Excess Argon-free Phengite in Ultrahigh-Pressure Metamorphic Rocks From the Lago di Cignana Area, Western Alps

In ⁴⁰Ar/³⁹Ar phengite analyses, UHP rocks have provided discordant ages. However, we have had a working hypothesis that a rock with a simple metamorphic history should have an insignificant amount of excess Ar. We have carried out Ar/Ar phengite analyses on six samples collected from three sites in an area (20m x 200m) by the Lake Cignana in the Zermatt-Saas zone consisting of Tethyan oceanic lithosphere which have suffered UHP metamorphism. Samples can be classified into three groups; garnet-phengite schist (01B, 02A), zoisite-clinozoisite schist (01A, 02B, 03B) and piemontite schist (03A). Sample 02A has coesite included in garnet. They have different mineral assemblage, but should have experienced the same P-T-t history during the UHP metamorphism and the subsequent exhumation. EPMA data indicate that phengites have more or less chemical variations in each sample and among samples. The former characteristics have formed during the retrograde chemical reaction and deformations, and the latter were due to difference of bulk chemistry of rocks. Ar/Ar analyses were carried out with a laser probe step-heating method using a single crystal and a spot dating method using thin sections. In age spectra by step heating, the zoisite-clinozoisite schists gave plateau ages of 38 and 42Ma, the garnet-phengite schist, 37-39Ma and the piemontite schist,37Ma, indicating that phengites have similar ages despite their chemical variations. This suggests that the closure temperature of phengite has insignificant dependence on its chemistry. Some phengites show disturbed age spectra with significantly young age fractions (ca. 30Ma) which are due to overcorrection of non-radiogenic Ar by an apparently large amount of ³⁶Ar. Spot dating was carried out on both matrix phengite and inclusion phengite in garnet. The result shows that the matrix phengite gave 39Ma, and the inclusion phengites yield 43-44Ma which are slightly older than the plateau ages by step heating method. Ages obtained in this study are consistent with SHRIMP age (44Ma: Rubatto et al., 1998), and Sm-Nd and Rb-Sr ages (40 and 38Ma, respectively: Amato et al., 1999). This indicates our working hypothesis was correct in a case of UHP rocks from this area.