much less soluble in silicate melts than H 2 S. Decreasing Fe content as a conse- quence of fractional crystallization will also decrease the solubility of S in the melt, increasing its con- centration in a coexisting fluid or gas phase. Isotope fractionation will occur between the three species (dissolved HS–, H 2 S, SO 2 ). The isotopic composition of the fluid (gas) will differ from that of the melt, and can be computed as:
δ^34 Sfluid = δ^34 Smelt − Δ HS − + Δ SO 2
R
R + 1
⎟ 9.1^7
where ∆HS is the fractionation factor between HS–^ and H 2 S, ∆SO 2 is the fractionation factor between H 2 S and SO 2. R is the molar ratio SO 2 /H 2 S and is given by:
R =
XSO 2
XH 2 S
K ν H 2 S ƒ O^3 / 22
Pf ν H 2 O XH 2 O ν SO 2
where ν is the activity coefficient, Pf is the fluid pressure (generally equal to total pressure), ƒO 2 is oxy- gen fugacity, and K is the equilibrium constant for the reaction:
H 2 S(g) +
2 O 2 ®^ H 2 O(g) + SO2(g) 9.1^9
Figure 9.36 shows the sulfur isotope fractionation between fluid and melt calculated from equations 9.17 and 9.18 as a function of function of temperature and ƒO 2 for PH 2 O = 1 kB. At the temperatures and ƒO 2 of most basalts, sulfur will be present primarily as H 2 S in the fluid (gas phase) and HS–^ in the melt. The fractionation between these species is small (~ 0.6 ‰), so the isotopic composition of fluid phase will not be very different that of the melt. For rhyolites and dacites, a significant fraction of the sulfur can be present as SO 2 , so that greater fractionation between melt and fluid is possible. An interesting feature of the above equations is that the fractionation between fluid and melt depends on the water pressure. Figure 23.11 is valid only for PH 2 O = 1 kB. A decrease in Pf or XH 2 O (the mole fraction of water in the fluid) will shift the SO 2 /H 2 S equal concentration boundary and the δ^34 S contours to lower ƒO 2. Conversely, an increase the in the water content will shift to boundary toward higher ƒO 2. Both the eruptions of El Chichón in 1983 and Pinatubo in 1991 release substantial amounts of SO 2. The SO 2 – rich nature of these eruptions is thought to result from mixing of a mafic, S-bearing magma with a more oxidized dacitic magma, which resulted in oxidation of the sulfur, and consequent increase of SO 2 in the fluid phase. There are a number of other processes that affect the solubility and oxidation state of sulfur in the melt, and hence isotopic fractionation. Wall rock reactions could lead to either oxidation or reduction of sulfur, crystallization of sulfides or sulfates could cause relatively small frac- tionations and additionally affect the SO 2 /H 2 S ratio of the fluid. Depending on the exact evolutionary path taken by the magma and fluid, δ^34 S of H 2 S may be up to – 13‰ lower than that of the original magma. Thus variations in the isotopic composition of sulfur are possible even in a mantle-derived magma whose initial δ^34 S was that of the mantle (~ 0‰). Variability of sulfur isotopic compositions do, however, give some indication of the ƒO 2 history of a magma. Constant δ^34 S of magmatic sulfides sug- gests ƒO 2 remained below the SO 2 /H 2 S boundary; variability in isotopic composition suggests a higher ƒO 2.
Sulfur Isotope Fractionation in Low-Temperature Systems
Many important ores are sulfides. A few of these are magmatic, but most sulfide ores were deposited by precipitation from aqueous solution at low to moderate temperature. At temperatures below about 400° C, sulfide species (H 2 S and HS–) are joined by sulfate (SO 4 2 − , HSO (^4) 1 − , KSO 4 1 − , NaSO 4 1 − , CaSO 4 and MgSO 4 ) as the dominant forms of aqueous sulfur. The ratio of sulfide to sulfate will depend on the
