The Questa porphyry Mo deposit in New Mexico provides a unique opportunity to study the relationship between pluton assembly and mineralization in a long-lived volcanic field. Magmatism along the caldera margin initiated at ~ 25.20 Ma and continued for ~ 770 ka. During this time, the emplacement of mineralizing intrusions progressed westward and culminated in the assembly of the Questa Mo deposit between 24.76 Ma and 24.50 Ma. Molybdenite Re/Os geochronology shows that mineralization occurred in multiple pulses without thermal resetting of the chronometer. Because most of the molybdenite samples used in this study are from previous fluid inclusion studies, we treat Re/Os molybdenite as a new thermochronometer. Molybdenite Re/Os ages are integrated with zircon U/Pb ages to evaluate the cooling histories within the Mo deposit. This study suggests that individual cycles of magma emplacement and mineralization cooled rapidly. In contrast to prior genetic models for the Questa Mo deposit, these data show that the mineralizing intrusions were generated via rapid melt generation, separation, and intrusion into the shallow crust without involvement in a long-lived magma chamber. It is proposed that the anomalously high magma flux event associated with ignimbrite eruption transferred materials (Mo, volatiles) from the upper mantle necessary for immediately subsequent mineralization. Partial melting and scavenging within a deep-crustal hybridized zone generated Mo-rich magma that ascended to form the Questa deposit. Moreover, this hypothesis predicts an important connection between caldera-forming systems and porphyry-style mineralization that could be incorporated into future exploration models.
 In 2010, the United States produced over $800 million of molybdenum oxide [USGS MCS, 2011], and the demand for Mo as an alloying agent in high-grade steel will likely increase as global economies recover. The majority of the world's Mo reserves are contained within porphyry Cu (± Mo) and porphyry Mo (± Cu) deposits. Although porphyry Mo deposits are often volumetrically smaller than porphyry Cu systems, they are still a significant source of Mo owing to higher ore grades.
 Understanding the timing of mineralization within a magmatic system is important for making accurate regional ore predictions. High-silica granite-related porphyry Mo deposits (Climax-type deposits) are often interpreted to form as cupolas above larger, more dynamic, magma chambers [Wallace et al., 1968; White et al., 1981; Carten et al., 1988a, 1988b, 1993; Lowenstern, 1994; Burnham, 1997; Klemm et al., 2008]. The depth and size of the magma body, as well as trends in chemical variation, remain topics of debate [Keith et al., 1986; Carten et al., 1993; Audétat, 2010]. Outstanding questions include: (1) How large is the magma chamber from which mineralizing cupolas originate? (2) At what depth in the crust do these magmas exist? (3) What is the mechanism for concentrating Mo?
 In a study of the chemistry, age, and tectonic setting of porphyry Mo deposits, Carten et al.  observed that mineralization typically occurs late within the lifespan of long-lived (> 1 Ma) magmatic provinces. For example, at Questa, New Mexico (Figure 1), Mo mineralizing magmas intruded along the southern margin of the Questa caldera and are interpreted to represent cupolas of an extensive late ring intrusion [Lipman, 1988]. Hydrothermal mineralization associated with ring faults is also observed in the Southern Rocky Mountain Volcanic Field in Colorado [Lipman, 2006] and at Round Mountain, Nevada [Henry et al., 1997] and the Borovista caldera in Bulgaria [Singer and Marchev, 2000].
 Climax-type deposits are thought to form during multiple magmatic-hydrothermal events that occur close in space and time [e.g., Wallace et al., 1968; White et al., 1981; Carten et al., 1988a]. Some models for Mo mineralization rely on simplified assumptions about the thermal history of plutons associated with the deposits. For example, Burnham  modeled the formation of porphyry ore deposits above plutons intruded as large bodies of mostly liquid magma with simple cooling histories. More recently, Seedorff and Einaudi [2004a] analyzed overprinted alteration assemblages along cross-cutting veins to derive a pseudocyclical thermal evolutionary model for the Henderson porphyry Mo deposit in Colorado. These authors note that unidirectional models (simple high- to low-temperature through time, [Lowell and Guilbert, 1970; Fournier, 1999]) are of limited use for understanding the complexities observed within these deposits.
 The suggestion that thermal cycling is important to development of Climax-type Mo deposits is consistent with recent studies that indicate plutons are assembled episodically and have more complex thermal histories than that of a single intrusion [e.g., Coleman et al., 2004; Seedorff and Einaudi, 2004a, 2004b; Davis et al., 2012]. However, studying the relationship between porphyry-style mineralization and sub-adjacent magma reservoirs is largely inhibited by the lack of exposure of cogenetic rocks. The Questa porphyry Mo system presents an opportunity to study this relationship because intrusions of varying volume and emplacement depths are exposed throughout the Latir volcanic field. Additionally, because the system is relatively young, abundant pre-, syn-, and (some) post-caldera volcanic rocks [Lipman et al., 1986] are preserved.
 Detailed geo-thermochronology allows evaluation of the link between magmatism and Mo mineralization at Questa. Combining zircon U/Pb with more recently developed molybdenite Re/Os geochronology allows for full characterization of the temperature-time (T-t) history of the Questa deposit and places the mineralization into the magmatic and thermal history of the system. Applying this approach: (1) lends insight into the timing and duration of magma emplacement relative to Mo mineralization, (2) characterizes how porphyry Mo systems relate to the evolution of a long-lived volcanic field and shallow pluton assembly, and (3) evaluates the applicability of Re/Os molybdenite as a new thermochronometer.
2 Geological Setting
 The Questa porphyry Mo deposit is located in the New Mexico portion of the Southern Rocky mountain volcanic field. It is part of the Latir volcanic field (~ 1200 km2) [Figure 1; Lipman et al., 1986]. Uplift and erosion along the eastern flank of the Rio Grande rift exposed a vertical section of volcanic strata and shallow plutons of the Latir field. Volcanism in the field began at 28.5 Ma with the eruption of dominantly andesite with local dacite and rhyolite [Zimmerer and McIntosh, 2012] and climaxed with the eruption of the ~ 500 km3 Amalia Tuff at 25.52 ± 0.06 Ma [Tappa et al., 2011].
 Exposed plutonic rocks at Questa span ~ 6 Ma of intrusive history (~ 25.6–19.1 Ma; Tappa et al., 2011; Zimmerer and McIntosh, 2012) and become younger to the south. Caldera margin plutons are intermediate in age among exposed plutonic rocks and are associated with mineralization [Meyer and Foland, 1991; Zimmerer and McIntosh, 2012]. From east to west, the caldera margin plutons include the Red River intrusive complex, Sulphur Gulch pluton, Southwest intrusive suite, and the Bear Canyon pluton. The Red River intrusive complex (~ 24.96 Ma; Zimmerer and McIntosh, 2012) is a high-silica rhyolite to aplite porphyry dike complex that cuts granodiorite and fine-grained monzonite [Lipman, 1988]. Epithermal Au-Ag veins occur near the contact of the Red River intrusive complex [McLemore and North, 1984]. The Sulphur Gulch pluton includes granite porphyry (~ 24.73 Ma, Zimmerer and McIntosh, 2012) and biotite-plagioclase porphyry phases. Drilling and cross-sections developed at the Questa mine site indicate that the two phases of the Sulphur Gulch pluton are connected at depth [SRK Consulting, Inc. report]. Also at depth, the Sulphur Gulch pluton is in contact with the Southwest intrusive suite, which includes granite, aplite, rhyolite, and latite associated with, and locally hosting, Mo mineralization [Ross et al., 2002]. The Bear Canyon pluton (~ 24.37 Ma, Zimmerer and McIntosh, 2012) is similar to the granitic phase of the Sulphur Gulch pluton and grades from aplite to granite porphyry.
 The roofs of the mineralizing intrusions at Questa dip at a low angle to the north and likely followed a pre-existing anisotropy [Ross et al., 2002] that may be related to the low-angle normal faults observed throughout the southern caldera margin [Meyer and Foland, 1991]. These faults are believed to have originally formed at high angles and were subsequently rotated to low angles above a batholith underlying the southern caldera margin [Meyer and Foland, 1991].
 The Questa porphyry Mo system is divided into three ore deposits: Spring Gulch, Central, and Log Cabin (Figure 1). The Central deposit is horseshoe shaped in map view and is the only one that is mined. Recent mining activity is underground along the southwest limb of the horseshoe (“Southwest zone”). Ore bodies are discontinuous along the Southwest zone and subdivided into the Vein zone, D ore body, and the Goat Hill ore body (Figure 2). The ore bodies within the Central deposit are mostly related to the emplacement of the Southwest intrusive suite. The Log Cabin ore body lies near the roof zone of the Bear Canyon pluton, whereas the Spring Gulch deposit lies above the Sulphur Gulch pluton. Major ore bodies at Questa are typically found at the intrusive contact with, or entirely within, pre-caldera volcanic units.
 The Questa ore system contains two separate styles of mineralization [Ross et al., 2002]. Both the Goat Hill and D ore bodies contain disseminated molybdenite within the matrix of a magmatic hydrothermal breccia. However, the majority of the ore is hosted within stockwork quartz + molybdenite veins. Ross et al.  subdivided the Goat Hill hydrothermal breccia on the basis of differences in matrix paragenesis, clast alteration, and breccia textures. Consequently, they observed five semi-stratified breccia facies (a through e, Figure 2) and interpreted them to represent evolution of the hydrothermal fluid away from the source. However, fluid inclusion work shows that there was little or no correlation between fluid evolution and the breccia facies [Rowe, 2012].
3.1 U/Pb Zircon Geochronology
 Samples were collected from units closely associated with epithermal mineralization (Red River intrusive complex) and the porphyry Mo deposit (the Sulphur Gulch pluton, Southwest intrusive suite, and Bear Canyon pluton; Figures 1c and 2). Samples were broken down with jaw crushers and a disc mill, and either processed on a water table, sieved, or both prior to heavy liquid and magnetic mineral separation.
 Zircons were thermally annealed in a 950 °C oven for 48 hours, and chemically abraded in HF + HNO3 at 220 °C for 12 hours [Mattinson, 2005]. The abraded zircons were spiked with a UNC in-house 205Pb-233U-236U tracer and equilibrated in HF + HNO3. In some cases, multigrain fractions were analyzed to counter relatively low U concentrations and the young age of these samples. After dissolution, fractions were dried and converted to chlorides. Anion exchange column chromatography was used to separate U from Pb. Uranium was loaded onto Re filaments. The majority of the U samples were loaded with graphite in order to ionize as a metal. However, several were ionized as UO2 late in the project in an attempt to raise ionization efficiencies. There are no systematic differences in analyses that could be attributed to how U was measured. Lead was loaded with Si-gel onto zone-refined Re filaments. All analyses were made on the Daly detector of the VG Sector 54 thermal ionization mass spectrometer at the University of North Carolina — Chapel Hill. In-run U fractionations were calculated based on the measured value for 233U/236U in the spike, and Pb fractionation was assumed to be 0.15%/amu. Analyses in which U was analyzed as an oxide were corrected for oxygen isotope interferences. Raw data were processed through Tripoli [Bowring et al., 2011] and U/Pb Redux [McLean et al., 2011]. Some analyses had high total common Pb (> 3 pg) that we suspect is derived from inclusions that were not removed during chemical abrasion (e.g., fluorite). Therefore, we assumed a maximum of 3 pg blank Pb and assigned all remaining common Pb to initial sample Pb. The isotopic composition of the sample common Pb was estimated using whole-rock data from Johnson et al.,  and Stacey and Kramers  where Pb isotope data were unavailable. All U/Pb ages are Th-corrected weighted mean 206Pb-238U ages. Correction for Th-disequilibrium was made using published whole-rock data for the plutons [Johnson et al., 1989]. The maximum Th correction was 0.085 Ma, and uncertainty in this correction is not significant for interpreting the results.
3.2 Re/Os Molybdenite Geochronology
 Molybdenite samples were collected from the Southwest zone of the Central deposit at Questa. We chose to date the five samples of Rowe  and two from Cline and Bodnar  because they determined the mineralization temperatures using fluid inclusion microthermometry for the Goat Hill and D ore bodies, respectively. The granite porphyry host for one of the samples from the Vein zone (QV11-02) was used for U/Pb zircon geochronology.
 Molybdenite was isolated with a diamond-tipped drill and chemically equilibrated with a mixed 185Re-188Os-190Os spike using the aqua regia Carius tube method [Shirey and Walker, 1995]. An attempt was made to select samples that did not contain a significant amount of pyrite, although the high Re and radiogenic Os concentrations in molybdenite overwhelm any non-molybdenite Re and Os. Osmium was extracted into chloroform (CHCl3) and then back extracted into HBr [Cohen and Waters, 1996]. Further Os purification was completed with micro-distillation [Birck et al., 1997] prior to being loaded onto Pt filaments with Ba(OH)2. Rhenium was isolated using Cl-based anion exchange column chromatography and loaded onto Pt filaments with Ba(NO3)2. Use of Ba(NO3)2 activators permits analysis as oxide anions (N-TIMS). Measurements were made at AIRIE (Colorado State University) on a Thermo-Fisher Triton TIMS using simultaneous Faraday cup collectors for Re and stronger intensity Os signals whereas weaker Os signals were collected in peak jumping mode with a secondary electron multiplier. Both Re and Os analyses were corrected for oxygen isotope compositions, and Os was further corrected for: (1) isobaric interferences (W, Re, and Pt), (2) common Os, (3) and mass fractionation based on the measured 190Os/188Os in the spike [Markey et al., 2003]. During data reduction, Re and Os abundances were corrected for analytical blank, although blank was negligible. Ages are reported with both 2σ analytical errors and combined analytical + 187Re decay constant uncertainty [Smoliar et al., 1996] errors (Table 2).
4.1 U/Pb Zircon Geochronology
 After Th-correction, all zircon U/Pb ages are concordant within analytical and decay constant uncertainties (Table 1 and Figure 3). Some samples show evidence for Pb-loss (e.g., JMR-34) which are identified where there are at least two outliers. A conservative age estimate was made for two samples from the Sulphur Gulch pluton (QM11-01 and JMR-6), the sample from the Bear Canyon pluton (JMR-34), and one from the Red River intrusive complex (MZQ-5) because they revealed a continuum of ages that could be related to inheritance, Pb loss and/or protracted zircon growth.
Table 1. U-Pb Data for Caldera Margin Rocks from the Latir Volcanic Field
 A granite porphyry (MZQ-5) and equigranular granodiorite (JMR-5) were collected from the southern and central portion of the composite Red River intrusive complex, respectively. Five zircon fractions from the granite porphyry yield a weighted mean 206Pb/238U age of 25.21 ± 0.055 Ma (N = 5; MSWD = 5.5). Four of the five fractions overlap within uncertainty. All four fractions analyzed from the granodiorite phase of the Red River pluton agree within uncertainty and define a weighted mean age that is identical to the granite porphyry (25.20 ± 0.036; MSWD = 0.79).
 A sample of the biotite-plagioclase aplite porphyry (sample QM11-01) of the Sulphur Gulch pluton was collected from within the open pit mine in the Central deposit. This sample yields a weighted mean age with statistically significant scatter (24.91 ± 0.069; N = 8; MSWD = 14). The second sample from the Sulphur Gulch pluton is a non-mineralized granite porphyry (JMR-6). The interpreted age for this sample is 24.44 ± 0.086 Ma (N = 9; MSWD = 11).
 A mineralized (pyrite and molybdenite) granite porphyry from the Vein zone orebody of the Southwest intrusive suite yields an age of 24.53 ± 0.045 Ma (MSWD = 8.3).
 Sample JMR-34 from the Bear Canyon pluton yielded zircons that were generally larger and more euhedral than those from more altered samples (e.g., Sulphur Gulch). Two fractions from this sample are significantly younger than the other analyses could possibly reflect Pb loss. However, we included all analyzed fractions, which yields weighted mean age of 24.46 ± 0.050 Ma (MSWD = 10.3).
4.2 Re/Os Molybdenite Geochronology
 Five vein-hosted and three breccia matrix molybdenite samples were analyzed from three separate ore bodies within the Central deposit (Goat Hill, D, and Vein; Table 2, Figure 4). All samples have moderate Re concentrations (< 70 ppm), and there is no correlation of Re concentration with mineralization style or ore body.
Table 2. Re-Os Data for Molybdenite of the Questa Porphyry Mo Deposit
Carius tube dissolution and sample-spike equilibration using a mixed-double Os spike Markey et al.  with sample weights 21–33 mg. All data are blank corrected, and Os isotopic measurements are fractionation corrected (using the double Os spike).
AIRIE run # refers to internal lab designation common to all AIRIE samples.
MHBX = Magmatic-hydrothermal breccia. Facies correspond to those of Ross et al. .
Percent molybdenite in drilled separate is based on a visual estimate under the binocular microscope. No other sulfides observed. Quartz-feldspar-biotite dilutant does not affect Re-Os age calculation.
Common Os is insignificant to the age calculation, and essentially zero for all analyzed molybdenites.
For MD-1254, MD-1255, MD-1256, Re blank = 24.22 ± 0.15 pg, total Os = 2.00 ± 0.02 pg with 187Os/188Os = 0.231 ± 0.001. For MD-1295, MD-1296, MD-1298, MD-1299, MD-1308, Re blank = 7.85 ± 1.48 pg, total Os = 1.86 ± 0.03 pg with 187Os/188Os = 0.322 ± 0.010.
Uncertainty on ages shown for both analytical error, and combined analytical and decay constant uncertainty for 187Re errors (0.31%, Smoliar et al., 1996).
Vein in aplite
Vein in MHBX facies A3
MHBX facies C
Vein in MHBX facies D
MHBX facies D
 The oldest molybdenite age obtained in this study is 24.76 ± 0.026 Ma from a quartz + molybdenite vein hosted in the aplite beneath the Goat Hill ore body. Two breccia matrix-hosted molybdenite samples yield analytically distinct ages of 24.61 ± 0.028 and 24.49 ± 0.043 Ma. Two vein-style molybdenite veins from within the same Goat Hill ore body yield ages of 24.59 ± 0.026 and 24.52 ± 0.028 Ma.
 Two samples from the D ore body yield ages of 24.61 ± 0.028 and 24.58 ± 0.027 Ma for vein and breccia matrix-hosted mineralization styles, respectively. One molybdenite vein that cuts granite porphyry within the Vein ore body was selected for both Re/Os molybdenite and U/Pb zircon geochronology. This sample yields a Re/Os age of 24.51 ± 0.032/0.085 Ma (uncertainties corresponding to analytical, and analytical with 187Re decay constant error, respectively) which is within uncertainty of the U/Pb zircon age obtained for the host granite (24.53 ± 0.052 Ma).
 Combining U/Pb zircon geochronology of intrusive rocks with Re/Os molybdenite ages from the Questa Mo deposit allows us to characterize the timing and duration of magma emplacement with respect to mineralization. Additionally, most of the samples used for Re/Os geochronology were analyzed for mineralization temperatures in earlier studies [Cline and Bodnar, 1994; Rowe, 2012], thus permitting combination of the new ages with temperature data to evaluate Re/Os as a thermochonometer. All the data can be considered within the framework of the evolution of the host Latir volcanic field to develop a model for porphyry Mo mineralization.
5.1 Intrusive History of the Caldera Margin Plutons
 The Red River intrusive complex is the oldest and easternmost of the caldera margin plutons. Equigranular granodiorite exposed in the central portion of the pluton was previously interpreted as the remnant magma chamber of a pre-caldera andesitic volcano [Lipman, 1988; Lipman and Reed, 1989]. However, the geochronologic data presented here indicate that the granodiorite crystallized approximately 300 ka after the eruption of the Amalia Tuff, nearly concurrently with high-silica granitic magma at ~ 25.20 Ma (Figure 5). However, field relations demonstrate that the Red River granite is part of a dike swarm that crosscuts the granodiorite [Lipman and Reed, 1989]. Therefore, although the units are the same age within uncertainty, the granite is demonstrably younger.
Johnson et al.  used major and trace element modeling to determine that the granodiorite within the Red River intrusive complex could not be a parental magma for the high silica aplite, thereby limiting the possibility that these magmas represent a single differentiated intrusion. Instead, the Red River pluton appears to have been assembled rapidly by intrusion of compositionally diverse magma pulses around 25.20 Ma.
 Following intrusion of the Red River intrusive complex, magmatism moved westward. New U/Pb zircon ages presented here suggest that the Sulphur Gulch pluton was assembled between 24.91 and 24.44 Ma with the biotite-plagioclase porphyry predating the granitic phase of the pluton. The granitic phases of the Sulphur Gulch pluton, Southwest intrusive suite and Bear Canyon pluton all intruded within uncertainty of each other at ~ 24.50 Ma. These high-silica granites were the primary contributors of metal-rich brines that resulted in breccia matrix-hosted and stockwork-style mineralization [Cline and Bodnar, 1994; Ross et al., 2002; Klemm et al., 2008; Rowe, 2012].
 Field mapping and gravity anomaly data indicate that the southern caldera margin plutons are upper portions of a partial ring intrusion along the caldera margin [Cordell et al., 1985; Lipman et al., 1986; Meyer, 1991]. The U/Pb zircon geochronologic data presented here indicate that assembly occurred over ~ 770 ka. We interpret the age range to represent a minimum assembly interval because thermal modeling demonstrates that crystallization of such a small magma volume would have occurred an order of magnitude more quickly [Spera, 1980]. This period corresponds to only a small interval of the ~ 6 Ma post-caldera pluton assembly history for the entire region [Tappa et al., 2011; Zimmerer and McIntosh, 2012], but brackets the time in which major magmatic hydrothermal systems were active [McLemore and North, 1984; Lipman et al., 1986; Lipman, 1988; Meyer and Foland, 1991; Ross et al., 2002]. These data are consistent with Zimmerman et al. , who observed that major porphyry-style mineralization episodes are short-lived relative to their host magmatic districts.
5.2 History of Mineralization
 Numerous occurrences of epithermal Au-Ag veins are present within the Red River intrusive complex and its wallrocks [McLemore and North, 1984]. New ages presented here therefore place a maximum age of 25.20 Ma on the epithermal-style mineralization. A younger age for mineralization cannot be ruled out; however, the epizonal textures associated with the pluton [Lipman, 1988] suggest that minor epithermal Au-Ag mineralization was synchronous with pluton assembly.
 Field evidence and fluid inclusion analysis [Rowe, 2012] demonstrate that the Mo mineralization at Questa involved multiple events and hydrothermal fluids. Major molybdenite mineralization at Questa was coeval with the emplacement of the Southwest intrusive suite, and the Sulphur Gulch and Bear Canyon plutons. The earliest mineralizing intrusion identified in this study is the biotite-plagioclase porphyry of the Sulphur Gulch pluton (24.91 ± 0.069 Ma) which is exposed in the open pit, just east of the Central ore body. The open pit mine, which is no longer active, targeted lode-style molybdenum mineralization [Ross et al., 2002] which probably received its metals from this intrusion. Generally, local ore grades are highest sub-parallel to the contact between the biotite-plagioclase porphyry and the pre-caldera andesite [B. Walker, per. comm.]. These field relationships suggest that this intrusion contributed ore to the system rather than acting as a host for later mineralization events.
 Molybdenite Re/Os ages record 250 ka of mineralization in perhaps three episodes (Figure 4). These data reflect the minimum duration of molybdenite mineralization because we targeted samples with independent temperature estimates rather than initial and final events. Seven of the eight molybdenite samples have ages between ~ 24.6 and 24.5 Ma, suggesting that major mineralization event occurred over this time interval. During this time period, both breccia matrix- and vein-style ore were deposited within the Goat Hill and D ore bodies [Ross et al., 2002]. These data demonstrate that the two distinct styles of mineralization were coeval in space and time for ~ 100 ka.
 The Goat Hill ore body alone was mineralized over a period of 250 ka, and molybdenite Re/Os ages become progressively younger upsection through its semi-stratified facies [Figure 2b; Ross et al., 2002]. These data could reflect time-progressive fluid evolution away from the source intrusion [Ross et al., 2002], but that requires the hydrothermal system to have been active for hundreds of thousands of years. This is unlikely given the short lifespan predicted by numerical models for an intrusion the size of the source (< 1 km3; Cathles et al., 1997; Ross et al., 2002). Ross et al.  noted that numerous dikes emanate from the roof of the aplite below the Goat Hill ore body. These intrusions may have been diachronous rather than extensions of a single intrusion, thereby allowing the possibility that the source aplite represents a series of amalgamated dikes. It is proposed that initial brecciation created an isotropic zone of weakness that re-fractured episodically during subsequent dike intrusions and fluid exsolution over a prolonged time period. Essentially, the Goat Hill hydrothermal breccia may have acted as a small trap that was easily shattered, thereby depressurizing small intrusions and favoring magma quenching and fluid exsolution [Candela, 1997].
Burnham  suggests that the presence of multiple populations of stockwork veins within porphyry-type systems could be related to the interplay between downward crystallization and repetitious crack-seal events within a single magma chamber. However, the total duration of Re/Os molybdenite ages observed in this study is too long for such a simplified process. It is proposed here that each mineralization stage corresponds to a complete cycle of magma injection, fluid exsolution, circulation, cooling, and mineralization. These data substantiate previous interpretations that Climax-type deposits form via complex multi-phase mineralization [Wallace et al., 1968; Seedorff and Einaudi, 2004a].
 In all, our data indicate syn-mineralization pluton emplacement over a period of about 770 ka (Red River intrusive complex through high-silica granite phase of Sulphur Gulch pluton) with early Au-Ag mineralization followed by at least 250 ka (and perhaps 400 ka if mineralization of the biotite-plagioclase porphyry was synchronous with intrusion) of Mo mineralization. Variable textures throughout caldera margin plutons with discrete intrusive contacts indicate at least a portion of the system was assembled by injection of numerous small intrusions. These data are consistent with growing body of evidence for incremental pluton assembly [e.g. Coleman et al., 2004; Matzel et al., 2006; Davis et al., 2012; Tappa et al., 2011; Leuthold et al., 2012], and links this concept with models for episodic porphyry mineralization [Wallace et al., 1968; Carten et al., 1988a; Maksaev et al., 2004; Seedorff and Einaudi, 2004a; Wilson et al., 2007].
5.3 Re/Os Molybdenite: A New Thermochronometer
 If molybdenum mineralization is demonstrably related to magma emplacement, such as in porphyry ore systems, the mineralization temperature can be used to help construct a system-wide thermal history. Fluid inclusion studies reveal that mineralization within porphyry systems occurs at temperatures between ~ 300 and 500 °C, with modes typically ~ 400 °C [Cline and Bodnar, 1994; Selby et al., 2000; Klemm et al., 2008; Rowe, 2012]. Because thermal modeling [e.g., Hanson and Glazner, 1995; Yoshinobu et al., 1998] demonstrates that incrementally assembled plutons “dwell” in the temperature window between hornblende and biotite Ar closure temperatures [~ 525 °C and ~ 325 °C, respectively; Harrison, 1981; Harrison et al., 1985], Re/Os molybdenite data can fill an important gap in our understanding of these systems [e.g., Stein et al., 2001; Markey et al., 2003].
 Most of the molybdenite separates used in this study are extracted from the same samples that were used in fluid inclusion studies, thereby allowing correlation of microthermometry results and Re/Os molybdenite ages. For the Goat Hill ore body, Rowe  determined a primary mineralization stage to occur at 380 °C, with another significant stage at ~ 280 °C. Rowe  observed evidence for halite trapping (e.g., multiple halite crystals in one inclusion), and as a result she reported liquid-vapor homogenization temperatures. This is in contrast to the studies from the D ore body [Cline and Bodnar1994; Klemm et al. 2008], for which halite homogenization temperatures are reported. Both of these studies determined mineralization temperatures to be > 410 °C for the D ore body. Because halite typically dissolves at a higher temperature than liquid-vapor homogenization in fluid inclusions from porphyry ore systems [e.g. Becker et al., 2008], the difference in temperature estimates between the Goat Hill and D ore bodies is likely minimal. Consequently, we use an intermediate temperature of ~ 400 °C for our discussion.
 The dispersion of Re/Os molybdenite ages presented here is interpreted to reflect episodic mineralization events. Therefore, chronologically later mineralization episodes with temperatures on the order of 400 °C did not disturb the Re/Os system in molybdenite. This observation is consistent with previous studies where Re/Os molybdenite ages were not reset by high-grade metamorphism [Stein et al., 1999, 2001, 2003; Stein and Bingen, 2002; Stein, 2006], and discrete Re/Os molybdenite ages that are concordant with U/Pb zircon ages from associated intrusions [Selby et al., 2007; Maksaev et al., 2004]. These data also support previous work that predicts the closure temperature of the Re/Os system in molybdenite to be > 400 °C [Suzuki et al., 1996]. As a result, Re/Os molybdenite is considered a useful new thermochronometer in geologic settings where molybdenite is the primary sulfide phase.
5.4 Integrated T-t History of the Questa Porphyry System
 Combination of new U/Pb and Re/Os ages allows evaluation of the T-t history of the caldera margin plutons and mineralization. Uncertainties in ages in this section include propagation of full analytical uncertainties and 238U and 187Re decay constant uncertainties. The geochronologic data place magmatic-hydrothermal mineralization along the Red River valley into a period of relatively low magma flux that existed following the eruption of the Amalia Tuff [Tappa et al., 2011]. Additionally, the timing of Mo mineralization is bracketed by the zircon ages of spatially associated plutons. Individual cycles of magma emplacement, fluid exsolution, and mineralization were short-lived from zircon saturation (~ 750 °C) through molybdenite mineralization (~ 400 °C). Consider sample QV11-02 from the Vein ore body (Figures 2 and 6a). Zircon crystallization within this sample occurred at 24.53 ± 0.052 Ma, and the pluton is cut by a molybdenite vein, dated at 24.51 ± 0.085 Ma. It is possible that molybdenite from the vein was not sourced from the same intrusive rock that was sampled; regardless, these data limit the duration of time between zircon crystallization and brittle-vein formation to have occurred within uncertainty of the ages, consistent with rapid cooling.
 Acknowledging these rapid cooling rates allows for a general interpretation of thermal cycling within the system (Figure 6b). Note, however, that our sampling strategy targeted samples where independent determinations of temperature were available and therefore may exclude higher frequency events. Identifying more equivalent U/Pb zircon and Re/Os molybdenite ages, such as those demonstrated for sample QV11-02, would strengthen this hypothesis. However, our sampling strategy was largely limited to samples previous fluid inclusion studies [Cline and Bodnar, 1994; Rowe, 2012], many of which do not contain enough material for U/Pb zircon analysis. Regardless, the combination Re/Os molybdenite and U/Pb zircon geochronology for sample QV11-02 demonstrate that, where available, such relationships can be identified. Thus, we infer multiple thermal cycles over 250 ka wherein each hydrothermal circulation event corresponds to a specific intrusion that cools rapidly [Cathles et al., 1997; blue line, Figure 6b].
 Evidence for rapid cooling from igneous through hydrothermal conditions is consistent with numerical models for shallow magmatic systems such as the one responsible for mineralization at Questa [3–5 km at time of emplacement; Cline and Bodnar, 1994; Cathles et al., 1997]. These data add to a growing body of evidence that suggest porphyry ore systems are characterized by multiple injections of magma where each pulse creates a hydrothermal circulation cell that remains active for < 100 ka [Marsh et al., 1997; Maksaev et al., 2004; Lawley et al., 2010; Braxton et al., 2012]. However, this study is the first to provide the timing and duration of multiple cycles for a Climax-type porphyry Mo system, and we suggest that individual cycles are possibly active for very short times (within uncertainty of the chronometers in this study).
5.5 Evaluation of Previous Genetic Models for the Questa Mo Deposit
 Previous thought on the origin of the Questa porphyry Mo deposit calls upon (1) chemical differentiation processes (assimilation, crystal fractionation, volatile fluxing) in a long-lived magma chamber within the middle to upper crust [Johnson et al., 1989; Carten et al., 1993] or (2) fractional crystallization in the lower crust [Klemm et al., 2008]. The mineralizing intrusions in these cases are believed to be cupolas that emanate from these long-lived reservoirs. Detailed geo-thermochronology presented here challenges these hypotheses.
 Agreement of U/Pb zircon ages with Re/Os molybdenite geochronology for the suite of high-silica granites precludes the possibility that these intrusions resided in a long-lived upper crustal magma chamber [e.g. Johnson et al., 1989; Carten et al., 1993]. Any fractionating chamber would have been below zircon saturation temperatures (750° to 780 °C) and crystallizing zircon [Johnson et al., 1989]. This predicts a gap between zircon saturation and cooling coincident with intrusion and mineralization (Figure 7a). Mafic recharge into the base of the system could reheat it above zircon saturation temperatures [e.g. Carten et al., 1993]; however, this requires a significant volume of mafic magma that is unrecognized in surface and subsurface data [Lipman and Reed, 1989; Meyer, 1991; 2004SRK Consulting, Inc. report].
 Alternatively, in situ magma differentiation could take place in the deep crust where ambient temperatures are higher. Klemm et al.  suggested that ~ 95% crystallization of a middle to lower crustal magma chamber is required to explain the rapid increase in Cs concentration within hydrothermal fluids between breccia matrix-style and vein-style mineralization. However, the Re/Os molybdenite ages in this study clearly show that these mineralization styles were concurrent over a ~ 100 ka period. Therefore, the data of Klemm et al.  cannot represent a time-progressive geochemical trend.
 The repetitious intrusion of mineralizing granitic magmas and the synchroneity of high- and low-temperature thermochronometers favors an evolution model in which the magmas are generated, ascend, and cool quickly (Figure 7b). We propose that, like other magmas associated with the Latir field, the origin of the mineralizing magmas was the lower crust [Johnson et al, 1990; Tappa et al., 2011].
5.6 Relationship Between Caldera Formation and Mineralization
 An intriguing set of observations from the detailed geochronology now available for the Latir volcanic field is that high-silica granites capable of mineralization were emplaced within only 0.25–0.40 Ma of the total 8.5 Ma history of the Latir field (~ 28.5 to 19.1 Ma; Zimmerer and McIntosh ) and that this short episode of mineralization occurred immediately after the eruption of the caldera-forming Amalia Tuff (~ 25.52 Ma, Tappa et al. ). Numerous authors discuss the relationship between large-volume calderas and their control on mineralization [e.g., Lipman, 1992; Rytuba, 1994; Henry et al., 1997; Singer and Marchev, 2000]; however, the majority of the discussion is centered on structural controls related to caldera subsidence (e.g. ring faults) and rarely on the relationship between ignimbrite generation and mineralization.
 Recent studies of volcanic-plutonic connections focus on magma flux variations responsible for ignimbrite and non-ignimbrite stages of volcanic field evolution [Glazner et al., 2004; Lipman, 2007; Annen, 2009; Tappa et al., 2011]. Although there are fundamental differences in interpretation among these papers, all agree that an unusually high flux of material (melt, fluid, gas) is required for formation of an ignimbrite as large as the Amalia Tuff (500 km3). Most models also focus on the role of a high flux of material from the upper mantle to either remobilize an existing magma body [Bachmann and Bergantz, 2003; Huber et al., 2012] or initiate deep crustal melting [Glazner et al., 2004; Tappa et al., 2011].
 It is hypothesized here that the period of high magma flux responsible for generating the Amalia Tuff set the stage for generating mineralizing intrusions at Questa (Figure 8). A high flux of juvenile mafic rocks may have exsolved hydrothermal fluids during crystallization and created a zone of hybridized crust (juvenile mafic rocks + old lower crust + hydrothermal assemblages). Stein  and Stein and Hannah  analyzed sulfur isotopes from molybdenite and concluded that all of the sulfur in porphyry Mo systems is derived from the intrusions. Observing the same data, Pettke et al.  emphasize that the 34S values are consistent with a mantle source. Moreover, Pettke et al.  used radiogenic Pb-isotopes to demonstrate that the mantle source was previously metasomatized, possibly during Proterozoic continental accretion. The high mafic magma flux required for formation of the ignimbrite rapidly transferred unusually high masses of material, including sulfur, into the lower crust as juvenile mafic rocks and hydrothermal fluids. Although mafic underplating throughout the history of the Latir field is hypothesized [Johnson and Lipman, 1988; Johnson et al., 1989, 1990] and thus, volatiles were presumably consistently transferred from the mantle to the crust, the unusually high flux necessary for formation of the ignimbrite may have prepared the system for economic mineralization.
 Partial melting of the hybridized zone immediately following eruption of the Amalia Tuff produced granitic magmas capable of mineralization. Geochemical and experimental studies show that partial melting of hydrous amphibolites (~ lower crust) can create appreciable volumes of granitic magma [Faure and Powell, 1972; Mahood and Halliday, 1988; Ratajeski et al., 2005]. Variable contributions from older lower crust, juvenile mafic rocks, and hydrothermal assemblages may explain slight compositional variations within productive intrusions at Questa (e.g., biotite-plagioclase porphyry [QM11-01] vs. quartz-alkali feldspar porphyry [QV11-02]). Metal endowment within the caldera margin plutons likely comes from a combination of (1) metal-rich lower crust source rocks and (2) release of volatiles during crystallization of underplating mafic magmas. These volatiles scavenge the metals and induce melting of the hybridized lower crust. According to this model, the halogen-rich volatilization that is typically inferred in porphyry Mo systems occurred during melting, rather than crystallization in an upper crustal magma reservoir [Hildreth, 1981; Keith et al., 1986; Carten et al., 1993; Figure 7b]. Moreover, this model can account for the observed components of both juvenile mafic rocks and much older (> 1 Ga) lower crust components within porphyry Mo magmas inferred from radiogenic isotope studies [Farmer and DePaolo, 1984; Stein, 1985; Pettke et al., 2010].
 The proposed model for hybridized source rocks is similar to models proposed by Richards [2009, 2011]. However, our model emphasizes the relationship of porphyry Mo deposits to silicic ignimbrite systems, and not necessarily arc or back-arc tectonic regimes. Regardless, the post-ignimbrite mineralization observed in this study suggests that metasomatic preparation of the lower crust may be a critical step in generating productive intrusions.
 This model implies the general possibility that porphyry Mo mineralization follows ignimbrite eruption [e.g., Lipman, 2007]. The world's largest Climax-type deposits in Colorado are located close in space and time to the Eocene-Oligocene flare-up [Chapin, 2012] and the Pine Grove deposit in Utah immediately post-dates a small ignimbrite eruption that ended an ~ 4 Ma hiatus from the voluminous (~ 10,000 km3) Indian Peak caldera complex [Best et al., 1989; Keith et al., 1986]. Drilling near the western rim of the ~ 1 Ma Valles caldera identified a shallow zone of molybdenum mineralization that is interpreted to be related to Climax-type mineralization at depth [Hulen et al., 1987]. Climax-type mineralization has been inferred from elevated F contents in soils along the partial-ring intrusion of the Oligocene Bonanza caldera in Colorado as well [Lipman, 2007; Rose and Pride, 2010]. These observations argue strongly for a genetic relationship between caldera-forming eruptions and Mo mineralization wherein a major hybridization event provides unusually high masses of readily mobile metals and “prepares” the lower crust for subsequent melting events that may produce productive magmas. Consequently, future exploration models should perhaps target caldera systems where post-ignimbrite intrusions can be identified. Additional detailed geothermochronology such as that presented here should be done to evaluate this hypothesis.
 New U/Pb zircon geochronology from the Latir volcanic field shows that mineralizing magmas were emplaced along the southern margin of the Questa caldera margin over 770 ka. Epithermal Au-Ag mineralization near the Red River intrusive complex, a possible source for mineral occurrences, post-dates caldera formation by at least 300 ka. Molybdenite Re/Os ages show that the Questa porphyry Mo deposit was assembled episodically over 250 ka between ~ 24.76 and 24.50 Ma, and the bulk of the mineralization was synchronous with the emplacement of suite of high-silica granites.
 This study demonstrates that Re/Os molybdenite geochronology can be used as a thermochronometer in cases where temperature can be determined independently. These data provide further evidence that the Re/Os chronometer remains closed to diffusion during repeated hydrothermal circulation events with temperatures in excess of ~ 400 °C. These data are used to show the complex thermal history of the incrementally assembled Questa porphyry Mo deposit.
 Detailed geo-thermochronology of the Questa Mo deposit is inconsistent with previous genetic models for the system that predict magma residence time in a long-lived system or simple evolution of mineralizing intrusions via fractional crystallization. The fact that Mo mineralization post-dates the eruption of the Amalia Tuff suggests a model wherein anomalously high magma flux mobilizes metals and sulfur from the mantle and lower crust. Scavenging and partial melting of the newly formed hybridized zone is hypothesized to create the magmas responsible for mineralization. This new model suggests that immediate post-ignimbrite plutons within calderas should be considered an important target for mineral exploration.
 We thank Amanda Rowe and Jean Cline for providing molybdenite samples from their personal collections. This project would not have been possible without the help of Bruce Walker and Chevron Mining Inc., who provided access to the mine property and advice throughout the project. Funding was provided by the National Science Foundation (EAR: 1050215), Geological Society of America, Sigma Xi Grants-in-Aid of Research, and the UNC Martin Fund. Thanks to Jez Inglis and Aaron Zimmerman for providing lab work and insight. Helpful reviews were provided by Eric Christiansen and an anonymous reviewer.