constitutes thick beds in the Franklin Marble and is associated with zincite .... used, with interatomic distances after Muller and Roy (1974). The. 1-180 indices ...
Mineral. Deposita 31, 98-103 (1996)
MINERALIUM DEPOSITA 9 Springer-Verlag 1996
Oxygen isotope fractionation in zinc oxides and implications for zinc mineralization in the Sterling Hill deposit, USA Y.-F. Zheng Department of Earth and Space Sciences, University of Sciences and Technology of China, Hefei 230026, Peoples' Republic of China Received: 12 April 1994/Accepted: 3 February 1995
Abstract. Oxygen isotope fractionation in the zinc oxides has been calculated by means of the modified increment method. The results suggest that zincite is slightly enriched in ~80 relative to the franklinite of the spinel-type structure but considerably depleted in 180 relative to the franklinite of the inverse spinel-type structure. The zinc oxides are significantly depleted in 180 relative to water under hydrothermal and metamorphic conditions. The oxygen isotope analyses of mineral pairs including the zinc oxides and the c o m m o n gangue minerals such as calcite and quartz can constitute a sensitive isotope geothermometer. Application of oxygen isotope geotherm o m e t r y to natural assemblages is attempted for the calcite-zinc ore mineral pairs from the Sterling Hill deposit in USA. The results indicate that the temperature of the zinc mineralization may be in the range from 410 ~ to 630~ and thus lower than the metamorphic temperatures of granulite facies. A metamorphic fluid could have been involved in the formation of the zinc ore minerals. Franklinite would structurally be an inverse spinel in the infancy of its formation, and thus could have originally evolved from Zn 2 + substitution to Fe 2 + of magnetite at the high temperatures.
Franklinite is the dominant ore mineral of zinc deposits at Franklin Furnace and Sterling Hill, New Jersey. It constitutes thick beds in the Franklin Marble and is associated with zincite, willemite and tephroite or rhodonite (e.g., Ridge 1952; Metsger et al. 1958; Dunn 1985; Johnson et al. 1990). Such an assemblage of the zinc oxide and silicate minerals is markedly different from those of zinc sulfide minerals which are the common constituents of other zinc deposits. In this regard, an investigation of oxygen isotope geochemistry can be a powerful means of quantifying the physico-chemical conditions under which the zinc ore minerals formed. For this purpose, it is essential to know the temperature dependence of oxygen isotope fractionation in the zinc ore minerals.
The oxygen isotope compositions of franklinite, zincite, willemite and associated calcite as well as tephroite were determined by Johnson et al. (1990) for the Sterling Hill deposit. Unfortunately, the quantitative interpretation of the oxygen isotope data was hampered by lack of knowledge of the temperature dependence of equilibrium oxygen isotope fractionation in the zinc ore minerals. This study presents a theoretical calculation of these by means of the modified increment method (Schiitze 1980; Zheng 1991, 1993a). Application of the calculated fractionations to the isotopic compositions of coexisting calcite and zinc ore minerals gives an insight into the geochemical processes of the zinc mineralization in the Sterling Hill deposit.
Calculation method and results Studies of the equilibrium oxygen isotope properties of solid minerals are essential to quantitative applications of oxygen isotope data to geological thermometry and geochemical tracing. There are the three approaches to determine the equilibrium oxygen isotope fractionation factors between mineral and fluid and between mineral and mineral: (1) theoretical calculations (e.g., Kawabe 1978; Hattori and Halas 1982; Zheng 1991, 1992): (2) experimental measurements (e.g., Clayton et al. 1972; Matsuhisa et al. 1979; Clayton et al. 1989; Zheng et al. 1994); and (3) empirical estimates (e.g., Bottinga and Javoy 1975; Agrinier 1991). All these approaches have advantages and disadvantages with respect to the methodology of calibrating stable isotope fractionation factors (O'Neil 1986; Zheng 1993b; Zheng et al. 1994). The statistical mechanical theory of isotopic fractionation as formulated by Urey (1947) and Bigeleisen and Mayer (1947) is severely limited when applied to systems containing condensed phases (e.g., Bigeleisen 1961; O'Neil and Clayton 1964; Kawabe 1978). The modified increment method has proven to be a very efficient approach for computing the thermodynamic isotope factors of oxygen in solid minerals (Schiitze 1980; Zheng 1991, 1992, 1993a, 1994). The increment method was primarily developed by Schiitze (1980) based on the observations of Taylor and Epstein (1962) and Garlick (1966) that the degree of 180-enrichment in a set of cogenetic silicate minerals can be correlated with bond strengths in the minerals. Richter and Hoernes (1988) applied the increment method to the calculation of oxygen isotope fractionation in silicate minerals. Although their calculated results are not accurate enough
99
Table 1. Calculation of normalized tsO-increments for zinc oxides Bond
CNc,
CNo
re, + ro(A)
mc~
W,-o
Cct-o
i,-o
i~,-o
Si4+-O Fe 3 + - O Fe a + - O
4 4 6 4 6
2 4 4 4 4
1.61 1.87 1.98 1.98 2.13
28.09 55.85 55.85 65.37 65.37
1.03748 1.04631 1.04631 1,04798 1.04798
0.62112 0.40107 0.25253 0.25253 0.15649
0.02285 0.01816 0.01146 0.01183 0.00733
1.0000 0.7946 0.5012 0.5179 0.3210
Zn 2+-O
Zn2+-O
Here the mean ionic radii of Fe 3 + in high spin and low spin states are taken for the calculation
Table 2. The calculated oxygen isotope fractionations in zinc ore minerals (10 ~ In c~= A x 106/T 2 + B x 103/T -I- C) Mineral
Franklinite-1 Franklinite-2 Zincite Willemite Tephroite
I- 1s o
0.5398 0.4818 0.4876 0.7284 0.6871
Quartz-mineral A
B
1.58 1.96 1.92 0.69 0.86
6.60 7.25 7.19 4.18 4.75
Calcite-mineral C -
2.81 3.11 3.08 1.75 2.00
A
B
1.13 1.49 1.35 0.23 0.39
6.70 7.35 7.29 4.28 4.85
Mineral-water C -
2.81 3.11 3.08 1.75 2.00
A
B
2.88 2.52 2.56 3.79 3.62
-
Reference C
11.37 12.02 11.95 8.94 9.51
2.89 2.99 2.98 2.50 2.60
This study This study This study Zheng (1993a) Zheng (1993a)
Franklinite-1 is of the inverse spinel-type structure whereas Franklinite-2 is of The spinel-type structure
to be theoretical geothermometers, their work has brought the unorthdox method to our attention. Zheng (1991, 1992, 1993a, b, 1994) has modified the increment method for calculating the thermodynamic isotope factors of oxygen in metal oxide, wolframate and silicate minerals. The results obtained are in fair agreement with existing experimental and/or empirical calibrations. The calculated fractionation factors between quartz and hematite have been successfully applied by Zheng and Simon (1991) to isotopic geothermometry in metamorphic ironformations. The theoretical calibration of the wolframite-water system has been corroborated by the hydrothermal experiments of Zhang et al. (1994). The quartz-wolframite mineral pair is demonstrated to be suitable as an isotopic geothermometer in hydrothermal tungsten mineralizations (Zheng 1992). The theoretical calibration of the futile-water system has been confirmed by the low temperature experiments of Bird et al. (1993). The corroboration of the modified increment method for the common rock-forming minerals has been presented by Zheng (1991, 1992, 1993a, b). This implies that all extensions of the methodology to other minerals are potentially valid. In this context, the modified increment method has been rationalized as an accurate approach for calculating oxygen isotopic fractionation in solid minerals as a function of statistical mechanical and crystal structural effects. Oxygen isotope fractionations in willemite Zn2SiO4) and tephroite (Mn2SiO4) have been calculated by Zheng (1993a). The present calculations deal with franklinite (ZnFe204) and zincite (ZnO). Zincite has a hemimorphic hexagonal structure, with Zn 2 § in fourfold coordination, which is closely similar to that of the hexagonal zinc sulfide, wurtzite (Berry et al. 1983). Two structures of franklinite are dealt with in this study. One is of a spinel-type (like the mineral spinel), with the structural formula Zn z + [Fe 3+ ]204. In the normal franklinite, Zn 2+ cations are distributed in the tetrahedral sites, whereas Fe 3 + cations occupy the octahedral sites. The other is of an inverse spinel-type like magnetite, with the structural formula Fe 3 + [Zn 2 +Fe 3 +]O4. In the inverse franklinite, Zn 2 + cations are distributed in the octahedral sites, and one half of Fe 3 + cations occupy the tetrahedral sites and the other half the octahedral sites. The method of calculating I-1 s 0 indices for metal oxides has been described by Zheng (1991) in detail and is not repeated here. In principle, the 1-180 index is calculated by summing the normalized 1sO-increment (i',_o) for different cation-oxygen bonds in the mineral structure. The i8 O-increment is determined by the effects of 9
cation-oxygen bond strength (C,-o) and cation mass on isotopic substitution (W,_o). The cation-oxygen bond strength is defined as a function of cation oxidation state (V), coordination number (CN,) and corresponding ionic radii (r, + ro). Substantially, the 1-180 index of a mineral results from a marriage of crystal chemistry with the relationship between vibrational frequency and reduced mass. Taking the Si-O bond in quartz as reference, the normalized tsO-increments of the cation-oxygen bonds in the zinc oxides are calculated and presented in Table 1, together with the parameters used, with interatomic distances after Muller and Roy (1974). The 1-180 indices obtained are listed in Table 2. According to the principles of the increment method, the greater the I-XSO index of a mineral, the more 1SO_enriched in it. Applying the reduced partition function ratios of quartz calculated by Kieffer (1982), the thermodynamic isotope factors of oxygen in the zinc oxides are computed. Figure 1 depicts the temperature dependence of oxygen isotopic fractionations between franklinites and zincite. Using the thermodynamic oxygen isotope factors for calcite, quartz and water (Zheng 1991, 1993a), the fractionation factors are obtained for the calcite- the zinc oxides, the quartz-the zinc oxides and the zinc oxides-water systems, respectively. Figures 2 and 3 show the temperature dependence of oxygen isotopic fractionation between the zinc oxides and water and between calcite and the zinc oxides, respectively. The algebraic expressions of the fractionation factors are listed in Table 2 for the temperature range from 0 ~ to 1200~ Zheng (1991, 1993b) has estimated uncertainties in the fractionation factors calculated by the modified increment method. The following error sources have been taken into account: (1) assignment in the parameters of crystal structure, (2) assumption of coupling coefficients, (3) accuracy of thermodynamic oxygen isotope factors for the reference systems, and (4) use of polynomial approximation. According to his estimation, errors in the fractionation factors are within --b 5% of the factor values,
Discussion Z h e n g (1991) c a l c u l a t e d t h a t m a g n e t i t e h a s a n I - l S O i n d e x of 0.5404 a n d h e m a t i t e has a n I - l s O i n d e x of 0.4809. I n c o m p a r i s o n w i t h t h e I - l S O indices of t h e zinc o x i d e s
100 5
-2 I-Inverse Spinel-type Franklinite
I-Inverse Spinel-type Franklinite
2-Spinel-type Franklinite
type Franklinite
4
3-Zincite
o
5:: 3 e-
~) E
~r 2
~-8
t-0
E O
"1"-"
-10 0
2
f ill,,
"10
I,,,
I,,.I,,,I,,i1,,,
I,,,
I,I
,l,,
200 400 600 800 1000 Temperature (~
-
1200 2400 600 ~
800 1000 Temperature (~
Fig. l. The calculated oxygen isotope fractionations between franklinite and zincite as a function of temperature
Fig. 2. The calculated oxygen isotope fractionations between the zinc oxides and water as a function of temperature
(Table 2), the franklinite of the inverse spinel-type structure behaves isotopically like magnetite, whereas the franklinite of the spinel-type structure and zincite behave isotopically like hematite. As shown in Fig. 1, zincite is slightly enriched in 180 relative to the franklinite of the spinel-type structure but significantly depleted in 180 relative to the franklinite of the inverse spinel-type structure. The Sterling Hill deposit, and its similar relative at Franklin Furnace, consist of metal-rich strata which are extraordinary. The stratigraphic sequence at Sterling Hill from bottom to top is as follows: (1) a willemite + franklinite + zincite + calcite unit, (2) a willemite + franklinite + calcite unit, (3) a calc-silicate unit composed of calcite • pyroxene • garnet • other calc-silicate minerals, and (4) a horizon of angular fragments of biotite and hornblende gneiss encased.in marble (Metsger et al. 1958; Johnson et al. 1990). The Sterling Hill deposit is an isoclinically folded sequence of the zinc-, iron-, and manganese-rich strata surrounded by the Franklin Marble. Johnson et al. (1990) observed that in coexisting zinc oxide minerals from the Sterling Hill deposit franklinite is enriched in 180 (1.0 to 1.9~o) relative to zincite. This could imply that the franklinite at Sterling Hill was of the inverse spinel-t3,pe structure in the infancy of its formation. In other words, the normal franklinite observed today would have been formed by electronic transfer from the inverse franklinite at high temperatures, with structural readjustment but without oxygen isotope reequilibration. The infant franklinite of the inverse spinel-type structure could have evolved from a magnetite precursor via cation substitution:
Johnson et al. (1990) mentioned that in the ore layers franklinite generally contains 10 tool % magnetite in solid solution. The authors showed that the ores at Sterling Hilt were metamorphosed to the granulite facies and the mineral assemblages are metamorphic in origin rather than primary. O'Neil (1977) suggested that stable isotope technique may be well-suited to the study of possible inheritance of structural units of precursor minerals during mineralogical reactions. In a series of experiments, O'Neil and Kharaka (1976) heated kaolinite in water at 350~ to produce pyrophyllite and diaspore. Hydrogen isotope exchange went essentially to completion during the profound change in mineralogy. However, only 33% of possible exchange of oxygen isotopes took place concomitantly. Because of chemical and structural similarities to kaolinite, the pyrophyllite could inherit the oxygen of intact units from the precursor, but the diaspore would undergo isotopic exchange with the water during its formation (O'Neil 1977). Oxygen isotope inheritance of scheelite from wolframite under hydrothermal conditions has been suggested by Zheng (1992). Zheng (1995) shows that the oxygen isotope composition of magnetite may be influenced by starting materials and reaction paths. The preservation of intact oxygen units and thus the oxygen isotope inheritance from precursor minerals may be common in the formation of magnetites. The present study indicates that the normal franklinite can inherit the oxygen isotope signature from the inverse franklinite of its precursor. As depicted in Fig. 2, the equilibrium fractionations for the zinc oxide-water systems are less than zero and have a minimum 103 In ~ values at temperatures between 100~ and 250~ Under hydrothermal and metamorphic
Fe30,~ + Zn 2+ ~ ZnFe204 + Fe 2+
101 1"
?
9
eq
t-
o
L) o
o r
o 9
)0
+.o
Temperature CC) #
Fig. 3. The calculated oxygen isotope fractionations between calcite and the zinc oxides. Franklinite-1 is of the inverse spinel-type structure whereas Franklinite-2 is of the spinel-type structure. Fractionations involving willemite and tephroite are after Zheng (1993a)
conditions, the zinc oxides deposited from a fluid can be significantly depleted in taO relative to the fluid. The ~~O value of the fluid can be about 7 to 11%ogreater than that of the zinc oxides. These results should be useful in estimating the oxygen isotope compositions of fluids responsible for zinc mineralizations, if the formation temperatures of the zinc oxide minerals are independently determined. As delineated in Fig. 3, the present calculations suggest that oxygen isotope analyses of calcite-zinc ore mineral pairs can constitute a sensitive isotope geothermometer, because there is a large difference in the computed temperature dependence of fractionation between calcite and the zinc ore minerals at high temperatures. So can the quartz zinc ore mineral pairs (Table 2). It must be pointed out that the temperature dependence calculated is for equilibrium partitioning of oxygen isotopes between the matters, such isotope equilibrium must be thus assumed when applying the calculations to the real world. At high temperatures, the isotopic equilibrium can be readily approached with a perfectly mobile component such as water and/or carbon dioxide in the system. However, there might be a risk of partial reequilibrium during the retrograde stages. Table 3 shows an example of isotopic temperatures calculated using the calcite-zinc ore minerals fractionation curves obtained from this study and Zheng (1993a). The oxygen isotope compositions of the coexisting calcite and franklinite pairs from the Sterling Hill deposit display a good positive correlation array with slope equal to one on an 6 ~ O versus 5~sO diagram (Fig. 4). This indicates oxygen isotope equilibrium between calcite and franklinite at the temperatures of about 425 ~ to 625 ~ in
I
~L
9
