V51C-0583 0800h
Isotopic Geochronology by Means of a Dynamic Simulation Model --- Part I. Uranium Series Method
This paper presents the hydrodynamic implications of traditional approaches to dating minerals and rocks by the U-Pb method, the development of an alternative dynamic simulation model, and the comparison of these two models. The proposed alternative model simulates dynamic mass transport and radioactive decay of the decay-chain members in a crystalline mineral deposit and in the surrounding host rocks. In conjunction with the results of mass spectrometer measurements, the simulation results can be used for geochronological dating of minerals and rocks. Because of the dynamic simulation approach, this model can avoid imposing two crude assumptions that are normally required by the traditional modeling approach. They are: (1) mineral deposits are confined in a closed system and (2) the decay chain is in secular equilibrium at the time of dating. Currently, these crude assumptions still remain controversial amongst geochronological scientists. The proposed simulation model is verified for the cases of diffusion and radioactive decay. Comparison of the simulation results and analytical solutions for the radioactive decay and the 3-D diffusion cases indicates that they are in excellent agreement. Eight scenarios are analyzed to demonstrate the effects of preexisting daughter nuclides, closed system assumptions, and secular equilibrium assumptions. By selecting the most conservative-values as model input parameters, all eight scenarios are analyzed. The results indicate that (1) analysis with preexisting daughter nuclides is always greater than those without, (2) closed system assumptions may considerably overestimate the age of minerals and rocks, and (3) secular equilibrium assumptions will result in an underestimation of the age of minerals and rocks. The results of an existing radiometric dating study conducted for the Alder Creek rhyolite by Getty and Depaolo are compared with the dynamic simulation model using the same mass spectrometer data. The results indicate that the age of the rhyolite is on the order of 11,300 to 11,900 years, as compared to the age of 1,030,000 years reported in Getty and Depaolo's study. This huge discrepancy in the Alder Creek rhyolite chronology indicates that it is necessary to reevaluate the accuracy of the conventional model, which utilizes a closed system assumption.
V51C-0584 0800h
FC-1: a Zircon Reference Standard for the Determination of Hf Isotopic Compositions via Laser Ablation ICP-MS
Zircon was extracted from a sample of gabbroic anorthosite of the Duluth gabbro complex (1099.0 ± 0.6 Ma) at Forest Center, MN. A previous collection from this site has been labeled FC-1 and has been used as a U-Pb age standard for SIMS analyses at the RSES of the ANU. Several other laboratories are also now evaluating FC-1 as a standard for ion probe (SIMS) and LA-ICP-MS methods. Its potential as a Hf isotopic standard for in-situ laser ablation (LA) MC-ICP-MS analyses is reported here. Zircons were divided into 6 fractions based on color and magnetic susceptibility, including: 1) clear, non-magnetic at 0deg, 2) brownish, non-magnetic at 0deg, 3) clear, magnetic at 0deg, 4) non-segregated, magnetic at 0deg, 5) non-segregated, magnetic at 1deg, and 6) non-segregated, magnetic at 3deg. Each fraction was dissolved and Hf purified on BioRad AG50W-X8 resin in 0.05N HF+0.5N HCl. Hf isotopic analyses of all 6 fractions yielded very consistent results with the average 176Hf/177Hf=0.282174 ± 12 (2 S.D.), relative to JMC-475 176Hf/177Hf=0.282162 ± 5 (2 S.D.). Laser ablation analyses of individual grains from the six fractions yielded similar results, with an average 176Hf/177Hf=0.282174 ± 15 (2 S.D.). The ablation experiments were performed on spots and raster lines (both 60 micron beam size), with 8 Hz repetition rate and 70% laser power. These parameters resulted in ion beam intensities around 11-16 V total Hf. The analyses were performed with on-line Lu and Yb isobaric interference corrections, using 176Lu/175Lu=0.02653 and 176Yb/172Yb=0.5870. All isotopic ratios were corrected for mass-bias using 178Hf/177Hf=1.46718. Although the Yb correction on mass 176 is significant (ca. 6% to 16%), the average Hf isotopic ratios obtained by LA are in excellent agreement with the average of the wet plasma analyses of purified Hf. No correlation between 176Hf/177Hf and Lu/Hf ratio is observed, suggesting no detectable inhomogeneity in the Hf isotopic ratios due to Lu decay. These results indicate FC-1 zircons that are 3 degrees magnetic or less are isotopically homogeneous with respect to Hf regardless of color or shape. This means that ample (about 20 milligrams per kg rock) homogeneous zircon can be separated from this sample by standard density/magnetic methods and that FC-1 will be a useful inter-laboratory standard reference material for hafnium isotopic measurements using laser ablation. FC-1 appears to be equivalent to other reported standards in terms of isotopic homogeneity and Lu/Hf ratio (e.g., Temora-1 and 2 and 91500 zircon reference standards) and will be further evaluated for its Hf and REE content.
V51C-0585 0800h
Ultra-high excess argon in kyanites - 15 Ga Ar/Ar age -
Almino-silicate of Al2SiO5 has crystalographicaly three different phases of andalusite, sillimanite and kyanite. Kyanite as high pressure phase has been an important key to indicate high pressure condition of metamorphic rocks in comparison with the andalusite - sillimanite type of metamorphic sequence. The Abukuma mountains in northeastern Japan, a typical andalusite - sillimanite type (Miyashiro, 1958) has been under debate from a geological viewpoint as the kyanite - staulorite assemblage has been observed by Kano and Kuroda (1968) who proposed a polymetamorphism of the sequence. Hiroi et al. (1998) also found kyanite included in garnet, and described Cretaceous rapid loading and unloading of the sequence in a high temperature environment. Although the kyanite-bearing rocks are extremely rare in the sequence, kyanite and staurolite have been observed commonly in the river sand, suggesting a paleo Barrovian type of metamorphic terrain. The similar approach to find a Barrovian terrain has also been carried out in the Kitakami mountains by Uruno and Kitakami River Sand Recearch Group (1997) who found new occurrence of kyanite and staurolite in the river sand around the Tono granitic mass 200km north of the Abukuma mountains. Thus, the dating of kyanite may make a constraint for the Barrovian type metamorphism. Kyanite was concentrated from the river sand in the Kitakami mountains using the systematic acid treatment and heavy liquid technique, and finally was handpicked under microscope. They occur generally as a single crystal with clear and planer shape with cleavage and sometimes as a crystal aggregate of kyanite. Ar-Ar analyses of each kyanite crystal were carried out using laser fusion. The obtained ages are 7.7+-0.4, 9.9+-0.4, 11.1+-0.4, 15.1+-0.7 and 16.3+-1.5Ga which are two or three times older than the age of the earth. The ages older than the earth's have been reported from the diopsides from eclogites, ultramafic rocks and Zaire cubic diamonds (e.g., Dalrymple and Lanphere, 1969; Kaneoka, 1974; Ozima et al., 1989). These materials including kyanites are extremely low in potassium, and could be affected strongly by excess argon. The kyanites having 15Ga indicate that they occurred in the ultra-high excess argon environment of the crust in comparison with the mantle. This specific environment suggests that kyanite crystallizes under a ultra-high argon pressure in the old rocks, which come from some potassium bearing phases of low argon retentivity. The recent information for the ultra-high pressure rocks and the associated gneisses provide us a new concept on metamorphic terrain in the world that some of Barrovian type rocks were retrograded from the ultra-high pressure ones. The kyanite of 15Ga may be recrystalized under a specific environment with ultra-high argon pressure derived from radiogenic argon in phengites during the Barrovian type retrogression of UHP rocks.
V51C-0586 0800h
A Reconnaissance Investigation of Isotopic Heterogeneity in Slowly-Cooled Plutonic Rocks of the Laramie Anorthosite Complex, Wyoming
The slowly-cooled, unmetamorphosed 1.43 Ga Laramie anorthosite complex in southeastern Wyoming offers a unique opportunity to investigate the possible extent of isotopic heterogeneity or disequilibrium present within plagioclase-rich plutonic rocks. The isotopic and trace element characteristics of four different magnetic fractions of feldspar separates from four different rocks were studied, including plagioclases from an anorthosite, leucogabbro, and troctolite, and mesoperthite from a monzosyenite. The Pb and Sr isotopic compositions were determined by MC-ICP-MS and TIMS, respectively. For the plagioclase, high concentrations of incompatible trace elements and the absence of inclusions in the less magnetic fractions suggest that these grains are from the rims of the crystals, while the inclusion-rich more magnetic fractions with lower concentrations of trace elements represent the cores. Stepwise leaching of the feldspars removed radiogenic Pb produced by the in-situ decay of U, so that the Pb isotopic composition of leached feldspars could be used to approximate the initial Pb isotopic composition of the whole rocks. The Pb and Sr isotopic compositions of the leached feldspars is distinctly lower than the isotopic compositions of the non-leached feldspars. The Pb isotopic compositions for the non-leached separates from the three plagioclase-rich samples span a large range of $^{206}$Pb/$^{204}$Pb (16.9711-17.2909, 16.8404-18.1773, and 17.0903-17.7679). In contrast, there is no significant Pb isotopic variation in the magnetic fractions of mesoperthite from the monzosyenite ($^{206}$Pb/$^{204}$Pb = 16.7365-16.7694). The less magnetic non-leached fractions are systematically more radiogenic than the more magnetic fractions, reflecting the higher concentration of U incorporated into the rims of the grains at the time of crystallization. The Pb-Pb age (1487 $\pm$ 62 Ma) of the non-leached plagioclases from the leucogabbro are within error of the crystallization age (1436.2 $\pm$ 0.6 Ma; U-Pb baddeleyite/zircon), indicating that the plagioclase crystals have remained a closed system since crystallization of the intrusion.
V51C-0587 0800h
Oxygen Isotopic Composition and U-Pb Discordance in Zircon
Zircon discordance is a common phenomenon in U-Pb geochronology, leading to difficulty in age assignment and uncertainties regarding geologic interpretation. Commonly invoked explanations for discordant zircon analyses include episodic lead loss and continuous lead diffusion. A likely mechanism for lead loss is via a fluid phase. In this study, we attempt to document fluid interaction with zircon through the use of high-spatial resolution oxygen isotopic measurements performed by ion microprobe. $\delta$$^{18}$O was analyzed in both concordant and discordant zircon grains, thus providing an indication of the relationship between discordance and $\delta$$^{18}$O. Results indicate that three characteristics of zircon appear to be interrelated: (1) U-Pb systematics and the associated age discordance; (2) $\delta$$^{18}$O and water-rock interactions that are implied therein; and (3) zircon texture as revealed by cathodoluminescence and BSE imaging. The key observation is that U-Pb disturbed zircons are often also depleted to various degrees in $^{18}$O. However, the relationship between discordance and $\delta$$^{18}$O is not systematic, as $\delta$$^{18}$O values of discordant zircons are lower, but irregular in their distribution. Textural differences between zircon grains also correlate with both U-Pb discordance and $\delta$$^{18}$O. Discordant grains exhibit either a recrystallized, fractured or strongly-zoned CL texture, possibly resulting from metamictization, and are characteristic of a lowered $\delta$$^{18}$O value. Concordant grains, in contrast, have less-expressed zoning and a smoother CL texture, and exhibit higher $\delta$$^{18}$O values. We interpret this to mean that various stages of water rock interaction, as evidenced by $\delta$$^{18}$O analyses, have the ability to leave their imprint on both the texture and U-Pb systematics of a zircon.
V51C-0588 0800h
Re-Os Isotopic Systematics of the Taitao Ophiolite, Southern Chile
The Os isotopic compositions of some ophiolitic materials may record the Os isotopic composition of the convecting upper mantle through time. Interpretation of Os data for old ophiolites can be compromised by post-formation open-system behavior, e.g. recent Re gain or loss. Such effects are particularly exacerbated in mafic rocks that dominate the upper crustal portions of most ophiolites. The high Re/Os of old mafic rocks also require large corrections in calculating initial $^{187}$Os/$^{188}$Os. Study of a very young ophiolite should reduce some of the age inherent issues. The Taitao Ophiolite is located just 50 km south of the present-day Chile Triple Junction and may be less than 10 Ma. We have examined a variety of ultramafic and mafic rocks from a large areal sampling of the ophiolite. The calculated initial $^{187}$Os/$^{188}$Os ratios of most mafic rocks are substantially enriched relative to chondritic, suggesting a transfer of radiogenic Os to the rocks or their mantle sources from either subducting crust or seawater. Peridotitic rocks all have Os concentrations consistent with concentrations present in the upper mantle. The calculated initial $^{187}$Os/$^{188}$Os ratios of these rocks range from 0.1236 to 0.1260. Such a range of isotopic compositions for a single ophiolite is typical and may reflect melt enrichment processes or fluid transport of radiogenic Os. Linking the Os isotopic compositions of whole rocks with Sr and O will help to constrain the processes involved. Additional analysis of separated chromites will help to define the absolute range in Os isotopic compositions throughout the ophiolite.
V51C-0589 0800h
Oxygen Isotope Fractionation between Synthetic Aragonite and Water: Influence of Solution Chemistry
Oxygen isotope analyses of natural aragonite, of both biogenic and abiogenic origin, are used frequently in studies of paleoclimatology, oceanography and carbonate diagenesis, but few experimental determinations of the temperature dependence of the oxygen isotope fractionation between aragonite and water have been carried out. In an attempt to obtain precise and reliable equilibrium oxygen isotope fractionation factors for the aragonite-water system and to resolve some of the inconsistencies in the literature, we developed a number of methods of slow precipitation of inorganic aragonite from Na$^{+}$-Ca$^{2+}$-Cl$^{-}$-HCO$_{3}$$^{-}$ solutions under laboratory-controlled conditions. On the basis of principles established in previous work (Kim and O'Neil, 1997), precipitation was assumed to occur at isotopic equilibrium and measurements were made of the aragonite-water fractionation factor as a function of temperature (5 - $40\deg$C). In addition, the influence of Mg$^{2+}$ concentration (20 to 100 mM) and pH ($\sim$8.2 to $\sim$10.7) of the precipitating solution on the oxygen isotope fractionation between aragonite and water at $25\deg$C was investigated. Under our experimental conditions, the measured aragonite-water oxygen isotope fractionation factor was independent of both the Mg$^{2+}$ concentration and pH, and the temperature dependence of the aragonite-water fractionation was similar to that of the calcite-water fractionation. More importantly, aragonite was always enriched in $^{18}$O relative to calcite when published aragonite or calcite acid fractionation factors were applied. Based on results of equilibrium experiments and the published acid fractionation factor for aragonite ($\alpha$=1.01034), a new expression is proposed for the oxygen isotope fractionation between aragonite and water at low temperatures: 1000ln$\alpha$(Aragonite-H$_{2}$O) = 17.32 (10$^{3}$T$^{-1}$) - 29.00 {\bf Reference} Kim S.-T. and O'Neil J. R. (1997) Equilibrium and nonequilibrium oxygen isotope effects in synthetic carbonates. {\it Geochim. Cosmochim. Acta} 61, 3461-3475.
V51C-0590 0800h
High-Precision Isotopic Analysis of Nanomole Quantities of Silicate Nitrogen
The N isotope system shows exciting potential to trace interactions among Earth's major reservoirs, in particular providing a unique tracer of the transfer of organic components through a wide range of biogeochemical processes. However, sample size requirements of conventional dual-inlet viscous-flow mass spectrometry limit routine N isotope work to relatively high-N systems. Analyses of smaller N$_{2}$ samples ($<$ 0.5 micromoles) are possible by static mass spectrometry, but in only a small number of labs and with some compromise in precision. We have developed a system for analyzing nanomole quantities of silicate N by carrier-gas methods, through combination of a metal high-vacuum extraction line fabricated at Lehigh University and a commercially available continuous-flow, gas chromatography interface (Finnigan Gas Bench II). Testing thus far has concentrated on analyses of aliquots of air and standard tank gases, and application of sealed-tube methods in the production of small N$_{2}$ samples from silicate standards (fuchsite, buddingtonite, whole-rocks). Procedural blanks of this method, employing 6 mm (o.d.) quartz tubes heated, then cracked onto the extraction line, are presently 0.5 to 1.0 nmoles. Practical minimum sample size, taking into account blanks and factors affecting N$_{2}$ transfer, is now approximately 70 nmoles, more than one order of magnitude smaller than that previously possible in our laboratory using dual inlet methods and a cold-finger/microvolume (for our previous method, procedural blanks are about 0.025 micromoles, and minimum sample size is about 1.0 micromole). Precision of the carrier-gas method for the silicate standards (1$\sigma$ for n $>$ 3) is at present 0.1 to 0.2 per mil and mean isotope values are within 0.1 per mil of accepted values based on previous analyses by dual-inlet mass spectrometry. Our ongoing research is aimed at reducing the minimum sample size to approximately 10 nmoles, further reducing the blanks, addressing some remaining minor nonlinearities in m/z 28 beam intensity and isotopic composition with N$_{2}$ sample size, incorporating IR laser heating in extractions, and producing silicate N-isotope standards. The small sample capability of these methods will afford analyses of microgram-sized samples of mineral separates (micas, ammonium-rich K-feldspars), thus allowing higher spatial resolution and opening up many new avenues of investigation previously impeded by the absence of sufficiently N-rich materials. Particular themes in our N isotope work include tracing organic signatures into the crust and mantle and understanding modern and ancient Earth volatile cycling and atmosphere evolution.
V51C-0591 0800h
Li Isotope Heterogeneity in Mantle-derived Xenoliths.
Variations in 7/6Li of mantle-derived rocks are a potentially powerful tracer of mantle geodynamic processes. Analysis of 7/6Li by secondary ion mass spectrometry (SIMS) avoids surface contamination and permits assessment of sample heterogeneity at high spatial resolution. We performed Li abundance and isotope analysis by SIMS of mantle-derived olivine and cpx in order to assess intragranular isotopic variability and explore the Li isotope effects accompanying contrasting types of metasomatism of Archean mantle lithosphere. Analytical techniques are described in the accompanying abstract by Hervig et al., with further investigations into matrix effects on instrumental fractionation in progress. The mineral grains were extracted from peridotites and megacrysts erupted in southern African kimberlites and basalts from the SW USA. Samples from southern Africa selected to represent varying degrees of reaction between refractory Archean lithosphere and deep-seated kimberlitic precursor melts revealed a spread in $\delta$7Li of ~20 per mil, correlated with indices of metasomatism. Assuming a calibration based upon San Carlos olivine (Seitz et al., Lithos, in press), Archean mantle olivine is characterized by 7/6Li below the MORB range. Within this group, high Li concentrations occur in olivines from pyroxenites and low Li in those from refractory dunites, consistent with Li introduction during ancient metasomatism. In contrast, Fe-rich olivine and cpx megacrysts crystallized from Mesozoic kimberlite precursor melts are comparatively Li-rich, with $\delta$7Li up to +16 per mil. Metasomatized, sheared peridotites span an intermediate range of compositions, consistent with the introduction of Li with high $\delta$7Li during reaction with asthenospheric megacryst magmas. Some megacrysts show extreme internal isotopic heterogeneity, up to 15 per mil, and olivines from peridotites are commonly zoned towards grain margins and internal cracks. We suspect these effects are due to recent metasomatism as well as late-stage Li diffusion during entrainment.
V51C-0592 0800h
Disturbed Sr and Nd Isotope Systematics in Zircons With Concordant SHRIMP U-Pb Ages
Little is known about the Sr- and Nd-isotopic systematics of zircon. With slow diffusion rates and a high resistance to weathering, zircon should preserve accurate age information and initial Sr and Nd isotopic ratios. As a common accessory mineral, it could provide petrogenetic information for rocks that have been altered, weathered, or metamorphosed. We have investigated the Sm-Nd and Rb-Sr systematics of zircons from unmetamorphosed granitic rocks that have yielded concordant U-Pb SHRIMP (Sensitive High Resolution Ion Microprobe) ages and have depleted mantle signatures for Nd and Sr isotopes. Zircon populations from mantle-derived igneous rocks with ages of 0.1, 1.7, and 3.8 Ga were chosen for Sr and Nd isotopic analysis. Low concentrations (Sr, 4 to 8 ppm and Nd, 6 to 12 ppm) and small grain size necessitate the use of multigrain aliquots. Meaningful results can be obtained only if all of the zircons in the rock are a coherent population with homogeneous ages throughout and among grains. Zircon U-Pb ages were characterized using the SHRIMP RG, and trace element concentrations were measured by LA-ICPMS. The populations are homogeneous and the material ablated by the ion beam ($\sim$~20 $\mu$m spot size) shows little evidence of lead loss. Results on zircons of 100 Ma and 1700 Ma indicate that both the Rb-Sr and Sm-Nd systems have been severely disturbed. For the 1700 Ma granitic rocks from the Yavapai sequence of Arizona, zircon Sm-Nd apparent ages are ca. 1000 Ma! Leaching was used to remove contributions from adhering or included minerals, but leached residues that presumably most closely approximate the composition of the pure zircon (e.g. have high Sm/Nd) are no less disturbed than unleached samples. Despite the U-Pb SHRIMP ages indicating a closed system, the zircons have failed to preserve a reasonable age or initial isotopic composition for Sr and Nd, indicating that parts of the crystal might be severely affected by radiation damage resulting in disturbed isotopic systems. The results are not compatible with ultra-low diffusivities in these natural zircons, and cast doubt on the reliability of zircons for determining geochemical parameters other than U-Pb age. To further evaluate these results, high T annealing followed by chemical abrasion of the zircons is being applied in order to remove radiation-damaged, disturbed zircon domains and allow more direct comparison of the Sm-Nd and Rb-Sr systems to U-Pb.
V51C-0593 0800h
SIMS Analyses for Li Isotope Ratios: From Olivine to Clay Minerals.
Secondary ion mass spectrometry (SIMS) is an extremely sensitive instrument for Li (up to ~4% of the atoms of Li sputtered from minerals are detected). While precision is high, the accuracy of Li isotope analyses by SIMS is critically dependent on the homogeneity of bulk-analyzed standards. We have used the Cameca ims 3f and 6f SIMS at Arizona State University to analyze crystalline and glassy bulk-analyzed materials and report on their homogeneity and how well the samples compare. We used a 5-20 nA (Ip) primary beam of O- and detected positive secondary ions with 0$\pm$20 eV excess kinetic energy. Mass resolving power was sufficient to separate 6LiH from 7Li, but the hydride species was never observed. 7/6 ratios on samples containing 1 ppm Li show precisions of $\sim$1.3$\permil$ (2 $\sigma$) in a 1 hour analysis. There is a strong positive correlation between Ip and 7/6 ratio in hydrous rhyolite glasses, indicating possible charge-driven Li diffusion. This relation is subtle in low-H basaltic glasses and not observed in olivines and clay minerals. The latter phases show minimal variation in the absolute 7/6 ratio measured by both SIMS instruments over a period of months. Intra-granular Li isotope heterogeneity, apparently related to microstructural imperfections, was observed in some cpx samples, and some olivines from mantle peridotites show strong zoning ($>$10$\permil$). These observations underscore the importance of developing microanalytical techniques of Li isotope analysis. Comparing bulk analyses of San Carlos olivine (Seitz et al., Lithos, in press) with our microanalyses provides a preliminary calibration of the SIMS for 7/6 ratios in olivine. Indirect evidence suggests this calibration applies between Fo80-95. When applied to SIMS analyses of MORB glasses (average $\delta$7Li $\sim$+3), the olivine calibration gives a 7/6 ratio $\sim$4$\permil$ heavier than expected. Using a homogeneous cpx as a standard produces MORB analyses near +3$\permil$. Whereas studies of more samples are required to understand matrix effects, these appear to be minor for many materials of interest.
V51C-0594 0800h
Nitrogen Isotopes in Olivine Separates from Volcanic Arcs, Hot Spots and Continental Mantle Xenoliths
We report the first nitrogen isotopic data of olivine separates from volcanic arcs (Cerro Negro, Nicaragua; Izalco, El Salvador; Turrialba, Costa Rica; Ichinomegata, Japan). In addition, we report nitrogen isotopic data of olivine separates from ocean islands (Hawaii, Reunion, Iceland) and continental mantle xenoliths (San Carlos, Arizona). Samples were processed by crushing and analyzed using a modified noble gas mass spectrometer (VG3400). N concentrations range from 0.6 to 22 micro ccSTP/g olivine. The $^{15}$N/$^{14}$N ratios (expressed in the $\delta$$^{15}$N notation where $\delta$$^{15}$N sample = {[($^{15}$N/$^{14}$N)sample/($^{15}$N/$^{14}$N)Air]-1} X 1000) of olivine separates are distinctly different from air (0.0$\permil$) and range from lower than mean MORB (- 5 $\permil$) to values characteristic of (subducted) oceanic sediments (+ 7 $\permil$). Positive $\delta$$^{15}$N values are found in olivines from volcanic arcs: Cerro Negro 1992 ash (+ 6.2 $\pm$ 1.6$\permil$), Izalco lava flow (+ 5.1 $\pm$ 0.7$\permil$), Ichinomengata spinel lherzolite (+ 1.1 $\pm$ 0.5 $\permil$) with the exception of Turrialba lava (- 1.7 $\pm$ 2.5$\permil$). Olivines from hot spots have both positive and negative $\delta$$^{15}$N signatures: Iceland, Theistareykir - northern rift zone (- 8$\pm$ 1.6 $\permil$), Hawaii, dunite from 1801 Kaupulehu flow of Hualuai volcano (+ 3.1 $\pm$ 0.3 $\permil$) and Reunion dunite (+ 0.2 $\pm$ 0.5$\permil$). The San Carlos mantle xenolith has a value of - 1.5 $\pm$ 2.5$\permil$. $^{40}$Ar/$^{36}$Ar ratios of the samples as determined in this study or reported in the literature are significantly higher than air (295.5) in olivines from Ichinomegata, San Carlos, Iceland, Reunion and Hawaii. The olivines from Cerro Negro have a $^{40}$Ar/$^{36}$Ar ratio of 306, close to that of air. The $^{3}$He/$^{4}$He ratios of the samples are higher than the MORB value of 8.0 R$_{A}$ (R$_{A}$ is the $^{3}$He/$^{4}$He of air), the exception being Cerro Negro (6.1 R$_{A}$). Hawaii, Reunion and Iceland have $^{3}$He/$^{4}$He of 10.3, 12.9 and 12.3 R$_{A}$, respectively. $\delta$$^{15}$N signatures of fumarole gas samples collected at Cerro Negro (+ 4.9 $\pm$0.1 $\permil$), Turrialba (- 1.0 $\pm$0.3 $\permil$) and localities close to Izalco (+ 4.0 $\pm$0.4$\permil$) are identical (within error) to those of the olivine separates. The results obtained in this study are significant and show that 1) N isotopes in olivine separates and volcanic/hydrothermal gas emissions both sample volatiles that are primarily derived from the magma source. 2) subduction of oceanic sediments and transfer of N through the mantle wedge controls the N isotopic composition in volcanic arcs as sampled by olivine separates and gas discharges. 3) the N isotopic signature of hot spot related volatiles is variable but the samples from Hawaii, and to a lesser extent from Reunion, support the idea that surficial N (from subducted sediments or oceanic basement) is recycled into the lower mantle. 4) N as sampled by the San Carlos mantle xenolith may record a greater extent of air contamination than the other samples. Alternatively, N data of San Carlos and the Iceland northern rift zone suggests that the N isotopic signature of the upper mantle may be more variable than previously inferred from MORB samples.
V51C-0595 0800h
The Impact of Diagenesis on the Sm-Nd Isotope Systematics of Black Shales
We investigate the impact of diagenesis on Sm-Nd systematics in black shales. Here we report Nd-isotopic compositions of different isolated shale components (organic, silicate, carbonate, sulfide) using thermal ionization mass spectrometry (TIMS). This study uses time correlative black shales with different grades of thermal maturity from the Utica Shale Magnafacies (middle Ordovician) of Quebec (immature: Tmax 20$^{0}$C to 50$^{0}$C Ro $<$0.5), Ontario (mature: Tmax 50 $^{0}$C to 140 $^{0}$C Ro 0.5 - 0.8), and New York (post mature: Tmax $>$140 $^{0}$C Ro $>$1). The units chosen for this study share a common provenance and range in thermal maturity from: Quebec (immature: Tmax 20$^{0}$C to 50$^{0}$C), Ontario (mature: Tmax 50 $^{0}$C to 140 $^{0}$C), and New York (post mature: Tmax $>$140 $^{0}$C; Hannigan and Basu, 1998). $\epsilon$Nd(0) values for the whole rock samples range from -15.3 to -9.0 for the immature shales and from -19.2 to -12.9 for the mature shales. The $\epsilon$Nd(0) values for the organic fractions range from -12.5 to -12.7 for the immature shales and from -23.5 to -14.0 for the mature shale samples. The carbonate fractions have $\epsilon$Nd(0) values ranging from -7.7 to -11.6 for the immature shales and from -10.7 to -12.5 for the mature shales. T$_{DM}$ values for the whole rock samples range from 2.13 to 2.15 Ga and 1.9 to 2.6 Ga for the immature and mature shales, respectively. $\epsilon$Nd(0) values for the immature shales become progressively more negative from the carbonates, organics to the whole rocks However, for the mature shales, the organic fraction shows most negative values followed by the whole rock and the carbonates. Our results suggest that during early diagenesis (thermal maturation of organic matter and formation of authigenic mineral phases) of black shales there is a clear fractionation between Sm and Nd - Nd being more mobile. Finally, the T$_{DM}$ model ages of all the whole rock samples both mature and immature indicate late Archean to early Proterozoic provenance ages distinctly older than Grenvillean sources.
V51C-0596 0800h
In-situ Production Rates of $^{53}$Mn in Antarctic Rocks
We present first successful measurements of terrestrial $^{53}$Mn from whole rock and pyroxene separates from surfaces with long-term exposure in the Antarctic Dry Valleys. The half-life for $^{53}$Mn (t$_{1/2}$ $\sim$3.7 Myr)is much longer compared to $^{36}$Cl (t$_{1/2}$ $\sim$301 Kyr), $^{10}$Be (t$_{1/2}$ $\sim$1.5 Myr) and $^{26}$Al ($_{1/2}$ $\sim$720 Kyr). Previous surface exposure ages based on $^{3}$He, $^{21}$Ne and $^{10}$Be from these rocks can be used to cross-calibrate the $^{53}$Mn production rate (PR) with modeled $^{53}$Mn production rates. With its half-life of $\sim$3.7 Myr and ubiquitous parent element (Fe), $^{53}$Mn may soon find a broad application to surface dating. Due to its long half-life, $^{53}$Mn is a valuable tool to quantify earth surface processes on time scales up to 10 Myr. $^{53}$Mn is a single target (Fe) product, therefore the production pathway and thus the PR model calculations are relatively simple. Because Fe is abundant in most rocks, whole rock samples can be processed on time scales exceeding the $^{36}$Cl method. Furthermore, the chemical separation of $^{53}$Mn from rocks is simple and fast compared to other cosmogenic nuclide methods such as $^{10}$Be, $^{26}$Al and $^{36}$Cl. Relatively high energies and advanced detector systems are indispensable for separating $^{53}$Mn from its isobar $^{53}$Cr. Furthermore, low production rates in terrestrial systems require a high analytical sensitivity. Recent advances at the tandem accelerator of the TUM and LMU in Munich have allowed an efficient suppression of $^{53}$Mn's interfering isobar, $^{53}$Cr, and reduced the detection limit for the $^{53}$Mn/$^{55}$Mn ratio to 10$^{-14}$. Our measurements confirm that the measurements of $^{53}$Mn are possible. All $^{53}$Mn data are normalized to a well-known meteorite standard. First results for six measurements in four Antarctic dolerites give promising results. The samples were taken from dolerite boulders or bedrock surfaces at elevations ranging from 1145 to 2555 m altitude. The Fe and Mn concentrations, used for activity calculations, were measured by atomic absorption spectroscopy or spectrophotometry. The terrestrial $^{53}$Mn activity of the whole rock sample and a pyroxene separate from the same rock are strikingly consistent with one another, demonstrating that $^{53}$Mn surface exposure dating (SED) is feasible for whole rock samples. Two independent chemical preparations of pyroxene sample from the same rock agree, suggesting that with increased counting times, the uncertainties of the $^{53}$Mn determinations should be reducible to about 10% or less. The modeled value for production rates in pyroxenes at 2555 m altitude is 1.7 dpm/kg Fe. This agrees well with the measured production rate of 1.66 dpm/kg Fe.
V51C-0597 0800h
Lithium Isotopic Composition of the Deep Continental Crust
We have investigated the Li isotopic composition of the deep continental crust by measuring composite samples from Archean high-grade metamorphic terranes in East China and granulite-facies xenoliths from East China (Hannuoba suite) and Queensland, Australia (Chudleigh and McBride suites). The 30 composite samples, including TTG gneiss, amphibolites and felsic to mafic granulites, have a narrow range of $\delta$$^{7}$Li values from +1.7 to +7.5$\permil$, with an average of +4 $\pm$ 1.4$\permil$ (1$\sigma$), which is indistinguishable from the upper mantle. In contrast, the three granulite xenolith suites display a much larger range in $\delta$$^{7}$Li, from +15.7 to -17.9$\permil$ with average values decreasing in the order: Hannuoba (-0.7 $\pm$ 4.9$\permil$ (1$\sigma$, 18 samples)), Chudleigh (-2.5 $\pm$ 5.5$\permil$ (1$\sigma$, 14 samples)) and McBride (-3.8 $\pm$ 7.6$\permil$ (1$\sigma$, 12 samples)). The xenoliths are, on average, lighter than the high-grade metamorphic terrane composites. The Li concentrations are also variable with xenoliths having lower Li concentration than high-grade metamorphic terrane composites (5 $\pm$ 4 ppm vs. 13 $\pm$ 6 ppm, 1$\sigma$). $\delta$$^{7}$Li correlates positively with H$_{2}$O for 12 granulite composites; mafic samples have the highest H$_{2}$O contents and $\delta$$^{7}$Li values while felsic ones have the lowest. This, together with an excellent positive correlation between Li concentration and Mg# for 13 TTG gneiss composites, suggests that both metamorphic dehydration and protolith lithology play important roles in determining the Li isotopic composition of metamorphic rocks. For the Hannuoba xenoliths, $\delta$$^{7}$Li correlates positively with Al$_{2}$O$_{3}$/CaO and negatively with Li, FeO, MgO, CaO and Co, suggesting that the protolith composition primarily controls the $\delta$$^{7}$Li values. No such correlations are found for the other granulite xenoliths. The difference in $\delta$$^{7}$Li between terranes and granulite xenoliths may reflect evolving lighter $\delta$$^{7}$Li with depth in the crust, since the former equilibrated at middle to upper lower crustal depths, while the latter equilibrated in the deep lower crust. Not surprisingly, the Li isotopic composition of the deep continental crust is very heterogeneous. Assuming the high-grade metamorphic terranes represent the middle to upper lower continental crust and the xenoliths represent the lowermost crust, we estimate the $\delta$$^{7}$Li of the lower crust to be +1 $\pm$ 2$\permil$ (1$\sigma$), which is similar to the upper continental crust and slightly lighter than the mantle. This probably is the result of metamorphic dehydration of different types of deep crustal rocks, which, together with surface weathering and low-T magmatic differentiation, drives the hydrosphere ($>$+30$\permil$) to heavier and bulk continental crust to lighter $\delta$$^{7}$Li values (0) than the mantle (+4$\permil$).
V51C-0598 0800h
Zircon U-Pb Discordance and its Relationship to Accumulated Alpha Dose: a Study of Natural and Experimentally Induced Pb Loss Using PDA and CA-TIMS Methods
It has long been known that radiation-damaged zircons are susceptible to Pb loss, but there has been little quantitative information on the relationship between Pb loss and accumulated radiation damage, or even on the exact mechanism(s) of Pb loss from radiation-damaged zircons. Multi-step zircon analyses provide new insights into this problem. Most typical igneous zircons are strongly zoned with respect to actinides on a scale that is short compared with typical grain radii, but long compared with alpha-recoil distances. This provides the opportunity to study the relationship of Pb loss and radiation damage on the sub-grain scale, on either multi-grain or single grain samples. Two analytical approaches were used: 1) "simple" PDA (multi-step zircon dissolution with no annealing); and 2) CA-TIMS (high-temperature annealing of radiation damage prior to multi-step dissolution analysis). All samples are Cretaceous and Jurassic, to minimize natural "fading" or annealing of radiation damage. The first approach was used on zircons that were concordant to near concordant. Lab-induced leaching effects in the early steps (that sampled highly radiation-damaged zircon zones) were followed by ca. concordant steps (that sampled less damaged zircon zones). The threshold of radiation damage for leaching (i.e., preferential Pb loss) is ca. 0.4 x 10E18 alphas per gram. The second approach was used on zircons that had lost significant amounts of Pb via natural discordance processes. The pre-dissolution annealing effectively eliminates laboratory leaching effects, and permits recovery of the natural discordance signature. With this approach, the initial dissolution step samples zircon that is the most highly radiation damaged and the most discordant. The next steps are progressively lower in radiation damage and discordance. Finally, still later steps are completely concordant. In these experiments, the boundary between concordance and discordance is ca. 0.2 +/- 0.1 x 10E18 alphas per gram; 10 percent Pb was lost at ca. 0.5 x 10E18 alpha dose; 50 percent Pb was lost at ca. 1.9 x 10E18 alpha dose. For reference, an accumulated alpha dose of ca. 2 x 10E18 represents the first "percolation point" at which amorphous regions in zircon become interconnected, and above ca. 8 x 10E18 zircons are completely metamict.
V51C-0599 0800h
In-situ Pb Isotope Ratio Measurements in Glasses and Melt Inclusions by LA-SF-ICPMS
We have developed a technique to determine Pb isotopes in glass fragments and melt inclusions by laser ablation (LA) - sector field (SF) - ICPMS. The measurements were done with a New Wave UP-213 laser system and an ELEMENT2 ICP mass spectrometer using the electrical scan mode. The geological MPI-DING reference glasses (Jochum et al., 2000), for which TIMS and MC-ICPMS Pb triple spike (TS) data are now available, were used to test our technique. Our LA-SF-ICPMS data for $^{208}$Pb/$^{206}$Pb and $^{207}$Pb/$^{206}$Pb agree with the high-precision Pb data within 0.2 %. Polished 150 $\mu$m thick sections were used for the analysis of melt inclusions from samples of the Hawaii Scientific Drilling Project (HSDP). Inclusions were ablated for 10 - 30 s in single spots (40 - 80 $\mu$m diameter). Each analysis consisted of 100 to 300 measurements of $^{206}$Pb, $^{207}$Pb and $^{208}$Pb. Typical in-run precision (1 RSE) ranged from 0.2 - 0.4 %. These values are similar to those obtained by SIMS (Saal et al., 1998). The $^{208}$Pb/$^{206}$Pb (1.91 - 2.13) and $^{207}$Pb/$^{206}$Pb (0.778 - 0.877) in the melt inclusions show large and systematic variations. The range of variability in Pb isotope ratios is similar to that reported in melt inclusions from Mangaia and Tahaa basalts (Saal et al., 1998). The Pb isotope arrays (in $^{208}$Pb/$^{206}$Pb vs $^{207}$Pb/$^{206}$Pb) can be explained by mixing of at least two end members. Groundmass values are uniform and similar to the whole rock data and plot within the melt inclusion fields. LA-SF-ICPMS has also been applied for determining Pb isotope ratios and trace element concentrations in carefully handpicked 200 - 500 $\mu$m large glass fragments from 19 samples of the submarine section of HSDP-2. Precision of the Pb isotope data was about 0.2 - 0.3 %. Most LA-SF-ICPMS data agree with high-precision TIMS data using aliquots of about 50 mg. Both data sets confirm the temporal Pb isotope variations found in the HSDP-2 core based on the whole-rock TS Pb isotope data (Eisele et al., 2003).
V51C-0600 0800h
Increasing the Linear Dynamic Detection Range of High Resolution ICPMS
High resolution ICPMS (HR-ICPMS) is widely used in geochemistry for interference-free multielement determinations in complex sample matrices and measurement of precise isotope and elemental ratios at low concentrations, and as a detector for laser ablation analysis. The linear dynamic detection range is of immense importance in ICPMS because of the wide range of elements and concentrations to be analyzed in a single analysis, and particularly so for analysis of laser ablation pulses. HR-ICPMS systems have typically used a discrete dynode detector with a linear dynamic range of 8-9 orders of magnitude which is, in general, not sufficient for quantitation of both matrix elements and ultra-trace elements in a single analysis. To overcome this limitation, we have developed a new detector system for the Finnigan ELEMENT2 HR-ICPMS which incorporates 3 modes of operation which span a linear dynamic range of 12 orders of magnitude. The new detector system has, in addition to the discrete dynode detector with analog and ion counting modes, a Faraday detector, with automatic ion signal-dependent switching between the three detection modes. The new electronics of fast detection system have a minimum dwell time of 1 ms, allowing analysis of transient signals. The presentation will describe the technical concept and the operation of the new "triple mode detector" and demonstrate its capability for laser ablation-HR-ICPMS analysis of matrix and ultra-trace elements in a single analysis.