A summary on the advances in published studies on the epidote mineral group, which have been achieved during the last years, will be given. These advances include several aspects: New studies about the structure of epidote minerals have been carried out, which show the significance of the Al-Fe3+ distribution. In crystal chemistry it has been demonstrated that a number of trace elements such as Sr, Pb, rare earth elements and others, can be incorporated in epidote minerals, and they behave as important carriers of petrogenetically significant elements and isotope systems. Thermodynamic data have been determined by drop solution calorimetry on the orthorhombic and on the monoclinic members in the solid solution series in the Al-Fe3+ system. All these studies have a large effect on our interpretation of the phase relations; the existence of two miscibility gaps in the solid solution series Al-Fe3+ seems to be established now and the phase transition of clinozoisite-orthozoisite is at least qualitatively known. The Al-Fe3+ distribution between epidote and other minerals has been studied experimentally. The high pressure stability of zoisite was determined, and there are new data of the breakdown reactions at high temperature of the solid solution series in the Al-Fe3+ system. Finally, the description of natural occurrences give new insights into the petrogenetical potential of the epidote minerals: as segregation and vein minerals they are a good monitor for fluid-rock interaction; their occurrence in rocks of granitoid composition is a potential indicator for crystallization conditions; and their occurrence in reaction assemblages with REE phosphate minerals is important for the REE content of the rocks. Finally, there is some information about the deformational behaviour of epidote minerals.
Enthapies of drop-solution have been measured for a suite of epidote group minerals in molten 2PbO*B2O3 at 975 K. The goal of the study was to measure the energetics of the Al = Fe3+ exchange and to discover if the enthalpy measurements provided support for the proposed miscibility gap in the monoclinic series. The sample set spans the compositional range from Ca2Al3Si3O12(OH) (XFe = 0) to Ca2Al2Fe3+Si3O12(OH) (XFe = 1.0) and includes four orthorhombic zoisites with XFe ranging from 0 to 0.15 and twelve monoclinic samples spanning the clinozoisite-epidote solid solution from XFe=0.28 to 0.95. For the orthorhombic series, the enthalpy of drop-solution is nearly constant, within experimental error, with an average value of 493.3 ± 3.3 kJ/mol. This yields an enthalpy of formation, Hof, for zoisite of -6881.1 kJ/mol.
For the monoclinic series, the measured enthalpies show a definite asymmetric deviation from ideality. The enthalpy of drop-solution is relatively constant, within experimental error, for samples between XFe = 0.28 and 0.65 with a value of 495 ± 5 kJ/mol. With further increases in Fe the enthalpy rises steeply to 521.1 kJ/mol at XFe = 0.95. In the absence of a pure Fe-free clinozoisite, we assumed an endmember enthalpy value equal to the average zoisite value + 4 kJ/mol. This was based on the estimated enthalpy of the polymorphic transformation zoisite = clinozoisite, and the relative stability of clinozoisite over zoisite at low temperature. The trend of the data was extrapolated from XFe = 0.95 to 1.0 for the Fe-rich endmember. The endmember enthalpies of drop solution are 496.3 ± 3.3 for clinozoisite and 525.8 ± 5.3 kJ/mol for epidote. The calculated Hof from the elements for clinozoisite and epidote are -6885.2 and -6461.5 kJ/mol, respectively. The drop-solution data are in excellent agreement with the two previously reported measurements (Kiseleva & Ogorodova,1987). The asymmetric deviation from ideal mixing supports the presence of a miscibility gap between XFe = 0.5 and XFe = 0.75. The enthalpy change associated with one full Al=Fe3+ exchange in epidote is about 25 kJ/mol, which is in good agreement with the same ionic exchange in garnets and spinels.
Kiseleva IA & Ogorodova LP, Geochem. Internatl, 24, 91-98, (1987).
Members of the clinozoisite-pistacite series (epidotes) are common minerals in calcsilicate rocks, rodingites, skarns and metabasites. The epidote structure contains 3 octahedral sites, two of them can be occupied by Al and Fe3+ (M1, M3). The occupancies lead to an intracrystalline disorder state and a non-convergent ordering process occurs which can be described by the reaction Fe3+,M3 + Al,M1 = Fe3+,M1 + Al, M3. As shown by Moessbauer spectroscopy this intracrystalline exchange reaction occurs in Fe-rich epidotes (Fe pfu > 0.7) only (Fehr & Heuss-Aßbichler, 1997). With increasing temperature epidote becomes more disordered at a given composition and the intracrystalline equilibrium constant Kd increases also due to increasing amounts of Fe3+ on M1. Discrepancies for the degree of disorder exist between heat treated natural epidotes and synthetic ones (Giuli et al., 1998). Therefore the kinetic constants for the disordering and ordering reactions were studied in the temperature range 500 to 650°C at 0.3 GPa/HM buffer. Equibrium is achieved rapidly within a week for the disordering and ordering reaction as well. The occurrence of shared edges between M1 and M3 sites obviously favors the site exchange of Fe3+ and is responsible for the fast kinetics. The data clearly supports the lower Kd values of Fehr & Heuss-Aßbichler (1997) and no hint was observed for occupation of Fe3+ on M2 site.In all epidotes with Fe pfu < 0.7 all Fe3+ is located on M3 exclusively. In the compositional range 0.5 < Fe pfu < 0.7 the Moessbauer spectra consist of two M3 doubletts implying the coexistence of two distinct epidote phases with different environments of their M3 sites. This solvus is the result of convergent ordering processes of Al and Fe3+ on M3 site. The kinetic constants for the disordering and ordering reactions were studied in the temperature range 500 to 650°C at 0.3 GPa/HM buffer by means of Moessbauer spectroscopy. The sizes of the exsolved epidote phases are in the nanoscale as shown by HRTEM and in microscale the samples are homogeneous according x-ray and microprobe data. The kinetic constants are quite different for this convergent disordering and ordering reactions in contradiction to the non-convergent ordering processes in Fe-rich epidotes. The ordering process takes place very much slower than the disordering process due to the formation of new domains with different local order. During disorder reaction progress the primary high amount of domains with different local order decreases and consequently the activation energy for the disorder process is smaller compared to the ordering reaction.
Fehr KT & Heuss-Assbichler S, N. Jb. Miner., Abh., 172, 43-67, (1997).
Giuli G, Bonazzi P & Menchetti S, Terra Abstracts, 10, 22, (1998).
Infrared powder-absorption spectra were measured in the region 400 to 4000 cm-1 for twelve natural, chemical homogeneous samples along the join clinozoisite-epidote [Ca2Al2AlSi3O12(OH)-Ca2Al2FeSi3O12(OH)], collected from a variety of geological environments. In addition five of these samples were heat treated for 20 days at 500° up to 650°C/ 0.3 GPa/ HM-buffer in order to achieve intracrystalline equilibrium conditions. The spectra of the initial samples show continuous changes in the appearance of both number and shape of the bands with changing composition. Two different trends were observed. Generally Fe(III)-content causes frequency shifts with the exception of three modes (at 951, 650 and 517 cm-1), in accordance with Langer & Raith (1974). Spectra of iron rich epidotes with Fe pfu. > 0.75 show broad lines. Various heating temperatures have no effect on the shape of the spectra and vibrational mode frequencies. However, it is known by Mößbauer spectroscopy of iron rich epidotes that with decreasing temperature an ordering process of Fe(III) between M1 and M3 octahedral site occurs (Fehr & Heuss-Aßbichler, 1997). Thus it is evident that within the measured frequency range Al-Fe(III)-ordering on M1 and M3 site has no effects on the observed vibration modes. A different trend was observed for clinozoisites and epidotes with compositions < 0.75 Fe pfu. Along the solid solution series the shape of the spectra changes significantly. A comparison of heat treated samples shows that with increasing temperature the line widths get broader, and at high temperatures some of them may only be observed as spectral shoulders. Additionally the intensity of the modes changes with temperature and at low temperatures a splitting of the absorption bands may occur. This effect becomes more evident in Al-rich samples. The observed bands within the region of 400-1200 cm-1 are thought to be caused by various stretching modes of SiO4-tetrahedra. We suppose that they are related to changes in Si-O-Si angles, as there is no evidence for changes of the tetrahedral size. The spectral changes as a function of composition and temperature indicates different local environments caused by changes of the structural order parameter. According to these results there is evidence for a further ordering phenomena: in addition to the phase transformation zoisite - clinozoisite for Al-rich compositions and a non-convergent Al-Fe(III) ordering process between M1 and M3 sites, we expect for intermediate compositions a convergent ordering process of Fe(III) and Al(III) at M3 at lower temperatures on M3 accompanied by exsolution processes (Heuss-Aßbichler & Fehr, 1998).The comparison of initial materials with intermediate AlFe(III)-composition with heat treated samples shows that IR-spectra may be used as a potential geothermometer.
Fehr KT & Heuss-Aßbichler S, N.Jb. Miner .Abh., 172, 43-67, (1997).
Heuss-Aßbichler S & Fehr KT, Beih. z. Eur. J. Miner., 10, 130, (1998).
Langer K & Raith M, Amer. Mineral, 59, 1249-1258, (1974).
The epidote-group minerals zoisite (orthorhombic) and clinozoisite (monoclinic) (general formula: Ca2Al2(Al,Fe3+)[Si3O11(O/OH)]) are solid solutions between the Al- and Fe-endmembers. Their compositions can be expressed as mol% Al2Fe (=Fe3+/(Fe3++Al-2)*100). Both are widespread in metamorphic rocks and can be viewed as minerals replacing the anorthite component in rocks where anorthite is no longer stable due to pressure and/or temperature. Despite their geological significance, their phase relations are only poorly understood, especially for Fe3+-poor compositions.
To study the phase relations between zoisite and clinozoisite, we performed synthesis experiments in CFASH at 0.5 (hydrothermal technique) and 2.0 (Piston-Cylinder apparatus) GPa and 500 to 800°C. Starting material was an oxide/hydroxide mixture with 10% excess SiO2 with bulk compositions ranging from 0 to 40% Al2Fe. To overcome kinetic problems and to enhance growth rate, we added a 1 molal CaBr2-solution. Natural zoisites (Al2Fe = 0%) and clinozoisites (Al2Fe = 45%) served as seed-crystals in all runs. Oxygen fugacity was buffered at HM by conventional double capsule technique with inner Pt- (0.5 GPa-runs) or AgPd- (2.0 GPa-runs) and outer Au-capsules. All run products were analysed by EMP and XRD.
Zoisite- and clinozoisite-seed-crystals show overgrowths up to 150 µ m wide. The compositions of these overgrowths could easily be determined by EMP and are homogenous. Depending on bulk compositions, orthorhombic and/or monoclinic overgrowths formed. The compositions of the overgrowths suggest the following phase relations:
At 0.5 GPa the maximum Fe3+ content in zoisite is ~7% Al2Fe at 500°C, ~10% Al2Fe at 600°C and ~12% Al2Fe at 650°C. Coexisting clinozoisite s generally more Fe3+-rich with Al2Fe ~12% at 500°C, ~16% at 600°C and ~19% at 650°C. In the monoclinic solid solution series our data indicate a miscibility gap between 25 and 45% Al2Fe with the coexistence of two monoclinic forms. At 600°C clinozoisite I with ~31% Al2Fe coexists with clinozoisite II with ~45% Al2Fe. The solvus seems to be closed at or slightly above 650°C with a critical composition of ~40% Al2Fe.
At 2.0 GPa, the two-phase field is shifted to higher Al2Fe-values. The maximum Fe3+ content in zoisite is ~16% Al2Fe at 700°C and ~19% Al2Fe at 800°C. Coexisting clinozoisite has ~22% Al2Fe at 700°C and ~26% Al2Fe at 800°C.
The fugacity of oxygen (fO2) is a critical component in chemical systems during metamorphism, particularly because of its potential influence on the composition of coexisting fluids. The fO2 for rock assemblages is calculated from oxide pairs thought to be in equilibrium, such as rutile-ilmenite or ilmenite-magnetite. Reactions involving orthopyroxene, spinel and olivine have also been employed in fO2 calculations for some mantle systems. However, many metamorphic rocks do not contain either the oxide pairs or the high temperature assemblages necessary to calculate the fO2. These methods are unreliable below about 600°C, further limiting their application to high temperature assemblages. Therefore, new assemblages are needed for calculating the fO2 in rocks from low to medium grades. The following equation between garnet and epidote provides a new oxybarometer:
2Ca2FeAl2Si3O12(OH) = 2Ca2FeAl2Si3O12 + H2O + 0.5O2
where epidote (Ps33) dehydrates to garnet (Alm33 Grs67), water and oxygen. This equation has then been rewritten using the unmixed garnet end-members of almandine (Alm) and grossular (Grs) by the reaction: 12 Epi = 8 Grs + 4 Alm + 6 H2O + 3 O2. We have applied this equilibrium for fO2 calculations to several suites of rocks including a high-pressure Bulgarian eclogite and a Barrovian calc-pelitic rock from the Berwick Fm. of New Hampshire. The free energy data for the garnet species is from Robie and Hemingway (1995) and Holland and Powell (1998) for epidote. A continuously disordering mixing model for epidote has been used (Bird and Helgeson, 1980) and the quaternary asymmetrical mixing model of Ganguly (1996) has been used for garnet. Holland and Powell's 1998 Thermocalc program was used in the calculations.
Evaluation of log fO2-T equilibria indicates that garnet-epidote equilibria represent a robust oxybarometer. The reaction has only weak dependence on pressure: the locus of the reaction shifts by less than 0.05 log fO2 / kbar. Small amounts of Fe3+ have a minor affect on the location of log fO2 curves. For example, 5 wt.% Fe3+ in garnet will shift curves by less than 0.1 log fO2 unit when treated as a inert diluent. Similarly, the propagation of statistical counting errors from EMPA change loci by less than 0.1 log fO2. The P-T equilibria can be generated from the same garnet-epidote reaction in systems where fO2 can be either fixed on an oxide buffer or constrained between two oxygen buffer curves. Dependence of this reaction on XCO2 is very low, approximately 10°C/ 0.1 XCO2. The new oxybarometer is applicable to the wide variety of metamorphic rocks from the greenschist through granulite facies in which garnet coexists with epidote.
Bird DK & Helgeson HJ, Am. J. Sci., 280, 907-941, (1980).
Holland TJB & Powell R, J. Metamoprphic. Geol., 16, 309-343, (1998).
Robie RA & Hemingway BS, US. Geol. Surv. Bull, 2131, (1995).
Ganguly J, Cheng W & Tirone M, Contr. Mineral Petrol., 126, 137-151, (1996).
The tonalite of the Bergell pluton builds an east to west oriented body extending over a distance of 65 km from Ticino (Switzerland) to Val Malenco (Italy). The depth of emplacement estimated by the help of the Al-in-hornblende geobarometer (Hammarstrom and Zen, 1986) is approximately 25 to 28 km (P = 7 to 8 kbar) for the western part and 15 to 18 km (P = 4 to 5 kbar) for the eastern part of the pluton in fairly good agreement with other estimations based on the nappe stacking sequence of the central alps (Trommsdorff and Nievergelt, 1983) and on the cooling history of the Bergell pluton (Jaeger, 1983). The occurrence of epidote has already been reported and interpreted as epidote of magmatic origin by Cornelius in 1915 based on textural criteria. Idiomorphic to hypidiomorphic grains of epidotes are up to 3 mm in diameter. When coexisting with biotite and quartz they preserve their original shape.In the presence of plagioclase however the epidote grains are strongly resorbed. Some grains contain a core consisting of allanite. A second generation of epidote is clearly of subsolidus origin growing along veins and small fractures or overgrowing plagioclase.
The modal abundance of epidote is variable and independent of the depth of emplacement. It ranges from 3 to 9%. At some localities the tonalite is epidote free (<1%). The absence of epidote can be attributed to variations in the bulk rock chemistry and to the varying degree of deformation. In the eastern part of the pluton however epidote is clearly absent due to the lower pressure of emplacement as described by Zen and Hammarstrom (1984). The chemistry of epidote is varying systematically as a function of the emplacement depth: in the eastern part epidote is enriched in pistacite component (Fe3+/(Al+Fe3+) = 0.27) compared to epidote from the western part (Fe3+/(Al+Fe3+) = 0.18).This can be explained by a plagioclase consuming reaction that produces zoisite and increases the Al and Na content of hornblende. Ilmenite and magnetite or rutile coexisting with epidote indicate a oxygen fugacity that corresponds approximately to the NNO buffer. These observations support a low-pressure stability limit for magmatic epidote of 5 kbar in good agreement with the results from the experiments of Schmidt and Thompson (1996).
Cornelius HP, Geologische Rundschau VI, 3, 166-177, (1915).
Hammarstrom JM & Zen EA, Amer. Mineralogist, 71, 1297-1313, (1986).
Jaeger E, Convegno internat. Padova, 11, (1983).
Schmidt MW & Thompson AB, Amer. Mineralogist, 81, 462-474, (1996).
Trommsdorff V & Nievergelt P, Mem. Soc. Geol. Ital, 26, 55-68, (1983).
Zen EA & Hammarstrom JM, Geology, 12, 515-518, (1984).
Iron is present only as ferric iron in clinozoisite-pistacite solid solutions in metamorphic rocks. This may not be the case, however, in magmatic, REE-free epidotes, as evidenced by the fO2-dependent supersolidus stability curve for epidote over a range of oxygen fugacities (Schmidt and Thompson, 1996). In contrast, the cores of epidotes in calc-alkaline granitoids, dacitic dikes, and volcanic tuffs commonly are REE-enriched. The presence of REE in epidote mandates the presence of ferrous iron because of the following two substitutions: Ca + Fe3+ = REE + Fe2+ and Ca + Al = REE + Fe2+, which describe the accommodation of trivalent REE. Similarly, incorporation of Th into epidote-group minerals, can be expressed either by the two substitutions Ca + 2Fe3+ = Th + 2Fe2+ and REE + Fe3+ = Th + Fe2+ (both these reactions are accompanied by a change in Fe2+/Fe3+) or by Ca + Th = 2REE.
In view of the above substitutions, it could be inferred that allanite only crystallizes in silicic melts as a result of elevated REE and Th abundances. The value of Fe2+/Fe3+, however, can also be changed via the oxy-reaction Fe2+ + OH- = Fe3+ + O2- + 0.5 H2 (Dollase, 1973; Bonze and Menchetti, 1994). This oxidation-dehydrogenation reaction is controlled by T, P, and fH2. According to the above substitutions, the Fe oxy-mechanism would thus also have an effect on the incorporation of REE and Th into the epidote structure.
In the Tertiary Bergell pluton (Eastern Central Alps) epidote-group solid solutions occur as zoned phenocrysts almost in all parts of the intrusion which can be sampled over a large range in depth. The phenocrysts typically exhibit an LREE-enriched core, and the total REE content as well as the Fe2+/Fe3+ value decrease systematically from core to rim. An analogous relationship exists between Th and Fe2+/Fe3+ as well as between Ti and Fe2+/Fe3+. Preliminary 57Fe Mössbauer spectroscopic analyses of density separates that proxy core to rim aliquots indicate that Fe3+/Fe(total) varies from approximately 0.4 to 1.0.
Much of the interest in magmatic epidotes has centered on whether there is a threshold pressure required for epidote to crystallize in silicic systems (e.g., Zen and Hammarstrom, 1986). If the oxidation state in natural allanites is controlled by the Fe oxy-reaction, then it can tentatively be suggested that the stability of allanite in magmatic environments is perhaps more closely related to the relationship between T and fH2 rather than to the threshold pressure and the REE content of the melt.
Bonze P, Menchetti S, Am Mineral, 79, 1176-1184, (1994).
Dollase WA, Zeits Krist, 138, 41-63, (1973).
Schmidt MW, Thompson AB, Am Mineral, 81, 462-474, (1996).
Zen E-AN, Hammarstrom JM, Geology, 12, 515-518, (1986).
High U/Pb minerals, such as zircon, allow for precise dating by conventional (i.e. TIMS) techniques, but often provide rather poor links to the PT and chemical environment in which they crystallize. Other actinide-enriched minerals such as allanite can yield nearly equally precise chronological information. Importantly, however, they may also provide superior information on their crystallization environment, depending on crystal-chemical and textural properties. This has been demonstrated by a geochronological study of allanite from the Tertiary Bergell pluton (Oberli et al., 1996), which intrudes the Alpine metamorphic nappe complex of the Central Alps immediately north of the Insubric Line (SE Switzerland/N Italy).
U-Th-Pb isotopic data obtained on fragments of zoned allanite, together with zircon U-Pb results, have been interpreted to trace approx. 5 My of magma crystallization/fractionation (33 - 28 Ma). This interpretation was based on the observed correlation of Th-Pb ages, which decrease from core to rim, with strongly decreasing Th/U and LREE/HREE, and on the preservation of excess radiogenic 206Pb due to enhanced initial 230Th. As the relatively long crystallization interval of 4 My traced by the allanite data alone may appear surprising, there is a need for a careful assessment of whether the observed variations in the isotopic ages document very slow crystal growth (Berger et al., 1997), or whether they might reflect variable partial resetting of the U-Th-Pb isotopic system during slow regional cooling or by a postmagmatic thermo-metamorphic event.
An answer to this question is provided by the zonal pattern of isotope systematics preserved in allanite. The marked contrast between high Th / low U cores and high U / low Th rims of these crystals provides a reference for the analysis of potential isotopic effects generated by the redistribution of 232Th-derived 208Pb and 238U- and excess 230Th-derived 206Pb. Quantitative modeling of radiogenic Pb production and redistribution by volume diffusion, for a variety of post-crystallization temperature-time paths, generates results which are incompatible with the measured distribution. This suggests that the observed correlation of U-Th-Pb isotopic trends and ages with the chemical zoning pattern traces crystallization in a fractionating melt and does not result from syn- to post-magmatic open-system behavior. Detailed studies of allanite by micro-analytical isotope/geochemical methods can, therefore, be used to establish accurate time scales for magma evolution and fractionation, and thus can provide rate information essential for the quantification of igneous processes.
Oberli F, Meier M, Berger A, Rosenberg C & Gieré R, J. Conf. Abs., 1, 439, (1996).
Berger A, Rosenberg C, Oberli F, Meier M & Gieré R, Abstracts EUG 9, Terra Nova, 9, 458, (1997).
Allanite occurs in several different associations within the Variscan metagranites and granites of Western Carpathians. It occurs as (1) a primary phase both included in monazite, and as discrete crystal grains, (2) a secondary phase, seen as overgrowths on monazite, and as (3) a metamorphic phase formed from monazite. First two of these associations are present in the peraluminious tonalite in Tribec Mts., located in the Tatric superunit, the third in the Veporic superunit of the Western Carpathians mountain range. The replacement of primary allanite by monazite is interpretated as being due to a decrease in Ca activity caused by the onset of plagioclase crystallization. Initial breakdown of monazite during subsolidus alteration produced crystallisation of secondary allanite rims. In some cases, in the granites undergoing amphibolite facies metamorphism, monazite is completely replaced by apatite, allanite or epidote possibly due to Alpine burial (Veporic superunit). For primary allanite, distinct compositional differences occur. The discrete allanite grains are relatively highly aluminious with a composition consistent with the S-type nature of the host rock. In contrast, allanite included in monazite is extremely variable in composition, particularly in light-REE, Al, Fe and Mg. Such variation is considered to result from disequilibrium within the melt during crystallization of the monazite. The allanite overgrowths on monazite, some which show an epitaxial relationship, have a composition different from both the primary allanites, being relatively poorer in FeO, MgO and TiO2 but richer in Al2O3.
The epidote-group minerals had been discovered among the authigenic hydrothermal mineral assemblages of the orogenic K- and K-Na rhyolites which occurs within the ore-bearing volcanic structures in South Sickhote-Aline and Okhotsk-Chukotka volcanic belts. These mineral assemblages form the capillary veins intersected rhyolite textures and are disposed at the wall of lithophysa vesicls in the form of different concentric zones. Mineral assemblages had been studied with an electron microprobe techniques and with a scanning electron microscope. Euhedral elongated prismatic crystals (up to 5 mm length) of epidote-group minerals occures at the hottest parts of the fossil geothermal fields, below the location areas of the main economic interest ores. In order of decreasing abundance, minerals in epidote-bearing assemblages are quartz, calcite, chlorite, sphene, sericite, feldspar, Mn-Fe oxide. Where present, garnet was the first mineral to precipitate in vesicls followed by epidote. Pink thulite-bearing mineral assamblages, including the quartz, calcite, Fe-chlorite, rutile, K-feldspar, todorokite, vernadite and spessartine, fluorite are typical for the Ag-polymetallic ore fields, green epidote -, including the quartz, Fe-Mg chlorite, sphene, K-Na-feldspar and andradite - for the Sn-polymetallic ore fields. The disappearance epidote facies asseblages in the upper part of the geothermal fields may result from decreasing of temperature of the geothermal fluids and minor changes of fugacity of CO2, O2, H2O. Chemical compositions variation of the epidote group minerals and other authigenic hydrothermal minerals are not controlled by host-rhyolite type and had been formed under certain physicochemical conditions from the hydrothermal fluids separated from various ore-bearing deep sources. Predominance of the epidote-group minerals among authigenic hydrothermal mineral assemblages of oroginic rhyolite exposed on the surface is the evidence of the considerable erosion cut of the ore-bearing volcanic structures uplift blocks.
Direct measurements of the rare earth elements in high temperature Icelandic geothermal fluids have revealed concentrations below detection limits (<200 pmol/kg for La). Hydrothermal epidotes collected from saline and meteoric water recharged geothermal systems in Iceland have been analysed for REE by ion microprobe and this has shown that epidotes can concentrate the REE (e.g. La concentrations to 500 ppm). Partitioning of REE into epidotes has been shown to be independant of epidote crystallographic lattice sizes, and that the epidotes reflect the fluid REE composition enabling the composition of geothermal fluids to be estimated. Despite the variation by more than 3 orders of magnitude in REE concentration across a single epidote grain, partition coefficients for REE into epidotes can be estimated between 3 x 10-4 to 1 x 10 -7. Eu/Eu* anomalies indicate that of the 8 and 9 fold co-ordinate sites in epidote, REE partition into the 8 fold co-ordinate site. Phase separation occurring within a geothermal well results in lower Eu/Eu* anomalies due to the loss of H2S from the water. Further evolution of the REE in geothermal fluids due to adsorption onto other hydrothermal secondary minerals may occur, however, there is no systematic change in REE composition of bulk altered rock with depth as a consequence of the low REE concentration of the hydrothermal fluids. Comparison of the REE profiles from the saline Svartsengi system with those of the meteoric water recharged systems shows systematically lower REE concentrations for epidotes from the saline systems, but similar patterns indicating lower REE concentrations in the host fluid. The similar REE patterns of epidotes between the saline and non-saline geothermal systems indicate similar REE patterns of the host fluids and similar REE speciation. The predicted REE pattern for the Svartsengi geothermal system is similar to that observed in submarine hydrothermal fluids from the mid-Atlantic ridge, indicating that epidotes can be used to predict REE patterns for the co-exitsing hydrothermal fluids.
Epidote [Ca2Al2FeSi3O12(OH)] crystallises in the monoclinic space group P21/m. The structure contains three crystallographically different octahedral sites M1, M2 and M3. As shown by X-ray structure analyses M1 and M2 form two chains of edge-sharing octahedra parallel to the b-axis. M2 is the most regular octahedral site and is occupied by Al exclusively, whereas Al and Fe(III) can be distributed over M1 and M3. M3 is the largest and most distorted octahedral site and shares common edges with M1. Mößbauer spectra of iron-rich epidotes exhibit symmetric lines, which can be evaluated by two strongly overlapping doublets (Dollase 1973, Fehr & Heuss-Aßbichler 1997). The doublet with both the largest quadrupole splitting of 1.97 mm/s and the highest intensity (~ 95%) is assigned to Fe(III) at M3 position, the one with the lowest QS = 1.58 mm/s and intensity Fe(III) at the M1 site, indicating that Fe(III) occupies preferentially the M3 site.Especially, the quadropule splitting of the M3 site is unusually large compared with the typical values in the range between 0.5 and 1 mm/s for high-spin ferric iron in octahedral position. In order to understand the reasons for this large splitting, molecular orbital calculations in local spin density approximation have been performed for model clusters of different size based on the experimentally determined structural data of Stergiou et al. (1987). Though the best calculated value of 1.51 mm/s for the quadrupole splitting of Fe(III) at the M3 is about 20% too small, a qualitatively correct picture of the bonding of iron in epidote can be derived. The electric field gradient of Fe(III) is negative, almost axially symmetric and dominated by the anisotropy of the Fe(3d) valence shell, whereas the contributions from the ligands play a minor role. This anisotropy can directly be correlated with certain structural features of the first coordination sphere of iron.
Dollase WA, Z. Krist, 138, 41-63, (1973).
Fehr KT & Heuss-Aßbichler S, N. Jb. Miner. Abh, 172, 43-67, (1997).
Stergiou A, Rentzeperis PJ & Sklavousnos S, Z. Krist, 178, 297-305, (1987).
F-rich zoisite with up to 2.4 wt% fluorine, corresponding to XF<0.58 (XF=F/[F+OH]), was synthesised as a byproduct in high-pressure piston cylinder experiments in the systems Ca-Al-Si-O-F and Ca-Al-Ti-Si-O-F at pressures and temperatures ranging from 20 to 35 kbar and 1000 to 1100°C, respectively. The synthetic F-bearing zoisite was found to coexist with the phases fluorite - kyanite, fluorite - kyanite - CaAlFSiO4, fluorite - kyanite - AlF-rich titanite (XF=1) and fluorite - anorthite - AlF-rich titanite (XF=1). The Al-content of titanite [Ca(Ti,Al)(O,F,OH)SiO4] in these experiments ranged from XAl=0.74 to XAl=1.0 (XAl=Al/[Al+Ti]).
Data on the fluorine content of natural zoisite are rare because the quantitative analysis of fluorine in minerals used to be difficult. However, since the (OH)-F--1 exchange is a common feature in most hydrous minerals, natural zoisite has the potential to contain considerable amounts of fluorine in the appropriate chemical environment. For example, zoisite was found with XF=0.03, coexisting with AlF-rich titanite (XF=1) and fluor-apatite (XF=1) in eclogite-facies calc-silicate rocks from Tromsø, Norway.
The stability of natural zoisite is therefore affected by the fluorine fugacity of a rock. Until the effect of the fluorine fugacity on the stability of zoisite is quantified, interpretations of real, F-bearing mineral assemblages should be treated with caution if they are solely based on experimentally determined F-free phase equilibria. For example, P-T conditions of equilibrium of eclogites and blueschists based on the reaction Anorthite + H2O = Zoisite + Kyanite + Quartz may be in error if OH-F substitution is present.
Note that in high-pressure assemblages with Al-rich titanite the equilibrium of the exchange reaction Zoisite-OH + CaAlSiO4-F = Zoisite-F + CaAlSiO4-OH (CaAlSiO4-(F,OH) is the aluminium component in titanite) appears to be biased towards the left side in the experiment as well as in nature. This can possibly be linked to the apparent inability of the titanite structure to incorporate more than about XOH=0.3 water.
A number of epidote - piemontite specimens exhibiting solid solution were found in marble near Nezilovo village, the upper part of the Babuna river, about 40 km SW of Skopje. Individual crystals are found in an unsoluble residue together with hematite, rutile, chlorite, muscovite, tourmaline, quartz and albite.
The minerals are commonly highly zoned in color and chemical content. Individual grains are idiomorphic and vary from homogeniously colored, to colour zoned examples, varying in color from deep red in the cores through pink, green and yellow to colorless on the rims. All ranges of colors are present in the same hand specimens, but they are not distributed evenly in the host rock.
Chemical formulas calculated from electron microprobe analyses varies from near ideal epidote, Ca2.02Al2.00(Fe0.93Al0.06Zn0.01)(SiO4)(Si2O7)O(OH) to intermediate epidote - piemontite, (Ca1.93REE0.03Pb0.02) (Al1.98Mg0.01)(Fe0.56Mn0.45)(SiO4)(Si2O7)O(OH).
Electron microprobe analyses also indicates that there is significant mixing of the zoisite end-member, up to 21 mol%. Most of the analyses show variablity in REE, but there are significant amounts of REE, as recorded previously in samples from the same region (Bermanec et al., 1994). Up to 0.3 apfu's (Ce+La+Nd) were detected. They do not, however, show a unique pattern of REE distribution. Small, but significant concentrations of Zn and Pb were also detected (up to 0.017 apfu Pb and 0.024 apfu Zn). There is also up to 0.02 apfu Mg, but its content is commonly less than 0.01 apfu.
Unit cell dimensions were calculated from indexed X-ray powder diffraction patterns and gave a=8.87(2)Å, b=5.636(8)Å, c=10.160(7)Å and ß =115.3(2)° for yellow to colorless crystals and a=8.87(1)Å, b=5.685(7)Å, c=10.160(5)Å and ß =114.7(2)° for pink to dark red crystals.
Bermanec V, Armbruster Th, Oberhaensli R, Zebec V, Schweiz. Mineral. Petrogr. Mitt, 74, 321-328, (1994).
Epidote occurs most commonly as a product of retrograde greenschist facies metamorphism. However, epidote breakdown reactions have intermediate P-T-slopes, which can also stabilize epidote at high pressures and intermediate temperatures. We present epidote parageneses in high grade rocks which are relicts that developed during decompression at medium to high grade conditions. The first example is an eclogite boudin from Sweden (Caledonides). This eclogite was strongly affected by amphibolite facies retrogression at approximately 0.9 GPa and 700°C. During this overprint, plagioclase/pyroxene symplectites (formerly omphacite) passed into coarser grained amphibole and plagioclase. The different assemblages mainly display the decompression history of the eclogite (omphacite->symplectite->amphibole). This history is also recorded in the epidote in these eclogitic relicts. Two types of epidote are present: (1) matrix epidote is characterized by XFe (Fe/(Fe+Al)) = 0.12-0.14. Some grains indicating local (µ m) retrogression documented by increasing XFe; (2) epidote inclusions in plagioclase with higher XFe (0.15-0.18) in comparsion to type (1) implying lower pressures.
Type (2) epidote is also observed in migmatites of the Central Alps (Bellinzona, Switzerland). The migmatites with tonalitic composition contain quartz, plagioclase, biotite, amphibole and epidote. The pressure and temperatures are approx. 0.6 GPa and 700°C, respectively. This type of epidote inclusions have XFe = 0.20-0.21 and often contain a significant amount of REE.
In both examples the type (2) epidote are characterized by increasing An-content in the plagioclase adjacent to the epidote. Al X-ray maps indicate similar Al2O3 content in the epidote and the directly surrounding plagioclase. This indicates an epidote-consuming reaction producing anorthite where Al is conserved. Such a reaction could be of the type: 2Ep1 + Ab = Ep2 + An + Na2+ (+H2O)with Ep1 being richer in Al compared to Ep2. A similar reaction may also involve the formation of amphibole. Both investigated examples indicate stability of epidote at relative high pressures. The XFe value is mainly controlled by changing pressures. This implies that epidote is an important hydrous phase at higher pressures and that it can control mineral reactions (and plagioclase composition) during decompression. The release of H2O from the epidote-consuming reaction could also lead to the formation of amphibole. Alternatively, a change in fO2 could also account for the observed compositional variation in epidote.
In the Oligocene acid volcanic rocks of the Murga caldeira, Borovitza caldeira, Eastern Rhodopes, Bulgaria (Yanev and Bardintzeff, 1997), epidote yielding high amounts of REE and containing 1 to 2.8 wt% MnO has been discovered. The volcanic rocks look like ignimbrites and are strongly altered into a quartz + adularia + albite assemblage. Epidote grains have a 10 to 30 micrometer size, present irregular shapes, except scarce rectangular skeletal crystals or clusters of agglutinated small crystals. Under optical microscope, they are pale green in colour.
REE amounts in discrete epidote grains are rather variable, ranging from 4.4 to 16.2 oxides wt% (Yanev et al., 1998). Sometimes, in the same grain, REE contents are higher in the core than in the rim. The crystals analysed fill the gap existing between the REE-bearing epidotes already described in the literature and allanite. REE are essentially La, Ce and Pr, with Ce/La ratios ranging between 1.2 and 1.9. The chondrite - normalised REE patterns are fractionated from LREE to MREE, with a Pr positive anomaly. Negative correlations between REE and Ca, as well as between Mn and (Ca + La + Y + Ce), suggest substitution schemes between these elements. The pistacite index Ps, 100 * Fe3+/(Fe3+ + Al), ranges between 26 and 37.
Epidote and allanite have identical crystal structures, REE being located in the A2 site, while Fe2+ and Mn2+ occupy the M3 site. They are commonly assumed to define an isomorphous series, but incompletely because of the possible occurrence of a miscibility gap between the two end-members. Other authors, however, consider they are two different minerals. In the (REE + Th) / (Ca + REE + Th + Mn + alk) vs. (Fe2+ + Mg) / (AlIV + Fetot + Mg + Ti) plot, the combination of the data available in the literature and our results illustrate a complete series exists between epidote and allanite.
REE-bearing epidotes were found previously in: (i) metamorphic rocks, e.g. gneiss, amphibolite and schist, and (ii) metasomatic zones, e.g. skarn, quartz-carbonate vein and albitite. We describe for the first time epidote growing within completely altered volcanic rocks and formed possibly at temperatures of about 300°C.
Yanev Y & Bardintzeff JM, Terra Nova, 9, 1-8, (1997).
Yanev Y, Bardintzeff JM, Rakalov K & Jelev G, N Jb Miner. Mh, 5, 221-233, (1998).
Recent studies differ on wheter occurrence of magmatic epidote in granitic rocks requiere high pressure granite solidification, or rapid magma transport. Epidote is a major mineral in hybrid rocks of tonalite composition in the Guitiriz syn-kinematic massif. In addition to epidote, the hybrid rocks consist of biotite, hornblende and sphene in a finer grained partially recrystallized quartzofeldspathic matrix, with minor allanite, apatite, opaque minerals and zircon. Coarse hexagonal epidote crystals (Ps26) are euhedral to subhedral and up to 1.5 mm in size. Epidote is euhedral against biotite but when it is in contact with plagioclase and quartz shows highly embayed, wormy contacts that are almost myrmekitic. Some epidote crystals exhibit allanite cores. The boundary between allanite and epidote is typically sharp. Epidotes often contain apatite, biotite or sphene inclusions. Blebs of hornblende completely included within the epidote are also present, a texture thought to be as a result of hornblende reabsortion into the melt from which epidote crystallized. The textural features here described for this mineral fit most of the criteria presented in the literature and are evidence of a magmatic origin for the epidote in the mafic-intermediate rocks of the Guitiriz massif. The lack of subsolidus alteration of the rocks and the evidences for magmatic hybridization with the granite host indicate a magmatic origin as well. Calculated crystallization conditions for the Guitiriz hybridization zone of ca. 700°C, 2.5 kbar are comparable with those estimated from regional metamorphism in the study area. As it may be observed, the occurrence of magmatic epidote in Guitiriz does not imply high pressures. The mafic-intermediate rocks studied are geographically related to a strong positive gravity anomaly which has been interpreted as one of the feeding zones of the Guitiriz massif corresponding to a fracture developed in an extensional regimen. The rapid ascent of the basic magma through this fracture should have favoured magmatic epidote preservation.
Index of EUG 10 Volume
Further EUG 10 Information
Index of the Journal of Conference Abstracts
Cambridge Publications Home Page
Last Updated on Wednesday, March 17, 1999.
© 1997 Cambridge Publications