FIRE and MUD Contents
Sulfur Isotopic Systematics of the June 1991 Mount Pinatubo Eruptions: A SHRIMP Ion Microprobe Study
By Michael A. McKibben,1 C. Stewart Eldridge,2 and Agnes G. Reyes3
1Department of Earth Sciences, University of California, Riverside, CA 92521.
2Geology Department and Research School of Earth Sciences, Australian National University, Canberra, A.C.T. 0200, Australia.
3Geothermal Division, Philippine National Oil Corporation, Fort Bonifacio, Metro Manila, Philippines; now at Institute of Geological and Nuclear Sciences, POB 30-368, Gracefield Road, Lower Hutt, New Zealand.
ABSTRACT
The SHRIMP ion microprobe has been used to make in situ analyses of individual anhydrite and sulfide crystals in the June 1991 eruption products of Mount Pinatubo. In air-fall pumice from the June 12 eruption, anhydrite crystals exhibit a broad, bimodal distribution of 
34S values (46 analyses on 23 crystals; range 3 to 16 per mil; modes of 6.5 per mil and 10.5 per mil). Chalcopyrite crystals have 34S values with a mode of -1 per mil (5 analyses of 5 crystals, range -2 to 0 per mil). In crystal-rich ash-flow pumice from the main June 15 eruption, anhydrite crystals exhibit a narrow, unimodal distribution of 34S values (44 analyses on 22 crystals, range 5 to 11 per mil, mode and average of 7 per mil). A single analysis of a chalcopyrite crystal yielded 34S=0 per mil.
Under the preeruptive temperature and oxygen fugacity conditions of both stages of the eruption, the chalcopyrites are in isotopic equilibrium with the 
7 per mil mode of anhydrites, and both phases are likely primary. The isotopically heavier anhydrites in the June 12 pumice may be xenocrysts acquired at a shallow level during this early vent-clearing eruption; the isotopically heaviest anhydrites have 34S values similar to those of hydrothermal vein anhydrites in a drillcore sample recovered from the geothermal system on the flank of Mount Pinatubo (9 analyses of 5 crystals, range 17-22 per mil, average 19 per mil).
The 34S value of primary anhydrite places constraints on the sulfur isotopic composition of SO2 gas resulting from the eruptions. If magma prior to the eruptions coexisted with an exsolved vapor phase containing SO2, then isotopic systematics require that the SO2 had a 34S value of 3.5 per mil. If most of this SO2 was erupted and quantitatively oxidized to H2SO4 in the stratosphere, then Mount Pinatubo aerosols could have a similar sulfur isotopic composition. Given estimates that a significant amount of the total sulfur in the eruption was present as a gas phase prior to eruption, the bulk sulfur isotopic composition of the eruption (crystals + melt + gas) would have been 5.0 to 6.5 per mil.
If, instead, most of the SO2 was generated by rapid irreversible breakdown of anhydrite during ascent decompression and eruption, then the sulfur isotopic composition of the total eruption and the erupted SO2 would have been similar to that of the primary anhydrite, 7 per mil. Quantitative conversion of this SO2 to H2SO4 in the stratosphere could have yielded sulfate aerosols with a similar isotopic composition.
Two analyses of a primary chalcopyrite grain in a quenched basalt inclusion from the June 7-14 hybrid andesite dome yielded 34S values of 1 per mil and likely reflect the sulfur isotopic composition of the mafic magma underplating the dacite. Long-term degassing of basalt magma may have been an important ultimate source of reduced sulfur to the dacite. However, Rayleigh effects upon degassing of a mixed SO2/H2S vapor from basalt magma would partition bulk sulfur into the dacite with a total 34S value no greater than 3 per mil, significantly less than the 5 to 7 per mil bulk sulfur isotopic composition of the eruption. The further 34S enrichment of the dacite is best explained by passive (noneruptive) steady-state degassing of H2S and SO2 from the dacite magma, in dynamic isotopic equilibrium with primary anhydrite, over long time periods.
Note to readers: Figures and tables open in separate windows. To return to the text, close the figure or table window or bring the text window to the front.
INTRODUCTION
Anhydrite is a characteristic and important mineralogical component of fresh tephra from Mount Pinatubo and other arc volcanoes that erupt oxidized, hydrous sulfur-rich magma (Luhr and others, 1984; Rye and others, 1984; Fournelle, 1990; Bernard and others, 1991). Anhydrite is not well preserved in older pyroclastic rocks because it weathers out of tephra soon after eruption. Experimental studies have demonstrated the relatively high solubility of anhydrite and the stability of dissolved SO42- species in oxidized melts (Carroll and Rutherford, 1987, 1988; Luhr, 1990). It is therefore commonly assumed that anhydrite in recent pyroclastic materials is a juvenile magmatic phase.
Bernard and others (1991) argued that anhydrites in pumices from the June 15, 1991 eruption of Mount Pinatubo were primary phenocrysts, mainly on the basis of their intergrowth textures with apatite. However, anhydrite and apatite are also common phases in hydrothermal veins within the geothermal systems that occur on the flanks of most Philippine volcanoes (Reyes, 1990). Such vein anhydrite ultimately owes its origin to degassing of magma batches: the magmatic SO2 (and H2S) becomes oxidized and deposited as hydrothermal anhydrite in the geothermal systems that commonly develop in the shallow portions of volcanic piles. An explored geothermal field containing abundant hydrothermal anhydrite existed on the northwest flank of Mount Pinatubo (Delfin and others, this volume; Cabel, 1990).
Beyond the question of a juvenile versus xenocrystic origin of anhydrite, there remains debate over its role in the mechanism of SO2 release from arc magmas. Three mechanisms have been proposed: (1) preeruptive exsolution of an SO2-bearing magmatic vapor phase from the dacitic melt (Luhr, 1990; Westrich and Gerlach, 1992; Gerlach and others, this volume); (2) rapid breakdown of anhydrite to SO2 in the magma during ascent decompression and eruption (Rutherford and Devine, 1991, this volume; Baker and Rutherford, 1992); and (3) decomposition of anhydrite to SO2 in the eruption plume (Devine and others, 1984, Sigurdsson, 1990).
For Mount Pinatubo, Westrich and Gerlach (1992), Matthews and others (1992), and Gerlach and others (this volume) have argued against mechanism 2 based on a reported lack of textural evidence for anhydrite breakdown and a lack of corresponding oxidation rims on iron-bearing phases. They also argue against mechanism 3 on the basis of the high temperatures required for decomposition (>1,300°C) and thermodynamic and kinetic barriers to anhydrite reduction in a cold, oxygenated plume.
Westrich and Gerlach (1992) and Gerlach and others (this volume) argue in favor of mechanism 1 for the June 1991 Mount Pinatubo eruption on the basis of mass balance considerations. Differences between the sulfur content of glass inclusions trapped in phenocrysts (preeruptive melt) and that of interstitial glass in pumice (degassed melt) are negligible and cannot account for the 20 million tonnes of "excess" SO2 that were estimated from satellite spectral data to have been erupted into the stratosphere (Bluth and others, 1991). (Luhr and others (1984) and Luhr (1990) encountered a similar "excess" sulfur problem in applying the sulfur-content difference method to glasses from the 1982 El Chichón eruption). Furthermore, they note that the amount of SO2 erupted from Mount Pinatubo is an order of magnitude greater than that which could have been dissolved in the melt at depth. Westrich and Gerlach (1992) and Gerlach and others (this volume) argue that this "excess" sulfur problem requires that magmatic water, CO2, and SO2 must have exsolved at depth to form a free gas phase within the magma, prior to eruption. They cited the presence of large vapor bubbles in glass inclusions in phenocrysts and the nearly sulfur-saturated composition of the inclusion glass as evidence for the presence of this early exsolved gas phase.
Whatever the mechanism for the shallow generation of the erupted SO2, it has been argued that the ultimate source of sulfur in the 1991 Pinatubo eruptions involved long-term degassing of sulfur from basaltic magma underplating the dacitic magma (Pallister and others, 1991, 1992, this volume; Matthews and others, 1992).
Knowledge of the sulfur isotopic compositions of sulfur-bearing minerals in the 1991 Mount Pinatubo eruption products may contribute to resolving the origin and release of sulfur in dacite from Mount Pinatubo, because different sources of melt sulfur and (or) different mechanisms of generating SO2 may result in distinctive isotopic signatures being preserved in these minerals. For this reason, we conducted a SHRIMP ion microprobe study of the in situ sulfur isotopic compositions of anhydrite and sulfide crystals in the June 1991 eruption products of Mount Pinatubo. We also analyzed sulfur-bearing hydrothermal vein minerals in drillcores from the Pinatubo geothermal field (Delfin and others, this volume; Cabel, 1990).
SHRIMP SULFUR ISOTOPIC MICROANALYSIS
Microbeam techniques of stable isotopic analysis offer unique insights into geochemical processes, because individual crystals and crystal growth zones often exhibit more isotopic variability than is found with conventional bulk isotopic sampling and analytical techniques. Sulfur isotopic microanalyses are routine on the SHRIMP (Sensitive High Resolution Ion Microprobe) facility at the Australian National University. Standards exist, fractionation is well understood, and correction procedures have been developed for the common metallic sulfides and sulfates (Eldridge and others 1987, 1989; McKibben and Eldridge, in press). Analytical work has been done on fine-grained natural sulfides and sulfates from numerous ore deposits (Eldridge and others, 1988, 1989, 1993; McKibben and Eldridge, 1995; McKibben and others, 1993), from sedimentary and volcanic rocks in the Salton Sea and Valles Caldera geothermal systems (McKibben and Eldridge, 1989, 1990), and from mantle rocks (Eldridge and others, 1991; Rudnick and others, 1993). Although the analytical precision for 34S on the SHRIMP is typically 2
at the 2
level (compared to 2 precisions of 0.2 for conventional bulk sampling and analysis), the SHRIMP studies have shown that intra- and intercrystalline 34S variations in minerals can be more than an order of magnitude greater than this over distances of less than 200 
m. Therefore, the advantages of the spatial resolution of the SHRIMP in resolving such variations far outweigh its limits in precision.
The SHRIMP ion beam excavates a sharp elliptical crater on the target mineral's surface that is typically 20-30 m wide and 5 m deep. The ability of the SHRIMP to achieve such a small sampling volume while maintaining the textural integrity of the sample makes it an ideal facility for analyzing individual crystals and crystal growth zones in fine-grained materials such as tephra.
Sulfur isotopic analysis on the SHRIMP requires samples that are polished to levels appropriate for reflected-light microscopy, usually with final polishing grits in the 0.05-m size range. Anhydrite-bearing pyroclastic samples should be prepared by using a minimum of water because of the high solubility of anhydrite under ambient conditions. However, there is negligible sulfur isotopic fractionation between anhydrite and aqueous SO42- at room temperature (Ohmoto and Rye, 1979), so any partial dissolution of anhydrite during sample preparation does not measurably modify the sulfur isotopic composition of the remaining material.
ANALYZED SAMPLES
We conducted SHRIMP 34S analyses on sulfide and sulfate minerals in four different Mount Pinatubo samples, as described briefly below and in table 1. More detailed petrographic and geochemical data from similar samples are found elsewhere in this volume (Pallister and others; Gerlach and others; Delfin and others). SHRIMP data for these samples are listed in table 1, and histograms of the 34S values are plotted for each of the four samples in figure 1.
Figure 1A. SHRIMP 34S data (table 1) for chalcopyrite and anhydrite from pumice of the June 12, 1991 eruption of Mount Pinatubo (sample collected on Mount Bagang). Chalcopyrite data shown as solid bars; anhydrite data as cross-hatched bars. The chalcopyrite data have a mode of -1; anhydrite data are bimodal at 6.5 and 10.5. B, SHRIMP 34S data (table 1) for chalcopyrite and anhydrite from pumice of the June 15, 1991 eruption of Mount Pinatubo (sample collected in Sacobia River Valley). Chalcopyrite data shown as solid bars; anhydrite data as cross-hatched bars. The anhydrite data are unimodal at 7.
Figure 1C. SHRIMP 34S data (table 1) for chalcopyrite in basalt inclusion from June 7-14 hybrid dome. D, SHRIMP 34S data (table 1) for hydrothermal vein pyrite and anhydrite from drillcore sample at 1,437-m depth in well PIN-2d. Pyrite data shown as solid bars; anhydrite data as cross-hatched bars. The anhydrite data have an average of 19.
Table 1. SHRIMP 34S analytical data for 1991 Mount Pinatubo samples (34S of samples are in per mil).
Page 1
Page 2
Page 3
PUMICE FROM THE JUNE 12, 1991, ERUPTION
The June 12 phase of the 1991 Mount Pinatubo eruptions was the major vent-clearing eruption that preceded the climactic June 15 main eruption (Wolfe and Hoblitt, this volume). We obtained 46 SHRIMP 34S analyses on 23 crystals of anhydrite and 5 SHRIMP 34S analyses on 5 crystals of chalcopyrite in the pumice from this eruption. Modally, the pumice contains 64.5% colorless glass, 20% plagioclase, 10% green hornblende, 1.3% K-feldspar, 1% biotite, 1% magnetite, 1% anhydrite, 0.7% augite, 0.4% quartz, and 0.1% olivine. The rare olivine xenocrysts are rimmed by hornblende laths. Pallister and others (1991, 1992, and this volume) have described the products of this stage of the eruption as hybrid mixed dacite-basalt, on the basis of incompatible phenocryst assemblages and disequilibrium textures.
The anhydrite crystals occur as isolated anhedral to subhedral grains from 100 to 500 m in diameter, with crystal edges often in direct contact with glass (fig. 2A). Some anhydrite crystals contain inclusions of euhedral apatite. Analyzed chalcopyrite crystals occur as anhedral grains from 50 to 80 m in diameter, either isolated in glass or associated with pyrrhotite within hornblende and other mafic phases (fig. 2B). Pyrrhotite, though generally more common than chalcopyrite, occurred as crystals that were too small (<30 m) to be analyzed.
Figure 2A. Representative photomicrographs (polished thin sections in reflected light) of SHRIMP analytical craters in anhydrite, in pumice of the June 12, 1991 eruption of Mount Pinatubo (sample collected from Mount Bagang). Numbers in boxes are 34S values (in relative to Canyon Diablo troilite (CDT) meteorite standard) for the craters identified by the arrows; craters are 20-30 m in diameter and about 5 m deep. Anhydrite crystals are anhedral to subhedral, with some crystal boundaries in direct contact with glass. Cleavage is often visible in the anhydrite crystals, as are hexagonal cross sections of apatite crystals. Dark areas are voids; dusty, less-reflective areas are voids filled with epoxy that was used to impregnate the samples prior to polishing.
Figure 2B. Representative photomicrographs (polished thin sections in reflected light) of SHRIMP analytical craters in chalcopyrite, in pumice of the June 12, 1991 eruption of Mount Pinatubo (sample collected from Mount Bagang). Numbers in boxes are 34S values (in relative to CDT standard) for the craters identified by the arrows; craters are 20-30 m in diameter and about 5 m deep. Chalcopyrite crystals are anhedral and occur as isolated crystals in glass or in mafic phases associated with subhedral embayed magnetite and sometimes pyrrhotite. Dark areas are voids; dusty, less-reflective areas are voids filled with epoxy that was used to impregnate the samples prior to polishing.
Most anhydrite crystals were large enough to permit analysis of both cores and rims (table 1, fig. 2A). In contrast, chalcopyrite crystals were too small to allow more than a single analytical crater, which nearly spanned the entire crystal (fig. 2B).
The anhydrite 34S values exhibit a 13 range in isotopic compositions (fig. 1A), more than six times the typical 2 analytical precision. The mean of the anhydrite 34S data is 8.3, with a standard deviation of 2.9. A bimodal distribution in the anhydrite 34S values is evident, with modes of 6.5 and 10.5. This apparent bimodality is real and statistically significant: a chi-squared test indicates a likelihood of less than 5/1,000 (e2=0.0045) for randomly deriving the observed distribution from a single normally distributed population of 34S values with the same mean and standard deviation. Though not shown explicitly, the data from this sample yield anhydrite core and rim 34S subpopulations that are statistically identical, each showing the same range and bimodality. There is no obvious petrographic difference between the two anhydrite modes.
A positive correlation exists between anhydrite core and rim 34S values (fig. 3A). These relations indicate that almost all of the anhydrite crystals are not isotopically zoned at a level detectable by SHRIMP (2 precision at 2 level): only two crystals exhibit a difference of more than 4 between core and rim values. Figure 3A also indicates that the observed 13 range in values is significant and not due to random analytical uncertainties: there are uniformly 34S-depleted and uniformly 34S-enriched anhydrite crystals.
Figure 3A. Plot of core versus rim SHRIMP 34S data (in , table 1) for anhydrite crystals from pumice of the June 12, 1991 eruption of Mount Pinatubo (sample collected from Mount Bagang). There are distinctly 34S-depleted crystals and distinctly 34S-enriched crystals. Only two crystals show a difference >4 between core and rim 34S values, indicative of intracrystalline isotopic zoning at a level detectable by SHRIMP. B, Plot of core versus rim SHRIMP 34S data (in , table 1) for anhydrite crystals in pumice from June 15, 1991 eruption of Mount Pinatubo (sample collected from Sacobia River valley). The data cluster much more tightly than those from June 12 anhydrites. No crystals show a difference >4 between core and rim 34S values.
Due to their smaller grain size and much lower modal abundance, the number of chalcopyrite analyses is few in comparison to anhydrite (table 1). The chalcopyrite 34S values cluster tightly, with an average of -1 (fig. 1A).
PUMICE FROM THE JUNE 15, 1991, ERUPTION
The main June 15 stage of the 1991 eruption sequence produced mostly phenocryst-rich dacite tephra and lesser phenocryst-poor dacite tephra (Wolfe and Hoblitt, this volume; Pallister and others, this volume). We obtained 44 SHRIMP 34S analyses on 22 crystals of anhydrite in a phenocryst-rich dacite pumice from this eruption (table 1). Owing to the small size (<30 m) of most sulfide grains in this sample, we obtained only a single SHRIMP 34S analysis for one chalcopyrite grain.
Representative photomicrographs of anhydrite and chalcopyrite from this pumice sample are shown in figure 4. In reflected light many of the anhydrite crystals are distinctly rounded, more so than in the June 12 pumice. This rounding occurs in direct contact with glass and thus is not due to partial dissolution of the anhydrite during sample preparation. Either the anhydrite partially dissolved in the melt or the silicic, viscous nature of the June 15 dacite melt promoted abrasion of the relatively soft anhydrite phenocrysts. This rounding differs from textures described by Bernard and others (1991), who observed sharp, euhedral contacts between June 15 anhydrite crystals and glass in transmitted light.
Figure 4A. Representative photomicrographs (polished thin sections in reflected light) of SHRIMP analytical craters in anhydrite, from pumice of June 15, 1991 eruption of Mount Pinatubo (sample collected from Sacobia River valley). Numbers in boxes are 34S values (in relative to CDT standard) for the craters identified by the arrows; craters are 20-30 m in diameter and about 5 m deep. Anhydrite crystals are anhedral to subhedral, with some rounded crystal boundaries in direct contact with glass. Cleavage is visible in some anhydrite crystals, as are hexagonal cross sections of apatite crystals. Dark areas are voids; dusty, less-reflective areas are voids filled with epoxy that was used to impregnate the samples prior to polishing.
Figure 4B. Photomicrograph (polished thin section in reflected light) of SHRIMP analytical crater in chalcopyrite, from pumice of June 15, 1991 eruption of Mount Pinatubo (sample collected from Sacobia River valley). Number in box is 34S value (in relative to CDT standard) for the crater identified by the arrow; crater is 20-30 m in diameter and about 5 m deep. Chalcopyrite crystal (ccp) is anhedral, with darker bornite (bn) exsolution lamellae, and occurs as isolated crystal in glass. Dark areas are voids; dusty, less-reflective areas are voids filled with epoxy that was used to impregnate the samples prior to polishing.
The analyzed chalcopyrite grain exhibits exsolution lamellae of bornite (fig. 4B), which likely formed by breakdown of magmatic Cu-Fe-S intermediate solid solution to these two phases plus pyrrhotite upon cooling to < 350°C (see Sugaki and others, 1975). Because equilibrium 34S differences between chalcopyrite and bornite are less than 1 at this temperature, and because the SHRIMP craters were placed so as to avoid the visible bornite lamellae, the SHRIMP 34S value of this chalcopyrite should be representative of the 34S value of the original magmatic sulfide (intermediate solid solution).
A single population of anhydrite 34S values was observed in this sample, with a mode of 7, a mean of 7.3 and a standard deviation of 1.6 (fig. 1B). This mode is virtually identical to the mode of the isotopically lighter anhydrites in the June 12 pumice samples (fig. 1A). No significant isotopic zoning is evident in any analyzed anhydrite crystal. A plot of core versus rim 34S values in June 15 anhydrite shows no statistically significant correlation (fig. 3B), exhibiting a tighter clustering than anhydrites in the June 12 pumice (fig. 3A).
Using conventional bulk sampling techniques, Imai and others (1993) and Bernard and others (this volume) obtained 34S values for the June 15 crystal-rich dacite that averaged 8.2 (Imai and others, acid leachates) and 7.4 (Bernard and others, water-soluble sulfate), in excellent agreement with our SHRIMP values (fig. 1B). The 34S value for chalcopyrite in the pumice from the June 15 eruption (fig. 1B) is within the range of 34S values for chalcopyrite from pumice of the June 12 eruption (fig. 1A).
BASALT INCLUSION
Pallister and others (1991, 1992, this volume) argue that the basalt inclusions found in some of the Pinatubo pyroclastic flows represent samples of an underplating mafic magma that may have slowly degassed sulfur into the overlying dacite and whose sudden intrusion to shallower levels in 1991 may have triggered the eruption of the dacite. Presumably, the sulfur isotopic composition of primary sulfide in the basalt inclusions may represent the isotopic composition of reduced sulfur present in the deep mafic melt.
Pyrrhotite and chalcopyrite grains are common in the basalt inclusion, but only one 25-m grain of chalcopyrite (fig. 5) was large enough to be analyzed during an analytical session with the new SHRIMP II, which has a spatial resolution of 10-15 m. Two SHRIMP craters on this grain yielded identical 34S values of 1 (fig. 1C), statistically indistinguishable from the 34S values for chalcopyrite in pumice samples from the June 12 and 15 eruptions (figs. 1A and 1B).
Figure 5. Photomicrograph (polished thin section in reflected light) of SHRIMP II analytical craters in chalcopyrite (ccp) (with minor bornite lamellae not clearly visible), in basalt inclusion from June 7-14 hybrid dome (sample collected from pyroclastic flow in Maraunot valley). Numbers in boxes are 34S values (in relative to CDT standard) for the craters identified by the arrows; each crater is 10-15 µm in diameter and about 5 m deep. Chalcopyrite crystal is anhedral and occurs as isolated crystal in feldspar. Dark areas are voids; dusty, less-reflective areas are voids filled with epoxy that was used to impregnate the samples prior to polishing.
VEIN IN DRILLCORE
Geothermal exploration wells drilled into the flank of Mount Pinatubo in 1988 encountered excess enthalpy fluids with highly variable compositions, gas contents, and generally acidic pH values (Buenviaje, 1991; Michels and others, 1991; Delfin and others, this volume). Drillcores from these wells preserve evidence of past hydrothermal mineral deposition from liquid-dominated fluids. Veins and massive alteration zones are generally dominated by calcite at <300°C and anhydrite at >300°C.
Within well PIN-2D, drillcore from a depth of 1,437 m contains a thick vein of intergrown coarsely crystalline anhydrite and lesser pyrite in bleached wall rock. Fluid inclusions in the anhydrite yield homogenization temperatures of 278-308°C and freezing point depressions of 1.1-4.8°C.
Pyrite crystals from the vein exhibit a narrow isotopic range with a mean 34S value of -2 (fig. 1D), statistically indistinguishable from 34S values of the chalcopyrites in the pumices and basalt inclusion. Coexisting vein anhydrite 34S values are slightly more variable, with an average of 18.7 (fig. 1D). They are significantly heavier than 34S values for anhydrites in pumice of the June 15 eruption, but the isotopically lightest values of the drillcore vein anhydrites are statistically indistinguishable from the isotopically heaviest values of anhydrites in pumice of the June 12 eruption.
DISCUSSION
PRIMARY ANHYDRITE AND SULFIDE
The unimodal anhydrite 34S population in pumice of the June 15 eruption (fig. 1B) implies that an isotopically uniform and well-mixed primary SO42- component was present in magma erupted during the main eruption. In contrast, the broader bimodal distribution of anhydrite 34S values in pumice of the June 12 eruption (fig. 1A) implies that isotopically distinct sources of SO42- contributed to this earlier, vent-clearing eruption.
A primary, juvenile origin for the 7 mode anhydrite population is supported by the apparent sulfur isotopic equilibrium between these anhydrites and coexisting sulfide in both pumice samples. In both the June 12 and June 15 magmas, oxygen fugacity was relatively high and melt SO42- was much greater than melt HS- (Rutherford and Devine, this volume), so the sulfur isotopic composition of primary melt sulfide should have been fixed (buffered) by the much larger masses of phenocryst CaSO4 and primary melt SO42-. Spectroscopic studies (Carroll and Rutherford, 1988) suggest that SO42- in melt and SO42- in anhydrite have the same chemical structure, so it can be assumed that sulfur isotopic fractionation between dissolved and crystalline SO42- should be negligible at magmatic temperatures.
In pumice of the June 15 eruption, the mean observed sulfur isotopic difference between anhydrite and chalcopyrite is 7-8 (fig. 1B). This is in excellent agreement with the expected equilibrium fractionation of 7-8 at 780°C (fig. 6), the estimated magma temperature prior to eruption as calculated from Fe-Ti oxide geothermometry (Rutherford and Devine, this volume).
Figure 6. Curves showing equilibrium sulfur isotopic fractionation (in ) between relevant S-bearing phases and H2S as a function of temperature (Ohmoto and Rye, 1979) over the range applicable to June 1991 Mount Pinatubo eruption products. Preeruptive temperatures for Mount Pinatubo magmas taken from Pallister and others (this volume) and Rutherford and Devine (this volume).
In pumice of the June 12 eruption, the mean observed isotopic difference between chalcopyrite and the isotopically lighter mode of anhydrite is 7-8, in reasonably good agreement with the expected equilibrium fractionation of 6-7 at 950°C (fig. 6), the estimated hybrid magma temperature prior to eruption (Rutherford and Devine, this volume). The mean observed difference of 11-12 between chalcopyrite and the isotopically heavier mode of anhydrite corresponds to a sulfur isotope equilibration temperature of <600°C (fig. 6), well below the estimated preeruption temperatures of Mount Pinatubo magmas.
ORIGIN OF ISOTOPICALLY HEAVY ANHYDRITE
Because the anhydrite in pumice of the June 12 eruption is not isotopically homogeneous, and because the isotopically heavier anhydrites are not in isotopic equilibrium with sulfide at the estimated preeruptive magma temperature, the heavier anhydrites must have been derived from extraneous sources. They must have been acquired relatively late (shallow) by the magma, such that complete mixing and isotopic reequilibration with primary SO42- and HS- could not occur prior to eruption.
One shallow source that could have provided late-stage extraneous SO42- is subvolcanic igneous rock derived from crystallization of earlier SO42--bearing dacite magma batches. However, primary anhydrite in such dacitic subvolcanic rocks would not necessarily be isotopically distinct from the primary anhydrite in the 1991 dacite magma. Also, the isotopically heavy anhydrites show no evidence of intergrowths with older dacitic material.
Another possible source of shallow extraneous SO42- may be local concentrations of hydrothermal anhydrite that occur within the flanks of Mount Pinatubo (Delfin and others, this volume; Cabel, 1990). Such anhydrite ultimately owes its origins to degassing of SO2 (and H2S) from earlier magma batches into overlying hydrothermal/ground-water systems, which commonly occur within the edifices of arc volcanoes in tropical latitudes (Reyes, 1990; Giggenbach, 1992).
This hydrothermal xenocryst hypothesis poses some textural difficulties, however. Hydrothermal anhydrite in such deposits often occurs intergrown with sulfide, apatite, calcite, and other authigenic phases, as well as acid-altered volcanic rock, but only apatite has been commonly seen intergrown with anhydrite crystals in Mount Pinatubo pumices. It may be possible that the dacite melt was saturated in anhydrite and apatite but strongly undersaturated in other hydrothermal phases, a situation that would have promoted the generation of partially resorbed hydrothermal xenoliths containing only anhydrite and apatite. Nonetheless, firm textural evidence for the xenolithic nature of the isotopically heavy June 12 anhydrites is lacking.
The isotopic compositions of vein anhydrite and pyrite in the drillcore sample are consistent with an ultimate origin by degassing of magmatic SO2 from earlier magma batches into hydrothermal fluids. Authigenic anhydrite forming in magma-hydrothermal systems is commonly enriched in 34S compared with the magmatic SO2 from which it was derived (Rye, 1993). In the presence of water at temperatures below 350°C, SO2 hydrolyzes according to the reaction
4SO2+4H2O=3H++3HSO4-+H2S
(Ohmoto and Rye, 1979). Because of the strong relative partitioning of sulfur isotopes among SO2, H2S, and SO42- at lower temperatures (fig. 6), aqueous SO42- generated by the breakdown reaction tends to be enriched in 34S relative to the original SO2 in magmatic vapor. Likewise, any aqueous H2S generated by the breakdown reaction tends to be depleted in 34S relative to the original SO2. The magnitudes of the relative enrichments and depletions of these decomposition products depend on the initial magmatic vapor SO2/H2S ratio and the temperature-oxygen fugacity path that the fluid follows as it cools (Rye, 1993). As temperatures cool to 400-200°C, the isotopically enriched SO42- and isotopically depleted H2S are often coprecipitated as hydrothermal anhydrite and sulfides.
The relatively enriched vein anhydrite 34S values and relatively depleted vein pyrite 34S values (fig. 1D) are consistent with such an origin for sulfur in the Mount Pinatubo drillcore vein minerals. The observed relative proportions of anhydrite being greater than pyrite in the vein are also consistent with the stoichiometry of the SO2 decomposition reaction. The observed sulfur isotopic fractionation of 21 between the vein anhydrite and pyrite corresponds to a sulfur isotopic equilibration temperature of 285°C (Ohmoto and Rye, 1979), in excellent agreement with the fluid inclusion temperatures.
If such isotopically enriched hydrothermal vein anhydrite was assimilated during the early June 1991 Mount Pinatubo eruptions, then this could explain the isotopically heavy mode of anhydrites in pumice of the June 12 eruption. Assimilation presumably would have sampled a variety of hydrothermal anhydrites that had been deposited over a range of SO2 decomposition conditions with a corresponding variety of 34S enrichments. For this reason the 34S range and mode of any anhydrite xenocrysts in pumice of the June 12 eruption should not necessarily coincide precisely with those of anhydrites in the single drillcore sample.
Because isotopically heavy anhydrites constitute nearly half of those analyzed in pumice of the June 12 eruption, a significant component of extraneous SO42- must have existed in this vent-clearing eruption. These heavy anhydrites are clearly out of isotopic equilibrium with the other anhydrites and sulfides. Their heaviest 34S values are statistically indistinguishable from the lightest values of hydrothermal vein anhydrites from one drillcore sample. Although a hydrothermal xenocryst origin for heavy anhydrites in the June 12 pumice is unsupported by firm textural evidence, no alternative hypothesis can meet all of the sulfur isotopic constraints.
In considering the feasibility of nonprimary sources of "excess" sulfur for the 1991 Mount Pinatubo eruptions, Gerlach and others (this volume) argue convincingly that any SO42- that was dissolved in shallow hydrothermal fluids is of insufficient mass to have been a significant source of sulfur to the main June 15 eruption. However, this does not mean that hydrothermal vein anhydrite, whose equivalent SO42- mass could have vastly exceeded the mass of SO42-dissolved in hydrothermal fluids, could not have contributed anhydrite xenocrysts to the vent-clearing June 12 and later stages of the eruption.
CONSTRAINTS ON ISOTOPIC COMPOSITION
OF SO2 GAS AND H2SO4 AEROSOLS
The sulfur isotopic composition of the erupted SO2 and resultant H2SO4 aerosols can be estimated from the primary anhydrite 34S values and preeruption temperatures. The estimated values depend on assumptions regarding the mechanism of SO2 generation.
Gerlach and others (this volume) and Westrich and Gerlach (1992) present a vapor saturation and accumulation model in which a hydrous CO2- and SO2-bearing vapor phase coexisted with the dacite magma prior to its ascent and eruption. If sulfur isotopic equilibrium between melt anhydrite and vapor SO2 was attained prior to eruption, and if irreversible anhydrite breakdown during eruption (Rutherford and Devine, this volume) did not contribute significantly to the SO2 content of the eruption, then the sulfur isotopic composition of the SO2 gas in the eruption cloud and the total sulfur isotopic composition of the bulk eruption (solids plus gas) can be estimated.
For a preeruption temperature of 780°C for the main June 15 eruption, the equilibrium isotopic fractionation between anhydrite and SO2 is 3.5 (fig. 6). If the 34S value of primary anhydrite in melt was 7 (fig. 1B), then any coexisting SO2 gas must have had a 34S value of 3.5.
Comparison of the total sulfur content of the erupted dacite with the amount of sulfur erupted as SO2 gas (from data in table 1 of Gerlach and others, this volume) implies that 16 to 57 mass percent of the total erupted sulfur was contained within the vapor phase prior to eruption, with the remainder occurring as anhydrite crystals in the melt. Therefore, the 34S value of total sulfur in preeruptive magma plus vapor would have been 5.0 to 6.5.
If the erupted SO2 gas was quantitatively oxidized to H2SO4 in the stratosphere, then sulfate aerosols from the June 15 eruption should have a 34S value identical to that of the SO2, 3.5. This estimated value is in excellent agreement with a 3.3 34S value for sulfate deposits collected from the engines of aircraft that experienced performance problems at altitudes of 10 to 12.5 km in June 1992 (Casadevall and Rye, 1994). However, there is no absolute certainty that these engine sulfate deposits were derived solely from quantitative oxidation of June 15 Mount Pinatubo SO2.
Rutherford and Devine (this volume) argue that decomposition of anhydrite during ascent decompression and eruption is another potential source of SO2 during the June 15 eruption. If such decomposition occurred rapidly and irreversibly, such as via the reaction
then the 34S value of the SO2 gas product would have been identical to that of the primary anhydrite, 7.
Bernard and others (this volume) analyzed two "fresh" Mount Pinatubo ash samples (collected on July 4, 1991) for soluble sulfate by using conventional bulk sampling techniques. They obtained 34S values of 9.4 and 9.7, significantly heavier than values for anhydrite in pumice of the June 15 eruption. They noted that the ash samples had total sulfur contents that were markedly higher than dacite pumice of the June 15 eruption and therefore concluded that the ash had adsorbed significant amounts of SO42- derived from SO2 in the eruption cloud. If so, this implies that either the SO2 was isotopically heavy to begin with (34S >9) and was quantitatively converted to SO42- or that oxidation of SO2 to SO42- in the eruption plume favors the heavier sulfur isotope and produces a significant sulfur isotopic enrichment in the SO42- product that is subsequently adsorbed on ash. Given the constraints on the 34S value of erupted SO2 as discussed above, only the latter hypothesis (preferential oxidation of 34SO2 over 32SO2) appears feasible as an explanation for the isotopically heavy SO42- adsorbed on ash of the June 15 eruption. This implies that quantitative conversion of SO2 to SO42- aerosols in the atmosphere may not occur, in which case the original sulfur isotopic signature of volcanic SO2 may not necessarily be preserved in aerosols.
Direct analysis of the 34S values of aircraft-collected stratospheric Pinatubo sulfate aerosols might help resolve the uncertainties and apparent conflicts over the isotopic composition of the erupted SO2 and the mechanism of SO2 generation. Such analyses might also indicate the extent of any isotopic fractionation during conversion of SO2 to adsorbed sulfate on ash in the eruption plume or to H2SO4 aerosols over longer residence times in the stratosphere. Once such processes are better known, it may prove possible to use the 34S composition of sulfate aerosols as a precise "fingerprint" to distinguish among sources such as volcanic eruptions, combustion of fossil fuels, terrestrial biogenic sources, and seasalt (see Calhoun and others, 1991; Nriagu and others, 1991).
SULFUR IN BASALT AND IMPLICATIONS
FOR SULFUR IN DACITE
The single analyzed grain of sulfide in the basalt inclusion (34S=1, fig. 1C) is primary in appearance, occurring within plagioclase (fig. 5). Given that such basalt is thought to be quenched inclusions of an eruption-triggering (fresh) basaltic magma batch (Pallister and others, this volume), the analyzed sulfide is likely representative of the sulfur isotopic composition of sulfide that was dissolved in the underplating basalt melt. The sulfur isotopic composition of sulfides in gabbros from the nearby and possibly underlying Zambales Ophiolite Complex is also 1 (J. Fournelle, Univ. of Wisconsin, and R. Carmody, USGS, written commun., 1994). Mafic to ultramafic igneous rocks typically contain primary sulfide minerals with 34S values between -1 and 2 (Ohmoto and Rye, 1979).
Because sulfur isotopic fractionation between melt HS-, chalcopyrite, pyrrhotite, and H2S vapor is negligible at >1200°C (fig. 6; Ohmoto and Rye, 1979), crystal fractionation of solid sulfides or loss of H2S gas during basalt magma crystallization should produce negligible Rayleigh effects. If the basalt became oxidized at some point (e.g., via interaction with the base of the more hydrous dacite "mush") such that SO2 could form in addition to H2S in an exsolving basaltic gas phase, then preferential partitioning of 34S-enriched SO2 into the gas could have occurred upon relatively low extents of basalt degassing, even at the high temperatures required (fig. 6). However, even under such oxidizing conditions, isotopic mass balance requires that the basalt could not have degassed sulfur (as a supercritical fluid with relatively high SO2/H2S) that was more than 2 heavier than the initial sulfide in the basalt melt at 1200°C (Ohmoto and Rye, 1979, eq. 10.6). Rayleigh effects with progressive degassing would make this constraint even less than 2. This means that any vapor degassed into the dacite by the basalt was constrained to have had a bulk 34S value no greater than 3, given a 1 value for sulfide in underplating basalt.
These constraints make the source of the isotopically enriched primary SO42- in the dacite problematical. If the vapor saturation and accumulation model (Westrich and Gerlach, 1992; Gerlach and others, this volume) is correct, then the bulk sulfur isotopic composition of the June 1991 dacite eruption (gas plus solids) was 5.0 to 6.5, as was calculated above. If the anhydrite decomposition model of Rutherford and Devine (this volume) is correct, then the bulk sulfur isotopic composition of the preeruptive dacite magma must have approached that of the primary anhydrites, 7. Either model yields a dacite bulk sulfur isotopic composition that is significantly heavier than the maximum possible (3) from degassing of the underlying basalt melt. Other processes in addition to degassing of basalt are thus required to explain the bulk sulfur isotopic enrichment of the dacite.
Given the high anhydrite/pyrite ratio and 34S-enriched nature of anhydrite in hydrothermal deposits occurring in the flanks of Mount Pinatubo, it is possible that repeated cycles of heating and noneruptive intrusion of the dacite into the volcanic pile could have allowed assimilation of such anhydrite by the magma. This could have provided sufficient time for mixing and isotopic homogenization of the resultant anhydrite "phenocrysts." However, if this process occurred many times and was a major means of sulfur isotopic enrichment of the dacite, then xenoliths of acid-altered volcanic rocks with other hydrothermal mineral phases should have been abundant; these are not clearly manifested in the dacite.
Crystal fractionation and gravity settling of 34S-depleted sulfide crystals from the dacite could result in limited 34S enrichment of the remaining magma at temperatures appropriate for dacite of the June 15 eruption (fig. 6). However, given that inclusions of quenched underplating basalt magma were entrained in the dacite eruption products, one could expect that inclusions of any gravity-settled basal cumulate sulfides should also be seen in the basalt inclusions; none have been reported.
If assimilation or crystal fractionation mechanisms are thus infeasible or unlikely, then the only other means of generating a preeruptive dacite with bulk primary 34S >3 is via passive (noneruptive) steady-state loss of SO2 or H2S vapor from the dacite, in dynamic equilibrium with melt SO42- and anhydrite phenocrysts.
A STEADY-STATE LONG-TERM
DEGASSING MODEL
Over long time periods, H2S and SO2 from underplating basalt (total 34S between 1 and 3) could have been degassed passively into the more hydrous dacite; such degassing could cause generation of SO2 and CaSO4 in the dacite melt via reactions such as
As shown in figure 6, if these redox reactions attain isotopic equilibrium under passive conditions, they result in the preferential partitioning of 34S from basaltic H2S and SO2 into SO42-; the SO42- then dissolves in the dacite melt and ultimately precipitates as anhydrite phenocrysts. The remaining relatively 34S depleted H2S and SO2 will also dissolve in the dacite melt initially, but at some point a slow buildup of these volatiles could cause passive vapor saturation in the dacite. The gases would then be able to diffuse slowly through and out of the dacite magma over long time periods. The dacite melt could maintain a high oxidation state because the hydrogen gas produced by the reactions above should also diffuse slowly out of the melt, once vapor saturation is obtained.
If basaltic heat and H2S-SO2 gas are thus added slowly to the base of the dacite, it might reach and maintain vapor saturation passively, allowing steady-state noneruptive loss of the 34S-depleted gases and hydrogen from the dacite over long time periods. This steady-state flux of 34S-depleted gas diffusing up through the dacite would be accompanied by a buildup of 34S-enriched SO42- in the melt as anhydrite phenocrysts; the result would be a dynamic isotopic equilibrium. Eventually, the sulfur isotopic compositions of H2S and SO2 in the dacite would be buffered by the large accumulated reservoir of 34S-enriched anhydrite phenocrysts.
A schematic illustration of this model is shown in figure 7. This steady-state passive vapor loss model is consistent with the model for preeruptive vapor saturation and accumulation proposed by Westrich and Gerlach (1992) and Gerlach and others (this volume) and provides an explanation for the otherwise problematic 34S-enriched nature of the dacite magma relative to the underplating basalt magma. Under conditions where a sudden large increase in basaltic heat and gas input causes the dacite to undergo catastrophic vapor oversaturation, a combination of SO2-bearing vapor and high 34S SO42- -- rich magma would erupt.
Figure 7. Schematic model for degassing of basaltic H2S-SO2 vapor into dacite and its subsequent partitioning into relatively 34S-depleted SO2-H2S vapor and relatively 34S-enriched melt SO42- and CaSO4 phenocrysts. Long-term steady-state passive (noneruptive) degassing of dacite, with vapor in dynamic isotopic equilibrium with SO42- in anhydrite, results in a dacite melt that is more enriched in 34S than the basaltic vapor. See text for detailed discussion.
CONCLUSIONS
Application of ion microprobe analytical techniques to the June 1991 eruption products of Mount Pinatubo has allowed the sulfur isotopic systematics of the volcanic system to be worked out in a manner that is not possible by use of conventional bulk techniques. The pumices contain a primary population of anhydrite phenocrysts with 34S 7 that are in sulfur isotopic equilibrium with primary sulfides having 34S -1. Pumice from the early vent-clearing June 12 eruption contains a significant component of isotopically heavy anhydrite (34S 10) that may represent xenocrysts derived from hydrothermal anhydrite deposits occurring within the flanks of the volcano.
Isotopic mass balance implies that the bulk sulfur isotopic composition of the preeruptive magma (solids + melt +- gas) was between 5 and 7 and that the SO2 cloud from the June 15 eruption had a 34S value between 3.5 and 7, depending on whether the SO2 existed in equilibrium with the magma in a preeruptive vapor phase or, instead, was generated rapidly by irreversible anhydrite decomposition during magma ascent and eruption. H2SO4 aerosols produced in the stratosphere from the erupted SO2 may have a similar isotopic composition, although subsequent 34S enrichment during atmospheric oxidation of SO2 cannot be ruled out.
Basalt inclusions in the June 7-14 hybrid dome contain sulfide with 34S 1, implying that any H2S/SO2 vapor degassed into the dacite by the basalt must have had a bulk 34S value between 1 and 3. Because the eruption was significantly more enriched in 34S than this, other sulfur isotopic partitioning processes must have affected the dacite magma over relatively long time periods prior to the June 15 eruption. In light of textural and sulfur isotopic constraints, assimilation of 34S-enriched hydrothermal anhydrite and (or) crystal fractionation of 34S-depleted sulfide seem to be unlikely mechanisms for generating 34S-enriched dacite. Instead, a mechanism involving long-term passive degassing of basaltic sulfur into the dacite and the attainment of steady-state dynamic isotopic equilibrium between a diffusing 34S-depleted vapor phase and 34S-enriched anhydrite phenocrysts can explain the sulfur isotope systematics.
ACKNOWLEDGMENTS
Research was sponsored by National Science Foundation grant EAR 93-04434 to M.A. McKibben. C.S. Eldridge acknowledges the support of the Australian Research Council and a Visiting Fellowship from the Research School of Earth Sciences. A.G. Reyes acknowledges past support from the Geothermal Division of the Philippine National Oil Corporation. This research project owes its origins to Werner F. Giggenbach, who first suggested that we apply the SHRIMP to the "sulfur problem" in volcanic eruptions. T.J. Casadevall, T.M. Gerlach, J.S. Pallister, and H.R. Westrich graciously provided many of the samples for this study. Comments on earlier versions of the manuscript by T.J. Casadevall, D.E. Crowe, A.Imai, J.F. Luhr, C.G. Newhall, and W.C. Shanks, and discussions with L. Baker, T.J. Casadevall, J.Fournelle, T.M. Gerlach, J.S. Pallister, R.O. Rye, M.J. Rutherford, R.I. Tilling, and H.R. Westrich, all resulted in a much improved manuscript. In particular, Chris Newhall prompted a critical reevaluation and improvement of many of the senior author's early ideas and biases.
REFERENCES CITED
Baker, L., and Rutherford, M.J., 1992, Anhydrite breakdown as a possible source of excess sulfur in the 1991 Mount Pinatubo eruption: Eos, Transactions, American Geophysical Union, v. 73 (Suppl.), p. 625.
Bernard, A., Demaiffe, D., Mattielli, N., and Punongbayan, R.S., 1991, Anhydrite-bearing pumices from Mount Pinatubo: Further evidence for the existence of sulphur-rich silicic magmas: Nature, v. 354, p. 139-140.
Bernard, A., Knittel, U., Weber, B., Weis, D., Albrecht, A., Hattori, K., Klein, J., and Oles, D., this volume, Petrology and geochemistry of the 1991 eruption products of Mount Pinatubo.
Bluth, G.J.S., Doiron, S.D., Schnetzler, C.C., Kreuger, A.J., and Walter, L.S., 1991, Global tracking of the SO2 clouds from the June 1991 Mount Pinatubo eruptions: Geophysical Research Letters, v. 19, p. 151-154.
Buenviaje, M.M., 1991, Geochemical characteristics of acid fluids in Mt. Pinatubo, Philippines: Stanford University, Proceedings, Sixteenth Workshop on Geothermal Reservoir Engineering, Report SGP-TR-134, p. 267-273.
Cabel, A.C., Jr., 1990, Petrology of the permeable section of drillhole PIN-3D, Mt. Pinatubo geothermal prospect, Luzon, Philippines: University of Auckland, New Zealand, Geothermal Institute, Report No. 90.04, 30 p.
Calhoun, J.A., Bates, T.S., and Charlson, R.J., 1991, Sulfur isotopic measurements of submicrometer sulfate aerosol particles over the Pacific Ocean: Geophysical Research Letters, v. 18, no. 10, p. 1877-1880.
Carroll, M.R., and Rutherford, M.J., 1987, The stability of igneous anhydrite: Experimental results and implications for sulfur behavior in the 1982 El Chichón trachyandesite and other evolved magmas: Journal of Petrology, v. 28, p. 781-801.
------1988, Sulfur speciation in hydrous experimental glasses of varying oxidation state: Results from measured wavelength shifts of sulfur X-rays: American Mineralogist, v. 73, p. 845-849.
Casadevall, T.J., and Rye, R.O., 1994, Sulfur in the atmosphere: Sources and effects on jet-powered aircraft [abs.]: Eos, Transactions, American Geophysical Union, v. 75 (Suppl.), p. 78.
Delfin, F.G., Jr., Villarosa, H.G., Layugan, D.B., Clemente, V.C., Candelaria, M.R., Ruaya, J.R., this volume, Geothermal exploration of the pre-1991 Mount Pinatubo hydrothermal system.
Devine, J.D., Sigurdsson, H., Davis, A.N., and Self, S., 1984, Estimates of sulfur and chlorine yield to the atmosphere from volcanic eruptions and potential climatic effects: Journal of Geophysical Research, v. 89, p. 6309-6325.
Eldridge, C.S., Compston, W., Williams, I.S., Both, R.A., Walshe, J.L., and Ohmoto, H., 1988, Sulfur isotope variability in sediment-hosted massive sulfide deposits as determined using the ion microprobe SHRIMP: I. An example from the Rammelsberg orebody: Economic Geology, v. 83, p. 205-211.
Eldridge, C.S., Compston, W., Williams, I.S., Harris, J.W., and Bristow, J.W., 1991, Isotope evidence for the involvement of recycled sediments in diamond formation: Nature, v. 353, p. 649-653.
Eldridge, C.S., Compston, W., Williams, I.S., and Walshe, J.L., 1989, Sulfur isotope analyses on the SHRIMP ion microprobe, in Shanks, W.C., III, and Criss, R.E., eds., New frontiers in stable isotope analysis, U.S. Geological Survey Bulletin 1890, p. 163-174.
Eldridge, C.S., Compston, W., Williams, I.S., Walshe, J.L., and Both, R.A., 1987, In situ microanalysis for 34S/32S ratios using the ion microprobe SHRIMP: International Journal of Mass Spectrometry and Ion Processes, v. 76, p. 65-83.
Eldridge, C.S., Williams, N., and Walshe, J.L., 1993, Sulfur isotopic variability in sediment-hosted massive sulfide deposits as determined using the ion microprobe SHRIMP: II. A study of the H.Y.C. deposit at McArthur River, Northern Territory, Australia: Economic Geology, v. 88, p. 1-26.
Fournelle, J., 1990, Anhydrite in Nevado del Ruiz November pumice: Relevance to the sulfur problem: Journal of Volcanology and Geothermal Research, v. 42, p. 189-201.
Gerlach, T.M., Westrich, H.R., and Symonds, R.B., this volume, Preeruption vapor in magma of the climactic Mount Pinatubo eruption: Source of the giant stratospheric sulfur dioxide cloud.
Giggenbach, W.F., 1992, Magma degassing and mineral deposition in hydrothermal systems along convergent plate boundaries: Economic Geology, v. 87, p. 1927-1944.
Imai, A., Listanco, E.L., and Fujii, T, 1993, Petrologic and sulfur isotopic significance of highly oxidized and sulfur-rich magma of Mt. Pinatubo, Philippines: Geology, v. 21, p. 699-702.
Luhr, J.F., 1990, Experimental phase relations of water- and sulfur-saturated arc magmas and the 1982 eruptions of El Chichón volcano: Journal of Petrology, v. 31, p. 1071-1114.
Luhr, J.F., Carmichael, I.S.E., and Varenkamp, J.C., 1984, The 1982 eruptions of El Chichón volcano, Chiapas, Mexico: Mineralogy and petrology of the anhydrite-bearing pumices: Journal of Volcanology and Geothermal Research, v. 23, p. 69-108.
Matthews, S.J., Jones, A.P., and Bristow, C.S., 1992, A simple magma-mixing model for sulfur behavior in calc-alkaline volcanic rocks: Mineralogical evidence from Mount Pinatubo 1991 eruption: Journal of the Geological Society of London, v. 149, p. 863-866.
McKibben, M.A., and Eldridge, C.S., 1989, Sulfur isotopic variations among minerals and fluids in the Salton Sea geothermal system: A SHRIMP ion microprobe and conventional study of active ore genesis in a sediment-hosted environment: American Journal of Science, v. 289, p. 661-707.
------1990, Radical sulfur isotopic zonation of pyrite accompanying boiling and epithermal gold deposition: A SHRIMP study of the Valles Caldera: Economic Geology, v. 85, p. 1917-1925.
------1995, Microscopic sulfur isotopic variations in ore minerals from the Viburnum Trend, southeastern Missouri, U.S.A.: A SHRIMP study: Economic Geology, vol. 90, p. 228-245.
McKibben, M.A., Eldridge, C.S., Bethke, P.M., and Rye, R.O., 1993, SHRIMP study of extreme bacteriogenic sulfur isotope enrichments in the moat sediments of the Creede Caldera [abs.]: Geological Society of America, Abstracts with Program, v. 25, no. 6, p. A-316.
Michels, D.E., Clemente, V.C., and Ramos, M.N., 1991, Hydrochemical features of a geothermal test well in a volcanic caldera, Mt. Pinatubo, Philippines: Stanford University, Proceedings, Sixteenth Workshop on Geothermal Reservoir Engineering, Report SGP-TR-134, p. 261-266.
Nriagu, J.O., Coker, R.D., and Barrie, L.A., 1991, Origin of sulfur in Canadian Arctic haze from isotope measurements: Nature, v. 349, p. 142-145.
Ohmoto, H., and Rye, R.O., 1979, Isotopes of sulfur and carbon, in Barnes, H.L., ed., Geochemistry of hydrothermal ore deposits (2d. ed.): Wiley, p. 509-567.
Pallister, J.S., Hoblitt, R.P., Meeker, G.P., Knight, R.J., and Siems, D.F., this volume, Magma mixing at Mount Pinatubo: Petrographic and chemical evidence from the 1991 deposits.
Pallister, J.S., Hoblitt, R.P., Newhall, C.G., and Scott, W.E., 1992, A basalt trigger for the 1991 eruptions of Pinatubo volcano? Yes. [abs.]: Eos, Transactions, American Geophysical Union, v. 73 (Suppl.), p. 347.
Pallister, J.S., Hoblitt, R.P., and Reyes, A.G., 1991, A basalt trigger for the 1991 eruptions of Pinatubo volcano?: Nature, v. 356, p. 426-428.
Reyes, A.G., 1990, Petrology of Philippine geothermal systems and the application of alteration mineralogy to their assessment: Journal of Volcanology and Geothermal Research, v. 43, p. 279-309.
Rudnick, R.L., Eldridge, C.S., and Bulanova, G.P., 1993, Diamond growth history from in situ measurement of Pb and S isotopic compositions of sulfide inclusions: Geology, v. 21, p. 13-16.
Rutherford, M.J., and Devine, J.D., this volume, Preeruption pressure-temperature conditions and volatiles in the 1991 dacitic magma of Mount Pinatubo.
Rye, R.O., 1993, The evolution of magmatic fluids in the epithermal environment: the stable isotope perspective: Economic Geology, v. 88, p. 733-753.
Rye, R.O., Luhr, J.F., Wasserman, M.D., 1984, Sulfur and oxygen isotope systematics of the 1982 eruptions of El Chichón volcano, Chiapas, Mexico: Journal of Volcanology and Geothermal Research, v. 23, p. 109-123.
Sigurdsson, H., 1990, Assessment of the atmospheric impact of volcanic eruptions, in Sharpton, V.L., and Ward, P.D., eds., Global catastrophes in Earth history: Geological Society of America Special Paper 247, p. 99-110.
Sugaki, A., Shima, H., Kitakaze, A., and Harada, H., 1975, Isothermal phase relations in the system Cu-Fe-S under hydrothermal conditions at 350° and 300°C: Economic Geology, v. 70, p.806-823.
Westrich, H.R., and Gerlach, T.M., 1992, Magmatic gas source for the stratospheric SO
2
cloud from the June 15, 1991, eruption of Mount Pinatubo: Geology, v. 20, p. 867-870.
Wolfe, E.W. and Hoblitt, R.P., this volume, Overview of the eruptions.
FIRE and MUD Contents
PHIVOLCS | University of Washington Press | U.S.Geological Survey
This page is <https://pubs.usgs.gov/pinatubo/mckibben/>
Contact: Chris Newhall
Last updated 06.11.99