节点文献

660km深度俯冲板块流变强度的高温高压实验研究及其动力学应用

Experimental Investigation on the Strength of the Material Near the660Km Discontinuity at High Pressures and Temperatures and Its Implications for Geodynamics

【作者】 赵素涛

【导师】 金振民; Harry W. Green II;

【作者基本信息】 中国地质大学 , 构造地质学, 2012, 博士

【摘要】 660km地震不连续面一般对应着地球深部林伍德石(ringwoodite,Mg2SiO4)向钙钛矿(perovskite, MgSiO3)和镁方铁矿(magnesiowustite, MgO)的相变,是地球内部分开上地幔和下地幔的重要地球动力学界面,深俯冲板块在这一深度附近发生了许多重要的演化,譬如深源地震的彻底消失,深俯冲板舌发生偏转后才进入下地幔等。因此,了解和把握660km地震不连续面附近物质的力学属性及其边界本身的化学/结构属性是当前地球深部物质科学和动力学研究的重要前沿课题。而高温高压实验研究是认识深部地球动力学过程的重要途径之一,它使得我们可用实验室的观察结果来代替仅仅是推测的假说,所以我们要充分认识实验研究工作在解决未来固体地球科学难题中的地位和意义。目前,国际上对上地幔橄榄岩等已经有较多的研究,而对660km地震不连续面附近的物质(钙钛矿和镁方铁矿)由于实验仪器的限制而研究很少,这在很大程度上限制了我们对660km深度俯冲板块的变形和演化及相关地球动力学问题的认识。本论文主要利用能很好控制变形并能精确确定其强度的加州大学河滨分校Green教授实验室的高温高压Griggs流变仪对与660km深度不连续面附近结构完全相同的物质(Co2TiO4→CoTiO3+CoO)进行了一系列开创性的初步研究工作,具体研究内容包括以下几个方面:a)660km地震不连续面附近物质分解前后强度变化的流变学实验研究;b)660km地震不连续面附近物质电子背散射衍射(EBSD)初步研究;以及c)以我国唯一的深源地震区——东北深震为例探讨了深源地震的发震机理及为什么大于660km深度以后没有地震发生的机制。对660km地震不连续面附近物质分解前后强度变化的流变学实验研究初步确定了1)分解产物(CoTiO3+CoO)的确具有后成合晶结构——表面上看是微小细粒的,在二维空间上看是独立的单个晶体,但在三维空间上它们很有可能是连在一起的。2)相同条件下后成合晶(CoTiO3+CoO)的强度比Co-Ti尖晶石(Co2TiO4)强约25%,说明这种后成合晶结构是一种使物质变强的结构而不是像前人提到的由于颗粒度变小而使物质变弱的结构;应用到地球科学中,俯冲岩石圈当其进入下地慢时并没有变弱而是比其下伏的地幔转换带具有更大的粘度,这与地球物理上的证据是一致的。利用我们的结果能够很好的解释俯冲陡倾板舌中(包括汤加以及我国的东北地区)由于下地幔粘度的加大导致的对插入板片的抵抗力致使其应力状态以下倾压缩应力为主的情况。之外,在尖晶石相分解的一系列试验中并没有发现任何的剪切不稳定性,那样深源地震在转换带底部停止的最可能解释或者是亚稳态橄榄石已经消耗殆尽(但已经在4个俯冲带内发现此深度橄榄石依然存在)或者就是像此文研究中的那样生成了不产生剪切不稳定性的这种后成合晶结构。另外,显微构造和岩石组构分析是构造地质学中一项重要基础工作,它通过研究变形岩石的显微构造和结晶学优选方位(LPO)来揭示宏观构造变形规律、应力状态、运动方式和形成机制等构造信息,对于研究地质体变形环境、变形机制和变形动力学过程非常关键,也是约束和建立地球深部地质与地球物理模型的重要支撑。最近十几年来,装备在扫描电镜上的电子背散射衍射(EBSD)技术日臻成熟,已经成为材料科学和地球科学显微组构分析的强有力手段。作为构造地质学中的革命性技术方法,EBSD能提供完整的量化显微构造和结晶学测量数据,实现矿物相鉴定、变形机制、位错滑移系、变质过程和岩石物理性质模拟等研究。我们对变形前后CoTiO3+CoO后成合晶的EBSD测定结果显示:1)变形前后成合晶结构中的CoTiO3和Co0在三维上的确是连接在一起的,并且它们同时生长还遵循一定的方向,即CoO的{111}//CoTiO3的{0001}、CoO的{110}//CoTiO3的{10-10}和CoO的{100}//CoTiO3的{2-1-14}或{1-214};2)后成合晶变形过程中,发生了动力学重结晶和颗粒生长,但并没有任何证据显示由颗粒边界滑移主导的“超塑性”流动使后成合晶结构弱化的现象;对变形后的Co-Ti尖晶石(Co2TiO4)和变形前后的分解产物(CoTiO3+CoO后成合晶)分别进行了电子背散射衍射(EBSD)晶格优选方位的初步研究,结果表明:1)变形后的Co-Ti尖晶石和变形前后的分解产物中的CoO其晶格优选方位都不明显,它们的极密图中均显示近随机分布的组构特征,对应的方位差角相对频率分布图也都接近理论随机分布曲线。可能这是由于我们的轴向压缩实验样晶变形量较小或者是其高对称性晶体结构(立方晶体)所导致的,因而不能用来推导其变形机制;2)变形后的分解产物中的具有三方结构的CoTiO3存在较强的晶格优选方位,{110}面和{100}面极点为近于平行面理(σ2和σ3组成的面)分布的大圆环带,其最大极密值对应轴向近于平行线理;{001}面极点垂直线理方向,最大极密值接近Z轴(σ1轴向压缩)方向;说明CoTiO3内的位错滑移很可能是主导其塑性变形和形成较强优选方位的主要变形机制,且(001)面是其主滑移系中的易滑移面;也就是说分解后这两种物质不是各自以颗粒边界滑移来进行变形,而是后成合晶以整体的位错蠕变来进行变形的。之外,深震是指震源深度大于300km的天然地震,研究深源地震的机理将有助于深入地了解板块构造的驱动机制、动力学特征;而通过深源地震体波的研究还可以获得许多与地球深部构造有关的物质状况和性质的重要信息,从而加深对地球内部结构的认识;同时深源地震也是人类潜在的自然灾害之一(例如1970年6月31日发生在南美哥伦比亚的深震和1977年6月17日发生于汤加地区(西南太平洋)的深震(震级5.6级,震源深度达684km)等);由此探索深源地震机制对于查明俯冲板块与地震空间成因关系和开展全球自然灾害地质研究具有十分重要的意义。因此本文对我国唯一的深震区——东北深源地震带的深源地震进行了初步性研究,研究结果显示:1)我国东北深震的发生与西北太平洋板块向欧亚大陆的俯冲直接相关,且其震源机制解显示的应力状态以下倾的压缩应力为主,这与上述的俯冲岩石圈当其进入下地幔时并没有变弱而是比其下伏的地幔转换带具有更大的粘度的结果相一致。证据主要有a,震源深度在SEE—NWW方向有依次加深的趋势,而在SSW—NNE方向的剖面上震源深度似乎没有明显变化,分布较SEE—NWW方向上的剖面是很均匀的;b,P轴的优势方位为北西西向,平均方位是277°,而P轴的平均仰角是31°,T轴的平均仰角为50.8°(与P轴方向接近垂直);也就是说东北深震震源机制解中的主压应力轴是与太平洋板块的下插方向相一致的,而主张应力轴与其垂直;c,从日本海沟到我国的东北地区震源深度逐渐增加,依次为浅源、中源和深源地震区,且中国东北深震几乎是从日本海沟沿直接倾斜下来;2)通过与东北深震区和西太平洋俯冲带的地球物理资料相对比,认为我国东北深源地震的发生很可能是由橄榄石的相变引起的,而没有发现任何超过660km的深源地震发生,或者与本区亚稳态橄榄石已经消耗殆尽有关,或者与林伍德石转换为钙钛矿和镁方铁矿从而形成稳定的后成结晶结构相关。最后值得一提的是本论文的第六章,从内容上来看它似乎与本博士论文的题目不太相关,但考虑到对较常用的高温高压Multianvil仪器操作和实验流程的学习也是我博士研究生期间主要目的之一,且单斜辉石中Ca-Eskola含量随温度、压力、化学成分的变化对理解许多超高压变质(UHPM)现象(例如在许多榴辉岩和石榴橄榄岩中的绿辉石和透辉石中显示有小的具有一定晶体学趋向的SiO2包裹体)有着重要的理论研究意义,所以在第六章中对我博士研究生期间重要的研究成果之一——“利用多压砧(Multianvil)高温高压仪器对单斜辉石中Ca-Eskola含量高温高压实验研究”进行了展示。本研究中我们利用加州大学河滨分校的Walker型多压砧高温高压仪器对具有绿辉石组成(Na2O=1.78wt.%,CaO=10.37wt.%,MgO=5.44wt.%,Al2O3=23.12wt.%,SiO2=59.29wt.%)的玻璃(glass)粉末样品在无水的、P分别为6、8、10、12GPa和T分别为900、1000、1100、1200℃状态下进行了一系列实验(共16个实验),并对实验后的样品进行了Ca-Eskola(简写为Ca-Esk)含量随温度、压力、化学成分变化的研究。研究结果显示:1)所有的反应产物中与单斜辉石共同存在的矿物有石榴石、蓝晶石和Si02(柯石英或斯石英),且单斜辉石都是非配比平衡的(nonstoichiometric),也就是说在6个氧的基础上总的阳离子数小于4;2)Ca-Esk含量对起始物质化学组成非常敏感,尤其是Si02的含量;3)等温条件下非配比平衡单斜辉石中Ca-Esk含量值在4-5和8GPa之间达到最大(38-42mol.%),随后随着压力的升高Ca-Esk含量在约9GPa-10GPa迅速降低直到斯石英变为稳定(在12GPa,1200℃时降低到约18mol.%),同时也伴随着辉石向石榴石的转换,直到15GPa所有的辉石全部转换为石榴石为止;4)等压条件下非配比平衡单斜辉石中Ca-Esk含量值随温度的变化较压力来说并不是很明显;且考虑到等温条件下单斜辉石中“过量Si02”从1.5到4-5GPa逐步升高,在5-8GPa达到最大,我们认为UHPM中Ca-Esk含量与从地球深部深达200-250km处(压力约为8GPa)折返上升来的辉石中SiO2的出溶相关,因为其在P≥9GPa后Ca-Esk含量迅速降低,这样系统中“过量SiO2”也将随着降低。只有UHPM岩石在“过饱和的Si”的状态下折返,其才能出溶SiO2(柯石英或石英),所以200-250km只是从本研究中推论出的岩石折返的最大深度。实验结果还说明辉石从地球深部上升过程是非常迅速的,因为发现在单斜辉石中出溶的SiO2是柯石英(如果上升过程很慢,其势必退变为石英)。

【Abstract】 The660km seismic discontinuity corresponds, at least at lower temperatures, to the transition of the spinel phase of (Mg,Fe)2Si04, ringwoodite, to (Mg,Fe)SiO3, perovskite (pv)+(Mg,Fe)O, magnesiowustite (mw). The rheological properties of material above and below this discontinuity as well as the chemical/structural nature of the boundary itself play an important role for the understanding of deeply subducted slabs, the termination of deep earthquakes, mantle convection, post-glacial rebound, and other geodynamic processes. Although the rheology of olivine, the dominant upper-mantle mineral, has been extensively studied, knowledge about the rheological properties of the material near the660km discontinuity, especially pv+mw is limited because of technical reasons. As a consequence, the deformation behavior of deep subduction slab at the660km seismic discontinuity and its role in the deeper levels of subduction zones and mantle remain poorly understood. To address this important problem, we have performed preliminary studies on a similar realistic analogue system (disproportionation of Co2TiO4spinel into CoTiO3ilmenite+CoO) with the material near the660km discontinuity under high temperature and high pressure, including an experimental investigation on the rheology of the material near the660km discontinuity, preliminary EBSD study on the deformation microstructure of the material near the660km discontinuity and their significances for the mechanism of deep-focus earthquakes in North-Eastern area of China.We have conducted experiments at strain-rates of10-1~10-5/s, pressures of2-4GPa, and temperatures of1273-1523K to determine and compare the flow strength of spinel with its dissociation product in detail through high pressure and temperature experiments in the5GPa piston-cylinder (Griggs) deformation apparatus in which the deformation can be well controlled, with accurate stress measurement with a realistic analogue system:disproportionation of Co2TiO4spinel into CoTiO3ilmenite+CoO. Our results show that (i) Such decomposition reactions yield typical symplectic microstructures--"wormy" intergrowths of daughter phases-in which the domain size viewed in2D can appear much smaller than the actual size of individual intergrowths of the two phases in3D.(ii) The symplectic high-pressure reaction product (CoTi03+CoO) is stronger than the Co2TiO4spinel parent about25%rather than much weaker as predicted, that is, symplectites are structurally strong, not structurally weak because each is a complexly intergrown2-phase bicrystal, not weak nanocrystalline aggregates. Application to Earth shows that subducting lithosphere entering the lower mantle is not weaker and can be expected to be significantly more viscous than the mantle transition zone as has been shown for the lower mantle in general by geophysical measurements. With our results, the down-dip compressive stress states of subducting slabs can be understood well, as observed in the slabs of western Pacific Ocean (including the north-eastern area of China). In addition, there is no indication of a shearing instability during incipient breakdown of the spinel phase under stress in this system. Thus, the most likely explanation of termination of deep earthquakes at the base of the mantle transition zone is that either metastable olivine (now having been observed in4subduction zones) has been exhausted or that, if carried into the lower mantle, reacts aseismically to form symplectites.In the past decade the electron backscatter diffraction (EBSD) technique has become into a available apparatus for the microstructure and crystallographic analysis of materials equipped on the scanning electron microscope. This method was widely used for measurement of texture and microstructure of minerals because of its convenience, high quality positionand, high orientation resolution, and so on. Recently, it have been applied to phase identification, studying deformation mechanisms, constraining dislocation slip systems, studying metamorphic processes, constraining metamorphic processes and numerical computations of petrophysical properties. Our preliminary results from electron backscatter diffraction (EBSD) study on the deformed or undeformed microstructure of symplectites of CoTiO3+CoO and CO2TiO4spinel phase show that:(a) From the scatter diagrams of orientation map of symplectites, it can been seen that all plane{111} of CoO//{0001} of CoTiO3,{110} of CoO//{10-10} of CoTiO3and{100} of CoO//{2-1-14} of CoTiO3, although in the second (right) symplectite show two groups orientation of CoO, suggesting that CoO and CoTiO3are intergrowths in3D and also follow a special direction.(b) During deformation of the symplectites, dynamic recrystallization and grain growth occurs, however, there is only slight diminution of grain size and no evidence for structural weakening of the symplectite due to’structural superplastic’ flow dominated by grain-boundary sliding.(c) There is no pronounce crystallographic preferred orientation (CPO) in deformed Co2TiO4and undeformed or deformed CoO of symplectites. All the misorientation angle distributions correspond to theoretically random distribution curves. The nearly random CPO may be attributed to the high symmetry of crystallographic structure (cubic) or inadequate deformation, and can not be used to imply active slip systems.(d) In contrast, a relative stronger crystallographic preferred orientation be developed in deformed CoTiO3of symplectites, the{110} plane tends to form a girdle in the plane subparallel to the lineation with a relatively weak point maximum subparallel foliation, and{100} plane nearly show the same pattern as{110}. The {001} plane poles normal to lineation and tend to parallel Z direction (σ1compression direction), all suggest dislocation glide on the plane of{001} maybe play a dominant role for the plasticity of CoTiO3.The results from the preliminary study of deep-focus earthquakes in Northeast area of China show that (i) The deep-focus earthquakes region of North-Eastern China is a part of Western Pacific Ocean subduction zones, and is the result of driving action from Western Pacific Ocean subducting slab. The down-dip compressive stress states is showed from the the focal mechanisms of the deep earthquakes in North-Eastern China. It is consensus with the greater viscosity of the lower mantle mentioned above,(ii) Most deep-focus earthquakes in North-Eastern China occur within metastable olivine wedge, so their occurrence should be related to phase transformation. The most likely explanation of termination of deep earthquakes at about660km is that either metastable olivine has been exhausted or that, ringwoodite breakdown yields such aseismically symplectites.It’s worth noting that there seems no relation between the chapter6and the title of this paper, however, learning how to operate MA apparatus and how to process the sample for MA is one of main prospective objects during my time as a graduated student, besides, it is important to understand the composition of starting materials, temperature and pressure implications of Ca-Eskola component in clinopyroxene, such as, the explanation of omphacite and diopside from many eclogites and garnet peridotites display tiny, crystallographically oriented, rod-shaped or lamellar inclusions of quartz. So as one of important results during my time as a graduated student, the " Ca-Eskola component in clinopyroxene:experimental studies at high pressures and high temperatures in multianvil apparatus" is shown at the chapter6.A series of anhydrous experiments using powdered glass of omphacite composition (wt.%):Na2O=1.78, CaO=10.37, MgO=5.44, Al2O3=23.12, SiO2=59.29were conducted in a Walker-style multianvil apparatus in the P-T range6-12GPa and900-1200℃to study activity of the Ca-Eskola (Ca-Esk) component. All clinopyroxenes synthesized in equilibrium with garnet, kyanite and SiO2(coesite or stishovite) are nonstoichiometric, with the sum of cations<4, calculated for6oxygens. We systematically have measured the values of the Ca-Esk component of clinopyroxenes as a function of pressure and temperature. The results yield that the highest value of the Ca-Esk component (38-32mol.%) is reached at6GPa, then it slightly decreases up to36-30mol.%at8GPa, and significantly decreases at10-12GPa from26to16mol.%when stishovite becomes stable in the assemblage and the pyroxene progressively dissolves into garnet. Comparison of our data with those available in the literature (which all together cover a wider range of P and T within pyroxenes stability field) show that the Ca-Esk activity strongly depends on bulk chemistry at all ranges of pressure and temperature. However, the sensitivity of the Ca-Esk component to pressure and temperature is not a simple negative or positive correlation, it is characterized by complex relationships which are strongly affected by bulk composition. In general, integrated analyses of published experiments suggest that the Ca-Esk component rises rapidly to its highest values (~36-38mol.%) from1to4-6GPa, then it slightly decreases between8to10GPa followed by remarkable decrease at highest pressure ranges of12to15GPa culminating in replacement of pyroxene by garnet. We conclude that the SiO2exsolution lamellae observed in clinopyroxenes from many ultrahigh pressure terranes can be explained by decompressions of the "Si-rich" nonstoichiometric clinopyroxene, however the depth from which such rocks are uplifted should not exceed200-250km, corresponding to pressure~8-10GPa.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络