震击高压变质作用:观察与启示

近六十多年来,关于造山带高压变质作用发生的原因一直存在争议。当前的主流观点认为高压变质岩形成于俯冲带深部以后折返到地壳。虽然早就认识到蓝片岩相变质作用总是与逆冲断裂相伴并且变质压力朝着冲断面增高,而标志地震的榴辉岩相假玄武玻璃也已经发现了二十多年,但是一般认为断裂或剪切带的作用只是引进流体促进深部岩石的变质反应。几年前,苏文辉等人(Su et al.,2006)通过实验演示了石英经过非晶化可以在柯石英稳定的压力下迅速转变为柯石英,提出无需深俯冲,浅源地震即可形成柯石英。本文介绍笔者等人对中国苏鲁超高压变质带仰口榴辉岩的部分工作结果,从地质观察角度为这一论点提供证据。例如,粒间柯石英仅出现在榴辉岩相角砾岩的角砾而非胶结物中,它必须在相当于地震的时间尺度快速降压冷却才能得以保存。变辉长岩中的榴辉岩相碎裂岩脉也记录了应力导致的瞬时高压和高温。角砾岩和碎裂岩中包含大量高压矿物的石榴石和针状蓝晶石以及局部出现的"显微花岗岩"都是类似地震熔体淬火的结果。它们指示震后高压矿物没有再生长。而如果地震发生在深部,流体浸入后矿物应当在高压下持续结晶,从而消除淬火结构。碎裂岩脉中星散的铬铁矿微粒极可能是附近超镁铁岩通过断裂迸溅进入基性岩的,是地震中不同岩石机械混合的证据;环绕它们的富铬榴辉岩矿物记录着一次无可争议的瞬时高压结晶事件。碎裂岩脉相对围岩更缺乏流体的事实反映应力而非流体在高压相变中的关键作用。榴辉岩相角砾岩和碎裂岩脉形成的时间尺度太小,来不及完成俯冲和折返过程。它们更可能是辉长岩在地壳原地因地震波引起的高压发生了榴辉岩化作用。目前的地震力学模型不能导出高压相变所需要的压力,只是因为建立模型时没有考虑涉及高压相变的地震。需要强调的是,假说或理论需要观察事实的检验而不是相反。已有资料显示,瞬时熔融和非晶化可能是震击高压相变的路径。

Coseismic high-pressure metamorphism: Observations and implications

The last six decades witnessed a continued debate between the hypotheses of subduction-exhumation and tectonic overpressure for the formation of high-pressure metamorphic rocks in orogenic belts. The currently widely held viewpoint is that crustal rocks were subducted to mantle depths, where they were subjected to high-pressure metamorphism and were then exhumed back in the crust. Although blueschist facies metamorphism has been recognized to be related to thrust faults and the metamorphic pressure increases towards the fault planes, and eclogite pseudotachylyte indicating fossil earthquake was found 20 years ago, it is generally considered that the role of faults or shear zones is to introduce fluid that enhances metamorphic reactions at depths. Recently Su et al. (2006) have demonstrated that quartz can transform rapidly to coesite at a pressure in the stability of coesite after being amorphized. They suggest that quartz may transform to coesite during a large earthquake in the crust and deep subduction may not be needed for high-pressure metamorphism to take place. The present review introduces the recent work by the writer and co-workers on some eclogite facies breccia and cataclasite dykes at Yangkou in the Chinese Su-Lu ultrahigh-pressure metamorphic belt, providing geological evidence for this argument. It is pointed out that intergranular coesite occurs only in the fragments but not in the matrix of the eclogite breccia; it can only be preserved by rapid decompression and cooling at the timescale of a seismic event. The eclogite facies cataclasite dykes also record transient higher pressure and temperature relative to the host metagabbro. The micropoikilitic garnet containing numerous high-pressure mineral inclusions, the kyanite needles, and the 'microgranites' in both types of rock are interpreted to be the results of quenching of seismogenic melts or amorphous phases. They indicate no further growth of the high-pressure minerals occurred after the seismic events. This is in contrast to the general idea that high-pressure crystallization occurs only in subducted dry rocks when fluid becomes available, in which continuous crystallization afterwards will erase the quenching textures. Sparse chromite particles in the cataclasite dykes indicate mechanical mixing between the metabasic rocks and the nearby ultramafic rocks during seismic faulting. The Cr-rich eclogite minerals around the chromite grains are unequivocal evidence for transient high-pressure crystallization. The fact that the cataclasite dykes are less hydrated than their host rocks implies the key role of stress rather than fluid during high-pressure phase transformation. The timescale of formation of these high-pressure rocks are too short for the rocks to complete the subduction-exhumation processes. They are more likely the results of eclogitization of a gabbro under high-pressures given rise to by seismic waves in the crust. It is noted that the necessary high pressure cannot be derived from the currently used mechanical models for earthquakes. This is simply because high-pressure phase transformation has not been taken into account in developing the models. It is emphasized that hypotheses and theories are to be tested by observations but not vice versa. The available data show that flash melting and amorphization are the probable mechanisms for transient high-pressure metamorphism.