| Summary: | 高温-超高温变质岩石的矿物组合及组构特点取决于不同的进变熔融反应,不同程度的熔体丢失以及不同程度的退变反应三种过程的综合效应.利用相平衡定量研究方法可以很好地模拟进变熔融反应的类型、P-T条件、熔体含量及其丢失行为、以及熔融过程中熔体与残余物的化学成分变化等,这对探讨高温-超高温变质作用过程以及花岗岩的成因非常重要.对平均泥质岩(APR)进行相平衡模拟表明变质泥质岩在等压(0.8GPa)升温熔融过程中可发生5种熔融反应:饱和流体固相线、白云母脱水熔融、黑云母熔融、钾长石-石榴石熔融和铝铁镁矿物熔融,后两种熔融反应主要发生在超高温条件下.减压过 程中发生怎样的熔融反应受减压温度控制:在麻粒岩相(如850℃)减压可发生钾长石熔融、黑云母熔融和钾长石-石榴石熔融反应;在高角闪岩相(如750℃)减压主要发生白云母脱水熔融和钾长石熔融;在超高温麻粒岩相(如950～1000℃)减压主要发生钾长石-石榴石熔融和铝铁镁矿物熔融.熔体成分受熔融反应和P-T条件控制,如在高角闪岩相发生的饱和流体固相线和白云母脱水熔融可形成弱过铝的奥长花岗质和二长花岗质熔体;在麻粒岩相发生的黑云母熔融和钾长石熔融形成的熔体具有强过铝的二长花岗岩成分;在中压超高温发生的钾长石-石榴石熔融和铝铁镁矿物熔融形成强过铝的二长(钾长)花岗岩质熔体,可形成石榴石花岗岩;在低压超高温下发生的铝铁镁矿物熔融可形成堇青石花岗岩.除了极端超高温下的铝铁镁矿物熔融外,其它熔融反应都会使残余物的成分更贫硅,贫Na2O和K2O,富FeO和MgO,但Al2O3和Mg#基本不变.高温-超高温下发生深熔的岩石只记录降温过程形成的固相线组合,但固相线的类型与温度条件取决于熔体的丢失行为.在不丢失熔体或者获得熔体的岩石中,岩石最后只记录流体饱和固相线组合;发生熔体部分丢失的岩石会记录缺流体固相线组合,并且熔体丢失越多,缺流体固相线的温度越高;发生全部流体丢失的岩石可记录岩石所达到的最高温度.因此,在一个麻粒岩相区,甚至一个野外露头上不同部位的岩石记录不同的P-T条件.熔体丢失是导致使麻粒岩相组合在升温过程中发生超高温变质,在降温过程中得以部分保存的重要条件.发生部分熔融的高级变质岩中随着温度升高,熔体含量增加,会发生锆石分解,只有在降温过程中发生锆石结晶,因此,麻粒岩中新生锆石只记录降温过程到固相线及以后的年龄,一般不会记录麻粒岩相峰期时代.对泥质高压麻粒岩来说,如果经历ITD型变质演化,会发生递进减压熔融,变质反应易于达到平衡,但如果减压速度快并使岩石直接抬升到地壳浅部,会出现一些ITD型结构标志,如残留金红石、蓝晶石,或在石榴石周围出现堇青石的反应冠状体等,此时锆石记录的退变质年龄会与峰期变质年龄相差不大(如10～30Myr);但如果泥质高压麻粒岩减压至中、深地壳,受其中有滞留熔体影响易于发育IBC型结构特征,表现为麻粒岩组合被(中压)角闪岩相组合叠加,在泥质岩中出现黑云母+夕线石构成的暗色条带,或者出现退变白云母和含白云母的浅色体.在中、深地壳经历IBC过程的麻粒岩锆石记录的退变质年龄会与峰期年龄相差很大(如～ 100Myr).高级变质岩中由于出现熔体使水流体活度降低,麻粒岩作为排除部分熔体的残余物,其水活度更低.从这一角度来说,水活度低是麻粒岩相变质作用的结果,而不是条件.某些麻粒岩区之所以出现多期麻粒岩相变质叠加受流体行为控制.在亚固相线下流体饱和岩石变质熔融作用从饱和水固相线开始,然后依次发生含水矿物的脱水熔融和无水矿物熔融,这一过程中流体是内部缓冲的,在麻粒岩相温度峰期形成一组平衡矿物组合,难以保留峰期之前的信息.而流体不饱和岩石(如已形成的麻粒岩或岩浆侵入体)变质作用受外部注入流体控制,与构造变形密切相关.如果发生两期麻粒岩相变质叠加变质,在强应变域会形成晚期麻粒岩组合;在弱应变域,会出现两期麻粒岩组合,其中晚期矿物表现为反应冠状体或细粒交生体;而在一些应变非常弱的区域,可能只保留早期矿物组合. The mineral assemblages and textural relations of high-temperature (HT) and ultrahigh temperature (UHT) metamorphic rocks depend on the prograde partial melting reactions, partial melt loss and partial retrogression. Using phase modeling approaches, the prograde partial melting reactions, P-T conditions, melt production and loss, and the chemical compositions of melt and residues during a melting process can be quantitatively modeled, which is very helpful for understanding the HT-UHT metamorphic processes and petrogenesis of granitoids. Phase modeling for an average pelite sample (APR) suggests that there are five types of melting reactions during an isobaric (0.8GPa) heating process, involving fluid-saturated solidus melting (HM), muscovite dehydration melting (MM), biotite dehydration melting (BM), K-feldspar-garnet melting (KGM) and the melting of Al-Fe-Mg silicates (AFM melting) after K-feldspar breakdown. The latter two types of melting mainly occur under UHT conditions. What types of melting reactions occur during decompression in metapelite depends on temperatures; the decompression melting under granulite facies (i.e. 850 degrees C) involves K-feldspar melting (KM), BM, and KGM; the decompression melting under high-amphibolite facies (i.e. 750 degrees C) involves MM and KM; and the decompression melting under UHT granulite facies (i.e. 950 similar to 1000 degrees C) involves KGM and AFM melting. The melt compositions are dependent on melting reactions and P-T conditions. For example, the HM and MM under high amphibolite facies can produce weak para-aluminous trondhjemitic to monzogranitic melt, the KM and BM under granulite facies can produce strong para-aluminous monzogranitic melt, the KGM and AFM under UHT granulite facies of medium-pressure can produce strong para-aluminous monzogranitic melt, forming garnet granite, and the AFM melting under UHT granulite facies of low-pressure may produce cordierite granite. Except for the AFM melting under UHT conditions, the other types of melting may cause the residues to be poor in SiO2, Na2O and K2O, rich in FeO and MgO, without obviously changing their Al2O3 and Mg-#. Partially melted rocks under HT-UHT conditions can only preserve the solidus assemblages in cooling processes, but the types and temperatures of the solidus depend on melt loss behavior in the rock. If there is no melt loss or there is melt gain, the rock can only retain the water-saturated solidus, and if there is melt loss, the rock will retain a fluid-absent solidus and the corresponding assemblage. The more the melt is lost, the higher the temperature of the fluid-absent solidus will be. If all melt is lost in a rock, it can record the highest temperature ever reached. As a consequence, rocks from different domains in a granulite facies region, or from different localities in a granulite outcrop, can record different temperatures due to different degrees of melt loss. In high-grade rocks with partial melting, zircon will be dissolved as melt increases with increasing temperature, and zircon growth can only occur during cooling. Thus, the newly grown zircon in granulites should record the age when cooled to the solidus, rather than the age of the peak pressure or temperature stages. A high-pressure politic granulite tends to reach equilibrium when there is progressive melting during an isothermal decompression (ITD), but if the decompression is so fast to uplift the rock to shallow crust in a short time, the textures characteristic of the ITD evolution involving residual rutile and kyanite, and cordierite corona around garnet can appear. In this case, the cooling age recorded by zircon may not be much younger than the peak stage (10 similar to 30Myr). However, if a high-pressure pelitic granulite is uplifted to mid- or lower crust and followed by an IBC evolution, it will develop characteristic IBC textures due to the presence of retained melt, with the granulite assemblages being overprinted by amphibolite facies, forming the biotite-sillimanite selvages, retrograde muscovite and muscovite bearing leucosomes. In this case, the cooling age recorded by zircon may be much younger than the peak stage (similar to 100Myr). Melting in high-grade rocks may cause decreasing of water activity, and thus, the water activity would be much lower in the granulites resulted from the residues after melt loss. From this point of view, the low water activity in most granulites should be a result of granulite facies metamorphism, rather than one of its prerequisites. Occurrence of multi-phase granulite facies metamorphism in a granulite region depends on fluid behaviors. For a fluid-saturated rock at subsolidus conditions,. its partial melting starts from the fluid-saturated solidus, and followed by the dehydration melting of hydrous phases and melting of anhydrous phases. These melting reactions are controlled by internally buffered fluid, forming one stage equilibrium assemblage at the granulite peak temperature, obscuring the pre-peak prograde information. However, metamorphism of the fluid-unsaturated rocks (i.e. the existed granulite or intrusion) is controlled by external fluid-infiltration, closely related to structural deformation. If a granulite terrane is overprinted by a late phase granulite facies metamorphism, in high-strain domains that are favorable for fluid flowing, may occur equilibrium granulite assemblages of the late phase; in low-strain domains, may appear two phases of granulite assemblages with the late phase minerals occurring as coronae or fine-grained intergrowths; while in some very low-strain domains, the early phase assemblages may be preserved. 国家自然科学基金项目 SCI(E) 中文核心期刊要目总览(PKU) 中国科技核心期刊(ISTIC) 中国科学引文数据库(CSCD) cjwei@pku.edu.cn 6 1625-1643 32
|
|---|