Calculation of effective bulk composition and its application in metamorphic phase equilibria modeling

Abstract

变质相平衡模拟是变质岩领域近几十年最重要的进展之一，它已经成为确定变质作用P-T-t轨迹和探索变质演化过程的有力工具。变质岩的矿物组合不但与其形成的温度（T）和压力（P）条件有关，而且受控于岩石的全岩成分（X）。但是变质岩通常是不均匀的并且往往保留两期以上的矿物组合，因此计算不同成分域或不同变质演化期次的有效全岩成分是模拟P-T视剖面图的核心问题之一。在中-低温变质岩中，石榴石变斑晶的生长会不断地将其核部成分冻结而不参与后续变质反应，这导致根据实测全岩成分计算的P-T视剖面图无法有效地模拟石榴石幔部或边部生长阶段的变质演化过程。瑞利分馏法和球体积法利用电子探针实测的石榴石成分环带可以模拟计算石榴石各个生长阶段所对应的有效全岩成分，本文推荐使用这两个方法来处理石榴石变斑晶的分馏效应问题。相比较而言，石榴石在高温变质岩中通常无法保留生长阶段的成分环带特征，这是因为石榴石成分在高温条件下会发生扩散再平衡，并同时与多数基质矿物达到热力学平衡，这时一般不需要考虑石榴石的分馏效应。但是高温变质岩通常会发生部分熔融并伴随熔体的迁移，进而改变岩石的有效全岩成分。因此，通过P-T视剖面图模拟熔体迁移前后的变质演化过程需要使用相平衡法计算迁移的熔体成分以及熔体迁移前后岩石的有效全岩成分。此外，后成合晶与反应边是变质岩中最常见的退变质反应结构，但是后成合晶或反应边中的矿物之间并未达到热力学平衡。这种情况需要结合岩相学观察和矿物成分，利用最小二乘法确定后成合晶或反应边中发生的平衡反应方程式，进而获取变质反应发生时的有效全岩成分并通过计算P-T视剖面图来估算退变质的温压条件。除此之外，岩石体系中三价铁（Fe₂O₃）和H₂O含量的估算一直以来都是相平衡模拟研究中的难点，本文推荐使用P/T-X（Fe³⁺/Fe^tot，MH₂O）视剖面图来确定这两个组分的含量，这是因为P/T-X图可以估算各个变质演化阶段或特定矿物组合的Fe₂O₃或H₂O含量。;Phase equilibria modeling is one of the most important advances in metamorphic petrology in recent decades,which has become a useful tool for determining the P-T-t paths and understanding the metamorphic evolution. The metamorphic mineral assemblages are not only determined by pressure (P) and temperature (T), but also controlled by the bulk-rock compositions (X). However, the metamorphic rocks are commonly heterogeneous and preserve two or more mineral assemblages. Thus, to perform P-T pseudosections, it is critical to calculate the effective bulk compositions of different compositional domains and/or mineral assemblages. In low-medium temperature metamorphic rocks,the growth of garnet porphyroblasts will continuously isolate the garnet core from the matrix. Thus, the P-T pseudosection calculated using the analyzed bulk-rock composition may not represent the metamorphic evolution related to the growth of garnet mantle or rim. Using the measured garnet zoning profiles, both Rayleigh fractionation and ball volume methods can be applied to calculate the effective bulk composition corresponding to each stage of garnet growth. Therefore, it is recommended to treat the garnet fractionation through these two methods. By contrast, garnet does not commonly show growth compositional zoning profile in high-temperature metamorphic rocks, because the garnet composition will suffer volume diffusion and re-equilibrated at high temperature conditions. This will lead to the thermodynamic equilibrium between garnet and matrix minerals. In this case, it is unnecessary to consider the garnet fractionation. However, high-temperature metamorphic rocks usually undergo partial melting and melt migration, which in turn changes the bulk-rock compositions. Therefore, to model the metamorphic evolutions before and/or after melt migration through the P-T pseudosection requires the phase equilibria method to calculate the migrated melt compositions and the effective bulk compositions before and/or after the melt migration. The symplectites and corona are the most common retrograde microstructures in metamorphic rocks, but the thermodynamic equilibrium has not been reached between the minerals in symplectites and corona. In this case, we suggest determining the equilibrium reaction equation that occurs in the symplectites and corona through the least square method combined with the petrographic observation and mineral compositions. Thus, the bulk composition of the reactants or products represents the effective bulk composition when the equilibrium reaction occurs. In addition,the estimation of ferric iron (Fe₂O₃) and H₂O content in bulk-rock composition is always critical to the phase equilibria modeling. It is suggested to apply the P/T-X (Fe³⁺/Fe^tot, MH₂O) pseudosections to determine the content of these two components. Because the P/T-X pseudosections are robust to estimate the Fe₂O₃ or H₂O content for any metamorphic evolution stage or mineral assemblage.