New Insights into Granite Petrogenesis From Experimental Studies of Hydrous Melting, Water Solubility, and Supercritical Fluids

花岗岩是地球上最重要的岩石之一。它们是大陆地壳的中心组成部分，承载着比地球最近2亿年历史更早的地质记录中几乎所有的信息。这些岩石有助于确定过去板块汇聚的位置。它们可能是地球形成时期最早形成的岩石之一？地球的早期演化，可能早在地球起源的3亿年之内。然而，尽管它们很重要，关于花岗质岩石的起源仍然存在令人惊讶的争议。我们知道形成花岗岩岩浆需要水。但岩石学家对这些水的来源意见不一。广泛接受的模型是，水是由含水矿物输送的，这些矿物在融化区域变得不稳定；然而，这并不能解释花岗岩成因的关键方面，也与大陆地壳趋同边缘岩浆作用的一般模式不一致。另一种说法是，花岗岩是在大陆地壳的开放体系熔融过程中产生的，而地壳深处流动的流体促进了花岗岩的形成。然而，人们对这种流体的基本化学性质知之甚少，这阻碍了对替代假设的适当测试。该项目旨在通过对富水流体在花岗岩成因中的作用的新实验限制来解决这一问题。这项工作将利用加州大学洛杉矶分校开发的热液活塞缸方法，在这种方法中，大量的水可以被密封并保留在贵金属胶囊中。初步研究表明，简单矿物-水或岩石-水二元体中各相区的结构诊断是可重复性的。在不明确的情况下，可以用简单的痕量元素监测器来检查纹理解释，这些监测器可以独立识别液体、蒸汽或液体+蒸汽场。当一起部署时，现在可以获得完整T-XH2O二元体的准确和精确等压相关系。将进行三项调查。第一个将确定钠长石- h2o和相关二元中的相关系。我们将首先完成本文所述钠长石- h2o的初步研究。这项工作绘制了整个二元体系的等压温度-组成关系，包括固体、液体、晶体在水中的溶解度、液气混相间隙及其闭合。当不同压力的二元组合时，可以建立临界曲线和第二临界终点，达到以前无法获得的精度和准确度。该研究将扩展到硬玉- h2o和石榴石- h2o二元结构，为建立NaAlSiO4-SiO2-H2O体系关键部分液体-水相互作用的模型奠定基础。在第二个实验项目中，我们将以矿物- h2o二元体为基础，研究单长花岗岩- h2o体系。最后，我们将研究当加入CO2或碱卤化物降低H2O活性时相关系的变化。每组实验的数据将作为h2o -硅酸盐混合热力学模型的基础，首次为理解液-气混相间隙的拓扑结构、它们与含水熔融的联系以及花岗岩体系中超临界硅酸盐熔体- h2o流体的热力学建立理论基础。完成后，这项工作将导致对花岗岩岩石成因中含水流体作用的新认识，这将为会聚边缘岩浆作用和大陆地壳演化模型提供信息。

Granites are one of our planet's most important rocks. They are a central component of the continental crust, which carries nearly all information in the geologic record older than the most recent 200 million years of earth history. These rocks help identify past sites of plate convergence. And they may have been among the first rocks to form during earth?s early evolution, perhaps as early as within 300 million years of Earth's origin. Yet despite their importance, there remains surprising controversy about the origin of granitic rocks. We know that water is required to create granitic magmas. But petrologists do not agree on the source of this water ? the widely accepted model is that water is delivered by hydrous minerals that become unstable in the melting region; however, this fails to explain key aspects of granite genesis and is inconsistent with more general models of convergent-margin magmatism in the continental crust. An alternative is that granites are generated during open-system melting in the continental crust, and that migrating deep-crustal fluids facilitate granite creation. However, the basic chemical properties of such fluids are poorly understood, hindering proper testing of the alternative hypothesis. This project seeks to address this problem through new experimental constraints on the role of water-rich fluids in the genesis of granites.The work will exploit hydrothermal piston-cylinder methods developed at UCLA, in which large volumes of H2O can be sealed and retained in noble metal capsules. Preliminary investigations demonstrate that textures diagnostic of each phase region in simple mineral-H2O or rock-H2O binaries can reproducibly be obtained. Where ambiguous, the textural interpretations can be checked with simple trace-element monitors that independently identify liquid, vapor, or liquid+vapor fields. When deployed together, accurate and precise isobaric phase relations across full T-XH2O binaries can now be obtained. Three lines of investigation will be pursued. The first will determine phase relations in albite-H2O and related binaries. We will first complete preliminary studies of albite-H2O described herein. The work maps isobaric temperature-composition relations across the full binary, including solidus, liquids, crystal solubility in H2O, and the liquid+vapor miscibility gap and its closure. When binaries at different pressures are combined, the critical curve and second critical end point can be established to previously unobtainable precision and accuracy. The studies will be expanded to the jadeite-H2O and nepheline-H2O binaries, forming a foundation for modeling liquid-H2O interactions in key parts of the system NaAlSiO4-SiO2-H2O. In the second experimental program, we will build on the mineral-H2O binaries by investigating the haplogranite-H2O system. Finally, we will examine the change in phase relations when H2O activity is reduced by addition of CO2 or alkali halides. Data from each set of experiments will be used as the basis for thermodynamic models of H2O-silicate mixing to establish a theoretical foundation for understanding, for the first time, the topologies of liquid-vapor miscibility gaps, their links to hydrous melting, and the thermodynamics of supercritical silicate melt-H2O fluid in granitic systems. When complete, the work will lead to new understanding of the role of aqueous fluids in granite petrogenesis, which will inform models of convergent margin magmatism and evolution of the continental crust.