Journal of Petrology Advance Access originally published online on August 18, 2005
Journal of Petrology 2006 47(1):97-118; doi:10.1093/petrology/egi069
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Monazite–Xenotime Thermochronometry and Al2SiO5 Reaction Textures in the Picuris Range, Northern New Mexico, USA: New Evidence for a 1450–1400 Ma Orogenic Event
1 DEPARTMENT OF GEOLOGY, BUCKNELL UNIVERSITY, LEWISBURG, PA 17837, USA
2 DEPARTMENT OF EARTH AND ENVIRONMENTAL SCIENCES, RENSSELAER POLYTECHNIC INSTITUTE, TROY, NY 12180–3590, USA
RECEIVED NOVEMBER 2, 2004; ACCEPTED JUNE 30, 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGEOLOGIC SETTING OF THE...DATA AND INTERPRETATIONSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Al2SiO5 reaction textures in aluminous schist and quartzite of the northern Picuris range, north-central New Mexico, record a paragenetic sequence of kyanite to sillimanite to andalusite, consistent with a clockwise P–T loop, with minor decompression near the Al2SiO5 triple-point. Peak metamorphic temperatures are estimated at 510–525°C, at 4·0–4·2 kbar. Kyanite and fibrolite are strongly deformed; some prismatic sillimanite, and all andalusite are relatively undeformed. Monazite occurs as inclusions within kyanite, mats of sillimanite and centimetre-scale porphyroblasts of andalusite, and is typically aligned subparallel to the dominant regional foliation (S0/S1 or S2) and extension lineation (L1). Back-scatter electron images and X-ray maps of monazite reveal distinct core, intermediate and rim compositional domains. Monazite–xenotime thermometry from the intermediate and rim domains yields temperatures of 405–470°C (±50°C) and 500–520°C (±50°C), respectively, consistent with the prograde to peak metamorphic growth of monazite. In situ, ion microprobe analyses from five monazites yield an upper intercept age of 1417 ± 9 Ma. Near-concordant to concordant analyses yield 207Pb–206Pb ages from 1434 ± 12 Ma (core) to 1390 ± 20 Ma (rim). We find no evidence of older regional metamorphism related to the
1650 Ma Mazatzal Orogeny. KEY WORDS: Al2SiO5; metamorphism; monazite; thermochronometry; triple-point
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGIC SETTING OF THE...DATA AND INTERPRETATIONSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Several recent studies (Karlstrom et al., 1997
, 2004; Pedrick et al., 1998; Read et al., 1999; Williams et al., 1999) hypothesize that Proterozoic rocks in north-central New Mexico (Fig. 1) experienced distinct tectonothermal events at 1650 and 1400 Ma. They propose that regional amphibolite to near-granulite facies metamorphism and north-vergent folding and thrusting was related to the Mazatzal Orogeny (1650 Ma). Subsequent overprinting by amphibolite facies metamorphism at 1400 Ma was accompanied by the reactivation of pre-existing (1650 Ma) regional deformational fabrics. The 1400 Ma thermal event is interpreted to reflect a major basaltic underplating event at the base of the crust (Williams et al., 1999; Karlstrom et al., 2004). This proposed thermal perturbation provides a mechanism to reset the argon systematics within older 1650 Ma metamorphic minerals, such as hornblende and muscovite, and cause new metamorphic minerals to grow or overgrowths to form on pre-existing minerals.
|
With respect to the Al2SiO5 ‘triple-point’ rocks of northern New Mexico (Holdaway, 1978
; Grambling, 1981; Grambling & Williams, 1985), Karlstrom et al. (1997) proposed "...polymetamorphism of the middle crust with P–T paths close to the alumino-silicate triple-point at both 1·65 and 1·4 Ga. A consequence of this complex thermal history is growth (and overgrowth) of amphibolite-grade minerals at both 1·65 and 1·4 Ga." The studies by Karlstrom et al. (1997), Pedrick et al. (1998) and Read et al. (1999) briefly discuss the possibility of a single regional tectonothermal event around 1400 Ma, but dismiss it in favor of the polymetamorphic hypothesis. Recent electron microprobe dating of monazite in the Tusas range (Fig. 1; Kopera et al., 2002a, 2002b) reveals a spread of ages from >1700 to 1400 Ma that are interpreted to reflect some component of 1450–1400 Ma deformation and metamorphism superimposed upon earlier Mazatzal (1650 Ma) deformation and metamorphism. Monazite core ages 
1700 Ma are interpreted as detrital. The primary goal of this work is to investigate the timing and nature of the Al2SiO5 ‘triple-point’ metamorphism in rocks of the northern Picuris Range (Fig. 1). To the best of our knowledge, this study is the first in northern New Mexico to utilize in situ analysis of monazite to constrain both the temperature and timing of monazite growth, yielding a relatively well-defined point in the T–t history of our samples. We also document the timing of monazite growth relative to the Al2SiO5 polymorphs and the regional deformational fabrics to constrain the absolute timing of the regional metamorphism and deformation. Such data are key in evaluating and refining models for the Proterozoic tectonic evolution of the southwestern USA.
|
GEOLOGIC SETTING OF THE PICURIS RANGE |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGIC SETTING OF THE...DATA AND INTERPRETATIONSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Numerous works document and discuss the geology (Fig. 2), metamorphism and deformational history of the Picuris range (Montgomery, 1953
; Long, 1974; Holdaway, 1978; Bell, 1985; Grambling & Williams, 1985; Bauer, 1988, 1993; Holdaway & Goodge, 1990; Bauer & Kelson, 1997; Williams et al., 1999). The ‘triple-point’ rocks examined in this study were collected from quartzite and aluminous schist within the Ortega Formation (Bauer & Williams, 1989), on the northern limb of the Hondo syncline (Fig. 2). This area is approximately 1 km north of the area studied by Holdaway & Goodge (1990). Metamorphic P–T conditions in this area were previously estimated at 530°C, 4 kbar based upon Grt–Bt thermometry and Al2SiO5 phase equilibria (Grambling & Williams, 1985; Holdaway & Goodge, 1990).
|
The Hondo syncline (Fig. 2) and other major map-scale and outcrop-scale folds were interpreted as second-generation folds (F2) by Bauer (1993)
. Ductile high-strain zones (Fig. 2) bound the Ortega and Rinconada formations to the north (Pilar shear zone) and south (Plomo fault). Bedding (S0) is well preserved within the cross-bedded quartzite of the Ortega and the overlying Rinconada formations. A near-bedding parallel schistosity (S1) and associated down-dip extension lineation (L1) are observed within the quartzites. S1 may be related to rootless, small-scale, intrafolial, isoclinal folds, preserved within the schists of the Rinconada Formation. Both S0, S1 and L1 are folded by F2-folds. Within schistose units, S1 is transposed into an east-striking, moderate to steeply south-dipping foliation (S2). S2 transects the F2 folds and is interpreted to have formed relatively late in the development of the F2 folds (Bauer, 1993). In the quartzites of the northern Picuris, Bauer noted that S1 and L1 in the northernmost quartzite continued development during F2 folding and probably represent a transposed composite foliation and lineation. Bauer (1993) proposed that the shearing and multiple generations of folds and foliations formed during a single progressive deformation. |
DATA AND INTERPRETATIONS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGIC SETTING OF THE...DATA AND INTERPRETATIONSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Petrography of the Al2SiO5 and associated minerals
Ten oriented samples of Ortega Formation quartzite and five oriented samples of schist were collected during fieldwork in the summers of 1994 and 2000. Quartzites were cut perpendicular to S1/S0 and parallel to L1. Schists were cut perpendicular to the intersection lineation (L1/2) of S1 and S2; in addition, a second cut was made perpendicular to S1 and parallel to L1/2. Quartzite samples (Fig. 3a and b) are characterized by the mineral assemblage quartz, kyanite, sillimanite, chloritoid, muscovite, hematite–ilmenite, tourmaline ± andalusite. Schist samples (Fig. 3c–f) were collected across a single layer approximately 4–5 m thick. These samples are characterized by large (4–10 cm diameter) andalusite porphyroblasts, abundant muscovite, both prismatic and fibrolitic sillimanite, kyanite, chloritoid, chlorite ± staurolite, hematite–ilmenite, and tourmaline. Accessory minerals, including monazite, xenotime, zircon and apatite, were observed in all samples.
|
Relative timing of the Al2SiO5 polymorphs
Kyanite is strongly deformed, fractured and embayed, and shows polymorphic replacement by sillimanite or andalusite (Fig. 3a, b and d). Many kyanite grains also show partial replacement by muscovite. Kyanite is aligned parallel to S0/S1 and L1 in the quartzite. Ilmenite–hematite inclusions within kyanite are common, and are also aligned parallel to S0/S1 and L1. Similar textures are observed within the schist samples, though kyanite and ilmenite–hematite are observed in both S1 and S2. These observations are consistent with earlier work by Holdaway (1978)
, Bauer (1988, 1993) and Holdaway & Goodge (1990), and suggest kyanite growth during progressive regional S1 and L1 deformation.
Sillimanite generally occurs as fibrolite or as fine, prismatic crystals in quartzites (Fig. 3a and b), and both fibrolite and large prismatic crystals up to 20 mm long in the schist samples (Fig. 3c–f). In the northern quartzite, sillimanite is generally aligned in S1 and L1. Elsewhere, sillimanite needles are generally undeformed (Holdaway, 1978; Holdaway & Goodge 1990; Bauer, 1993). Fibrolite in the schist occurs within large andalusite porphyroblasts (Fig. 3c and f), and within muscovite in the surrounding matrix. Bent and folded fibrolite are observed within andalusite porphyroblasts (Fig. 3f). Prismatic sillimanite crystals are oriented both parallel to and at a high angle to S2 (Fig. 3d and e), and contain inclusions of titanhematite aligned parallel to S2. We interpret these relations to indicate that prismatic sillimanite growth was syn- to post S2. Prismatic sillimanite cross-cuts kyanite; it is only observed as inclusions within andalusite, and typically shows fractures and embayments, patchy extinction and partial replacement by muscovite or andalusite. The crenulated fibrolite (Fig. 3f) pre-dates the prismatic sillimanite (Fig. 3e).
A large (8–10 cm diameter) andalusite porphyroblast with inclusions of kyanite and sillimanite aligned parallel to L1 was observed in a thin (1–2 cm) compositional layer within one quartzite sample. Within the schist layer, andalusite poikiloblasts (5–10 cm diameter) form up to 70% of the rock. They contain abundant inclusions of kyanite, sillimanite, quartz, muscovite, chloritoid and ilmenite–hematite (Fig. 3c–f). Andalusite replaces both kyanite and sillimanite, consistent with direct polymorphic replacement; in addition, andalusite also pseudomorphs muscovite and both kyanite and sillimanite are commonly rimmed by muscovite, consistent with the multi-step reaction mechanism proposed by Carmichael (1969). Poikiloblasts are optically continuous and overgrow a spaced crenulation cleavage defined by muscovite. These textures suggest that andalusite overgrows an intermediate to late stage of S2 (Fig. 3e and f). There is a weak deflection of the matrix foliation around these andalusite porphyroblasts that reflects minor, post-andalusite deformation.
Locally, kyanite, sillimanite and andalusite show minor, patchy alteration to pyrophyllite, consistent with observations by Bauer (1993). The occurrence of strongly deformed, embayed kyanite, crenulated fibrolite and embayed prismatic sillimanite within relatively undeformed andalusite shows that andalusite growth post-dates both fibrolite and prismatic sillimanite for these aluminous bulk compositions.
Relative timing of chloritoid and staurolite
Chloritoid (Fig. 3b) is present in many samples of schist and quartzite but the modal abundance is always small, visually estimated at 
1–2%. Chloritoid is commonly in contact with kyanite, sillimanite or andalusite; less common are chloritoid pseudomorphs after muscovite, and grains with partial retrograde alteration to chlorite. Chloritoid is generally deformed and aligned in S1/S2. Staurolite is observed in the schist samples but not the quartzite samples, consistent with observations by Holdaway (1978) and Holdaway & Goodge (1990). Staurolite occurs as millimetre-scale, subhedral to euhedral crystals, with a pale bluish-grey colour in plane-polarized light and very weak pleochroism; modal abundance estimated visually is 1–2%. Staurolite shows little evidence of strain or dissolution and is interpreted to be late or post S2.
Chemistry of the Al2SiO5 and associated minerals
Methods
Given the nearly identical mineral assemblages within the quartzite samples and the schist samples, we selected one quartzite (94–26) and one schist (00–2a) for major element analyses of kyanite, sillimanite, andalusite, muscovite, staurolite, chloritoid and ilmenite–hematite. Average Al2SiO5 analyses are given in Table 1 and representative chloritoid and staurolite analyses are given in Table 2. Major element analyses of the rock-forming minerals and all monazite and xenotime chemical analyses were performed on the JEOL 733 Superprobe electron microprobe housed in the Department of Earth and Environmental Sciences at Rensselaer Polytechnic Institute (RPI). Major element silicate and oxide mineral analyses were conducted at 15 kV, 20 nA with a spot size of 1–5 µm using natural standards for calibration.
|
|
Results
Kyanite and sillimanite within the quartzites contain 0·61 and 0·58 wt % Fe2O3, respectively. Within the schist samples, kyanite and sillimanite contain 0·76 and 0·87 wt % Fe2O3; andalusite contains 1·67 wt % Fe2O3. Mn2O3 was below detection limits in all Al2SiO5 polymorphs. Ilmenite–hematite compositions in the quartzite are variable, with analyses that range from Hem95Ilm5 to Hem53Ilm47, suggesting the presence of exsolved ilmenite and hematite domains. Ilmenite–hematite compositions from the schists are uniformly titanhematite (Hem91–94Ilm9–6). Ilmenite–hematite from both samples showed no measurable Mg or Mn. Chloritoid Mg/(Fe+Mg) varies from 0·043 in the quartzites to 0·121 in the schists. Chloritoid from the schist contains significantly higher MnO concentrations (MnO = 3·85 wt %) relative to chloritoid from the quartzites (MnO = 0·26 wt %). Staurolite contains 5–6 wt % Zn,
1 wt % MnO, with Mg/(Fe+Mg) = 0·12. Mineral compositions are generally consistent with analyses previously reported by Holdaway & Goodge (1990) for similar bulk compositions. The Al2SiO5 polymorphs analysed in this study contain significantly more Fe3+ than analyses reported by Grambling & Williams (1985) for the Picuris range.
P–T estimates of the ‘triple-point’ rocks
The occurrence of sillimanite + chloritoid but no staurolite in the quartzite samples places an upper limit on the temperature of metamorphism (Fig. 4; reaction (1)):
 | (1) |
; Rao & Johannes, 1979) place it at temperatures below the Pattison (1992) Al2SiO5 triple-point (Fig. 4), inconsistent with the interpreted equilibrium assemblage of sillimanite + chloritoid. Thus, we retain the Berman (1988) Al2SiO5 phase diagram (Fig. 4) utilized by Holdaway & Goodge (1990) and follow their methods and assumptions to calculate the temperature offset (at 4 kbar) for the reaction that results from the substitution of Mg and Mn into chloritoid in equilibrium with a hypothetical staurolite.
|
In the Fe-rich quartzites, chloritoid compositions of XFe = 0·96 correspond to a temperature shift of approximately –20°C (continuous line, Fig. 4), consistent with previous estimates by Holdaway & Goodge (1990)
, who estimated an error of 50% on 
T. This method of determining reaction offset is best applied to minerals with only small deviations from end-member compositions, and is not appropriate for the Zn-rich staurolite. Clearly, the incorporation of Zn+Mn, and possibly Li, stabilized staurolite to a temperature below 525°C.
Offsets for the Al2SiO5 phase equilibria were determined in a similar manner using the substitution of Fe3+ for Al3+. Based upon the average partition coefficients from Holdaway & Goodge (1990), a hypothetical andalusite in equilibrium with kyanite and sillimanite from the quartzite samples would have XFe2SiO5= 0·012. This amount of Fe substitution in andalusite shifts the andalusite–sillimanite reaction up approximately 500 bar. The lack of andalusite in the quartzite sample requires metamorphic pressures greater than approximately 4·0 kbar (Fig. 4). In the schist sample, measured andalusite compositions of XFe2SiO5 = 0·018 produce an offset of approximately 700 bar and require pressures below 4·2 kbar for the growth of andalusite. Holdaway & Goodge (1990) estimated errors of 50% on P. Our results (Fig. 4) are nearly identical to those of Holdaway & Goodge (1990), and constrain peak metamorphic temperatures to between 510°C (lower limit for sillimanite stability) and 525 ± 10°C, at pressures near 4·1 kbar ± 100 bar.
Proposed P–T–D path
The aluminium silicate textural relationships show early kyanite growth followed by sillimanite and, finally, andalusite growth, consistent with a clockwise loop around the Al2SiO5 triple-point with a component of decompression (Fig. 4). The early nature of the P–T loop is not well constrained, but probably falls somewhere between the two dashed lines. The P–T path of near-isobaric heating is similar to one proposed by Holdaway (1978), though his P–T path was placed below the triple-point based upon an inferred Al2SiO5 reaction sequence of kyanite to andalusite to sillimanite. The looping P–T path of heating and compression followed by approximately 1–1·5 kbar of near-isothermal decompression is based upon Gibbs method modelling of garnet zoning from the Picuris, and both Fe3+ and Mn3+ compositional zoning patterns in andalusite from the Rio Mora area (Daniel, 1992). We favor the P–T path of heating and compression, given the compressional style of deformation; however, the amount of decompression is poorly constrained.
Kyanite, sillimanite and chloritoid are commonly aligned in S1/S2 and L1. S2 develops late in the history of folding and is the dominant regional foliation. Andalusite and some prismatic sillimanite overgrow S2, showing that peak metamorphic temperatures outlasted much of the regional shortening (D2) indicated by F2 and S2. Based upon 40Ar–39Ar ages in the region (Karlstrom et al., 1997), these rocks experienced very slow cooling and probably remained at midcrustal depths of 12–15 km for 100–200 Myr. This extended residence time at elevated temperatures is responsible for the establishment of the sub-horizontal Al2SiO5 isograds documented by Grambling (1981) and Grambling & Williams (1985) in the adjacent Truchas Peaks and Rio Mora areas.
It is important to note that Holdaway & Goodge (1990) interpreted the offset of the andalusite–sillimanite reaction to be the result of a difference in fluid pressure (Pf) between quartzite and schist. Based upon the observed Al2SiO5 reaction textures from our study, we interpret the difference in pressure to be the result of different bulk-rock compositions equilibrating at slightly different pressures along a near-isothermal decompressional P–T path.
Monazite and xenotime petrography
Three samples of quartzite (94–23, 94–26 and 94–27) and two schist samples (00–2a and 00–3c) were selected for detailed analysis of monazite and xenotime. Monazite occurs as subhedral crystals ranging in size from 5 x 10 to 30 x 100 µm. They are observed as inclusions within kyanite (Fig. 5a), mats of fibrolite (Fig. 5b), andalusite (Fig. 5d and e), titanhematite and matrix muscovite (Fig. 5f). Matrix monazite commonly occur along quartz grain boundaries, Al2SiO5–quartz grain boundaries (Fig. 5c) and titanhematite–quartz grain boundaries. In the quartzites, titanhematite inclusions within monazite are aligned parallel to S1 and L1, and the monazite grains themselves are also aligned parallel to S1 and L1 (Fig. 5a–c). Monazite grains in schist samples are commonly included within andalusite, oriented parallel to S2 (Fig. 5d) and contain titanhematite inclusions aligned in S2 (Fig. 5e). A few grains are oriented perpendicular to S2. Matrix monazite are commonly aligned in S2 and included within foliated muscovite (Fig. 5f).
|
Monazite and xenotime chemistry
Methods
Monazite and xenotime grains were located by back-scattered electron (BSE) mapping of polished thin sections and energy dispersive spectroscopy (EDS) analysis. BSE mapping was performed both at RPI and with an FEI Quanta 400 ESEM equipped with an EDAX Genesis 4000 EDS system, located in the Department of Geology at Bucknell University. Monazite and xenotime grains were then examined with a petrographic microscope and photographed to document textural relationships with the rock-forming minerals and deformational fabrics. Selected monazite grains were mapped to characterize the distribution of Y, U, Th and Pb, or Ce, Ca, Y, Th and U using the program X-ray Map, coded by F. Spear and D. Wark. Mapping conditions included an accelerating voltage of 25 keV, Faraday cup current of 200–250 nA, pixel size of 0·2–2·0 mm and dwell times of 50–100 ms/pixel. Element maps were processed with the freeware program NIH Image.
Analytical conditions for quantitative monazite and xenotime analyses include 15 kV accelerating voltage, 50 nA Faraday cup current and count times of 20–100 s. For monazite, analysed elements (with approximate detection limit in wt %) include Si (0·02), P (0·47), Ca (0·01), Y (0·04), La (0·14), Ce (0·14), Pr (0·13), Nd (0·17), Sm (0·10), Gd (0·11), Tb (0·10), Dy (0·05), Er (0·05), Th (0·09), U (0·06) and Pb (0·11). Elements analysed in xenotime include Si (0·02), P (0·67), Ca (0·01), Y (0·11), Nd (0·07), Sm (0·06), Eu (0·06), Gd (0·07), Tb (0·08), Dy (0·08), Ho (0·10), Er (0·08), Tm (0·06), Yb (0·07), Th (0·12), U (0·07) and Pb (0·11). Standards include synthetic REE and Y phosphates, apatite, Th and Pb silicate, and U oxide.
The complexity of wavelength dispersive spectroscopy (WDS) analysis of minerals with high numbers of L-line and M-line generating elements, and potential interferences, is well known (Roeder, 1985). X-ray lines common to both the monazite and xenotime analytical protocols are Si K
, P K, Ca K, Y L, Th M1, U Mß1, and Pb Mß1. Y L and P K are analysed on a PET crystal to obtain larger resolution than is possible on TAP; this avoids the P K–Y Lß interference inherent in xenotime. Analysis of Pb Mß (on a Xe detector) avoids the interference from second-order Ce L on Pb Mß inherent in 140 mm Rowland circles coupled with Ar detectors, and the coupled Th Mz1,2 plus Y L
2,3 interference on Pb M. Interference of Th M on U Mß is corrected by application of an empirical correction factor combining measurement of Th M intensity plus apparent U Mß intensity in the U-free Th standard and the Th M intensity in the unknown (Pyle et al., 2005).
For the REE, the high concentrations of La, Ce and Nd in monazite generate problematic interferences between the Lß and L lines of the above elements and the L lines of the MREE (Nd–Tb). To avoid the need for interference corrections, intensities of the Lß lines of Pr, Nd, Sm, Gd and Tb are measured in monazite. The Lß line of Gd has a near-total overlap with Ho L, but extremely low (generally 0·1 wt %) concentrations of Ho in monazite obviate the need for a Gd Lß–Ho L interference correction.
Conversely, in xenotime, the generally low concentrations of LREE (La–Sm) result in negligible interference between Lß and L lines of the LREE and the L lines of the MREE. Therefore, the La lines of the MREE (Nd, Sm, Eu, Gd, Tb) in xenotime are measured. Ho Lß is measured in xenotime so as to avoid Gd Lß–Ho L interference. A final interference to be addressed is that of Tb Lß on Er L. Er is a major component in xenotime, and usually below detection limits in monazite, whereas Tb is a minor component (0–1 wt %) in both REE phosphates, but slightly more abundant in xenotime (Scherrer et al., 2000). It is preferable to correct Er L for Tb Lß interference rather than analyse Er Lß, because of the lower intensity of Er Lß and the partial overlap of Er Lß by Th Lß2. Tb L is analysed in xenotime to avoid Er L interference. Tb Lß is chosen for monazite analysis despite its lower intensity, as Tb L in monazite is subject to partial overlap by the tail of Sm Lß3, and is close also to the Sm Lß1 peak.
Results
Back-scatter electron imaging and X-ray compositional maps of monazite from all samples reveal up to three distinct compositional domains that we interpret as core, intermediate and rim (Fig. 6). In the quartzite samples (Fig. 6a–d), the core domains are defined by high Th, with the intermediate domains having low U and Y, and the rim domains defined by moderate to high U, and Y. The irregular and embayed nature for many of the grains and truncated internal zoning suggests at least two periods of dissolution interspersed between the growth events (Fig. 6a and b). Several mapped grains did not display such a core, and probably reflect a cut through the edge of the grain. Yttrium zoning is relatively flat across the core and intermediate domains, and increases slightly along the rims. A few subhedral monazite inclusions in kyanite show only localized development of a high-U rim, suggesting the grain was well armoured from subsequent dissolution and reprecipitation events (Fig. 6c). Anhedral monazite inclusions in fractured kyanite (Fig. 6d) commonly show well-developed high-U rims. We suggest that such grains were not isolated from the matrix and experienced the same dissolution and growth events that are recorded by the matrix monazite. Monazite grains from the schist samples (00–2A and 00–3C) possess U and Th compositions and zonation patterns distinctly different from those seen in the quartzite samples. Chemical zoning is characterized by a high-U core, a low-U outboard zone and an intermediate-U rim (Fig. 6e). Elevated Th content generally corresponds to high U. Relatively high yttrium corresponds to the increasing uranium in the rim domains.
|
Multiple analyses were taken from selected monazite grains to document the compositional variation in monazite and xenotime grains from both the quartzite and schist samples (Fig. 7). Representative monazite and xenotime analyses are given in Tables 3 and 4, respectively. (A complete listing of the 87 monazite analyses used in this study is available online from the Journal of Petrology website at http://www.petrology.oupjournals.org.) Xenotime analyses (Fig. 7a) vary between XYPO4 = 0·70–0·80 and XHREEPO4 = 0·20–0·30. A rescaled view (Fig. 7b) of monazite analyses from Fig. 7a more clearly shows the compositional differences across the different domains. Core compositions are the most enriched in HREE and the intermediate domain is the most depleted in HREE. Significant enrichment in huttonite + brabantite components (up to 15–25%) characterize the core domains in both quartzite and schist samples (Fig. 7c).
|
|
|
Analyses (Table 3) indicate that actinide enrichment in the quartzite samples is due primarily to Th substitution (up to 21·9 wt % ThO2). Thorium enrichment is not as extreme in the aluminous schist samples (up to 10·5 wt %), but monazite grains are highly enriched in U (up to 3·9 wt % UO2), relative to monazite from the quartzite. The incorporation of actinides (Th, U) is accomplished by the brabantite, Ca(Th,U)REE–2 and huttonite (Th,U)SiREE–1P–1 substitutions (Förster, 1998
). Ce2O3 concentrations fall to as little as 13 wt % in Th-rich monazite, with concomitant depletion in La2O3 (minimum 2·3 wt %).
Xenotime analyses from both quartzite and aluminous schist samples (Table 4) show some compositional variability, though back-scatter electron images show no detectable contrast variations within grains. XYPO4 varies between 0·71 and 0·78. Major HREE components (Dy, Er, Yb) vary by no more than 2 wt % in analysed grains. XGdPO4 ranges from 1·88 to 5·98 wt % and the higher concentrations correlate with lower XYPO4, consistent with xenotime growth during changing temperature conditions (Gratz & Heinrich, 1998).
Relative timing of monazite growth
Petrographic observations from the quartzite samples (Fig. 5a–c) and compositional zoning patterns observed in both matrix monazite and monazite included within kyanite (Fig. 6a–c) suggest that the monazite cores formed prior to the formation of kyanite. Intermediate and rim domains are interpreted to have grown in the kyanite stability field with rim domain growth probably extending into the sillimanite stability field for the schist samples. The alignment of titanhematite inclusions within aligned monazite (Fig. 5c) and aligned monazite included within aligned kyanite (Fig. 5a) strongly supports a history of progressive deformation during the formation of S1 and L1 in the quartzite samples.
Monazite–xenotime geothermometry
Recent experimental and empirical studies have produced several calibrations of a monazite–xenotime solvus geothermometer (Gratz & Heinrich, 1997, 1998; Heinrich et al., 1997; Pyle et al., 2001; Seydoux-Guillaume et al., 2002). The ability to determine both temperature and time from the same monazite provides a relatively well-constrained point on the T–t history of a metamorphic rock (Viskupic & Hodges, 2001). Furthermore, the monazite geothermometer can potentially be applied to metamorphic rocks that lack appropriate minerals for Fe–Mg exchange thermometry. The high closure temperature for REE diffusion in monazite (Cherniak et al., 2004) potentially allows for the recovery of pre-peak metamorphic temperatures. We note that monazite is susceptible to partial (or complete) dissolution and regrowth, and examples of pelitic schist with multiple generations of monazite are well documented (Pyle & Spear, 2003).
Monazite–xenotime equilibrium
Application of the thermometers requires the equilibrium coexistence of monazite with xenotime. Xenotime is present in all samples, though only rarely in contact with monazite; monazite inclusions were also observed within xenotime in one sample. Experimental results in the system YPO4–GdPO4–CePO4 show systematic partitioning of Y and Gd between coexisting monazite and xenotime such that, in monazite, both Y and Gd increase with increasing temperature, whereas in xenotime, Gd decreases and Y increases with increasing temperature (Gratz & Heinrich, 1998).
Given the chemical variations shown in the monazite grains from the Picuris, the potential for disequilibrium was carefully considered. A plot of XYPO4 vs XGdPO4 (Fig. 8a) shows monazite and xenotime compositions for all five samples. The high Gd, low Y xenotime are interpreted as early, relatively lower-temperature compositions; xenotime with low Gd and high Y are interpreted as later, relatively higher-temperature compositions. The presence of monazite with core Gd concentrations higher than most of the coexisting xenotime suggests monazite core growth in the absence of xenotime.
|
Figure 8b and c shows a rescaled view of the monazite compositions from Fig. 8a. Core Gd compositions in monazite from quartzite vary significantly from XGdPO4 = 0·049 to near 0·015, at near-constant XYPO4 = 0·02–0·025. Monazites from schist samples show a similar trend, with XGdPO4 = 0·02–0·04 at near-constant XYPO4 = 0·03–0·035 for the schist samples (Fig. 8b). When core domains and three points interpreted as mixed domain analyses are removed, the intermediate and rim domains show a systematic increase in Gd with increasing Y, suggesting growth in the presence of xenotime (Fig. 8c). Based on these observations only the intermediate and rim temperature estimates are interpreted as geologically significant.
Monazite–xenotime geothermometers
Monazite–xenotime temperature estimates from three calibrations are included in Table 3 for each intermediate and rim spot analysis, and are summarized in Table 5. The thermometer of Gratz & Heinrich (1998) is based upon the partitioning of Gd (XGd = Gd/
REE) between monazite and xenotime, and was experimentally calibrated for the system YPO4–CePO4–GdPO4. Intermediate domain temperatures were calculated by pairing monazite compositions with the highest-Gd (low-temperature) xenotime analysed in the quartzite (XGd = 0·24) and schist sample (XGd = 0·12). Rim domain temperatures were calculated by pairing monazite compositions with the lowest-Gd (highest-temperature) xenotime analysed in the quartzite (XGd = 0·10) and schist (XGd = 0·11) samples. The accuracy of this calibration is given as approximately ±50°C.
|
The second calibration, by Pyle et al. (2001)
, is an empirical calibration (Y+HREE vs T) of the monazite limb of the monazite–xenotime solvus based upon selected monazite–xenotime pairs from New England. The third calibration, by Seydoux-Guillaume et al. (2002), is based upon experimental work in the ThSiO4–CePO4–YPO4 system, and on earlier experimental results from Gratz & Heinrich (1997) in the binary system CePO4–YPO4. Results from Seydoux-Guillaume et al. (2002) suggest that YPO4 in monazite increases with increasing ThSiO4, producing a significant shift in the monazite limb of the solvus, and a narrowing of the monazite–xenotime miscibility gap. The calibration of Seydoux-Guillaume et al. (2002) has no constraints for temperatures below 700°C. The accuracy of these two solvus calibrations is estimated at approximately ±50°C. This uncertainty is a minimum value because of the poorly constrained lower-temperature portions of these calibrations, and the increasing steepness of the monazite limb of the solvus with decreasing temperature.
Comparison of intermediate and rim domain temperatures
The three calibrations show considerable variation between intermediate and rim domains and between quartzite and schist samples (Table 5). Rim temperatures for the schist samples from all three calibrations are reasonable and fairly consistent with phase equilibria and petrographic observations. Rim temperatures from the quartzite samples are more variable with the calibration of Pyle et al. (2001), giving significantly lower temperature estimates. The quartzite samples also yield significantly lower intermediate domain temperatures, with very large standard deviations: 315 ± 50°C and 290 ± 68°C for the calibrations of Pyle et al. (2001) and Seydoux-Guillaume et al. (2002), respectively. These low temperatures are not consistent with monazite growth in the kyanite stability field, and probably reflect the lack of low-temperature data in the calibrations. These geologically unreasonable temperatures make comparison between domains and samples for the calibrations of Pyle et al. (2001) and Seydoux-Guillaume et al. (2002) unreasonable, and further discussion is limited to the results from the calibration of Gratz & Heinrich (1998).
The thermometer of Gratz & Heinrich (1998) yields intermediate domain temperatures of 405 ± 18°C for the quartzite samples and 470 ± 7°C for the schist samples. The difference in temperature is a result of the difference in the high-Gd (lowest-temperature) xenotime compositions measured in the quartzite (XGd = 0·24) and schist (XGd = 0·12) samples. This compositional difference and the corresponding temperature difference may be real. If this is the case, the temperature difference probably reflects a difference in bulk REE chemistry and reaction history. Alternatively, the temperature difference may reflect geological uncertainty in analysing and correlating the proper monazite and xenotime compositions. Xenotime grains are small (5–15 µm long) and much less common than monazite in these samples; a total of eight xenotime grains were analysed in the two schist samples. It is possible that we did not locate and analyse the lowest-temperature, highest-Gd xenotime. If the schist sample did contain xenotime with a Gd concentration comparable with that of xenotime in the quartzite, then the average calculated temperature from 12 intermediate domains would be 422 ± 3°C, indistinguishable within precision from the quartzite sample.
Temperature estimates from the quartzite samples suggest that the intermediate monazite domain formed at temperatures near the kyanite-in isograd. Intermediate domain monazite growth in the schist samples may have occurred at similar temperatures or at somewhat higher temperatures of 470°C. Regardless of this uncertainty, the schist samples yield results consistent with monazite growth at the lower to intermediate temperatures of the kyanite stability field, consistent with petrographic and chemical zoning observations.
Average temperatures for monazite rim domains are 501 ± 18°C and 517 ± 9°C, for the quartzite and schist samples, respectively. The maximum rim temperatures in the quartzite and the schist sample are 536 and 538°C, respectively. These temperatures are indistinguishable within analytical precision and consistent with peak temperature estimates previously determined in this study from phase equilibria. Monazite rim temperatures are also consistent with petrographic interpretations that suggest growth in the kyanite or possibly the sillimanite stability field.
Matrix monazite rims and monazite inclusions within fractured kyanite yield similar temperatures, suggesting partial dissolution and regrowth of monazite in kyanite. In the schist samples, monazite inclusions within andalusite and matrix monazite yield indistinguishable temperatures. The occurrence of monazite aligned in S2, and included within andalusite that overgrows S2, suggests that all monazite growth occurred prior to andalusite. The nearly continuous spread in monazite intermediate and rim compositions (Fig. 8c), and their corresponding temperatures, may reflect a process of dissolution and reprecipitation during progressive deformation and metamorphism rather than distinct monazite-forming events; minor disequilibrium at the thin-section scale may also be a factor. Three monazite rim analyses adjacent to fractures give temperatures of 480 ± 8°C, and may reflect retrograde monazite growth.
Ion microprobe age determinations
Methods
Ion microprobe analyses of monazite were performed using the IMS 1270 ion microprobe at the Northeast National Ion Microprobe Facility at Woods Hole Oceanographic Institution. Raw data were reduced with ZIPS vs 2.4 and results imported into IsoPlot (Ludwig, 1999) for plotting and error analysis. Monazite standard UCLA76 (336 Ma) was used for calibration of masses 232Th, 248ThO, 238U, 254UO–238U, 204Pb, 206Pb, 207Pb and 208Pb. Reproducibility of the standard age based upon nine calibration points is 335 ± 7 Ma. Analytical time was 60 s. No cycles were rejected for the unknowns.
Results
The ion microprobe analyses are summarized in Table 6; all ages are interpreted to represent the time of monazite growth. Two monazites from sample 00–2A and three monazites from sample 00–3C were analysed. Analysed monazite grains and compositional maps are shown in Fig. 9. U, Th and Pb concentrations from the analysed domains are presented in Table 7. All 15 analyses are shown on a U–Pb concordia (upper intercept age of 1417 ± 9 Ma) diagram (Fig. 10). The analysed monazite grains occur as inclusions within andalusite and matrix muscovite. Many of the ion probe analyses sampled across compositional domains and possibly age domains, and we are cautious in using these data to document the relative timing of core, intermediate and rim monazite growth. We note that the younger, discordant ages are typically from the second (or later) spot analysed on each grain, suggesting that sample charging may be responsible for the discordant analyses. 207Pb–206Pb age determinations from concordant and near-concordant analyses yield ages that range from 1434 ± 12 Ma (core) to 1390 ± 20 Ma (rim) and suggest monazite growth occurred over several million to a few tens of million years.
|
|
|
|
These ion microprobe ages are somewhat younger than the
1450 Ma monazite ages reported by Williams et al. (1999), and the garnet (1461 ± 13 Ma) and staurolite (1454 ± 7 Ma) ages reported by Lanzirotti & Hanson (1997) for the southern Picuris. Electron microprobe chemical age determinations from more than 50 additional monazite grains from these samples, including monazite grain ‘n’ included within kyanite (Figs 6a and 7b), yield ages from 1450 to 1390 Ma (Pyle & Daniel, in preparation), and are consistent with the ion microprobe isotopic age determinations. We find no evidence of older, 1650 Ma monazite in any of these five samples. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGIC SETTING OF THE...DATA AND INTERPRETATIONSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Discussion of monazite ages and a proposed P–T–t–D path
The absence of older (
1650 Ma) monazite cores or inclusions requires that (1) we simply missed the older detrital or metamorphic cores or grains; or (2) these rocks experienced only a single, regional, amphibolite-facies metamorphic event beginning at 1450–1435 Ma; or (3) older detrital (>1700 Ma) or metamorphic (1650 Ma) monazite grains dissolved completely prior to, or during, the 1400 Ma metamorphism. Given the X-ray mapping and analysis of more than 50 monazite grains (includes EMP age analyses, from Pyle & Daniel, in preparation), from five samples and two different bulk compositions, we suggest that the first possibility seems remote.
Alternatively, these rocks experienced only a single P–T loop at 1450–1400 Ma, inconsistent with the models of polymetamorphism proposed by Karlstrom et al. (1997, 2004), Pedrick et al. (1998), Read et al. (1999) and Williams et al. (1999). In this case, the older monazite grains from the Tusas peaks may reflect contact metamorphism from 1690 to 1660 Ma plutons. The occurrence of aligned titanhematite within monazite indicates that S1 was well established prior to 1435 Ma. It is possible that this foliation reflects, in part, an older, 1650 Ma deformation; however, the progressive nature of S1 interpreted from the petrographic observations suggests otherwise. We interpret S1, the regional, km-scale, F2 folding and the S2 transecting cleavage (Bauer, 1993) to have formed between 1435 and 1400 Ma. A summary P–T–t–D diagram (Fig. 11) places the monazite time and temperature data into context with respect to the Al2SiO5 polymorph reaction sequence and deformational fabrics.
|
Given recent work by Kopera (2002a
, 2002b) that documents monazite with 1400 Ma rims and either 1650 or >1700 Ma cores from the adjacent Tusas range, we do not rule out the possibility that early (1650 Ma) monazite was present in the Picuris range. If older monazite was present, then the very efficient dissolution of monazite, in the northern Picuris, presents an interesting problem regarding monazite stability and the processes of regional metamorphism and fluid flow. If the northern Picuris range did experience regional metamorphism near 1650 Ma, we suggest that peak metamorphic temperatures were probably much cooler (400°C) than proposed by previous workers (Karlstom et al., 1997, 2004; Williams et al., 1999). We stand by our conclusion that the regional folding and thrusting occurred between 1450 and 1400 Ma.
Recent and continuing studies suggest that deformation related to the 1400 Ma orogenic event is widespread. Proterozoic rocks in the Taos, Cimarron, Tusas, Rincon and Needle uplifts record the penetrative deformational effects of this event (Pedrick et al., 1998; Read et al., 1999; Williams et al., 1999; Daniel & Pyle, 2001; Hallett et al., 2002; Kopera et al., 2002a, 2002b). Regional metamorphism and deformation also extends to the south at least as far as the Manzano uplift (Marcoline et al., 1999; Ralser, 2000). Effects of this orogenic event also extend north into central Colorado (Selverstone et al., 2000; Shaw et al., 2001) and west into northern Arizona (Dumond et al., 2004), where evidence is present for 1400 Ma reactivation of older Proterozoic shear zones.
Crustal evolution of the Picuris from 1690 to 1630 Ma
Our favoured hypothesis is that these rocks experienced only a single major tectonothermal event between 1450 and 1400 Ma. Contrary to earlier tectonic interpretations (Bauer, 1993; Karlstrom et al., 1997, 2004; Williams et al., 1999), we propose that the Vadito–Hondo supracrustal sequence remained at relatively shallow crustal levels (2–8 km) between 1690 and 1630 Ma, and recorded no significant regional metamorphism or ductile deformation related to the Mazatzal Orogeny.
Field relations and proposed emplacement depths for the Cerro Alto metadacite, the Puntiagudo granite porphyry, the Rana quartz monzonite and the Penasco quartz monzonite exposed in the southern Picuris range support this hypothesis. Long (1974) and Bell (1985) reported a thin discontinuous, fine-grained border zone around the Puntiagudo pluton and a relatively uniform grain size approaching this zone. Long interpreted this zone as a chilled margin; however, Bell interpreted the border zone as an area of contact metamorphism. Both studies reported generally discordant intrusive contacts and locally near-concordant intrusive contacts. Long (1974) concluded that the emplacement depth for the Puntiagudo granite porphyry (1685 Ma) was between 1 and 4 km, although Bell proposed a slightly deeper emplacement depth of 6–10 km. Both studies agreed that the Rana quartz monzonite (1673 Ma) displays a chilled margin and sharply discordant contacts, and that emplacement depth was 3–6 km.
The Cerro Alto metadacite (1630 Ma) shows a fine grain size and appears to be interlayered with amphibolites interpreted as basalt flows (Bell, 1985). Both Long (1974) and Bell (1985) estimated emplacement depths no greater than 2 km, and both agreed this unit was probably a subvolcanic intrusion. However, Long (1974) and Bell (1985) disagreed about the timing of the metadacite. Long (1974) interpreted the Cerro Alto metadacite to be the oldest of the igneous bodies, coeval with the formation of the amphibolites in the Vadito Group. Bell (1985) interpreted the metadacite as the youngest based upon an inferred unconformity and a preliminary U–Pb zircon age of 1630 Ma. Based upon the interpretations of Bell (1985), the three intrusions record an overall decrease in emplacement depth between approximately 1685 and 1630 Ma. There is little or no evidence of significant contact metamorphism from these plutons. Long (1974) and Bell (1985) estimated an emplacement depth of 8–13 km for the younger 1435 Ma Penasco quartz monzonite, requiring burial between 1630 and 1435 Ma.
The Puntiagudo quartz monzonite (1685 Ma), Rana quartz monzonite (1673 Ma) and Cerro Alto metadacite (1630 Ma) are all penetratively foliated, commonly display S–C fabrics and mylonitic zones, and are folded by F2 (Bauer, 1993). The younger Penasco granite (1436 Ma) shows much less deformation, characterized by the development of a moderate to weak foliation along the margin of the granite. Based upon these observations, regional deformation in the southern Picuris was probably waning during the emplacement of the 1435 Ma Penasco pluton or partitioned into different structural domains