Journal of Conference Abstracts

Geochronology and Isotope Geology

EUG 10

Geochronology and Isotope Geology

Session O09:3B

0886

O09 : 3B/25 : H3

A Next Generation High Precision Thermal Ionization Mass Spectrometer (TIMS)

Johannes Schwieters (JSchwieters@compuserve.com,

Reinhold Pesch (RPesch@compuserve.com),

Dietmar Tuttas (DTuttas@compuserve.com) &

Bernd Windel (BWindel@compuserve.com)

Finnigan MAT GmbH, Barkhausenstr.2, D-28197 Bremen, Germany

To address recent research needs (e.g. the putative terrestrial ¹⁴²Nd anomaly (Harper et al., 1992)) as well as a number of anticipated future needs in geo- and cosmochemistry (e.g. ultra small Marsian samples) we developed a completely new TIMS instrument: the TRITON TI. The new instrument incorporates a range of new technological innovations which serve as a key to higher precision, accuracy and reliability for both, static as well as multidynamic measurement strategies. The electronics incorporate a new concept for electronic gain calibration: the "virtual amplifier concept". It completely eliminates inter amplifier bias and improves external precision in static TIMS measurements to a level which was not attainable previously. The new ion optics incorporate "zoom optics" to overcome compromises in peak overlap and peak flatness which were one limiting factor for multidynamic measurements in the past. The mass dispersion of the fast laminated magnet has been substantially increased to 81 cm to achieve optimum peak flatness and to reserve more space for eight high precision moveable detector carriers. The new high precision variable multicollector gives maximum flexibility to easily adapt the detector configuration to the specific needs of the application with any combination of "plug in" Faraday cups and "plug in" channeltron detectors (multi ion counting). The multi ion counting option is the method of choice for the analysis of ultra small sample volumes (Richter et al.,1998).The dynamic range of the current amplifiers has been increased to 50 V . The accessible relative mass range has been increased to 15%, to allow static collection of ⁶Li-⁷Li, ⁴⁰Ca - ⁴⁶Ca, U-Pu and ²³⁸U-²⁰⁴Pb. The new Windows NT based software supports networking and remote control. It's intuitive structure sets new standards in TIMS operation and data evaluation. Measurements demonstrating the performance of the new instrument will be presented.

Harper CL, Jacobsen SB, Nature, 360, 728, (1992).

Richter S, Ott U, Begemann F, Nature, 391, 261, (1998).

1787

O09 : 3B/26 : H3

Isotope Ratio Measurements with a Single Collector Sector-Field ICP-MS

Meike Hamester (MeikeHamester@compuserve.com)¹,

Julian Wills (wills@finnigan.de)¹,

Wolfgang Kerl (113615.3362@compuserve.com)¹ &

Dan Wiederin (DanWiederin@compuserve.com)²

¹ Finnigan MAT, Barkhausenstrasse 2, Bremen, D-28201, Germany

² Elemental Consulting, 2004 N.52nd Street, Omaha, NE 68104, USA

The ICP has a number of advantages as an ion source for isotope ratio mass spectrometry. These include the ability to ionize a wide range of elements, relatively low interelement matrix effects, easy sample handling, minimal sample preparation requirements, and fast analysis which are significant advantages when compared to thermionization mass spectrometry (TIMS). However, the precision obtained by TIMS is normally better than that possible with ICP-MS. Noise derived primarily from the ICP ion source and sample introduction system limits isotope ratio precision for conventional single collector ICP-MS instruments. One solution to this problem has been the development of multiple collector ICP-MS instruments which are capable of simultaneously measuring the ion beam from several isotopes. The present work investigates a second approach, namely to stabilize the ion source in order to dramatically reduce the sources of noise at the detector. This results in external reproducibilities of, for instance, 0.01% for lead isotope ratios. Furthermore, the use of a single collector sector-field ICP-MS has several advantages for certain "difficult" isotope ratios: One of the limitations of low resolution ICP-MS is the occurrence of spectral interferences. Sector-field ICP-MS (Finnigan MAT ELEMENT) with a resolution of more than R = 10000 extends the possibilities to measure ratios of spectrally interfered isotopes. Recent advances of the mass stability in high resolution, the scan technique and a new sample introduction system including a high precision nebulizer significantly improves the precision for isotope ratio measurements for spectrally interfered isotopes. Another challenge is the measurement of low mass elements like lithium and boron. Because of the large high voltage scans (30% of magnet rest mass) capability, the isotopes of these elements can be measured without changing the magnetic field, which is mandatory for precise isotope ratio measurements in order to scan fast. Using this scan technique, precision better than 0.1% for lithium and boron can be obtained. The possibility of simultaneous measurement of the analog and counting signal and precise calibration between the two modes of the discrete dynode detector delivers precision of less than 0.1% for large isotope ratios like ²³⁵U/ ²³⁸U in natural uranium. This work investigates the use of a high resolution ICP-MS for isotope ratio measurements of spectrally interfered isotopes (copper, calcium, selenium), for low mass elements (lithium and boron) and for large isotope ratios (uranium, thorium).

3502

O09 : 3B/27 : H3

Multicollector Magnetic Sector ICPMS with a HEXAPOLE Interface

Zenon Palacz (zenon.palacz@micromass.co.uk) &

Simon Meffan-main (simon.main@micromass.co.uk)

Micromass UK Ltd., Floats Road, Wythenshawe, Manchester M33 9LZ, UK

ICP magnetic sector multicollector instruments combine the flexibility of an ICP ion source with the isotope ratio performance of a Thermal Ionisation Multicollector. These instruments are typically double focusing systems with an energy analyser between the ICP source and the magnetic sector. The energy analyser focuses the ion energy spread of ^20-30ev from the ICP, down to 1ev and allows the magnet to accurately mass resolve. A hexapole collision cell is employed on the MICROMASS ISOPROBE to reduce the energy spread by colliding the ions from the ICP with trace amounts of an inert gas within the hexapole. The ions assume the energy of the collision gas. They are then pushed through the transfer optics and then accelerated through the magnetic sector into the multicollector analyser. The collision process also has an added benefit, which was not originally anticipated. It was found that Argon based molecular species and in some circumstances Argon itself can be reduced or removed entirely by charge transfer within the hexapole. The advantage is that very little of the plasma Argon ions are passed into the analyser, and isotopes which were usually interfered with by Argon species were now measurable at low resolution and maximum sensitivity. These include ⁴⁰Ca, ⁵⁶Fe, ⁵⁴Fe, ⁸⁰Se, ⁵²Cr.

We will present isotope ratio data obtained on the ISOPROBE for a variety of elements in solution and by laser ablation. Pb in particular is extremely sensitive on this system with ion currents of 400volts/ppm (1e11 ohm resistor) being obtained. With a CETAC MCN6000 nebuliser this corresponds to ion yields of 1%. Similar efficiencies are also obtained by laser ablation. This means that intense Pb ion signals of the order of 1e6cps/ppm can be obtained from 50 micron spots in zircons, and multicollection with Faradays can be used to obtain Pb-Pb dates. We will also present data on simultaneous U-Pb measurements. The high sensitivity of the instrument also extends to other elements such as Hf where ion/atom efficiencies of 0.4% can be obtained. This is substantially higher than TIMS. Thorium also has a similar efficiency and precise (<0.2% 1SD) data can be obtained in solutions of ~100 ppb, with sample consumptions of <50ng. This opens up the real possibility of obtaining meaningful U series data by insitu analysis in minerals, and also by measuring a range of U series isotopes at high abundance sensitivity (<50 ppb) with a retarding filter.

The hexapole also allows for the first time in ICPMS, the precise measurement of Fe isotopes. We will present data which shows that ⁵⁴Fe/⁵⁶Fe ratios can be obtained on samples of a few hundred nanograms, to precisions of <50 ppm. This can only be achieved because of the removal of argide interferences, the high sensitivity (>0.1%), and the possibility of normalising Fe with Cu which has a similar mass and mass bias.

1980

O09 : 3B/28 : H3

The MultiFlow: A new Continuous Flow Prep System for Isotopic Analyses of D/H and ¹⁸O in Waters as Well as ¹³C and ¹⁸O in Carbonates and DIC

Andy Phillips (andy.phillips@micromass.co.uk),

Francois Fourel,

John Morrison &

Tom Merren

Micromass UK ltd, Floats Road - Wythenshawe, Manchester M239LZ, UK

The use of stable isotope as tracers has become a more and more powerful tool in geochemical interpretations. Originally analyses were performed with single or dual inlets, the sample preparations being completely offline. Then automated preparation devices were developed to be connected online to the mass spectrometers like for instance the automatic carbonate devices. More recently continuous flow techniques appeared where the sample is measured as a gas pulse in a helium stream. This was first applied to EA-IRMS and then to GC-IRMS. The advantage of the continuous flow technique is generally simpler systems, high throughput and smaller sample amounts. Partly because of the shorter measurement times the precision that can be achieved is lower than with dual inlet.What is presented here is a new continuous flow prep system that allows different types of measurements that used to be performed through dual inlet prep systems: 1/ The analyses of D/H and ¹⁸O using water equilibration with respectively H2 and CO₂ on 200 microlitre aliquots of water. Data presented here illustrate the reproducibility of the technique and analyses of international standards (SMOW, GISP, and SLAP) show the accuracy that can be expected from such a system. 2/ The analyses of ¹³C and ¹⁸O on CO₂ from carbonates reaction with phosphoric acid. Reproducibility runs are presented here to evaluate the technique. International references analyses (NBS19, NBS18) are shown to illustrate the performances of this configuration. Data are presented also to illustrate the capacities of the system for small carbonate samples. 3/ The analyses of ¹³C on Dissolved Inorganic Carbonates with acidification techniques. Data are presented here to illustrate the reproducibility of the MultiFlow in that configuration. Experiments are shown using artificial DIC made from known bi-carbonates to demonstrate the quality of the analyses performed with that configuration.The new MultiFlow prep system addresses the major applications of stable isotopes in geology. It allows higher sample throughput and much shorter analyses times using a much simpler system with a good level of precision on the measurements.

1933

O09 : 3B/29 : H3

Requirement for Hydrogen Isotopic Measurement in Continuous Flow Mode. Application to Compound Specific Determinations

Francois Fourel (francois.fourel@micromass.co.uk),

Andy Phillips,

Tim Brockwell,

John Morrison &

Tom Merren

Micromass UK ltd, Floats Road - Wythenshawe

The use of stable-hydrogen isotope measurements is becoming an effective tool in conjunction with other stable isotopes such as ¹³C in tracing natural phenomena. It has been demonstrated for instance that ¹³C and D/H signatures could be successfully used to determine the history of methane occurrences between bacterial, thermal or migratory origins (Friedrich and Juntgen 1972, Prinzhofer and Pernaton 1997). Whereas ¹³C determination is now commonly available with automated systems, the preparation technique currently used for D measurements is classical combustion of organic samples in a sealed tubes followed by cryogenic separation of the waters of combustion, and reduction of water to hydrogen gas on hot zinc (Coleman et al 1982). However, cryogenic purification of samples and the D zinc reduction technique requires considerable manual labor and cost, and generally suffers from low sample throughput The technique further requires the use of classical dual inlet IRMS for measurement of hydrogen isotopes. Here we describe a new on-line analytical method for determining D of waters or organic materials that uses continuous-flow GC-pyrolysis technology. The use of He as a carrier gas for on-line pyrolysis IRMS introduces a potential detector problem due to very large ion beam at m/z 4 from helium, which may not be totally resolved from HD at m/z 3. The small HD signal would therefore have superimposed upon it a significant interference from the tail of the mass 4 peak. Prior to the commencement of this work a modified Micromass Isoprime collector geometry was developed to attain complete baseline separation of the m/z 3 ion beam from that of m/z 4. The isotope ratio of interest is HD/H₂ at m/z 3/2. Unfortunately, an ion reaction occurs in the mass spectrometer ion source such that an interfering mass, H₃, is indistinguishable from HD and requires the familiar H₃₊ correction. Under continuous-flow conditions, however, the pressure of hydrogen entering the source is constantly changing over the width of the peak and the correction is no longer simple. An algorithm for the H₃ correction of hydrogen GC peaks was developed and tested for both accuracy and linearity over a wide range of delta values and sample sizes. A GC-IRMS system configured in pyrolysis mode (Brockwell et al 1993) was interfaced to an Isoprime stable-isotope ratio mass spectrometer with modified analyzer in continuous flow mode. Samples of calibrated references were analyzed to illustrate reproducibility, accuracy and linearity performances of the system. Then natural samples were analyzed with the same configuration to illustrate the application of such a technique to solve the question of the origin of natural fluids.

Friedrich HU & H Jüntgen, Advances in Organic Geochemistry, 639-646, (1972).

Prinzhofer A, Pernaton E, Chemical Geology, 142, 193-200, (1997).

Max LColeman, Thomas J Shephard, John J Durham, John E Rouse, Gillian R Moore, Analytical Chemistry, 54-6, 993-995, (1982).

Brockwell TG Freedman PA, Jumeau EJ, Poster, (1993).

3714

O09 : 3B/30 : H3

Intercrystalline 3-Dimensional Isotopic Profiling of Minerals: A New Era for Geochronology?

Simon Meffan-Main (simon.main@micromass.co.uk)¹,

Zenon Palacz (zenon.palacz@micromass.co.uk)¹ &

Bob Cliff (bob@earth.leeds.ac.uk)²

¹ Micromass UK Ltd., Floats Road, Wythenshawe, Manchester M33 9LZ, UK

² School of Earth Sciences, University of Leeds, Leeds LS2 9JT

In recent years it has become apparent to geochronologists that many igneous and metamorphic processes can result in intercrystalline isotopic heterogeneity. Such variations can allow multiple ages to be obtained using some isotopic systems (for example U-Pb in zircon) but may hinder reliable age estimation in others (such as the Rb-Sr system in feldspars and white mica). Previously, intercrystalline isotopic signatures have been obtained by SHRIMP for zircon U-Pb and microsampling followed by TIMS for Sr in feldspars. However, new instruments and techniques herald a new age in intercrystalline isotope measurements. Herein we present data obtained with a single focusing multi-collector inductively-coupled plasma mass spectrometer (the Micromass IsoProbe) combined with laser ablation sample introduction (LA-MC-ICPMS).

Figure 1 illustrates spatially controlled 3-D ²⁰⁷Pb/²⁰⁶Pb data obtained from within a single zircon crystal (rims show evidence of ²⁰⁷Pb excess). The data was obtained by continuous ablation into the grain; each individual ratio represents an ablation depth of circa. 2µm. The advantages of such an approach are immense as rapid, precise and accurate 3-D age profiling becomes possible. The Pb data in Figure 1 will be expanded upon and simultaneous U-Pb measurements from within single zircon grains will be presented.

LA-MC-ICPMS is not restricted to measurement of Pb isotopic ratios; indeed it is possible to analyse a great number of elements. Thus, we have also investigated the Sr isotopic systematics of metamorphosed feldspar grains. Isotopic heterogeneity is prevalent in many metamorphic feldspars which can result in anomalous Rb-Sr ages (e.g. Meffan-Main & Cliff, 1997). Data will be presented showing how the Sr isotopic system of feldspars can be mapped at a sub-millimeter scale in just a few minutes using LA-MC-ICPMS.

Photomicrograph of ablated zircon with ²⁰⁷Pb/²⁰⁶Pb data. Ablation pits are 50 µm in diameter. Each plot shows data for an individual 50 µm ablation pit with total depth of circa. 200 µm. The y-axis of plots shows ²⁰⁷Pb/²⁰⁶Pb ranging from 0.15 to 0.45 and the x-axis shows depth from 0 µm to 200 µm.

Meffan-Main, S. & Cliff, R.A., Journal of Conference Abstracts, EUG9

1929

O09 : 3B/33 : H3

LA-PIMMS: A New Method of U-Th-Pb Geochronology Using Micro-Sampling Techniques

Randall Parrish (r.parrish@nigl.nerc.ac.uk)¹,

Geoff Nowell²,

Steve Noble²,

Matt Horstwood²,

Hilke Timmermann²,

Phil Shaw³ &

Ian Bowen³

¹ University of Leicester and NERC isotope Geosciences Laboratory, British Geological Survey, Keyworth, Notts, NG12 5GG, UK

² NERC Isotope Geosciences Laboratory, BGS, Keyworth, Notts, NG12 5GG, UK

³ VG Elemental, Ion Path, Road Three, Winsford, Cheshire CW7 3BX, UK

We have used a VG P54 plasma ionisation multicollector mass spectrometer (PIMMS) equipped with multiple Faraday detectors and a computer controlled 266nm UV laser to measure the full spectrum of U, Th, and Pb isotopes in the minerals monazite, xenotime and zircon, including corrections for mass bias using thallium with a natural isotopic composition. Using two different standards of 55 and 1876 Ma age, a linear, zero-intercept calibration function relating measured to known ratios has been empirically verified using both the ²⁰⁸Pb/ ²³²Th and ²⁰⁶Pb/ ²³⁸U decay systems for monazite, and the ²⁰⁶Pb/ ²³⁸U decay system for xenotime and zircon. Repeated measurement of our standards over the course of a day's operation has produced uncertainties in the slope of the calibration line of between 3-8% (standard deviation), within sampled areas of approximately 30 microns across by 10-20 microns depth. The small variation in inter-element ratios and the fact that the slope of the measured vs. actual ratios are within 10% of 1.0, implies limited and fairly constant differential ionisation/volatilisation in our procedure; this result is in contrast to many previous LA-ICP-MS zircon analyses from a number of laboratories who have found it difficult to control the inter-element laser-induced fractionation to these levels of uncertainty using UV lasers. In a separate experiment, we measured ²⁰⁸Pb/ ²³²Th and/or ²⁰⁶Pb/ ²³⁸U from zircon, xenotime and monazite from the 55 Ma standard and found that, irrespective of the mineral analysed, the measurements agree to within 3.3%, proving that there is no significant 'matrix' effect. By contrast, ion microprobe-based U-Pb geochronology (i.e. SHRIMP or CAMECA 1270) involves complex calibration functions as well as the necessity to match the matrices of sample and unknown. Because of the linearity of the calibration, its zero intercept, and its non-dependence on matrix, our method elegantly overcomes many of the problems inherent to ion microprobe isotopic micro-analysis for geochronology, and facilitates rapid sampling and isotopic measurement, simpler standardisation procedures, and reduced cost. We believe it will be an extremely promising method in future geochronological and isotopic research.

2785

O09 : 3B/34 : H3

In Situ non Matrix Matched U/Pb and Pb/Pb Dating of Zircons by EXCIMER LA-ICP-MS

Ingo Horn (horn@eps.harvard.edu),

William F McDonough (mcdonoug@eps.harvard.edu) &

Roberta L Rudnick (rudnick@eps.harvard.edu)

Harvard University, Earth and Planetary Science Depart., 20 Oxford Street, Cambridge, MA, USA

Over the past few year our understanding of laser interaction with minerals has increased rapidly resulting in the development of Laser Ablation systems with controlled ablation characteristics. This enables us today to perform non matrix matched U/Pb and Pb/Pb dating of zircons with precision and accuracy of better than 2% RSD (2<sigma>). Previous attempts, using LA coupled to a single collector quadrupole ICP-MS, involved the use of a matrix matched calibration standard (zircons) (Jackson et al., 1996; Hirata and Nesbitt, 1995) and have been limited to a single spot size between calibration and the determination of unknown zircons. Other techniques, such as "active focusing" (Hirata and Nesbitt, 1995) could increase the precision and accuracy of U/Pb ratio determinations only slightly due to the reduction (but not elimination) of "elemental fractionation". However none of these features are needed if an appropriate laser beam handling is applied to the LA system, which reduces "elemental fractionation" during the ablation process and therefore makes it possible to work without matrix match. We have dated the concordant 91500 zircon (Ontario, Canada - TIMS U/Pb, Pb/Pb age: 1065 Ma, 1062 Ma, Wiedenbeck et al. 1995) using our in house build 193nm Excimer laser system coupled to a VG PQII. The analysis were carried out under static laser focusing conditions with the silicate glass SRM NIST 612 as calibration standard, while using different spot sizes on the sample. The determined ages (1072 Ma ± 16 ²⁰⁶ Pb/ ²³⁸ U, 1075 ± 23 ²⁰⁷Pb/²³⁵U) and 1064 Ma ± 70 Ma (²⁰⁶Pb/²⁰⁷Pb) are in agreement with TIMS U/Pb , Pb/Pb and ion probe Pb/Pb ages. The latter shows higher variability (786 to 1136 Ma) and uncertainties (± 100-300 Ma) when compared to LA ages. We will demonstrate the effects of laser beam handling with respect to ablation time and spot size and characterise the level of "elemental fractionation" on zircons of different ages.

Jackson SE, Longerich HP, Horn I, Dunning G, VM Goldschmidt Conf. ,Cambridge Publication, 1, 283, (1996).

Hirata T, Nesbitt RW, Geochimica et Cosmochimica Acta, 59, 2491-2500, (1995).

Wiedenbeck M, Alle P, Corfu F, Meier M, Oberli F, von Quadt A, Roddick JC, Spiegel W, Geostandards Newsletter, 19, 1-23, (1995).

1268

O09 : 3B/35 : H3

On the Microstructure of Phases Used for U-Pb Dating

Ljuba Kerschhofer (kerschh@uni-muenster.de)¹,

Alex Deutsch (deutsca@uni-muenster.de)¹ &

Urs Schaerer (scharer@ipgp.jussieu.fr)²

¹ Inst. f. Planetologie, Univ. Muenster, Wilhelm-Klemm-Str. 10, D-48149 Muenster, Germany

² Lab. de Geochronologie, Universite Paris VII, 2 place Jussieu, F-75251 Paris Cedex 05, France

The U-Pb method is most commonly used for high-precision age dating in igneous and metamorphic rocks. Minerals suitable for this method should not only contain as little non-radiogenic lead as possible but also be retentive with respect to U, Pb, and the intermediate daughters. In reality, however, the dated minerals often yield discordant ages, indicating an open system behaviour with Pb loss. Several processes have been proposed to explain the discordance of U-Pb data, such as continuous volume diffusion of lead, leaching, and recrystallisation. So far, however, the mechanism responsible for the discordance is still unconstrained, mainly due to (i) our insufficient knowledge of microstructural properties of dating minerals, and (ii) the lack of detailed Pb diffusion data. In this study, we systematically investigate the microstructures of zircon (ZrSiO₄), monazite (CePO₄), titanite (CaTiOSiO₄), rutile (TiO₂), and baddeleyite (ZrO₂). Of these five minerals, representing the most commonly used phases in U-Pb dating, we have selected in each case two populations with concordant or only slightly discordant (i.e., <5%) U-Pb ages. These ages range from "old" to "young" (e.g., for zircon 1852 M and 79 Ma, respectively). U concentrations vary strongly between the different populations (from a few up to several 1000 ppm). Thin sections of single crystal fragments of each population are analysed using optical microscopy and cathodoluminescence scanning electron microscopy (JSM-6300F equipped with an ASK detector). Subsequently, these samples are glued on copper grids and ion-milled for investigation with transmission electron microscopy (TEM, JEOL3010). Among the interesting microstructural features are the zonation patterns frequently occurring in zircon. The combination of cathodoluminescence and TEM data enables us to resolve these zonations on a very fine scale and to interpret the zonation in terms of structural or chemical inhomogeneities. TEM is further used to evaluate the degree of metamictisation of the phases and its dependence of U content and crystal structure, to image fission tracks, and to characterise crystal defects, such as dislocations or twins (e.g. in rutile). We presume that our comparative microstructural study together with the well defined U-Pb systematics of the investigated fractions will result in a better understanding of the ultimate reasons for discordancy.

0563

O09 : 3B/36 : H3

Some Good Reasons for Monazite to be Concordant

Jean-Marc Montel

(montel@opgc.univ-bpclermont.fr)

CNRS-UMR6524, Universite B. Pascal, 5 rue Kessler, Clermont Ferrand, France

Monazite, contrarily to zircon is most of the time concordant in the U-Pb concordia diagram. This means that during a geological event monazite is either totally retentive or totally reset for Pb. Several properties of monazite explain this feature. (1) Pb can be easily incorporated in the monazite structure. PbTh(PO₄)² is a Pb-equivalent of brabantite (CaTh(PO₄)², with a monazite structure. This means that there not a strong driving force for expelling Pb out of monazite. In particular, it cannot be expelled toward other minerals, which cannot accept Pb better than monazite. This is a major difference with zircon, in which there is no suitable site for Pb. (2) Monazite is rarely metamict. Even old, radioactive monazite still crystalline, despite the enormous amount of radiation damages they suffered. So Pb is always linked to the crystal net, and cannot be simply leached out of the crystal. This is also a major difference with zircon, which is commonly metamict. (3) Diffusion in monazite is slow. Then high temperature or very long event a necessary to promote Pb diffusions out of the crystal. Because of (1) and (2) it is the only way for Pb to go out of the Crystal, and it is only possible if the surrounding medium is a better Pb acceptor than monazite (fluid or magma for example). (4) The solubility of monazite in hydrothermal fluids and in silicate melt is not very low and strongly temperature-dependant. This favours dissolution-precipitation processes, which are faster than diffusion. As a conclusion, actual discordance in monazite (generated by Pb diffusion) must be restricted to some specific conditions. Most of the time, the resetting of monazite must occur by dissolution-precipitation. This in agreement with the spatial distributions of ages observed in monazite by in-situ dating techniques.

0308

O09 : 3B/37 : H3

Experimental Study of the Resetting of the U-Th-Pb Geochronological System of Monazite

Anne-Magali Seydoux

(seydoux@gfz-potsdam.de)¹,

Jean-Marc Montel

(montel@opgc.univ-bpclermont.fr)²,

Jean-Louis Paquette

(paquette@opgc.univ-bpclermont.fr)² &

Moacyr Marinho³

¹ GeoForschungsZentrum Potsdam, Telegrafenberg - Projektbereich 4.1, 14473 POTSDAM, Germany

² Département des Sciences de la Terre, 5, rue Kessler, 63038 Clermont-Ferrand Cedex, France

³ Université Fédérale de Bahia, Brasil

Monazite is a LREE orthophosphate used in geochronology to precise the timing of magmatic and metamorphic events. Monazite give generally ²⁰⁶Pb/ ²³⁸U and ²⁰⁷Pb/ ²³⁵U concordant ages, nevertheless discordant U-Pb ages have also been reported: eg. presence of inherited cores, mixing of different populations of grains, Pb diffusion profile...Although the dating methods are now very precise, it is sometimes very difficult to interpret the obtained ages. This is partly due to the poor knowledge of the mechanisms of U-Th-Pb system resetting in minerals. Furthermore, the concept of closure temperature is generally not appropriate to interpret U-Th-Pb ages. We studied experimentally the resetting of the U-Pb and Th-Pb systems in monazite. We used abraded fragments (200-400 µm) of an homogeneous monazite from Brazil. U-Pb analyses of the starting material resulted in concordant U-Pb ages of 474 ± 1 Ma. The grains were treated experimentally under hydrothermal conditions. Runs were carried out at 800 to 1200°C, 7 Kb and for 5 to 60 days. Then, single grains of monazite were analyzed with S.E.M, electron microprobe and dated by isotopic dilution and mass spectrometry. Modifications occured only for temperatures higher than 1000°C, but only the outermost part of the grains is affected. Morphologically, the grains tend to be re-facetted. This is achieved by incongruent dissolution, of the monazite, letting ThO₂ crystals as a solid residue, and precipitation of a Th-poor layer of monazite. Electron-probe analyses showed no modification of the crystals composition at the scale of the microprobe, in particular no diffusion profile. In the concordia diagram, the analytical points are not significantly discordant, confirming that only a very thin layer of the crystal is affected by the transformation. Despite the severe P-T conditions we used (1200°C, 7 kbar, 5 days), the U-Pb system in monazite is not modified. Transformations are limited to the surface of the crystal and no visible diffusion occured.

0557

O09 : 3B/38 : H3

The Influence of Hydrothermal Alteration on the U/Pb-System of Zircons ­ EMS and TIMS Studies

S. Kurz (skurz@gwdg.de) &

B. T Hansen (bhansen@gwdg.de)

IGDL, SFB 468, Goldschmidtstr. 3, 37077 Göttingen, Germany

The U/Pb isotopic system of zircons is a powerful tool to date rocks. Nevertheless interpretation of discordant U/Pb ages is again and again subject of discussions (MEZGER & KROGSTAD, 1997). In the centre of discussion is mostly the preferential leaching of radiogenic Pb (PIDGEON et al., 1966). HANSEN & FRIEDRICHSEN (1989) however showed that leaching is not only restricted to the radiogenic Pb but can also be proofed for the radioactive parent U. Especially Ca-dominated basement brines, which are closely bound to crystalline environment in a depth of 10-15 km, are considered to play an important role on the mobilisation of both elements.

EMS-studies on discordant zircon fractions of different localities in Sweden, the German Fichtelgebirge and Erzgebirge show considerable influence of percolating Ca-rich fluids. Enrichment of Ca is predominatly observed along the rims and at tips of microfractures within the crystalls. It can be shown that incorporation of Ca is restricted to disordered metamict zones with high U-content. It seems to be obvious that hydrothermal solutions are responsible for a remarkable alteration of zircons. One of our samples even shows that alteration may be severe enough to produce geological meaningless U/Pb-ages (WIEGAND, 1997). To determine the degree of influence of these reactive fluids on the U/Pb-systematics of zircons we carried out a series of hydrothermal experiments with a 100-125 µm zircon fraction of the Gunnerstan-granite (SW-Sweden) using NaCl-CaCl₂-solutions to simulate the reactive fluids. According to PERSSON et al. (1987) U/Pb-analyses of five zircon fractions of this granite result in a discordia with an upper intercept age of 1552±38 Ma and a lower intercept age of 181 Ma (no errors stated). Our hydrothermal experiments lead to considerable changes in the ²⁰⁷Pb/²³⁵U and ²⁰⁶Pb/²³⁸U ratios of the treated zircon fraction. The magnitude of change varies with experiment conditions. We were able to create a discordia which intersects the concordia at 1541±36 and the origin. The ²⁰⁷Pb/²⁰⁶Pb ratios remain nearly unaffected and can also be determined by TIMS-analyses in the NaCl-CaCl₂-solutions after leaching experiments. Our experiments show furthermore that leaching is not only restricted to Pb but also affects the U-content of the zircons. Leaching of U can mostly be observed in short run experiments (<500 h) and in the presence of concentrated NaCl-CaCl₂-solutions.

Hansen, BT & Friedrichsen, JD, Lithos, 23, 209-223, (1989).

Mezger, K & Krogstad, EJ, J. metamorphic Geol, 15, 127-140, (1997).

Persson, P-O, Malmström, LR & Hansen, B, T, Geol. Rundsch, 76/2, 389-406, (1987).

Pidgeon, RT, O´Neil, JR & Silver, LT, Science, 145, 1538-1540, (1966).

Wiegand, B, Geotektonische Forschung, 88, 1-177, (1997).

Session O09:4A

1884

O09 : 4A/01 : H3

U-Th Isotope Evidence for Magma Residence Times beneath a Large Rhyolitic Caldera Volcano, Taupo, New Zealand

Bruce Louis Alexander Charlier

(b.l.a.charlier@open.ac.uk)¹,

Steve Blake (s.blake@open.ac.uk)¹,

Peter van Calsteren (p.v.calsteren@open.ac.uk)¹ &

Colin J. N. Wilson (C.Wilson @gns.cri.nz)²

¹ Department of Earth Sciences, The Open University, Milton Keynes, England

² Institute of Geological and Nuclear Sciences, Private bag 2000, Wairakei, New Zealand

Taupo volcano is situated in the central portion of the North Island of New Zealand, and is one of the most productive rhyolitic systems in the world. It is located in an extensional arc produced by the subduction of the Pacific plate beneath the Australian plate. This isotopic study attempts to place constraints on the rates of magma production and residence times at Taupo: Do eruptions tap a single large and long-lived magma body, or are they the products of transient magma chambers? What are the timescales involved with the replenishment of magma in this system?

U-Th disequilibria can be used to constrain rates and timescales of melt generation and magma residence times: Selected samples from the twenty-eight eruptions since the caldera forming Oruanui eruption have been analysed for U and Th isotopes by TIMS. With (²³⁰Th/²³²Th) and (²³⁸U/²³²Th) activity ratios of 0.69-0.75, they show little disequilibrium (<4%) and both Th and U excess.

Four magma batches may be identified in this post-caldera sequence on the basis of stepwise changes in whole rock and mineral compositions and Fe-Ti oxide temperatures. Variations in U-Th ratios demonstrate the influences of apatite and zircon during the genesis of each magma batch, whereas variations in ²³⁰Th/²³²Th can be attributed to crustal contamination and are consistent with variable ⁸⁷Sr/⁸⁶Sr. In order to investigate possible changes in magma residence times, mineral separates are being analysed for U-Th isotopes.

2679

O09 : 4A/02 : H3

U-Pb and Pb-Pb Dating of Diagenetic Processes in Black Shale

Volker Liebetrau (vliebet@ugcvax.dnet.gwdg.de)¹,

Robert Frei (robertf@geo.geol.ku.dk)² &

Andreas Kronz (akronz@ugcvax.dnet.gwdg.de)¹

¹ Geochemisches InstitutUniversität Göttingen, Goldschmidt Straße 1n, 37077 Göttingen, Germany

² Geologisk Institut, Københavns Universitet, Øster Voldgade 10, DK-1350 København, Danmark

Isotopic dating of sedimentary rocks can provide the most direct method of calibration the geologic time scale. Organic-rich black shale commonly contain excellent biostratigraphic markers and sometimes high U concentrations. Enrichment of U and the retentive nature of this reducing lithology provides hope for the utilization of the U-Pb system in sedimentary geochronology. Apart from dating of diagenetic calcite our combined U-Pb, Pb-Pb isotope and electron-microprobe (EMP) study of carbonates from Late Cambrian black shale (peltura scarabaeoides biozone, southern Sweden) attacks the problem of precise but incorrect Pb-Pb isochrons. Bulk leachate data (Israelson et al., 1996) and results of various sequential dissolution experiments (Liebetrau et al., 1997; and this study) on material from the same sampling area and sample type are in parts contradictory. The combination of the sequential dissolution isotope data with wavelength dispersive x-ray spectroscopy (WDX)-EMP element mappings provides an extended interpretation of the geologic setting. We selected samples from lenticular calcite-rich concretionary bodies and the surrounding Upper Cambrian sedimentary Alum Shale strata as well as from cone-in-cone calcite crystals (grown between the two other phases). Israelson et al. (1996) have shown that some samples are highly enriched in U. Their approach to bulk leaching for carbonates has yielded distinctly different U-Pb ages of 478 ± 5 Ma and 510 ± 5 Ma which, combined with sedimentary constraints, enabled an estimation of the compaction rate for these black shale. A Pb-Pb fluid inclusion-host experiment on syn-diagenetic cone-in-cone calcite (cm scale crystals, mechanically homogenized) resulted in a two point reference line of 476 ± 9 Ma. This corresponds to a Pb-Pb isochron interpretation of Israelson et al. (1996) with 473 ± 15 Ma (MSWD = 63.6) for a petrographical comparable sample. Leachates of aliquoted fragments from the same sample, separated into light and dark split pieces, and using the same carbonate attacking reagent show a divergence towards higher U-Pb ages around 490 Ma. Between dark and light fragments the ²⁰⁶Pb/²⁰⁴Pb ratio varies by factor 2, the younger ground aliquot lies in between, but the age information of light and dark is similar. Grinding is increasing the leachability of a source for relative younger radiogenic Pb. Probably some Pb left in a more retentive kerogenic and/or detrital phase during syndiagenetic calcite crystallization and the related U mobilization. An age of 490 Ma for early diagenesis fits well into the geological model and the stratigraphic succession. Sediment samples containing synsedimentary calcite fillings of the primary porous volume (Israelson et al., 1996) were also treated by sequential dissolution. On mechanically homogenized material a precise but by paleontologic and stratigraphic reasons meaningless, at least for the sedimentation age (ca. 510 Ma), Pb-Pb isochron age of 598 ± 4 Ma (MSWD = 0.57) was obtained. This isochron has to be interpreted as a mixing line. About 75% of U is easily leachable and the rest remains in a more retentive silicic and/or kerogenic phase. Consequently, without prior homogenization the leachates of different single whole fragments scatter slightly from this isochron, indicating different Pb sources. The definiton of heterogeneities in the U-Pb system is supported by the detection of specific U spots using detailed WDX-EMP element mappings.

Israelson et al, EPSL, 141, 153-159, (1996).

Liebetrau et al, EOS Transactions AGU, 78, 381, (1997).

2946

O09 : 4A/03 : H3

(U+Th)/He Thermochronometry of the Ethiopian Basement

Raphaël Pik (rpik@crpg.cnrs-nancy.fr),

Bernard Marty &

Jean Carignan

CRPG-CNRS, BP 20, 54501 Vandoeuvre, France

The (U+Th)/He method of dating underwent some recent developments that have clearly demonstrated the usefulness of apatite datings for low-temperature thermochronometry of rocks (Farley et al., 1996). Indeed, owing to numerous diffusion experiments the closure temperature of Helium in apatite is now well constrained at about 80°C (Warnock et al., 97). Titanite is also very ubiquist in basement rocks and potentially interesting for (U+Th)/He thermochronometry. Until now, the diffusion of He in titanite was only reported by Reiners and Farley (1998), which determined a closure temperature at about 190-210°C. The aim of this work is (1) to better constrain the closure temperature of He in titanite by additional diffusion experiments. Our preliminary results confirm the range previously determined. (2) to highlight the power of apatite-titanite-dating combination to record the low-temperature cooling history of rocks in the shallow part of the crust. In particular, the combination of (U+Th)/He dating of apatite and titanite in basement sections where the paleostructural position of samples is well constrained allows to better determine the cooling history because : (a) this thermal history must fit two different sets of data ; (b) the temperature range where He is partially retained is different for apatite and titanite and then this combination allows to discriminate ambiguous ages affected by long stays in the partial retention zone of each mineral. Results that will be presented concern apatite and titanite datings of Ethiopian basement rocks. Ethiopia which is located at the junction of three rifts is characterized by its important and continuous magmatic activity since Oligocene. The first and major volcanic phase consists of a thick pile of fissural basalts (up to 2000 m) erupted 30 Ma ago before the main phases of rifting. This volcanic crisis was clearly linked to the activity of the Afar plume (Marty et al., 1996). Moreover, all this area and more generally the periphery of the Red Sea and gulf of Aden undergone a variable uplift. Fission track datings from the Arabian border of the Red Sea indicate that this margin was rapidly cooled and unroofed 20-25 Ma ago (Menzies et al., 1996). We have sampled uplifted Ethiopian basement far away from the Red Sea border (Blue Nile Gorges) in order to characterize its thermal history and to discuss the chronological relations between basement uplift - trapps eruption - plume impingement

Farley KA, Wolf RA & Silver LT, Geochim. Cosmochim. Acta, 60, 4223-4229, (1996).

Marty B, Pik R & Yirgu G, Earth Planet. Sci. Lett., 144, 223-237, (1996).

Menzies M, Gallagher K, Yelland A & Hurford J, Geochim. Cosmochim. Acta, 61, 2511-2527, (1997).

Reiners PW & Farley KA, Mineralogical Magazine, 62A, 1249-1250, (1998).

Warnock AC, Zeitler PK, Wolf RA & Bergman SC, Geochim. Cosmochim. Acta, 61, 5371-5377, (1997).

2751

O09 : 4A/04 : H3

U-Series Dating of Last Inter-Glacial Corals from the Florida Keys; Implications for Diagenetic Age Disturbance

Ceciel Fruijtier,

Tim Elliott (ellt@geo.vu.nl) &

Wolfgang Schlager (schw@geo.vu.nl)

Faculteit der Aardwetenschappen, Vrije Universiteit, de Boelelaan 1085, 1081 HV Amsterdam, Netherlands

Mass-spectrometric ²³⁴U/²³⁰Th determinations on last inter-glacial corals from the Key Largo formation, Florida Keys, U.S.A., yield apparent ages from 125-138 ka. Samples show initial ²³⁴U/²³⁸U activity ratios, (²³⁴U/²³⁸U)^T, ranging from 1.168-1.176. Such values are significantly greater than last inter-glacial seawater (~1.15) This implies open system behaviour of U-series nuclides, as has been commonly observed in other last inter-glacial corals. Thus, despite being some of the freshest material available for the Key Largo Formation, including samples with >99% aragonite, the accuracy of the precise (±1 ka) ²³⁴U/²³⁰Th ages of these corals is in doubt. This study, however, provides constraints on processes that perturb U-series ages, the understanding of which is key to reconstructing palaeo sea-level in the many areas where completely pristine corals are unavailable. The sample with the oldest age (138 ka) is also petrographically the most altered. This suggests that diagenesis tends to increase apparent ages, and indeed most of the ²³⁴U - ²³⁰Th ages are older than would be anticipated from glacio-hydro-isostatic modes for this location. A more robust indication of the magnitude of age perturbations comes from geological control. We took samples either side of a marked erosion feature, that possibly represents a short-lived sea-level drop within the last inter-glacial. Notably, the stratigraphically younger samples have ²³⁴U/²³⁰Th ages 5 ka older than the underlying corals. Although perturbed by at least 5 ka, the stratigraphically younger samples have a lower (²³⁴U/²³⁸U)^T) than the presumably less altered samples beneath the erosion surface. Likewise, for the dataset as a whole clear relationships between age, (²³⁴U/²³⁸U)^T) and petrography are not evident. We feel a plausible scenario that can reproduce the observed U-series data is combined ²³⁴U and ²³⁰Th addition to corals of younger true age (e.g. ~120 ka). A range in both ²³⁴U/²³⁰Th ratios and amounts of contaminant are required. We suggest this might reflect differences in volume and composition of a percolating fluids that carry in U-series nuclides.

2886

O09 : 4A/05 : H3

TIMS Analyses of ²³⁴U-²³⁰Th Disequilibrium in Calcretes

Arnaud Chiquet (achiquet@cerege.fr) &

Bruno Hamelin (hamelin@cerege.fr)

CEREGE - UMR 6635, Europôle Méditerranéen de l'Arbois, BP 80 Aix-en-Provence, France

Calcretes consist on secondary calcite horizons precipitated in soils of semi-arid regions. The vertical distribution of calcite in the soil profiles strongly depends on climatic conditions. Due to the large fractionation between U and Th during precipitation, calcretes are expected to be appropriate to U-series dating. However, the authigenic calcite generally contains a large ²³²Th-rich silicate fraction. This requires correction of the detrital ²³⁰Th contribution, based on the obtention of isochrons on coeval samples, either through total sample dissolution (TSD), or through sequential extractions on single samples.

We collected six impure carbonate samples (3 to 10 cm³ each) from a sub-vertical vein (30 - 370 cm) crossing a calcrete profile on the Toledo granite (Central Spain). Previous Sr isotopic data have demonstrated the allochtonous origin of calcium in this profile (Chiquet et al., in press a). For each sample, four to six fractions were selected with variable proportion of detrital phase (4.1 - 28.2%). These samples were totally dissolved by a sequential chemical procedure using HNO3-HCL-HF acids. Isotopic ratios were measured by TIMS on a VG 54-30, using static multi-collection between faraday cups and ion counting Daly.

The results, expressed in classical isochron plots (²³⁰Th/²³²Th vs ²³⁴U/²³²Th and ²³⁴U/²³²Th vs ²³⁸U/²³²Th), suggest that most samples are close to isotopic equilibrium. Only two samples show clear disequilibrium, giving ages ranging from 75 to 90 kyr, consistent with previous estimates from calcite mass-balance at this site (40 to 260 kyr; Chiquet et al., in press b).

Calcrete genesis results from a dynamic process of seasonal dissolution and precipitation of calcite, due to alternating humid and dry soil conditions. Such annual redistribution probably affects continuously the ²³⁰Th - ²³⁴U - ²³⁸U chronometer. Partial dissolution of calcite mobilises U, while authigenic ²³⁰Th probably reprecipitates onto silicate and carbonate phases. During this process, the carbonate is artificially enriched in ²³⁰Th, that induces an apparent aging of the calcrete deposit. However, no excess of ²³⁰Th are found, suggesting that this geochemical fracitonation occurs only on a local scale within the profile.

Chiquet A, Michard A, Nahon D & Hamelin B, Geochem. et Cosmochem. Acta, (in press a).

Chiquet A, Colin F, Hamelin B, Michard A & Nahon D, Chem. Geol, (in press b).

2754

O09 : 4A/06 : H3

²³⁰Th and ¹⁴C Dating of Speleothems from the Bahamas: Implications for Calibration of the Radiocarbon Timescale

David A. Richards (David.Richards@bristol.ac.uk)¹,

J. Warren Beck (wbeck@physics.arizona.edu)²,

Douglas J. Donahue²,

R. Lawrence Edwards³ &

Peter L. Smart (P.L.Smart@bristol.ac.uk)¹

¹ School of Geographical Sciences, University of Bristol, Bristol, UK

² Department of Physics, University of Arizona, Tucson, AZ, USA

³ Department of Geology and Geophysics, University of Minnesota, Minneapolis, MN, USA

It is well known that there have been shifts in the offset between ¹⁴C ages and true calendar ages for the last 50 ka as a result of substantial variation in ¹⁴C production rate and changes in the global carbon cycle. Calibration of the radiocarbon timescale prior to 11 ka is based on a relatively small number of coupled ²³⁰Th and ¹⁴C ages of corals and ¹⁴C ages of annually laminated sediments. Here, we present TIMS ²³⁰Th-²³⁴U-²³⁸U, ²³¹Pa-²³⁵U and AMS ¹⁴C ages for speleothems from submerged caves of Grand Bahama that offer considerable potential for an accurate and continuous calibration curve from 11 to 45 ka.

Low U concentrations in speleothems from the Bahamas (typically less than 500 ppb) demand that ¹⁴C and U-series isotopic analyses are performed on sub-samples of different mass (10-20 and 500-1000 mg, respectively) and sampling interval (1-2 mm and 5 - 20 mm, respectively) from the principal growth axis of the stalagmites (up to 620 mm). Growth rates (> 10 mm/ka) were non-uniform and drip loci often moved during periods of continuous growth, causing variable stalagmite morphology. The sampling strategy is therefore critical, and implications for the "²³⁰Th-age growth model" and estimates of the true calendar age will be discussed.

Accurate calibration of the radiocarbon timescale is dependent on a reasonable knowledge of the initial conditions of 'dead carbon fraction' and initial ²³⁰Th. Comparison of the offset between ²³⁰Th and ¹⁴C ages for Bahamas speleothems and coral/laminated-sediment records for the period 11 to 20 ka indicates that there is a small, but relatively constant 'dead carbon' effect from dissolution of ancient carbonates. Th-232 concentrations are generally low (< 0.6 ppb), however in a few sub-samples, where ²³⁰Th/ ²³²Th atomic ratios were < 1 x 10^-3, correction for initial Th conditions had a significant effect. In addition, we find no evidence for post-depositional mobility of the isotopes of interest, and are therefore confident that past activities of atmospheric ¹⁴C can be estimated accurately.

Using the derived calendar-age model for stalagmite growth and corresponding AMS ¹⁴C ages, substantial variation in past atmospheric ¹⁴C is observed, with three peaks of ¹⁴C > 700 per mil during the period 35 to 50 ka. Estimates of ¹⁴C prior 35 ka are subject to large errors and we have paid careful consideration to the effects of variable growth rate, initial ²³⁰Th and dead carbon contribution on their precisions. Similar, large and abrupt peaks have been observed in numerous independent records of cosmogenic nuclide production (eg. ¹⁰Be and ³⁶Cl in the polar ice cores) and inferred from records of geomagnetic intensity in volcanic and sedimentary rocks, yet there is no consensus on their timing, magnitude and causes. The continuous and radiometrically dated record of ¹⁴C activity presented here should help to resolve this.

2000

O09 : 4A/09 : H3

Evidence from ¹⁰Be and U-Series Disequilibria on the Possible Contamination of MORB Glasses by Sedimentary Material

Bernard Bourdon (bourdon@ipgp.jussieu.fr)¹,

Steve Goldstein (sgoldstein@lanl.gov)²,

Didier Bourles (bourles@cerege.fr)³ &

Claude Allegre⁴

¹ Lab Geochimie Cosmochimie IPGP-CNRS T14-24, 4, Place Jussieu, 75252 Paris cedex 05

² J514 Los Alamos National Laboratory, Los Alamos NM 87545, USA

³ CEREGE BP 80, 13545 Aix-en-Provence, France

⁴ Lab Geochimie Cosmochimie IPGP-CNRS T14-24, 4, Place Jussieu, 75252 Paris cedex 05, France

It has often been suggested that U-series disequilibria measured in mid-ocean ridge basalts (MORB) can be perturbed by contamination with sedimentary or hydrothermal material found near the ridge axis where the basalts are erupted. Various leaching procedures are generally used to remove surficial contaminants such as Mn oxides. Here, we provide an independant way of estimating the maximum degree of contamination by sediment using constraints from ¹⁰Be. Since ¹⁰Be is a cosmogenic nuclide (half life of 1.5¥10⁶ years) produced in the atmosphere, there is no reason to believe that it should be present in the mantle underlying the ridge axis. Thus, any ¹⁰Be found in a MORB glass must result from contamination by sedimentary material where cosmogenic ¹⁰Be is enriched.

We have selected four MORB samples with a wide range in Th concentrations (87 to 550 ppb) and for which enough fresh glass was available. One of the samples was not picked and showed evidence for surficial contamination (DO_3-1B-d). The same sample and two other samples were carefully picked under microscope. The samples were leached with a hydroxylamine+acetic acid mixture for 6 hours to dissolve the hydrous phases that contain ¹⁰Be and excess nuclides such as ²³¹Pa and ²³⁰Th. The glasses were then rinsed and ground in a agate mortar prior to ¹⁰Be and U-series separation and mass spectrometry.

All the samples are characterized by large ²²⁶Ra excess which indicates that they must be younger than 10 kyr (unless excess ²²⁶Ra results from contamination). There is no ¹⁰Be present in all the glasses above the detection limit (6x10⁴-1x10⁵ atoms) except for sample DO_3-1B-d. Considering that the sedimentary contaminant contains 10⁹ to 10¹⁰ atoms per gram, the maximum fraction of contaminant in the picked glasses is 10^-4 to 10^-5. There is a large difference between sample DO3-1B-c and DO3-1B-d in ¹⁰Be and (²³¹Pa/²³⁵U) ratios (2.45 versus 3.44) which suggests that leaching and sample selection are essential for obtaining reliable measurements of U-series in MORB glasses. Preliminary mass balance calculations between surficial contaminant and an hypothetical pristine glass with (²³¹Pa/²³⁵U)=1 and ¹⁰Be /⁹Be =0 shows that for the unpicked sample (DO3-1b-d) nearly all the the excess ²³¹Pa can be accounted for by contamination. For the other samples, 15% to less than 1% of the excess ²³¹Pa can be explained by incorporation of sedimentary material. This study reemphasizes the need for careful sample preparation for U-series measurements in MORB, especially at low U levels (<50 ppb).

2823

O09 : 4A/10 : H3

Calibration of Time Integrated Cosmogenic Production Rates Over the Last 1.25 Million Years: Results from Lava Flows on Lanzarote, Spain

Tibor J. Dunai (dunt@geo.vu.nl) &

Jan R. Wijbrans

Isotopengeologie, Vrije Universiteit, De Boelelaan 1085, The Netherlands

A critical issue for exposure age dating utilizing in situ produced cosmogenic nuclides is the accurate knowledge of production rates. Previous calibrations used lava flows younger than 17.5 ka (e.g. Cerling and Craig, 1994). Only recently this time span was increased to 125 ka (Singer et al. 1998). The time-integrated production rates between 0 and 17 ka and 0 and 125 ka are virtually identical (115 atoms ³He/g/yr olivine, >60º latitude, sea level). Given the high variability of the intensity of the geomagnetic field throughout the calibration periods (ranging 20-100% of present day intensity, Frank et al. 1997) and the fact that the geomagnetic field intensity is the single most important factor for temporal variations of production rates, this agreement is fortuitous. Thus calibrations covering different time spans, shorter and longer, are still very much needed. In the present study we focus on the long term production rates. We choose the volcanic island Lanzarote as a target area as its arid climate is likely to preserve lava flow tops for long periods of time.

Olivine separates from samples of three basaltic-lava flow tops of were analyzed for their cosmogenic ³He and ²¹Ne. Leached groundmass of the same samples were dated with the ⁴⁰Ar/³⁹Ar incremental-heating technique. The ⁴⁰Ar/³⁹Ar ages of the samples are 130±70 ka, 270±20 ka and 1.24±0.02 Ma (errors are 2<sigma>). Using the previously derived production rate for ³He (115 atoms/g/yr olivine) the exposure ages of the two younger locations are indiscernible their ⁴⁰Ar/³⁹Ar ages. The exposure age of the oldest lava flow is 10% lower than its ⁴⁰Ar/³⁹Ar age. We see three potential causes for this deviation: i) variable production rate ii) erosion of the highly vesicular flow top and iii) incomplete degassing of argon at the time of eruption. The estimated magnitude of the latter two effects is such that interference corrected ages would be near concordant (±5%). Calculated ²¹Ne exposure ages of all three locations agree well with the corresponding ³He-ages.

The results suggest that both the scaling factors of Lal (1991) and the production rate for ³He (115 atoms/g/yr olivine, >60° latitude, sea level) have an uncertainty < 5% for the period between 125 ka and 1.24 Ma. Moreover, the results of our study indicate that the time averaged intensity of the geomagnetic field over the last 270 ka and most probably also over the last 1.24 Ma was constant within ±5%. This is remarkable as three magnetic reversals occurred in that period. This implies that the process generating the geomagnetic field is very conservative, i.e. has near-identical strength after reversals. In that light we postulate similar time integrated production rates for in-situ produced cosmogenic nuclides for periods exceeding 1.24 Ma.

Cerling TE & Craig H, Geochim. Cosmochim. Acta, 58, 249-255, (1994).

Singer BS et al, GSA Annual Meeting Toronto, (1998).

Frank Met al, Earth Planet. Sci. Lett, 149, 121-129, (1997).

2238

O09 : 4A/11 : H3

Helium Isotope Fractionation in Sedimentary Rocks Related to Specific Behaviour of Tritium, ³He Precursor

Igor Tolstikhin¹,

Igor Kamensyky,

Valentine Nivin,

Bernhard Lehmann (lehmann@climate.unibe.ch)² &

Hugo Loosli

¹ Geological Institute, Kola Scientific Centre, Apatity 184200, Russia

² Physics Institute, University of Bern, CH-3012 Bern, Switzerland

Reduced ⁴He/³He ratios, e.g., down to ~1/100 times those expected from radiogenic production, were observed in sedimentary rocks which formation and history eliminate a contribution of a mantle ³He -bearing fluid. This anomaly could originate from different behaviour of helium and tritium in damage tracks produced by emission of these nuclides from a-decay and exothermic 6Li(nt , a)3H + 4.5 Mev reaction. Generally the tracks cross grain boundaries or some imperfection(s) within a rock or mineral allowing a fast loss of noble ⁴He and ³He atoms. However radiogenic ³He has the precursor, 3H. If 3H is chemically bounded within a track, and the track is able to recover via some underground process on the time scale ~10 years (of the order of 3H life-time), then 3H and related ³He are trapped within the recovered track. This mechanism implies the shorter residence time of ⁴He in the rocks/minerals than ³He; therefore ⁴He/³He ratios could decrease through time. To check this mechanism ⁴He, 3H and ³He(3H) were produced by the above reaction in the targets, i.e., layered composites of thin sections of: quartz, sample, Li-bearing cover, sample and quartz. The samples were these very rocks showing reduced ⁴He/³He ratios. Each target was situated in a quartz ampoule pumped out and sealed off and then exposed to the thermal neutron flux in the reactor. After 4-months-long storage, the nuclides produced during (3H, ⁴He) and after (³He) irradiation were measured in a gas phase above the targets (gas) and compared with their total quantities (total) expected from the Li abundance, the neutron flux, and the time. The expectations were confirmed by measurements of 3H and ³He in some targets. The ratios obtained were 3H(gas)/3H(total) <0.05 and ³He(gas)/³He(total) varying from 0.3 to 0.9. The average half-lifes of 3H and He within the targets were estimated at ~13 and ~0.25 years respectively. Preliminary measurements of 3H and ³He loss from the targets under fully controlled conditions (without the irradiation period when heat liberated could stimulate faster loss of these species) result in longer half-lifes both 3H and ³He in the targets, ~400 and ~1.25 years, respectively. These results confirm the scenario envisaging longer retention of 3H than He in damage tracks. Within the frame of this scenario the life-time of 3H bear a time constraint on a rate of recovery of radiation tracks: to recover the tracks at least one to several newly formed atomic layers should appear before 3H decay, e.g., on the time scale of ~10 years.

2312

O09 : 4A/12 : H3

The Distribution of Rhenium and Osmium in Garnet-Pyroxenite Minerals

Abdelmouhcine Gannoun (gannoun@ipgp.jussieu.fr)¹,

Kevin William Burton

(burton@opgc.univ-bpclermont.fr)²,

Jean-Louis Birck¹ &

Claude-Jean Allegre¹

¹ Lab. de Géochimie et Cosmochimie, IPG de Paris, 4, place jussieu, 75252 Paris Cedex 05, France

² Dept de Sciences de la Terre, Univ. Blaise-Pascal 5, rue Kessler, 63038 Clermont-Ferrand Cedex, France

Rhenium and Osmium (Re-Os) isotopic and elemental data have been obtained for minerals (clinopyroxene, garnet) separated from garnet pyroxenites of orogenic massifs. The Os concentration for the garnet and clinopyroxene range from 1 to 16 ppt and from 9 to 24 ppt respectively. This range is lower than Os concentration in the corresponding whole rock (59 to 1224 ppt) and indicate that the major silicates phases do not control the Os budget. This suggests that minor phases such as sulphide or platinum group minerals dominate the Os budget in mantle rocks. The Re concentration for the garnet and clinopyroxene range from 40 to 300 ppt and from 149 to 470 ppt respectively, while that for the whole rock range from 100 ppt to 4 ppb. The distribution and behaviour of Re is not clear. It appears that silicate minerals partially control the Re budget. Modal calculations for the pyroxenite M82-8(P1) from Beni-Bousera indicate that garnet and clinopyroxene can only account for about 15% of the total Re budget. In other samples garnet can only account for about 30% for the garnet pyroxenite Lh6 (Lherz), and clinopyroxene in the sample La13d can account for the total Re budget. In contrast, garnet from the pyroxenite GB9 accounts for less than 1% of the total Re budget. These results suggest that the siderophile and chalcophile behaviour of Re during mantle melting may be similar to that of the moderately chalcophile elements such as Cu or Ni rather than the highly siderophile Pt group elements.Garnet and clinopyroxene in two garnet-clinopyroxenites from Beni-Bousera massif yield mineral isochrons of 24±2 Ma and 29±3 Ma respectively. This age is in good agreemont with those determined previously for the emplacement age of betico-rifain massifs in the crust (Polvé, 1983; Zindler, 1983), while the Os isotopic composition of whole rock, yield model age of 638 and 639 Ma respectively. This age is older than the time of Re-Os equilibration between host silicate phases suggesting that some other phase in the rock controls the Os budget and that this phase is not in equilibrium with garnet or clinopyroxene. Taken together, these results suggest a spatial and temporal decoupling of Re and Os although at this stage it is not clear if this is a primary feature of these lithologies, rather than a secondary effect (such as diffusion) that post dates crustal emplacement.

Polvé M, Ph D dissertation, (1983).

Zindler A, Staudigel H, Hart SR, Enders R, Goldsteïn S, Nature, 304, 226-230, (1983).

3O09 : 4A/13 : H3

The ¹⁸⁷Os/¹⁸⁸Os Distribution of the Oceanic Mantle is Bimodal!

Jonathan Snow (jesnow@mpch-mainz.mpg.de)

Max-Planck-Institut für Chemie, Mainz

According to current thinking, oceanic peridotites should show a range of ¹⁸⁷Os/¹⁸⁸Os depending on the age and degree of Re depletion of the upwelling mantle (Allègre and Luck, 1980). Other isotope systems behave in this way, and the source of MORB is heavily influenced by this reservoir. However, seafloor weathering is capable of altering ¹⁸⁷Os/¹⁸⁸Os ratios in peridotite (Snow and Reisberg, 1995) . Thus it is important to select samples based on independent geochemical criteria (Snow and Reisberg, 1995; Snow and Dick, 1995) for determining that a rock is unweathered, even for drilled samples. I have compiled all available published and unpublished unweathered abyssal peridotite Os results (Snow and Reisberg, 1995; Roy-Barman and Allègre, 1994; Brandon et al., 1998; Snow et al., in prep.), currently 30 analyses, with a mean of 0.1251+.002, as shown in Figure 1.

Histogram of oceanic mantle ¹⁸⁷Os/¹⁸⁸Os measurements ENRf₈ (2, 4-6) .

¹⁸⁷Os/¹⁸⁸Os

Figure 1.

¹⁸⁷Os/¹⁸⁸Os values greater than Bulk Earth (0.1292; Meisel et al., 1996) do not appear in the new data set. Such values are thus due to alteration (Snow and Reisberg, 1995) and do not reflect mantle Os isotopic compositions. Figure 1 also shows a surprising result: an apparent bimodal distribution of ¹⁸⁷Os/¹⁸⁸Os values in the oceanic mantle. There are no reported ¹⁸⁷Os/¹⁸⁸Os between 0.1240 and 0.1250! This is not a geographic distribution, as there are low (<0.124) and high (>0.125) values in almost every data set. There is also no correlation with any known geochemical parameter.

These observations suggest that there are two fundamentally different yet spatially intertwined Os isotopic reservoirs in the depleted mantle. Each has a different mean age and degree of Re depletion. It is conceivable that other geochemical differences between these reservoirs are lost during the process of MOR partial melting. Such a scenario is consistent with "plume-centric" mantle convection models (Morgan and Morgan, 1997; Morgan and Morgan, submitted).

Allègre C, Luck J, EPSL, 48, 148-154, (1980).

Snow J, Reisberg L, EPSL, 136, 723-733, (1995).

Snow J, Dick H, GCA, 59, 4219-4235, (1995).

Roy-Barman M, Allègre C, GCA, 58, 5043-5054, (1994).

Brandon A, et al, Fall AGU, (1998).

Snow J, et al, (in prep).

Meisel et al., Nature, 383, 517-520, (1996).

Morgan W & Morgan JP, EUG 9, 57, (1997).

Morgan JP & Morgan W, EPSL (submitted).

2O09 : 4A/14 : H3

Trace Elements, U and Sr Isotopes Behaviour in Two African Laterites

Olivier Dequincey (odequinc@illite.u-strasbg.fr)¹,

François Chabaux (fchabaux@illite.u-strasbg.fr)¹,

Norbert Clauer (nclauer@illite.u-strasbg.fr)¹,

Jean-Claude Leprun (leprun@melusine.mpl.orstom.fr)² &

Nicole Liewig (nliewig@illite.u-strasbg.fr)¹

¹ CGS-EOST, 1, rue Blessig, 67084 Strasbourg Cedex, France

² LCSC Centre ORSTOM, 911 Avenue Agropolis BP 5045, 34032 Montpellier Cedex 1, France

Dating of laterites and evaluation of their evolution rates lead to contradictory results. The use of long half-life chronometers applied to separated mineral phases provides ages of several 10⁷ years, while Quaternary ages are generally obtained with short half-life chronometers often applied to whole rock samples. The mineralogical heterogeneity of the lateritic formations could explain these contradictions. The different minerals can be of different ages, some being million years old and others very young. Thus, depending on the kind of material analysed, the constitutive phases lead to variable "ages". The study of the evolution steps of laterites would therefore require to distinguish the presently active mineral phases from the non active ones and also to decipher the transformation steps of each phase. The aim of this study is to work on separate mineral phases and to apply on each of them chronometers of different half-life. Two sites are studied : the Goyoum laterite, Cameroon, and the Kaya laterite, Burkina Faso.

For the Goyoum laterite, the < 0.2 µm size fraction, containing almost exclusively authigenic minerals, was analysed with both the Rb-Sr and the (²³⁴U/²³⁸U) radioactive disequilibria methods. The samples are all in radioactive disequilibrium and hence record recent (< 1 Ma) fractionation processes. The results seem also to underline two main active U fractionation fronts in the profiles with trends towards equilibrium above. In assuming a closed-system behaviour except within the fractionation horizons, evolution rates of 6-8 m/Ma can be calculated for these two fronts and an age of ca. 1.5 Ma can be proposed for the whole profile. In this case the observed variations of the Sr isotope ratios would record the Sr isotopic heterogeneity of the fractions analysed.

For the Kaya site (Burkina Faso), four nearby profiles, from top of a hill (Kaya 1) to a creek (Kaya 5), were studied. The < 0.2 µm size fractions and the whole-rock samples were analysed. The isotopic Rb-Sr and U methods are coupled with the study of trace elements, in order to estimate the part of the geochemical fractionation due to the recent evolution of the weathering profiles. Trace element data point out the existence of chemical fractionations relative to depth within each profile but also lateral variations among the profiles. These lateral variations could indicate lateral fluxes of elements. The U disequilibria results on the < 0.2µm fractions from profile near the creek also demonstrate the existence of recent U fractionation within the profiles. From these data it is proposed that the basal part of the profile has evolved as a closed system above a bottom fractionation front, while the upper part records lateral U fluxes. These results emphasise the interest of the multi-methods geochemical approach to understand laterites evolution.

Session O09:4B

3509

O09 : 4B/25 : H3

How old are Haüy's Type-Locality Eclogites? A Review of Radiogenic Isotope Data in Metabasites, Metapelites and Metapegmatites in SE Austria

Martin Thöni (Martin.Thoeni@univie.ac.at)

Institut für Geologie, Geozentrum, Althanstrasse 14, A-1090 Wien, Austria

Recent attempts to constrain time relations for the type-locality eclogites and their encasing rocks in SE Austria (Haüy, 1822) were faced with the following inherent limitations: a) very low REE concentrations in Grt, Omp and Amp and very low U contents in Zrn of the metabasites; b) pre-high-P memory effects, due to incomplete resetting of the isotopic systems during eclogitization; c) re-equilibration and/or retrogression of the high-P assemblages after the pressure peak. Some petrological, geochemical and isotopic results are summarized below.The metabasic eclogites were derived from gabbros and/or basalts of N-MORB-type composition. The time of crystallization for some of the eclogite protoliths was constrained by Sm-Nd Pl-Cpx and U-Pb Zrn (multigrain) ages to c. 240-280 Ma (Miller & Thöni, 1997; Heede, 1997). The generation of abundant pegmatites and a correlated low-P metamorphism in the eclogite host rocks (mica schist-pegmatoid gneiss association) is probably of the same, Permian age (Habler & Thöni, 1998).

Eclogitization of the mafic rocks resulted in high-P assemblages composed of Grt+Omp (Jd 40)+Zo+Qtz+ Rt±Ky±Amp±Phe. Pyropic Grt (25-27% Prp) from pervasively recrystallized Ky-St-mica schists proves deep subduction of at least part of the metasedimentary host rocks as well. Petrological estimates for the high-P event are 18-20 kbar/580-680°C (Miller, 1990; Thöni & Miller, 1996). Sm-Nd and Rb-Sr mineral-mineral (Grt, Omp, Zo, Amp, Phe) or mineral-whole rock isochrons give ages for the high-P event between 109±8 and 86±1 Ma (Fig. 1). Zrn U-Pb "ages" are significantly older (Heede, 1997). The results show that, contrary to traditional views, this part of the "pre-Alpine Austroalpine basement" experienced a distinct, high metamorphic Alpine evolution. The Permian magmatic and metamorphic pulses, which generated the eclogite protoliths, may probably be related with lithospheric extension, initial Tethyan rifting and transtensional movements along the future Adria-Austroalpine microplate boundaries. In the course of Alpine collisional processes, dismembered oceanic fragments and continental Austroalpine crustal slices were involved in Cretaceous (eo-Alpine) subduction to depths of c. 60 km. Exhumation of the high-P rocks was fast (2-5 mm/a), but discontinuous, and cooling below 300°C was largely completed by the end of the Cretaceous.

Saualpe-Koralpe-Pohorje Sm-Nd Mineral Ages

Habler G, Thöni M, Mitt Österr. Miner. Ges, 143, 291-293, (1988).

Haüy RJ, Traité de Minéralogie, ii ed., Paris (Bachelier), (1822).

Heede H-U, Münster. Forsch. Geol. Paläont, 81, 168, (1997).

Miller C, Schweiz. Mineral. Petrogr. Mitt, 70, 287-300, (1990).

Miller C, Thöni M, Chem. Geol. (Isotope Geosci. Sect. ), 137, 283-310, (1997).

Thöni M, Miller C, J. metamorphic Geol, 14, 453-466, (1996).

1262

O09 : 4B/26 : H3

Deformation Effects on K-Feldspar ⁴⁰Ar/³⁹Ar Ages

Steven Reddy (sreddy@lithos.curtin.edu.au)¹,

Graham Potts (gpotts@liv.ac.uk)²,

Simon Kelley (S.P.Kelley@open.ac.uk)³ &

Nick Arnaud (arnaud@opgc.univ-bpclermont.fr)⁴

¹ Tectonics Special Research Centre, Curtin University, Perth WA 6102, Australia

² Dept. of earth Sciences, University of Liverpool, Liverpool L69 3BX, U.K.

³ Dept. of Earth Sciences, The Open University, Milton Keynes MK7 6AA, U.K.

⁴ Laboratoire de Geologie, Univesite Blaise Pascal, Clermont Ferrand, 63038, France

A combination of forescatter scanning electron microscopy, ultra-violet (UV) laser and cycle stepped-heating ⁴⁰Ar/³⁹Ar dating techniques have been applied to several deformed grains of K-feldspar to investigate the relationship between microstructure and ⁴⁰Ar/³⁹Ar isotope systematics. The sample studied is a Neoproterozoic (c. 800 Ma) arkose from Skye, Scotland, that contains detrital Archean feldspars metamorphosed at amphibolite-facies conditions c.1900 Ma and cooled slowly. The sample was subsequently deformed c. 420 Ma ago during Caledonian orogenesis. The form and distribution of deformation-induced microstructures has been mapped using intragrain variations in lattice orientation. Variations are related to thin section and regional structural data to provide a well-constrained deformation history for the feldspar grains. Over 100 in situ ⁴⁰Ar/³⁹Ar ages were measured from three grains using a UV laser and vary from 1030 to 317 Ma, reflecting the internal microstructure of individual feldspars. The grain showing the least strain records by far the greatest range of ages from 1030 to 420 Ma. The oldest ages are found at the centre of the grain with ages generally decreasing concentrically towards the grain rim. This distribution is thought to reflect partial resetting due to heating around the time of deformation. This simple pattern is truncated by younger ages spatially associated with fractures and internally deformed areas. The relationship between deformation and age variations is best seen in the most deformed grain where there is no spatial relationship between age and detrital grain boundaries and ages preserved range from 452 to 317 Ma. These ages are generally younger than the inferred age of deformation but overlap with ages obtained from the least deformed grain. The oldest ages from this grain are recorded from orientation domain boundaries (ODBs) and are related to excess Ar in these microstructures. Cyclically-heated data from the remainder of the same deformed grain (age profiles, ³⁸Ar/³⁹Ar correlations and log (r/r₀) plots) indicate the presence of three distinct Ar reservoirs. These reservoirs probably represent (1) excess Ar within fluid inclusions within the fractures cutting the feldspar grain, (2) excess Ar released from the lattice adjacent to the ODBs and (3) radiogenic Ar from the feldspar lattice. The data show that Ar loss in the deformed feldspar reflects the presence of a deformation-induced population of small diffusion domains in combination with 'short-circuit' diffusion along deformation-induced defects. Our study illustrates that deformation appears to play a significant role in modifying the diffusion characteristics of individual mineral grains even when the structural history of the rock is relatively simple. The evolution of intragrain microstructure during deformation of K-feldspar plays an important role in intragrain Ar diffusion and therefore to resultant apparent ages.

0423

O09 : 4B/27 : H3

Evolution of the SE-Asian Continent: U-Pb, Hf, and Pb Isotope Data for Single Mineral Grains from Major Rivers

Francois Bodet (bodet@ipgp.jussieu.fr) &

Urs Scharer (scharer@ipgp.jussieu.fr)

Laboratoire de géochronologie, Université Paris 7 et IPG-Paris, 2,place Jussieu, F-75251 Paris, France

U-Pb ages and Hf isotope signatures were simultaneously determined on single grains of zircon and baddeleyite, and single grains of K-feldspar were analyzed for Pb isotopes. The rivers investigated are the Mekong, the Red River, and the Irrawaddy. For the Mekong, which essentially samples the sedimentary cover of Eastern Tibet, Yunnan, Laos and Thailand, about 100 U-Pb ages document magmatic events at 2.7-2.4 Ga (minimum ages of inherited components), and at 1.90, 0.40 and 0.22 Ga (concordant and subconcordant grains). A series of about 40 initial epsilon Hf values (<epsilon>Hf_i) were determined on the most concordant grains showing that major crustal growth occurred at 1.9 Ga (<epsilon>Hf_1.9 Ga: -1 to -4), whereas the 0.40 and 0.22 Ga magmatic events are dominated by melting of 1.9 Ga crust (<epsilon>Hf_0.4-0.22 Ga : -4 to -14). For the Red River, which crosses Yunnan and Vietnam, about 80 zircon and baddeleyite grains define magmatic events at 2.2, 1.9, 0.8, 0.4, 0.22, 0.08 and 0.03 Ga. Most of these grains are concordant or subconcordant yielding about 60 <epsilon>Hf_i values substantiating that all magmatic events reflect to various extents formation of new continental crust in combination with limited melting of crustal material (<epsilon>Hf_2.2-1.9 Ga : +4 to -3; <epsilon>Hf_0.8 Ga: +7 to -4; <epsilon>Hf_0.4 Ga: +1; <epsilon>Hf_0.22 Ga: +5 to -8, <epsilon>Hf_0.08 Ga: +6 to -3; <epsilon>Hf_0.03 Ga: +4 to 0). In strong contrast to the Mekong and Red River, the Irrawaddy zircons are dominated by young U-Pb ages indicating erosion of essentially Cenozoic rocks, formed to a large extent by mantle extraction at that time (<epsilon>Hf_0.05 Ga: +13 to -4). Initial Pb isotopic compositions (Pb_i) of about 150 single grains of K-feldspar reveal significant differences in source characteristics, indicating the dominance of old crust for the Mekong (> 1.9 Ga; high ²⁰⁷Pb/²⁰⁶Pb, whereas the other two rivers are characterized by erosion of crust continuously forming during Precambrian and Phanerozoic times. These Pb_i data are in agreement with the U-Pb ages and <epsilon>Hf_i found for the same river, and the entire set of data shows the source terrains of the three rivers to belong to crustal segments that experienced significantly different evaluations. This observation strengthens the view that SE-Asia is composed of a series of rather narrow terranes, subsequently attached to stable Asia in Phanerozoic times.

3647

O09 : 4B/28 : H3

A Thermal Record of the Assembly and Break-Up of Rodinia and Gondwana in the Palaeoproterozoic Mt. Isa and Georgetown Terranes, North-Eastern Australia

Richard A. Spikings (spikings@eurasia.ethz.ch)

Australian Geodynamics Cooperative Research Centre, School of Earth Sciences, La Trobe University, Melbourne, Vic. 3083, Australia.

The paucity of preserved Neoproterozoic and Phanerozoic units within the Palaeoproterozoic Mt. Isa and Georgetown Inliers of north-eastern Australia renders it difficult to determine their post orogenic (~¾1500 Ma) tectonic histories. Hornblende, white mice, biotite and alkali feldspar ⁴⁰Ar/³⁹Ar data and apatite fission track (AFT) data reveal several distinct periods of post ~1500 Ma accelerated cooling. Plateau and weighted mean ⁴⁰Ar/³⁹Ar ages range between ~1580-380 Ma. AFT ages vary from 390-132 Ma and corresponding mean track lengths range from 13.55 to 11.76 µm. These results record a non-linear, protracted cooling history below ~550°C.

T-t paths for the Three Main Fold Belts of the Mount Isa Inlier and the Georgetown Inlier Derived from ⁴⁰Ar/³⁹Ar and AFT Thermochronology (this study) and at Various Times pre ~1500Ma (U/Pb, Rb/Sr and K/Ar data Obtained Elsewhere). Australian Orogenic Episodes Focused Elsewhere in the Australian Crust are Shown, ASO: Alice Springs Orogeny, R-D: Ross-Delamarian Orogeny, Ms-Gr: Musgravian - Grenvillian Orogeny

Periods of cooling revealed by the data occurred at ~1400 Ma, 1160-1100 Ma, 750-700 Ma, 550-500 Ma, 400-300 Ma and post 100 Ma (Figure 1). These events resulted in the exhumation of ~20 km of crust from the Mt. Isa Terrane since the Isan Orogeny (1550 Ma) and were responses to tectonism focused elsewhere in the contemporaneous Australian plate. Denudation rates varied between 0.9-0.02 km/Ma and were greatest during the ~1400 Ma event when ~¾14 km of crust was denuded from various sub-regions. The periods in time correspond to major events associated with the assembly and break up of Rodinia and Gondwana. A maximum of ~4 km of crust was exhumed from the Georgetown Inlier during ~750-700 Ma. Similarly, the exhumation of ~5 km of crust from the Mt. Isa Inlier during ~550-450 Ma, may have been in response to the reactivation of a Rodinian break-up structure. The thermal record of the Inliers also records pre- Rodinian and post Gondwanan activity.

The spatial variation of ⁴⁰Ar/³⁹Ar and AFT data across individual inliers indicates that major structurally defined sub-regions show similar thermal histories on a regional scale. However, smaller scale fault-bounded blocks demonstrate variable cooling histories. This suggests that reactivation of favourably orientated minor scale faults within the Inliers, including segments of major faults, have occurred since ~1500 Ma. Variations in ⁴⁰Ar/³⁹Ar data from central regions of the Mount Isa Inlier suggest shearing and ¾5 km vertical displacement across a major fault zone during ~1180-1150 Ma.

1051

O09 : 4B/29 : H3

Zircons of the Polymetamorphic Complexes of the Urals

Yulia Pystina (reg@geo.komi.ru)

Institute of Geology, 54, Pervomayskaya st., Syktyvkar, 167000, Russia

Nyartinsk gneiss-migmatitic complex is located in the northern part of Lyapinsk anticlinorium which is part of Central-Uralian uplift. The complex is composed of strongly dislocated and metamorphosed rocks. The most abundant in the section of the complex are garnet and garnet-bearing gneisess. Accessory zircon found in these rocks can be divided into three morphological types: (1) dark-pink rounded zircon 2210 mln. years of age; (2) brownich-yellow opaque short prismatic zircon 670 mln years of age; (3) pale to colourless zircon with age estimates 700 mln years. The obtained age estimates for the zircons are consistent with the geological environment: 2210 mln years can be attributed to the age of granulitic metamorphism, while 700 mln years and 600 mln years can suggest the age of the final stage of metamorphism and metasomatism, respectively.

1979

O09 : 4B/30 : H3

On the Source of the Great Dyke Layered Ultramafic to Mafic Intrusion, Zimbabwe: A Re-Os Study on Chromites

Ronny Schoenberg (ronny@mpi.unibe.ch),

Thomas F. Nägler (naegler@mpi.unibe.ch) &

Jan D. Kramers (kramers@mpi.unibe.ch)

Isotopengeologie, Min.-Pet. Inst., Univ. Bern, Switzerland

We present Re-Os results of different chromitite seams from the lower ultramafic sequence of the Great Dyke, Zimbabwe. The Great Dyke (age 2452±14 Ma, recalc. after Hamilton, 1977) is a layered ultramafic to mafic complex some 550 km in length and presently 4 to 12 km wide, that intruded gneisses and parts of greenstone belts of the up to 3.6 Ga old Zimbabwe Craton. Although geological studies and chromite mining began several decades ago, knowledge about the origin and evolution of the Great Dyke is still limited.

This study is aimed at evaluating the mantle source of the ultramafic sequence of the Great Dyke and to test for potential crustal contamination. This is done by comparing the initial Os isotopic compositions of chromitites to those of a) the chondritic evolution as proxy for the average mantle, 2) other, 2.7 Ga to 3.6 Ga old ultramafic intrusions of Zimbabwe, tracing the subcontinental lithospheric mantle evolution beneath the craton (Nägler et al. 1997) and 3) the layered intrusion of the Bushveld complex, a case where crustal contribution is significant (e.g. Schoenberg et al, 1998).

Samples were clean (hand picked) separates of chromites from different chromitite seams, as in our earlier studies we found this mineral to have unaltered Re/Os and Os isotopic ratios. For this study Re and Os measurements were partly carried out using N-TIMS and partly on MC-ICP-MS, using direct distillation in the case of Os.

Our results clearly indicate that the source of the Great Dyke ultramafic magma was average mantle, and that it is clearly unrelated to the subcontinental lithospheric trend. Further, no contamination from the intruded, significantly older crust can be found within the ultramafic sequence. Samples from different chromitite seams gave identical isotopic compositions, indicating a homogeneous source.

Hamilton, J. Petrol, 18, 24-52, (1977).

Nägler, ThF, Kramers JD, Kamber BS, Frei R & Prendergast MDA, Geology, 25, 983-986, (1997).

Schoenberg R, Kruger FJ & Kramers JD, Min. Mag, 62A, 1349, (1998).

2O09 : 4B/33 : H3

¹⁸⁷Re-¹⁸⁷Os Isotopic and Abundance Systematics of Proterozoic Volcanic Arc Derived Ultramafic Cumulates from SW Sweden

Anders Scherstén (asch@gvc.gu.se)¹,

John W. Morgan (jmorgan@cnr.colostate.edu)² &

Richard J. Markey (rmarkey@cnr.colostate.edu)²

¹ Göteborg University, Earth Sciences Centre - Geology, Box 460, Sweden

² AIRIE group, Department of Earth Resources, Colorado State University, USA

Mafic-ultramafic bodies from Lerum and Kedum in SW Sweden were examined for Re and Os abundance and isotopic composition. Both Lerum and Kedum are located West of and within the Mylonite Zone, a major shear zone that separates the Lerum and Kedum rocks from rocks of higher metamorphic grade to the East. Eight samples were selected from ultramafic cumulates and one relatively unfractionated intrusion that is thought to be the best estimate of the primary magma.

The rocks show a wide compositional range in major and compatible trace elements such as MgO, Cr and Ni. The magmas were hydrous as indicated by cumulus hornblende and this observation together with trace element chemistry supports a supra subduction zone origin. The crystallisation trend is olivine, clinopyroxene ±orthopyroxene, hornblende and plagioclase. Primary oxides and sulphides are rare in the Lerum samples except for abundant ilmenite exsolutions in clinopyroxene. All Kedum samples contain disseminated primary magnetite.

In our samples Re concentration remains fairly constant during fractionation of the magmas whereas Os shows great variation. In the Lerum rocks Os increases with increasing MgO, Cr and Ni. Samples that are thought to represent the initial magmas had about 30 ppt Os. The most basic cumulates, those that contain the most olivine, are enriched about 10 times with respect to Os. Subsequent cumulates are quickly depleted in Os and the lowest measured level in one ultramafic cumulate is 9 ppt. The Kedum samples are less basic than Lerum but has higher Os concentrations. Disseminated magnetite occurs to various extent but has no effect on Re or Os concentrations. No systematic Os correlation with MgO, Cr and Ni is observed but mineral separates of olivine, pyroxene, hornblende and magnetic minerals indicate that olivine and pyroxene control Os concentration. Re does not appear to partition preferentially into magnetite as has been recently proposed (Rigther et al. 1998).

The Lerum samples show large disturbances but plot along a 1620 Ma reference line which is the U-Pb age.All Lerum samples are disturbed with moderate to extreme negative <gamma>Os values. This is interpreted as late Re-addition and supported by a serpentine separate plotting to the right of the reference line. The Kedum samples are less affected and six analyses form a 1880 Ma isochron with a +15 <gamma>Os initial. The suprachondritic initial for the Kedum isochron is probably due to addition of fluids from the descending slab or crustal contamination.

Righter K, Chesley JT, Geist D & Ruiz J, Journal of Petrology, 39, 785-795, (1998).

2572

O09 : 4B/34 : H3

U-Pb Zircon Geochronology of Silicic Tuffs from the Old Red Sandstone of the Dingle and Munster Basins (SW Ireland): Refining the Devonian Time Scale

Edward A. Williams (eaw@erdw.ethz.ch)¹,

Sergei A. Sergeev (sergeev@erdw.ethz.ch)²,

Iwan P. Stoessel (ivan@erdw.ethz.ch)¹,

Mary Ford (ford@ensg.u-nancy.fr)³ &

Kenneth T. Higgs (k.higgs@ucc.ie)⁴

¹ Geologisches Institut, ETH-Zentrum, 8092 Zurich, Schweiz

² Institut fur Isotopengeologie und Mineralische Rohstoffe, ETH-Zentrum, 8092 Zurich, Schweiz

³ Ecole Nationale Supérieure de Géologie, CRPG-CNRS, Rue du Doyen Marcel Roubault - BP40, 54501 Vandoeuvre-lès-Nancy Cedex, France

⁴ Department of Geology, University College, Cork, Ireland

New high-precision U-Pb isotope dilution geochronology of single magmatic zircons from silicic airfall and ground surge tuffs are correlated with new and re-appraised miospore records to refine the recent Devonian time scale of Tucker et al. (1998).

Magmatic zircons have been recovered from a 1.3 m thick trachy-andesitic tuff (the Cooscrawn Tuff Bed) from the fluviatile Lower Old Red Sandstone (Dingle Group) of the Dingle Basin. Dating of these grains implies a revision of the age of the (Early Devonian) Pragian-Emsian stage boundary to be pre-411.2 Ma, since the Cooscrawn Tuff post-dates late Pragian-early Emsian miospores from the same section. This increases the duration of the Emsian stage over the estimate of Tucker et al. (1998), but does not conflict with their Esopus Formation determination (408.3 ±1.9 Ma) which has a considerable biostratigraphic latitude.

The ages of silicic tuffs from fluviatile Old Red Sandstone successions in the NW of the Munster Basin confirm an interval of volcanicity early in its history, and are relevant to the Givetian-Frasnian stages of Devonian time. In the west of the basin depocentre, individual isotopic ages of the 6.3 m thick Keel Tuff and the 11.6 m thick Enagh Tuff, and their very similar REE geochemistry, confirm their correlation, and give a resultant age of 384.9 ± 0.4 Ma. The Keel-Enagh Tuff Bed post-dates a discrete horizon (Fish bed 1 of Russell, 1978) containing the vertebrate Bothriolepis and a new miospore assemblage by *550 m of fine-grained alluvium (Valentia Slate Formation). The chronostratigraphic constraints (late Givetian-early Frasnian/Givetian for Bothriolepis, and a mid Givetian maximum for the miospores) yield an error bar which intersects with the time line of Tucker et al. (1998, fig. 2), implying approximate agreement with this scale.

Isotopic dating of magmatic zircons from thick-bedded acidic tuffs of the Killeen Volcanic Centre (Lough Guitane Complex) in the east of the basin depocentre give an age of 384.5 ± 1.0 Ma, indistinguishable from the Keel-Enagh Tuff Bed. A similar chronostratigraphical constraint is provided by the Moll's Gap Quarry microflora (Higgs & Russell, 1981). This assemblage pre-dates the oldest age estimate of the Lough Guitane Volcanic Complex (385.4 Ma), but by a poorly-known ORS section, yielding a similar agreement with the latest time scale.

The isotopic dates also have significant implications for local basin histories including: basin subsidence with relation to late Caledonian (Acadian) orogenesis, the age of ORS groups in the Dingle Basin, and time-averaged compacted accumulation rates for the Munster Basin (c. 0.16-0.18 mm/yr over c. 23 Myr).

Higgs K & Russell KJ, Geological Survey of Ireland Bulletin, 3, 17-50, (1981).

Russell KJ, Journal of Earth Sciences Royal Dublin Society, 1, 151-162, (1978).

Tucker RD, Bradley DC, Ver Straeten CA, Harris AG, Ebert JR & McCutcheon SR, Earth and Planetary Science Letters, 158, 175-186, (1998).

1173

O09 : 4B/35 : H3

Single Zircon Ages of Migmatites from the Menderes Massif, SW Turkey

Christiane Dannat

(dannc000@mail.uni-mainz.de) &

Thomas Reischmann

Institut für Geowissenschaften, Johannes Gutenberg-Universität, 55099 Mainz, Germany

The Menderes Massif in SW Turkey is consist of a central gneiss core overlain by an envelope of schists and marbles. It is bordered by the Izmir-Ankara suture zone to the North and the Lycian nappes to the South. Previous geochronological studies had shown that the core gneisses were largely derived from granites which intruded during the late Precambrian to early Cambrian, the main igneous phase in the Menderes Massif. Minor magmatic events are documented by Triassic granites and Tertiary intrusions in the central and northern part of the Menderes Massif. This study concentrates on the geochronology of the migmatites that rarely occur within the core gneisses to established the age relationship between the igneous activities and the high grade metamorphic event that produced the migmatites. The investigated migmatites were sampled in the central and the northern part of the Menderes Massif. Zircons were separated from the migmatic parts of the rocks such as leucocratic dikes and melt patches. The morphologies of the zircons of these migmatites consist mainly of euhedral, long-prismatic grains which suggest an igneous origin, but also of rounded grains typical for a sedimentary source. Only euhedral grains were measured using the single zircon Pb/Pb evaporation method (Kober, 1986). The ages of the zircons show a major peak at about 540 Ma. This indicates that the partial melting documented in the migmatites is contemporary to the main magmatic phase of the Menderes Massif. Some older zircons with ages from 686 Ma to 2336 Ma remained unaffected by the high grade metamorphic phase. It is noteworthy that the migmatites do not contain zircons of Triassic or Tertiary age. Our zircon data suggest that migmatisation in the Menderes Massif was related to a pre-alpine metamorphic event that most probably took place at the Precambrian /Cambrian transition.

Kober, B., Contrib.to Min.and Petrol., 93, 482-490, (1986).

1834

O09 : 4B/36 : H3

The Age of the Sakarya Continent in W Anatolia: Implications for the the Evolution of the Aegean Region

Fazil Özmen (oezmf000@mail.uni-mainz.de) &

Thomas Reischmann

(Thomas.Reischmann@uni-mainz.de)

Institut für Geowissenschaften, Johannes Gutenberg-Universität, Becherweg 21, Germany

The Sakarya continent is a crustal segment in NW Anatolia with exposures of the pre-alpine crystalline basement. To the SE it is separated from the Menderes Massif by the Izmir-Ankara zone. To the N and W it is bordered by the Marmara and the Aegean Sea. The continuation to the E is camouflaged by Neogene cover. Outcrops of the crystalline basement are rare and therefore information about the pre-alpine evolution is scarce. This study is concentrates on radiometric dating of this basement using the single zircon Pb/Pb evaporation technique.

The basement immediately S of Bursa comprises grey, banded gneisses of granitic composition. The zircons in these gneisses are euhedral, mainly yellow in colour and slightly turbid. The zircon population documents a heterogeneous and complex history. Three different groups of ages could be identified. One age is 462±6 Ma, which is the oldest age in the Sakarya continent dated so far. The other two ages are 313±1 Ma and 294±4 Ma, the latter being the most frequent age. In the northern part of the Biga peninsula close to the Marmara Sea a granodioritic gneiss was sampled. The zircons are euhedral, clear and yellow in colour. The age obtained is 372±5 Ma. In the S of the Biga peninsula near Edremit we sampled a series of granitic-granodioritic-trondhjemitic gneisses. The granitic-granodioritic rocks contain euhedral, yellow zircons that are frequently turbid. Their ages are 373±3 Ma and 370±3 Ma. A trondhjemitic gneiss had turbid-grey, euhedral zircons that are 304±5 Ma. From the Kozak Dag we analysed clear, euhedral zircons of a granodioritic gneiss. Their mean age is 293±8 Ma.

Because of the euhedral, typical igneous morphologies of the zircons, the ages of the homogeneous popoluations are interpreted as the magmatic crystallisation ages of the granitoid host rocks. The results provide strong evidence that the Sakarya continent formed during the Palaeozoic. The major crust forming magmatic episodes could be recognized at about 370 Ma and at about 300 Ma. The upper Carboniferous phase is also known from the Cyclades and the Pelagonian zone in Greece (Engel & Reischmann, this volume; Stampfli et al., 1998), and therefore has to be considered as an important episode in the evolution of the Aegean region.

Engel M & Reischmann T, J., Conf., Abs., 4 (1999).

Stampfli GM, Mosar J, De Bono A & Vavassis I, Bull. Geol. Soc. Greece, 22, 113-120, (1998).

1906

O09 : 4B/37 : H3

Geochronology of the Pre-Alpine Basement of the Central Cyclades, Greece

Marcus Engel (engem000@mail.uni-mainz.de)¹ &

Thomas Reischmann

(Thomas.Reischmann@uni-mainz.de)²

¹ Institut für Geowissenschaften, Johannes Gutenberg-Universität, Becherweg 21, 55099 Mainz, Germany

² Max-Planck-Institut für Chemie, Becherweg 27, 55020 Mainz, Germany

The Cyclades belong to a belt of exposures of pre-alpine crystalline basement that extends from the mainland of Greece into Turkey. The geological structure of these islands is characterised by basal orthogneisses, overlain by nappes comprising Late Palaezoic to Mesozoic metasediments as well as Jurassic to Cretaceous ophiolites. The best exposures of the gneissic basement are on the central islands of the Cyclades such as Delos, Paros, Antiparos, Naxos, Sikinos, and Ios. In this contribution we present geochronological data obtained by the single zircon Pb/Pb evaporation technique from various gneisses of these islands and discuss their implications for the evolution of the pre-alpine basement of the region.

The rock samples from the central Cyclades are coarse grained, leucocratic metagranites, often with augengneiss textures. They contain the mineral assemblages typical for granites with K-feldspar, plagioclase, quartz, muscovite, biotite, and frequently small amounts of tourmaline. The geochemical composition of these gneisses is akin to the composition of modern S-type granites.

Cathodoluminescence investigation of the zircons show magmatic zoning and some rare inherited cores. Scanning electron microscope studies show that the morphologies of the zircon grains are euhedral without indications of sedimentary transport and thus support the igneous origin of the zircons. Zircon ages of augengneisses from Delos vary from 295±4 to 327±4 Ma. The basement of Paros has ages of 302±2 to 325±4 Ma, similar to the neighbouring island of Antiparos (292±1 to 308±3 Ma). Zircons from the augengneisses of Ios yield 302±3 to 311±2 Ma, and gneisses from Sikinos are dated at 301±2 to 325±4 Ma.

We interpret these ages as the igneous formation ages of the granitoid gneisses. Some grains with older ages between 500 Ma and 2303 Ma are interpreted as inherited zircons. This implies that the granites formed from a source which is at least partly of Early Proterozoic age.

The prevailing late Carboniferous ages of about 300 Ma document a major igneous event in the central Aegean region. This event is not known in the Menderes Massif in SW Turkey which had previously been considered to be the eastern continuation of the Cyclades. However, such Carboniferous ages are already known from gneisses from the Pelagonian zone (Stampfli et al., 1998) and the Sakarya continent in NW Turkey (Okay et al., 1996; Özmen & Reischmann, this volume). This implies that the pre-alpine basement of the Cyclades forms the central part of an arcuate orogenic belt between the Pelagonian zone and the Sakarya continent which was largely formed by magmatic activities in the Late Carboniferous.

Stampfli G, Mosar J, De Bono A & Vavasis I, I Bulletin of the Geological Society of Greece, 1, 113-120, (1998).

Okay A I, Satir M, Maluski H, Siyako M, Monie P, Metzger R, Akyüz S, The Tectonic evolution of Asia. Camridge University Press., 420-441, (1996).

Özmen F, Reischmann T, Terra Nova, J., Conf., Abs. 4, (1999).

2646

O09 : 4B/38 : H3

Permotriassic Dike Magmatism in the Bormio-Livigno-Reschenpass Area (Eastern Alps)

Judit Zöldföldi

(judit.zoeldfoeldi@student.uni-tuebingen.de)¹,

István Dunkl (dunkl@uni-tuebigen.de)² &

Wolfgang Frisch (frisch@uni-tuebigen.de)²

¹ Institut für Mineralogie, Petrologie und Geochemie, Universität Tübingen, D-72074 Tübingen, Wilhelmstr. 56., Germany

² Institut für Geologie, Universität Tübingen, D-72074 Tübingen,Sigwartstr. 10., Germany

Porphyritic dikes cut Austrolapine micaschists and augengneisses in the Bormio-Livigno-Reschenpass area. The hypabyssal dikes are nonmetamorphic and may be divided into four types, according to their mineral modes: rhyolite, gabbrodioritporphyrite, diorite-microdiorite and andesite. SiO₂ ranges from 50 wt-% to more than 70 wt-%.

AFM-plots and the PEACOCK-index suggest a calc-alkaline (partially with high potassic affinity) and a tholeiitic trend. The gabbrodioritporphyrites, diorites-microdiorites have subdoleritic to porphyritic texture. They consist of plagioclase (51-55 v%), olivine (0-2 v%), clinopyroxene (up to 5 v%), biotite (5-10 v%) and Mg-hornblende (5 v%), quartz (0-5 v%) and accessory ilmenite, apatite and zircon. Sericite, epidote, chlorite, calcite and uralitic amphibole from the hydrotermal assemblage. Until now there have not been deep researches in the area of the Reschenpass, Bormio and Livigno. Gatto et al. (1976) and Beccaluva et al. (1979) place the dike-rocks of the eastern Alps in the oligocene. Until now the theory was that these dike-rocks belong to the tertiary magmatism along the Periadriatic-Lineament. K-Ar determinations were performed on 11 samples. The available isotopic age data on dikes of the Bormio-Livigno-Reschenpass area suggest three groups: (i) 264.0±15 (hornblende) - 237.3±8.8 (biotite) Ma; (ii) 185.7±7.1 - 89.6±3.7 Ma (w.r.); (iii) 23.7±1.4 Ma (w.r.).The main part of the rocks contignous to the Northern part of the Periadriatic-Lineament are not to be ascribed to the tertiary magmatism. The dikes were probably formed in the permotriassic. This is witnessed by the age-dating of the biotite and hornblende, as well as by the similarity of the classical Ti/Zr diagrams of the triassic dikes of the southern alpine area with the samples wich we have analysed.In comparison with similar rocks from other localitis in the eastern parts of the central Alps, these dikes may be attributed to an permotriassic igneous event during a tensional tectonic regime. As an important geodynamical consequence has to be stated that the prevailing volcanism of the southern-alpine triassic can also be found in the cristalline part of the Austroalpine unit.

Beccaluva L, Gatto GO, Grenanin A, Piccirillo EM & Scolari A, N Jb. Geol Pal Mh, 6, 321-339, (1979).

Gatto GO,Grenanin A, Piccirillo EM & Scolari A, Studi Trentini Sci. Nat., 53, 21-47, (1976).

Session O09:4P

3560

O09 : 4P/01 : PO

High-Precision Isotope Ratio Measurements with the IsoProbe ­ A New Multiple-Collector ICP Magnetic Sector Mass Spectrometer

K. Mezger (klaush@nwz.uni-muenster.de),

M. Rehkämper (markr@nwz.uni-muenster.de),

J. Aggarwal (aggarwa@nwz.uni-muenster.de),

S. Weyer,

F. Wombacher &

S. Rochnowski

Zentrallaboratorium für Geochronologie, Universität Münster, Corrensstr. 24, D-48149 Münster, Germany

The IsoProbe is a new, single focusing magnetic sector mass spectrometer with an ICP source. A hexapole collision cell, situated between the skimmer cone and the entrance to the magnetic sector analyzer is employed to reduce the energy spread of the ions from >15 eV to <1 eV. This enables the instrument to achieve accurate mass focusing and symmetric, flat topped peaks for all of the elements thus far investigated (e.g., B, Sr, Ag, Nd, Pb, U). The sensitivity of the IsoProbe is higher than that of previous magnetic-sector ICP-MS instruments at similar operating conditions. Using a CETAC MCN 6000 desolvating nebulizer, we have been able to obtain total ion beam currents of >1 x 10^-10 A for 100 ppb solutions of Nd and Pb. The instrument at the Universität Münster is equipped with 9 individually adjustable Faraday cups that permit high-precision isotope ratio measurements for a wide range of elements with multiple collection in the static mode.

Repeated isotope analyses of B, Nd and Pb standard solutions have been performed in order to investigate the analytical performance of the instrument for a variety of different applications. For multiple runs of 100 ng of La Jolla Nd, we have been able to achieve external reproducibilities (2<sigma>) of <25 ppm for ¹⁴²Nd/¹⁴⁴Nd, ¹⁴³Nd/¹⁴⁴Nd, and ¹⁴⁵Nd/¹⁴⁴Nd following correction of mass discrimination by internal normalization relative to ¹⁴⁶Nd/¹⁴⁴Nd. Multiple runs of the NIST SRM 981 Pb yielded 2<sigma> reproducibilities of <50 ppm for ²⁰⁸Pb/²⁰⁶Pb, ²⁰⁷Pb/²⁰⁶Pb and of <400 ppm for all Pb isotope ratios that include the low-intensity ²⁰⁴Pb beam. In this case we applied an external method for the correction of mass discrimination, which uses the known isotopic composition of admixed Tl to correct for the mass bias observed for Pb. Boron isotope measurements using the IsoProbe have shown strong mass discrimination, resulting in measured ratios approximately 40% higher than those observed with either negative or positive ion TIMS. Precise correction for mass discrimination, however, is possible by normalization of the results to NIST SRM 951 B. Repeated analyses of standards indicate that B isotope ratios can be measured with a precision of better than 1‰, which is superior to the negative ion TIMS technique.

Following complete validation of the analytical performance of the IsoProbe, we plan to investigate the feasibility of isotope ratio and isotope dilution concentration measurements for a number of additional elements that are of interest for geo- and cosmochemical studies.

0449

O09 : 4P/02 : PO

Automated Fission Track Dating

John Alfred Miller (jam2@esc.cam.ac.uk)¹,

Daniel Costich (dec21@cam.ac.uk)² &

Anthony James Hurford (t.hurford@ucl.ac.uk)³

¹ Bullard Laboratories, Madingley Road, Cambridge, UK

² Cavendish Laboratory, Madingley Road, Cambridge, UK

³ University College London, Gower Street, London, UK

Attempts at developing means of automated fission rack dating have to a large extent been frustrated by the difficulty of obtaining good images of the tracks over their whole lengths using conventional high power microscopy. Work in Cambridge has shown that it is possible to produce clear images of tracks by means of the confocal microscope.

Conventional methods of fission track dating cannot take into account tracks lost from the sample and detector as a result of etching. Attempts have been made to counter this problem by endeavouring to standardise the conditions of etching. Unfortunately there is no way of knowing whether or not it has proved successful.

Track losses can be quantified if the minimum retained angle (<theta>c) of the tracks can be found. Using confocal microscopy, a series of planar images can be obtained from different depths in the target material to a resolution of 0.2µm. The procedure for determining <theta>c is described. The image analysis system allows reconstruction of the tracks in three dimensions and their lengths to be recorded. A robust algorithm is used to identify tracks even in noisy images and measure the dip angles and lengths of the tracks. This technique does not however detect every track present and cannot therefore be used reliably for track counting. However, the mode angle of the tracks can be found with accuracy.

Automatic track counting is achieved by counting the number of holes in the surfaces of the sample and the detector. Size limits are set so that scratches and other such surface features are not counted. The program can distinguish overlapping holes on shape criteria. Questionable targets show up coloured.

These methods have been applied to a sample and its detector. The results are presented.

The validity of a new method of correcting for the number of tracks lost when an external detector is etched is explored.

1256

O09 : 4P/03 : PO

The PAZ of Zircon Alpha Damage and Fission Tracks: Field Evidences and Consequences for FT Annealing Experiments

Meinert K. Rahn (rahn@ruf.uni-freiburg.de)¹,

Mark T. Brandon (brandon@hess.geology.yale.edu)² &

John I. Garver (garverj@idol.union.edu)³

¹ IMPG, Albertstrasse 23b, 79104 Freiburg, Germany

² Department of Geology and Geophysics, Yale University, New Haven, CT 06520-8109, U. S. A.

³ Geology Department, Union College, Schenectady, NY 12308-2311, U. S. A.

Accumulated alpha damage in zircons has a distinct influence on the annealing behaviour of fission tracks (FT). Despite the fact that in experiments there is no clear relationship between present alpha damage density and FT annealing temperatures, minor amounts of alpha damage seem capable to markedly shift the annealing of FTs toward lower temperatures.

For the validity and extrapolation range of existing annealing experiments, it is important to know the thermal relationship between the alpha damage and FT partial annealing zone (PAZ). In the worst case, these two PAZs could cross between laboratory and geologic conditions, thus rendering any extrapolation of annealing experiments invalid. In our study, six physically different relationships between the two PAZs are discussed in the light of various experimental and field observations. It turns out that only one relationship with the alpha damage PAZ separated and at higher T than the FT PAZ for the complete time range is compatible with all existing observations. This relationship implies that zircons entering the FT PAZ from the low-T side will always show an alpha damage density in accordance to their grain-specific P-T history and U content. Zircons approaching the FT PAZ from the high-T side will accumulate some alpha damage before starting to accumulate FTs. Only in the case of very rapid cooling, zircons might remain unaffected by alpha damage and thus show an annealing behaviour as found in experiments done with induced FTs in alpha damage-free zircons. The result implies that all experimental studies on FT annealing have to realte to a data set of induced FT experiments as an upper T limit for zircon FT annealing conditions, and only material with very homogeneous age and U content should be used for natural FT annealing experiments.

2040

O09 : 4P/04 : PO

C/N and ¹⁵N Evolution During Graphitization of Organic Matter

Magali Ader (ader@ipgp.jussieu.fr)¹,

Jean-Paul Boudou (boudou@cicrp.jussieu.fr)¹,

Marc Javoy (mja@ipgp.jussieu.fr)¹ &

Jae-Ho Oh (jhoh@rock25t.kigam.re.kr)²

¹ Institut de Physique du Globe de Paris, Univ Paris 7, France

² Korea Institute of Energy and Resources, Daejon

Using conventional techniques (Ader et al., 1998) showed that, during meta-anthracitization / semi-graphitization (triggered by regional as well as by contact metarmorphism), isotopic composition of organic nitrogen remains almost constant, in spite of a large decrease of organic nitrogen content until about 0.5 (atomic N/C ratio). In the present work, using a more sensitive technique, i.e. static mass spectrometry, organic N/C and ¹⁵N in semi-graphite and graphite series, formed by contact metamorphism of anthracite seams (Oh et al. 1991), were determined in order to study (1) the effect of higher metamorphism grade on the organic atomic N/C ratio and the isotopic ratio ¹⁵N of organic nitrogen, and (2) to check if the nitrogen content and isotopic composition of coal (metasedimentary) graphite are different of those of fluid-deposited graphite. Results show that, as for anthracitization, isotopic composition remains the same all along the graphitization whereas denitrogenation leads to very low nitrogen content (> 10 ppm) - though higher than in fluid-deposited graphites which contain only traces of nitrogen (<3 ppm). Preliminary conclusions are that (1) nitrogen isotopic composition remains constant from anthracite to graphite and, as a result, that nitrogen is released from the parent organic matter without any isotopic fractionation, (2), isotopic studies dealing with nitrogen in green schist facies metasedimentary rocks must take organic nitrogen, as well as ammoniacal nitrogen into account, and (3), because ultimate graphite contain only traces of nitrogen, mantle diamonds (some of them contain up to 3000 ppm of N) cannot originate directly from the phase transformation of graphite to diamond in the subducted slabs.

Ader M, Boudou J-P, Javoy M, Goffé B & Daniels E, Org. Geochem, in press, (1998).

Oh J-H, Rouzeau J-N, Oberlin A, Deubergue A & Kwak Y-H, Bull. Soc. géol. France, 162, 399-407, (1991).

3259

O09 : 4P/05 : PO

Microdomain Oxygen Isotope Analysis of Zircon by Ion Microprobe Multicollection

Stefan Claesson (stefan.claesson@nrm.se)¹,

Stephen J. Mojzsis (sjm@argon.ess.ucla.edu)² &

Chris D. Coath (coath@argon.ess.ucla.edu)²

¹ Swedish Museum of Natural History, Box 50 007, SE-104 05 Stockholm, Sweden

² University of California Los Angeles, Dept of Earth and Space Sciences, Los Angeles, CA 90024-1567, USA

Individual zircon crystals can contain complex internal structures which retain both the memory of their formation, and subsequent events in their evolution. U-Pb isotope ion microprobe (SIMS) microanalysis of selected domains within single zircon crystals has opened new possibilities to unveil the chronology of polymetamorphic terranes; analysis of other major and trace elements, and isotope systems, can help us to understand how an age determination relates to a specific geological event or process. Furthermore, O isotope compositions of zircon can convey information about formation conditions, and also has the potential to discriminate between magmatically and metamorphically formed zircon. Published ¹⁸O_SMOW values for zircon vary between +1 and +15‰, and typical values for igneous crystals are between +4 and +8‰. Both experimental results and investigations of rocks with well constrained P-T history have demonstrated that the O-isotope system will remain closed at >500°, depending on details of the metamorphism.

A limiting factor to conventional and laser fluorination oxygen isotope analysis of zircon is the large sample size requirement; such requirements have previously made it difficult to resolve the sometimes large complexities found within individual crystals. SIMS analysis can be done on selected areas of 10-20 micrometer in diameter on single crystals, yet it is limited by a relatively poor analytical precision relative to conventional techniques. The use of an ion microprobe multicollection (SIMS-MC) system has the potential to improve this precision and to resolve small differences in e.g. oxygen isotopic compositions.

We present here results from SIMS-MC oxygen isotope analysis of zircon using the Cameca ims1270 at UCLA. Analysed samples include the standard zircon 91500, Mud Tank (courtesy of T.M. Harrison, UCLA), and the kimberlite zircon Kim-5 (courtesy of J. Valley, UMich.). We have also finished a pilot study of the oxygen isotope variability in