40 Ar/ 39 Ar Geochronology of Magmatic-Steam Alunite from Alunite Ridge and Deer Trail Mountain, Marysvale Volcanic Field, Utah: Timing and Duration of Miocene Hydrothermal Activity Associated with Concealed Intrusions

: Porphyry and epithermal deposits are important sources of base and precious metals. Most actively mined deposits have been exhumed such that ore bodies are relatively close to the surface and are therefore locatable and economic to extract. Identifying and characterizing concealed deposits, particularly more deeply buried porphyry deposits, represents a far greater challenge for mineral exploration, and will become progressively more important as near-surface resources are gradually exhausted over time. We report high-precision 40 Ar/ 39 Ar dates for coarsely crystalline alunite that precipitated from magmatic steam in open fractures in Oligocene dacitic volcanic rocks, and a SHRIMP 206 Pb/ 238 U zircon date for one of several rhyolite dikes present at Alunite Ridge and Deer Trail Mountain, Utah. Both the magmatic-steam alunite and rhyolite dikes are related to concealed intrusions. The rhyolite dike yielded an age of 30.72 ± 0.36 Ma, which is older than a commonly cited 27.1 Ma age estimate for the Three Creeks Tuff Member of the Bullion Canyon Volcanics that is cut by the dike. 40 Ar/ 39 Ar data for samples of magmatic-steam alunite and sericite from six mines and prospects provide evidence for at least two periods of episodic hydrothermal activity at ca. 15.7 – 15.1 Ma and ca. 14.7 – 13.8 Ma, with the older and younger pulses of activity recorded at the more eastern and western sites, respectively. These two periods of hydrothermal activity are consistent with previous interpretations that Alunite Ridge and Deer Trail Mountain are underlain by two concealed porphyry stocks. 40 Ar/ 39 Ar analyses of individual bands in a sample of massive, centimeter-scale banded vein alunite yield indistinguishable ages with a weighted mean of 13.98 ± 0.12 Ma, consistent with a short-lived ( ≲ 250 ka) magmatic event with episodic vapor discharge recurring on short timescales ( ≲ 36 ka). 40 Ar/ 39 Ar geochronology of magmatic-steam alunite is a valuable tool to constrain the timing and duration of magmatic hydrothermal activity associated with unexposed intrusions and potentially porphyry deposits, and therefore may be useful in exploration.


Introduction
Porphyry and epithermal deposits are the most significant global resources for copper (Cu), molybdenum (Mo), and rhenium (Re), and are major sources of gold (Au), silver (Ag), tellurium (Te), selenium (Se), mercury (Hg), and other metals [1]. Both types of deposits are hydrothermal in origin; high-temperature (700-350 °C) porphyry deposits generally form at depths between 2 and 10 km from predominantly magmatic fluids, while low-temperature (ca. 300-150 °C) epithermal deposits form at depths <1.5 km from mixtures of magmatic and meteoric fluids [1,2]. Multiple lines of evidence from intermineral intrusions (where preserved; [3]), numerical modeling (e.g., [4,5]), and geochronologic studies (e.g., [6][7][8][9][10]) indicate that the timing and duration of (potentially episodic) magmatic and hydrothermal activity are generally closely linked, though under some conditions (e.g., depending on the permeability of the host rocks and the size of the primary intrusion) hydrothermal systems may persist for nearly 1 Ma after an episode of magmatism. In addition, numerical modeling of porphyry Cu systems indicates that the overall volume and duration of magmatic (and, therefore, hydrothermal) activity is the primary control on the endowment of a deposit [5], with secondary factors such as the efficiency of ore mineral precipitation contributing to the variability of endowment observed in particular types of porphyry deposits (e.g., Au-rich porphyry deposits; [11]). If these controls also apply to other types of porphyry deposits (e.g., Climax-type Mo deposits), then high-precision geochronology (e.g., 40 Ar/ 39 Ar) to establish the timing and duration of magmatic and hydrothermal activity in epithermal deposits is a powerful exploration tool for concealed porphyry deposits.
Most actively mined deposits have been exhumed such that the ore bodies are relatively close to the surface and economically extractable. More deeply buried and undiscovered porphyry systems may have hydrothermal alteration halos that extend to the surface. Alunite, which forms by a variety of processes in supergene and hypogene environments and is indicative of different types of alteration and/or different types of mineralization, has potential for mineral deposit prospecting, e.g., [12]. Magmatic hydrothermal replacement alunite forms as SO2 laden magmatic vapor cools and contracts at the top of the magmatic vapor plume and equilibrates with wall rock, with alunite replacing aluminosilicates in the host rocks and disseminated pyrite forming as a result of H2S in the fluids replacing Fe-bearing minerals. In contrast, magmatic-steam vein alunite forms as highly oxidizing magmatic vapor rapidly expands unimpeded along open fractures all the way to the surface without significant interaction with wall rock [12]. Vein alunite generally precipitates as banded sequences of fine-grained to dendritic to terminated crystals growing from the fracture walls towards the center, with trace to locally minor amounts of hematite.
Both types of alunite can constrain the timing and duration of hydrothermal (and, indirectly, magmatic) activity above concealed intrusions. Published geologic, structural, and geochemical data from Alunite Ridge and Deer Trail Mountain indicate a strong likelihood that a concealed porphyry system is present at depth [13,14]. In this investigation, we analyzed samples of nearly pure magmatic-steam vein alunite (most of which are coarsely crystalline; two are fine-grained) using the 40 Ar/ 39 Ar method as a means of establishing the timing and duration of magmatic degassing at Alunite Ridge and Deer Trail Mountain, Utah. Here, we document 40 Ar/ 39 Ar data for magmatic-steam alunite that record at least two periods of episodic hydrothermal activity. We show that magmaticsteam alunite-which formed above two concealed intrusions-records numerous volatile exhalations that occurred over short durations (≲250 ka), with a recurrence interval ≲36 ka. We also test whether age variations are recorded between cm-scale bands of alunite, and place the timing of magmatic-steam alunite within a larger tectonic context of migrating Miocene bimodal magmatic activity following rollback of the previously subducted Farallon slab. We also report a zircon 206 Pb/ 238 U date for one of several felsic dikes that cut the Bullion Canyon Volcanics at Alunite Ridge and Deer Trail Mountain to explore how these intrusions relate temporally to the concealed porphyry system and potentially to mineralization in the area. Our results demonstrate the utility of highprecision 40 Ar/ 39 Ar geochronology of alunite as an exploration tool, particularly for locating and characterizing deeply concealed porphyry systems.

Magmatic-Steam Alunite at Marysvale, Utah
The Marysvale volcanic field is located in the High Plateaus of Utah where the Colorado Plateau transitions to the Basin and Range (Figure 1). It is situated near the northeastern terminus of a broad belt of Paleogene and Neogene igneous rocks and associated mineral deposits that stretch more than 350 km from southeastern Nevada into central Utah [15]. The Marysvale volcanic field is dominated by an upper Eocene-lower Miocene sequence of calc-alkaline intermediate rocks (andesite to low-silica rhyolite lava flows, ash-flow tuffs, volcanic breccias, and debris flows) erupted between about 35 Ma and 22 Ma, and a Miocene bimodal rhyolite and basalt suite erupted between 22 Ma and 14 Ma (e.g., [16]). These upper Eocene, Oligocene, and Miocene volcanic rocks unconformably overlie Paleozoic, Mesozoic, and lower Cenozoic sedimentary rocks [13]. At Alunite Ridge and Deer Trail Mountain, units of the ca. 34-22 Ma Bullion Canyon Volcanics are exposed atop the older sedimentary rocks (e.g., [17]).   [14] (see also Cunningham et al. [16] for detail of the Central intrusion). See Figure 1 for location.
The rhyolitic rocks of the younger Mount Belknap Volcanics that are exposed in the Marysvale area were erupted from two distinct igneous centers: a western source area dominated by the mostly Oligocene Mt. Belknap caldera located in the central Tushar Mountains, and an elongate (ca. 24 km by 5 km) eastern source area in the Antelope Range that stretches from northeast of Marysvale to Alunite Ridge and Deer Trail Mountain ( Figure 1) [13,16]. Igneous activity in the eastern source area began ca. 21-19 Ma northeast of Marysvale with shallow intrusions and the eruption of rhyolite tuffs, then migrated to the southwest over several Ma and culminated in the formation of the Deer Trail Ag-Aubase metal mantos and alunite veins at Alunite Ridge and Deer Trail Mountain, likely associated with the intrusion of two concealed stocks ( Figure 2) [13,14,16,18]. However, no bimodal volcanic or intrusive rocks are exposed at Alunite Ridge or Deer Trail Mountain.
Alunite Ridge, located ca. 12 km southwest of Marysvale in the Tushar Mountains (Figures 1 and 2), hosts a series of small mines and excavations including the world's largest known deposits of pure, coarsely crystalline K-alunite in banded veins up to 20 m thick [13]. Mining of K-alunite at Alunite Ridge first began when Germany restricted exports of potash in 1914. Between 1915 and 1920 the principle mine, Mineral Products, produced an estimated 250,000 tons of potash from an estimated 3 million total tons of vein-type alunite. After a hiatus, mining resumed briefly during World War II when an additional 12,000 tons of replacement-type alunite was processed to recover alumina due to a shortage of bauxite for aluminum production [19].
Advanced argillic alteration is common in rocks of the Marysvale area, and replacement-and vein-type alunite formed in distinct epithermal environments and locations [13]. The replacement alunite is fine-grained, and formed in near-surface, highly oxidizing conditions during hydrothermal alteration of volcanic rocks associated with intrusions of quartz monzonite stocks with published K/Ar and 40 Ar/ 39 Ar ages of 15,20]. These replacement deposits are mostly located near the periphery of the Central intrusion ( Figure 2) [16], though several smaller deposits are associated with the intrusive core of the Monroe Peak caldera [13].
In contrast to replacement alunite in the Marysvale area, vein-type alunite at Alunite Ridge and Deer Trail Mountain is coarsely crystalline, banded at the centimeter scale, and precipitated from "magmatic steam", which is generally defined as a low-density fluid that ascends rapidly from a degassing magma to shallow levels and that may contain one or two subcritical phases (liquid and vapor; [21]). At Alunite Ridge, magmatic-steam alunite filled extensional fractures to form veins that are steeply dipping and cut through the previously erupted dacitic rocks of the Bullion Canyon Volcanics, and cut or parallel rhyolitic quartz-porphyry dikes of unknown age. Some of the alunite veins are enveloped by magmatic hydrothermal replacement alunite. The vein alunite (and associated enveloping replacement alunite, where present) is likely genetically related to two inferred stocks beneath Alunite Ridge and Deer Trail Mountain. These intrusions produced concentric and radial fractures superposed on regional (generally N-S to NE-SW trending) Basin and Range structures and zoned alteration halos that are most intense at Alunite Ridge and generally grade outward from advanced argillic alteration to propylitic alteration [13,14]. The vein alunite contains gas-rich fluid inclusions, indicating precipitation from a low-density SO2-rich vapor that had little interaction with the wall rocks, and S, H, and O-isotope data indicate a magmatic origin of the steam with no significant isotopic exchange with the wall rocks (e.g., as evidenced by disequilibrium d 34 S compositions) or meteoric water [12,13,21,22]. In comparison, typical replacement alunite in Marysvale area (and worldwide) has relatively gas-poor fluid inclusions (with fluids formed from more dense, relatively SO2-poor and H2S-rich fluids) and isotopic compositions (e.g., for d 34 S) that more closely approach equilibrium conditions [12,21].
Bassett et al. [23], Caskey and Shuey [24], and Steven et al. [15] reported K/Ar dates for numerous igneous rocks, alunite, and sericite in the Marysvale area, including two dates for vein-type alunite from the Mineral Products mine of 14.3 ± 0.6 (0.7) Ma and 13.8 ± 0.6 (0.7) Ma (2σ, Table S1). (If available, we show external uncertainties, which include decay or decay plus monitor age uncertainties for K/Ar and 40 Ar/ 39 Ar dates, in parentheses following internal uncertainties for each date in the paper.) The samples of sericite studied by Bassett et al. [23] were collected from the Deer Trail mine workings, and indicate hydrothermal activity altered the Permian country rocks at ca. 14.6-13.2 Ma. The data reported by Caskey and Shuey [24] and Steven et al. [15] include dates for the Three Creeks Tuff Member of the Bullion Canyon Volcanics and the underlying Wah Wah Springs Tuff Member of the Needles Range Group, which have been interpreted (along with zircon fission track dates) as indicating the Three Creeks Tuff Member is ca. 27.1 Ma [15,25]. More recently, Best et al. [26] published a weighted mean of 11 plateau dates of 30.06 ± 0.10 (0.16) Ma (2σ) for plagioclase separates from samples of the Wah Wah Springs Tuff Member (Table S2). Waltenberg [27] and Ren and Vasconcelos [28] analyzed several neutron-irradiated aliquots of one sample of coarsely crystalline pink alunite from Alunite Ridge (Berkeley mineral collection, sample number 5915) to study Ar diffusion and 39 Ar recoil effects in alunite. Because the samples had been irradiated, they were also able to compute plateau 40 Ar/ 39 Ar dates ranging from 15.96 ± 0.11 (0.12) Ma to 15.553 ± 0.072 (0.081) Ma (2σ, Table S2). (The K/Ar and 40 Ar/ 39 Ar dates cited here have been recalculated using decay, isotopic, and monitor age parameters consistent with those we used to compute our results; see Methods for details.)

Current Mineral Exploration at Deer Trail
At the time of writing this manuscript, MAG Silver Corp. holds most of the land and claims around Deer Trail Mountain and Alunite Ridge. The company is actively exploring for carbonate replacement deposits (CRD) at the Deer Trail mine. They have completed an initial drilling campaign in which they confirmed the presence of the Redwall Limestone (a favorable lithology for the formation of CRDs), traced feeder structures to depth, and intercepted high-grade sulfide mineralization within host rocks at greater depths than was previously known [29]. All three of the holes that were drilled as part of this initial campaign bottomed out in "multi-phase porphyritic granitoids" [29], which may be part of one of the hypothesized porphyry stocks. The age of the porphyritic granitoid rocks has not yet been reported. A second phase of drilling is currently underway. We also note that geologists at MAG Silver Corp. have sampled "felsite" dikes from their property around Alunite Ridge and Deer Trail Mountain, and reported U/Pb and 40 Ar/ 39 Ar dates ranging from 31-28 Ma, though detailed analytical data are not yet available [30]. From publicly available images and limited access by the authors to hand specimens, these "felsite" dikes appear to have similar coloration, mineralogy, and textures to the quartz-porphyry dikes we have observed in the field at the Mineral Products, Christmas, and Bradburn mines.

Samples
Samples of vein-type magmatic-steam alunite were collected over multiple field seasons from veins exposed at four alunite mines and workings, and at a mine dump at the Close In landslide deposit, which slid from the top of Deer Trail Mountain and is currently located at the edge of the Sevier Valley to the northwest of the Deer Trail mine and on the eastern side of the Tushar fault [19].  Products mine for zircon U/Pb analyses. Samples of magmatic-steam alunite were collected from the L&N, Christmas, Edna, Mineral Products, and Close In mines/prospects, and sericite was separated from a sample collected from the Deer Trail mine workings for 40 Ar/ 39 Ar dating. Note, the location shown for the Deer Trail mine marks the portal to the workings, which extend ca. 3200 m to the west beneath Deer Trail Mountain [14]. The cross section from A to A' is shown in Figure 9.
Aliquots of all samples except ROR MV-1-06 and 002-05, including twelve individual aliquots extracted from seven of the bands in the alunite slab (M341), were analyzed using the incremental heating 40 Ar/ 39 Ar method at the U.S. Geological Survey (USGS) in Denver, CO. Aliquots of M939A, ROR MV-1-06, and 002-05 were analyzed using the single-grain total fusion (SGTF) 40 Ar/ 39 Ar method, also at the USGS in Denver. All aliquots of alunite were prepared for 40 Ar/ 39 Ar analysis by gently crushing the coarse-and fine-grained alunite specimens, washing in deionized water, and hand-picking fresh mm-and submm-sized fragments under a binocular microscope. A sample (M992) was also collected from the workings of the Deer Trail mine ( Figure 3, Table S3) from which sericite was separated by physically scraping mineral surfaces and interiors of vugs. Three separates of sericite were analyzed by the incremental heating 40 Ar/ 39 Ar method at the USGS in Denver.
A sample of a highly altered N30°W-trending rhyolitic quartz-porphyry dike (M945-R) was also taken from the upper Mineral Products mine at the same location as the alunite sample M945 ( Figure 3, Table S3) to determine the timing of this generation of intrusive activity relative to the hydrothermal system that produced the magmatic-steam alunite (and potentially produced the CRD at the Deer Trail mine [13,14]). This felsic dike shares one contact with the host rocks of the Three Creeks Tuff Member of the Bullion Canyon Volcanics [31], and one contact with the vein alunite that formed parallel to the rhyolitic dike. Sample M945-R was, crushed, pulverized, and sieved to obtain the 32-200 mesh fraction, which was then processed to concentrate zircon by density separation with methylene iodide (ρ = 3.32 g/cm 3 ). The "heavies" were briefly washed in acetone, followed by a final Frantz magnetic separation. The zircons were not air-abraded or acid-washed during preparation for U-Pb analyses. Approximately 60 grains were handpicked under a binocular microscope for inclusion onto an epoxy mount. The mount was ground to expose the approximate center of the grains, polished, and then imaged with transmitted light, reflected light, and cathodoluminesence (CL) to identify internal structures, zoning related to chemical composition, and defects. CL imaging ( Figure S65) was performed on the JEOL 5800 scanning electron microscope at the USGS Microbeam Laboratory, Denver. Nine zircon grains from M945-R were analyzed for U-Th-Pb isotopic values, and U-Th-Pb, REE, and Hf concentrations on the SHRIMP-RG at the Stanford-USGS Micro Analysis Center (SUMAC) at Stanford University.

40 Ar/ 39 Ar Geochronology Methods
Selected alunite fragments and sericite separates were loaded along with the neutron fluence monitor Fish Canyon sanidine (FCs) into precise positions within 18 mm-diameter Al disks, stacked, wrapped in high-purity Al foil and encapsulated under vacuum in quartz tubes. The quartz tubes were sealed into Al canisters with Cd shielding to reduce thermal neutron interactions and rotated at 1 rpm during six separate neutron irradiations (DT-2 and DT-4 each totaled 20 MWh; DT-6 totaled 10 MWh; DT-7 totaled 1 MWh; DT-11 totaled 2 MWh; DT-47 totaled 6 MWh; Table S4) in the central thimble position of the USGS TRIGA reactor [32]. Following irradiation, the alunite and sericite samples and grains of the FCs neutron fluence monitor were loaded with stainless-steel tweezers into a stainless-steel or copper sample holder and placed into a laser chamber with an externally pumped ZnSe window, which is attached to a custom-built ultra-high vacuum extraction line. For alunite analyses, pure silver wool was inserted into the exit tube of the sample chamber to react with any free sulfur gas liberated during sample heating. The volume of the mostly stainless-steel extraction line, including a cryogenic trap operated at −130 °C and two SAES GP50 getters (one operated at room temperature, one operated at 2.2A), is estimated at ~450 cc. A combination of turbo molecular pumps and ion pumps maintain steady pressures within the extraction line of <1.33 × 10 −7 Pa.
Individual alunite fragments were either incrementally heated or fused in a single extraction by controlling the power output of a 50 W CO2 laser equipped with a beam homogenizing lens, which results in uniform energy deposition over the entire sample surface. Aliquots of sericite were incrementally heated. During laser heating (for durations between 20 and 90 s), the gas evolved from the sample was exposed simultaneously to the silver wool (alunites only) and the cryogenic trap, and the gas was further purified in a second stage by additional exposure to the SAES getters for 120 s. For samples analyzed in 2009-2011, the sample gas was then expanded into a Mass Analyzer Products (MAP) 215-50 noble gas mass spectrometer in static vacuum mode. Isotopes were measured using an electron multiplier detector operating in analog mode by sequentially peak hopping over the mass range 36 to 40 for 10 cycles of data collection on each isotope. The MassSpec computer program written by A. Deino of the Berkeley Geochronology Center was used for data acquisition (v. 7.76). For samples analyzed in 2022, purified sample gases were expanded into a Thermo Scientific Argus VI noble gas mass spectrometer in static vacuum mode. Isotopes were simultaneously collected using four Faraday cups (for masses 40, 39, 38, and 37 on H1, AX, L1, and L2, respectively) and one compact discrete dynode (CDD) secondary electron multiplier detector (for mass 36). The Faraday detectors were intercalibrated by peak hopping a beam of 40 Ar over each detector using an automated routine in MassSpec (v. 7.91); the CDD was intercalibrated relative to the H1 Faraday cup using the Isotopic Reference Ratio Intercalibration (IR-IC) method [33] where the measured 40 ArH1/ 36 ArCDD ratio is calibrated to the atmospheric value of 298.56 ± 0.31 (1σ) [34].
All data reduction was performed with MassSpec (v. 8.159). Time zero intercepts were fit to the data (using parabolic, linear, and/or average best fits) and corrected for detector baselines, operational blanks (where the instrument was run as if to analyze a sample, but without firing the laser), mass discrimination (for peak hopping data collected on the MAP 215-50, monitored by analyzing air pipettes) or combined detector intercalibration/mass bias (for data multicollected on the Argus VI, using air pipette analyses and the IR-IC method of Turrin et al. [33]), radioactive decay of 37 Ar and 39 Ar, and nucleogenic interferences. Apparent ages for SGTF grains and steps in incremental heating experiments were also corrected for trapped atmosphere (see additional details below).
Laser fusion of individual FCs crystals at each position within the irradiation packages resulted in neutron flux ratios reproducible to ≤1.1%, with most ≤0.5% (2σ). Isotopic production ratios were determined from irradiated CaF2 and KCl salts; ( 36 Ar/ 37 Ar)Ca, ( 38 Ar/ 37 Ar)Ca, ( 39 Ar/ 37 Ar)Ca, ( 40 Ar/ 39 Ar)K, and ( 38 Ar/ 39 Ar)K values are given in Table S5. Cadmium shielding during irradiation prevented any measurable ( 40 Ar/ 39 Ar)K. 40 Ar/ 39 Ar plateau dates (and uncertainties) are considered the best estimate of the cooling age of the minerals and were calculated from samples if three or more consecutive heating steps released ≥ 50% of the total 39 ArK and that had indistinguishable apparent 40 Ar/ 39 Ar ages at 2σ [35]. In a few instances where the Fleck et al. [35] algorithm did not identify a plateau, we forced a plateau for exploratory purposes but prefer the integrated (total gas) dates in these cases. 40 Ar/ 39 Ar plateau and total gas (integrated) dates are reported in Table  1 (see Figures S1-S29), which were computed using an age of 28.201 ± 0.012 (0.046) Ma for FCs (2σ; [36]), the 40 K decay constants of Min et al. [37], and an atmospheric 40 Ar/ 36 Ar ratio of 298.56 ± 0.31 (1σ) [34]. Note, all K/Ar and 40 Ar/ 39 Ar dates quoted from the literature have been recalculated to use decay, isotopic, and monitor age parameters that are consistent with the Min et al. [37] decay constants and the age for FCs of Kuiper et al. [36] using the ArAR tool of Mercer and Hodges [38]. For the parameters used here, the effects of recalculating most published K/Ar and 40 Ar/ 39 Ar dates are well below the reported uncertainties, with age shifts of ≤0.2% and <0.005%, respectively (Tables S1 and S2). The only exceptions are for two K/Ar dates of sericite reported by Bassett et al. [23] and two K/Ar dates of biotite reported by Caskey and Shuey [24], which were shifted by ca. 2.80%-2.84% when recomputed with our preferred decay and isotopic parameters (Table S1).
Dates for SGTF analyses were pooled for each sample to compute an inverse variance-weighted mean (hereafter simply a "weighted mean"; Table 2, Figures S30-S32).
Where the fell outside of the 95% CI of the distribution, we performed outlier detection using an algorithm based on the Hampel identifier, which is based on deviations from the median rather than the mean (e.g., for more commonly applied Extreme Studentized Deviation or "2-sigma exclusion" identifiers) [39] (pp. 280-311). Here, we applied the Hampel identifier to the values (where and are the value and uncertainty of each date, respectively, and ̄ is the weighted mean), which are effectively the weighted residuals of the individual dates from the weighted mean. Thus, the analytical uncertainties of each date influence the identification of outliers, since relatively precise dates have a stronger effect on the value of the weighted mean than more imprecise dates. For each sample, we started with a permissive cutoff value of 6 (beyond which the OLi values computed using the Hampel identifier are considered potential outliers) and then moved in integer steps to lower, more aggressive cutoff values to identify the fewest number of outliers necessary to achieve an acceptable or until we reached a cutoff value of 2. (Note, in some cases relatively imprecise dates on the margins of a population can be retained by this process while more precise dates closer to the center of the distribution can be identified as potential outliers.)  We also computed inverse isochrons for each sample of alunite (Table S6, Figures  S33-S61). (We did not compute inverse isochrons for aliquots of Deer Trail sericite since all the evolved gasses were dominated by radiogenic 40 Ar.) Alunite data were excluded from York regressions (implemented following York et al. [41]) if they represented <2% of the total 39 ArK released or had 36 Ar signals that were indistinguishable from the background (one step from M341 G2). We consider SGTF weighted mean, plateau, and isochron dates to be acceptable if the value of the mean squared weighted deviation ( ; [42]) falls within the 95% CI of the , which is sensitive to the degrees of freedom (DOF) in the calculation [40]. In instances where the value of the falls outside the 95% CI of the , we have expanded the uncertainties by √ under the assumption that the uncertainties of individual steps should explain the scatter of data about the SGTF weighted means and plateaus (DOF = N − 1) or York regressions for inverse isochrons (DOF = N − 2).

SHRIMP U/Pb Geochronology Methods
Prior to SHRIMP-RG analysis, the epoxy mount with zircon grains from M945-R was acid-washed (1N HCl), rinsed (distilled water), dried, and coated with 10 nm of Au for maximum surface conductivity. Using a ~4-6 nA primary O2 ion beam, a 50 × 70 µm region was rastered for 120 s prior to each analysis. Sensitivity ranged from 5 to 30 cps per ppm Pb. Analysis time ranged from 14 to 16 min with the primary beam producing clean, sharp, 20 to 30 µm elliptical, 1 to 2 µm deep, flat-bottomed, analytical pits. Element concentration data were standardized against zircon standard MAD 4.6 (4200 ppm U; Madagascar green zircon) and are accurate to better than 10 percent, and age-data were standardized against zircon standard R33 (419.26 ± 0.39 Ma, 95% confidence level; from the Braintree Complex monzodiorite, Vermont; [43]), analyzed approximately every fifth analysis. Our foremost objective during the SHRIMP analyses was to measure zones of primary magmatic growth (i.e., to determine the crystallization age). However, we also measured zircon cores and rims (where identified with CL) to define possible ages of inheritance and metamorphic growth. The choice of location of an individual SHRIMP analysis was determined by factors including CL zoning pattern, absence of imperfections such as cracks or inclusions (liquid or solid), and purpose of the analysis (i.e., age of inheritance [source], igneous protolith, and/or metamorphic overgrowth). All analyses were conducted on a single epoxy mount during one analytical session, and no age inconsistencies were observed that could be related to instrument condition nor calibration.
Data reduction followed the methods of Williams [44] and Ireland and Williams [45], using SQUID 1 [46]. We computed a conventional inverse variance-weighted mean using the individual 206 Pb/ 238 U dates for individual spot analyses (using the methods applied to SGTF 40 Ar/ 39 Ar data). Both internal (analytical + 238 U decay constant) and external uncertainties (also including the age uncertainty of the standard R33) are provided for the weighted mean 206 Pb/ 238 U date. Rare-earth elements La through Yb and Hf were measured concurrently with the U-Th-Pb analyses as additional masses on each pass through the mass range. The concentration of U, Th, Hf and REE were calibrated using an in-house zircon standard MAD 4.6 and are reproducible at 2-4% (1σ), except for La (15%) because of its typically very low concentration (30 ppb). Pr was calculated from its neighbor elements because CeH (cerium hydrides) cannot be resolved from the Pr peak and contributes significantly to the counts at 141 Pr, given the low abundance of Pr and the relatively higher Ce concentration. Trace element concentrations were normalized to chondrite values of McDonough and Sun [47]. 40 Ar/ 39 Ar results for magmatic-steam alunite and sericite are shown as release spectra (Figure 4, Figures S1-S32) and probability density plots (PDPs, Figures S33-S35); threeisotope correlation (inverse isochron) plots of alunite are provided in the Supplementary Materials ( Figures S36-S64); inverse isochrons were not computed for M992 sericite aliquots (released gasses were dominated by radiogenic 40 Ar, denoted 40 Ar*). We report the 95% confidence interval (95% CI) for all ages and discussions below unless otherwise noted. External uncertainties, which include decay and monitor age uncertainties, are shown in parentheses immediately following the internal uncertainties. With few exceptions, the inverse isochron regressions yielded 40 Ar/ 36 Ar intercepts that are indistinguishable from modern atmosphere. Consequently, isochron dates (Table S6) overlap with the 40 Ar/ 39 Ar plateau dates summarized in Table 1. Detailed 40 Ar/ 39 Ar data are tabulated in Table S7.

Results
Samples of magmatic-steam alunite from the upper Mineral Products mine yielded plateau, total gas, and weighted mean dates ranging from 15.83 ± 0.26 (0.69) Ma to 15.445 ± 0.018 (0.59) Ma, and alunite from the Close In locality yielded 40 Ar/ 39 Ar plateau and weighted mean dates ranging from 15.60 ± 0.24 (0.66) Ma to 15.153 ± 0.058 (0.64) Ma. Our dates for the Mineral Products samples are statistically older than previously determined, but less precise, K/Ar ages (recalculated with decay constants of Min et al. [37]) of 14.3 ± 0.6 (0.7) Ma and 13.8 ± 0.6 (0.7) Ma (2σ) reported by Steven et al. [15] for different samples of vein alunite from the same locality. However, our Close In upper Mineral Products dates are comparable to the ca. 16-15.5 Ma range of dates reported by Waltenberg [27] and Ren and Vasconcelos [28] for multiple aliquots of a sample of vein-type alunite from Alunite Ridge. In addition, sericite from the Deer Trail mine yielded plateaus ranging from 15.643 ± 0.087 (0.64) Ma to 15.400 ± 0.090 (0.61) Ma. Ma. In addition, twelve individual aliquots of alunite were analyzed from seven distinct bands of sample M341 from the Christmas mine locality to test for apparent age variations ( Figure 5). Two aliquots, A2 and D2, yielded dates that are younger (13.3 ± 1.0 (1.2) Ma) and older (16.1 ± 5.5 (5.5) Ma), respectively, and relatively imprecise compared to dates from the remaining ten aliquots. Note that gas released from aliquot D2 was dominated by atmospheric argon (the median and mean of the 40 Ar* yields are 2.5% and 4.7%, respectively; Figure S10), which contributed significantly to the imprecision of the plateau date for D2. The ten aliquots of M341 excluding A2 and D2 yielded overlapping plateau (and isochron) dates ranging from 14.28 ± 0. 30 Figure 5). All alunite and sericite dates are shown in Figure 6.  Table 1 for external uncertainties of the individual plateau dates.) Two aliquots (A2 and D2) were excluded from a weighted mean of the remaining 10 aliquots that yielded a date of 13.98 ± 0.12 (1.1) Ma (95% CI). The uncertainty has been expanded by √ because the value of the (3.04) falls outside of the 95% CI of the ([0.3, 2.11]) for (N -1) = 9 degrees of freedom. Uncertainties are all plotted at 2σ. U-Th-Pb isotope data are summarized in Table 3 and illustrated in Figure 7 (see Figure S65 for CL images with analytical spot locations). Three analyses of xenocrystic cores of two zircon grains yielded 206 Pb/ 207 Pb dates in excess of 1700 Ma, while all other analyses (N = 10) of magmatic zones in 9 zircon grains yielded Oligocene 206 Pb/ 238 U dates. The conventional weighted mean computed using these 10 analyses is 30.70 ± 0.35 (0.45) Ma (95% CI, = 1.16; external uncertainty includes the error in the decay constant for 238 U and the age of the standard R33 [43]; Figure 7). Note that the 10 analyses of magmatic zones in the M945-R zircons have very similar 232 Th/ 238 U ratios with an average value of 0.60. Chondrite-normalized REE data (normalized using the composition of McDonough and Sun [47]; Figure S66), as well as Hf, U, and Th abundances are given in Table S8. All 40 Ar/ 39 Ar and SHRIMP-RG U-Th-Pb and trace element data are available in Mercer et al. [48].

Timing of Felsic Intrusive Activity Relative to the Formation of the Three Creeks Tuff Member of the Bullion Canyon Volcanics
The quartz-rhyolite porphyry dike and the Three Creeks Tuff Member of the broadly 34-22 Ma Bullion Canyon Volcanics on each side of the alunite vein at upper Mineral Products have been altered to quartz, alunite, and pyrite. Our weighted mean zircon 206 Pb/ 238 U date of 30.70 ± 0.35 (0.45) Ma for the rhyolite dike demonstrates that it was emplaced during the period of Oligocene calc-alkaline volcanism (Figure 7). The 31-28 Ma "felsite" dikes that have been dated by geologists from MAG Silver Corp. [30] are clearly also part of this period of Oligocene calc-alkaline igneous activity. However, there are conflicting dates published for the Three Creeks Tuff Member of the Bullion Canyon Volcanics. Plagioclase and biotite K/Ar dates published by Caskey and Shuey [24] and Steven et al. [15] (recalculated to use parameters consistent with those used here) range from 32.16 ± 0.70 (0.82) Ma to 26.95 ± 1.10 (1.16) Ma, and zircon fission track dates are 27.4 ± 1.3 and 27.0 ± 1.2 Ma [15,25]. Steven et al. [15] considered the older dates to be unreliable since they are older than the underlying Wah Wah Springs Tuff Member of the Needles Range Group, which they considered to be 28.9 Ma, and because the 27.5 ± 0.4 Ma date originally reported by Caskey and Shuey [24] was for a sample at the base of the Three Creeks Tuff Member. Rowley et al. [25] concurred with this assessment and considered a 27.1 Ma average of the younger 40 Ar/ 39 Ar and zircon fission track dates to be the best estimate for the age of the Three Creeks Tuff Member.
However, Steven et al. [15] and Rowley et al. [25] may not have accounted for differences in decay and isotopic abundance values, which can be significant (e.g., [38]). When recalculated using the parameters of Steven et al. [15], the date for the sample of the Three Creeks Tuff Member published by Caskey and Shuey [24] becomes 28.22 ± 0.41 (0.84) Ma, and when recalculated using our preferred parameters it becomes 28.27 ± 0.41 (0.84) Ma (Table S1). Both recalculated values are older than the ca. 27 Ma age interpreted by Steven et al. [15] and Rowley et al. [25] as the likely age of the Three Creeks Tuff Member, but are younger than our weighted mean zircon 206 Pb/ 238 U date for the rhyolite dike. While the 30.06 ± 0.10 (0.16) Ma weighted mean of 11 plateau dates reported more recently by Best et al. [26] (Table S2) for samples of the Wah Wah Springs Tuff Member is permissive of an older age for the overlying Three Creeks Tuff Member, it is also younger than our 206 Pb/ 238 U date for the rhyolite dike (though it overlaps with the young end of the distribution of SHRIMP-RG spot dates). Furthermore, a recalculated date of 31.45 ± 0.31 (0.87) Ma (Table S1) was reported by Caskey and Shuey [24] for the Wah Wah Springs Tuff Member, so the 28.9 Ma value that was used to dismiss older dates for the Three Creeks Tuff Member may be questionable. Regardless, since the rhyolite dike we sampled from the upper Mineral Products mine cuts what has been mapped as the Three Creeks Tuff Member (Figure 3) [31], it must be older than ca. 30.7 Ma or the rocks exposed at Alunite Ridge may in fact be older and miscorrelated with younger rocks in the region. We recognize our data set is limited to one dike, but the 31-28 Ma dates reported by geologists from MAG Silver Corp also indicate that older dikes are present cutting the Bullion Canyon Volcanics. Additional work is needed to refine geologic and geochronologic constraints for volcanic and intrusive rocks in the Alunite Ridge and Deer Trail Mountain area.

Timing and Spatial Distribution of Miocene Hydrothermal Activity at Alunite Ridge and Deer Trail Mountain
Magmatic-steam alunite from Alunite Ridge and Deer Trail Mountain near Marysvale preserves a record of at least two periods of magmatic degassing from ca. 15.7-15.1 Ma and from ca. 14.7-13.8 Ma (Figures 6 and 8). The older episode of magmatic-steam alunite formation is recorded at the Close In and Mineral Products localities, which are located to the east and southeast, respectively, of the other localities (Figures 3 and 8; recall that the Close In deposit is in a landslide that originated from Deer Trail Mountain). The 15.643 ± 0.087 Ma to 15.400 ± 0.090 Ma sericite dates for sample M992 from the Deer Trail mine provide additional evidence for this older period of hydrothermal activity and consequent alteration of the country rocks. In addition, the ca. 16-15.5 Ma dates published by Waltenberg [27] and Ren and Vasconcelos [28] from diffusion experiments on multiple irradiated aliquots of one sample of vein-type alunite support the timing of this initial episode of hydrothermal activity. Unfortunately, it is unclear precisely from where the sample these authors analyzed was collected, though its age suggests it was from the Mineral Products or Close In mines at Alunite Ridge. Steven et al. [15] did not recognize the earlier ca. 15.7-15.1 Ma pulse of magmatic degassing as the two samples they analyzed from the Mineral Products mine yielded younger dates of ca. 14.3-13.8 Ma. However, given the approximately 20 m width of this alunite vein, it is not surprising that both magmatic episodes are recorded in this single vein system.  Table S1 for external uncertainties). The location shown for the Deer Trail mine marks the portal; the workings extend ca. 3200 m to the west beneath Deer Trail Mountain [14].
The younger episode of magmatic degassing is clearly recorded at the Christmas and L&N mines and the Mt. Edna prospect. Given that 10 of the 12 aliquots (all M341 aliquots excluding A2 and D2) from seven alunite bands in the banded slab from the Christmas mine yielded a weighted mean of 13.98 ± 0.12 Ma (Figure 5), we interpret the bands to have completely formed within a period of ca. 250 ka (based on the 95% CI of the weighted mean). While the older apparent age of aliquot D2 of the banded slab (M341) can be explained by very low 40 Ar* yields, the younger apparent ages of aliquot M341 A2 and sample 003-05 from the Mt. Edna prospect are more enigmatic. Closure temperatures reported for both supergene and hypogene alunite, assuming a range of cooling rates from 5-100 °C/Ma, are consistently higher (≳200 °C) [28,49,50] than the 90-180 °C temperatures expected for the formation of magmatic-steam alunite [12,22]. Diffusion modeling also indicates that prolonged heating at the lower steam-heated temperatures would require excessive durations to cause any significant Ar loss (e.g., 100 Ma if alunite is held at 100 °C) [28]. In addition, the preservation within banded alunite of multicomponent gas signatures with contributions from crustal and meteoric sources (e.g., [21,51]) precludes significant thermal overprinting.
Therefore, there is no obvious natural mechanism for thermally driven Ar loss at Alunite Ridge, implying the younger apparent ages of M341 A2 and 003-05 either (1) reflect a later pulse of hydrothermal activity, (2) are due to Ar loss resulting from physical alteration/defects in the mineral structures of the analyzed aliquots of alunite, or (3) are simple analytical artefacts. Since the band from which M341 A2 was derived is situated between two other bands with ca. 14 Ma dates ( Figure 5), and given the large uncertainties associated with both M341 A2 and 003-05, we consider these dates to be unreliable (for either of reasons 2 or 3 above) and find it most likely that hydrothermal activity ceased by 14 Ma or shortly thereafter (within a few hundred ka). Furthermore, given the retentivity of alunite in the steam-heated environment and lack of evidence for subsequent thermal overprinting at Alunite Ridge, individual cm-scale bands of alunite are likely to accurately record their formation age. Thus, the relatively short (likely ≲250 ka) period over which the seven alunite bands developed in the slab sampled from the Christmas mine is consistent with a short recurrence interval of ≲36 ka for magmatic steam exhalation. Additional high-precision 40 Ar/ 39 Ar geochronology of individual bands in a more extensive sample of banded alunite (e.g., from the 20 m-wide vein at the upper Mineral Products mine), combined with robust statistical analysis (such as the methods employed by Morgan et al. [52]), could further refine constraints on the duration of each period of magmatic degassing at Alunite Ridge.

Relation to Regional Tectonic and Igneous Activity
The geologic model in which a concealed composite pluton was emplaced with one stock beneath Deer Trail Mountain and another beneath Alunite Ridge is supported by multiple lines of geologic and geochemical evidence [13,14,16,53]. Our 40 Ar/ 39 Ar results provide additional high-precision geochronologic support for this model. The 15.7-15.1 Ma dates for vein-type alunite and sericite alteration from the eastern Close In, Mineral Products, and Deer Trail mines may define an initial phase of magmatic degassing associated with initial intrusion of a porphyry stock beneath Deer Trail Mountain. This episode of activity may also be associated with the CRD deposit at the Deer Trail mine (e.g., [13,14]). Then, some 0.5-1.5 Ma later, a second phase of intrusion occurred beneath Alunite Ridge to form a second porphyry stock, and consequent hydrothermal activity produced the alunite veins at the Mineral Products, Christmas, L&N, and Mt. Edna mines over several hundred ka. These Miocene intrusions likely overprinted an Oligocene rhyolitic intrusion that was the source for the rhyolite dikes we observed at the Mineral Products, Christmas, and Bradburn mines (Figure 9).  [14]). See Figure 3 for the location of the cross section. An Oligocene rhyolitic stock is inferred to be the source of rhyolitic quartzporphyry dikes like the one dated at the upper Mineral Products mine and others at the Bradburn and Christmas mines. The two Miocene stocks (likely similar to those in Climax-type porphyry Mo systems) are inferred to be the sources for the magmatic volatiles that episodically produced magmatic-steam alunite during two periods at ca. 15.7-15.1 Ma and ca. 14.7-13.8 Ma. These periods of hydrothermal activity may have produced (or overprinted) the carbonate replacement deposit at the Deer Trail mine.
The timing of Miocene intrusive and hydrothermal activity at Deer Trail Mountain and Alunite Ridge is consistent with the migration of bimodal basalt-rhyolite magmatism across the Marysvale volcanic field (Figure 2) [18], and with broader patterns of Miocene bimodal igneous activity in southwestern Utah (e.g., [54]). Furthermore, the change around 22 Ma from calc-alkaline to bimodal basalt-rhyolite magmatic activity in the Marysvale volcanic field is consistent with the more regional compositional shifts within the Basin and Range Province that occurred during the late Oligocene to Miocene [55][56][57][58][59]. These shifts are thought to reflect the coincident evolution from subduction of the Farallon plate to subduction of a transform boundary of the East Pacific Rise along the west coast of California during the Miocene (e.g., [26,[55][56][57][58][59]). Prior to this shift in composition, the dominantly calc-alkaline magmatism that swept from north to south in the northern Basin and Range Province and from south to north in the southern Basin and Range Province was a consequence of the progressive removal of the Farallon plate that exposed the chilled and hydrated continental lithosphere to hot upwelling asthenosphere (e.g., [57][58][59][60][61]).
Within this tectonic context, the last calc-alkaline events to occur in the Marysvale-Deer Trail area were the eruption of the Monroe Peak caldera and emplacement of the Central intrusion at ca. 23 Ma (Figure 2), and the eruption of the Delano Peak Tuff Member of the Bullion Canyon Volcanics ca. 22 Ma [15,16]. Bimodal igneous activity then promptly followed as asthenospheric mantle drove melting in the hydrated and thinned continental lithosphere, migrating from the northeast towards the southwest in the Marysvale area ( Figure 2). As this activity terminated near present day Alunite Ridge, felsic magmas were generated in part through partial melting associated with intrusion of subcontinental lithospheric mantle-derived basalts into the crust. These felsic magmas episodically released supercritical fluids that rapidly ascended, variably entraining crustal volatiles and meteoric water, and finally precipitated magmatic-steam alunite [51].

Implications for Mineral Exploration
Published geologic, geochemical, and geochronologic data (e.g., [13,14]) and our 40 Ar/ 39 Ar results for magmatic-steam alunite and sericite from the mines and prospects at Alunite Ridge and Deer Trail Mountain suggest that the dominant period of hydrothermal activity was in the Miocene as a result of the intrusion of two porphyry stocks. However, our data do not preclude the possibility that Miocene intrusive and hydrothermal activity overprinted a previous Oligocene hydrothermal system associated with the emplacement of the rhyolite dikes in the Bullion Canyon Volcanics. Additional work is needed to conclusively determine the age, or ages, of the CRD in the Deer Trail mine.
While most porphyry Cu systems are located in continental and oceanic arcs and associated with subduction-related igneous rocks (e.g., Dilles and John [1] and references therein), the inferred stocks beneath Alunite Ridge and Deer Trail were emplaced in an extensional "postsubduction" tectonic setting that produced Climax-type porphyry Mo systems (e.g., [1,13,16]). From our results, the durations of individual periods of magmatic and associated hydrothermal activity (with episodic discharge of magmatic steam) at Alunite Ridge and Deer Trail Mountain are comparable to those observed in other Climaxtype systems for which high-precision geochronology data are available. For example, at the Questa porphyry Mo deposit (~400 kt MoS2), the total duration of mineralizing magmatism was ca. 770 ka with Mo mineralization lasting ≲500 ka [9,10]. The durations of mineralizing magmatism at the Pine Grove (~450 kt MoS2) and Victorio (~250 kt MoS2) porphyry Mo deposits are <2 Ma (e.g., [62,63]). In addition, the cumulative duration of the two periods of Miocene magmatic/hydrothermal activity at Alunite Ridge and Deer Trail Mountain is approximately 1.5 Ma. This is comparable to the durations of intrusive and hydrothermal activity at the Los Pelambres (ca. 2 Ma, 36 Mt Cu) and El Teniente (ca. 1.6-2 Ma, 95 Mt Cu) porphyry Cu deposits (see Figure S11 of Chelle-Michou et al. [5]).
Since the duration of intrusive activity exerts a first-order control on the endowment of porphyry Cu deposits [5], this might be cautiously viewed as an indicator that the concealed (and potentially relatively long-lived) Climax-type porphyry system below Alunite Ridge and Deer Trail Mountain may not be barren. At the least, the episodic release of magmatic S over the course of ~1.5 Ma of felsic magmatism makes the Alunite Ridge/Deer Trail Mountain area an appealing exploration target for porphyry-style sulfide mineralization. However, we caution that this is speculative, as the method of Chelle-Michou and coworkers [5] was developed for well-exposed and well-characterized porphyry Cu deposits in subduction-related tectonic settings. Furthermore, the link between the durations of magmatic/hydrothermal activity and the endowment of Climaxtype porphyry Mo deposits is not yet well-understood [62], and mineralization may be limited by the availability of external S (e.g., from mafic magmas; [64]) during the period when rapidly cooling (i.e., short-lived) felsic porphyries are releasing Mo-rich fluids (e.g., [9]). Regardless, 40 Ar/ 39 Ar geochronology to determine the age and duration of magmaticsteam alunite formation is clearly a valuable tool for identifying concealed porphyry systems in either subduction-related or extensional tectonic settings.

Supplementary Materials:
The following supporting information can be downloaded at: www.mdpi.com/article/10.3390/min12121533/s1, Supplementary Figures S1-S66: All 40 Ar/ 39 Ar release spectra, inverse isochrons, and probability density plots, as well as zircon CL images, and SHRIMP trace element figures; Table S1: Published K/Ar data for igneous rocks and vein and replacement alunite in the Marysvale area, Utah;

Data Availability Statement:
The data presented in this study are available in the Supplementary Material, and in an accompanying USGS data release by Mercer et al. [48] (https://doi.org/10.5066/P9IF4UZP).