Transport of Lithophile Elements in Magmatic-Hydrothermal Fluids

人们早就认识到富水（含水）和富二氧化碳（碳酸）流体在广泛的地质过程中起着重要作用。特别重要的是，这种流体有选择地将化学成分（如金属）从一个地质储层输送到另一个地质储层的能力。因此，含水流体和碳酸流体在世界上一些最具经济价值的矿床的形成中发挥了关键作用。岩浆热液矿床是地壳浅层富挥发物岩浆体冷却、相分离的产物。流体的化学性质取决于岩浆的性质，而岩浆的性质又受到岩浆活动的构造环境的影响。与破坏性板块边缘相关的岩浆活动往往以含水流体为主，而在板内岩浆活动中碳流体更为普遍。流体还含有各种阴离子（如F、Cl、S等）或阴离子络合物（CO3、SO4等），它们在金属运输中起着重要作用。尽管人们普遍认识到流体在成矿过程中的重要性，但令人惊讶的是，我们对它们的物理化学知之甚少，这反过来又限制了我们预测矿床形成的方式和地点的能力。我们的无知在很大程度上源于研究高温含水或碳流体在实验上的困难。与硅酸盐或碳酸盐熔体不同的是，流体在室温和室温下不会淬火成固体，因此很难在化学或物理上对其进行表征。我们开创了一种新的实验方法来解决这个问题，用激光钻穿冷冻实验胶囊的壁，直接分析冷冻液体，在切片过程中没有污染的风险。将激光与ICP-MS设备耦合意味着我们可以分析冷冻流体中的各种微量元素。共存的硅酸盐或碳酸盐熔体可以淬火并从胶囊中取出以供后续分析。我们可以系统地改变流体的组成及其浓度，使我们能够探索金属在流体中如何复杂的关键控制。通过观察熔体-流体分配随流体组成的变化，我们可以对存在的金属-配体复合物的类型进行假设。然而，我们不能在压力和温度下直接观察到这些。为此，我们开发了另一种实验方法，在电阻加热的金刚石砧压力池中，将一小滴已知成分的液体放在两颗金刚石的扁平尖端之间。这些钻石对同步加速器产生的x射线是透明的，这意味着这种溶液可以在高压和高温下进行原位研究。这种方法允许我们评估在划分实验的基础上做出的预测。最后，我们可以使用计算量子化学（经典和从头算分子动力学）来预测在高压和高温下液体中阳离子的水合作用和络合作用。最近实现的一种称为元动力学的技术使我们能够从分子动力学模拟中得出自由能，从而得出金属配合物形成的平衡常数。总之，我们正在从三个完全不同但又互补的方向来研究金属运输问题。就其本身而言，每种方法都有局限性；这些方法结合起来，将使我们能够在与矿体形成完全相同的物理条件下，对含水流体和碳流体产生全面的了解。我们将从研究一类重要但相对简单的金属开始，即碱金属、碱土和稀土，尽管我们的方法最终可以扩展到涵盖所有具有重要经济意义的金属。

It has long been recognized that H2O-rich (aqueous) and CO2-rich (carbonic) fluids play a fundamental role in a wide range of geological processes. Of particular importance is the ability of such fluids to selectively transport chemical components, such as metals, from one geological reservoir to another. As a consequence, aqueous and carbonic fluids play a key role in the formation of some of the most economically important ore deposits in the world. Magmatic-hydrothermal ore deposits result from the cooling and phase-separation of volatile-rich magma bodies in the shallow crust. The chemistry of the fluids depends on the nature of the magma from which they exsolved, which in turn is influenced by the tectonic setting of the magmatism. Magmatism associated with destructive plate margins tends to be dominated by aqueous fluids, whereas carbonic fluids are more prevalent in intraplate magmatism. Fluids also contain a variety of anions (e.g. F, Cl, S etc) or anionic complexes (CO3, SO4 etc) which play an important role in metal transport. Despite the universal recognition of the importance of fluids in ore formation, we have surprisingly little understanding of their physical chemistry, which in turn limits our ability to predict how and where ore deposits may form. A large part of our ignorance stems from the experimental difficulties of studying high-temperature aqueous or carbonic fluids. Unlike silicate or carbonate melts, fluids do not quench to a solid at room temperature and pressure, making it difficult to characterise them chemically or physically. We have pioneered a novel experimental approach to this problem, in which a laser is used to drill through the walls of a frozen experimental capsule, directly analysing the frozen fluid, without risk of contamination during sectioning. Coupling the laser to an ICP-MS apparatus means that we can analyse the frozen fluid for a wide variety of trace elements. The coexisting silicate or carbonate melt can be quenched and retrieved from the capsule for subsequent analysis. We can systematically vary the composition of the fluid and its concentration, allowing us to explore the key controls on how metals are complexed in fluids. By looking at the variations in melt-fluid partitioning with fluid composition we can hypothesise about the types of metal-ligand complexes that are present. We cannot, however, directly observe these at pressure and temperature. To do this, we have developed an alternative experimental methodology in which a small droplet of fluid of known composition is held between the flattened tips of two diamonds in a resistance-heated diamond anvil pressure cell. The diamonds are transparent to synchrotron-generated X-rays, meaning that the solution can be studied in situ at elevated pressure and temperature. This approach allows us to evaluate the predictions made on the basis of the partitioning experiments. Finally, we can use computational quantum chemistry (classical and ab initio molecular dynamics) to predict the hydration and complexation of cations in fluids at at elevated pressure and temperature. Recent implementations of a technique call metadyanamics enables us to derive free energies and, hence, equilibrium constants, for the formation of metal complexes from molecular dynamical simulations. In summary, we are approaching the problem of metal transport from three quite different, but complementary directions. In its own right, each approach has limitations; in combination these approaches will enable us to generate a comprehensive picture of aqueous and carbonic fluids under precisely the same physical conditions as ore bodies form. We will begin by studying an important, but relatively simple, class of metals, the alkalis, alkaline earths and rare earths, although our methodology can ultimately be extended to encompass the entire range of economically important metals.