The Fate of Magmatic Sulfides During Intrusion or Eruption, Bingham and Tintic Districts, Utah

Transcription

1 2006 Society of Economic Geologists, Inc. Economic Geology, v. 101, pp The Fate of Magmatic Sulfides During Intrusion or Eruption, Bingham and Tintic Districts, Utah WILLIAM J. A. STAVAST,, * JEFFREY D. KEITH, ERIC H. CHRISTIANSEN, MICHAEL J. DORAIS, DAVID TINGEY Department of Geology, Brigham Young University, Provo, Utah ADRIENNE LAROCQUE, Department of Geological Sciences, University of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2 AND NOREEN EVANS CSIRO Exploration and Mining, P.O. Box 136, North Ryde NSW 1670, Australia Abstract Magmatic sulfides in 97 samples of volcanic and intrusive rocks from the Tertiary Bingham (Cu-Au-Mo) and Tintic (Ag-Pb-Zn-Cu-Au) districts, Utah, were examined to help better understand the fate of magmatic sulfides during intrusion and eruption. Our findings show that shallowly emplaced dikes and sills have erratic but locally high concentrations of sulfides. Volcanic rocks and large porphyry intrusions from these districts typically have at least two orders of magnitude fewer sulfides than the dikes. Sulfide concentrations vary dramatically across these dikes and sills; for example, in one sill in Castro Gulch, Bingham district, sulfide abundance increases from 9 ppm by volume in the center to more than 2,000 ppm near the margin. Chalcophile metals show corresponding changes in abundance. For example, the whole-rock copper content of the sill ranges from 23 ppm in the center to 35 ppm along the margins. The textures of sulfide grains (interpreted to reflect recrystallization, resorption, and degassing) even in the most sulfide-rich samples, commonly have been modified, suggesting that no sample preserves all of its original magmatic sulfide content. Immiscible liquids of monosulfide solid solution crystallized as pyrrhotite, pyrrhotite and chalcopyrite, or pyrite and chalcopyrite with declining temperature and pressure. These locally recrystallized to pyrite and chalcopyrite or to pyrite and an Fe oxide as they are oxidized. The alteration and preservation textures change from subspherical sulfide blebs near the margins of dikes and sills, to partially altered sulfides farther in, to complete absence of sulfides in the vast majority of intrusions (except where small sulfides are completely enclosed by phenocrysts). Sulfide concentrations appear to vary according to cooling rate and inferred pressure at the time of quenching or crystallization of the matrix. Most of the sulfides along the quenched margins of these dikes and sills are in the matrix. Slower cooling coupled with removal of magmatic volatiles, including sulfurous gases (e.g., H 2S, SO 2), allows the resorption or oxidation of magmatic sulfides to occur during final crystallization of a magma. Together, these processes remove greater than 90 percent of the original endowment of magmatic sulfides. This probably explains the low-magmatic sulfide abundances of slowly cooled, large porphyritic intrusions, and most importantly, allows metals and sulfur to participate in the formation of porphyry deposits. The relative abundances of base metals lost from the center of the sill are similar to the relative abundances of the metals in the Bingham deposit (production and reserves), suggesting that these processes also may have operated at a larger scale. Introduction MAGMATIC SULFIDES, when present in igneous rocks, are known to host the major portion of chalcophile metals due to the large partition coefficients for metals (Stimac and Hickmott, 1996) and this controls the availability of Cu and Au during development of porphyry copper deposits (Candela and Holland, 1986). For example, the partitioning of Cu into magmatic sulfides relative to silicate melts (D Cu is ) is orders of magnitude greater than the partitioning into silicate or oxide minerals, which have crystal and/or liquid partition coefficients of about one (Stimac and Hickmott, 1996). Consequently, the formation of magmatic sulfides has the potential to deplete the silicate melt in Cu, as well as Ni, Co, Zn, As, Se, Mo, Pb, and precious metals (Stimac and Hickmott, 1996). If sulfides form and persist in a magma chamber during episodes of hydrothermal fluid release (due to equilibrium Corresponding author: , bill@stavast.us *Present address: 2625 West Dove Ln, Thatcher, AZ conditions with the fluid phase), no significant amount of metals could be collected by the fluid and incorporated into an orebody. Thus high concentrations of magmatic sulfides should not be present in rocks associated with orebodies. In order to form a large porphyry deposit such as Bingham, Utah, Cu, Au, and S would need to be scavenged from ~250 km 3 of magma (assuming typical magmatic concentrations) and transported to the site of ore formation near the top of the magma chamber (Dilles, 2000; Hattori and Keith, 2001). Previous workers observed high concentrations of magmatic sulfides in some quenched dikes from the Tintic and Bingham mining districts, Utah (Cannan, 1992; Hook, 1995; Waite et al., 1997; Pulsipher, 2000; Maughan, 2001). Preliminary data from Keith et al. (1997) showed that slowly cooled intrusions have only 1/100 of the sulfides in dikes. Hook (1995) studied the composition of magmatic sulfides in the volcanic rocks south of the Bingham mine. He observed that the sulfides were mainly monosulfide solid solution (Mss) or pyrrhotite and intermediate solid solution (Iss) /06/3578/ $

2 330 STAVAST ET AL. or chalcopyrite. Two unmineralized intrusions contained pyrite as well. Hook (1995) noted that there were relatively few sulfide grains with degassing textures (spongy Fe oxides formed as sulfide blebs lose sulfur; Larocque et al., 2000) in the volcanic rocks, but such degassed grains occur more commonly in the intrusions near the Bingham mine. In a similar study in the Tintic district, Cannan (1992) noted that the sulfides in vitrophyres, latite lava flows, and dikes of the district are pyrrhotite, pyrite, and chalcopyrite with abundances ranging from to 0.1 vol percent. In this study, we examine the high concentrations of magmatic sulfides in the Bingham and Tintic districts and how the distribution and abundance of magmatic sulfides vary across dikes. We compare these data to comagmatic volcanic rocks and large intrusive bodies. Castro Gulch in the Bingham district and Little Valley in the Tintic district are two areas with high concentrations of magmatic sulfides in dikes and were chosen for this study (Fig. 1). We hypothesize that the high magmatic sulfide concentrations are caused by sulfide preservation (due to quenching under pressure) and that these sulfide-bearing dikes are typical of preore magmatic sulfide concentrations in ore-related porphyries in both mining districts. Geologic Setting The Tintic and Bingham districts lie on the eastern side of the Basin and Range province in central Utah (Fig. 1A). The Bingham district is 25 km south-southwest of Salt Lake City. The Bingham deposit is a Cu-Au-Mo (-Ag-Pb-Zn) porphyry FIG. 1. A. Location of Bingham and Tintic districts. B. Regional geology of the Bingham pit area (after Phillips et al., 1997; Pulsipher, 2000). The dashed line shows the outline of high-grade ore (<0.35% Cu). C. Regional geology of the Tintic district (after Keith et al., 1991). Dashed line shows hidden porphyry Cu deposit (Morris, 1975) /98/000/ $

3 FATE OF MAGMATIC SULFIDES, BINGHAM & TINTIC DISTRICTS, UTAH 331 deposit with reserves of 2,820 million metric tons (mt) of 0.73 percent copper (Ballantyne and Smith, 1997). The area was volcanically active between 39 and 33 Ma; mineralization occurred between 39.2 and 37.7 Ma (Babcock et al., 1995). In the Bingham district, abundant latite dikes occur in Castro Gulch, southeast of the Bingham pit (Fig. 1B). According to Deino and Keith (1997), these latite dikes are comagmatic with the porphyry-related magma but predate the main pulse of mineralization ( 40 Ar/ 39 Ar age of ± 0.19 Ma). In Castro Gulch, there are 10 dikes and sills that range in outcrop length from 25 to 700 m and cut quartzite of the Pennsylvanian Butterfield Peaks Formation (Stavast, 2002). Block and ash flows south of Castro Gulch have similar ages to the dikes (Maughan, 2001) and have clasts that are indistinguishable in composition from the Castro Gulch intrusions, indicating that the dikes and extrusive units are comagmatic. Both units contain rare accessory sapphire crystals, approximately 1 mm in diameter, of the same color and composition in terms of included trace elements (Pulsipher, 2000). Several lines of evidence indicate that the level of emplacement of the Castro Gulch intrusions was relatively shallow. First, extrapolation of the prevolcanic erosion surface that is exposed along ridge tops adjacent to Castro Gulch indicates that no more than 40 to 100 m of Paleozoic sedimentary rocks were present above the dikes that are most deeply incised by the present-day ravine. If approximately 100 m of volcanic rocks was also present, at the time of dike emplacement, as indicated by the paleotopographic reconstruction of Waite et al. (1997), then perhaps 200 to 300 m of cover may have existed. Other features of the rocks support this shallow level of dike emplacement. For example, the texture and size of matrix crystals differ little from extrusive units, and the margins of the dikes, where well exposed, contain glass and devitrified glass and retain the general appearance of vitrophyres. The Tintic district is 60 km south of Bingham. The district has produced Ag, Pb, Zn, Cu, and Au for more than 100 years from vein and replacement deposits related to monzonite intrusions (Morris and Lovering, 1979). According to Moore (1993), the area was volcanically active from 35 to 32 Ma. In the Tintic volcanic field, both quenched latite dikes and more slowly cooled, coarser grained monzonite dikes occur in and around Little Valley. This area is 7 km southeast of the Silver City stock, the ore-related intrusion (Morris, 1975), and 5 km east-southeast of a concealed porphyry Cu-Mo-Au prospect (Morris, 1975; Fig. 1C). The dikes in Little Valley range from 10 to 100 m in length. There are approximately 28 dikes and six larger intrusions (Stavast, 2002). Keith et al. (1991) reported a 40 Ar/ 39 Ar age of 33.0 ± 0.2 Ma for one of the larger intrusions. The dikes and intrusions cut latite lava flows, shoshonite lava flows of similar composition, and intercalated Tertiary limestone and shale beds. These dikes can be inferred to be shallowly emplaced because they cut coeval lava flows of the same composition as well as lake sediments (caldera lakes?) that are interbedded with the lava flows. Analytical Techniques Seventy-four samples were collected for this study from the major dikes in Castro Gulch (36 samples) and Little Valley (38 samples). Seven dikes were sampled near their edges and centers to examine changes across the dike. Only dikes that were not cut by veins and had no noticeable alteration of minerals were sampled. Fifteen samples of extrusive rocks and eight samples of large intrusions that were previously collected were also studied. Thirty-nine samples of dike rocks were analyzed by X-ray fluorescence analysis (XRF) to determine major and trace element compositions in each district and to determine differences in composition across specific dikes (method in Maughan, 2001). Platinum group element and gold analyses were performed on whole-rock samples (across one sill) at CSIRO Exploration and Mining by isotope dilution and external calibration ICP-MS as described by Evans et al. (1993). Phenocryst and sulfide compositions from representative dike samples were determined with a Cameca SX50 electron microprobe that has been upgraded to SX100 standards. Silicates and oxides were analyzed with a 20-nA beam and a 15- kv acceleration voltage. Sulfides were analyzed with a 20-nA beam and a 25-kV acceleration voltage. Sulfide grains larger than 5 µm in diameter were analyzed to minimize possible contamination from excitation of the surrounding silicate. Sulfides were analyzed for S, Fe, Co, Ni, Cu, Zn, As, Ag, and Pb. Although the sulfide grains are not homogeneous, most of the individual phases were small enough that a bulk composition of the grain was obtained. In larger phases, compositions vary by less than 45 percent. Sulfides were also analyzed by secondary ion mass spectrometry (SIMS) for Au and Pt (method in Larocque et al., 2000). Modal sulfide abundances were obtained by automated wavelength dispersive element mapping on the electron microprobe and by petrographic examination of grain size and number in a 5-28-mm area. For samples that have high abundances of sulfides measurements were made in an area 3 28 mm; typical numbers of sulfide grains in these samples ranged from about 600 to 1,200. Grains were classified as sulfides or partially degassed sulfides that have lost sulfur. Grains classified as partially degassed sulfides contain at least 20 percent spongy Fe oxide (described by Larocque et al, 2000) and have some residual sulfide and sulfur within the grain. Grains with less than 20 percent spongy Fe oxide were counted as sulfides. Spongy Fe oxide grains with no remnant sulfide were not counted. Percent sulfide was then determined by comparing the area of sulfides with the total area. We assumed that the sulfides are all spherical, they are evenly distributed in three dimensions, and that the thin section cuts every bleb down the center of each sulfide (e.g., a vol % of would contain 44 blebs of 1-µm size, 17 blebs of 2-µm size, and 3 blebs of 5-µm size in a 5-28-mm area). Modal analyses using the electron microprobe were done by mapping 50 random, 1-mm 2 areas per thin section using a continuous scan and a 5-µm beam. These areas were mapped for S, Fe, and Ba. The Cameca image analysis software was used to calculate the percent of the area that contains Fe and S, but not Ba and S to avoid including barite. Modal analyses of five samples were determined twice using both the petrographic counting method and electron microprobe element maps to determine reproducibility (Table 1); typical uncertainties of counting were in the range of vol percent for samples with the lowest concentrations of sulfide and vol percent for samples with the highest concentrations /98/000/ $

4 332 STAVAST ET AL. TABLE 1. Volume Percent Sulfide in Samples of the Castro Gulch Sill 1 Sample Count 1 Count 2 Count 3 Microprobe 2 Range Std. Dev. Castro Castro Castro-6a Castro Castro-15b Each count is the volume percent sulfide determined by image analysis of three different 5-28-mm areas of the same sample 2 Volume percent determined by elemental maps of S, Fe, Ba, and Si The ability to distinguish between hydrothermal and magmatic sulfides when both may be present in a sample is an important issue. Several common characteristics of magmatic sulfides are (1) their rounded shapes, (2) they commonly occur as inclusions in early formed silicates and oxides, and (3) they have higher concentrations of Co, Ni, Cu, Zn, and other trace metals than hydrothermal sulfides. Hydrothermal sulfides, on the other hand, form along fractures, are not rounded, and tend to have stoichiometric compositions. The classification of magmatic versus hydrothermal sulfides was based on these criteria. Only magmatic sulfides were counted in this study. Results Castro Gulch Two types of dikes occur in Castro Gulch, fragmental and latite (Fig. 2). The fragmental dikes are related to subvertical fractures and contain a variety of igneous lithologic units. The latite dikes occur as both dikes and sills. The sampled fragmental dike contains clastic material (Fig. 3), rounded FIG. 2. Geologic map of Castro Gulch, Bingham district, showing locations for the fragmental dike and the sill sampled in this study. pebbles (Fig. 3B, D), silt- to coarse sand-size matrix (Fig. 3C), flow banding in the matrix (Fig. 3C), and fragments that appear to have been somewhat plastic or molten at the time of emplacement (Fig. 3A-B). Four distinct clasts types are present, represented by samples Castro-22a1, 29a, 29b, and 29c. Castro-29d is a sample of the matrix. The dike is of particular interest because some of the fragments contain distinctly higher concentrations (1 vol %) of magmatic sulfides than do the adjacent clasts or matrix ( vol %). All clasts in this fragmental dike, except Castro-22a1, contain rounded quartz, sieve-textured plagioclase, altered clinopyroxene, altered biotite, Fe-Ti oxides, and a very fine grained matrix. Clasts 29a and 29b range in size from 1 to more than 20 cm across. Two of the clasts are latites and indistinguishable from each other and from other dikes in the Castro Gulch area. One is lavender (Castro-29a) and the other is light gray in color (Castro-29c). These clasts have a similar mineralogy, except that 29c has less amphibole (Table 2). Castro-29a was apparently plastic at the time of emplacement (Fig. 3A), whereas Castro-29c was brittle (Fig. 3B). Castro-29a contains vol percent sulfide and percent oxidized sulfide (Table 3). A third clast is darker gray (Castro-29b) but indistinguishable from others in its phenocryst proportions (Table 2). This clast contains vol percent sulfide and oxidized sulfide (Table 3). The fourth clast (Castro-22a1) is a crystal-rich clot (Fig. 3D), about 3 cm across, that appears to be cumulate or glomeroporphyritic in nature. It contains plagioclase, amphibole, clinopyroxene, titanite, quartz, Fe oxide, and 1 percent sulfide (Table 2). The sulfides appear to be magmatic based on the presence of pyrrhotite cores; they are interstitial to other cumulate crystals, and they show no evidence of sulfidizing the edges of mafic silicates. The matrix of the fragmental dike is tan colored and composed mainly of fine (0.1 1 mm) fragments of igneous rocks, quartzite, and limestone (Castro-29d). No microphenocrysts are present. The crystals and sand-sized fragments in this matrix are distinctly graded and flow aligned (Fig. 3C). The matrix contains no sulfides. Whole-rock analyses of two of the clasts (Castro-29a and 29b; Table 3) show few significant chemical differences between clasts. Differences in color between clast types may be partially related to differences in the degree of devitrification. However, the matrix (Castro-29d) has significantly lower concentrations of TiO 2, Al 2 O 3, Fe 2 O 3, MgO, Na 2 O, K 2 O, V, Cr, Ni, Cu, Zn, Rb, Sr, Ba, Ce, and Pb and higher amounts of SiO 2 and CaO (Table 3; Stavast, 2002). The two other clasts (sulfide-rich Castro-22a1 and Castro-29c) were not large /98/000/ $

5 FATE OF MAGMATIC SULFIDES, BINGHAM & TINTIC DISTRICTS, UTAH 333 FIG. 3. Combined brittle and ductile magma mixing shown in fragmental dike Castro-22a. Five rock types are indicated by different colors: lavender (29a), light gray (29c), dark gray (29b), tan matrix (29d), and white (22a1) A. This sample shows the heterogeneity of the dikes, with a plastically deformed fragment present. B. Light gray, dark gray, and white fragments can be seen. Lavender rock wraps around white fragment. Rounded quartzite clasts are present. C. Close-up of size-graded matrix, which coarsens upward. D. Close-up of sulfide rich (1 vol %) white fragment (Castro-22a1). enough to allow separation of the clasts from the matrix for bulk-rock analysis. Whole-rock compositions of the nonfragmental dikes and sills of Castro Gulch plot near the corner of the latite, andesite, trachyte, and dacite fields (Fig. 4). These dikes are hereafter referred to as latites. Samples were taken across one sill to determine variations across the latite dikes (Fig. 5), and these data show that S, Cu, Ni, Zn, Au, Pt, Pd, Rh, and Ir generally increase toward the margins of the sill (Table 3, Fig. 6). Typical nonfragmental dike material contains approximately 47 percent phenocrysts: 10 percent plagioclase (An ), 25 percent amphibole (magnesiohastingsite), 7 percent biotite (Phlog 54 Sid 27 Ann 19 ), 2 percent Fe-Ti oxides, <2 percent clinopyroxene (Wo 44 En 51 Fs 5 ), <1 percent quartz, and trace amounts of quartzite fragments. Some samples contain no clinopyroxene or quartz phenocrysts. Minor apatite and barite and trace amounts of olivine and a dark blue to mottled blue and white euhedral sapphire are also present. Barite sometimes occurs as inclusions in phenocrysts. Plagioclase and amphibole phenocrysts are generally 2 to 3 mm. Biotite and pyroxene phenocrysts are <1 mm and oxide phenocrysts are <0.5 mm. Resorption and magmatic-reequilibration textures are common in all samples from Castro Gulch. For example, quartz TABLE 2. Mineralogy of Units in Fragmental Dike and Latite Dikes and Sills from Castro Gulch Sample no. Castro 22a1 Castro 29a Castro 29b Castro 29c Average latite Sample type clast clast clast clast dike Amphibole (vol %) Plagioclase Clinopyroxene <2 Biotite 2 <1 2 7 Quartz 1 <1 <1 <1 ± Fe oxides 2 <1 <1 <1 2 Titanite 2 Sulfide <0.2 Olivine tr Barite tr Apatite tr Sapphire tr Fine-grained matrix /98/000/ $

6 334 STAVAST ET AL. TABLE 3. Whole-Rock XRF Analyses of Dikes Sample Castro Castro Castro Castro Castro Castro Castro Castro Castro Castro a 15b 29a 29b Easting Northing Rock Type LS LS LS LS LS LS LS LS FD FD Position Margin Center Center Center Center Center Center Margin Clast Clast SiO 2 (wt %) TiO Al 2O Fe 2O MnO MgO CaO Na 2O K 2O P 2O LOI Total F (ppm) S Cl Sc V Cr Ni Cu Zn Ga As Rb Sr Y Zr Nb Mo <1 1 2 <1 0 <1 1 2 <1 <1 Ba La Ce Nd Sm Pb Th U Pt 2 (ppb) Pd Ir Rh Au Sulfide (vol %) Oxidized sulfide Total Abbreviations: FD = fragmental dike, HAL = alkali-rich latite, HAS = alkali-rich shoshonite, LAL = alkali-poor latite, LAS = alkali-poor shoshonite, LS = latite sill, MAT = matrix 1 UTM NAD 27, zone 12 2 ICP-MS analyses 3 Oxidized sulfide is any sulfide having greater than 20 percent spongy Fe oxide phenocrysts, where present, are commonly rounded and embayed. Plagioclase crystals are often broken and have a sieve texture in most samples. In addition, small anhedral inclusions of plagioclase are commonly present within large euhedral to subhedral amphibole crystals. Hydrothermal alteration is present to varying degrees in Castro Gulch, with most outcrops showing little or no effect from alteration. Most dike margins contain darker matrix (glass or very fine grained crystals), whereas the groundmass in the center of the dike is coarser grained. Higher degrees of /98/000/ $

7 FATE OF MAGMATIC SULFIDES, BINGHAM & TINTIC DISTRICTS, UTAH 335 from the Bingham and Tintic Districts Castro Tintic Tintic Tintic Tintic Tintic Tintic Tintic Tintic Tintic Tintic 29d a 28b 28c 29a 29b 29c FD HAL HAS LAS LAL HAL HAL HAL HAL HAL HAL MAT Center Center Center Center Margin Center Margin Margin Center Margin <1 1 <1 < alteration in dike interiors are indicated by some phenocrysts. For example, in one dike, pyroxenes are completely altered in the center of the dike but are only partly altered to clays and chlorite along the edge of the dike. Amphiboles in the center of the dike are altered along fractures and cleavage planes, whereas those along the edges of the dike show no alteration. Plagioclase in the center of the dike is partially altered, creating small cavities that have been filled with carbonate and minor sericite, whereas the plagioclase along the edges of the dike is broken but relatively unaltered. Sampling for this /98/000/ $

8 336 STAVAST ET AL. FIG. 4. Total alkalis vs. silica diagram showing IUGS classification of Bingham and Tintic samples normalized to 100 percent on a volatile-free basis. study focused on outcrops where macroscopic alteration was minimal or negligible. Thin sections with no microscopic evidence of alteration were used for the study. Sulfides from Bingham district and Castro Gulch: Very few samples of volcanic and intrusive rocks in the Bingham area contain high concentrations of magmatic sulfides. Most volcanic rocks in the Bingham district contain to vol percent sulfides with an average of percent (Fig. 7). Typical large intrusions (Last Chance stock) in the district are generally comparable to volcanic rocks in terms of sulfide abundances and contain only to vol percent sulfides (Fig. 7; Borrok et al., 1999, Stavast, 2002). However, unusually high magmatic sulfide abundances occur in some Castro Gulch intrusions, ranging from to 0.21 vol percent (Fig. 7). Samples taken from the margins of the dikes have the highest concentrations of magmatic sulfides (Fig. 7). The three order of magnitude range of concentration cannot be an artifact of uncertainties in the modal analyses or hydrothermal alteration (although some recrystallization of sulfides and sulfidation of silicates proximal to degassed sulfides does occur, as explained below). A quenched sill margin (Castro-15a, located in Fig. 5) that is rich in magmatic sulfides FIG. 5. Detailed map of Castro Gulch sill, showing sample locations. This well-exposed sill has one exceptionally well preserved glassy margin and was chosen to examine changes in sulfide abundances and elemental chemistry across the sill. FIG. 6. Variation in concentrations of S, Cu, Ni, Zn, Au, Pt, Pd, Rh, and Ir across Castro Gulch sill, shown in Figure 5 (distance from top margin). Concentrations of chalcophile elements and total S are generally higher toward the margins of the sill. contains approximately 500 sulfide blebs between 1 and 5 µm, 350 blebs between 5 and 20 µm, and 60 blebs that are 20 µm or larger in a 5-28-mm area of the thin section. A typical latitic volcanic rock (e.g., Tick-41, collected 12 km southeast of Castro Gulch, at UTM NAD 27 zone 12 E403820, N ) contains 300 sulfide blebs between 1 and 5 µm, 20 blebs between 5 and 10 µm, and a few blebs larger than 10 µm (Fig. 8). Immiscible sulfides in all samples occur as (1) spherical or ovoid blebs that are completely encased in silicate or oxide phenocrysts (Fig. 9F), (2) spherical or ovoid blebs that are partly encased in silicate or oxide phenocrysts, and or (3) spherical or ovoid blebs in the matrix (Fig. 9A-E). The sulfide inclusions that are completely encased by phenocrysts consist entirely of pyrrhotite that apparently crystallized from monosulfide solid solution (i.e., the pyrrhotite is not homogeneous in composition) or they have exsolved to pyrrhotite with minor pyrite (~5%) and chalcopyrite (~10%; Fig. 9F). The blebs that are only partly encased in phenocrysts or are in the matrix are mainly pyrite with traces of chalcopyrite and spongy Fe oxides (as described by Keith et al., 1991; Cannan, 1992; Larocque et al., 2000). Larocque et al. (2000) identified 15 chemical and textural criteria for distinguishing spongy Fe oxides formed by degassing of sulfide blebs from Fe oxides formed by other processes. The spongy Fe oxides from Castro /98/000/ $

9 FATE OF MAGMATIC SULFIDES, BINGHAM & TINTIC DISTRICTS, UTAH 337 FIG. 7. Vol percent of sulfide, not including oxidized sulfides, in dikes, volcanic rocks, and stocks from the Bingham and Tintic districts. Dike Margins (shown within ovals) have a significantly higher concentration of sulfides than dike interiors, volcanic rocks, and stocks. Concentrations were determined by petrographic examination of grain size and number of blebs in a 5-28-mm area (e.g., a vol % of would contain 44 blebs 1 µm in size, 17 blebs 2 µm in size, and 3 blebs 5 µm in size in a 5-28-mm area, assuming sulfides are all spherical, they are evenly distributed in three dimensions, and that the thin section cuts every bleb down the center of each sulfide). FIG. 8. Cumulative number of sulfide blebs counted by petrographic examination vs. bleb diameter, showing relative sulfide abundances for a quenched dike margin in Castro Gulch (Castro-15A) and a latitic volcanic rock (Tick-41) from the Bingham district. Blebs were measured in an area of 5 28 mm. Both samples show a very uniform distribution of bleb size, with smaller blebs more readily preserved than larger blebs. Gulch fulfill all of these criteria, including globular forms, frothy, porous, or spongy textures, gradation from massive cores to porous, frothy rims, low concentrations of Ti, similar minor and trace element concentrations to coexisting, more pristine magmatic sulfides, and pyrite lamellae replacing the original sulfide. The degassed blebs have been oxidized to pyrite, and almost all have traces of chalcopyrite and or spongy Fe oxide rimming the pyrite. A trace of covellite is also present in a few blebs. In some samples, flow layering curves around the degassed blebs. Elemental compositions of pyrite in the matrix of one sample, Castro-15b, are similar to hydrothermal pyrite from the ore zone in the Bingham Cu deposit in that it has low or no chalcophile metals. Three out of 11 analyzed pyrites in the matrix have some traces of Cu (0.05 wt %) or Zn (0.03 wt %), suggesting that they may have been magmatic, although they recrystallized during cooling. All of the sulfides analyzed from Castro Gulch samples, except those from Castro-15b, have trace amounts of Ni, Co, Cu, Zn, As, and Ag (80% have >0.2 cumulative wt % of these elements, Table 4, Fig. 10). Castro Gulch sulfides also contain Au and Pt (Table 5). Sulfide blebs that occur as inclusions in phenocrysts have the highest concentrations of trace metals. The sulfides that occur in the matrix tend to have slightly lower concentrations of trace metals but still distinctly higher concentrations than in hydrothermal sulfides (Table 4). Average compositions of each phase were used to calculate the initial composition of the bleb before exsolution (Fig. 10, Table 4). Spatial variations in Castro Gulch sulfides: Ten samples were taken across one well-exposed, unaltered sill to document the variations in sulfide concentrations (Fig. 4). Variations in sulfide concentrations across the sill can be seen in Table 3 and Figure 11. Castro-15b is 25 cm from the top edge of the sill and contains 0.21 vol percent sulfides and 0.03 percent partially degassed sulfides, whereas the center of the sill (Castro-14) contains percent sulfides, all of which occur as inclusions within silicates or magmatic oxides. The bottom edge of the sill (Castro-1) contains vol percent sulfide and percent partially degassed sulfides. The sulfide phases that are present also vary across the sill. Samples near the margin contain pyrite + chalcopyrite + Fe oxide (Castro-1, -10, -11, -15a, -15b, and -15c). Samples closer to the center contain pyrite + Fe oxide (Castro 2), and samples at the center contain no sulfide in the matrix (Castro- 12, -13, and -14). All samples contain pyrite + chalcopyrite ± pyrrhotite ± Fe oxide as inclusions in phenocrysts. Oxidation and degassing textures in sulfides also vary across the sill. In the quenched margins of the sill (Castro- 15b), the sulfides show little or no oxidation or degassing effects other than the conversion of magmatic sulfides to pyrite (Fig. 9A), which may happen simply due to equilibration to lower temperatures in an open system, as discussed below. Grains near the margins of the sill (Castro-1), but not in the outermost sample, have preserved sulfide lamellae and Fe oxide (Fig. 9B). Samples closer to the center of the sill contain partially oxidized and degassed sulfides with Fe oxides along the edges (Fig. 9C-D). Samples near the center of the sill contain Fe oxides with a spongy texture (Fig. 9E), but, as noted above, samples in the center do not contain remnant sulfides or spongy textured Fe oxides, except /98/000/ $

10 338 STAVAST ET AL. FIG. 9. Reflected light (RL) and transmitted light (TL) photomicrographs of sulfides in Castro Gulch sill. A. Round bleb of pyrite in the matrix along the margin of the sill (Castro-15b) (combined RL and TL). B. Sulfide bleb on the other margin of the sill (Castro-1), with pyrite and Fe oxide lamellae and hematite and pyrite that has recrystallized from the bleb (RL). C. Two sulfide blebs near the margin of the sill (Castro-15b). The upper sulfide has exsolved to pyrite, pyrrhotite, and covelite. The curved outer boundaries indicate the original shapes and extent of the blebs (RL). D. Degassed sulfide bleb indicated by the spongy texture of Fe oxide surrounding the remaining pyrite (Castro-15a; RL). E. Spongy Fe oxide near the center of the sill (Castro-11; RL). F. Sulfide inclusion in amphibole containing pyrrhotite, pyrite, and chalcopyrite (Castro-15a; RL). Abbreviations: Amp = amphibole, Cov = covellite, Cp = Chalcopyrite, Hem = hematite, Mag = magnetite, Py = pyrite. TABLE 4. Representative Microprobe Analyses of Sulfides from Castro Gulch Inclusions 1 Partly encased inclusions 2 Matrix 3 Castro Castro Castro Castro Castro Castro Castro Castro Castro Castro Castro Castro Sample 1 29B B B 1 1 Mineral Po Po Py Cp Py Py Py Cp Py Py Cp Cp S (wt %) Fe Co 0.07 < < <.02 <.02 <.02 <.02 <.02 <.02 Ni < < <.02 <.02 <.02 Cu < < < Zn <.03 < <.03 <.03 <.03 <.03 < <.03 As < <.10 < < <.10 < Ag Pb Total Average Average Hydrothermal sulfides inclusion matrix from Bingham Pit Sample composition composition D D Mineral Po Py Cp bleb 4 Py Cp Py Py Cp Cp S (wt %) Fe Co < <.02 <.02 <.02 <.02 Ni < <.02 <.02 <.02 Cu < < <.04 < Zn <.03 <.03 <.03 <.03 < <.03 <.03 <.03 <.03 As < <.10 < <.10 <.10 <.10 <.10 <.10 Ag <.04 < <.04 <.04 < <.04 <.04 Pb Total Abbreviations: Cp = Chalcopyrite, Po = Pyrrhotite, Py = Pyrite, - = not analyzed 1 Ovoid blebs completely encased by phenocryst 2 Ovoid blebs partly encased by phenocryst 3 Ovoid blebs in matrix 4 Average bleb is 85 percent Po, 10 percent Cp, and 5 percent Py /98/000/ $

11 FATE OF MAGMATIC SULFIDES, BINGHAM & TINTIC DISTRICTS, UTAH 339 FIG. 10. Compositions of end-member sulfides (open squares), sulfides within blebs (diamonds), the average of sulfide blebs as inclusions in phenocrysts (open circle), and the average of sulfide blebs in the matrix (closed circle) from Castro Gulch, plotted in a Cu-Fe-S diagram. Sulfide phases within the blebs are pyrite, chalcopyrite, and sulfur-rich pyrrhotite. where the original blebs were protected as inclusions in silicate or oxide minerals (Fig. 9F). Partial resorption is readily apparent among crystalline sulfides. The observed embayments and irregular rims around sulfides are similar to those on resorbed quartz and other minerals (Fig 9E). Sulfides in the matrix near the edge of the sill tend to have more spherical shapes than those closer to the center, which have been resorbed. Little Valley Four types of dikes occur in Little Valley (Fig. 12): an alkali-poor and an alkali-rich shoshonite and an alkali-poor and an alkali-rich latite (Fig. 4). The alkali-poor shoshonite dikes (with K 2 O + Na 2 O <6.4 wt %) contain approximately 12 percent plagioclase (An ), 12 percent clinopyroxene (Wo 46 En 44 Fs 10 ), 2 percent biotite (Phlog 58 Sid 18 Ann 24 ), 2 percent oxide, a trace of orthopyroxene (Wo 2 En 66 Fs 32 ), and 72 percent matrix (Table 6). They are poorer in P, Ba, K, Th, Zr, Cu, Pb, and Sr and richer in Cr, Ni, and Nb than the alkali-rich shoshonite dikes (Table 3). These dikes do not contain glass, and the matrix is composed of larger microlites than in other dike samples from the area, indicating slower cooling. The dikes contain small vesicles (1 5 mm) that are lined with limonite. The dikes contain only traces of magmatic sulfide. FIG. 11. Variations in the abundance of sulfide and total sulfide (including degassed sulfide with >20% spongy Fe oxides) across Castro Gulch sill from Figure 5 (distance from upper margin). Sample numbers are included for reference. Concentrations increase toward the margins of the sill. The alkali-rich shoshonite dikes (with K 2 O + Na 2 O >7.1 wt %) have a texture and composition similar to the shoshonite lava flows on the west side of Little Valley. They contain approximately 15 percent plagioclase (An ), 7 percent clinopyroxene (Wo 45 En 44 Fs 11 ), 7 percent biotite (Phlog 56 Sid 21 Ann 23 ), 2 percent oxide, trace orthopyroxene (Wo 2 En 70 Fs 28 ), and 69 percent matrix (Table 6). In contrast to the alkali-poor shoshonite dikes, these dikes have a glassy to very fine grained matrix and higher concentrations of magmatic sulfides ( vol %). Crumbly perlitic dike margins prevented collection of the outermost margins and complete definition of sulfide abundances across these dikes. The alkali-poor latites (with K 2 O + Na 2 O <7.5 wt %) contain approximately 20 percent plagioclase (An ), 8 percent biotite (Phlog 58 Sid 17 Ann 24 ), 5 percent clinopyroxene (Wo 45 En 43 Fs 12 ), 2 percent oxide, a trace of orthopyroxene (Wo 2 En 66 Fs 31 ), and 65 percent matrix (Stavast, 2002). These dikes are poorer in Cu, Rb, Sr, Zr, Ce, Nd, Pb, and Th, and richer in Ni, Cr, and Nb than the alkali-rich latite dikes (Table 3). They have a coarser grained matrix (with crystals about 0.25 mm and no apparent glass) and low sulfide abundances. One sample (Tintic-1), however, has high concentrations of TABLE 5. Au and Pt Analyses for Castro Gulch Sulfides Sample no. Castro-1 Castro-11 Castro-15a Castro-15b Location B A B D C A A B C Mineral Py Py Cp Po Py Py Py Py Py Au (ppb) Total Pt (ppb) Sample no. Castro-16 Castro-22a1 Location A C C D A B C C D E E E Mineral Py Py Py Py Py Py Py Py Py Py Py Py Au (ppb) Total Pt (ppb) Notes: Abbreviations: Cp = Chalcopyrite, Po = Pyrrhotite, Py = Pyrite; SIMS analyses on sulfide grains bigger than 50 µm /98/000/ $

12 340 STAVAST ET AL. fine grained. These dikes are similar in chemical composition and texture to the latite lava flows that cover most of Little Valley (Moore, 1993). They tend to weather more rapidly than the dikes that have a coarser-grained matrix, which may be due to perlitization and fracturing of glass. These dikes contain the highest abundances (up to 0.11 vol %) of magmatic sulfides. Two dikes (Tintic-28 and -29, Fig. 12) were sampled to determine variations in chemistry across the dikes, and these data show higher concentrations of S, Cu, Ni, and Zn along the margins of the dikes (Table 3, Fig. 13). However, the concentration of sulfur is higher in the center of dike Tintic-29 (Table 3, Fig. 13). FIG. 12. Geologic map of Little Valley, Tintic district, showing the distribution of dikes and dikes sampled in this study. sulfides (0.22 vol %) as grains of pyrite that occur along fractures and do not have spherical shapes. The alkali-rich latite dikes (with K 2 O + Na 2 O >8.3 wt %) contain 10 percent plagioclase (An ), 10 percent biotite (Phlog 59 Sid 20 Ann 21 ), 7 percent clinopyroxene (Wo 47 En 43 Fs 10 ), 2 percent Fe-Ti oxide, and 71 percent matrix (Table 6). No orthopyroxene was present. The matrix is glassy or very TABLE 6. Mineralogy of Dikes from Little Valley Rock type Alkali-poor Alkali-rich Alkali-poor Alkali-rich shoshonite shoshonite latite latite Plagioclase (vol %) Clinopyroxene Orthopyroxene tr tr tr Biotite Fe oxides Fine-grained matrix FIG. 13. Sulfide and total sulfide (including degassed sulfide with >20% spongy Fe oxide), S, Cu, Ni, and Zn concentrations across dikes Tintic-28 and Tintic-29 (distance from west margin). Sulfides and elements generally increase toward the margin of the dikes /98/000/ $

13 FATE OF MAGMATIC SULFIDES, BINGHAM & TINTIC DISTRICTS, UTAH 341 Sulfides from Tintic district and Little Valley: The dikes from Little Valley contain to 0.11 vol percent magmatic sulfides, whereas volcanic rocks and a large ore-related intrusion (Silver City stock) contain to and to vol percent, respectively (Fig. 7). The order of magnitude difference cannot be an artifact of uncertainties in the modal analyses. A dike margin (Tintic-28c, Fig. 14) contains approximately 300 sulfide blebs between 1 and 5 µm, 220 blebs between 5 and 20 µm, and 25 blebs that are 20 µm or larger; a typical large intrusion (Silver City stock, located at UTM NAD27 zone 12 E412824, N in Fig. 1) contains 60 sulfide blebs between 1 and 5 µm and 5 blebs between 5 and 10 µm (Fig. 14). Immiscible sulfides in Little Valley dikes are present in the same three textural sites noted for the Castro Gulch dikes and include variable proportions of pyrrhotite, pyrite, and chalcopyrite (Fig. 15). The sulfide inclusions completely enclosed by phenocrysts are pyrrhotite (or monosulfide solid solution) or pyrrhotite with pyrite and chalcopyrite. Blebs that are only partly encased in phenocrysts are composed of pyrite with chalcopyrite, with only some containing pyrrhotite. The sulfide blebs in the matrix are either pyrite, pyrite with rimming chalcopyrite, pyrrhotite with rimming pyrite, or rarely pyrrhotite (all of the blebs have spongy Fe oxide rims). Chalcopyrite, when present normally makes up 10 to 20 percent of the sulfide bleb in all bleb types. Traces of covellite are also present in some of the blebs. The fracture-hosted hydrothermal sulfides (in Tintic-1) are distinct from magmatic sulfides in that they contain very low concentrations of trace metals (Table 7). All of the magmatic sulfides analyzed from Little Valley had trace amounts (>0.3 cumulative wt %) of Ni, Co, Cu, Zn, As, and Ag (Table 7). Pyrrhotite is sulfur rich (Fig. 15). Sulfides within blebs in the matrix have slightly lower concentrations of these trace metals than sulfides that are sealed in FIG. 15. Compositions of end-member sulfides (open squares), sulfides within blebs (diamonds), the average of sulfide blebs as inclusions in phenocrysts (open circle), and the average of sulfide blebs in the matrix (closed circle) from Little Valley (Tintic district) plotted in a Cu-Fe-S diagram. Sulfide phases within the blebs occur as pyrite, chalcopyrite, and sulfur-rich pyrrhotite. These data confirm the empirical observation that blebs hosted as inclusions in phenocrysts are generally richer in pyrrhotite due to less extensive oxidation (see Figs. 16 and 17). phenocrysts (Table 7). The distribution of trace elements in individual sulfide phases is not homogeneous. Average compositions of each phase were used to calculate the initial composition of the bleb before exsolution (Fig. 15, Table 7). Spatial variations in Little Valley sulfides: Two of the glassy alkali-rich latite dikes (Tintic-28 and -29, Fig. 12) were sampled at three points across their widths, at each margin and in the center. They show an increase in sulfide abundance and metal content toward the dike margins (Fig. 13). In Tintic vol percent sulfide and 0.05 percent oxidized sulfide occurs along one margin and 0.01 percent sulfide and 0.07 percent oxidized sulfide occurs at the other, whereas percent sulfide and percent oxidized sulfide occurs at the center (Fig. 13). Sulfide concentrations along the two margins of Tintic-29 are 0.02 vol percent sulfide and 0.07 percent oxidized sulfide and 0.01 percent sulfide and 0.02 percent oxidized sulfide, respectively, whereas percent sulfide and percent oxidized sulfide occurs at the center (Fig. 13). Depletions in sulfide abundances in dike cores are not as great as at Castro Gulch. This may reflect that the dikes are much narrower (1.5 and 2 m vs. 34 m) in Little Valley and/or the inability to collect samples of the extreme margins due to poor preservation of perlitic glass. Discussion FIG. 14. Cumulative number of sulfide blebs counted by petrographic examination vs. bleb diameter, showing relative sulfide abundances and sizes for a quenched dike margin in Little Valley, Tintic district (Tintic-28C), and in the Silver City monzonite (SCID). Blebs were measured in an area of 5 28 mm. A total of 586 blebs were present in Tintic-28c and only 64 blebs in SCID. High concentrations of sulfides in dikes The most significant observation of this study is the order of magnitude difference in sulfide abundance between quenched dikes and more slowly cooled equivalent igneous rocks. This result was not unanticipated, based on the similar findings of Moore and Schilling (1973) in a survey of sulfide abundances in midocean ridge basalts. They showed that magmatic sulfides are well preserved below a water depth of 200 m and that a 2- to 3-cm-thick quenched rind on the pillows had the highest concentrations of sulfur and sulfides. In their interpretation, quenching under pressure inhibits volatile loss and magmatic sulfides are preserved. An implication /98/000/ $

14 342 STAVAST ET AL. TABLE 7. Representative Microprobe Analyses of Sulfides from Little Valley Partly encased Inclusions 1 inclusions 2 Matrix 3 Sample Tintic Tintic Tintic Tintic Tintic Tintic Tintic Tintic Tintic Tintic Tintic Tintic Mineral Po Po Py Py Cp Py Py Cp Po Py Py Cp S (wt %) Fe Co 0.04 < Ni Cu < Zn <0.03 < < <0.03 < <0.03 <0.03 < As < < Ag <0.04 <0.04 < < <0.04 <0.04 < Pb Total Average Average Hydrothermal sulfides inclusion matrix from Tintic district Sample composition composition Tintic Tintic 1 1 Mineral Po Py Cp bleb 4 Po Py Cp 4 bleb Py Py S (wt %) Fe Co <0.02 <0.02 Ni <0.02 <0.02 Cu <0.04 <0.04 Zn <0.03 < <0.03 <0.03 < <0.03 <0.03 <0.03 As < < Ag <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 Pb Total Abbreviations: Cp = Chalcopyrite, Po = Pyrrhotite, Py = Pyrite, - = not analyzed 1 Ovoid blebs completely encased by phenocryst 2 Ovoid blebs partly encased by phenocryst 3 Ovoid blebs in matrix 4 A verage bleb is 85 percent Po, 10 percent Cp, and 5 percent Py drawn from both studies is that most unquenched magmatic rocks rarely preserve a significant fraction of the magmatic sulfides that were present prior to eruption or intrusion. At Castro Gulch and Little Valley, the outermost margins of the dikes (with a glassy to very fine grained groundmass) have the highest concentrations of sulfides, analogous to the glassy rims of pillow basalts. Dikes that have a coarser groundmass have lower concentrations of sulfides, similar to large intrusions. Sulfide accumulation may occur in some instances and may have occurred in the cumulate clast in the fragmental dike in Castro Gulch that contains 1 vol percent sulfide. Thus, the removal of magmatic sulfides from normal dike cores can be related to the slow cooling and loss of volatiles that also occurs in large intrusions. All three of the dikes profiled in this study have lower concentrations of sulfides in their centers. Furthermore, the larger sill from Castro Gulch has lower concentrations (0.001 vol %) of sulfide in its center than the narrower dikes (0.01 vol %). A slower rate of cooling may allow time for the removal of sulfides and the escape of sulfur-rich fluids containing chalcophile elements. Based on the comparison of magmatic sulfides in dike margins to large intrusions, dike interiors, and volcanic rocks, 90 to 99 percent of magmatic sulfides may be destroyed and removed from intrusions and volcanic rocks in this manner. Sulfide crystallization The types, abundances, and compositions of magmatic sulfides that eventually crystallize from high-temperature Fe-S- O immiscible liquids depend largely on the sulfur and oxygen fugacities of the system. Whitney (1984) has shown that in order to achieve sulfide saturation in highly oxidized magmas, high sulfur fugacities are required. Reduced magma (1 log unit < NNO) crystallizes pyrrhotite and magnetite from the monosulfide solid solution. Oxidized magma (3 log units > NNO) crystallizes pyrite and an Fe oxide from monosulfide solid solution. Pyrite would crystallize at about 600 to 700 C. The Bingham and Tintic dikes are examples of oxidized systems. The oxygen fugacities of the Bingham and Tintic intrusions were 2 to 4 log units > NNO (Kim, 1992; Tomlinson, 1997). The fact that the magmas were sulfide saturated indicates that sulfur fugacities were also high. This is consistent with the presence of 0.2 vol percent sulfide and 1,000 ppm whole-rock sulfur (Castro-15b). A consequence of high sulfur fugacities is the formation of S-rich monosulfide solid /98/000/ $

15 FATE OF MAGMATIC SULFIDES, BINGHAM & TINTIC DISTRICTS, UTAH 343 solution (Whitney, 1984) that reequilibrates to S-rich pyrrhotite, intermediate solid solution), pyrite, and Fe oxide, all of which are found in the Bingham and Tintic systems (Figs. 10, 15). Work by Davis et al. (1991) and Nilsson and Peach (1993) indicated that intermediate magmas normally have low sulfur concentrations (<500 ppm). However, the latite magmas at Bingham have high sulfur concentrations (1,000 ppm) as a result of magma mixing or volatile transfer from underplated mafic magmas (Keith et al., 1997; Waite et al., 1997; Hattori and Keith, 2001; Maughan, 2001). Competing processes during dike emplacement During dike emplacement, four processes may simultaneously affect the immiscible sulfide blebs: crystallization of sulfide melt, resorption (accompanying degassing of magma), exsolution during cooling, and oxidation and/or degassing. For those sulfide blebs that were not resorbed before the silicate melt quenched, a common sequence is noted that involves crystallization, exsolution, and reequilibration (Fig. 16). Our observations suggest that immiscible monosulfide solid solution liquid (Fig. 16A) crystallized initially to form pyrrhotite or pyrrhotite + intermediate solid solution + pyrite (Fig. 16B). These phases then exsolved to pyrrhotite + chalcopyrite + pyrite (seen in inclusions and in some matrix sulfide from Little Valley) as they cool (Fig. 16C). This sequence is predicted in the phase diagram of Barton and Skinner (1979; Fig. 17). Oxidation of sulfides occurs in several steps but may occur before or after exsolution of discrete sulfide phases (Fig. 16). Iron in pyrrhotite is oxidized to Fe oxide as S and some chalcophile metals escape. Some S may combine with the remaining pyrrhotite to form pyrite (Fig. 16D-E). When oxidation effects are more severe, pyrite may then be oxidized to Fe oxide, releasing additional sulfur and chalcophile metals to the magmatic fluid (Fig. 16E-F). Chalcopyrite is often the last major sulfide component to be removed by oxidation (Fig 16F-G). Sulfides in the matrix or those hosted by fractured phenocrysts are the most vulnerable to oxidation. Resorption of crystallized sulfides is evidenced by the common embayments and irregular rims in the sulfides (Figs. 9H, 16E, H). As noted, sulfides near the edge of the dikes tend to have more spherical shapes than those nearer the dike interior. The dramatic decrease in the number of sulfides and lack of spongy textured Fe oxides toward the center of the dikes suggests that most resorption of sulfides occurred while silicate melt was present and sulfides were immiscible melts (Fig. 16I). Resorption appears to be the major process for removal of sulfide melt. Destabilization of sulfides could have been caused by decompression and removal of sulfurous gases from the melt during dike emplacement and open-system behavior. Whole-rock composition and metal ratios The majority of chalcophile elements are hosted by sulfides (Candela and Holland, 1986; Halter et al., 2002). If sulfides are resorbed, oxidized, and removed by an aqueous fluid in response to open-system behavior, then whole-rock S, Cu, Zn, Ni, and PGE should dramatically decrease as well. This is seen in the samples from Castro Gulch and Little Valley. In FIG. 16. Schematic drawing illustrating one possible sequence of crystalistallization (A-C) and destruction of sulfides (D-G) by resorption and oxidation. The immiscible monosulfide solid solution (Mss) segregates from the silicate melt (A). It crystallizes as pyrrhotite (po), pyrite (py), and intermediate solid solution (Iss) (B). Iss crystallizes to chalcopyrite (cp) and pyrrhotite (po) (C). Some resorption of the bleb back into the melt also occurs at this stage. Pyrrhotite is subsequently oxidized to pyrite and Fe oxide (D). Resorption continues and pyrite is oxidized to Fe oxide (E). Pyrite and chalcopyrite are oxidized to Fe oxide (F), leaving spongy Fe oxide remnants (G). Fe oxide also is resorbed (H). If the sulfide bleb is completely resorbed, no trace of sulfide or Fe oxide is present (I). Complete resorption of sulfide into melt could occur at any point along the sequence and is not necessarily the last stage. FIG. 17. Plot of logƒ S2 vs. temperature for the system Cu-Fe-S-O (after Barton and Skinner, 1979), showing the approximate path of crystallization of the sulfides in Castro Gulch and Little Valley (arrow) /98/000/ $