激光剥蚀-扇形磁场电感耦合等离子体质谱法同时测定锆石U-Pb年龄和微量元素含量
2024-03-19赵令浩孙冬阳胡明月袁继海范晨子詹秀春
赵令浩,孙冬阳,胡明月,袁继海,范晨子,詹秀春
(中国地质调查局元素微区与形态分析重点实验室,国家地质实验测试中心,北京 100037)
锆石、榍石、磷灰石、独居石等副矿物在地质样品中广泛存在,这些矿物普遍铀含量较高,因此成为U-Pb定年的目标矿物[1-5]。同时相对于普通造岩矿物,副矿物是各种关键微量元素如Zr、Hf、P、Ti、Nb、Ta、稀土元素(REEs)、U、Th等的主要载体,通过副矿物微量元素特征可以为划分年龄期次、解译年龄意义提供参考。例如,利用锆石Th/U比值及REEs配分模式特征区分岩浆及变质成因锆石[6-8];根据独居石中Y含量区分其与石榴子石或磷钇矿共生[9];根据碎屑磷灰石中微量元素含量区分磷灰石成因及来源,进而实现磷灰石U-Pb年龄溯源统计[10-11];另一方面副矿物微量元素含量可以提供矿物形成温度、氧逸度等信息,例如锆石Ti温度计[12]、榍石Zr温度计[13]、Ce氧逸度等[14]。变质或部分熔融过程中,副矿物的参与可能对岩石与流体系统的关键微量元素或同位素体系产生明显的影响[15-19],结合现代地质体中副矿物U-Pb年龄、关键微量元素特征及同位素特征构建变质矿物P-T-t轨迹,反演矿物生长过程,是“岩石年代学(Petrochronology)”研究中的重要组成部分[20]。因此,采用微区分析技术在有限的分析空间内获得更多的元素或同位素信息至关重要。
LA-ICP-MS技术因其高效、准确的特征成为副矿物U-Pb定年的主要技术方法之一,近年来随着含普通铅副矿物U-Pb定年数据处理方法不断完善和仪器测试能力的提高,U-Pb定年目标矿物领域不断拓展,尤其低/超低U矿物U-Pb定年方法的建立为解决传统疑难地质问题提供有力的技术支撑[21-25]。
根据质谱接收系统差异,激光剥蚀与多接收质谱联用构成LA-MC-ICP-MS,该仪器具有高灵敏度特征,可以同时采集数种元素或同位素信号,用于同位素比值的准确测定,但无法进行微量元素定量分析。为实现原位U-Pb同位素定年和微量元素含量同时检测,Kylander-Clark等(2013)[26]提出采用多接收质谱和单接收质谱串联(LASS)设计,分别用于U-Pb定年和微量元素定量分析。
激光剥蚀与单接收质谱连接构成LA-ICP-MS,尽管只有一个检测器,但可实现大范围质量数元素/同位素快速顺序检测,目前多用于微区微量元素定量分析和副矿物U-Pb同位素定年。单接收质谱根据质量分析器差异又可分为四极杆质谱(Q-ICPMS)和扇形磁场质谱(SF-ICP-MS),Hattendorf等(2003)[27]详细对比了两种质谱结构和性能的差异。总体上,Q-ICP-MS顺序扫描速度较快(2~50Hz),但灵敏度相对较低,广泛应用于锆石、榍石、磷灰石等相对高U、Pb矿物U-Pb年龄与关键微量元素含量同时分析。SF-ICP-MS具有高灵敏度、高分辨率特征,其灵敏度可达1000~10000cps/(μg/g),根据副矿物中的U-Pb含量,扇形磁场质谱更利于高空间分辨率和超低U含量样品的U-Pb定年分析。例如Kooijman(2012)[28]采用LA-SF-ICP-MS实现了12μm锆石U-Pb定年;Wu等(2020)[29]通过改造LA-SF-ICP-MS进样系统,大幅度提升了质谱仪的灵敏度,并且实现了高空间分辨率(5~16μm)锆石U-Pb定年。另外,SF-ICP-MS与激光系统联用实现了低/超低U矿物U-Pb定年,例如对U含量在ng/g级的碳酸盐矿物进行准确的U-Pb定年等[23,30]。但是受限于磁场定位时间较长,扫描速度较慢,根据选择分析元素质量范围,典型扫描速度1~5Hz,远低于Q-ICP-MS。尽管Latkoczy等(2002)[31]采用LA-SF-ICP-MS准确测定了样品中多元素含量,并提出通过优化设置可以使分析不受质谱扫描速度的影响,但相对于元素定量分析,尤其是在相对较少数据统计量的情况下,同位素比值的测定对于仪器准确性和稳定性提出更高的要求。为保证年龄结果的准确性与稳定性,目前报道的采用高分辨质谱(SF-ICPMS)测定副矿物U-Pb年龄方法中一般仅测试与U-Pb年龄相关的同位素(202Hg~238U),无法同时进行关键微量元素定量分析。在多期次生长副矿物研究中,U-Pb年龄与微量元素含量分别测试一方面可能会受到矿物生长空间影响,不便开展多次取样;另一方面非原位测试可能会导致测定年龄与微量元素所反映的温度、压力等地质环境信息耦合困难。
基于锆石U-Pb定年与微量元素含量同时测定在“岩石年代学”领域的重要性,本文采用激光剥蚀扇形磁场等离子体质谱(LA-SF-ICP-MS),以25μm激光斑束对7种常见的锆石U-Pb标准样品,包括91500[32]、GJ-1[33]、Tanz[34]、SA01[35]、Temora1[36]、Plešovice[37]和Qinghu[38]进行U-Pb年龄和Ti、Hf、REEs等关键元素含量的同时定量分析,探讨LA-SF-ICP-MS同时进行U-Pb定年和微量元素定量分析方法的可行性及其对年龄结果的影响。
1 实验部分
1.1 锆石U-Pb样品及处理
实验中采用的锆石U-Pb样品按其年龄由高到低包括:91500[32]、GJ-1[33]、Tanz[34]、SA01[35]、Temora1[36]、Plešovice[37]和Qinghu[38],年龄范围为1064~159Ma。这些矿物样品具有稳定的ID-TIMS U-Pb年龄,在世界范围内被广泛用作U-Pb定年标准样品。本文通过分析这些样品以验证所建立测试方法的准确性。
91500锆石是应用最为广泛的锆石U-Pb定年标准样品。该样品的ID-TIMS206Pb/238U年龄为1062Ma[32]。GJ-1锆石是澳大利亚MacQuarie大学大陆地球化学与成矿作用研究中心实验室的U-Pb测定标准锆石[33]。该锆石的TIMS年龄结果不谐和,206Pb/238U和207Pb/235U年龄分别为599.8±1.7Ma和601.6±1.3Ma,但是LA-ICP-MS分析结果谐和,目前该样品LA-ICP-MS分析中采用参考年龄为~603Ma。Tanz锆石是中国地质大学(武汉)团队开发的锆石U-Pb定年和Zr-O同位素组成标准样品。该样品的ID-TIMS年龄为566.16±0.77Ma,SIMS和LA-ICP-MS分析结果都介于564~569Ma,表明该样品具有良好的U-Pb同位素均一性[34]。SA01锆石是中国科学院地质与地球物理研究所近年来开发的锆石U-Pb定年、Hf-O同位素组成微区测试标准物质。该样品的ID-TIMS年龄为535.1±0.3Ma[35]。Temora1锆石产自澳大利亚Lachlan造山带镁铁质岩,目前常用作SHRIMP锆石U-Pb定年标准样品。该样品的ID-TIMS年龄为416.75±0.24Ma[36]。Plešovice锆石产自捷克富钾麻粒岩,该样品ID-TIMS测定的206Pb/238U年龄为337.13±0.37Ma[37]。Qinghu锆石是中国科学院地质与地球物理研究所离子探针实验室标准锆石。该样品的TIMS谐和年龄为159.38±0.12Ma[38]。
将所有锆石样品粘在PVC模具底部,然后向模具中注入环氧树脂和固化剂,制备成直径约24mm锆石靶,并对样品靶表面进行打磨、抛光,直至样品露出光洁表面。采用光学显微镜结合扫描电镜拍摄样品在透射光和放射光下照片及阴极发光(CL)图片,观察样品内部结构特征,避免样品裂隙及包裹体等对测试结果造成影响。在激光剥蚀分析前,利用去离子水及无水乙醇擦拭锆石表面,并采用高压N2流吹扫样品,去除样品表面的普通Pb污染。
1.2 仪器设置
锆石U-Pb定年和微量元素含量分析在中国地质调查局元素微区与形态分析重点实验室完成,采用ESL NWR 193UCArF准分子激光器及ELEMENTⅡ扇形磁场高分辨电感耦合等离子体质谱仪(SFICP-MS,美国ThermoFisher Scientific公司)。
本实验中激光剥蚀采用25μm激光斑束,频率8Hz。以He气作为吹扫气体提高剥蚀气溶胶传输效率[39],并通入1mL/min氮气以提高仪器灵敏度[40]。气路上采用信号匀化装置,增大气溶胶扩散空间,可有效地提高样品剥蚀信号稳定性。
SF-ICP-MS分析采用低分辨模式(M/△M=300)。实验前采用25μm激光线扫描NIST612进行仪器调谐,使La和Th信号>1.5×105cps,信号稳定性(RSD)为1%~2%,同时监测ThO+/Th+控制氧化物产率<0.2%。选择分析锆石U-Pb同位素和微量元素,包括29Si、49Ti、89Y、91Zr、139La、140Ce、141Pr、146Nd、147Sm、151Eu、157Gd、159Tb、163Dy、165Ho、166Er、169Tm、172Yb、175Lu、178Hf、206Pb、207Pb、208Pb、232Th和238U,该质量数范围内质谱磁场定位4次,设置首次磁场定位时间为0.1s,其余磁场定位时间0.05s。206Pb、208Pb、232Th和238U测试时间10ms,207Pb测试时间20ms,其余各元素测试时间5ms,每次扫描总时间0.87s,有效分析时间占71%。多数锆石中不含或含极低普通铅,且采用单接收质谱难以准确测定204Pb用于普通铅校正,为提高有效分析时间的比例,本方法中未检测202Hg和204Pb。仪器工作条件见表1。
1.3 数据处理
LA-ICP-MS测试前采用大激光斑束对样品表面进行预剥蚀,去除样品表面可能存在的Pb同位素污染。样品分析采用点剥蚀模式,点分析时间90s,包括仪器背景信号采集时间20s,激光剥蚀信号采集时间40s,以及吹扫时间30s。
采用91500锆石和NIST610分别作为U-Pb同位素比值和微量元素定量分析的标准物质。每分析10个未知样品点插入分析一组标准样品(2点91500锆石和1点NIST610)以校正分馏效应。
锆石U-Pb年龄和微量元素含量数据处理采用GLITTER 4.0软件完成[41],选择标准样品传递误差1%(Std Uncertainty),所有样品与91500标准锆石截取相同信号区间。锆石年龄谐和图和加权平均图绘制采用Isoplot[42]。锆石微量元素含量计算分别以Si(SiO2含量32.8%)和Zr(ZrO2含量67.2%)作为内标。本文所有年龄值和同位素比值误差均为2σ。
2 结果与讨论
扇形磁场质谱(SF-ICP-MS)具有低背景、高灵敏度特征,在锆石U-Pb定年过程中可以有效地提高空间分辨率和分析的准确度。因此,LA-SF-ICP-MS在低U-Pb含量年轻锆石及其他的低U矿物如石榴子石、方解石等矿物定年中有独特的优势。但由于SF-ICP-MS扫描多元素过程中磁场定位时间的影响,当测量元素质量数大范围变化时,其扫描速度明显降低,因此采用LA-SF-ICP-MS测定锆石年龄时一般仅检测202Hg、204Pb、206Pb、207Pb、208Pb、232Th和238U等与U-Pb定年相关的7个同位素。本文中采用的方法在检测U-Pb相关同位素的基础上,同时检测锆石中的关键主、微量元素含量,包括Si、Ti、REEs、Hf等(表1),并对采集与未采集微量元素两种方法定年结果进行对比,测试结果如图1、图2、表2和表3所示。
图1 LA-SF-ICP-MS锆石U-Pb定年结果(激光斑束25μm)Fig. 1 Zircon U-Pb dating results by LA-SF-ICP-MS; Red data measured only U-Pb isotopes, while blue data determined U-Pb and key trace elements simultaneously (laser spot size 25μm).
表3 LA-SF-ICP-MS 锆石多元素同时分析微量元素定量结果Table 3 Trace element results measured by LA-SF-ICP-MS, determining U-Pb and key trace elements simultaneously.
2.1 LA-SF-ICP-MS锆石U-Pb定年和多元素含量同时分析方法定年结果
91500锆石:测试样品30点,获得单点206Pb/238U和207Pb/235U年龄分别为1045.7~1080.5Ma和1044.0~1093.0Ma,单点误差分别为20.3~22.7Ma(1.9%~2.1%)和14.1~21.7Ma(1.7%~2.0%)。在谐和图上所有样品点呈现谐和特征,获得U-Pb谐年龄1063.3±2.4Ma(图1a),206Pb/238U加权平均年龄为1062.8±7.3Ma(图1b),MSWD=0.24,与推荐值在误差范围内一致。
GJ-1锆石:测试样品37点,获得单点206Pb/238U和207Pb/235U年龄分别为599.1~610.4Ma和563.4~618.3Ma,单点误差分别为11.4~12.4Ma(~1.9%)和9.8~12.7Ma(1.6%~2.1%)。该样品整体谐和年龄为603.8±1.6Ma(图1c),206Pb/238U加权平均年龄为604.0±3.8Ma(图1d),与推荐年龄在误差范围内一致。
Tanz锆石:测试样品24点,获得单点206Pb/238U和207Pb/235U年龄分别为550.3~569.7Ma和542.5~577.1Ma,单点误差分别为9.7~11.9Ma(1.7%~2.1%)和8.1~11.5Ma(1.5%~2.0%)。该样品整体谐和年龄为559.5±1.7Ma(图1e),206Pb/238U加权平均年龄为562.7±4.2Ma(图1f),与推荐年龄在误差范围内一致。
SA01锆石:测试样品22点,获得单点206Pb/238U和207Pb/235U年龄分别为526.0~541.6Ma和518.8~547.8Ma,单点误差分别为10.4~12.1Ma(1.9%~2.3%)和9.5~17.2Ma(1.7%~3.3%)。该样品整体谐和年龄为533.8±2.0Ma(图1g),206Pb/238U加权平均年龄为534.5±4.7Ma(图1h),与推荐年龄在误差范围内一致。
Temora1锆石:测试样品26点,获得单点206Pb/238U和207Pb/235U年龄分别为405.8~426.6Ma和404.2~456.4Ma,单点误差分别为8.1~10.0Ma(1.9%~2.3%)和7.0~16.5Ma(1.6%~3.8%)。该样品整体谐和年龄为417.5±1.5Ma(图2a),206Pb/238U加权平均年龄为416.7±3.4Ma(图2b),与推荐年龄在误差范围内一致。
Plešovice锆石:测试样品30点,其中2个点分析结果可能受到蜕晶作用或包裹体影响,207Pb/235U明显偏高外,其余28点单点206Pb/238U和207Pb/235U年龄分别为326.4~339.6Ma和321.0~342.4Ma,单点误差分别为5.2~7.3Ma(1.5%~2.1%)和5.1~6.8Ma(1.4%~2.1%)。谐和图上除了受影响的2个点偏离谐和线,其余各点的谐和年龄为334.6±1.0Ma(图2c),206Pb/238U加权平均年龄为335.8±2.5Ma(图2d),MSWD=0.29,与推荐值在误差范围内一致。
Qinghu锆石:测试样品20点,获得单点206Pb/238U和207Pb/235U年龄分别为156.3~164.9Ma和154.3~176.4Ma,单点误差分别为2.9~3.2Ma(1.8%~1.9%)和3.1~4.1Ma(2.0%~2.5%)。该样品整体谐和年龄为160.1±0.6Ma(图2e),206Pb/238U加权平均年龄为160.5±1.3Ma(图2f),与推荐年龄在误差范围内一致。
2.2 LA-SF-ICP-MS仅测定锆石U-Pb年龄结果
为探讨LA-SF-ICP-MS锆石U-Pb年龄和微量元素含量同时检测对于年龄结果的影响,本文在相同实验条件下,仅检测U-Pb定年相关同位素,每个样品测试20点,U-Pb年龄结果见图1、图2和表2。
91500锆石:单点206Pb/238U和207Pb/235U年龄分别为1045.3~1076.8Ma和1047.7~1106.4Ma,单点误差分别为15.0~16.2Ma(~1.4%)和10.8~11.7Ma(~1.1%)。该样品整体谐和年龄为1065.9±2.4Ma(图1a),206Pb/238U加权平均年龄为1065.4±6.8Ma(图1b)。
GJ-1锆石:单点206Pb/238U和207Pb/235U年龄分别为598.2~603.9Ma和587.6~603.5Ma,单点误差分别为8.6~8.9Ma(~1.4%)和6.9~7.7Ma(1.2%~1.3%)。该样品整体谐和年龄为600.5±1.3Ma(图1c),206Pb/238U加权平均年龄为601.7±3.8Ma(图1d)。
Tanz锆石:单点206Pb/238U和207Pb/235U年龄分别为556.3~568.1Ma和554.6~569.4Ma,单点误差分别为8.3~8.6Ma(~1.5%)和6.9~7.3Ma(1.2%~1.3%)。该样品整体谐和年龄为563.2±1.3Ma(图1e),206Pb/238U加权平均年龄为564.6±3.7Ma(图1f)。
SA01锆石:单点206Pb/238U和207Pb/235U年龄分别为525.4~539.6Ma和527.8~537.1Ma,单点误差分别为8.2~9.2Ma(~1.5%)和8.3~8.8Ma(1.5%~1.7%)。该样品整体谐和年龄为532.9±1.5Ma(图1g),206Pb/238U加权平均年龄为533.2±3.8Ma(图1h)。
Temora1锆石:单点206Pb/238U和207Pb/235U年龄分别为413.8~421.9Ma和412.6~427.2Ma,单点误差分别为6.5~6.8Ma(~1.5%)和5.6~7.3Ma(1.3%~1.7%)。该样品整体谐和年龄为418.2±1.1Ma(图2a),206Pb/238U加权平均年龄为418.2±2.9Ma(图2b)。
Plešovice锆石:单点206Pb/238U和207Pb/235U年龄分别为335.0~341.8Ma和333.7~340.9Ma,单点误差分别为4.9~5.8Ma(1.3%~1.5%)和4.6~5.2Ma(1.5%~1.7%)。该样品整体谐和年龄为337.7±0.9Ma(图2c ),206Pb/238U加权平均年龄为337.8±2.3Ma(图2d)。
Qinghu锆石:单点206Pb/238U和207Pb/235U年龄分别为161.8~156.6Ma和165.5~156.2Ma,单点误差分别为2.6~3.1Ma(1.6%~2.0%)和2.7~3.2Ma(1.7%~2.0%)。该样品整体谐和年龄为159.4±0.5Ma(图2e),206Pb/238U加权平均年龄为159.1±1.3Ma(图2f)。
2.3 LA-SF-ICP-MS定年与多元素含量同时测试方法和仅U-Pb定年方法对比
受SF-ICP-MS磁场定位时间的影响,本文中仅测U-Pb相关同位素方法质谱单次扫描时间0.306s,在40s的激光剥蚀样品时间内,采集元素信号强度数据134组;而U-Pb年龄与微量元素含量同时分析方法中质谱单次扫描时间0.857s,在40s激光采样时间内获得有效信号强度数据47组。
图1和图2对两种方法定年结果的精密度和准确度进行了直观地比较,相较之下,仅检测U-Pb同位素方法定年结果数据点误差更小,且数据点更为集中。
LA-SF-ICP-MS仅检测U-Pb同位素方法定年结果206Pb/238U和207Pb/235U年龄误差分别为1.5%和1.3%,单点206Pb/238U和207Pb/235U年龄一致性较好,RSD值分别为0.2%~0.9%和0.5%~1.3%;相比之下,U-Pb定年和多元素含量同时检测方法获得206Pb/238U和207Pb/235U年龄误差略有增大,分别为1.9%和1.7%,单点年龄离散度增大,206Pb/238U和207Pb/235U年龄RSD值分别为0.4%~1.4%和1.2%~3.3%,其中207Pb/235U年龄RSD值增大明显。
为探讨两种方法测定年龄结果精密度变化可能原因,本文以91500锆石为例,统计了相同实验条件下,两种方法测定206Pb、207Pb信号强度及238U/206Pb信号比值的相对标准偏差(RSD)。其中,仅检测U-Pb同位素方法采集206Pb、207Pb信号和238U/206Pb信号比值的RSD分别为13%、13%和5.2%,而多元素含量同时检测方法获得206Pb、207Pb信号和238U/206Pb同位素信号比值的RSD分别为14%、19%和9.3%。多元素同时采集延长了质谱单次扫描时间,在一定程度上对于检测同位素的信号强度及比值稳定性造成影响,尤其是对于低含量同位素如207Pb的影响更为明显。两种测试方法同位素信号及比值变化特征与U-Pb定年结果变化特征相一致。综合考虑数据处理过程中年龄不确定度计算方法和影响因素[43-46],因此,多元素同时检测对于同位素信号强度稳定性的影响可能是造成最终定年结果误差变大的主要原因。
锆石LA-ICP-MS定年过程中,多种因素都会对测试结果的精密度和准确度造成影响,例如测试仪器状态、标准样品推荐值、数据处理方法及软件应用等。目前一般认为LA-ICP-MS锆石U-Pb定年结果精密度为1%~2%,相对于推荐年龄LA-ICP-MS测试年龄结果偏差(准确度)可达到1%[43-46]。本文实验结果表明,尽管多元素同时检测可造成单点锆石U-Pb年龄结果的变化范围增大,但数据结果精密度仍优于2%,并且多元素同时检测对于样品的谐和年龄和206Pb/238U加权平均年龄的准确性没有影响——仅检测U-Pb同位素方法获得各样品谐和年龄和206Pb/238U加权平均年龄相对TIMS年龄偏差均小于0.5%,多元素同时检测分析方法获得各样品谐和年龄和206Pb/238U加权平均年龄相对TIMS年龄偏差分别小于1.0%和0.7%,完全满足U-Pb同位素地质年代学测试要求,同时,高精度和高空间分辨率的定年方法为精细刻画复杂地质过程提供技术支持[46]。
2.4 LA-SF-ICP-MS锆石定年同时分析微量元素定量结果准确性分析
采用LA-SF-ICP-MS测定锆石U-Pb年龄的同时,对锆石中Si、Ti、Y、Zr、Hf、REEs、Pb、Th和U等关键主、微量元素进行定量分析。实验采用NIST610作为微量元素含量分析外标物质,数据处理选择Si(SiO2含量32.8%)或者Zr(ZrO2含量67.2%)作为内标元素,分析结果列于表3。
稀土元素检出限在11~73ng/g之间,大部分小于20ng/g,其中La和Pr检出限分别为20ng/g和13ng/g,这两个元素在多数锆石中极度亏损,较低的检出限有利于这两个元素的检测准确。Ti检出限最高达302ng/g,Pb、Th、U检出限最低,分别为15ng/g、4ng/g和2ng/g。本方法的检出限能够满足锆石中关键微量元素准确定量需求。
由于天然锆石中微量元素分布的不均匀性,因此,对锆石微量元素含量的测定结果难以反映分析方法的精确度和准确度。本实验中随测NIST612和KL2-G作为质量监控样品。10组监控样品结果表现良好的精密度(图3a),NIST612中除Ti元素RSD略高外(4.8%),其余元素RSD值小于3%;KL2-G中含量较低的Tm和Lu的RSD值~5%,其余元素RSD值均小于5%。检测结果平均值与标准样品推荐值相比较,NIST612中Ti的相对误差(8.5%)较高,其余元素的相对误差均小于5%,准确度较高;而KL2-G中各元素的相对误差多在5%~10%之间,U的相对误差最大达~17%。虽然KL2-G中各元素含量相对较低,但良好的测试精密度表明较高的相对误差更可能是由于NIST610与KL2-G之间的基体差异的影响。尽管目前锆石微量元素分析中常采用NIST610作为标准样品,但研究表明NIST系列标准样品与天然矿物基体差异明显[47],其微量元素含量明显高于锆石,尤其是LREEs含量,因此对于锆石微量元素含量准确测定有待于进一步研究。总体上,本文建立的方法对于NIST612和KL2-G可获得稳定、准确的分析结果。
内标元素的选择是影响LA-ICP-MS分析结果准确性的重要因素。尽管研究表明LA-ICP-MS测定锆石微量元素含量过程中,在NIST610为外标条件下,Zr相对于Si更适合作为内标元素[47]。本实验中分别采用Si和Zr作为内标元素进行定量分析,所有锆石样品定量结果显示采用Si作为内标元素计算结果总体上略高于Zr作为内标计算结果(图3b)。对于91500和SA01样品,采用Si内标各元素分析结果与文献[35]和[48]推荐值相对误差多小于10%,而采用Zr内标结果相对误差在10%~20%之间。另外,由于锆石与采用的标准样品中Zr含量差异悬殊,在使用高扩散空间匀化装置条件下,锆石分析后需要长时间吹扫以降低仪器中Zr背景值,且可能影响标准样品中Zr检测准确性。尽管检测Si会增大SF-ICP-MS磁场变化范围,在有效时间内减少扫描次数(约减少5组数据),但综合实验结果,对于U-Pb定年结果及误差没有明显影响,因此本实验方法中检测Si作为内标元素。
91500、SA01、GJ-1和Tanz颗粒内部微量元素均一性较好,检测结果除Ti、La误差略高外,其余元素误差均小于10%;Plešovice颗粒内部元素均一性较差,所有检测元素的相对误差均大于10%;Qinghu和Temora1锆石由于检测点位于不同颗粒,因此微量元素含量结果相差较大。总体上,按照锆石中的U平均含量,91500 在球粒陨石标准化图解上,所有锆石样品具有一致的亏损LREEs富集HREEs的特征,明显的Ce正异常,Eu异常差别明显,Temora1和Plešovice具有明显的负Eu异常(图4中a,c),Tanz锆石呈现弱负Eu异常(图4b),GJ-1和SA01中Eu呈现弱负异常或无异常(图4中a,b)。 微量元素替代进入锆石中可能对锆石结构特征造成影响,LA-ICP-MS分析过程中影响激光剥蚀效率,从而影响U-Pb年龄结果[49-50],因此,在锆石U-Pb定年过程中可以根据以上各个锆石标准样品的微量元素特征,结合仪器工作条件选择与待测锆石样品基体更为匹配的标准,以降低由此可能带来的误差。 采用激光剥蚀-扇形磁场高分辨等离子体质谱(LA-SF-ICP-MS)建立了锆石U-Pb年龄与微量元素含量同时检测分析方法,并对常用的锆石U-Pb定年标准样品进行分析。分析结果表明:相对于仅检测U-Pb相关同位素方法,U-Pb定年及微量元素同时检测方法在一定程度上影响了U-Pb同位素信号稳定性,尤其是对低含量元素/同位素信号稳定性的影响更为明显,进而对U-Pb定年结果精密度产生影响。相对仅检测U-Pb同位素分析方法结果,多元素同时分析206Pb/238U年龄和207Pb/235U年龄变化范围变大,其中206Pb/238U年龄RSD值从0.2%~0.9%增大至0.4%~1.4%;207Pb/235U年龄RSD值从0.5%~1.3%增大至1.2%~3.3%,单点年龄误差从~1.5%增大至~2.0%。尽管如此,多元素同时检测方法不会对样品最终年龄结果产生明显影响,该方法测定各样品谐和年龄和206Pb/238U加权平均年龄与仅检测U-Pb同位素方法测定年龄结果在误差范围内一致,与TIMS年龄结果偏差均小于1%,完全满足地质年代学分析需求。同时,测定各锆石样品中微量元素含量值与文献报道范围吻合,其中91500和SA01样品中各元素测定值与文献报道值相对误差小于10%。综合以上分析结果,采用LA-SF-ICP-MS建立的分析方法克服了高分辨质谱仪扫描速度慢的缺点,可以同时准确测定锆石的U-Pb年龄和关键微量元素含量。 尽管LA-ICP-MS锆石U-Pb定年方法已十分成熟,但提高其空间分辨率,并且在有限的分析空间获得更多的原位元素/同位素信息是研究者不变的追求。LA-SF-ICP-MS同时进行U-Pb定年和微量元素含量分析,一方面可以充分利用高分辨质谱高灵敏度特征实现高空间分辨率分析,另一方面原位微量元素含量的准确测定可以获得更多的地质信息,为精细刻画地质演化环境和过程提供丰富地球化学数据。 致谢:中国科学院地质与地球物理研究所王浩副研究员提供SA01锆石样品;中国地质大学(武汉)罗涛副研究员提供Tanz锆石样品,在此表示感谢! BRIEF REPORT Significance:Zircon (ZrSiO4) is a common accessory mineral in both terrestrial and extraterrestrial rocks. It serves as the most frequently utilized mineral in determining the age, origin, and thermal history of rocks through U-Pb geochronology, primarily due to its high closure temperature, resistance to alteration, high uranium content, and minimal incorporation of common lead during its crystallization. Apart from age determination, zircon also provides various geochemical information, including trace elements (Ti, REEs), O isotopes, and Zr isotopes. By combining all these data, it assists in constructing metamorphicP-T-tpaths, which helps in inferring the mineral growth history,making it an integral component of “Petrochronology” research[20].Natural zircon, however, generally exhibits small grains and often possesses intricate internal structural features. Hence, higher analytical spatial resolution and acquiring more in-situ elemental/isotopic information within limited analytical space is crucial. Laser ablation inductively coupled plasma-mass spectrometry (LA-ICP-MS) is a routine technique for zircon UPb dating. The sector field mass spectrometer (SF-ICP-MS) possesses high sensitivity, offering possibilities for high spatial resolution U-Pb dating of zircon[28-29]and ultra-low uranium mineral content[21-25,30]. However, the usage of the magnetic sector mass analyzer results in slower scanning speeds[27]. Consequently, in the reported zircon U-Pb dating by LA-SF-ICP-MS, typically, only U-Pb-ages-related isotopes are detected, and simultaneous quantitative analysis of key trace elements cannot be conducted. This study optimized instrument parameters and enhanced signal stability to enable simultaneous determination of U-Pb dating alongside quantification of key elements such as Ti, REEs, and Hf. Seven commonly employed zircon U-Pb standard samples were measured to assess the feasibility of the method and its influence on U-Pb dating results. Methods:The experiments were carried out at the Key Laboratory of Elemental Microanalysis and Morphology of the China Geological Survey using an ESL NWR 193UC ArF excimer laser and the ELEMENT Ⅱ sector field inductively coupled plasma-mass spectrometer (SF-ICP-MS, ThermoFisher Scientific, USA). Helium gas was used as the carrier gas[39], while 1mL/min of nitrogen was introduced to boost instrument sensitivity[40]. A signal homogenization device with higher aerosol diffusion space was utilized to improve signal stability. Laser spot size and frequency were 25μm and 8Hz, respectively, with a laser energy density of 3J/cm2. SFICP-MS was employed in low-resolution mode (M/ΔM=300), with sensitivity of U about 5000cps/(μg·g-1), and signal stability RSD of 1% to 2%, oxide production rates (ThO+/Th+) below 0.2%. U-Pb isotopes and key trace elements were measured, including29Si,49Ti,89Y,91Zr,139La,140Ce,141Pr,146Nd,147Sm,151Eu,157Gd,159Tb,163Dy,165Ho,166Er,169Tm,172Yb,175Lu,178Hf,206Pb,207Pb,208Pb,232Th, and238U. The measuring time for206Pb,208Pb,232Th, and238U was 10ms,207Pb was 20ms, and the remaining elements were 5ms each, resulting in a total measuring time of 0.87s per reading, with an effective analysis time of 71%. Since most zircons contain no, or extremely low, common lead and single-collector mass spectrometry it makes it challenging to accurately measure204Pb for common lead correction.202Hg and204Pb were not detected to improve the proportion of effective analysis time. Detailed instrument operating condition parameters are provided in Table 1. The 91500 zircon and NIST610 were used as standard materials for U-Pb isotope ratios and trace element quantitative analysis, respectively. For every 10 unknown sample analyses, a set of standard samples was inserted to correct for fractionation effects. Each spot analysis consisted of a 20s background collection and 40s sample data acquisition and 30s flushing. Raw data were reduced offline using GLITTER 4.0 software package[41]. All samples shared the same signal interval as the standard zircon 91500 for analysis. Age calculations and Concordia diagram construction were performed by Isoplot/Ex version 2.23[42]. The uncertainty for all isotope ratios and age values in this study were reported at a 2σlevel. Two methods were used to determine zircon U-Pb ages. Method 1: Only U-Pb-related isotopes were measured,including202Hg,204Pb,206Pb,207Pb,208Pb,232Th, and238U. Single measuring time was 0.306s per reading, and within the 40s sampling time, data for 134 sets of element signal intensities were collected. Method 2: Simultaneous determination of U-Pb ages and trace elements content, including Ti, REE, Hf, and U-Pb isotopes206Pb,207Pb,208Pb,232Th,238U, etc. Single measuring time was 0.857s, and within the 40s sampling time, effective signal intensity data for 47 sets were obtained. Data and Results:The U-Pb dating results are shown in Fig.1 and Table 2, while the measured trace element contents are presented in Fig.3 and Table 3. (1) U-Pb dating results comparison Fig.1 provides a visual comparison of precision and accuracy of the results obtained by these two dating methods. It shows that the dating results from Method 1 exhibit smaller errors, and the data points are more concentrated. Using LA-SF-ICP-MS with solely U-Pb-related isotope detection, the dating results show an error of 1.5% and 1.3% for206Pb/238U and207Pb/235U, respectively. The consistency of single-point206Pb/238U and207Pb/235U ages is good, with RSD values ranging from 0.2% to 0.9% and 0.5% to 1.3%, respectively. In contrast, when employing Method 2, the errors for206Pb/238U and207Pb/235U ages increase slightly to 1.9% and 1.7%, respectively. The dispersion of single-point ages increases as well, with RSD values for206Pb/238U and207Pb/235U ages ranging from 0.4% to 1.4% and 1.2% to 3.3%, respectively. Particularly, the RSD value for207Pb/235U ages notably increases. To investigate potential causes for the variability in precision between the two methods, the RSD of the signal intensities of206Pb,207Pb, and238U/206Pb signal ratios of zircon 91500 obtained under the same experimental conditions were statistically analyzed. For Method 1, the RSD values for the collection of206Pb,207Pb signals, and238U/206Pb signal ratios were 13%, 13%, and 5.2%, respectively. Meanwhile, for method 2, the RSD values for206Pb,207Pb signals, and238U/206Pb isotope signal ratios were 14%, 19%, and 9.3%, respectively. Simultaneously collecting multiple elements extended the mass spectrometer’s single-scan time, which to some extent affected the stability of the detected element/isotope signals and their ratios, especially for low-content isotopes like207Pb. The fluctuation characteristics of isotopic signals and ratios corresponded to the changes in U-Pb dating results. Considering factors during data processing, age uncertainty calculation methods, and influencing elements[43-46], it can be inferred that the impact of simultaneous multi-element detection on the stability of isotope signal intensity might be the primary cause for the increased final dating result errors. Generally, the precision of LA-ICP-MS zircon U-Pb dating results is considered to be within 1% to 2%, while the accuracy of the measured age can reach 1% or better[43-46]. The experimental results in this study indicate that although simultaneous multi-element detection can increase the range of changes in single-point zircon U-Pb ages,the precision of the data is still better than 2%. Furthermore, simultaneous multi-element detection does not affect the accuracy of the sample’s concordia age and the206Pb/238U weighted average age. For method 1, the relative deviations of concordia ages and206Pb/238U weighted average ages for each sample compared to TIMS ages are less than 0.5%, while for method 2, these figures are less than 1.0% and 0.7%, respectively. (2) Quantitative results of trace elements The results of trace element analysis in zircon samples are shown in Table 3. The detection limits for REE range between 11ng/g and 73ng/g, with the majority being below 20ng/g. Among these, the detection limits for La and Pr are 20ng/g and 13ng/g, respectively. Ti has the highest detection limit, reaching up to 302ng/g, while Pb, Th,and U have the lowest detection limits at 15ng/g, 4ng/g, and 2ng/g, respectively. Si and Zr were employed as internal standard elements for quantitative analysis. All quantitative results of zircon samples using Si as the internal standard element show slightly higher values overall than those calculated using Zr as the internal standard element (Fig.2b). For the 91500 and SA01 samples, the analysis results using Si as the internal standard exhibit relative errors of less than 10% compared to the recommended values in literature [35]and [48]. However, when using Zr as the internal standard, the relative errors range between 10% and 20%.Furthermore, due to significant differences in Zr content between zircon and the selected standard samples,prolonged flushing is required after zircon analysis to decrease the background Zr levels in the instrument under conditions of high aerosol diffusion space. This prolonged flushing might impact the accuracy of Zr detection in the standard samples.3 结论