|
|
|
|||||||||||||||||
| JOURNAL HOME | HELP | CONTACT PUBLISHER | SUBSCRIBE | ARCHIVE | SEARCH | TABLE OF CONTENTS |

Via Corvetto 33, 09014 Carloforte, Italy
Department of Geosciences, Oregon State University, Corvallis, Oregon 97331
Dipartimento di Geoingegneria e Tecnologie Ambientali, Università di Cagliari, Italy
Istituto di Geologia Ambientale e Geoingegneria del C.N.R., Sez. di Cagliari, Università di Cagliari, Italy
Corresponding author: e-mail, rossana.simeone{at}tiscali.it
| Abstract |
|---|
|
|
|---|
Tresnuraghes is characterized at depth by chalcedony ± quartz ± barite veins within a 50-m-wide zone of K-feldspar-quartz-illite alteration and overlying local occurrences of chalcedony sinter, which define the paleosurface. Kaolin deposits near the paleosurface are characterized by zonation outward and downward from an inner shallow zone of kaolinite 1T-opal ± dickite ± alunite (<20-µm-diam grains) to an outer deeper kaolinite 1M-montmorillonite-cristobalite. This zonation indicates formation by descending acidic fluids. The system evolved from ascending weakly acidic or neutral fluids that boiled to produce H2S-rich vapor, which condensed and oxidized within the near-surface vadose zone to form steam-heated acid-sulfate waters and kaolin alteration.
At Romana, veins at depth contain chalcedony or quartz and minor pyrite and are enclosed in up to 20-m-wide zones of kaolinite 1T-quartz alteration. Near hydrothermal vents along the paleosurface, chalcedonic silica is enclosed within a zone of kaolinite 1T-alunite (<50-µm-diam grains)-quartz-opal ± dickite ± cristobalite. Kaolin quarries near the paleosurface display outward and downward zoning to kaolinite 1T-opal ± cristobalite and then to montmorillonite-kaolinite 1T ± opal, consistent with formation by descending low pH fluid. The siliceous and advanced argillic alteration along steep conduits formed from acidic ascending magmatic-hydrothermal fluids, whereas the near-surface kaolin formed from steam-heated meteoric waters.
Alteration mineral assemblages and stable isotope data provide evidence of
the temperature and source of hydrothermal fluids. Barite from Tresnuraghes
(average
18O = 17.1
,
34S
= 18.8
), one alunite sample from Romana (
18O =
12.0
,
D = 3
,
34S = 16.7
),
and quartz from both localities (
18O = 15.922.0
)
formed in hydrothermal feeders. Source fluids were likely mixtures of meteoric
water and minor magmatic fluid, similar to other epithermal systems.
Kaolinite-dickite minerals from the kaolin deposits (
18O
= 16.621.4
,
D = 43 to 53
) formed from
steam-heated meteoric water having
D = 20 per mil,
consistent with the presence of anomalous Hg and fine-grained Na- and Fe-poor
alunite. The laterally extensive kaolin deposits in Sardinia, and possibly
similar deposits elsewhere in the world, appear to represent the uppermost parts
of large hydrothermal systems that may be prospects for gold at depth.
| Introduction |
|---|
|
|
|---|
|
| Regional Geology |
|---|
|
|
|---|
Pre-Tertiary rocks are not exposed in these areas but are likely Hercynian-age granitoids, which are unconformably overlain by Carboniferous-Permian volcano-sedimentary rocks and Mesozoic carbonate rocks elsewhere in the region (Carmignani et al., 1989, 1992).
Younger rocks in the region are composed of late Miocene sedimentary sequences made up of volcaniclastic and marine deposits (Cherchi and Montadert, 1982), and these are unconformably overlain by Pliocene-Quaternary basalts.
| Tresnuraghes Area |
|---|
|
|
|---|
|
|
|
Silicified rock, siliceous veins, and sinter: Silica minerals at Tresnuraghes occur in veins and sinter and as pervasive silicification of the pyroclastic rocks. These include opal (as opal CT), chalcedony, and microcrystalline quartz. Opal and some chalcedony also closely associated with kaolinite. Opaline layers a few centimeters thick occur at the top of the alteration zones. At Porto Alabe, a chalcedony sinter, more than 10 m thick, is exposed in a gravel pit. It is characterized by well-developed horizontal laminations with columnar structures perpendicular to and between laminations (possibly evidence of bacterial activity during silica deposition). Plant stems 1 cm in diameter and up to 5 cm in length were also recognized. These stems are unequivocal evidence of a sinter origin (White et al., 1989). At Nuraghe Tepporo, a silicified breccia crops out discontinuously along an east-westtrending fracture. It is characterized by angular centimeter-sized clasts of white opal in a red opaline matrix.
Quartz and chalcedony chiefly occur in veins that appear to underlie opal-, alunite-, and kaolinite-bearing zones and are here referred to as siliceous feeders. At Ischia Ruggia, veins and silicified zones of wall rock up to 3 m thick occur in north-southtrending structures displaced by late east-west faults. The veins are composed of chalcedony and quartz. At San Marco, stockwork veinlets of chalcedony and barite are hosted by a silicified pyroclastic flow. Barite commonly fills vugs in the chalcedony veinlets.
Rock samples from the different silicified bodies contain high Hg (up to 800 ppb) and As (up to 200 ppm). Enrichments in Hg and As are associated with anomalous Au in some rock samples at Porto Alabe (up to 0.5 ppm Au) and at San Marco (up to 0.1 ppm Au). Silicified rock at San Marco is also characterized by high Ba, ranging from 100 to 1,000 ppm (Gold Mines of Sardinia, 1999).
Kaolinite 1T-opal ± quartz ± dickite ± alunite: This alteration is pervasive in the Salamura, Punta Allò, and Nuraghe Tepporo areas and is typical of the kaolin quarries. It extends vertically up to 20 m and laterally for less than 30 m at Nuraghe Tepporo and up to 30 m vertically and up to 100 m laterally at Salamura and Punta Allò quarries (Fig. 3). Differential thermal analyses (Garbarino et al., 1991a) indicate a bulk composition of up to 50 wt percent silica minerals and up to 50 wt percent of kaolinite group minerals + alunite. Alunite and dickite increase in abundance where the alteration is more intense and destroys the texture of the primary ignimbrite. Alunite and kaolinite with lesser opal CT and quartz form very fine grained powdery aggregates and pervasively replace feldspar phenocrysts and matrix. These mixtures are white to tan in hand sample, but in thin section, most alunite-quartz and alunite-kaolinite mixtures are light to medium reddish-brown and consist of grains that are <20 µm in diameter. Individual crystals commonly cannot be discerned (Fig. 4B, sample A-6). Opal occurs as aggregates within alunite and kaolinite and also in veinlets that crosscut the kaolin deposits. Fine-grained (<20-µm) opal-alunite mixtures are also found at Punta Allò. Alunite is K rich and Fe and P poor. Reconstructed alunite compositions from <20-µm-diam grains mixed with quartz, opal, and kaolinite from Tresnuraghes are shown in Table 2.
|
|
Kaolinite 1Md-montmorillonite-cristobalite: This alteration assemblage forms a wide halo extending up to 500 m from the quarries.
K-feldspar-quartz-illite: This alteration occurs in the ignimbrites as a selvage to the siliceous (quartz-chalcedony) veins that crop out at Ischia Ruggia and extends for several meters (up to 100 m) away from the veins.
Montmorillonite-quartz-illite: This alteration occurs as a pervasive outer alteration halo peripheral to the siliceous veins and inner K-feldspar-quartz-illite alteration halo at Ischia Ruggia (from 50500 m). Rock texture is preserved, and primary feldspar occurs as relics. Similar alteration occurs as an envelope to the Porto Alabe sinter.
| Romana Area |
|---|
|
|
|---|
|
Hydrothermal alteration
Hydrothermally altered rocks in the Romana area cover 4 km2. There
are four major kaolin quarries: Locchera, Scanu 1, Scanu 2, and Donigazza (Fig.
5), and Locchera is the largest kaolin deposit in Italy with measured reserves
of approximately 100,000 t and production of 20,000 t/y.
Common minerals are kaolinite 1T, dickite, and silica minerals (from opal CT to chalcedony to quartz). Alunite occurs in the Locchera and Scanu quarries and as a minor component at DAlgata. Montmorillonite occurs in the less altered areas and as the capping alteration at the Scanu quarry. The only sulfide present is pyrite. The altered rock extends over a vertical interval of at least 300 m (Fig. 6).
|
|
Kaolinite 1T-quartz: This is a pervasive alteration that occurs as an envelope of up to 20 m in width around the silicified bodies. Kaolinite is present throughout but decreases in abundance away from the kaolinite-quartz zone and grades outward into a kaolinite-montmorillonite ± opal zone.
Kaolinite 1T-alunite-quartz-opal ± dickite ± cristobalite: This mineral association is found at the Locchera quarry. Alteration is pervasive and intense and appears to be controlled by subvertical faults. At present the quarry face is 50 m high. Within the exposures no vertical zonation of the minerals has been observed except for a few silicified layers (1 m thick) that occur at different levels in the quarry. However, alunite decreases in abundance laterally away from the controlling fluid conduits. Alunite is intergrown with kaolinite and quartz and produces a white to yellow aspect to the rock. The alunite-kaolinite mixtures fill and surround euhedral prismatic quartz grains up to 0.3 mm long. Locally, vuggy cavities where feldspar has been removed by hydrothermal alteration are lined with a <0.3-mm-thick zone of euhedral rhombic alunite crystals (<1050 µm in diam) projecting into vugs (Fig. 4A). Electron microprobe analyses of the alunite indicate that it is K rich and contain 4 to 8 mole percent end-member natroalunite, about 1 mole percent phosphorus in the aluminum site, and about 1 mole percent jarosite component (e.g., Fe/Fe + Al; Table 2). Centimeter-wide veins of dickite crosscut this alteration from the center to the margin of the quarries. Pervasive supergene minerals consisting of Fe and Mn oxides occur in the shallower part of the deposit (up to 45 m below the surface) and fill fractures.
At the Scanu quarry a similar association occurs. The alteration consists of quartz-opal-kaolinite ± cristobalite and rare alunite. A 3-m-thick montmorillonite-rich bed occurs as a horizontal layer at the surface and is interpreted to be late and unrelated to the main hydrothermal alteration event.
Kaolinite 1T-opal-montmorillonite ± cristobalite: At the Donigazza quarry vertical zonation of hydrothermal minerals is observed, within the 20-m-high quarry face. The intensity of alteration decreases with depth. The upper part (<10 m in thickness) of the alteration zone is characterized by pervasive kaolinite-opal-cristobalite, but the primary rock texture is preserved. At the base of the quarry face the alteration grades into opal-kaolinite-montmorillonite with relatively high silica content. Primary quartz and plagioclases are preserved.
Kaolinite 1T-montmorillonite-opal: This alteration occurs as a pervasive zone in the pyroclastic units of Romana. The alteration is more intense close to the silicified bodies and the kaolin deposits. Montmorillonite-opal occurs in veinlets filling fractures that crosscut the kaolinite 1T-montmorillonite-opal alteration and is commonly associated with Fe oxides and Fe hydroxides.
|
|
| Stable Isotopes |
|---|
|
|
|---|
|
34S
value of the Romana pyrite is 1.1 per mil, similar to magmatic sulfur (3
to +3
; Rye et al., 1992). Alunite is finely intergrown with kaolinite and
quartz, so sulfur isotope determinations are from mixtures. The
34S
values of the two alunite samples from Romana are 5.0 and 12.0 per mil; alunite
from Tresnuraghes has a
34S value of 12.0 per mil.
The
34S values of the two barite samples from
Tresnuraghes are 17.6 and 19.0 per mil. The
34S
values of barite and alunite differ significantly from that of pyrite, so
alunite cannot be formed solely from the weathering of pyrite (cf. Field and
Lombardi, 1972; Field and Gustafson, 1976; Rye et al., 1992). The difference
between the
34S values of pyrite and alunite are
less than would be expected for low-temperature (<200°C) isotopic
equilibrium (Ohmoto and Lasaga, 1982), so these minerals are not in isotopic
equilibrium.
The combined oxygen and sulfur isotope compositions of alunite have been used
by Rye et al. (1992) to distinguish between alunite (and barite) of steam-heated
and magmatic-hydrothermal origins. The
18O values
of two Tresnuraghes barite samples are 15.4 and 18.9 per mil. The
34S
and
18O values of the barite are similar to those
of alunite in well-documented epithermal systems formed at 200°C or less and
having H2S/SO24 ratios in the hydrothermal
fluid of about unity (Fig. 7; Rye et al., 1992).
The
18O value of a mixture of alunite >
kaolinite > quartz in volume proportions of ca. 60/30/10 from sample Rom-7 is
16.7 per mil, from which we can estimate a range of
18O
for the combined oxygen in the sulfate and hydroxyl sites of alunite of about 16
to 17.5 per mil. This estimate is based on the proportions of the three
minerals, the constant alunite-kaolinite fractionation factor of 3.3 per mil in
the range 25° to 100°C, a quartz-alunite fractionation of 10 per mil at 25°
to 100°C, and alunite composition in which oxygen is 4/7 in the sulfate site
and 3/7 in the hydroxyl site (Table 4, Fig. 4A). We assume that the
18O
composition of the water in equilibrium with the assemblage can be modeled using
the alunite-water fractionation factor of Stoffregen et al. (1994). They report
that the exchange rates for oxygen in the hydroxyl site and the sulfate site are
the same within error, which permits us to estimate the total alunite-water
fractionation factors as the sum of the fractionation factors for the hydroxyl
and sulphate sites multiplied by the respective mole fractions of oxygen in
these sites. When this estimated
18O value and the
corresponding
34S value of 12.0 per mil are plotted
in Figure 7, the composition lies close to that of low-temperature alunite from
steam-heated environments, such as Tolfa, Italy (Lombardi and Sheppard, 1977).
Oxygen and hydrogen isotopes of kaolinite minerals
The
18O values of kaolinite (n = 3) and
dickite (n = 3) have a range of 16.6 to 21.4 per mil, whereas the
D
values fall in a narrow range of 43 to53 per mil. The compositions lie on
a
18O versus
D plot between the
"kaolinite line" of Savin and Lee (1988) and the field of Tolfa
kaolinites (Fig. 8). Three kaolinite values from Romana lie very close to the
kaolinite line, consistent with the formation of kaolinite during weathering
(e.g., soil formation) or supergene processes in equilibrium with meteoric
waters at ambient temperatures of ca. 25°C. Dickite, has slightly lower values
of
18O of 16.6 to 19.0 per mil and higher values of
D of 43 to 47 per mil, compared to the Romana kaolinites
(Table 4, Fig. 7). The single alunite sample Rom7 is a mixture of
alunite-kaolinite-quartz from Romana and has a
18O
value of 16.7 per mil and a
D value of 3 per mil.
Therefore, the alunite and dickite are not in isotopic equilibrium, which
suggests that they formed at different times or one of the minerals later
exchanged with a water of different isotopic composition or temperature.
Oxygen in silica minerals and barite
Hydrothermal quartz from Romana and Tresnuraghes has
18O
values ranging from 15.9 to 22.0 per mil (Table 4, Fig. 8). The
18O
value of unaltered, optically clear sanidine from the rhyolitic host rock is 6.7
per mil, which should be similar to the original bulk magmatic
18O
value of rhyolite and quartz in rhyolite. This value of 6.7 per mil is lower
than the
18O values of 8.5 to 13 per mil reported
for 18O-enriched Quaternary silicic volcanic rocks in the Latium area
of Italy (Turi and Taylor, 1976; Ferrara et al., 1985). Thus, the hydrothermal
quartz is strongly enriched in 18O relative to unaltered host rock. A
slightly lower
18O value of 13.0 per mil was
measured in opal from Tresnuraghes.
| Environment of Formation |
|---|
|
|
|---|
The surface exposures at Tresnuraghes and Romana generally lack sulfides, and there is no drill information, so the sulfidation state of the hydrothermal fluids cannot be directly determined. However, we can estimate other characteristics of the hydrothermal systems based on the silicate and sulfate mineralogy.
Tresnuraghes
At Tresnuraghes the occurrence of kaolinite associated with alunite indicates
an acid environment with a pH between 2 and 4 (Hemley et al., 1969; Stoffregen,
1987). By comparison with active geothermal systems, the kaolinite-opal
association suggests temperatures up to 120°C (Reyes, 1990). At low pH,
hydrogen ions inhibit the polymerization of dissolved silica (Fournier, 1985). Therefore, neutralization of the fluids was required to cause silica
precipitation in the form of opal (Sillitoe, 1993).
The mineral assemblage grades horizontally from proximal kaolinite 1Md and kaolinite 1T-opal into distal kaolinite 1 Md-montmorillonite-cristobalite. Although there are no geologic data to indicate the significance of the crystal structure of kaolinite (cf. Bailey, 1993), the disordered monoclinic form (1Md) most likely occurs at lower temperature than the ordered triclinic form (1T), by analogy with the muscovite structure. Montmorillonite forms at temperatures up to 140°C and at a pH of ca. 5 to 6 in modern geothermal systems (Reyes, 1990; Simmons and Browne, 1998), and therefore may reflect alteration by fluids that had previously been partly neutralized by reaction with the host rock. Because montmorillonite-bearing assemblages occur at depth in the quarries, we infer that fluid flow, cooling, and neutralization was lateral and downward (cf. Schoen et al., 1974). This may explain the shallow depth extent of these deposits.
Based on the S-O-H isotopes and the alteration zoning at Tresnuraghes the kaolinite-opal ± alunite deposits likely formed in a steam-heated environment where hot spring-derived, H2S-rich steam condensed in the near-surface vadose zone, mixed with local ground water, cooled, oxidized, acidified, and descended (cf. Schoen et al., 1974).
The fine-grained nature of alunite (<20 µm) and its intimate intergrowth with light brown kaolinite and opal are also consistent with a steam-heated environment. Thompson et al. (1999) noted that low-temperature alunite of both steam-heated and supergene origin is typically <50 µm in diameter, whereas coarser alunite (50100 µm) is characteristic of high-sulfidation quartz-alunite assemblages formed at >100°C (Hedenquist et al., 2000). The low Na, P, and Fe contents and high K contents of the Tresnuraghes alunite contrast with mixed Na-K-alunite with substantial P in quartz-alunite alteration associated with porphyry copper deposits (Watanabe and Hedenquist, 2001; Lipske, 2002) where temperatures exceed 100°C.
The fluids that formed the central and deeper siliceous feeders were hotter and geochemically different from the fluids that formed the overlying and distal kaolin deposits. The occurrence of K-feldspar-quartz-illite alteration proximal to the vein systems with quartz-chalcedony and local barite suggests circulation of weakly acidic to near-neutral pH fluid and temperature higher than 220°C (cf. Reyes, 1990). Sinter, which occurs at Tresnuraghes, commonly forms from boiled, alkali-chloride waters where the paleowater table intersects the paleosurface (Hayba et al., 1985; Browne, 1991). Thus, the fluid-dominated hydrothermal system of Tresnuraghes most likely represents an upflow zone of neutral alkali-chloride fluid that is overlain by a steam-heated, acid-sulfate environment containing kaolinite deposits. The Tresnuraghes system can be classified as adularia-sericite type and, due to the lack of sulfides, is inferred to have formed under low-sulfidation conditions based on common sulfide-silicate associations (Hayba et al., 1985).
Rock samples from the kaolin quarries show an enrichment of As and Hg in the siliceous feeders with generally low amounts of chloride-complexed metals such as Pb, Sb, and Zn values (Gold Mines of Sardinia, 1999). Because Hg is readily transported in the vapor phase and As is not (Barnes and Seward, 1997), the Hg-As association suggests fluctuation from aqueous transport of As in the hydrothermal fluid to vapor transport of Hg in the steam-heated vadose environment. Possibly, this is the result of a late steam-heated overprint upon an earlier and underlying As-rich rock produced by condensed fluids.
|
D value of 20 per mil at
temperatures of 40° to 50°C (Table 4, Fig. 8). These data suggest dickite
formation from local meteoric water at about 25° to 50°C, but dickite
typically forms at a temperature of >120°C in geothermal systems (Reyes,
1990). At 120°C, Tresnuraghes dickite would be in equilibrium with fluid having
a
18O value of 8 per mil and a
D
value of 25 per mil, similar to arc-type magmatic water. A magmatic water
origin is considered unlikely because of the low-temperature and near-surface
environment. Exchange between meteoric water and Paleozoic carbonate rocks at
250°C could also have caused 18O enrichment of water to
18O
values of 8 per mil but is also considered unlikely because it requires a
water/rock mass ratio of about 0.3 (Fig. 8), which is lower than the water/rock
ratio required for strong hydrothermal alteration (cf. Taylor, 1979). Our
preferred model is that the dickite likely originally formed at >120°C from
meteoric water with a small component of magmatic water, but reequilibrated
isotopically at 25° to 50°C in the steam-heated environment. The isotopic data
do not rule out a supergene origin for the dickite and associated kaolinite by
acid produced by weathering of pyrite at 25°C. However, we know of no geologic
examples of dickite that has formed via supergene processes. Furthermore, the
34S
value of 12.0 per mil for a single alunite sample from the
kaolinite-alunite-opal assemblage is unlikely to have been produced only by
weathering of pyrite with a
34S value of ~O per
mil.
Barite from central siliceous feeders a Tresnuraghes plots on the
34S
versus
18O diagram in the field of
magmatic-hydrothermal alunite and sulfate produced by ~200°C fluids with H2S/SO4
ratios of about unity (Fig. 7; Rye et al., 1992). Equilibrium between barite and
quartz and the local meteoric water would imply that these two minerals formed
at temperatures >75° and >100°C, respectively (Fig. 8B). At 100°C the
quartz from the sinter sample would be in equilibrium with meteoric water with a
18O value of 3.7 per mil, similar to the dickite
minerals (Table 4). We assume a higher temperature for the deeper quartz
veinlets and calculate equilibrium
18O values for
water of 0.3 to + 2.8 per mil at 160°C (e.g., >57-m depth at hydrostatic
pressure) and 3.8 to 6.9 per mil at 220°C (>250-m depth). These 18O-enriched
compositions would indicate a component of magmatic water but also could be
interpreted as meteoric water enriched by exchange with Paleozoic carbonate or
Cenozoic volcanic rocks at water/rock ratios of ~5 to 0.1 (Fig. 8A).
Romana
Blanketlike kaolin deposits at Romana are characterized by hydrothermal
alteration proximal to hydrothermal veins and feeders with the association
kaolinite 1T-alunite-quartz-opal ± dickite ± cristobalite. The mineralogy
suggests a formation temperature of <120°C for kaolinite-rich (1T)
assemblages and 120° to 200°C for kaolinite- and dickite-bearing assemblages
based on comparisons with active geothermal systems (Reyes, 1990). However,
similar to Tresnuraghes, the S-O-H data indicate temperatures of <120°C. The
coexistence of cristobalite, opal, and quartz in these deposits could be
explained by rapid change of fluid-flow rate, or temperature along the flowpath
or with time (Dove and Rimstidt, 1994), or partial transformation of amorphous
silica to quartz and chalcedony (Fournier, 1985). The latter requires that
proximal fluids had a temperature below 120°C. More distal kaolinite 1T-rich
alteration at Romana has a steam-heated origin at <50°C, similar to
Tresnuraghes, and is consistent with the fine-grain size (<50 µm)
and K-rich composition of coexisting alunite. At the Scanu and Donigazza
quarries, the vertical zonation of the alteration minerals is similar to that
observed in the Tresnuraghes quarries and supports the origin of the kaolin
deposits by descending fluids.
Silicified feeders, including silicified rocks and the quartz-, chalcedony-, and pyrite-bearing veins, at Romana are enveloped in rock that has been altered to kaolinite 1T, alunite, quartz, and local dickite and therefore formed from fluids with lower pH than the fluids that produced the quartz-chalcedony veins enveloped by K-feldspar-illite at Tresnuraghes. Hypogene assemblages with kaolinite, alunite, and pyrite require both low pH (<2) and a high-sulfidation state (cf. fig. 5 of Heald et al., 1987). The siliceous layers in the DAlgata body are similar to the shallow parts of high-sulfidation epithermal systems where geothermal water (Browne, 1991) discharged on the surface after extensive lateral flow and dilution (Sillitoe, 1993). Deeper parts of hydrothermal systems at Romana are locally exposed in the 250 m of relief (Fig. 5). High Hg and As values of silicified bodies are consistent with fluctuating vapor- and liquid-dominated conditions, as at Tresnuraghes. The Rattari and Locchera silicified bodies contain weakly anomalous Cu and Au, similar to the common Cu-As-Au metal association in the high-sulfidation environment (Hayba et al., 1985). Based on the mineralogical and geochemical features and the S-O-H data, Romana is a shallow epithermal system in which ascending acidic fluids produced selvages of kaolinite-rich alteration along hydrothermal feeders that were overlain by near-surface blanketlike kaolin deposits. The deep silicified feeders and hydrothermally altered rocks therefore are interpreted to have formed in an acid-sulfate (Heald et al., 1987) or high-sulfidation epithermal environment (Stoffregen, 1987).
The hydrogen and oxygen isotope values of three Romana kaolinites and one
dickite from the blanketlike near-surface kaolin deposits plot along the
kaolinite line and are consistent with either a steam-heated or supergene origin
because they are in isotopic equilibirium at 25° to 50°C with meteoric water
having a
D value of 20 per mil. The sulfur isotope values
of alunite (
34S = 5 and 12
) are consistent with
partial oxidation of H2S at <100°C, similar to that observed in
other steam-heated localities such as Tolfa, Italy. The sulfur isotope values of
alunite also differ from Romana pyrite (
34S = 1.1
),
and therefore all sulfate cannot be derived from supergene oxidation of pyrite.
The alunite with the
34S of 12 per mil can be
interpreted to be of magmatic-hydrothermal origin because it is in isotopic
equilibrium at 100° and 220°C with water having
18O
values of 0.3 and 8.8 per mil, respectively, and a
D value of
1 per mil (Table 4 shows the calculation for 100°C). The
18O
of this fluid is consistent with a mixture of magmatic and meteoric water, and
the
D is similar to local magmatic waters sampled from
volcanoes in Italy and the Mediterranean basin (Fig. 8A;
D ~ 0
± 10
; Panichi and Noto, 1992; DAmore and Bolognesi, 1994; Giggenbach,
1997). Quartz from the central zone of silicified rock at DAlgata and
kaolinite-alunite-opal ± dickite alteration at Locchera has
18O
values with isotopic equilibrium temperatures higher than 120°C. At 160°C this
quartz would have been in equilibrium with water having
18O
values of 1.9 and +4.2 per mil, respectively, which is consistent with the
central silicified feeders having formed from a mixture of meteoric and minor
magmatic waters (Fig. 8B, Table 4).
Isotopic composition of Miocene meteoric water
The Sardinian kaolinites and dickites are calculated to have formed in
isotopic equilibrium with meteoric water having a
D value of
20 per mil, which is higher than present-day Sardinian meteoric water (
D
= 38 to 54
; Caboi et al., 1994). The Sardinian kaolin minerals most
likely formed from Miocene meteoric water, which is significantly heavier than
present waters, and their formation is possibly also attributable to the wetter
or cooler climate of the western Mediterranean in the Miocene than the climate
of today. This situation is similar to the one proposed by Arribas et al. (1995)
for the Miocene Rodalquilar epithermal Au district of southeastern Spain.
| Conclusions |
|---|
|
|
|---|
In the Romana area silicified bodies (chalcedony or quartz ± pyrite) with locally anomalous As, Hg, Cu, and Au formed along faults at 100°C and possibly as high as 220°C, from ascending acidic water. These bodies are flanked and overlain by rock alteration zoned from inner kaolinite 1T-alunite-quartz ± dickite at Locchera quarry to outer kaolinite 1T-quartz. The S-O-H isotope composition of alunite and quartz from the central silicified zone suggests contributions of magmatic-hydrothermal sulfur and water, whereas the data for near-surface kaolinite, dickite, and alunite suggest equilibration at 25° to 50°C with meteoric water. The mineral zonation in the kaolin quarries from inner kaolinite 1T + opal to montmorillonite suggests that fluids descended from the surface and were gradually neutralized by host rocks. Mineralogical and isotopic features of Romana suggest the silicified feeders represent the shallow part of a boiling acidic and high-sulfidation hydrothermal system, similar to the quartz-alunite veins in the Virginia City area, Nevada (Vikre, 1998), Nansatsu district, Japan (Hedenquist et al., 1994), and San Juan Mountains, Colorado (Larson and Taylor, 1987). The uppermost part of the system formed kaolin and dickite at temperatures up to ~120°C in an environment that likely included both steam-heated conditions and local mixing and cooling of meteoric and magmatic fluids.
The Sardinian kaolin deposits are remarkable for the large amount of clay and
alunite produced from feldspar and volcanic glass by hydrolysis, which requires
large inputs of acid and sulfur from magma-derived fluids. The
18O
of the kaolinite minerals in the blanketlike kaolin deposits is higher than
other steam-heated alunite-bearing occurrences such as Marysvale, Utah
(Cunningham et al., 1984). The 18O-enriched kaolin minerals in the
Sardinian deposits suggest that the kaolin alteration occurred at very low
temperatures of ~25° to 50°C from fluids that originated in the steam-heated
environment but were subsequently cooled (cf. Rye et al., 1992). The stable
isotope data argue against a supergene (weathering) origin for the kaolin and
alunite and suggest that the low-temperature steam-heated waters were modified
Miocene meteoric waters. These waters had a
D of 20 per
mil, which is distinguishable from modern meteoric water with a
D
of 38 to 54 per mil (Caboi et al., 1994). The isotopic data suggest a
contribution of magmatic sulfur, oxygen, and hydrogen in both the Tresnuraghes
and the Romana systems along steeply dipping faults flanked by steam-heated
environments.
Future studies of these deposits are needed to understand in more detail the paragenesis of different silica minerals and to correlate trace metal enrichments with sulfide and ore mineral occurrences in the hydrothermal environment. At present, the lack of information on sulfide minerals precludes rigorous definition of the sulfidation state. However it is evident that the current exposures of kaolin deposits represent the ancient Miocene paleosurface and steam-heated environment, with only limited exposure of the underlying fluid-dominated epithermal environment. Hence, the deeper parts of these systems may be a prospect for precious metals.
| APPENDIX |
|---|
|
|
|---|
One hundred and twenty samples were collected for X-ray diffraction analyses. The mineralogical composition of the specimens was determined using a Rigaku geigerflex diffractometer at 30 kV, 30 mA, Cu tube, and Ni filter. The structural order of the kaolinite phases was determined on the basis of peak morphology, taking into account both intensity and neatness of the basal reflection or the presence of near-continuous diffraction bands. All the peaks obtained were compared with the JCPDS (Joint Committee on Powder Diffraction) datafile (1985).
Twenty samples of mineral separates of quartz, kaolinite, barite, alunite,
and pyrite were prepared for
18O,
D
and/or
34S analyses by standard techniques in the
laboratories of Oregon State University. Oxygen gas was extracted from silicate
via laser fluorination with ClF3 following the method of Sharp
(1990). Hydrogen gas was liberated via heating to 1,400°C in vacuum and
reduction of water via uranium at 650°C (Bigeleisen et al., 1952). Sulfur was
extracted from sulfide via a cupric oxide combustion method and from sulfate via
conversion to Ag2S, using the method of Thode et al. (1961), followed
by cupric oxide combustion.
Oxygen and hydrogen isotopes were analyzed in the laboratories of Washington
State University. Oxygen isotope analyses of sulfate were done in the U.S.
Geological Survey laboratories at Denver, using graphite-furnace reduction and
continuous flow mass spectrometry. Replicate analyses of NBS and laboratory
standards yielded reproducibilites of ±0.2, ±4, and ±0.1 per mil for
18O,
D, and
34S, respectively. The
18O
and
D analyses are reported per mil relative to V-SMOW.
Analyses of NBS-30 biotite standard gave a value of
D = 65
per mil and
18O = +5.1 per mil. The
34S
values are reported relative to Canyon Diablo Troilite.
Calculation of water-rock exchange
Calculations of equilibrium isotopic compositions of water and water/rock
ratios were made following the standard methodology and equations of Taylor
(1979, 1997) and Field and Fifarek (1985) at 250°C (lower temperatures would
yield final water compositions little-shifted from starting composition).
Isotopic exchange reactions for hydrogen are governed by
where
DH2O
is the final water
composition,
D H2Oi is the initial water
composition,
Drocki is the initial rock
compostion,
r-w is the rock-water isotopic
fractionation, and w/r is the atomic water/rock ratio for hydrogen [= water/rock
mass ratio x (wt % H in water)/(wt % H in rock)]. Equation (1) allows
calculation of final isotopic composition of waters reacted with rock at various
w/r ratios. Analogous equations are used for oxygen.
Initial parent Miocene meteoric water with
18O
of 3.75 per mil and
D of 20 per mil is reacted with
carbonate (50 wt % O and 0.1 wt % H having the isotopic composition
18O
of 23
and
D of 60
), and all exchangeable oxygen and
hydrogen, respectively, are assumed to be in calcite (
r-w
~1,000 ln
calcite-water = 2.78 x 106/(T°K)2
2.89
; Friedman and ONeil, 1977) and illite/muscovite/smectite (1,000 ln
mica-water = 25
; Sheppard and Gilg, 1996). Similarly, meteoric
water is reacted with volcanic rock (50 wt % O and 0.1 wt % H having the
isotopic composition
18O of 10
and
D
of 10
), where we assume that oxygen contained in volcanic glass, its
devitrification products, alkali feldspar, and Na-rich plagioclase can by
modeled by K-feldspar (1,000 ln
K-feldspar-water = 2.39 x 106/(T°K)2
2.51; Matsuhisa et al., 1979). Hydrogen fractionation in volcanic rocks is
also assumed to be governed by illite/muscovite/smectite mineral fractionation.
| Acknowledgments |
|---|
May 10, 2002; November 4, 2004
May 10, 2002; November 4, 2004
| References |
|---|
|
|
|---|
Arribas, A. Jr., Cunningham, C.G., Rytuba, J.J., Rye, R.O., Kelly, W.C., Podwysocki, M.H., McKee, E.H., and Tosdal, R.M., 1995, Geology, geochronology, fluid inclusions, and isotope geochemistry of the Rodalquilar gold-alunite deposit, Spain: Economic Geology, v. 90, p. 795822.[Abstract][ISI][GeoRef]
Bailey, S.W., 1993, Review of the structural relationships of the kaolin minerals, in Murrary, H.H., Bunday, W.M., and Harvey, C.C., eds., Kaolin genesis and utilization: Boulder, CO, Clay Minerals Society, Special Publication 1, p. 2542.
Barnes, H.L., and Seward, T.M., 1997, Geothermal systems and mercury deposits, in Barnes, H.L., ed., Geochemistry of hydrothermal ore deposits, 3rd ed.: New York, John Wiley, p. 699736.
Beccaluva, L., Brotzu, P., Mac