Negative 187Os/186Os vs. 143Nd/144Nd Correlation Observed in the Continental Crust

Qingzhu Yin Max-Planck-Institut für Chemie, Abteilung Kosmochemie, Postfach 3060, D-55020 Mainz, Germany; present address: Harvard Center for Isotope Geochemistry, Department of Earth and Planctary Sciences,

Harvard University, 20 Oxford Street, Cambridge, MA 02138, USA

qingzhu@geochemistry.harvard.edu

Alex Rocholl Mineralogisches Institut, Univ. Heidelberg, Im Neuenheimer Feld 236, D-69120 Heidelberg, Germany

Emil Jagoutz Max-Planck-Institut für Chemie, Abteilung Kosmochemie, Postfach 3060, D-55020 Mainz, Germany

We have studied for the first time the Os concentration and isotopic composition in two "new" types of samples that act as integrators of the local continental crust.

Ferromanganese coatings

These form on the sand grains used as filter banks
in water plants during the oxidative extraction of
dissolved Fe+2. Groundwater is flushed through the sand
and intensively mixed with air, oxidizing Fe+2 to Fe+3 and precipitating it as Fe-hydroxide onto the sand grains along with Mn and other ions.

One sample is from the water plant of Schweinfurt,
N-Bavaria
. The water source consists of 2/3 of Main River water infiltrating the sandy and gravel beds on the side of the river bank, and 1/3 of ground water from the Triassic "Muschelkalk-Unit". The measured 187Os/186Os ratio is 10.226±0.007, very similar to what was typically measured in river sediments.

Similar coating sample from the Wasserwerk Rheinwald, Karlsruhe, SW-Germany, has also been analyzed for both
Os concentration and isotopic composition. The sample consists of 60% MnO2, 20% Fe2O3, both in the amorphous hydroxides form, and 20% carbonates plus other minor components. The water source consists of 100% Rhein River water infiltrating the sandy and gravel beds on the side of
the river bank. The Os concentration of the bulk sample
(i.e. including the sand core) is ~10 ppt, while the Os concentration on the sand grain-free basis is ~200 ppt. Similar range of concentration has been observed for the modern marine sediments and the marine Mn nodules. It is noteworthy that the Os isotopic composition of this sample is significantly lower than any continental samples so far being measured for Os isotopes. The measured 187Os/188Os = 0.62237 ± 57. This corresponds to 187Os/186Os = 5.19 relative to 186Os/188Os = 0.1199. Since about 20 g of the sample has been used for this run, 1 pg of Os blank in the total chemical procedure will not alter the measured isotopic composition of the sample. The interference on mass 233 would result in an uncertainty
in the measured 187Os/188Os ratio of 1.6% maximum (233/236 = 0.01). The industrial contamination of the Rhein River by noble-metals is a possible explanation.

Red Beds

The other type of sample studied is the Red Bed sample from Sri Lanka. It is a natural ferric oxide soil that formed as a chemical weathering products. Various sources of
Os may become homogenized and concentrated by Fe-oxide in Red Beds. Our sample covers the Vijayan (Wanni) Complex, in the Northwest Sri Lanka, the type locality for middle to lower crustal section. The measured 187Os/188Os =1.0440±.0025. The corresponding 187Os/186Os = 8.71 relative to 186Os/188Os = 0.1199.

Conclusions

Osmium isotopic measurements in ferromanganese oxide coatings on sand grain filters from water-plants in Germany and in ferric oxide from Sri Lankan Red Beds show that these samples concentrate Os and act as integrators of local continental crust. The 187Os/186Os ratios of 8.71-10.226 are comparable with those of river sediments and demonstrate the feasibility of mapping the Os isotopic composition of the upper continental crust with this type of sampling.

A negative correlation between Os and Nd isotopic compositions observed for the first time in the upper continental crust is probably a natural consequence of the extraction of the crust from the mantle. The Os and Nd isotope data for manganese nodules from Indian Ocean are found to fall on the extension of the continental Os-Nd correlation line due to the similarity in 187Os/186Os ratios between detrital inputs around Indian Ocean and global seawater average. The new crustal Os parameters are used to assess the role of crustal recycling in the genesis of mantle components. It provides a new constraint for the present day mixing processes in the source region of EM II type basalts, i.e. the mixture of depleted mantle and recycled oceanic crust (HIMU source) could not mix with terrigenous sediments to generate less radiogenic-Os EM II source. We estimate the average 187Os/186Os ratio for the upper, whole, and lower continental crust to be 14, 7.5, and 6 respectively