3.1   样品采集

在项目执行过程中，沿甘肃兰州—合作一线的开展了详细地野外地质调查，采集了有代表性的岩石样品（图 2, 图 3），其中锆石U-Pb年代学样品7件，包括花岗岩、钾长花岗岩、花岗闪长岩、闪长岩、英安岩和安山岩。

图 2.  安山岩（ZHZ09）侵入到花岗闪长岩（ZHZ08）

下载: 全尺寸图片 幻灯片

图 3.  花岗岩0430（A）和逆冲断层（B）的野外照片

下载: 全尺寸图片 幻灯片

3.2   样品测试方法及精度

3.2.1   锆石LA-MC-ICP MS分析方法和锆石特征

选择代表性样品，粉碎至60目，通过淘洗—磁选—电磁选—手工挑取等一系列方法分离锆石。在双目镜下，选择干净透明的单矿物颗粒，逐粒整齐排放于双面胶带上，按上PVC环，将环氧树脂和固化剂混合均匀后注入PVC环，然后放入烘箱，60摄氏度恒温48 h。取出样品靶，利用砂纸磨至样品中心部位并用抛光布抛光。靶制备完成后对矿物进行反射光和透射光显微照相，以查明样品表面的裂隙和内部的包裹体。之后，在扫描电镜实验室对锆石进行阴极发光图像采集，检查矿物颗粒的内部结构和成分环带，以确定合适的位置进行分析测试。在分析测试前，用酒精超声清洗样品靶，除去可能的表面污染。采用离子探针方法分析的样品靶需在分析前进行表面镀金。阴极发光成像观察在北京离子探针中心进行，在中国地质科学院地质研究所，利用扫描电镜进行了BSE图像和锆石内部包裹体的成分测试。通过阴极发光和BSE图像来查明锆石内部生长层的分布和结构，选取测试点。锆石U-Pb同位素定年测试在中国地质科学院矿产资源研究所MC-ICP-MS实验室完成，锆石定年分析所用仪器为Finnigan Neptune型MC-ICPMS及与之配套的Newwave UP 213激光剥蚀系统。激光剥蚀所用斑束直径为25μm，频率为10 Hz，能量密度约为2.5 J/cm²，以He为载气。信号较小的²⁰⁷Pb，²⁰⁶Pb，²⁰⁴Pb(+²⁰⁴Hg), ²⁰²Hg用离子计数器(multi-ion-counters)接收，²⁰⁸Pb，²³²Th，²³⁸U信号用法拉第杯接收，实现了所有目标同位素信号的同时接收并且不同质量数的峰基本上都是平坦的，进而可以获得高精度的数据，均匀锆石颗粒²⁰⁷Pb/²⁰⁶Pb，²⁰⁶Pb/²³⁸U，²⁰⁷Pb/²³⁵U的测试精度(2)均为2%左右，对锆石标准的定年精度和准确度在1%(2)左右。LA-MC-ICP-MS激光剥蚀采样采用单点剥蚀的方式，数据分析前用锆石GJ-1进行调试仪器，使之达到最优状态, 锆石U-Pb定年以锆石GJ-1为外标，U、Th含量以锆石M127(U:923×10^-6; Th:439×10^-6; Th/U: 0.475. Nasdala et al, 2008)为外标进行校正。测试过程中在每测定5~7个样品前后重复测定两个锆石GJ1对样品进行校正，并测量一个锆石Plesovice，观察仪器的状态以保证测试的精确度。数据处理采用ICPMSDataCal程序(Liu et al. 2010)，测量过程中绝大多数分析点²⁰⁶Pb/²⁰⁴Pb > 1000, 未进行普通铅校正，²⁰⁴Pb由离子计数器检测，²⁰⁴Pb含量异常高的分析点可能受包体等普通Pb的影响，对²⁰⁴Pb含量异常高的分析点在计算时剔除，锆石年龄谐和图用Isoplot 3.0程序获得。完成了7件锆石U/Pb地质年代学测试。

样品0412为英安岩，采自于西秦岭造山带内，锆石呈自形、长柱状，棱角清晰，粒度在100~250 μm，长宽比一般为2:1，个别可达3:1。锆石阴极发光和背散射图像都显示锆石没有核部，较干净，基本上不含包裹体，具明显的韵律环带结构（图 4a），核部环带密度小，而边部环带密度大，为岩浆锆石。

图 4.  岩石样品代表性锆石阴极发光照片

下载: 全尺寸图片 幻灯片

ZHZ08为花岗闪长岩，采自于西秦岭造山带内。锆石呈自形、柱状，棱角清晰，粒度在100~250 μm，长宽比一般为2:1~3:1。锆石阴极发光和背散射图像都显示锆石没有核部，较干净，个别含包裹体，具弱化的韵律环带结构（图 4b），为岩浆锆石，但可能受到后期热事件的影响。

ZHZ09为安山岩，采自于西秦岭造山带内，锆石呈自形—半自形，大部分为长柱状，个别为浑圆状，粒度在50~200 μm，长宽比一般为2:1~3:1。锆石阴极发光和背散射图像都显示锆石没有核部，较干净，个别含包裹体，具弱化的韵律环带结构（图 4c），个别锆石边部由于含有较高的U和Th，显示暗色发光的韵律环带，但是与核部锆石为同一期，可能是由于同期事件中锆石生长时U、Th变化引起的。

0419为钾长花岗岩，采自于积石山。锆石呈自形—半自形，长柱状，粒度在100~200 μm，长宽比一般为2:1，个别达4:1。锆石阴极发光和背散射图像都显示锆石具有核-幔-边结构，白色模糊环带的核部，灰色弱环带的幔部和黑色弱发光的边部，个别锆石的幔部含有包裹体（图 4d）。

0429B为闪长岩，采自于临夏盆地的北侧的唐旺镇。锆石呈半自形—他形，柱状、浑圆状，粒度在50~150 μm，大部分长宽比为2:1，少量为1:1。锆石阴极发光和背散射图像都显示锆石具有核-幔-边结构，弱振荡环带的核部，灰色无环带的幔部和白色的窄边（图 4e）。

0429C为钾长花岗岩，采自于临夏盆地的北侧的唐旺镇。锆石呈自形，长柱状，粒度为100~150 μm，长宽比一般为2:1。锆石阴极发光和背散射图像都显示锆石具有核-边结构，白色含有包裹体的核部，明显的振荡环带的边部（图 4f）。

0430-2为花岗岩，采自于临夏市北3 km处。锆石呈自形，柱状，粒度在150~200 μm，长宽比为2:1。锆石阴极发光和背散射图像都显示锆石显示均一的明显振荡环带（图 4g）。

3.2.2   主量元素、微量元素分析方法

为确定岩石的地球化学特征，分析了岩石的全岩主量、微量元素和Rb-Sr、Sm-Nd同位素组成。主量及微量元素的测试在国土资源部国家地质实验测试中心进行。全岩主量元素采用X荧光光谱(XRF)玻璃熔片法进行分析。流程如下：首先将全岩粉末在105 ℃的烘箱中烘烤2 h，去除样品中的吸附水。将样品从烘箱中取出后迅速放入干燥器中冷却。待样品冷却至室温后，准确称取0.5 g (0.5000±0.0007g)样品放入已恒重的坩埚中，之后将盛有样品的坩埚放入马弗炉中加热至1000 ℃灼烧1.5 h。取出灼烧后的样品置于干燥器中，冷却至室温后称重，计算样品的烧失量。之后，准确称取混合试剂(成分为溶剂Li₂B₄O₇、助溶剂LiF、氧化剂NH₄NO₃) 5 g，与样品混合并研磨至均匀。将混合样品倒入铂金坩埚中，加入3滴溴化锂（脱模剂），在高频熔样机内1000 ℃下充分熔融后倒出，冷却形成玻璃熔片。最后，利用XRF(X荧光光谱仪3080E)对样品进行主量元素分析。在分析过程中，选用国家标准物质中心提供的岩石标样GSR.1 (花岗岩)、GSR.2（安山岩），GSR.3（玄武岩）作为标准参考物质。分析结果中，分析精度为5%。

全岩微量元素的分析采用混合酸溶法溶样。分析测试采等离子质谱仪(ICP-MSExcell)完成。分析流程如下：首先将样品放入烘箱内在105 ℃下烘烤2 h，除去吸附水。将样品取出后置于干燥器中冷却至室温。准确称量50.00 mg (49.00~51.00 mg)样品放入Teflon有盖溶样弹中，并加入1.5 mL高纯HNO₃、1.5 mL高纯HF和0.01 mL高纯HClO₄。将盛有样品的溶样弹置于140 ℃的电热板上开盖蒸干，以除去大部分SiO₂。蒸干后，再向溶样弹中加入1.5 mL高纯HNO₃、1.5 mL高纯HF。随后将溶样弹加盖并装入钢套密封, 放入烘箱中190 ℃恒温120 h。取出溶样弹，蒸干样品，然后加入3 mL高纯HNO₃再次蒸干，以去除残余的HF。之后，加入3 mL 1:1高纯HNO₃，放入钢套中置于烘箱中150 ℃恒温12 h，以保证完全提取样品。冷却后，将样品倒入100 mL PET瓶中，并加入1 g Rh内标，加水定容至100 g，待上机测试。微量元素和稀土元素（REE）含量大于10×10^-6的元素的测试精度为5%，而小于10×10^-6的元素精度为10%。个别在样品中含量低的元素，测试误差大于10%。

3.2.3   全岩Sr-Nd-Pb同位素分析方法

Rb-Sr和Sm-Nd同位素分析在中国地质科学院地质研究所同位素实验室进行。首先称取100~150 mg样品放入Teflon有盖溶样弹中，加入1.5 mL高纯HNO₃，1.5 mL高纯HF和0.01 mL高纯HClO₄，放在140 ℃的电热板上开盖加热蒸干。蒸干后加入1 mL高纯HNO₃，2 mL高纯HF，加盖装入钢套中，放入烘箱中190 ℃温度下加热120 h。取出样品后置于140 ℃电热板上蒸干。加入1 mL 6N HCl，再次蒸干并升温至200 ℃直至白烟冒净。加入1 mL 3N HCl，保持80 ℃温度下静置保温过夜，待化学分离。通过同位素稀释法，利用Finnigan MAT-262质谱仪测试Sr同位素组成及Rb、Sr、Sm和Nd的浓度。利用Nu Plasam HR MC-ICP-MS多接收等离子质谱仪(Nu Instruments)进行Nd同位素分析。Nd和Sr分析结果通过分别标准化到¹⁴⁶Nd/¹⁴²Nd = 0.7219和⁸⁶Sr/⁸⁸Sr = 0.1194进行质量分馏校正。在分析样品期间，Sr同位素测试标准为NBS987，测试值为0.710247±12 (2σ)。Nd同位素标准为JMC Nd，测试值为0.511127±12(2σ)。Sr和Nd同位素的测试精度分别为±0.000010 (n=18), 和±0.000011 (n=18)。

3.1   Sample collection

During the implementation of the project, a detailed field geological survey was conducted along the Gansu Lanzhou—Hezuo line, and representative rock samples were collected (Fig. 2 and Fig. 3), of which 7 pieces are zircon U−Pb geochronology samples, including granite, K−feldspar granite, granodiorite, diorite, quartz andesite, and andesite.

Figure 2.  Field photographs showing the andesite ZHZ09 intruding into granodiorite ZHZ08

下载: 全尺寸图片 幻灯片

Figure 3.  Field photographs showing the granite 0430 (A) and thrust fault (B)

下载: 全尺寸图片 幻灯片

3.2   Sample analytical procedures and accuracy

3.2.1   Zircon LA−MC−ICP−MS analysis method and zircon characteristics

A representative sample was selected, crushed into 60 mesh, and the zircon separated by a series of methods, namely washing, magnetic separation, electromagnetic separation, and manual separation. Under a binocular microscope, clean and transparent single− mineral particles were selected, and placed one by one on double-sided tape, and then a PVC ring pressed on. After mixing epoxy resin and its curing agent evenly, the mixture is poured into the PVC ring, and then placed into an oven, and kept at 60℃ for 48 hours. Taking out the sample target, it was sanded with a piece of sandpaper to the central portion and polished with a piece of polishing cloth. After the target preparation was completed, reflected and transmitted light photomicrography was conducted on the mineral to identify surface cracks and internal inclusions of the sample. Thereafter, cathodoluminescence (CL) image acquisition was conducted for the zircon in the scanning electron microscope laboratory to examine the internal structure and composition zoning of mineral particles, to determine the appropriate position for analysis and testing. Prior to the analysis and testing, the sample target was cleaned ultrasonically with alcohol, to remove possible surface contamination. The sample target was plated with gold on the surface and then was analyzed using the ion probe method. The CL imaging observation was carried out at the Beijing SHRIMP Center, and the backscattered (BSE) imaging and the test of components of inclusions in zircon were conducted using scanning electron microscopy at the Institute of Geology, Chinese Academy of Geological Sciences, Beijing. The distribution and structure of the growth layer in zircon were determined by CL and BSE images, to select analysis points. The zircon U−Pb analyses were conducted in the MC-ICP-MS laboratory of the Institute of Mineral Resources, Chinese Academy of Geological Sciences. The equipment used for the zircon dating analysis is the Finnigan Neptune MC -ICP -MS and its accompanying Newwave UP 213 Laser Ablation system. In laser ablation, the diameter of the beam spot used is 25 μm, the frequency is 10 Hz, the energy density is about 2.5 J/cm², and the carrier gas is He. Multi-ion-counters were used to receive signals for ²⁰⁷Pb, ²⁰⁶Pb, ²⁰⁴Pb(+²⁰⁴Hg) and ²⁰²Hg with weak signals, and the signals of ²⁰⁸Pb, ²³²Th, and ²³⁸U were received using Faraday cups; therefore, all target isotope signals were simultaneously received and all peaks with different mass numbers were substantially planar, to obtain data with high accuracy. The test accuracies (2s) of ²⁰⁷Pb/²⁰⁶Pb, ²⁰⁶Pb/²³⁸U and ²⁰⁷Pb/²³⁵U of uniform zircon particles are all about 2%, and the standard dating accuracy for zircon is about 1% (2s) accordingly. The single point ablation method was used for LA−MC−ICP− MS laser ablation sampling, and the zircon GJ−1 was used for adjustment of the equipment so that it achieved its optimal state before data analysis. The zircon GJ−1 was taken as an external standard for zircon U−Pb dating, and the zircon M127 (U: 923 ppm; Th: 439 ppm; Th/U: 0.475. Nasdala et al., 2008) was taken as an external standard for calibration for the U, Th contents. During the test, two zircons GJ1 were measured repeatedly for calibration of samples before and after every 5–7 samples were measured. One zircon Plesovice was measured, and the equipment state should be observed to ensure the test accuracy. The ICPMSDataCal program (Liu et al. 2010) was used for data processing. ²⁰⁶Pb/²⁰⁴Pb > 1000 for vast majority of analysis points, and the common Pb correction was not performed during the test. ²⁰⁴Pb was detected by an ion counter. The analysis points with abnormally high content of ²⁰⁴Pb might be affected by the common Pb in inclusions, etc., so the analysis points with abnormally high content of ²⁰⁴Pb were excluded from the calculation. The zircon age concordia diagram was obtained with Isoplot 3.0 program. The U−Pb geochronology analysis was completed for 7 zircon samples.

Sample 0412 is quartz andesite, collected from the western Qinling orogenic belt. The zircons are euhedral and long columnar with clear edges and corners, with grain sizes being between 100–250 mm and with length−width ratios being 2:1 generally and reaching 3:1 occasionally. Both CL and BSE images of the zircons show that the zircons have no evident cores, and are relatively clean and substantially free of inclusion, with distinct rhythmic zonal structure (Fig. 4a). The core zone density is low, but the edge zone density is higher, so the zircons are magmatic zircons.

Figure 4.  CL images of representative zircons for magmatic rocks

下载: 全尺寸图片 幻灯片

Sample ZHZ08 is a granodiorite, collected from the western Qinling orogenic belt. The zircons are euhedral and columnar with clear edges and corners, with grain sizes being between 100–250 mm and with length-width ratios being generally 2:1–3:1. Both CL and BSE images of the zircons show that they have no evident cores and are relatively clean, and several of them have inclusions, with weakened rhythmic zoning structure (Fig. 4). The zircons are magmatic zircons, but they might have been affected by late thermal events.

Sample ZHZ09 is an andesite, collected from the western Qinling orogenic belt. The zircons are euhedral–subhedral, and most of them are long cylindrical and several are perfectly round. The grain sizes are between 50–200 mm, and the length-width ratios are generally 2:1–3:1. Both CL and BSE images of the zircons show that they have no nucleus and are relatively clean, and several of them have inclusions, with weakened rhythmic zoning structure (Fig. 4c). Some zircons have edges containing high U and Th and showing rhythmic zoning structure with dark luminescence, which are, however, in the same period with the core zircons. This may have been caused by the changes in the contents of U and Th in the growth of the zircon during the contemporaneous event.

Sample 0419 is a k-feldspar granite, collected from Jishishan Hill. The zircons are euhedral–subhedral, and long columnar, with grain sizes being between 100–200 mm and with length-width ratios being generally 2:1 and occasionally reaching 4:1. Both CL and BSE images of the zircons show that they have a core−mantle−edge structure, including a white fuzzy zoning core, a gray weak zoning mantle and a black weak luminescence edge. Several zircons contain inclusions in their mantles (Fig. 4d).

Sample 0429B is a diorite, collected from Tangwang Town on the north of the Linxia basin. The zircons are subhedral–anhedral, and columnar, perfectly round, with grain sizes being between 50–150 mm, and with length-width ratios being mostly 2:1 and 1:1 rarely. Both CL and BSE images show that the zircons have a core−mantle−edge structure, including a weakly oscillating zoning core, a gray unzoned mantle and a white narrow edge (Fig. 4e).

Sample 0429C is a k-feldspar granite, also collected from Tangwang Town. The zircons are euhedral, and long columnar, with grain sizes being between 100–150 mm and with length-width ratios being generally 2:1. Both CL and BSE images of the zircons show that they have a core−edge structure, including a white core containing inclusions, and an edge with distinct oscillating zone (Fig. 4f).

Sample 0430-2 is a granite, collected from a place about 3 km from the north of Linxia city. The zircons are euhedral and columnar, with grain sizes being between 150–200 mm, and with length–width ratios being 2:1. Both CL and BSE images show that the zircons have a homogeneous distinct oscillating zoning structure (Fig. 3g).

3.2.2   Analysis methods for major elements and trace elements

In order to determine the geochemical characteristics of the sample rocks, the compositions of the whole rock major elements, trace elements and Rb−Sr, Sm−Nd isotopes were analyzed. The testing of the major and trace elements was conducted at the National Geological Experiment and Testing Center of the Ministry of Land and Resources. The whole rock major elements were analyzed using the X−ray fluorescence spectroscopy (XRF) melting glass plate method. The procedure was as follows: frst, put the whole rock powder into an oven to bake at 105℃ for 2 hours so as to remove the adsorbed water from the sample. Remove the sample from the oven and quickly put it into a desiccator to cool. After the sample is cooled to room temperature, accurately weigh 0.5 g (0.5000 ± 0.0007g) of the sample into a crucible dried to a constant weight, and then put the crucible with the sample into a muffle furnace and heat to 1000℃ to burn for 1.5 hours. Take out the burned sample and put it in a desiccator to cool to room temperature, and then weigh it and calculate the loss on ignition of the sample. After that, accurately weigh 5 g of mixture reagent (composed of solvent Li₂B₄O₇, cosolvent LiF, oxidant NH₄NO₃), mix with the sample, and grind till even. Pour the mixed sample into a platinum crucible, and add 3 drops of lithium bromide (release agent); pour it out after being suffciently melted at 1000℃ in the high frequency melting machine, and cool down to form a glass plate. Finally, the major element analysis of the sample was carried out by using XRF (X−ray fluorescence spectrometer 3080E). In the process of analysis, the rock standard samples GSR.1 (granite), GSR.2 (andesite) and GSR.3 (basalt) provided by the National Standard Material Center were selected as standard reference materials. In the analysis results, the analysis accuracy is 5%.

For analysis of the whole rock trace elements, the mixed acid dissolution method was used for sample dissolving. The analytical test was performed with an inductively coupled plasma mass spectrometer (ICP−MS−Excell). The analysis flow was as follows: first, put the sample into an oven to bake at 105℃ for 2 hours so as to remove the adsorbed water. Take out the sample and place it in a desiccator to cool to room temperature. Weigh accurately 50.00 mg (49.00–51.00mg) of sample and put into a Teflon sample-dissolving vessel with cover, and add 1.5 mL of highly pure HNO₃, 1.5 mL of highly pure HF and 0.01 mL of highly pure HClO₄. Put the sample-dissolving vessel with cover opened on an electric heating plate at 140℃ to evaporate to dryness, to remove most of the SiO₂. After drying, add 1.5 mL of highly pure HNO₃ and 1.5 mL of highly pure HF again into sampledissolving vessel. Then cover the sample-dissolving vessel, and load it into the steel jacket to seal; put into the oven at 190℃, and keep the constant temperature for 120 hours. Take out the sample-dissolving vessel, evaporate the sample to dryness, and then add 3 mL of highly pure HNO₃. Evaporate to dryness again to remove the residual HF. After that, add 3 mL of highly pure HNO₃ (1:1), load into steel jacket, and put into the oven at 150℃ and keep the constant temperature for 12 hours to ensure complete extraction of the sample. After cooling, pour the sample into a 100 mL PET bottle, add 1 g of Rh internal standard, and add water into the bottle to make 100 g, ready for test on the machine. The testing accuracy is 5% for trace elements and rare earth elements (REEs) with content greater than 10×10^-6, and 10% for trace elements and REEs with content less than 10×10^-6. The test error is greater than 10% for several elements with low content in the sample.

3.2.3   Analysis methods for whole rock Sr−Nd−Pb isotopes

The analysis of Rb−Sr and Sm−Nd isotopes was carried out at the isotope laboratory of the Institute of Geology, Chinese Academy of Geological Sciences. First, weigh 100–150 mg of sample into a Teflon sample-dissolving vessel with cover, add 1.5 mL of highly pure HNO₃, 1.5 mL of highly pure HF and 0.01 mL highly pure HClO₄, and put the vessel with cover opened on an electric heating plate at 140℃ to evaporate to dryness. After drying, add 1 mL of highly pure HNO₃ and 2 mL of highly pure HF, load into steel jacket, and put into the oven at 190℃ and keep the constant temperature for 120 hours. Take out the sample and put on the electric heating plate at 140℃ to evaporate to dryness. Add 1 mL of 6N HCl, evaporate to dryness again, and heat to 200℃ until white smoke is fully released. Add 1 mL of 3N HCl, and keep at 80℃ overnight, waiting for chemical separation. With the isotope dilution method, the composition of Sr isotopes and the concentrations of Rb, Sr, Sm and Nd were measured using Finnigan MAT−262 mass spectrometer. Nd isotope analysis was conducted using Nu Plasam HR MC−ICP−MS Multi-collector plasma mass spectrometer (Nu Instruments). The analysis results of Nd and Sr were normalized to ¹⁴⁶Nd/¹⁴²Nd = 0.7219 and ⁸⁶Sr/⁸⁸Sr = 0.1194, respectively, for mass fractionation correction. During the analysis of the samples, the Sr isotope test standard was NBS987 and the test value was 0.710247±12(2σ). The Nd isotope test standard was JMC Nd, and the test value was 0.511127±12(2σ). The test accuracies of Sr and Nd isotopes are 0.000010 (n = 18) and 0.000011 (n = 18), respectively.