V23B-1426
Results from the (U-Th)/He dating systems in Japan Atomic Energy Agency
Japan Atomic Energy Agency (JAEA) has jointly set up the lab of the (U-Th)/He dating in cooperation with Kyoto University and National Research Institute for Earth Science and Disaster Prevention. We use the MM5400 rare gas mass spectrometer and the SPQ9000 ICP quadrupole mass spectrometer, belonging to JAEA, and built a new vacuum heater using infrared laser to extract helium. HF decomposes zircon after the alkali-fusion method using XRF bead sampler and LiBO3 in the preparation of ICP solution. Helium is quantified using sensitivity method. Uranium and thorium are using standard addition method. Quantifications of uranium-238 and thorium-232 are only need for parent isotopes to date samples because they are expected that the state of secular equilibrium becomes established and samarium does not compose the samples. At the present stage, we calibrate our systems by dating some standards, such as zircon from the Fish Canyon Tuff and apatite from the Durango, those are the international age standard, and apatite and zircon from the Tanzawa Tonalite Complex, that was dated in Yamada's PhD thesis, as a working standard. We report the results and detailed views of the dating systems.
V23B-1427
Evidence for diffusive loss of cosmogenic 3He during vacuum crushing of mafic phenocrysts
In vacuum crushing is an efficient technique to selectively release the primordial helium component trapped within olivine and pyroxene phenocrysts. However, contrary to previous assumptions, recent studies have shown that this method may cause significant release of matrix sited cosmogenic 3He (3Hec). Because this loss may bias both the determination of magmatic 3He/4He ratios (Yokochi et al., 2005) and the accuracy of 3Hec measurements (Blard et al., 2006), it is essential to understand what mechanism is responsible and under what conditions 3Hec loss is manifest. For this reason, olivines and pyroxenes with various amounts of matrix-sited 3He (from 107 to 1011 at.g-1) were crushed in air or in vacuum using different crushing devices. The sample temperature was carefully controlled during each crushing experiment, and ranged from 25 to 325 ° C. The resulting powders were then sieved to obtain several homogeneous grain fractions ranging between <10 microns and >300 microns. The 3Hec concentrations measured in each fraction clearly show that significant 3Hec loss (>20%) affects only the finest fraction (<10 microns) and, importantly, only under hot conditions (T>300 ° C). Even the smallest fractions (<10 microns) quantitatively retain matrix-sited 3Hec when crushed under cold conditions (T<25 ° C), regardless of the duration and energy of crushing. These results invalidate the previously proposed mechanism that involved spallation tracks and implied a purely grain size control of the magnitude of loss (Yokochi et al., 2005). Moreover, new diffusion experiments were carried out to constrain the diffusivity of matrix-sited helium in crushed olivines. When used to model diffusive 3Hec loss as a function of grain size during crushing, these new data predict the observed release fairly well. Therefore, we conclude that temperature-enhanced volume diffusion is the main mechanism controlling the release of 3Hec during crushing. For future applications, special attention should thus be paid to control both the grain size and the temperature of the sample during crushing. References Blard, P.-H. et al. (2006) EPSL 247, 222-234. Yokochi, R. et al. (2005) G-cubed 6, doi:10.1029/2004GC000836.
V23B-1428
4He Implantation in Natural Diamond: Implications for Apatite (U-Th)/He Thermochronometry
Current apatite (U-Th)/He thermochronometry protocols correct for ejection of α-particles from grain margins. However, the potential for implantation of 4He into apatite grains, from primary or secondary actinide minerals, has received more limited attention. Evidence for significant natural α-fluxes in the near- surface environment is provided by surface feature and He abundance studies on diamond. Intense α- damage induces a green colour centre in diamond, enabling visual assessment of natural α-implantation doses. Diamonds with transparent green coats and/or green spots occur in most primary and detrital diamond deposits worldwide, indicating that α-implantation rates into upper crustal minerals may be more significant than previously envisaged. Experiments on transparent green-coated natural diamonds reveal implanted αHe concentrations up to 0.015 cc/g, attributed to secondary uranium phases deposited by circulating groundwater (Shelkov et al., 1998). Implantation of similar α-dosages into apatite grains would increase (U-Th)/He ages by up to several hundred percent, dependent on α-dose rate, grain dimensions and actinide content. Investigation of actinide-rich granites in Australia has revealed the common juxtaposition of apatite and actinide phases such as monazite and zircon. In addition, secondary actinide-bearing phases (e.g. uraninite) are observed along joints, fractures, miarolitic cavities and weathering fronts, thus providing additional α-sources. These results demonstrate that (U-Th)/He thermochronometry analyses of apatite, particularly from actinide-rich, weathered granites and sediments, need to evaluate the potential for 4He implantation in the near-surface environment. Insight into the extent of this problem may be achievable through multiple analyses of single grains, in situ laser probe analyses, 4He/3He step-heating experiments, abrasion of grains and/or complementary apatite fission track analyses. Reference: Shelkov, D.A., Verchovsky, A.B., Milledge, H.J. and Pillinger, C.T., 1998, The radial distribution of implanted and trapped 4He in single diamond crystals and implications for the origin of carbonado: Chemical Geology, v. 149, p. 109-116.
V23B-1429
Three new ways to calculate average (U-Th)/He ages
Traditionally, the "average" age of multiple (U-Th)/He analyses has been calculated as the arithmetic mean age. There exist at least three alternative methods: (a) in analogy with the fission track method, the pooled age is calculated by adding the respective U, Th and He abundances of several grains together, thereby generating one "synthetic" multi-grain measurement; (b) the isochron age is the slope of helium concentration versus present- day helium-production; (c) the central age is computed from the geometric mean U-Th-He composition. Each of these methods is more appropriate than the arithmetic mean age in certain applications. The pooled age is useful for comparing single-grain with multi-grain analyses, whereas the isochron age can be used to detect "parentless helium". The central age is the most accurate way to calculate a sample average of several single- grain analyses because U, Th and He form a ternary system and only the central age adequately captures the statistics of this compositional data space. Fortunately, the expected difference between the arithmetic mean age and the central age is relatively small, less than 1% if the external age reproducibility is better than 15% (1se). The (U-Th)/He age equation can be visualized on a ternary diagram to illustrate that the alpha-ejection correction should be applied before, and not after age calculation, in order to avoid a partial linearization of the age equation. To facilitate the calculation of the central age, a web-based calculator is provided at \ttt{http://pvermees.andropov.org/central}
V23B-1430
The relationship between garnet growth and MSWDs in garnet Lu-Hf dating
The advent of high-precision Lu-Hf analysis by MC-ICP-MS has led to unprecedented age resolution for garnet. Typical analytical uncertainties of ±0.005% in 176Hf/177Hf combined with a bulk 176Lu/177Hf ratio in metapelitic garnets of ~10 (±0.5% uncertainty) result in theoretical age uncertainties of only ±2 Myr and ±200 kyr for 400 and 30 Ma garnets respectively. Yet some garnet Lu-Hf isochrons have MSWD values higher than desired; traditionally these would be viewed as indicating "bad data" due to such factors as inclusions in the garnet or problems in analytical methods. Although several considerations are important in evaluating Lu-Hf garnet data, here we show that high MSWD's may readily result from typical (c. 10 Myr duration) garnet growth. We modeled growth of 5-10 wt% garnet, assuming a linear increase in garnet volume with time and simple Rayleigh distillation. Growth durations of 5-15 Myr were based on previous publications and on combined theoretical thermal and thermodynamic models. Garnet-matrix partition coefficients were determined from natural garnet core vs. whole rock compositions from metapelites in Great Smoky Mountains, and were 25-100 for Lu and 0.04-0.2 for Hf. The models show the following: (1) Garnets exhibit bell-shaped profiles in Lu, but nearly constant Hf contents. (2) Plots of 176Hf/177Hf vs. 176Lu/177Hf for different garnet splits are concave upward. (3) In metapelites (high Lu/Hf) and amphibolites (low Lu/Hf) modeled MSWD values are commonly >10 and >2 respectively. (4) Higher MSWD's result from better separation of different stages of garnet growth (i.e., the best mineral separation actually results in the worst statistics) and from longer durations of garnet growth. (5) The matrix Lu/Hf ratio evolves because of garnet growth, but the effects on isochrons are minor. (6) The most radiogenic split provides a minimum estimate of the age of garnet nucleation; the least radiogenic split provides a maximum estimate of the termination of garnet growth. Predictions 1-3 are directly corroborated by natural data, supporting the theoretical models and interpretations based on previously collected Lu-Hf data. Prediction 6 implies that high MSWD's are not detrimental, rather the complexities of the Lu-Hf system permit inference of garnet growth rates independently of other isotopic systems, allowing improved interpretation of metamorphic processes.
V23B-1431
Using Thermal Subsidence Of The Seafloor To Validate The Geologic Timescale To 80 Ma
Age calibration of the geologic timescale is based largely on cyclostratigraphy and radiometric ages, but validation (as opposed to calibration) of the timescale is difficult because most time dependent geologic processes that might serve as clocks are frought with ambiguities and uncertainties. An exception is the well- understood thermal-isostatic subsidence of the sea floor after it forms at mid-ocean ridges. To a very good approximation, depth is a linear function of the square root of sea floor age for ages less than about 80 Ma. Known sea floor depths and ages have been widely used to study global and regional subsidence, but if the model parameters (thermal diffusivity, mantle temperature, mantle density, and lithosphere density) can be independently determined, the subsidence model can be used to estimate the age of oceanic basement from its depth t = sqrt\{(d - dr)/S)\} where d is depth, dr is the depth of the mid-ocean ridge crest, and S is the rate of subsidence. We have used reported depths and ages from a suite of 41 sites where normal oceanic basement was penetrated by drilling during the Deep Sea Drilling Project (DSDP) and Ocean Drilling Program (ODP) to test this approach. Basement ages were estimated from magnetic lineations using the geomagnetic reversal timescale and/or from the biostratigraphic age of the oldest sediment overlying basement. The correlation between observed and predicted depths is 0.97, with an RMS difference of 8 Ma. Much of the misfit is from the ten oldest sites; the misfit for 31 sites with basement ages less than 55 Ma is 4 Ma. We conclude that this approach can be used to validate the geologic timescale. The next steps in this study are to update the basement ages based on the most recent calibrations, and to include the temperature dependence of the thermal diffusivity in the subsidence model.
V23B-1432
New Analytical Developments in K-Ca Geochronology Using MC-ICP-MS
While the decay of 40K to 40Ar has been used extensively in geochronology, the other arm of this branched decay, the beta decay of 40K to 40Ca, has received much less attention, despite the system's potential application to important geologic problems. Unfortunately, measuring calcium isotopic ratios has proven to be challenging with conventional single collector TIMS instruments given the relative signal intensity of 40Ca with respect to all other calcium isotopes. Additionally, TIMS analyses take many hours (and a steady signal) to complete and the thermal source does a relatively poor job of ionizing calcium. Here we present a technique that make it possible to precisely and reproducibly measure calcium isotopic ratios in solution using a multicollector ICP-MS. The success of this technique was predicated upon overcoming two significant analytical problems: 1) the interference of argon (the carrier gas) at mass 40 and 2) the general problem of mass distribution of calcium isotopes, in which 40Ca is significantly more abundant than the other naturally - occurring isotopes. Through repeated measurement of 40Ca/42Ca in the NIST SRM 915b we achieve an external precision of ~ 0.05% at the 95% confidence interval. Internal errors on individual analyses are ~ 0.02% at the 2 σ level. We are applying our MC-ICP-MS technique for calcium isotope analysis to K-rich salts and authigenic glauconites of known age, as well as to igneous whole rock and mineral separates with variable K/Ca ratios. Measured 40Ca/42Ca ratios in these materials match well with modeled values based on their ages and K/Ca ratios.
V23B-1433
High-Precision 40Ar/39Ar Geochronology and Geology of St. George Island, Pribilof Islands, Alaska: Implications for Eruption Rates in the Bering Sea Basalt Province
New high-precision 40Ar/39Ar ages and geologic mapping establish an eruptive chronology for St. George Island, Pribilof Islands, Alaska. St. George is part of the Bering Sea basalt province (BSBP), a group of 15 late Cenozoic (mostly < 6 Ma) alkalic to tholeiitic basaltic volcanic fields widely distributed on islands in the Bering Sea, along the west coast of Alaska, and along the coast of northeast Russia. Twelve samples of washed, but otherwise untreated, whole-rock basalts from St. George were cut with a micro-wire saw into chips ~3 mm3 in size and irradiated for 40Ar/39Ar analysis. The chips were incrementally heated with a CO2 laser equipped with an integrator lens, and analyzed using a NU Instruments Noblesse mass spectrometer equipped with a Faraday cup and two ion counting electron multipliers. Detector intercalibration was done using automated air pipettes. A minimum of 20 heating steps were measured per sample, with the data often defining age plateaux. Isochron plots of the data yield ages ranging from 1.57 ± 0.04 to 2.89 ± 0.11 Ma, with trapped 40Ar/36Ar ratios ranging from 312 to 330. The stratigraphic positions of the dated rocks are known directly from field relations and there are no discrepancies between the 40Ar/39Ar ages and this sequence. Geochemical data combined with the age data indicate no progressive petrologic trends during evolution of the magmatic system, except for intermittent eruption of distinctive plagioclase-phyric basalts with low to moderate MgO contents (7 - 5 wt%) beginning at ~2.0 Ma. The new age data combined with volume estimates indicate an average subaerial eruption rate of ~107 m3km-2yr-1, which is adjusted for 3% sedimentary and ultramafic basement rocks beneath the volcanic pile, an average vesicularity of 5%, and an assumed surficial erosion value of 20%. This rate is identical to the estimate (110 m3km-2yr-1) by Mukasa et al. (JGR 112, 2007) for St. George Island. Both estimates, however, do not account for significant coastal erosion and faulting that has modified St. George since the end of volcanism, as evidenced by numerous E\-W to NE\-SW striking fault ridges and nearly continuous marine terraces and wave-eroded sea cliffs that surround the island, the highest of which rise over 300 m above narrow rocky shorelines. In this regard, geologic, structural, and geomorphic relations suggest that as much as 25% of the island (by area) may have been removed due to faulting and coastal erosion, which significantly impacts estimated eruption rates and comparison to other intraplate basaltic volcanic fields in the BSBP and elsewhere.
V23B-1434
Interpretation of 40Ar/39Ar Age-spectra in Low-grade Polymetamorphic Rocks: The Importance of Petrologic constraints
40Ar/39Ar age-spectra of white micas from low-grade polymetamorphic terranes can be difficult to interpret, because these low-grade rocks frequently have multiple generations of cleavage-forming white mica that grew at temperatures below their closure for diffusion of argon (~350°C). Under such conditions, each generation of cleavage-forming white mica will retain its original growth age. In addition, rocks from metamorphic terranes can also contain populations of detrital and diagenetic white micas which also have distinct ages associated with them. A final complication, frequently found in these low-grade polymetamorphic samples is the presence of inseparable, intergrown chlorite and the problems with associated 39Ar recoil during irradiation of the samples, prior to sample analysis. The shapes of 40Ar/39Ar age spectra of white micas separated from such samples are frequently complex and are quite variable in shape. Some age spectra show a steady increase in age with increase in the temperature of release, while others have sigmoidal or saddle-shaped patterns. The age of most of the steps in such age spectra are mixtures of the various age populations of white mica in the sample, and as such are geologically meaningless. Nonetheless, useful constraints on the ages of at least some events recorded in these complex low-grade polymetamorphic rocks can be discerned with the help of petrographic observations, and by placing the results in a spatial context to test for reproducibility and/or the occurrence of meaningful patterns. If no chlorite is present and the sample contains no detrital white mica, an estimate of the maximum age of the youngest component and the minimum age of the oldest cleavage forming white mica populations can frequently be discerned. If the sample contains detrital white mica, a minimum age for its cooling or crystallization can be estimated. The presence of a small amount of intimately intergrown chlorite masks the age of the youngest white mica population, but the minimum age of the oldest generation of white mica may still be estimated. Samples from the Blue Ridge anticlinorium in northern VA indicate cleavage formation during both the Devonian and the Pennsylvanian. Similar, but somewhat more complex samples from the Potomac terrane in the eastern Piedmont of northern VA and MD indicate discrete tectonothermal events in the Ordovician, Silurian, and Pennsylvanian, while samples from the Westminster terrane in the westerm Piedmont of MD suggest cleavage formation in the middle Silurian, the Devonian and the Pennsylvanian. While the results are not of high precision, they are very useful at the orogenic level.
V23B-1435
The new 39Ar/40Ar dating facility of the LSCE, background and performances.
Precise and accurate timescales are increasingly needed in most disciplines of the Earth sciences. To contribute to this challenging task, a new 39Ar/40Ar laboratory, specifically devoted to the dating of very recent (down to 1 ka) volcanic products, has been developed at the LSCE (CEA-IPSL-UVSQ, France). In this contribution, we present the first data obtained from this new facility. The laboratory comprises a VG 5400 mass spectrometer coupled with a high sensitivity and high dynamic range ion counting system. Gas extraction is achieved with a 25 watts C02 laser or a double vacuum furnace depending of the analyzed samples. The full metal vacuum and purification line feature a GP50 and a compact Ti flash getters which permit extremely low blank for all Argon isotopes (e.g. ~ 3.0 10e-19 Moles for 36Ar). Both analytical protocols and hardware were specifically developed and optimized to date extremely young samples. Analytical performances including protocols, flux monitoring as well mass spectrometer discrimination correction method will be presented in the light of data obtained over the last 10 months. All samples were irradiated, under cadmium, in the Β-1 position (~1.0 10e+13 fast n cm-2 s-1) of the 70MWh-1 OSIRIS reactor (Pierre Süe laboratory, CEA-Saclay, France). Irradiation package is composed of home-design Aluminium disks constituting a 4 cm stack (10 to 30 unknowns/irradiation). Analyzed neutrons flux standards indicate less than 1% variation along the 4cm stack and validate the use of this reactor for high-precision 39Ar/40Ar dating. The precision and accuracy of the facility has been checked from cross-comparison of international single grain standards including FCs (28.02Ma), ACR-2 (1.194Ma) and TCR (28.32Ma) using the most recent recommended values for these monitors. A total of 80 grains in two irradiations (10 and 120 minutes) will be presented in details. Results from single-grain analyses agree within errors with those proposed by Renne et al., 1998. These preliminary results show that the new 39Ar/40Ar facility housed at the LSCE meets with state-of-the-art performances required in high-precision 39Ar/40Ar dating of very young sample, as well as for the interlaboratory calibration of irradiation monitors across the whole range of the K/Ar chronometer.
V23B-1436
Multicollector 40Ar/39Ar Dating: Protocols, Reproducibility and Intercalibration of Mineral Standards
The recent availability of commercial multi-collector noble gas mass spectrometers provides new opportunities for improved precision in 40Ar/39Ar dating, particularly for young Quaternary aged samples, where precise measurement of the 40Ar/36Ar ratio is critical. Multicollection reduces the errors introduced through signal decay and sequential measurement of different isotopes species on a single collector instrument. QUADLAB is equipped with a NU-Instruments multicollector Noblesse noble gas mass spectrometer, which is configured with a faraday detector and three ion-counting electron multipliers. The instrument has the capability to measure several noble gas isotopes simultaneously and to change measurement configurations instantaneously by the use of QUAD lenses (zoom optics). The mass spectrometer, laser and extraction line are fully automated using a modified version of the program "MASS SPEC". Detector intercalibration, deadtime determination, instrument mass fractionation and long-term data reproducibility are evaluated by repeated measurement of air aliquots from a calibrated air pipette. We present analytical protocols for the measurement of irradiated samples and results of intercalibration of different monitor minerals. http://www.quadlab.dk
V23B-1437
Advances in Hardware and Data Reduction Protocol in 40Ar/39Ar Dating; On Single and Multi- Collector Mass Spectrometer Systems
The two limiting factors in obtaining accurate high precision 40Ar/39Ar ages are the accurate and precise determination of the mass spectrometer mass discrimination (usually expressed as 1 a.m.u.) and of the 36Ar background corrections. These two corrections have a direct affect on the accuracy of the neutron flux monitor (J which is applied to the unknowns) and the sample unknowns. The mass discrimination correction has the most significant influence on samples with low concentrations of radiogenic Ar because measured 36Ar/39Ar ratio is multiplied by 1 a.m.u. to the third power resulting in a correction of up to 20 per mil to 36Ar/39Ar ratio. This ratio if further increased by the atmospheric 40Ar/36Ar (295) to correct for atmospheric Ar. However, the mass discrimination is not trivial for samples with high radiogenic 40Ar concentrations either because the measured 40Ar/39Ar ratio is directly multiplied by a 1 a.m.u correction, resulting in a 2 to 10 per mil correction to this ratio. In some cases this variation in mass discrimination is hidden in J, however, this is not the case for sample measured at different time frames or methods. Similarly, the 36Ar background corrections has the most significant influence on samples with low concentrations of radiogenic Ar because measured 36Ar/39Ar ratio is multiplied by 1 a.m.u. to the third power and then increased by the atmospheric 40Ar/36Ar (295) to correct for atmospheric Ar. Improvements in the determination of these two corrections, is required to improve the accuracy and precision 40Ar/39Ar dating. We have developed a protocol that closely tracks any changes in the mass spectrometer mass discrimination during the time frame of the experiments thereby improving the accuracy of our age determinations. To further improve the accuracy of the 36Ar measurements, we have also implemented digital ion-counting and multi-collector data acquisition, significantly improving our low-level signal measurements resulting in a major improvement in the accuracy and precision of 40Ar/39Ar age determinations.
V23B-1438
Establishing the Limits of Yellowstone Hotspot Volcanism at the Time of the Steens Mountain Reversal
One of the best records of magnetic field behavior during a geomagnetic polarity reversal is found in the Steens basalts of the Oregon Plateau. Lavas preserving transitional field directions were first discovered near the top of a 1000 m flow-on-flow section at Steens Mountain, Oregon. Other Steens lavas of equivalent age (16.5 Ma) recording transitional field paths consistent with the record at Steens Mtn. are also found at Catlow Peak (70 km to the SSE) and at the Poker Jim Ridge (80 km to the W). These lavas are the earliest eruptions of the Columbia River Basalt Group and field relationships show they are the equivalent to the R0-N0 Imnaha members. Although others have found the Imnaha Basalt to be older than the Steens basalts, our 40Ar/39Ar age of 16.62+-0.28 Ma (2 sigma) from a normally magnetized lava near the top of a reverse to normal (R-N) polarized section of Imnaha Basalt at Squaw Butte, Idaho (225 km to the NE of Steens Mtn.) indicates Steens and Imnaha basalts are coeval. This section is known to have one lava recording a transitional field direction and it is highly likely this flow erupted during the Steens reversal. Therefore, Yellowstone Hotspot Volcanism (YHV) was occurring at locations at least 300km apart during the relatively short 4000 year duration of the Steens reversal. Some authors have suggested extending YHV during the Steens reversal further south. A Steens like series of flows in the Santa Rosa range (125 km to the SE of Steens Mtn.), show a reversal path similar to Steens. Our preliminary 40Ar/39Ar age of 18.09+-0.26 Ma (2 sigma) from a plagioclase separate suggest the lavas are older than the Steens reversal. As the stratigraphy of the section is complicated by many dikes of undetermined age, we have sampled the section to reexamine the reversal path. At Sheep Creek, Nevada (260 km to the SW of Steens Mtn.) a volcanic section with Steens-like basalts record a transitional-to-N geomagnetic polarity change with transitional field behavior similar to Steens. While a date by others (15.58+-.20 Ma, 2 sigma) of whole rock basalt from the bottom lava of the section excludes the possibility that the Steens reversal is found at Sheep Creek, whole rock dating is fraught with problems and leaves open the possibility that eruptions occurred at this distal southern location during the Steens reversal.
V23B-1439
40Ar/39Ar Ages for the Sentinel-Arlington Volcanic Field, Southwestern Arizona
The Sentinel Plains lava field and proximate small (<10 km diameter) shield volcanoes, collectively referred to as the Sentinel-Arlington Volcanic Field (SAVF) are composed of mostly basaltic lava flows with a small percentage of magmatic and phreatomagmatic tephra deposits. SAVF is located ~75 km southwest of Phoenix, Arizona, and covers ~600 km2. SAVF lies on the eastern terminus of the Gila River graben within the Basin and Range physiographic province. A series of northwest-trending normal faults cut across the surrounding terrain, indicating that the loci of the SAVF eruptive centers could be controlled by structural trends. The volcanic centers of SAVF erupted near the Gila River channel, damming and diverting the river at least twice, forming small ephemeral lakes. The relative timing of the SAVF eruptions was determined in order to unravel the SAVF eruptive history as well as the timing of the ancient Gila River interactions that led to the development of the Painted Rock transverse drainage. The absolute timing was determined in order datermine causal relationships with local tectonism. The SAVF basal contact is ~30 m above the Holocene surface where exposed along the current river channel; and the lavas show similar amounts mantling by aeolian dust, development of pedogenic calcium carbonate, and subsequent incision by radial ephemeral drainages. Relative timing of eruptive events was determined by stratigraphic and embayment relationships. Continuity of distal flows exposed in cross-sections to their source vents could be established using field work, and confirmed using geomagnetic secular variation and geochemical analyses. Edifices generally corresponded to discreet geomagnetic inclination, declination, and paleointensity values. Older eruptive events exhibited normal polarity, while stratigraphically younger events exhibited reversed polarity. Most lavas were alkali olivine basalt with a range of unnormalized SiO2 weight percentages ranging from 47.16-51.48. Geochronology using 40Ar/39Ar method revealed an age of 1.94 +/- 0.85 Ma for Painted Rock Low Shield (New Mexico Geochronology Research Laboratory), 1.64 +/- 0.14 Ma for Theba Low Shield (Rutgers University) and 1.24 +/- 0.040 Ma for Wild Horse Low Shield (Rutgers University). Some ages were precise enough to correspond to the Matuyama reversed polarity epoch, with SAVF initiation possibly within the Olduvai normal polarity event. These dates represent an overall improvement in precision and accuracy over previous dates (values corresponding to 6.20 Ma to 1.28 Ma) collected in the late 1970s and early 1980s using K-Ar technique. The 40Ar/39Ar ages correspond to expected magnetic polarities and stratigraphic sequences.
V23B-1440
Lunare Mare Basalt Meteorite NEA003-A: Chronology, Chemical and Petrological Composition
Lunar mare basalt Northeast Africa (NEA) 003-A is an unbrecciated, coarse-grained, low-Ti, low-Al, low-K olivine- rich basalt similar to olivine basalts from Apollo 12 and olivine-normative basalts from Apollo 15. This lunar basalt is chemically and petrographically distinct from the previously described lunar mare basalt meteorites Asuka 881757, Yamato 793169 LaPaz (LAP) 02205 clan, Northwest Africa (NWA) 032/479, Dhofar 287A, Miller Range (MIL) 05035 and NWA 773clan. NEA003-A has a coarse-grained magmatic texture consisting mainly of olivine, pyroxene and plagioclase grains. It presents the second highest modal abundance of olivine and the lowest modal abundance of plagioclase (all converted to maskelynite) compared to other unbrecciated mare basalt meteorite samples and is similar to Apollo 12 and Apollo 15 basalts. The distribution of silicate minerals in NEA 003-A is relatively homogeneous. Only small areas of the sample represent the late-stage mineral association composed mainly of elongated plagioclase, Fe-rich pyroxene and ilmenite, troilite and rare SiO2 phase and K-rich glass. Other minerals typical of the mesostasis, described for lunar basalt samples, (e.g. fayalite, fluorapatite, whitlockite, etc.) are not present in this sample. The conspicuous feature of olivine and pyroxene grains is the presence of numerous cracks and fractures that are likely due to shock. Some of the larger fractures near the surface of the meteorite are filled by the products of terrestrial weathering (mainly secondary Ca-carbonate). Shock event(s) seems to have converted all plagioclase into maskelynite. The Sm-Nd isotope systematics of four mineral fractions composed of different proportions of pyroxene and feldspar have been analysed. All four data points show a good collinear array. If pooled together in one single regression calculation, the result is t = 3.089±0.064 Ga, Ndi = 0.508600±0.000095 (corresponding to εNd3089 = -0.4±0.3), and MSWD = 1.04 (n = 4). The K-Ar systematics are disturbed and suggest an Ar-Ar age of 1.762±0.054 Ga (2σ) for pyroxene separates and of 2.315±0.040 Ga (2σ) and 2.293±0.042 Ga for bulk and maskelynite separates respectively. The incongruent Ar-Ar and Sm-Nd ages suggest prolonged post-shock temperatures (<900°C) that permitted the total to partial loss of the initial radiogenic 40Ar.
V23B-1441
Kalahari 009: One of the Oldest Lunar Mare Basalts - Chronology, Chemical and Petrological Composition, and Source Region
The meteorite Kalahari 009 is a monomict mare basalt breccia consisting of fragments of basaltic lithologies embedded in a fine-grained, heterogeneous matrix. The basaltic clasts have a coarse-grained subophitic texture. The main constituents of the breccia are predominantly pyroxene followed by plagioclase with lesser amount of olivine. Accessory minerals include ilmenite, chromite, troilite, chromian ulvöspinel, baddelyite, phosphates and Fe, Ni metal (having about 0.6 wt% Ni) (Sokol and Bischoff, 2005). The modal abundance is as follows: 50-55% pyroxene, ~35% plagioclase, ~10% olivine and <5% of opaque accessories. Symplectitic intergrowths of hedenbergite\+fayalite\+SiO2, representing late-stage assemblages, are ubiquitous (Sokol and Bischoff, 2005). The lunar origin of this sample is documented by bulk rock chemistry, Fe/Mn ratios in olivines and low 87Sr/86Sr ratio of 0.699242±43 measured in the third leachate of plagioclases (Sokol et al. 2007). Chemically Kalahari 009 has lower chondrite-normalized REE abundances and the lowest Th than the majority of mare basalts (Sokol et al. 2007). This lunar basalt shows positive Eu anomaly, which was previously only observed in some Luna 24 basalts (Ryder and Marvin, 1978) and suggested by Neal and Taylor (1992) to be due to plagioclase contamination of the source, or contamination from the anorthositic wallrock during magma ascent, or due to differentiation of ferrobasalt magma (Laul et al., 1978). Pyroxene compositions show strong variation and follow a typical mare basalt fractionation trend showing compositions in the miscibility gap of the pyroxene quadrilateral as well as zoning indicating and initial rapid cooling. On the other hand, the complete breakdown of the pyroxferoite, exsolution lamellae in pyroxene and the coarse grained nature of the basalt indicate slow cooling (second cooling stage). A crystallisation age of 4.286±0.095 Ga (2σ) was determined by Lu-Hf systematics (Sokol et al. 2007). The Ar-Ar age suggests a major impact at 1.721±0.030 Ga (2σ), and combined with the petrologic observations, it must have been an impact of ~30 GPa followed by an extended post-shock high temperature (<900°C) that caused the resetting of the K-Ar systematics. The very low Th content of Kalahari 009 (lower than that in Luna 24 basalts) points to a possible source far from the Procellarum-KREEP terrain. It is possible that this lunar basalt may come from a source on the lunar farside (e.g. Mare Moscovensis).
V23B-1442
Status Report on the 40Ar/39Ar and U/Pb Dating of Tuffs in the Dewey Lake Formation of West Texas Towards Constraining the Permo-Triassic Magnetostratigraphic Time Scale
A detailed magnetic polarity time scale for the Permo-Triassic Boundary interval, critical for correlating events in marine and terrestrial paleoenvironments, is not yet well-established. Recently, late Permian magnetostratigraphic studies have been reported for non-marine sections in Europe and South Africa (Szurlies et al., 2003; Nawrocki, 2004; Ward et al., 2005). However, these sections are devoid of index fossil suitable for correlation with marine successions and also lack age constraints from radioisotopic dating methods. In other words, it is dubious to correlate these magnetostratigraphic data with the GSSP Permo-Triassic boundary and mass extinction. The Dewey Lake red beds formation of West Texas, believed to be the youngest Permian formation in North America, has yielded high-quality paleomagnetic data (Molina-Garza et al., 1989; Steiner, 2001) and contains several silicic tuffs potentially enabling high-resolution calibration of the magnetic polarity time scale in this critical age range. The tuffs have yet to be placed into a regional stratigraphic or magnetostratigraphic framework, and it is unclear exactly how many distinct eruptive units are represented by the 7 distinct samples collected to date from widely separated (>160 km) localities. 40Ar/39Ar (sanidine and biotite) and U/Pb (zircon) studies reveal that all 7 sampled tuffs were probably erupted within several hundred ka of the Permo-Triassic boundary as dated at the Meishan GSSP section (Renne et al., 1995; Mundil et al., 2004) but results thus far are inadequate to convincingly resolve age differences between the various samples. U/Pb dating of some samples is severely challenged by Pb-loss from the zircons despite application of the Mattinson (2005) annealing/chemical abrasion technique. 40Ar/39Ar data have been obtained from as many as four different irradiations in order to reduce neutron fluence related error. We observe the familiar ~1% bias between U/Pb and 40Ar/39Ar ages. Biotite microprobe data, zircon U/Th TIMS data, and the absence of sanidine from some samples serve to help correlate or distinguish some samples despite irresolvable age differences; existing data suggest that 4 distinct tuffs are present in the Dewey Lake Formation. Resolving their ages convincingly will require further work, but it is clear from our results combined with previous magnetostratigraphic data that magnetic polarity reversals were relatively frequent in the latest Permian. Thus the uniqueness of correlations elsewhere with the Permo-Triassic boundary based on magnetostratigraphy alone are not well-founded.
V23B-1443
Precision and Reliability of 40Ar/39Ar Dating of Thin Tephra in Lacustrine Paleoclimate Archives
Definitive interpretation of paleolake sediments in terms of climate history strongly depends on the availability of a precise chronostratigraphy. In East Africa, lake sediments often contain intercalated ash and pumice layers, which can usually be dated with the 40Ar/39Ar method. However, dating of tephra will provide valuable age control only if (a) fresh volcanic material was deposited into the paleolake at or near the time of eruption, and (b) multiple volcanic events can be dated. The small amount of material typically available when tephra is obtained from sediment cores constrains the dating precision and accuracy, especially on the <100 ka timescale, due to the limited radiogenic accumulation of 40Ar. Total-fusion experiments on sanidine and anorthoclase phenocrysts from two paleolake records in Kenya (the Soysambu diatomite deposit and the Lake Nakuru sediment core) illustrate that considerable effort must be undertaken to optimize dating results. Valuable age control can be obtained from even limited quantities of tephra if sample preparation is conducted carefully and 36Ar-background abundances are minimized during argon extraction.
V23B-1444
NanoSIMS 207Pb-206Pb dating of monazite, xenotime and baddeleyite
A variety of geochronological problems call for in-situ dating with high spatial resolution (i.e., scales of tens of microns or less). Such techniques are particularly important for studies that focus on the relationships between intracrystalline age variations and petrologic or diagenetic processes. Precise, in-situ geochronology of particularly small grains or domains (ca. 1-5 microns) may be feasible with the Cameca NanoSIMS ion microprobe, which has a spatial resolution as good as tens of nanometers and sufficient mass resolution at high transmission to be useful for geochronologic systems involving isotopes of trace elements. We have used the Cameca NanoSIMS 50L housed in the Caltech Microanalysis Center to determine 207Pb/206Pb ages of monazite, xenotime and baddeleyite standards ranging in age from ~500 to 2000 Ma. Initial efforts focused on Pb-Pb geochronology because it is analytically straightforward: Pb is ionized almost entirely to Pb+ and instrumental mass fractionation of Pb isotopes is negligible. We compare our results to previous age determinations made on these standards using TIMS, SHRIMP and electron microprobe techniques. Our NanoSIMS 207Pb/206Pb age of 2047±35 Ma (2σ) for a baddeleyite crystal from the Phalaborwa carbonatite overlaps with the previously determined TIMS and SHRIMP dates of ca. 2060 Ma. We measured an age of 1018±20 Ma on xenotime standard x6413, slightly older than the TIMS 207Pb/206Pb date of 996.7±0.8 Ma. For the high-Th Moacyr monazite standard, variations of common-Pb corrected ages and 206Pb/204Pb ratios for 5 measurements spanning ~100 microns suggest varying amounts of interference from a mass-204 isobar (probably doubly ionized ThNdO2). The single measurement least-affected by this isobar corresponds to a common-Pb corrected 207Pb/206Pb age of 489±92 Ma, within error of the electron microprobe chemical age of ca. 505 Ma. Spatial variation of the mass-204 isobar intensity suggests that Th may be distributed inhomogeneously within the crystal. These results demonstrate that NanoSIMS Pb isotope data can be used to determine accurate and usefully precise 207Pb/206Pb ages, particularly for Proterozoic and older minerals, and that intracrystalline isotopic variations can be detected at scales of several microns. Future incorporation of U and Th measurements is expected to increase the precision of age determinations, particularly for high-Th and younger minerals.
V23B-1445
Isotopic Exchange in Igneous Zircons
Igneous zircons are a powerful tool for determining crystallization age and metamorphic history of rocks spanning Earth's entire geologic history. High concentrations of uranium and exclusion of lead in zircons make the U-Pb system the primary choice for radiogenic dating. This system has the added advantage of revealing potential parent or daughter loss as discordance on a concordia diagram. For a single mineral grain the U-Pb system reveals not only the age, but an estimate of post-crystallization alteration. As the range of isotopic analyses applied to ancient zircons broadens (oxygen isotopes, Lu-Hf), the ability to evaluate mobility in radiogenic systems other than U-Pb becomes more important. Low concentrations of many trace elements in zircon make them sensitive recorders of post-crystallization isotopic exchange, but their behavior is poorly understood. Comparison of the Sr- and Nd-isotopic systematics of zircon and its host rock can be used to evaluate the retentivity of zircon for the isotopes of these elements . We examined zircon populations from granitoids ranging in age from 100 Ma to 3.7 Ga. Zircon populations were dated via SHRIMP and both multigrain zircon separates and whole rock powders were analyzed for trace element concentrations, 87Sr/86Sr and 143Nd/144Nd. Analyses of other mineral separates (feldspar, apatite) allow the construction of whole- rock mineral isochrons. In each sample the zircons display disturbance in both the Rb-Sr and Sm-Nd systems. In fact, when compared to the all the other minerals represented on the isochron, the zircons display the highest amount of discordance. Likely promoted by radiation damage, the degree of open-system behavior of the Rb-Sr and Sm-Nd systems increases with the age of the samples. With greatly increased effective diffusivities and trace element mobility, general acceptance of zircons as impervious isotopic time capsules should be regarded with caution.
V23B-1446
Possible Transitional Process in Fission-Track Annealing of Zircon, Inferred by Remodeling of Laboratory Based Kinetics
Annealing kinetics of fission tracks (FTs) in zircon was remodeled to fit data from 43 laboratory experiments in three previous papers using spontaneous FT lengths in zircon of the Nisatai Dacite. The ranges of heating temperature and time are 350 - 912 deg. C and 0.001 - 10000 h, respectively. Because the transition in annealing process appears at a certain temperature zone in the data plots, two types of models are created to describe the annealing behavior involving the possible transitional process with improved goodness of fit; (a) the hybrid-linear model that consists of the fanning-linear model at low temperature and the parallel-linear model at high temperature, connected with a transitional temperature zone, and (b) the parallel-curvilinear model that traces smoothly the transition in the annealing process. The validity of the extrapolation of these models to geological time scale is assessed by comparison with FT data from some deep borehole core samples. The length estimates by the hybrid-linear model agree well with the borehole data, while the estimates by the parallel- curvilinear model show a significant disagreement with highly annealed borehole samples although this model provides good agreement with the experimental data.