How do metals escape from magmas to form porphyry-type ore deposits?

Abstract

Many precious and base metals (Cu, Au, Ag, Mo, W, Sn) in porphyry-type deposits are extracted from a granitic to granodioritic silicate melt and transported by magmatic-hydrothermal fluids to the site of ore deposition. Metals abundance increases from some ppb in the melt to % in ore bodies. Assuming a unique magmatic source for metals, it implies an exceptional combination of enrichment process acting over up to 4 orders of magnitude. Previous models of magma differentiation and hydrothermal circulations omitted to consider the role of metal segregation during magma chamber evolution. Models for the formation and evolution of intrusions have recently shifted from the so-called melting-storage-assimilation-homogenization (MASH) paradigm, basically steady state, to the mantle-melting-segregation-ascent-emplacement (m(M-SAE)) paradigm, dynamic and discontinuous in time. Successive magma injections of variable compositions progressively build the magma chamber. This is supported by field relations (e.g. cross-cutting dykes and stocks), textures (e.g. partially resorbed enclaves, or mineral fabrics), geochemistry (e.g. hybridization signatures), and age dating aswell as istopic studies. As melt crystallizes and forms a mush (>50% crystals), volatiles also exsolve forming a magmatic volatile phase (MVP) while melt motion slows down. Metals partition between the three phases: melt; crystals; and MVP. They generally prefer the MVP owing to more favorable partition coefficients. This is indicated by metal content in coeval melt and fluid inclusions with homogenization temperatures above 600 °C, or in volcanic fumaroles. Here, by simulating the physical interactions between the three phases we suggest a model of fluid sparging, during which metals segregate towards the MVP by diffusion and are further transported by advection as metal complexes. The model also provides estimates of metal enrichment. An undimensional Péclet number rules the competition between diffusion and advection, basically scaling with respect to the inverse of the product of melt viscosity (η) by metal diffusivity (D). The threshold value of the Péclet number between diffusion and advection is roughly 10−9. Fast diffusive metals (Au, Cu, Ag, W) readily diffuse from silicate melt and towards bubbles of the MVP. To escape the mush through tubular structures, the MVP must overcome a critical gas saturation level (about 20% vol.) usually reached after several magma injections. Advection then takes over and transports the metal-enriched MVP towards the top of the magma chamber. This leads to nearly coeval (1) separation between a high salinity-liquid phase and a low-salinity vapor phase, (2) fluid-rock interactions resulting in potassic, advanced argillic and phyllic alterations and (3) metal deposition. The metal enrichment scales as the ratio between partition coefficient (i.e. related to the gradient in chemical potential), diffusivity (i.e. related to the gradient of concentration), and melt viscosity (i.e. related to the gradient of momentum). The rates at which all such gradients relax determe metals enrichment, inducing chemical and physical instabilities, leading to a cyclic process. The whole cycle also encompasses the case of a partial, non-completed, full enrichment yielding to barren intrusions. A tentative model generalizes the sparging fluid model to other metal deposits linked to an intrusion. Such generalization should be interpreted as predicting metal enrichment by 3–4 orders of magnitude, rather than predicting an exact value.