Helium and Argon Thermochronometry of the Gold Butte Block, South Virgin Mountains, Nevada

Earth and Planetary Science Letters 178 (2000) 315^326 www.elsevier.com/locate/epsl

Helium and argon thermochronometry of the Gold Butte block, south Virgin Mountains, Nevada

Peter W. Reiners a;*, Robert Brady a, Kenneth A. Farley a, Joan E. Fryxell b, Brian Wernicke a, Daniel Lux c

a Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA 91125, USA b Department of Geological Sciences, California State University, San Bernardino, CA 92407, USA c Department of Geological Sciences, University of Maine, Orono, ME 04469, USA Received 19 November 1999; received in revised form 13 March 2000; accepted 13 March 2000

Abstract

One of the largest exposures of Precambrian crystalline rock in the Basin and Range province of the southwestern USA is the Gold Butte block of the south Virgin Mountains, about 15 km west of the Colorado Plateau. It has been interpreted as a largely continuous crustal cross-section about 15^20 km thick that was exhumed by a deeply penetrating normal fault during Miocene extension. To test this interpretation as well as the use of the newly developed titanite (U^Th)/He thermochronometer, we examined the low temperature thermal history of the Gold Butte block with the apatite and titanite (U^Th)/He and muscovite 40Ar/39Ar thermochronometers. Apatite He ages average 15.2 þ 1.0 (2c) Ma throughout the block, indicating that the entire section was warmer than 70³C prior to Miocene exhumation. Titanite He ages increase from 18.6 þ 1.5 Ma near the paleobottom (west) end of the block, to 195 þ 15 Ma near the paleotop (east) end. A rapid change from mid-Tertiary to increasingly older titanite He ages to the east is observed at about 9.3 km paleodepth, and is interpreted as a fossil He partial retention zone for titanite. Assuming a pre- exhumation geotherm of 20³C/km (consistent with earlier apatite fission track work), this depth would have corresponded to 196³C prior to exhumation, indicating that laboratory-derived He diffusion characteristics for titanite that yield a closure temperature of about 200³C are applicable and correct. Muscovite 40Ar/39Ar ages are 1.0^1.4 Ga near the paleotop of the block, and 90 Ma near the paleobottom. Together with 207Pb/206Pb ages on apatite and titanite, and an earlier apatite fission track transect across the Gold Butte block, our data indicate that the continental crust at the western edge of the Colorado Plateau resided at moderate geothermal gradients (and slowly declined in temperature) from 1.4 Ga to about 100^200 Ma. A 90 Ma cooling event clearly affected the mid-crust (deepest portions of Gold Butte), which may reflect accelerated cooling or a brief heating and cooling cycle at this time, after which gradients returned to about 20³C/km prior to rapid exhumation in the Miocene. This work thus supports previous structural and thermochronologic studies that suggest that the Gold Butte block is the thickest largely continuous cross- section of crust exposed in the southwestern USA. ß 2000 Elsevier Science B.V. All rights reserved.

Keywords: thermochronology; Basin and Range Province; exhumation; helium; Ar/Ar; extension

* Corresponding author. Present address: Department of Geology, Washington State University, Pullman, WA 99164, USA. Tel.: +1-509-335-3009; Fax: +1-509-335-7816; E-mail: [email protected]

EPSL 5439 10-5-00 316 P.W. Reiners et al. / Earth and Planetary Science Letters 178 (2000) 315^326

1. Introduction pressure estimates from a variety of thermoba- rometers, (2) abundance and size of cumulate phe- The Gold Butte block is a large exposure of nocrysts in the large Gold Butte granite pluton, Precambrian crystalline rock in the south Virgin and (3) deformation style, from brittle faulting Mountains of southeastern Nevada, about 15 km through brecciation and mylonitization [3]. Some west of the largely undeformed and £at lying sedi- workers, however, have used sedimentologic and mentary rocks of the Colorado Plateau (Fig. 1). erosional evidence to argue that most of the Pre- The block is bounded on the east by steeply dip- cambrian rock in the Gold Butte block is £at-ly- ping (50^65³) Phanerozoic sedimentary rocks sim- ing upper crust that was not strongly tilted or ilar to those of the Grand Canyon, on the north exhumed from deep levels during Miocene exten- by a large-displacement tear fault, and on the sion [4]. Resolving between these two contrasting west and south by unconformably overlying Ter- interpretations is of critical importance for at least tiary sediments [1]. Structural, petrologic, and two reasons. First, the former interpretation im- thermobarometric lines of evidence have been plies that the Gold Butte block is one of the larg- used to argue that the block represents a 15^20 est intact crustal sections in the Cordillera, and km thick section of Proterozoic crust that was thus provides valuable insights into the petrologic, uplifted, tilted to the east, and exhumed, by thermal, and structural evolution of the crust in deeply penetrating west-dipping normal faulting, the region of the Colorado Plateau. The latter exposing a thick continuous section through the suggests that large regions of crystalline mid-crus- mid-crust [1^3]. Evidence for the largely continu- tal rocks were at shallow levels in the Gold Butte ous nature of the pre-exhumation crustal section area prior to Miocene extension. Second, the ¢rst includes systematic east-to-west changes in: (1) model implies deep exhumation and high angles

Fig. 1. Location and generalized geologic map, after [3], of the Gold Butte block. Numbers beside (U^Th)/He and 40Ar/39Ar sample locations refer to sample numbers in Tables 1^3. Samples dated by apatite FT by Fitzgerald et al. [4] were collected along the northern and western margin of the block. The locations of these and the Ar-dated samples have been projected onto the line of the (U^Th)/He samples in Fig. 2. Because of the potential for north^south tilting of the block, the projected paleodepths of the FT samples may not be exactly the same as the He samples. `MHWF' is the Million Hills Wash fault that may disrupt the pre-exhumation structural sequence between samples 18 and 20 (see text for details).

EPSL 5439 10-5-00 P.W. Reiners et al. / Earth and Planetary Science Letters 178 (2000) 315^326 317 of tilting of exhumed crustal blocks, suggesting depths less than 5 km (which show a fossil apatite large amounts of extension in this portion of the partial annealing zone), Gold Butte is a relatively Basin and Range, while the second model implies intact crustal section that was rapidly exhumed at only limited or localized exhumation and exten- about 15^16 Ma. sion in this region. A number of recent studies have shown that Crystalline rocks in the interior of the Gold apatite (U^Th)/He thermochronometry provides Butte block include metamorphic rocks (primarily reliable cooling ages through temperatures of orthogneiss and garnet gneiss), and the intrusive about 70³C [11^20]. Most applications of this Gold Butte granite, a large pluton of rapakivi technique have focused on age^elevation relation- granite [3,5]. Previous Rb/Sr and U/Pb dating in- ships in vertical transects of mountain ranges to dicates that the gneissic lithologies have ages of constrain the timing and rate of exhumation. This about 1.7^1.8 Ga [6], while the 1.45 Ga Gold study instead uses (U^Th)/He thermochronometry Butte granite is representative of the widespread to characterize the timing, rate, and structural 1.4^1.5 Ga anorogenic magmatism at this time [7]. style of exhumation of extremely tilted, now More recent magmatism in the lower part of the sub-horizontal, crustal sections. This study also block has also been documented by 64^66 Ma represents the ¢rst application of the newly devel- monazite U/Pb ages from a two-mica granite in oped titanite (U^Th)/He thermochronometer the western portion of the block [1,8]. [21,22], which, according to laboratory di¡usion Previous ¢ssion track (FT) studies indicate late measurements, provides cooling ages correspond- Cretaceous through Miocene cooling in the Gold ing to closure temperatures (Tc) of about 200³C Butte block. A single biotite FT analysis by Vol- (for a cooling rate of 10³C/Myr). Together with borth [5] yielded an age of 84.7 þ 0.2, and zircon apatite He ages and muscovite 40Ar/39Ar dates and apatite samples from the central portion an- that provide cooling ages corresponding to about alyzed by Parolini et al. [9] yielded FT ages of 350^425³C [23], titanite He ages in Gold Butte 84.9 and 27.2 Ma (no errors given) respectively. provide an important test of the closure temper- In the most detailed thermochronologic study of ature of this newly developed chronometer. the block to date, Fitzgerald et al. [10] showed We collected and analyzed apatite, titanite, and that apatite FT ages from samples below about muscovite samples from a transect running 5 km paleodepth (assuming an overlying sedimen- through a roughly paleo-vertical section of the tary thickness of 2.5 km prior to exhumation) Gold Butte block (as suggested by the continuous were essentially invariant at about 14^17 þ 1^3 crustal section model) for (U^Th)/He, 40Ar/39Ar, Ma, whereas ages of samples towards the east and 207Pb/206Pb dating. Age^paleodepth relation- (decreasing paleodepth) increased to about ships for each of these systems provide long-term 50 þ 3 Ma near the unconformity. Fitzgerald et thermal histories throughout di¡erent regions of al.'s work supports previous structural and ther- the block that are consistent with the model of mobarometric evidence that, at least at paleo- Gold Butte as a largely continuous section of Pre-

Table 1 Muscovite 40Ar/39Ar ages of Gold Butte samples Sample Paleodepth Total gas age Plateau age Intercept age (km) 52-1-89 4.0 1368.5 þ 11.7 1374.4 þ 13.0 1362.7 þ 12.9 47-5-89 12.2 999.8 þ 8.8 994.8 þ 11.4 999.3 þ 9.4 5-10-88 16.0 88.5 þ 1.2 none 89.9 þ 2.0 16-1-88 16.5 92.6 þ 1.0 89.8 þ 1.1 92.0 þ 2.6 1-10-88 17.4 92.7 þ 1.2 none 92.8 þ 3.7 Samples were run in nine step-heating extractions, from 730³C to fusion (above 1280³C).

EPSL 5439 10-5-00 318 P.W. Reiners et al. / Earth and Planetary Science Letters 178 (2000) 315^326

Table 2 (U^Th)/He ages of apatite and titanite from Gold Butte Sample name Paleodeptha Apatite Average apatite Titanite Average titanite He age He age He age He age (km) þ 2c þ2c 98PRGB4 14.4 16.6 þ 1.0 17.4 þ 1.0 18.9 þ 1.5 18.6 þ 1.5 18.1 þ 1.1 22.5 þ 1.8 16.6 þ 1.3 16.4 þ 1.3 98PRGB6 12.6 14.3 þ 0.9 14.4 þ 0.9 14.5 þ 0.9 98PRGB9 10.9 12.8 þ 0.8 13.6 þ 0.9 14.5 þ 0.9 98PRGB12 9.0 14.7 þ 0.9 14.6 þ 0.9 26.6 þ 2.1 25.7 þ 2.1 14.6 þ 0.9 24.8 þ 2.0 98PRGB14 7.6 15.4 þ 0.9 17.1 þ 1.0 72.5 þ 5.8 73.2 þ 5.9 17.9 þ 1.1 74.0 þ 5.9 17.9 þ 1.1 98PRGB16 6.7 107 þ 8.5 104 þ 8.3 112 þ 8.9 94.8 þ 7.6 98PRGB15 6.1 11.1 þ 0.7 12.0 þ 0.7 139 þ 11 138 þ 11 13.0 þ 0.8 137 þ 11 98PRGB20 4.8 190 þ 15 192 þ 15 194 þ 16 98PRGB18 3.4 17.2 þ 1.0 17.0 þ 1.0 150 þ 12 150 þ 12 16.7 þ 1.0 aPre-exhumation paleodepth is depth beneath basal Tapeats sandstone (distance from crystalline rock Tapeats unconformity corrected for 65³ east dip of the overlying sedimentary rocks) plus an assumed pre-exhumation sedimentary thickness of 2.5 km. cambrian basement. These data also support the Muscovite 40Ar/39Ar dates were obtained by experimentally derived 200³C (U^Th)/He Tc for step-heating experiments at the University of titanite. Maine (Table 1). Apatite and titanite (U^Th)/He dates (Table 2) were obtained by standard proce- dures at Caltech (e.g., [18] for apatite; [21,22] for 2. Samples and methods titanite), involving heating of inclusion-free grains in a resistance furnace and measurement of re- To characterize (U^Th)/He age^depth relation- leased 4He by isotope dilution on a Balzers quad- ships in a steeply dipping pre-exhumation pro¢le, rupole mass spectrometer. 4He contents of some we collected samples of the Gold Butte granite titanite grains were also measured by peak height from a roughly east^west transect, perpendicular comparison to standards on a sector mass spec- to the strike of the overlying sedimentary rocks trometer (MAP 215-50). Grains were then recov- on the east side of the block (Fig. 1). Many sam- ered, spiked with 230Th and 235U, dissolved in ples from the western and central portions of the HNO3, HCl, and/or HF. Th and U isotope ratios block did not contain apatite or titanite. Samples were then measured on a single-collector double- collected for muscovite 40Ar/39Ar dating were col- focussing ICP^MS (Finnigan Element). lected from pegmatite dikes and gneissic litholo- Raw apatite ages were corrected for the e¡ects gies in the eastern and western portions of the of alpha-ejection ([24]; FT parameters were 0.68^ block. We show these results along with locations 0.79), and estimated uncertainties on apatite He of samples analyzed for muscovite K/Ar dating ages are 6% (2c). Titanite ages were not corrected (Fig. 1) by Wasserburg and Lanphere [6]. for alpha-ejection because of uncertainties in orig-

EPSL 5439 10-5-00 P.W. Reiners et al. / Earth and Planetary Science Letters 178 (2000) 315^326 319 inal grain sizes and shapes. We do not expect these corrections to exceed about 8% however, based on typical titanite crystal sizes and replicate analyses of other samples with known ages (primarily Fish Canyon Tu¡ and Mt. Dromedary specimens [22]). An additional complication with titanite He ages is the possibility of di¡usive rounding of He pro¢les within individual crystals [22]. Crystals with histories involving slow cooling or prolonged residence at temperatures near the closure temperature will result in strongly zoned He concentrations from core to rim. Because our sample preparation procedure produced irregular fragments (about 200U500 Wm) of larger titanite crystals (about 300U800 Wm), it is possible that fragments with a large fraction of material from either the original core or rim of a whole crystal record ages that are di¡erent from those of bulk grains [22]. This is especially true for samples from the eastern part of the block, where low pre-exhumation temperatures and partial He re- Fig. 2. Ages (left: linear scale; right: log scale) and paleo- tention would be expected. Nonetheless, based depths (assuming a 2.5 km thick sedimentary cover dipping on the relative sizes of observed whole crystals 65³, i.e., paleodepth is the distance of the sample from the and the fragments we selected for analyses, as unconformity along a perpendicular transect multiplied by well as the reproducibility of ages from individual sin[65³], plus 2.5 km) of samples from this study and those of Fitzgerald et al. [10] and Wasserburg and Lanphere [6]. samples, we do not expect these e¡ects to exceed Dashed lines are isotherms corresponding to the Tcs of apa- about 8%. Our estimated analytical uncertainty tite and titanite (U^Th)/He and the lower limit of muscovite on titanite He ages of 8% is based on the repro- 40Ar/39Ar, assuming a geothermal gradient of 20³C/km and ducibility of these and other samples, and is great- ambient surface temperature of 10³C. The westernmost, 77 er than the propagated uncertainty on U, Th, and Ma, apatite FT sample of Fitzgerald et al. is not shown here, due to uncertainties in its structural position. The total He content determinations (V2^3%). gas ages of the muscovite 40Ar/39Ar analyses (Table 2) and Apatite and titanite U/Pb and Pb/Pb dates (Ta- average 207Pb/206Pb ages of replicate samples (Table 3) are ble 3) were obtained by dissolving and spiking shown here. In the linear plot (left side) 2c error bars are approximately 0.1 mg of inclusion-free apatite smaller than symbols, except for titanite He ages, which are and titanite grains with 235U and 205Pb, and mea- shown. Lines through points of the same phase and system are not formal regressions and are meant as guides for the suring peak intensities on a single-collector, dou- eye only. ble-focussing ICP^MS. Initial Pb isotopic compo- sitions were assumed to be those of potassium

Table 3 Pb/Pb and U/Pb ages of apatite and titanite from Gold Butte Sample name Apatite 207Pb/206Pb age Titanite 207Pb/206Pb age Titanite 206Pb/238U age Titanite 207Pb/235U age 98PRGB4 1130 1384 1457 1450 1417 1439 1452 98PRGB18 1389 1391 1435 1439 1392 1467 1459 Estimated analytical errors, based on reproducibility of ages on these and other samples on the element, are 2% and 4% for titanite and apatite Pb/Pb ages, respectively, and about 4% for titanite U/Pb ages.

EPSL 5439 10-5-00 320 P.W. Reiners et al. / Earth and Planetary Science Letters 178 (2000) 315^326 feldspar we measured from a sample of the Gold 98PRGB12. Replicate analyses of the western- Butte granite (98PRGB4; 206Pb/204Pb = 16.25 and most titanite bearing sample yielded an average 207Pb/204Pb = 15.45) While not as precise as con- age of 18.6 þ 1.5 Ma, slightly older than the apa- ventional TIMS Pb dates, based on our experi- tite He ages at the same depth. Titanite in ence with these and other Precambrian age sam- 98PRGB12, in the central portion, yielded an ples, after correction for mass fractionation, average age of 25.7 þ 2.1 Ma. Titanites east of isobaric interferences, and blanks, we estimate this point become progressively older to the that the method provides suitably precise 207Pb/ east, from 73.2 þ 5.9 to 192 þ 16 Ma. An exception 206Pb ages (V2% for titanite and V4% for apa- to this eastward-increasing age rule is the eastern- tite), for samples greater than several hundred most sample, 98PRGB18, with an age of 150 þ 12 million years old. Titanite U/Pb ages were slightly Ma, which is younger than the adjacent sample to more variable (V4%) than titanite Pb/Pb dates the west (Fig. 2). As a whole these apatite and (above). Apatite U/Pb ages were considerably titanite (U^Th)/He ages are consistent with apa- more variable (in some cases greater than 20%); tite FT ages (with the exception of the 77 Ma this method was not fully evaluated and devel- westernmost sample of Fitzgerald et al.; not oped, and these results are not discussed here. shown here) which average 16.1 þ 1.5 Ma from estimated paleodepths of 16.5^5 km, then increase to 50 þ 3.0 Ma from about 5 to 3.5 km (Fig. 2). 3. Results Muscovite 40Ar/39Ar and K/Ar ages (total gas) from samples at paleodepths of 18^16 km are Fig. 2 shows all ages as a function of location tightly constrained to 91.3 þ 1.3 Ma (Table 2, in the block, expressed as pre-exhumation paleo- Fig. 2). Samples at shallower paleodepths are con- depth, with sample locations projected onto our sistently older, increasing from 1000 þ 8.8, transect parallel to the strike of the overlying sedi- through 1196 þ 35, and 1385 þ 39 Ma to the mentary rocks. We have assumed a pre-exhuma- east; the easternmost sample is 1368 Ma (the tion sedimentary cover of 2.5 km thickness, based 1196 and 1385 Ma ages are recalculations of Was- on thicknesses of the stratigraphic section on the serburg and Lanphere's [6] dates). east side of the block (Cambrian through Permian Apatite and titanite 207Pb/206Pb ages from the units [3]), which is similar to sections in adjacent easternmost sample (98PRGB18) are 1389 þ 56 areas, including the Grand Canyon. Also shown Ma, and 1392 þ 28 Ma, respectively (Table 3), in Fig. 2 are the apatite FT ages of Fitzgerald et and are concordant with 40Ar/39Ar ages from al. [10] and K/Ar ages of Wasserburg and Lan- samples at comparable paleodepths (Table 2). Ti- phere ([6]; ages recalculated using constants from tanite U/Pb ages are slightly older, between [25]). The (77 Ma) southwesternmost sample of 1435 þ 57 and 1467 þ 57 Ma. The westernmost Fitzgerald et al. [10] is not shown in Fig. 2 be- sample (98PRGB4) also yielded titanite Pb/Pb cause it was inferred to have been excised from a and U/Pb ages between 1384 þ 55 and 1457 þ 58 higher structural position and transported by mo- Ma, similar to the ages of the easternmost sample. tion along the overlying fault that exhumed the The apatite Pb/Pb age of the westernmost sample, block. however, is 1130 þ 45 Ma, distinctly younger than Apatite (U^Th)/He ages from the Gold Butte any of the other Pb ages. block show no consistent relationship with location, averaging 15.2 þ 1.0 Ma (Table 1, Fig. 2). In detail, the youngest ages (as young as 12.0 þ 0.7 4. Discussion Ma) are found in the central part of the block, and older ages are found on the western and Age^distance relationships among the di¡erent eastern edges (17.0 þ 1.0 and 17.4 þ 1.0, respec- chronometers in this study provide a range of tively). Titanite was not found in samples to the constraints on the structure of Gold Butte, the west of 98PRGB4, nor between 98PRGB4 and timing and rate of exhumation, and the pre-exhu-

EPSL 5439 10-5-00 P.W. Reiners et al. / Earth and Planetary Science Letters 178 (2000) 315^326 321 mation thermal structure of the block. The ¢rst companied exhumation at this time. Assuming important result is that with the possible excep- that the approximately 13 km section of paleo- tion of unexpectedly young titanite He ages in the vertical relief represented by the eastern- and easternmost portion of the block, which may rep- westernmost apatite samples was exhumed in resent duplication of section by faulting (as dis- 2 Myr yields an exhumation rate of 7 mm/yr. This cussed below), all of these data are consistent with is only a minimum estimate of the exhumation the interpretation of Gold Butte as a single, in- rate however, because it relies on the estimated tact, steeply dipping section of crust that was rap- range of apatite He ages as the temporal con- idly exhumed and cooled at 15^16 Ma. The pre- straint (i.e., cooling of the block was too rapid exhumation vertical thickness of the block is con- to produce a non-in¢nite age^paleodepth trend strained by the correlation between paleodepth that could be used to constrain exhumation and titanite He age through paleodepths shal- rate). Thermochronologic constraints on other lower than about 14 km. These age^paleodepth large-o¡set extensional faults in the southwestern relationships suggest a continuous depth^temper- USA have determined slip rates that are similar or ature increase, which we interpret as a continuous slightly higher than this (3^9 mm/yr; generally 7^ crustal section, to at least 14 km. Thus the inter- 8 mm/yr) [26^31]. Due to the inferred steep dip of nally consistent age results of each chronometer in the exhuming normal fault [1^3], in the case of this study and those of Fitzgerald et al. [10] Gold Butte the exhumation rate was probably strongly support the interpretation made by pre- only slightly less than the slip rate. vious workers that the Gold Butte block is a In detail, apatite He ages do show a variation largely intact, continuous crustal section with pa- between 12.0 þ 0.7 Ma and 17.4 þ 1.0 Ma, with the leodepth increasing towards the west [1^3]. This youngest ages in the central part of the block. recognition also provides an important basis for This may re£ect the topography and sur¢cial more detailed thermochronologic interpretations drainage patterns of the block following exhuma- of the ages. tion and tilting to an approximately 50^65³ east- Generally uniform V15^16 Ma apatite (U^ ward dip. Th)/He ages throughout the block indicate that In contrast to apatite, titanite shows a wide at this time, within less than about 2 Myr (the range of (U^Th)/He ages through the section. standard deviation of apatite He ages in the The average age of 18.6 þ 1.5 Ma for the western- block) all the samples in this transect were rapidly most sample is slightly older than that of apatite cooled below 70³C. This is consistent with Fitz- throughout the block, but if the single sample gerald et al.'s [10] evidence for rapid exhumation with the 22.5 Ma age is removed, the average at 16 Ma from the age of the base of the apatite age is 17.3 þ 1.0 Ma, which is indistinguishable FT partial annealing zone (PAZ) [10]. Assuming from the apatite He ages. Assuming that the 2.5 km of overlying sediments, a 10³C ambient 22.5 Ma age is spurious (possibly because of in- surface temperature, and a 20³C/km geothermal clusions of undissolved U- and Th-bearing miner- gradient (consistent with the 5 km paleodepth of als within the grains), these data indicate that the Fitzgerald et al.'s apatite FT PAZ), the depth of temperature at this pre-exhumation depth was the 85³C isotherm that is typically assumed to be higher than 200³C, the closure temperature of ti- the base of the apatite He partial retention zone tanite [21,22], consistent with crustal depths great- (PRZ [14,18]) would be 3.8 km, slightly deeper er than 10 km for a geothermal gradient of 20³C/ than the inferred paleodepth of the easternmost km. sample (3.4 km; Fig. 2). Thus the average The 25^27 Ma titanite He ages at 9.3 km pale- 17.4 þ 1.0 Ma age determined for this sample in- odepth are consistent with cooling below 200³C at dicates that this pre-exhumation geotherm was at this depth signi¢cantly before 15^16 Ma. How- least 20³C/km. The approximately 2 Myr spread ever, rocks at this depth were probably not sig- in apatite ages also provides a crude constraint on ni¢cantly cooler than 190³C just prior to exhuma- the maximum duration for the cooling that ac- tion, because: (1) apatite He ages show that at

EPSL 5439 10-5-00 322 P.W. Reiners et al. / Earth and Planetary Science Letters 178 (2000) 315^326 least up to 3.4 km paleodepth, rocks were still FT age on our sample also suggests that this east- hotter than about 85³C, and (2) apatite FT ages ernmost sample in the transect may be out of show that at least up to 5 km paleodepth, rocks structural sequence. were still hotter than about 110³C. Both of these With closure temperatures in the 350^425³C results are consistent with a 20³C/km geotherm, range [23] muscovite 40Ar/39Ar and K/Ar ages which predicts a temperature of 196³C at 9.3 km. provide higher temperature thermochronologic Isothermal holding of titanite at this temperature constraints than (U^Th)/He or apatite FT ages. prior to 15^16 Ma exhumation would result in The 1.37^1.39 Ga ages at the top of the section su¤cient He retention to produce an age greater are consistent with slow regional cooling of the than 15^16 Ma. For example, Wolf et al. [14] crust following intrusion of the 1.45 Gold Butte have shown that all crystals, including those granite, while the decreasing 1.0 and 1.2 Ga ages open to He di¡usion due to residence at temper- in the next two samples to the west probably re- atures near Tc, will achieve steady state He ages, £ect partial Ar retention over long time periods. 2 31 teqV(1/15)U(D/a ) . Assuming an activation en- Together with the tight cluster of 89^93 Ma ages ergy and frequency factor for titanite of 44.6 kcal/ towards the western end of the block, this age^ mol and 59 cm2/s [22], an average grain radius of paleodepth relation resembles a fossil PRZ [33] 200 Wm, and an isothermal holding temperature with a break-in-slope somewhere between 12 and of 196³C, the titanite teq is 8.7 Ma. Additional 16 km paleodepth. A 20³C/km geotherm predicts isothermal holding for 16 Ma at ambient surface a 350³C isotherm at about 17 km depth, so 15^16 temperatures following exhumation yields an age Ma Ar ages re£ecting zero Ar retention prior to of 25 Ma, in good agreement with the observed exhumation would only be expected at greater titanite ages of 25^27 Ma at 9.3 km paleodepth. crustal depths than observed here. Most impor- The rapidly increasing titanite He ages above tantly, the cluster of 90 Ma ages at the 18^16 9 km are consistent with a fossil titanite He PRZ km paleodepths clearly represents a cooling event established some time prior to about 192 Ma. The at this time not recorded by the (U^Th)/He chro- upper 9 km of the Gold Butte block has been nometers. It is unlikely that heating from a young cooler than 200³C since about this time, although (64^66 Ma) two-mica granite [8] in the west-cen- it is possible that it was signi¢cantly cooler than tral part of the block signi¢cantly a¡ected the 200³C well before then, and was subsequently muscovite Ar ages, because all three samples are heated and cooled again before 192 Ma. The un- concordant at 90 Ma, and Parolini et al. [9] also expectedly young 150 Ma age of the easternmost reported a single zircon FT age from the Gold sample is problematic in the interpretation of a Butte townsite (north-central portion of the continuous fossil He PRZ. One possibility is block) with a similar age of 85 Ma. Dumitru et that this sample represents a slightly greater struc- al. [34] also observed late Cretaceous apatite FT tural depth than the one immediately to the west, ages from V65^90 Ma in samples from the due to motion along an extension of the Million Grand Canyon region. Possible causes of this Hills Wash fault, a fairly large o¡set ( s 1 km), late Cretaceous cooling include cessation of Sevier shallowly dipping, top-to-the-west normal fault thrusting [35], the end of intense plutonism in the with a southern projection between these two Sierra Nevada and a regional decrease in geother- samples [1,3]. This would mean that sample mal gradients at this time [16,35,36], or erosion 98PRGB20 is the structurally shallowest sample caused by Laramide deformation [34]. In any in the He transect. This interpretation is sup- case, Ar age^depth relationships also suggest a ported by a 15.6 þ 2.3 Ma apatite FT age for structurally continuous block, and indicate a 98PRGB18 [32]. Projection of the apatite FT mid-crustal depth for the western end of the block ages of Fitzgerald et al. to our transect indicate from the late Cretaceous through Miocene. that samples in a continuous vertical sequence at Similar 1389^1392 Ma 207Pb/206Pb ages of both the paleodepth of sample 98PRGB18 should have titanite and apatite for the easternmost sample are ages of about 50 Ma. Thus the 15.6 Ma apatite also consistent with regional cooling of the crust

EPSL 5439 10-5-00 P.W. Reiners et al. / Earth and Planetary Science Letters 178 (2000) 315^326 323 following 1.45 Ga plutonism. The approximately 20 Myr age di¡erence between the Pb and Ar ages from the easternmost samples probably re£ect the higher closure temperature of the U/Pb system (approximately 700³C for titanite [37] and 450³C for apatite [38]). The markedly younger 1130 Ma apatite 207Pb/206Pb age for the westernmost sample probably re£ects signi¢cantly slower regional cooling at 15 km crustal depths. Combining U/Pb, Pb/Pb, Ar/Ar, FT, and (U^ Th)/He dates on samples from the eastern and western ends of the Gold Butte block provides a detailed time^temperature history for the di¡erent crustal levels of the block (Fig. 3). Zircon U/Pb [7] and titanite Pb/Pb dates indicate rapid cooling of the pluton from magmatic temperatures at 1.45 Fig. 3. Long-term thermal history of Gold Butte, based on data from all available geo- and thermochronometers. Zircon Ga to temperatures less than that of the titanite U/Pb is 1.45 Ga age of the Gold Butte granite [7]; an arbi-

Pb Tc (V700³C [37]) by about 1.4 Ga. Apatite trary high Tc of 900³C is assumed for this age. Other Tcs Pb/Pb ages indicate that rapid cooling of the east- are assumed to be 700³C for Pb in titanite, 450³C for Pb in ern portion through about 450³C [38] continued apatite, 350³C for Ar in muscovite, 200³C for He in titanite, quickly (1.39 Ga), while the western portion did 110³C for apatite FTs, and 70³C for He in apatite. The trend labeled `west end' includes He and Pb ages from sample not cool below the apatite Pb Tc until about 1.13 98PRGB4, and the average Ar age of the three westernmost Ga. The largest di¡erence in eastern and western muscovite samples. The trend labeled `east end' includes the ages for the same system is muscovite 40Ar/39Ar; easternmost (50 Ma) apatite FT sample of Fitzgerald et al. the eastern part of the block cooled through this [10], sample 52-1-89 for muscovite Ar age, 98PRGB18 for blocking temperature (assumed to be the median, Pb ages and the apatite He age, and 98PRGB20 for the titanite He age (due to the anomalously young age in the east- 388³C, of the 350^425³C range suggested by ernmost titanite). The bottom of the block clearly has a Hodges [23]) by 1.37 Ga, while the western por- slower cooling history until about 16 Ma, re£ecting its great- tion did not close with respect to Ar di¡usion er depth. The possible increase in temperature at 90 Ma until about 90 Ma. Assuming a pre-Miocene pa- could also be modeled as simply a gradual cooling through leodepth of 17 km for the muscovite samples at the muscovite Ar Tc, but the fact that all three of the westernmost samples yield 90 Ma ages suggests that a signi¢cant the west end of the block, geotherms of 20^25³C/ section of the mid-crust closed to Ar di¡usion at this time, km predict temperatures from 350 to 435³C at and this may be more consistent with cooling from a higher this depth. The 90 Ma muscovite ages in this por- geotherm (25³C/km), or regional exhumation, at this time. tion of the block could therefore represent either a The top of the block clearly cooled more quickly than the decrease in the geotherm below 25³C/km at this bottom, as indicated by the di¡erence in muscovite Ar ages between the east and west ends. The 192 Ma titanite He age time, or a transient heating and cooling cycle probably re£ects prolonged residence at a temperature near from a 20³C/km geotherm (Fig. 3). Subsequent 160³C (gray line) for some interval between the late Protero- to this late Cretaceous cooling event, the only zoic and early Tertiary. This could have been caused by ei- other cooling event recorded in rocks of the west- ther a slightly greater thickness of overlying sediments than ern portion of the block is the 15^16 Ma Miocene the pre-Miocene 4.8 km, or a higher geothermal gradient of V30³C/km, rather than the 20³C/km inferred for the more exhumation re£ected in all of the titanite (U^Th)/ recent history by the other data. He, apatite FT, and apatite (U^Th)/He systems. The timing of cooling through 200³C is more di¤cult to ascertain for the eastern portion of the Assuming the simplest possible cooling path, block, as the location of the top of the titanite He judging by the rapid Proterozoic cooling indicated PRZ is not clear, possibly due to structural com- by Pb and Ar ages and assuming the pre-Miocene plications in the easternmost portion of the block. paleodepth of about 5.1 km and inferred geother-

EPSL 5439 10-5-00 324 P.W. Reiners et al. / Earth and Planetary Science Letters 178 (2000) 315^326 mal gradient of 20³C/km, the east end of the at middle crustal depths at the western edge of the block near sample 98PRGB20 should have been Colorado Plateau resided at moderate geothermal about 112³C by the late Proterozoic, well below gradients (and slowly declined in temperature) the titanite He Tc. However, sample 98PRGB20 from 1.4 Ga to about 90 Ma. The 90 Ma event yields a 192 Ma titanite He age. A modestly in- may re£ect accelerated cooling, a brief heating creased geotherm (30^33³C/km) or slightly greater and cooling cycle (after which gradients returned paleodepth (V6 km, rather than 5.1 km) any time to a 20³C/km prior to rapid exhumation in the between the late Proterozoic and about 90 Ma Miocene), or a regional exhumation event as sug- could have caused considerable He loss and de- gested by Dumitru et al. [34]. Similar pre-exten- creased the titanite He ages in the eastern portion sional geothermal gradients of 20^25³C/km have of the block. For example, assuming the same also been interpreted from thermochronologic titanite He di¡usion parameters as above, an iso- data in other areas of the Basin and Range thermal holding temperature of 163³C (achievable [41,42]. The shallower crustal history is more dif- by either 5.1 km paleodepth at 30³C/km, or 6.1 ¢cult to interpret between 1.4 Ga and 192 Ma, km paleodepth at 25³C/km) would generate a teq but these data are consistent with rapid cooling of 184 Ma. With an additional 16 Ma of post- through the muscovite Ar Tc (350^425³C) follow- Miocene holding at 10³C, this would produce a ing intrusion of the Gold Butte granite by V1380 titanite He age of 200 Ma, not far from the ob- Ma, followed by residence at temperatures typical served age of 98PRGB20. Thus the mid-Jurassic of the upper 5^6 km of crust (V110^160³C) titanite He ages for the eastern portion of the probably until the Tertiary. A Laramide-age cool- Gold Butte block probably re£ect low but partial ing event may have a¡ected this shallow part of He retention at V150^170³C temperatures over the crust, but the most clearly expressed cooling long time periods (any time between V500 Myr event is the Miocene exhumation that rapidly de- and 1 Gyr) prior to Miocene exhumation, rather creased the temperature of this part of the crust to than any speci¢c geologic event. below 70³C at 15^16 Ma. The 50 Ma apatite FT age for the eastern portion of the block [10] may also be related to isothermal holding at temperatures at which FTs 5. Conclusions anneal at slow rates, rather than any particular cooling event. Alternatively, it could represent Systematic spatial variations in U/Pb, Ar/Ar, the latest cooling associated with Laramide-age FT, and (U^Th)/He ages of samples from Gold tectonism, as suggested by other studies of the Butte are consistent with the interpretation of the region [34]. Laramide uplift near Gold Butte has block as a single, largely intact, steeply dipping been inferred from evidence for a basement high cross-section of crust s 15 km thick. Apatite to the south (e.g., the Kingman uplift of [4,39], and titanite He ages throughout the block indi- Paleozoic and Mesozoic strata erosionally bev- cate that prior to rapid (at least 7 mm/yr) exhu- elled southward towards Gold Butte, Eocene mation at about 15^16 Ma, crust just west of the gravel deposits ¢lling northeastward-draining pa- present-day Colorado Plateau resided at geother- leocanyons near the edge of the Hualapai plateau mal gradients of about 20^25³C. A cooling event [40], and drainage patterns to the northeast in at 90 Ma a¡ected the mid-crust in this area, rep- southeastern Nevada and northwestern Arizona resented by samples from s 15 km depth, bring- [4]). Because of the 15^16 Ma apatite He ages in ing rock to a temperature between the Tc of Ar in the eastern portion of the block however, one ¢rm muscovite and He in titanite (200 to V388³C). constraint we can place on the eastern (shallow Prior to this event the time^temperature path of paleodepth) end of the block is that the temper- the block is di¤cult to interpret, but taken as a ature was above 70³C until the Miocene exhuma- whole, the He, Ar, and Pb ages from the upper tion event. and lower portions of the block suggest di¡erent Taken as a whole these data suggest that rocks cooling patterns from near magmatic tempera-

EPSL 5439 10-5-00 P.W. Reiners et al. / Earth and Planetary Science Letters 178 (2000) 315^326 325 tures at 1.45 Ga that re£ect di¡erent pre-exhuma- straints from ¢ssion track analysis, Geology 19 (1991) tion depths. The lower portion of the block did 1013^1016. [11] P.K. Zeitler, A.L. Herczig, I. McDougall, M. Honda, U- not cool through 450³C until about 1.13 Ga, and Th-He dating of apatite: A potential thermochronometer, probably remained at temperatures near 350³C Geochim. Cosmochim. Acta 51 (1987) 2865^2868. until about 90 Ma, consistent with a geotherm [12] R.A. Wolf, K.A. Farley, L.T. Silver, Helium di¡usion and of about 20^25³C/km for this pre-exhumation low-temperature thermochronometry of apatite, Geochim. depth. In contrast, the upper portion of the block Cosmochim. Acta 60 (1996) 4231^4240. [13] R.A. Wolf, K.A. Farley, L.T. Silver, Assessment of (U- cooled rapidly from near magmatic temperatures Th)/He thermochronometry The low-temperature history through 350^425³C by 1.37 Ga, consistent with a of the San Jacinto mountains, California, Geology 25 much shallower depth. Finally, the agreement be- (1997) 65^68. tween the apparent depth of the fossil He PRZ for [14] R.A. Wolf, K.A. Farley, D.M. Kass, Modeling of the titanite and the inferred depth of the pre-exhuma- temperature sensitivity of the apatite (U-Th)/He thermochronometer, Chem. Geol. 148 (1998) 105^114. tion 200³C isotherm supports the laboratory-de- [15] A.C. Warnock, P.K. Zeitler, R.A. Wolf, S.C. Bergman, rived di¡usion measurements indicating a 200³C An evaluation of low-temperature apatite U-Th/He ther- Tc for this chronometer.[RV] mochronometry, Geochim. Cosmochim. Acta 61 (1997) 5371^5377. [16] M.A. House, B.P. Wernicke, K.A. Farley, T.A. Dumitru, Cenozoic thermal evolution of the central Sierra Nevada, References California, from (U^Th)/He thermochronometry, Earth Planet. Sci. Lett. 151 (1997) 167^179. [1] R. Brady, B. Wernicke, J. Fryxell, Evolution of a large- [17] M.A. House, B.P. Wernicke, K.A. Farley, Dating topog- o¡set continental normal fault system, South Virgin raphy of the Sierra Nevada, California, using apatite (U- Mountains, Nevada, Geol. Soc. Am. Bull., in press. Th)/He ages, Nature 396 (1998) 66^69. [2] B.P. Wernicke, G.J. Axen, On the role of isostasy in the [18] M.A. House, K.A. Farley, B.P. Kohn, An empirical test evolution of normal fault systems, Geology 16 (1988) of helium di¡usion in apatite: borehole data from the 848^851. Otway basin, Australia, Earth Planet. Sci. Lett. 170 [3] J.E. Fryxell, G.G. Salton, J. Selverstone, B. Wernicke, (1999) 463^474. Gold Butte crustal section, South Virgin Mountains, Ne- [19] J.A. Spotila, K.A. Farley, K. Sieh, Uplift and erosion of vada, Tectonics 11 (1992) 1099^1120. the San Bernardino Mountains associated with transpres- [4] L.S. Beard, Paleogeography of the Horse Spring Forma- sion along the San Andreas fault, California, as con- tion in relation to the Lake Mead fault system, Virgin strained by radiogenic helium thermochronometry, Tec- Mountains, Nevada and Arizona, in: K.K. Beratan tonics 17 (1998) 360^378. (Ed.), Reconstructing the History of Basin and Range [20] K.A. Farley, M.A. House, B.P. Kohn, Laboratory and Extension using Sedimentology and Stratigraphy, Geol. natural di¡usivity calibrations for apatite (U-Th)/He ther- Soc. Am. Spec. Pap. 303, Boulder, CO, 1996, pp. 27^60. mochronometry, Miner. Mag. 62A (1998) 426^427. [5] A. Volborth, Rapakivi-type granites in the Precambrian [21] P.W. Reiners, K.A. Farley, Helium di¡usion and (U-Th)/ complex of Gold Butte, Clark County, Nevada, Geol. He thermochronometry of titanite, Miner. Mag. 62A Soc. Am. Bull. 73 (1962) 813^832. (1998) 1249^1250. [6] G.J. Wasserburg, M.A. Lanphere, Age determinations in [22] P.W. Reiners, K.A. Farley, Helium di¡usion and (U-Th)/ the Precambrian of Arizona and Nevada, Geol. Soc. Am. He thermochronometry of titanite, Geochim. Cosmochim. Bull. 76 (1965) 753^758. Acta 63 (1999) 3845^3859. [7] L.T. Silver, M.E. Bickford, W.R. Van Schmus, J.L. An- [23] K.V. Hodges, Pressure-temperature-time paths, Annu. derson, T.H. Anderson, L.G. Medaris Jr., The 1.4^1.5 Rev. Earth Planet. Sci. 19 (1991) 207^236. b.y. transcontinental anorogenic plutonic perforation of [24] K.A. Farley, R.A. Wolf, L.T. Silver, The e¡ects of long North America, Geol. Soc. Am., Abs. with Prog. 9 alpha-stopping distances on (U-Th)/He ages, Geochim. (1977) 1176^1177. Cosmochim. Acta 60 (1996) 4223^4229. [8] M. Martin, unpublished data, Massachusetts Institute of [25] R.H. Steiger, E. Ja«ger, Subcommission on geochronology: Technology, Cambridge, MA, 1998. Convention on the use of decay constants in geo- and [9] J.R. Parolini, E.I. Smith, J.R. Wilbanks, Fission track cosmochronology, Earth Planet. Sci. Lett. 36 (1977) dating of rapakivi granite in the Rainbow Gardens and 359^362. in the South Virgin Mountains, Clark County, Nevada, [26] G.A. Davis, Rapid upward transport of mid-crustal my- Isochron/West 30 (1981) 9^10. lonitic gneisses in the footwall of a Miocene detachment [10] P.G. Fitzgerald, J.E. Fryxell, B.P. Wernicke, Miocene fault, Whipple Mountains, southeastern California, Geol. crustal extension and uplift in southeastern Nevada: Con- Rundsch. 77 (1988) 191^209.

EPSL 5439 10-5-00 326 P.W. Reiners et al. / Earth and Planetary Science Letters 178 (2000) 315^326

[27] B.E. John, D.A. Foster, Structural and thermal con- [36] T.A. Dumitru, P.B. Gans, D.A. Foster, E.L. Miller, Re- straints on the initiation of angle of detachment faulting frigeration of the western Cordilleran lithosphere during in the southern Basin and Range: The Chemehuevi Laramide shallow-angle subduction, Geology 19 (1991) Mountains case study, Geol. Soc. Am. Bull. 105 (1993) 1145^1148. 1091^1108. [37] D.J. Scott, M.R. St-Onge, Constraints on Pb closure tem- [28] J.E. Spencer, S.J. Reynolds, Tectonics of mid-Tertiary perature in titanite based on rocks from the Ungava extension along a transect through west-central Arizona, orogen, Canada: Implications for U-Pb geochronology Tectonics 10 (1991) 1204^1221. and P-T-t path determinations, Geology 23 (1995) 1123^ [29] P.B. Gans, Extensional strain rates in the Basin and 1126. Range Province, Geol. Soc. Am., Abs. with Prog. 23 [38] K.R. Chamberlain, S.A. Bowring, Apatite-feldspar U-Pb (1991) 28. thermochronometer; a reliable, mid-range (approximately [30] D.A. Foster, A.J.W. Gleadow, S.J. Reynolds, P.G. Fitz- 450³C), di¡usion-controlled system, Geol. Soc. Am., Abs. gerald, Denudation of metamorphic core complexes and with Prog. 30 (1998) 215. the reconstruction of the Transition Zone, west central [39] A.F.H. Goetz, F.C. Billingsley, A.R. Gillespie, M.J. Arizona: Constraints from apatite ¢ssion track thermo- Abrams, R.L. Squires, E.M. Shoemaker, I. Lucchita, chronology, J. Geophys. Res. 98 (1993) 2167^2185. D.P. Elston, Applications of ERTS images and image [31] P.G. Fitzgerald, S.J. Reynolds, E. Stump, D.A. Foster, processing to regional geologic problems and geologic A.J.W. Gleadow, Thermochronologic evidence for timing mapping in northern Arizona, Jet Propulsion Laboratory of denudation and rate of crustal extension of the South Technical Report 32-1597, 1975, 188 pp. Mountains metamorphic core complex and Sierra Estrel- [40] R.A. Young, Laramide deformation, erosion and pluton- la, Arizona, Nuc. Tracks Rad. Meas. 21 (1993) 555^563. ism along the southwestern margin of the Colorado pla- [32] R. Donelick, unpublished data, Donelick Analytical Labs, teau, Tectonophysics 61 (1979) 25^47. 1999. [41] D.A. Foster, K.A. Howard, B.E. John, Thermochrono- [33] S.L. Baldwin, G.S. Lister, Thermochronology of the logic constraints on the development of metamorphic South Cyclades Shear Zone, Ios, Greece: E¡ects of duc- core complexes in the lower Colorado River area, in: tile shear in the argon partial retention zone, J. Geophys. M.A. Lanphere, G.B. Dalrymple, B.D. Turrin (Eds.), Ab- Res. 103 (1998) 7315^7336. stracts of the Eighth International Conference on Geo- [34] T.A. Dumitru, I.R. Duddy, P.F. Green, Mesozoic-Ceno- chronology, Cosmochronology, and Isotope Geology, zoic burial, uplift, and erosion history of the west-central U.S.G.S. Circ. 1107, 1994, p. 103. Colorado Plateau, Geology 22 (1994) 499^502. [42] K.A. Howard, D.A. Foster, Thermal unroo¢ng history of [35] T.A. Dumitru, Subnormal Cenozoic geothermal gradients a thick, tilted Basin-and-Range crustal section in the Tor- in the extinct Sierra Nevada magmatic arc: Consequences tilla Mountains, Arizona, J. Geophys. Res. 101 (1996) of Laramide and post-Laramide shallow-angle subduc- 511^522. tion, J. Geophys. Res. 95 (1990) 4925^4941.

EPSL 5439 10-5-00