拉萨地块西段中新世赛利普超钾质火山岩富集地幔源区和岩石成因:Li同位素制约

摘要

As a kind of "non-traditional stable isotopes", lithium isotope system has been used to study some important geological or geocbemical problems. Recently, the research on Li isotope has became one of the most fast developing fields iff the study of crust-mantle interaction. Compared with other countries, the study in this aspect remains in the preliminary stage in China. The authors have successfully applied lithium isotope to the typical study of Sailipu ultrapotassic rocks in southwestern Tibetan Plateau. Lithium concentrations of 14 whole-rock samples from Sailipu show a range from 11.2 × 10-6 to 22.9 X 10 6. Lithium isolopic compositions exhibits a variable range of δ7Li values (- 1.2%o to + 3.5%o) with an averageδ7Li of 0.2%o that corresponds to the average of upper continental crust. Lithium isotopic compositions of UK from Sailipu do not show any significant correlations with the degree of magmalic differentiation, as inferred from various compositional parameters (e.g. , SiO2, Li, Rb and Ga). This suggests insignificant Li isotope fractionation during ultrapotassic rock differentiation, reflecting the source characteristics. Lithium isotopic compositions of these samples vary by 4.7%o and do not correlate with radiogenic isotopic compositions or chemical and mineralogical parameters. The Li i.sotopic heterogeneity therefore likely reflects heterogeneous source rocks. Based on calculation modeling and a comparison with previous similar results, the authors hold that the most probable metasomatic agents were melts or fluids derived from subducted Indian crust instead of from Tethyan crust (including sediments). Therefore, the authors put forward a petrogenetic model of ultrapotassic rocks as follows: With the Neo-Tethyan o-ccan closure and onset of the Indian-Asian continent collision, the heavy oceanic lithosphere might have dragged mainly Indian plate into the subduction zone. Upon heating under high-pressure conditions the subducted Indian crust probably melted, or at least released high potassium/lithium fluids that metasomatized the overlying Tibetan lithospheric mantle. Slab-derived melts or fluids would have to produce extensive regions in the Tibetan lithospheric mantle of low δ7Li values which would be exchanged with the surrounding lithospheric mantle via diffusive disequlibrium, so successive injections of alkali/Li-rich fluid/melt should ultimately generate a mantle source with an average S Li of the slab influx. The northward subducted Indian lilhosphere-mantle might have experienced slab break-off beneath the Lhasa block along the Yarlung Zangbo suture zone in early Miocene, which caused the asthenospheric upwelling under the Indian plate through the slab window. Upwelling of the hot asthenosphere following the break off presumably caused partial melting of overlying metasomatized and contaminated lithospheric mantle domains that could have produced ultrapotassic primitive magmas, which formed the ultrapotassic volcanic rocks along the weak structure belts.%作为一种“非传统稳定同位素”,锂同位素地球化学研究已经成为近年来国际上研究的热点之一.文章成功应用锂同位素对青藏高原西南部赛利普超钾质火山岩进行了示范研究.研究表明,赛利普超钾质火出岩的w(Li)为11.2×10-6～22.9× 10-6,同位素组成δ7Li为1.2‰～+3.5‰,平均值为0 2‰,与平均上地壳的相当.超钾质火山岩的锂同位素组成与岩浆结晶分异程度参数之间不存在任何相关性,这表明在超钾质火山岩结晶分异过程中没有发生明显的锂同位素分馏,锂同位素组成特征反映了其形成时的源区特征.超钾质火山岩的锂同位素组成变化范围达4.7‰,并且与pb-Sr-Nd同位素和岩浆结晶分异参数之间亦无任何相关性,表明锂同位素异常可能反映了不均匀源区岩石特征.通过计算模拟以及与前人的类似研究成果进行对比,笔者认为俯冲印度地壳而不是特提斯洋壳(包括沉积物)的流体/熔体参与了超钾质火山岩的源区富集,并在此基础上提出了超钾质火山岩成因模式.