Single mineral dating by the Pb-Pb stepwise leaching (PbSL) technique has been employed to date rock-forming minerals such as titanite (Frei and Kamber, 1995), staurolite (Frei and Biino, 1996), pyroxene (Frei, 1995) and garnet (Frei and Kamber, 1995; Frei, 1995).
One extremely important question at this point is assessing the mechanism by which large spreads in the 206Pb/204Pb (and 207Pb/204Pb) ratios are obtained, and under what conditions 207Pb/206Pb isochrons can be obtained at all. Only when these points are explained will a generalized use of PbSL be legitimate.
Our sample was a gem-quality museum specimen of titanite from Otter Lake, Canada. We split the crystal in
4 pieces on which we performed a number of investigations.
1. A conventional bulk titanite U/Pb concordant age of 999±7 Ma was obtained on the 100-200 µm sieve fraction of a gently ground crystal piece.
2. PbSL on two size fractions (50-100 µm; 200-300µm) with HNO3 gave an isochron age of 1002±12 Ma comprising all five leach steps. Grain-size dependence and time dependence of the Pb release rate follow a first-order kinetics law expected for dissolution. The release of Th-derived Pb (given by the 208Pb/204Pb ratio) followed closely that of U-derived Pb, with small but significant variations only in the residue (step 5).
3. PbSL on the same fractions using HBr duplicated the HNO3 age results (1000±10 Ma). All 25 leach steps together give 998.0±5.5 (2*).
4. ICP element analysis on the HNO3 and HBr leach solutions showed that Fe and Ca were dissolved following first-order kinetics but Ti and Si were depleted in the leach solutions.
5. SEM photographs of grains from each leaching step show that reprecipitation of a palisade of minute euhedral Si+Ti-rich particles accounts for observation (4).
6. Three-isotope and three-element correlation diagrams (206Pb/204Pb vs 1/204Pb, Th/Pb* vs U/Pb*) and substitutent /major element ratio diagrams (Pb/Ca vs U/Fe) show that at least 3 (possibly 4) different reservoirs are required to account for inter-step variation trends. In these graphs, leach fractions of oxidizing and reducing acids follow slightly but significantly different trajectories. Both observations, as well as the 208Pb/204Pb ratio of the residue (point 2 above),
support the existence of heterochemical inclusions in the titanite.
7. Electron microprobe traverses of a polished crystal slice show Y concentrations ranging from 0.12% to 0.40% across the grain, compatible with both the presence of Y-rich inclusions and of large zonations.
8. Proton microprobe (PIXE) imaging of untreated and leached grains shows, in addition to the fine structure of the reaction front, the existence of Pb+U-rich inclusions having a size smaller than 1 pixel (×2 µm). Some other "inclusions" show no U enrichment paralleling that of Pb and may be either noise in the Pb counting or true Pb-only inclusions. Multivariant analysis of the Pixe data indicate that elements usually considered geochemically analogous are allocated to distinct principal components. This might be coupled to the obsevation of sub-µm particles with very uncommon
element associations.
Submicroscopic inclusions are present even in gem-
quality Otter Lake skarn titanite. Distinct, selective dissolution of host titanite and inclusions is able to account for the increased spread along the axes of the Pb/Pb isochron. Heterochemical inclusions can be presumed from electron microprobe data on high concentration trace elements (Y: > 1”) and is confirmed both by isotope correlation
systematics and by direct PIXE imaging.
A necessary and sufficient condition for the presence of a Pb/Pb isochron is that these inclusions achieved isotopic equilibrium during the metamorphic reaction. Small
inclusion size alone does however not guarantee complete equilibration, as very retentive inclusions preserved isotopic disequilibrium in some samples of Frei and Kamber (1995).
Frei, R., Extended abstract, Centennial Geocongress, Johannesburg, 64-67 (1995).
Frei, R., Biino, G.G., Prospert C: Geology 23 (1996, in press).
Frei, R., Kamber, B.S., Earth Planet. Sci. Lett. 129, 261-268 (1995).