ALBERT

Notes: Abstract High-silica rhyolites of the Sierra La Primavera, a late Pleistocene center near Guadalajara, are extremely Sr-poor (0.3–1.3 ppm), yet (with one exception) values of 87Sr/86Sri are relatively low at 0.7041–0.7048. Values of 143Nd/144Nd for all the rhyolites are (within errors) identical to a basalt at 0.5129. These surprisingly primitive values, along with feldspar ∂18O of +6.6‰, are consistent with an origin by fractional crystallization of mantle-derived basalt. However, absence of the large volume of associated intermediate rocks that would be expected if the 40 km3 of erupted rhyolite were produced mainly by fractional crystallization suggests alternative processes involving partial melting of Mesozoic or Tertiary mafic intrusive rocks (or lower-crustal metamorphic equivalents). The latter interpretation is preferred, especially in light of comparative data for other North American, Cenozoic, high-silica rhyolites. Isotopic compositions correlate with basement age, but generally lie between values for associated basalts and the underlying crust. Nearly all can be interpreted as containing both a young mantle-derived component and a crustal component, probably derived by partial melting at intermediate to deep levels of the crust. No matter what the proportions of mantle- and crust-derived material in parental magmas, the extremely low concentrations of Sr and Ba in the high-silica rhyolites require extensive fractional crystallization of feldspar-rich assemblages after parental liquids attain rhyolitic compositions. At La Primavera, contamination by shallow roof rocks probably led to the 0.708 87Sr/86Sri ratio of the earliest postcaldera lava dome, which is thought to have erupted through the same vent as the caldera-forming pyroclastic flows. Contamination associated with collapse apparently affected only a small volume of magma in contact with brecciated wall rocks close to the vent, as nearby lavas that erupted during the same episode about 95 ky ago are unaffected. No identifiable lowering of ∂18O took place on caldera collapse. Rhyolitic lavas that erupted 75, 60, and 30 ky ago document postcaldera chemical recovery of the chamber to progressively more evolved compositions in its upper reaches, but show little variation in ∂18O, 87Sr/ 86Sri, or 143Nd/144Nd with time, suggesting that the bulk of the rhyolitic magma within the chamber was isolated from significant wall-rock contamination. Most of the small range of 87Sr/86Sri among the rhyolites can be attributed to pre-eruptive, in situ decay of 87Rb, resulting in a measurable secular increase of 87Sr/86Sr in these Sr-poor magmas. The 87Sr/86Sri of the youngest rhyolite, however, is somewhat lower than predicted, suggesting that the silicic magma chamber was at times open to interaction with more-mafic magmas from below.

Notes: Abstract We have developed a modified Carius Tube design that permits the low-blank digestion of geological samples prior to platinum-group element analysis. The new Carius Tubes incorporate a liner of high-purity quartz glass that retains the sample and acids during the digestion procedure. This dramatically reduces the comparatively high Pt blank associated with dissolutions in conventional Carius Tubes. Using the new Carius Tube design we are able to achieve total procedural blanks for the determination of the PGE in geological samples that are at the 1–15 pg/g level for Ru, Pd, Ir and Pt. This constitutes a reduction of blank values by a factor of ∼10–100 compared to standard NiS fire assay sample preparation techniques.

Notes: Abstract  Alkali basalts and nephelinites from the volcanic province of northern Tanzania contain pyroxene and nepheline that show evidence for chemical and/or isotopic disequilibria with their host magmas. Olivine, pyroxene, nepheline and plagioclase all appear to be partially xenocrystic in origin. Five whole rock/mineral separate pairs have been analyzed for Sr, Nd, and Pb isotopic compositions. The 206Pb/204Pb ratios are distinct by as much as 20.94 (whole rock) vs. 19.10 (clinopyroxene separate). The Sr and Nd isotopic disequilibria vary from insignificant in the case of nepheline, to Δ 87Sr/86Sr of 0.0002 and ΔɛNd of 0.7 in the case of clinopyroxene. The mineral chemistry of 25 samples indicates the ubiquitous presence of minerals that did not crystallize from a liquid represented by the host rock. The northern Tanzanian magmas are peralkaline and exhibit none of the xenocrystic phases expected from crustal assimilation. The disequilibria cannot be the result of mantle source variations. Rather the xenocrystic phases present appear to have been derived from earlier alkali basaltic rocks or magmas that were contaminated by the crust. Material from this earlier magma was then mixed with batches of magma that subsequently erupted on the surface. Disequilibrium in volcanic rocks has potentially serious consequences for the use of whole rock data to identify source reservoirs. However, mass balance calculations reveal that the 206Pb/204Pb isotopic compositions of the erupted lavas were changed by less than 0.25% as a result of this indirect crustal contamination.

Notes: Abstract Concentric zoning in the Criffell pluton takes the form of a discontinuous outer margin of metaluminous hornblende granodiorite and an inner core of increasingly peraluminous muscovite granite. Previous investigations using major and selected trace elements have shown the variation to consist of both smooth and abrupt trends. This study of 15 samples for the rare earth elements shows patterns which strongly correlate with Sr and O isotope data. The principal feature of these data is a progressive decrease in total rare earths with approach to the geochemical centre of the pluton, and evolution to more radiogenic Sr and more silicic and peraluminous compositions. No significant europium anomaly is developed. The slope of light to heavy rare earths using La/Yb ratios varies in a complex manner showing no significant correlation with any of the main indices of bulk composition, but with peak values occurring within the inner part of the outer portion of the pluton. A map of Ce/Y variation based on 172 Ce and Y determinations is essentially identical. These data are considered in terms of various petrogenetic models and it is concluded that the data can only be interpreted in terms of a major and progressive involvement of crustally-derived anatectic magma towards the pluton interior. Trace element modelling favours processes of the assimilation-fractional crystallisation (AFC) type for the generation of this example of I-type to S-type granitoid zonation.

Notes: Abstract Models for the mechanisms of accretion of the terrestrial planets are re-examined using the experimental technique of high-precision isotope ratio mass spectrometry of tungsten (W). The decay of 182Hf to 182W (via 182Ta) provides a new kind of radiometric chronometer of planet formation processes. Hafnium and W, the parent and daughter trace elements, are highly refractory; however, Hf is lithophile and strongly partitioned into the silicate portion of a planet, whereas W is moderately siderophile and preferentially partitioned into a coexisting metallic phase. More than 90% of terrestrial W has gone into the Earth's core during its formation. The residual silicate portion, the Earth's primitive mantle, has a Hf/W ratio in the range 10−40, an order of magnitude higher than chondritic (∼1.3). Tungsten isotopic data for the Earth and the Moon suggest that we can date a major event of planet formation: The Moon formed about 50 Myrs after the start of the solar system, providing strong support for the Giant Impact Theory of lunar origin. Recent simulations of this event imply that the Earth was probably only half formed at the time. From this we can deduce the planetary accretion rate. Tungsten isotope data for Mars provide evidence of a much shorter accretion interval, perhaps as little as 10 Myrs, but the rates for the Earth over the same time interval could have been comparable. The large W isotopic heterogeneities on Mars could only have been produced within the first 30 Myrs of the solar system. Large-scale mixing, e.g. from convective overturn, as is thought to drive the Earth's plates, must be absent from Mars. Limitations of the method such as 1) cosmogenic 182Ta effects on lunar samples, 2) incomplete mixing of debris to cause W isotope heterogeneity on the Moon, and 3) initial 182Hf/180Hf heterogeneities of the early solar system are critically discussed.