地质矿物中锶同位素检测方案(激光剥蚀进样)

智能文字提取功能测试中

Chemical Geology 479 (2018) 10-21Contents lists available at ScienceDirectChemical Geology W. Zhang et al.Chemical Geology 479 (2018) 10-21 journal homepage： www.elsevier.com/locate/chemgeo Improved in situ Sr isotopic analysis by a 257 nm femtosecond laser incombination with the addition of nitrogen for geological minerals Wen Zhang，Zhaochu Hu*， Yongsheng Liu ， Tao Wu， Xiaodong Deng， Jingliang Guo， Han Zhao State Key Laboratory of Geological Processes and Mineral Resources， China University of Geosciences， Wuhan 430074， ChinaSchool of Earth Sciences， Zhejiang University， Hangzhou 310027， China 1. Introduction The radiogenic 87Sr/ 86Sr ratio is an important geochemical tracer insolid earth sciences (Davidson et al.， 2001； Jackson and Hart， 2006).High precision and bulk-rock Sr isotopic ratios have been determinedmainly by thermal ionization mass spectrometry (TIMS) (Yang et al.，2010； Koornneef et al.， 2015) and multiple collector-inductively cou-pled plasma mass spectrometry (MC-ICP-MS) (Waight et al.， 2002； Galler et al.， 2007； Yang et al.， 2011a). However， bulk-rock Sr isotopicratios generally reflect a mixed signature of various end-members thatmight not reach isotope equilibrium in the targeted rock (Davidsonet al.， 2001； Davidson et al.， 2007； Ramos and Tepley， 2008). In thisregard， in situ Sr isotopic measurements using laser ablation (LA)-MC-ICP-MS offer the spatial resolution to identify and distinguish the inter-and intra-crystalline isotopic variations of mineral grains on the scale oftens coff microns. As aresult， LA-MC-ICP-MSfor Sr isotopic ( . ht tps： // d o i . o r g/ 1 0.101 6/ j.c he mgeo.2017. 1 2.018 ) ( R eceived 1 8 J uly 2017； Received in revised form 13 November 2 017； A ccepted 2 1 December 2 0 17Available online 28 December 2017 ) ( 0009-2541/C 2017 Elsevier B.V . All ri g hts r e served ) measurements has received increasing interest in earth science， en-vironmental science and archaeology applications， including： (1)magmatic minerals such as carbonate， apatite， plagioclase， clinopyr-oxene， and perovskite (Christensen et al.， 1995； Davidson et al.， 2001；Waight et al.， 2002； Bizzarro et al.， 2003； Schmidberger et al.， 2003；Ramos et al.， 2004； Woodhead et al.， 2005； Yang et al.， 2009； Kimuraet al.， 2013； Tong et al.， 2016) and melt inclusion (Jackson and Hart，2006)；(2) biogenic carbonate such as otoliths (Outridge et al.，2002；Barnettjohnson et al.， 2005) and shells (Christensen et al.， 1995； Ramoset al.， 2004； Horstwood et al.，2008)； and (3) archaeological tooth en-amel (Balter et al.， 2008； Copeland et al.， 2010). Christensen et al. (1995) first demonstrated the feasibility of in situSr isotopic analysis using LA-MC-ICP-MS on both carbonate and feld-spar samples. Further studies reported highly precise (30-100 ppm) andaccurate 87Sr/86sr isotope ratios with the nanosecond (ns) laser abla-tion system (Davidson et a.， 2001； Waight et al.， 2002； Ramos andTepley， 2008； Kimura et al.， 2013； Chen et al.， 2015). Most of inter-esting and measured minerals for in situ Sr isotopic analysis weretransparent minerals. The previous studies reported that the interactionbetween transparent minerals and ns laser pulses was a non-linear ab-sorption process known as avalanche ionization (Liu et al.， 1997；Shaheen et al.， 2012). In this process， ablating transparent materials byns laser pulses depends on the pre-existence of seed electrons， which aremetallic impurities or thermal ionizations of shallow energy levels inthe transparent materials used to initiate laser-induced breakdown.However， the seed electrons are normally only present at low con-centrations and are randomly distributed in transparent materials， re-sulting in stochastic and inefficient ablation behavior (Liu et al.， 1997).In the previous studies， therefore， large spot sizes (100-300 um) wererequired to obtain sufficient signal intensity for Sr/86Sr isotopicanalyses of geological samples. Recently， femtosecond (fs) laser ablation systems with short pulsewidths have become available (Russo et al.， 2002； Hergenroder et al.，2006； Fernandez et al.， 2007； Pisonero and Giinther， 2008； Shaheenet al.， 2012). With fs laser ablation， the pulse duration (<1ps) isshorter than the phonon relaxation time (ca. 10 ps) (von der Lindeet al.， 1997； Mao et al.， 2004). Ultrafast energy deposition with fs laserpulses vaporizes and ablates the illuminated volume before thermalrelaxation sets in. The heat diffusion， which occurs during the ns laserpulse and results in a significant heat affected zone， is significantlyinhibited (von der Linde et al.， 1997； Mao et al.， 2004). In addition， thepulse intensity (irradiance or power density) reaches much larger va-lues for fs pulses (1014-1015Wcm-2) compared to ns pulses(10°-101°w cm-2)， which would improve the analytical quality ofablated materials. Due to these advantages， fs-LA-ICP-MS has beensuccessfully applied to the analyses of element ratios and elementconcentrations (Poitrasson et al.， 2003； Bian et al.， 2006； Koch et al.，2006； Horn and von Blanckenburg，2007； Borisova et al.， 2008； Jochumet al.， 2014； Li et al.， 2015； Li et al.， 2016) as well as isotope ratios， suchas Sr (Campos-Alvarez et al.， 2010； Yang et al.， 2011b)， U-Pb (Freydieret al.， 2008； Hirata and Kon， 2008； Kimura et al.， 2014)， Pb (Shaheenand Fryer， 2010； Chen et al.， 2014； Ohata et al.， 2015)， Mg (Oeser et al.，2014)， Si (Chmeleff et al.， 2008； Steinhoefel et al.， 2011； Schuessler andvon Blanckenburg， 2014)， Fe (Horn et al.， 2006； Steinhoefel et al.，2009b；Steinhoefel et al.， 2009a； Oeser et al.， 2014)， and Cu (Ikehataet al.， 2008； Ikehata and Hirata， 2013). However， the detailed studiesfor ablating transparent minerals by fs laser are not sufficient. There isgreat potential in elemental quantitative and isotopic ratio analyseswith high spatial resolution for transparent minerals using a fs lasersystem. An additional challenge for Sr isotope analysis using LA sampling isthe variety of isobaric interferences on the Sr isotope spectrum， in-cluding Rb， Kr， doubly charged rare earth elements (REEs)， calciumdimers and argides and possibly oxides (Ramos et al.， 2004； Woodheadet al.， 2005； Vroon et al.， 2008； Yang et al.， 2011b；Kimura et al.， 2013；Yang et al.， 2014). The strategies of eliminating or reducing these isobaric interferences have been widely explored and still attract sig-nificant attention. In recent decades， mixed-gas plasma has been ex-tensively investigated to overcome analytical limitations of the Arplasma， including the addition of nitrogen (N2)， hydrogen (H2)， me-thane， carbon-containing solvents and water (Durrant， 1994； Guillongand Heinrich， 2007； Hu et al.， 2008； Fliegel et al.， 2011； Hu et al.，2012a； Shaheen et al.， 2012； Lin et al.， 2014； Liu et al.， 2014； Xu et al.，2015； Fu et al.， 2016； Tong et al.， 2016). N2 is one of the most widelyused molecular gases in ICP-MS and has been utilized mainly for theimprovement of sensitivity and stability， the decrease in polyatomicinterferences and the attenuation of matrix effects. Recent studies showthat adding N2 to the Ar carrier gas can significantly enhance signalintensity during LA-MC-ICP-MS analysis， such as for Hf， Nd and Pb (Huet al.， 2012a； Shaheen et al.， 2012； Xu et al.， 2015). The N2-relatedintensity enhancement was attributed to the higher thermal con-ductivity of N2， which results in higher plasma temperature and theincreased excitation and ionization efficiency of many elements (Huet al.， 2012a； Shaheen et al.， 2012). However， the suppression of signalintensity resulting from N2 addition has been observed in some analysesof B， S and Sr (Lin et al.， 2014； Fu et al.， 2016； Tong et al.， 2016). Thus，the enhancement or suppression of signal intensity caused by the N2addition is element-specific. On the other hand， the more conclusiveobservation is that the addition of N2 can reduce polyatomic inter-ferences in the ICP. For example， Durrant (1994) first reported that theaddition of approximately 1% (v/v) N2 to the coolant flow or the ad-dition of approximately 12% N2 to the cell gas reduced CeO+/Cet andTho+/Th ratios by a factor of 2-3 in LA-ICP-MS. Hu et al. (2008)noted that the addition of 5-10 ml min-N2 to the central channel gasin LA-ICP-MS increased the sensitivity for most of the 65 investigatedelements by a factor of 2-3 while reducing the oxide ratios (ThO*/Th) (by one order of magnitude) and the hydride ratio (ArH*/Ar*)(by a factor of 3). Shaheen et al. (2012) showed increased abundancesof doubly charged ions (U++/U and Ce++/Ce) but reduced oxide ra-tios (UO+/U+ and ThO+/Tht) and mass bias of Pb and Tl upon theaddition of 5 ml min-1N2. Fu et al. (2016) reported that the additionof N2 efficiently reduces polyatomic interferences (O0*， SH* andOOH*) for in situ S isotope analyses in sulfides. Feldspars have high Sr abundances， ubiquitously occur in igneousrocks and exhibit a large range of crystallization temperatures(Gagnevin et al.， 2005； Charlier et al.， 2006； Davidson et al.， 2007).Therefore， feldspars were selected in this study as the main researchmineral for in situ Sr isotope analysis. Comparisons of the ablation rates，elemental sensitivities and analytical precision of ’Sr/86Sr ratios be-tween a 257 nm fs laser and a 193 nm ArF excimer ns laser were per-formed. In addition， we systematically investigated the effects andlimitations of small amounts of N2 mixed to the central channel gas flowof LA-MC-ICP-MS. The suppression of the sensitivities towards Rb andKr and yields of polyatomic interferences were found with the additionof N2. Combining the high ablation efficiency of the fs laser with thecapacity of suppressing the interference factors by adding N2， the im-proved in situ Sr isotopic analysis method using the 257 nm fs laser wasestablished. The feasibility and flexibility of the proposed method weredemonstrated by analyzing common plagioclases， K-feldspars with highRb/Sr ratios and low-Sr clinopyroxenes. 2. Analytical procedure 2.1. Instrumentation In situ Sr isotope analyses were performed on a Neptune Plus MC-ICP-MS (Thermo Fisher Scientific， Bremen， Germany) coupled with twolaser ablation systems at the State Key Laboratory of GeologicalProcesses and Mineral Resources (GPMR)， China University ofGeosciences (Wuhan)， China. The Neptune Plus， a double focusing MC-ICP- MS， was equipped with seven fixed electron multiplier ICs， andnine Faraday cups fitted with 10 resistors. In addition， a large dry Table 1 Summary of the operating parameters for the 193 nm excimer laser， the 257 nm fs laser，the MC-ICP-MS and the desolvation nebulizer system. Cup-configuration L4 L3 L2 L1 H2 H3 H1 83Kr 167E* 86Sr *4Sr 85Rb 87Sr 88Sr 173Yb RF Power 1250 W Cool gas flow 16.0 Lmin Auxiliary gas flow 0.10 L min Argon make-up gas flow 0.80 L min Helium carrier gas flow 0.50 L min Nitrogen gas flow 0，6， 12 ml mint Interface cones X skimmer cone + Jet sample cone Instrument resolution ~400 (low) Block number Cycles ofeach block 160 Integration Time (s) c. 0.524 sec Laser ablation system Laser type ArF excimer laser Yb：YAG femtosecond laser Wavelength 193 nm 257nm Pulse length 15 ns 300 fs Energy density 2.3-13.3 J cn 0.50-3.85 Jcm Spot size 60， 90.120 um 45， 60 pm Laser frequency 10 Hz 10-200 Hz Line scanning rate 5 um s Aridus Ⅱ desolvation nebuliser system Membrane temperature 160℃ Spray chamber temperature 110℃ Sweep gas flow 2.5 L min Ar Sample uptake rate 50 pL min' PFA nebuliser Nebuliser flow 0.95 L min A LA-ICP-MS combination， consisting of an Agilent 7500a ICP-MS(Agilent Technology， Tokyo， Japan) coupled with the ns- or fs-laserablation systems， was used for the measurements of major and traceelement concentrations in natural minerals and reference glasses.Detailed operating conditions for the laser ablation system and the Q-ICP-MS instrument and data reduction are given in SupplementaryInformation A. The CETAC Aridus IIM desolvation nebulizer system (CETACTechnologies， Omaha， USA) was used to investigate the influence ofinterferences on Sr isotopic determination. A series of NIST SRM 987solutions doped with interfering elements were measured using MC-ICP-MS. The solutions were self-aspirated at an uptake of 100 ml minthrough the PFA nebulizer and desolvated by the Aridus II system. The Surface Profiler (P16*， KLA-Tencor， California， USA) was usedto measure the depth of the ablated crater. Each crater was profiledacross the centerline over a range of 200 um using a speed of10 um s-. The P16+Surface Profiler uses a diamond ：stylus(radius= 2 um). Tests were performed at the Center of Micro- Fabrication and Characterization (CMFC) of the Wuhan NationalLaboratory for Optoelectronics (WNLO)， China. 2.2. Samples and reagents The NIST reference glasses SRM 610 and SRM 612， USGS referenceglasses BHVO-2G and GSE-1G and MPI-DING reference glasses ATHO-G， StHs6/80-G and T1-G were analyzed. Three feldspar megacrysts， YG0440 (albite)， YG0383 (albite) andYG4301 (anorthite)， were collected from Hebei Province (China) andinvestigated in this study， A K-feldspar megacryst (Tuyk) from a peg-matite deposit in Henan province (China) with a high Rb/Sr ratio (0.46)was used. Feldspar megacrysts were crushed into 2-5 mm fragmentsand cleaned with deionized water. Some of the fragments weremounted in epoxy resin discs for elemental and Sr isotope analysesusing LA-ICP-MS and LA-MC-ICP-MS， respectively. The remainingfragments were milled to < 200 mesh and were then acid-digested forSr isotope ratio measurements by TIMS. A natural clinopyroxenemegacryst (Cpx， HNB-8) with a low Sr concentration (89.2ugg) wasanalyzed. The chemical and Sr isotopic compositions of Cpx HNB-8have been reported by He et al. (2013) and Tong et al. (2016). Samples of natural pyrite and garnet from the Rock and MineralTeaching Section in the China University of Geosciences (Wuhan) wereused to investigate the difference in the laser rate between ns and fslaser ablation. Their chemical and isotope compositions were not de-termined. A Sr isotope-certified reference material (NIST SRM 987) and a Rbisotope-certified referencenmaterial (NISTSRM 984) (NIST，Gaithersburg， USA) were prepared using 2% HNO3. Single elementstandard solutions of Ca， Er and Yb (National Center for Analysis andTesting of Steel Materials， China) were used for the experiments withdoped interfering elements. Commercially available nitric acid (GRgrade) was further purified twice using a DST-1000 acid purificationsystem (Savillex， Eden Prairie， USA). 2.3. LA-MC-ICP-MS measurement and data reduction For in situ Sr isotope measurements， the combination of the high-sensitivity X skimmer cone and JET-sample cone was employed， and themass spectrometer was operated in the low mass resolution mode. Asmall amount of N2 (6-12ml min) was added to the carrier gas flowbehind the signal-smoothing device by a simple Y connector.NIST SRM610 was used to optimize the instrumental parameters， including the Heand Ar gas flow rates， the torch position， the RF power setting， and thesource lens settings for maximum sensitivity and optimum peak flat-ness. The routine data acquisition consisted of one block of 160 cycles(0.524 s integration time per cycle)， with the first 50 cycles being forbackground collection (no laser ablation) and the remaining 110 cyclesfor signal collection. The detailed setup parameters of the laser systemand MC-ICP-MS are listed in Table 1， including the Faraday-cup setup inthe mass spectrometer. The data reduction for LA-MC-ICP-MS analysis was conducted usingExcel spreadsheets. The interference correction strategy was the sameas the one reported by Tong et al. (2016). In summary， the regions ofintegration for both gas background and sample were selected first.Following background correction， which removes the background Krsignals， no additional Kr peak stripping was applied. Interferences werecorrected in the following sequence： (1) the interferences of 168Er++on84sr，170Er++and 170Yb++ on 85Rb，172Yb++ on 86Sr， and 174Yb++on8’Sr were corrected based on the measured signal intensities of167Er++，173Yb++84， 86-88Sr and 85Rb and the natural isotope ratios ofEr and Yb (Berglund and Wieser， 2011)； (2) the isobaric interference of8Rb on Sr was corrected by monitoring the Rb signal intensity anda user-specified Rb/Rb ratio using an exponential law for mass bias.The user-specified &Rb/8Rb ratio was calculated by measuring somereference materials with a known 8Sr/86Sr ratio. Following the interference corrections， mass fractionation of Sr isotopes was correctedby assuming 88Sr/86Sr=8.375209 (Jochum et al.， 2009) and applyingthe exponential law (Russell et al.， 1978). 3. Results and discussion 3.1. Ablation rate In this study， a range of natural minerals (pyrite， garnet， albiteYG0383 and Cpx HNB-8) and NIST SRM 610 glass were used to in-vestigate the ablation rates of ns and fs laser pulses. Fig. 1 shows therelationship between the number of laser pulses and crater depth invarious samples obtained by ns and fs laser ablation. When using the nslaser at a repetition rate of 10 Hz， a laser spot of 60 um and a laserfluence of ~9.5 J cm-2， the depth of ablation craters differed sig-nificantly in different types of materials (Fig. 1a). The crater depth inpyrite was significantly higher compared to silicate minerals and NISTSRM 610 glass. For albite YG0383， the lowest ablation rate of 0.026 umper pulse was found， which was approximately 3.3 times lower thanthat of NIST SRM 610 (0.086 um per pulse) (Fig. 1a). In contrast， for fslaser ablation with a repetition rate of 10 Hz， a laser spot of 60 um and alaser fluence of ~3.8 J cm-2， the ablation rates of all samples wereapproximately consistent (0.08-0.11 um per pulse) (Fig. 1b). Moreover，the lower fluence of the fs laser (~3.8Jcm) contributed the higherablation rates for feldspar (albite YG0383)， Cpx (HNB-8) and garnet，indicating the better ablation efficiency of the fs pulse relative to the nslaser for natural transparency minerals. Although the 193 nm ultraviolet (UV) can be absorbed by mostnatural minerals (Giinther et al.， 1999)， there is a fundamental differ-ence in the ablation mechanism between ns and fs laser ablation. Forthe ns laser， ablation starts through the absorption of the incident laserenergy by the free electrons in the target material (Liu et al.， 1997；Hergenroder et al.， 2006； Shaheen et al.， 2012). The ablation rate of thens laser would be influenced by the physical and chemical properties ofthe samples (Fig. 1a). Pyrite can effectively absorb laser energy andproduce a high ablation rate during ns laser ablation (0.144 um perpulse) because it contains many free electrons. However， free electronsnormally present at low concentrations and randomly distributed intransparent minerals， such as natural silicate minerals， result in a sto-chastic and inefficient ablation (Liu et al.， 1997； Hergenroder et al.，2006； Shaheen et al.， 2012). Feldspars are transparent minerals andoften lack metal elements， such as Fe. Therefore， albite YG0383 showsthe lowest ablation rate (0.026 um per pulse). For fs laser ablation， dueto the high pulse intensity， the interaction of fs laser pulses with solidmaterials is mainly dominated by two types of non-linear absorptionprocesses： avalanche ionization and multi-photon ionization. In multi- photon ionization， bound electrons of the transparent material can bedirectly ionized by simultaneously absorbing multiple photons andforming a mass of free electrons to initiate the avalanche ionization.Then， the energy transfer of the fs pulses into the samples does notdepend on already pre-existing free electrons， and the ablation event ismore deterministic and initiated by a similar laser-material interactionmechanism for various materials with different physical and chemicalproperties (Liu et al.， 1997； Shaheen et al.， 2012). Therefore， the ab-lation rates of the fs laser are consistent for various samples， includingpyrite and feldspar (Fig. 1b). This advantage not only shows the benefitof eliminating or weakening the matrix effect during the laser ablationprocesses but also helps to improve the analytical sensitivity fortransparent minerals， such as feldspars. 3.2. Signal responses in ICP-MS Fig. 2 shows the Sr responses (cps/ug g) obtained from LA-ICP-MS for reference glasses (NIST SRM 610， NIST SRM 612， ATHO-G，BHVO-2G) and natural feldspars (albites YG0383 and K-feldspar Tuyk)with changing laser fluences of the ns and fs lasers. For the ns laser， Srresponses were obviously different among the NIST glasses， USGSglasses and natural feldspars (Fig. 2)， which were consistent with theabove studies about ablation characteristics. The Sr responses of naturalfeldspars obtained by the ns laser were two times lower than those ofthe NIST glasses at the same fluence. The lower ablation rate of feld-spars was considered as the main reason. In contrast， for fs laser abla-tion， a similar response trend with increasing laser fluence was ob-served in all the samples (Fig. 2). This is in line with the expectation ofthe similar ablation rates for various materials ablated by the fs laser， asmentioned above. The direct and exact comparison between the ns laserand fs laser used in this study cannot be achieved due to the differentenergy distributions (flat peak vs Gaussian peak) and energy outputs.However， as a rough comparison， the Sr responses in feldspars from fslaser ablation at 3.8 J cm-2were 3.4 times higher than those from thens laser at 4.1 J cm-2. In addition to higher ablation rates of the fs laserpulses， the high sensitivity observed in fs laser ablation could be relatedto the size of laser-generated particles. The particles produced by fslaser ablation have small sizes distributed over a narrow range (D'Abzacet al.， 2012， 2013). This improves particle transport efficiency and io-nization inside the ICP and consequently improves the sensitivity andsignal reproducibility. However， the direct evidences between the im-proved signal sensitivity and the particle morphology were not pro-vided in this study. Further investigations are needed to validate thisassumption. The promotion of signal intensity benefits the improvement of theprecision for isotope ratio measurements. The within-run precisions Fig. 1. Depth vs applied number of laser pulses for various samples using the 193 nm ns laser system with a repetition rate of 10 Hz， a laser spot of 60 um and an energy density of~9.5 J cm-2(a) and using the 257 nm fs laser system with a repetition rate of 10 Hz， a laser spot of 60 um and an energy density of ~3.8 J cm-2(b). Fluence (J cm) Fig. 2. The Sr signal responses (cps/ug g) for reference glasses (NIST SRM 610， NSIT SRM 612， ATHO-G， and BHVO-2G) and nature feldspars (albites YG0383 and K-feldspar Tuyk)with changing laser fluences for the 193 nm ns laser (a) and 257 nm fs laser (b). The other laser parameters were the repetition rate of 10 Hz and the laser spot of 60 um in both laserablation system. (standard error， SE， k=2) of the Sr/86Sr ratio in BHVO-2G andnatural albite (YG0440) by the ns and fs lasers are shown in Fig. 3. Thesame laser repetition rate of 10 Hz and a spot size of 60 um were used，but the fluences were ~9.5 J cm-2and 3.8 J cm-2for the ns laser andfs laser， respectively. The concentrations of Sr and Rb are 396 uggand 9.20 ug g；-1 for BHVO-2G and 390ugg and 0.23 uggforYG0440， respectively. The theoretical precisions， Poisson countingstatistics errors (SDp， Yang et al.，2011b)， were calculated. As seen inFig. 3， the observed 2SE of the 87Sr/ 86Sr ratio correlates well with theSD. For the fs laser， the 88sr signal intensity and 2SE of the 87Sr/86srratio were approximately similar for BHVO-2G and YG0440. However，for the ns laser， the 88Sr signal intensity of YG0440 was approximately0.88V and ~42%lower than that of BVHO-2G (1.52V)， resulting inthe obvious deterioration of the 2SE of the ’Sr/86Sr ratio (Fig. 3).Obviously， the fs laser improves the analytical sensitivity and the pre-cision for transparent minerals (feldspars) due to the better ablationefficiency. This advantage will become more apparent for the low-Srminerals or high spatial resolution cases. 3.3. Influence of N2 addition on interference correction LA-MC-ICP-MS is a solid sample direct injection technique. Manyinterfering elements will be introduced into ICP-MS due to the absenceof a chromatographic purification process. The main interferences of Srisotope analysis for feldspar are Rb， Kr， calcium dimers and argides， anddoubly charged Er and Yb (Ramos et al.， 2004). In this section， we Fig. 3. Correlations between the within-run precisions (standard error， SE， k=2) ofsingle spot analyses and Sr signal intensities of the reference glass BHVO-2G and naturalalbite (YG0440) analyzed by ns and fs laser ablation. describe our investigation of the effect of inhibiting sensitivities of in-terfering elements and yields of polyatomic ions by the addition of N2. 3.3.1. Calcium dimers and argides， doubly charged Er and Yb In this study， NIST SRM 987 solutions (100ug1-) doped with in-terfering elements (Er and Yb) and a Ca solution (Ca= 500 ug1)were measured using MC-ICP-MS with a desolvation sample introduc-tion system (Aridus II) and GE on mode. The ion beams on masses of 84，83.5 and 86.5 were collected to represent the polyatomic ions ofCalcium (CaAr*/CaCa*)， doubly charged Er (Er++) and Yb(173Yb++)， respectively. The signal intensities of 88sr，167Er++，173Yb++and CaAr+/CaCa+ as a function of N2 addition are illustratedin Fig. 4. Increasing N2 from 0 ml min-1 to 12 ml min-1leads to agradual decrease in the signal intensity of 88Sr. Sr with a low ionizationpotential (IP=5.7 eV) can be completely ionized in a typical ICP；therefore， its sensitivity cannot be improved by increasing the tem-perature of ICP by adding N2. 167Er++and 173Yb++ were suppressed ata similar extent of Sr (Fig.4b and c， respectively). However， the signalintensity of CaAr/CaCa shows a significant decrease associated withan increase in the N2 flow rate， such as the signal suppression of 6.5times and 11.7 times for N2 additions of 6 ml min-1 and 12 ml min-1，respectively. This may be due to the higher ICP temperature by addingN2， which could result in decomposition of the polyatomic ion. 3.3.2. Krypton (Kr) The noble gas krypton (Kr) interferes with masses 84Sr and 86Sr andis therefore an important interference that needs to be corrected. Thesource of the Kr is the argon (and helium) gas used to transport thesample into the plasma. The amount of krypton in the argon (and he-lium) gas is supplier dependent. Woodhead et al. (2005) observed thatthe Kr abundances in the Ar supply vary largely between differentbatches， with a total Kr contribution of approximately 20 mV. Ramoset al. (2004) reported that Kr is present in low levels as an impurity inthe argon gas (typically <1 mV for 83Kr) and in the helium transfer gasused during laser ablation (< 3 mV for 83Kr). Kimura et al.(2013)suggested that both signal suppression and enhancement are possibleon Kr baselines due to (1) the suppression by laser aerosol loading and(2) the enhancement from Kr in the ablated samples. In this section， the natural albite (YG0383) was analyzed by fs-LA-MC-ICP-MS in GE on mode. Fig. 5 presents the average signal intensitiesof 88Sr* from YG0383 and Kr* in the background without laser firingas a function of N2 addition. With the addition of N2 from 0 ml minto 6 ml min- and 12 ml min-， the signal intensities of 88sr* inYG0383 decreased from 3.3 Vto~2.5V(1.32 times) and~1.9 V (1.73times)， respectively (Fig. 5a). However， the signal intensities of 83Kr+decreased significantly from 0.00163 V to 0.00027 V (6.0 times) and0.00013 V (12.5 times)， respectively (Fig. 5b). Kr has a high first ionization energy (14.0 eV) relative to Sr (4.2 eV)， which could theo-retically be improved the sensitivity due to the increase in ICP tem-perature by adding N2. However， the reverse experimental results inFig. 5 indicated that another important parameter must be considered.The evaporation enthalpies of Kr and Sr are 9.0kJ mol-and144 kJ mol，respectively. With the increase in ICP temperature byadding N2， Kr could evaporate and ionize earlier than Sr near the torch，resulting in a separation of the optimum-ICP parameters between Krand Sr. Then， the shorter sampling depth (the Z-axis value) and thehigher sample gas flow rate should increase the sensitivity of Kr. Fig. 5bshows that the signal intensities of Kr increase gradually with thesample gas flow rate from 0.10 ml minto 0.58 ml min， which isconsistent with our speculation. To the best of our knowledge， thisstudy shows for the first time that the addition of nitrogen dramaticallysuppressed the Kr interference， which would significantly benefit for the in situ Sr isotopic analysis by LA-MC-ICP-MS. Further investigationsare needed to reveal the exact mechanism behind this phenomenon. 3.3.3. Rubidium (Rb) Rubidium (Rb) is a well-known interference with &Sr for the87Sr/86Sr ratio analysis. Feldspar samples often contain a large amountof Rb， which can substitute for K in the way of isomorphism. The iso-baric interference of Rb with Sr was corrected by the measured86Sr/88Sr ratio and a natural 7Rb/85Rb ratio of 0.38571 (Ramos et al.，2004). In this method， the mass bias of Rb is assumed to be consistentwith that of Sr. However， it has been recognized in recent years thatdifferent elements can isotopically fractionate differently in MC-ICP-MS， such as Pb and Tl (Zhang et al.， 2016). An alternative approach isusing a certified reference material to calibrate a user-specifiedRb/Rb ratio (Jackson and Hart， 2006； Jochum et al.， 2009； Tong Fig. 5.Fig.5(a) The average signal intensities of 88Sr*obtained from YG0383 by fs-laser ablation MC-ICP-MS as a function of N2 addition (0， 6， 12 ml min). (b) The average signalintensities of 83Kr* in the gas background without laser firing as a function of N2 addition (0， 6， 12 ml min-). The vertical lines in (b) represent the optimum range of sample gas flowrate for 88sr* signal intensity at different N2 gas flow rates. Rb/Sr Fig. 6. (a) The accuracy of ’Sr/ 86Sr for a series of NIST SRM 987 doped with increasing amounts of Rb at normal and N2 (N2 =12 ml min) modes. The interference of Rb on ’Srwas corrected by the natural isotopic composition of Rb and fsr(b) The same data was corrected using a user-specified fRb， which was obtained from a NIST SRM 984 solution(100 ug1 1). Gray fields represent the accuracy of Sr/86Sr <0.0002. The black lines in each panel are simulated lines with different Afrb values. et al.，2016). The essence of the latter approach is the use of two dif-ferent mass bias factors (f) for Sr and Rb.Through determining moreaccurate Rb mass bias factors (fRb)， we were able to measure high Rb/Srratio materials. In this section， we first simulated the effect of AfRb at different Rb/Srratio conditions， as shown in Fig. 6.Afrb is the deviation between theused fRb and the true fRb. As seen in Fig. 6a， the black lines represent theaccuracies of the 87Sr/ 86sr ratios calibrated by different Af&b valuesfrom - 10% to 10%. If AfRb values are >10% or- 10%， the accuraciesof the 87Sr/86Sr ratios will be>0.0002 for the samples with Rb/Srratios > ~0.01， while the superior accuracies of < 0.0002 can be ob-tained for the samples with higher Rb/Sr ratios >0.10 if the AfRb va-lues are smaller than 1%. In other words， the higher Rb/Sr ratio sam-ples can be measured accurately when the used frb is closer to the truefb· After the mathematical simulation， we have undertaken a series oftests using Rb-doped (NIST SRM 984 Rb solution) NIST SRM 987 so-lutions (Fig. 6). In Fig. 6a， the mass bias factor of Sr (fsr) was used tocorrectttheImass bias(of Rb. Inthenormal condition(N2=0 mlmin-)， the deviation of f&b and fsr (AfRb) was up to~10%，resulting in the deviation of the 87Sr/86Sr ratios of> 0.0002 for Rb/Srratios >0.05. The measured results were well overlapped with the si-mulated lines. In Fig. 6b， the same data in Fig. 6a were processed againusing a user-specified fRb， which was obtained from a Rb standard so-lution (NIST SRM 984，100 ug1 ). The results show that the accuracyof the 87Sr/86Sr ratios significantly improved to <0.0002， even withRb/Sr ratio =0.2. Conclusively， the accuracy of feb and the Rb/Sr ratioare important parameters for the exact Rb-interference correction. An interesting finding was that the sensitivities of Rb in the normalcondition (N2=0ml min-1) were always higher than those of Sr. Themeasured Rb/Sr signal ratios were thus higher than the reference Rb/Srratios in the samples (Fig. 7). With the addition of N2 (12 ml min )，the sensitivities of Rb and Sr suffered an obvious suppression. However，Rb was worse， reducing the Rb/Sr signal ratios by approximately 1.47times (Fig. 7). The first ionization energy of Rb (5.7 eV) is similar tothat of Sr (4.2 eV)， but the evaporation enthalpy of Rb (72.2kJ mol-)is 2 times lower than that of Sr (144 kJ mol). Therefore， the differ-ence of evaporation enthalpies could be the main reason for the dif-ferent degrees of signal suppression of Rb and Sr with the addition ofN2. The advantage of reducing the Rb/Sr signal ratio with the additionof N2 is the more accurate correction of Rb interference for the rich-Rbsamples. This is demonstrated in Fig. 6b； with the addition of N2(12ml min-1)， the accuracies of the 87Sr/86Sr ratios were better than0.0002 even with Rb/Sr ratio = 0.5. Rb / Sr (reference values) Fig. 7. Effect of N2 addition on Rb/Sr signal ratios evaluated by solution-MC-ICP-MS. Thedata was obtained by analyzed a series of NIST SRM 987 doped with increasing amountsof Rb at normal and N2 (N2 = 12 ml min ) modes. 3.3.4. The effect of adding N2 for Sr isotope analysis Three feldspars (YG0383， YG0440， Tuyk) were analyzed by fs-LA-MC-ICP-MS to evaluate the effect of interference suppression by addingN2. Fig. 8 presents the accuracies of the 87Sr/86sr ratios and the84Sr/86Sr ratios as a function of N2. With the addition of N2， both sta-bility and accuracy of the measured 8Sr/ 86Sr ratios improved， espe-cially for Tuyk， which is Rb-rich K-feldspar and has a high Rb/Sr ratio(0.46) (Fig. 8a). More significant improvements occurred in theSr/8oSr ratios. In the normal condition (N2=0 ml min)， residualanalytical biases of 4Sr/86Sr in three feldspars were observed after on-peak background subtractions and mass-fractionation corrections(Fig.8b). With the addition ofN2， the accuracies of the 84Sr/ 86Sr ratiosimproved， indicating that either Ca polyatomic interferences or Kr ef-fects (such as Kr baseline suppression or release of trace Kr from thesample) have been inhibited efficiently. Though the addition of N2 re-duces the sensitivity of Sr in both solution and LA modes， higher sup-pression of Kr， Rb and polyatomic interferences improves the signal-to-noise ratio and thus increases the accuracy and precision of Sr isotopeanalysis. 4. Application to natural samples According to the investigation of ablation characteristics of the fslaser and the effects of the addition of N2， an improved in situ Sr iso-topes analytical method was established. The feasibility and flexibilityof the proposed method were verified by analyzing various naturalsamples. 4.1. Feldspar crystals Four natural feldspars were analyzed， and the results are shown inTable 2 and Fig. 9. The first three feldspars are plagioclases with dif-ferent An values (CaO/ (2×Na20+2×K20 + CaO) in mol.%) andSr concentrations. The concentrations of Rb in the three feldspars arevery low (<0.67 ugg-1). The measured results of the Sr/86Sr ratiosfor the three feldspars using fs-LA-MC-ICP-MS with N2 =12 ml minwere 0.713724±0.000030 (YG0440： 0.713718 using TIMS)，0.710909±0.000040 (YG0383：0.710919 usingTIMS)and0.703421 ±0.000030 (YG4301： 0.703426 using TIMS). The ac-curacies of the averaged values were from - 0.000014 to 0.000009.The measurement reproducibility， defined as two times of the relativestandard deviation (RSD， k =2) of repeated analyses， was better than0.000042. In addition， the 84sr/86sr ratios were relatively constantamong the three feldspars and similar to the natural ratio of 0.05657(Tong et al.， 2016). These data confirmed the availability of the pro- posed method for the common and low Rb/Sr ratio plagioclase samples. The fourth feldspar sample in Table 2 is a K-feldspar phenocryst(Tuyk)， which has been used as an in-house reference material for in situPb isotope analysis (Zhang et al.，2015).The concentrations of Rb andSr in the Tuyk are 285 ug gand 620 ug g，respectively. The Rb/Srratio is approximately 0.46， which is higher than the critical value of0.2 defined by Davidson et al.(2001) and Jackson and Hart (2006). Inaddition to using fs-laser ablation and N2， we selected three referenceglasses with high Rb/Sr ratios， StHs6/80-G (Rb/Sr ratio =0.06)， T1-G(Rb/Sr ratio = 0.28) and GSE-1G (Rb/Sr ratio =0.80)， to calculate theuser-specified fRb for correcting Rb interference. The average value forSr/86Sr acquired by fs-LA-MC-ICP-MS was 0.710325 ±0.000297 (SD， k=2)， which was significantly higher than the TIMS value(0.710206). The analytical accuracy and reproducibility (RSD， k=2)were 0.000167 and 0.000417， respectively. Obviously， for samples withhigh Rb/Sr ratios， the interference of Rb seriously inhibited the accu-racy and precision of 87Sr/86Sr ratio analysis. However， our methodpresents a significant improvement in the accuracy and reproducibilityfor high Rb/Sr feldspar samples relative to previous studies. For ex-ample， Davidson et al. (2001) reported accuracy and reproducibilitywere 0.0059 and 0.0011， respectively， for plagioclases with a high Rb/Sr ratio (0.52). Jackson and Hart (2006) reported that the basalt glasssamples with high Rb/Sr ratios (0.14) had large errors (0.000505). The above results also demonstrated that the four feldspars， whichhave various concentrations of the major elements， Sr and Rb， hadhomogeneous Sr isotope compositions and thus are suitable as referencematerials for in situ Sr isotope analysis， including feldspar samples withhigh Rb/Sr ratios (< 0.5). Our research interest is to decipher the origin of mafic micro-granular enclaves (MMEs) in granitoids. In past decades， several modelshave been proposed to explain the petrogenesis of these MMEs， in-cluding residues after partial melting， xenoliths of the country rocks，cumulates formed by early crystallization， or mafic magma enclavesgenerated during magma mixing (Davidson et al.， 2007). Commonly，MMEs contain lots of plagioclases， which are sensitive to the geo-chemical variations of the magma chamber. Therefore， tracing the8Sr/86Sr ratios that vary from core to rim of those plagioclases canprovide unique information for understanding the petrogenesis of theMMEs. However， plagioclases in MMEs usually have small grain sizes(200-300 um) with wide ranges of Rb/Sr ratios. Thus， it is a greatchallenge for high-precision Sr isotope measurements. To verify ourexperimental method， we chose MMEs from the Ganze-Daochenggranitic belt in the Yidun arc terrane as a case to study (Wu et al.，2016). Based on careful petrographic observation， two plagioclasegrains from thin sections were analyzed by fs-LA-MC-ICP-MS. Theanalytical results are listed in Fig. 9. As shown in Fig. 9， each plagio-clase has a diameter of approximately 200-400 um. The line scan modewas used with spot size of 45 um， repetition rate of 30 Hz and linescanning speed of 5 um s-1. Both plagioclases have distinct 87Sr/86Sr Table 2Sr isotope ratios for four natural feldspars measured by TIMS and fs-LA-MC-ICP-MS. TIMS (n=3) fs-LA-MC-ICP-MS An% Sr Rb Rb/Sr 87Sr/86Sr 87Sr/86Sr 2SD 84Sr/86Sr 2SD n Hgg 1 YG0440 0.713718±34 0.713724 0.000030 0.05636 0.00017 23 9 390 0.23 0.0006 YG0383 0.710919±18 0.710909 0.000040 0.05646 0.00020 32 10 1511 0.00 0.0000 YG4301 0.703426±24 0.703421 0.000030 0.05642 0.00024 23 51 1058 0.67 0.0006 Tuyk 0.710206±11 0.710325 0.000297 0.05641 0.00015 13 0 620 285 0.46 Fig. 9. Sr isotope ratios for four natural feldspars measured by TIMS and fs-LA-MC-ICP-MS. The error bars represent ± 2 standard errors (SE). The gray shadows represent the referencerange obtained by TIMS. Ave. val.， average value； TIMS， values determined by TIMS. precisionand accuracy..A natural upx megacryst (HNB-8，Sr =89.2 ug g) which has been analyzed for Sr isotope ratios usingTIMS and ns-LA-MC-ICP-MS by Tong et al. (2016)， was selected in thisstudy. Fig. 11 shows the variation of the Sr/86Sr ratios in HNB-8 as afunction of laser frequency and laser fluence. The line scanning modefor a 60 um spot size was used. The measured Sr/8Sr ratios obtainedat low ablation frequencies (20 Hz and 50 Hz) and a low ablation flu-ence (1.17 J cm2) were quite scattered due to the low Sr signal in-tensity. In contrast， at high frequencies (> 100 Hz) and high fluences(> 2.06 J cm-3)， the measured 87Sr/ 86Sr ratios in HNB-8 were con-sistent and agree with the value of TIMS within uncertainty. Comparedto previous research， the reproducibility of the 87Sr/86Sr ratios of HNB-8 in this study was 0.0001 (2SD) and 3 times lower than that of ns laserablation (0.0003， 2SD) (Tong et al.， 2016). However， using the line scanmode in this study produced a 60 ×400 um laser trough with an areaof 24，000 um， which was 2.1 times larger than that produced by nslaser ablation (spot size of 120 um， 11，304 (um). Although this com-parison is not exact， we can understand that the fs laser has a strongflexibility to address different types of samples.) Fig. 10. 87Sr/86Sr ratios of two plagioclases from MMEs determined using fs-LA-MC-ICP-MS combined with the addition of N2 (12 ml min). The line scanning mode was used. Thedetailed laser parameters are listed in each panel. The back linear regions in each panel are laser ablated troughs. Fig. 11. Sr/86Sr ratios of Cpx crystal HNB-8 determined by fs-LA-MC-ICP-MS as a function of laser frequency (a) and laser fluence (b). The data for the Sr/86Sr ratios from lowfrequency and low fluence were not collected to calibrate the average value. The error bars represent ± 2 standard errors (2SE). The black dotted lines represent the solution values (Tonget al.， 2016). The data of ns laser ablation for HNB-8 published by Tong et al. (2016) was plotted. In this study， we first compared the differences in ablation char-acteristics， signal responses and precisions of Sr isotope ratio analysesbetween ns LA and fs LA. The results indicated that the advantages of fsLA were not only the higher ablation rates relative to ns LA but also theelimination of matrix-dependent ablation behavior， producing con-sistent ablation rates and signal responses in a variety of materials. Thenatural transparent minerals that are difficult to ablate by ns laserpulses， such as feldspars， gain the maximum benefit from fs LA. As arough comparison， the Sr responses in feldspars from fs laser ablation at3.8 J cm- were 3.4 times higher than those from the ns laser at4.1 J cm-2，resulting in a significant improvement of the precision of Srisotope analysis， especially for the low-Sr minerals or high spatial re-solution cases. We demonstrated that sensitivities of Sr， Rb and Kr and yields ofpolyatomic interferences (CaAr*/CaCa) and doubly charged Er(167Er++) and Yb (173Yb++) were simultaneously suppressed by theaddition of N2. However， the effect of suppression on Sr was obviouslylower than that on Rb， Kr and polyatomic interferences (CaAr*/CaCa)， resulting in the enhancement of the signal-to-noise ratio. Thisimproved the accuracy of the 84Sr/86Sr ratios， which are easily inter-fered with by Ca polyatomic interferences and Kr and the by externalprecision of the 87Sr/86Sr ratio analysis. In addition， due to the decreasein the Rb/Sr signal ratio， the samples with a Rb/Sr ratio <0.5 can bemeasured within the accuracy of 7Sr/86Sr <0.0002 using a user-specified fRb. Combining the high ablation efficiency of the fs laser with the ca-pacity of suppressing the interference factors by adding N2， the im-proved in situ Sr isotopic analysis method using a 257 nm fs laser wasestablished and employed to analyze four natural feldspars， includingK-feldspars with high Rb/Sr ratios (0.46). The good accuracy and ex-ternal reproducibility of Sr/86Sr for the four feldspars not only con-firmed the availability of the proposed method but also suggested thatthe four natural feldspars have homogeneous Sr isotope compositionsand are suitable candidates for matrix-matched feldspar reference ma-terials. Despite the somewhat lower accuracy and precision of laser ablationanalyses of Sr isotopes compared with the conventional solution mea-surements， especially for the small samples with low-Sr concentrationor high Rb/Sr ratios， LA-MC-ICP-MS can provide useful information onsmall-scale Sr isotopic variations in many geological applications. Inthis study， two plagioclases in mafic microgranular enclaves (MMEs)with small grain sizes (200-300 um) and wide ranges of Rb/Sr ratioswere analyzed and showed obvious variations of the ’Sr/ 86Sr ratiosfrom core to rim. In addition， the high frequency of the fs laser was usedto analyze the natural Cpx megacryst with a low-Sr concentration (HNB-8， Sr=89.2ug g ). The analytical reproducibility of the&Sr/8Sr ratios in HNB-8 analyzed by the fs laser was 3 times lowerthan that of the ns laser.Both cases showed the feasibility and flexibilityof the fs-LA-MC-ICP-MS analysis method established in this study. Acknowledgements This research is supported by the National Key Research andDevelopment Project of China (2016YFC0600309)， the NationalScience Fund for Distinguished Young Scholars (41725013) and theNational Nature Science Foundation of China (Grants 41730211，41603002 and 41573015)， the China Postdoctoral Science Foundation(2015M580677， 2016T90741)， the Science Fund for DistinguishedYoung Scholars of Hubei Province (2016CFA047) and the most specialfund from the State Key Laboratories of Geological Processes andMineral Resources， China University of Geosciences (MSFGPMR04 andMSFGPMR08). Appendix A. Supplementary data Supplementary data to this article can be found online at https：//doi.org/10.1016/j.chemgeo.2017.12.018. References Balter， V.， Telouk， P.， Reynard， B.， Braga，J.， Thackeray，F.， Albarede， F.， 2008. Analysis ofcoupled Sr/Ca and 87Sr/86Sr variations in enamel using laser-ablation tandemquadrupole-multicollector ICPMS. Geochim. Cosmochim. Acta 72 (16)， 3980-3990. Barnettjohnson， R.， Ramos， F.C.， Grimes， C.B.， Macfarlane， R.B.， 2005. Validation of Srisotopes in otoliths by laser ablation multicollector inductively coupled plasma massspectrometry (LA-MC-ICPMS)： opening avenues in fisheries science applications. Can.J. Fish. Aquat. Sci. 62 (11)， 2425-2430. Berglund， M.， Wieser， M.E.， 2011. Isotopic Compositions of the Elements 2009 (IUPACTechnical Report). pp. 1102-1103. Bian， Q.， Garcia， C.C.， Koch， J.，Niemax， K.， 2006. Non-matrix matched calibration ofmajor and minor concentrations of Zn and Cu in brass， aluminium and silicate glassusing NIR femtosecond laser ablation inductively coupled plasma mass spectrometry.J. Anal. At. Spectrom. 21 (2)， 187-191. Bizzarro， M.， Simonetti， A.， Stevenson， R.K.， Kurszlaukis， S.， 2003. In situ 87Sr/86Sr in-vestigation of igneous apatites and carbonates using laser-ablation MC-ICP-MS.Geochim. Cosmochim. Acta 67 (2)， 289-302. Borisova， A.Y.， Freydier， R.， Polve， M.， Salvi， S.， Candaudap， F.， Aigouy， T.， 2008. In situmulti-element analysis of the Mount Pinatubo quartz-hosted melt inclusions by NIRfemtosecond laser ablation-inductively coupled plasma-mass spectrometry.Geostand.Geoanal. Res. 32 (2)， 209-229. Campos-Alvarez，N.O.， Samson， I.M.， Fryer， B.J.， Ames， D.E.， 2010. Fluid sources andhydrothermal architecture of the Sudbury structure： constraints from femtosecondLA-MC-ICP-MS Sr isotopic analysis of hydrothermal epidote and calcite. Chem. Geol.278 (3)， 131-150. ( Ch a rlier ， B. L .A. ， G in ibre ， C . ， M or ga n ， D . ， N o wel l， G . M. ， Pe a r s o n， D.G . ， D a vi d so n， J . P. ， O ttl ey ， C .J.， 2006. M e t hods f o r the m i cr os am pling an d high -p r ec ision analy s i s o f s tron t ium a n d r ubidi u m i s o t o pes a t si n g le c r y s t a l sca l e fo r pe t ro l ogi c a l a n d geo - c h r ono l og i cal a p pl i c a t io ns . C hem. G e ol. 232 ( 3- 4)， 1 1 4-1 3 3. ) ( Ch e n ， K . Y . ， Y uan ， H . L . ， B ao， Z . A . ， Z ong， C.， D a i ， M. N . ， 2014 . P r e c ise a n d ac c ur a t e i n si t u ) determination of lead isotope ratios in NIST， USGS， MPI-DING and CGSG glass re-ference materials using femtosecond laser ablation MC-ICP-MS. Geostand. Geoanal.Res. 38 (1)，5-21. Chen， W.T.， Zhou， M.F.， Gao， J.F.， Zhao， T.P.， 2015. Oscillatory Sr isotopic signature inplagioclase megacrysts from the Damiao anorthosite complex， North China： im-plication for petrogenesis of massif-type anorthosite. Chem. Geol. 393-394，1-15. Chmeleff， J.， Horn， I.， Steinhoefel， G.， von Blanckenburg， F.， 2008. In situ determinationof precise stable Si isotope ratios by UV-femtosecond laser ablation high-resolutionmulti-collector ICP-MS. Chem. Geol. 249 (1-2)， 155-166. Christensen， J.N.， Halliday， A.N.，Lee， D.C.， Hall， C.M.， 1995. In situ Sr isotopic analysisby laser ablation. Earth Planet. Sci. Lett. 136 (1-2)，79-85. Copeland， S.R.， Sponheimer， M.， Lee-Thorp， J.A.， Roux， P.J.L.， Ruiter， D.J.D.， Richards，M.P.， 2010. Strontium isotope ratios in fossil teeth from South Africa： assessing laserablation MC-ICP-MS analysis and the extent of diagenesis. J. Anal. At. Spectrom. 37(7)， 1437-1446. D'Abzac， F.， Seydoux-Guillaume， A.， Chmeleff， J.， Datas， L.，Poitrasson， F.， 2012. In situcharacterization of infrared femtosecond laser ablation in geological samples. Part B：the laser induced particles. J. Anal. At. Spectrom. Vol.27， 108-119. D'Abzac， F.， Beard， B.L.， Czaja， A.D.， Konishi，H.， Schauer， J.J.， Johnson， C.M.， 2013. Ironisotope composition of particles produced by UV-femtosecond laser ablation of nat-ural oxides， sulfides， and carbonates. Anal. Chem. 85， 11885-11892. Davidson， J.， Tepley Iii， F.， Palacz， Z.， Meffan-Main， S.， 2001.Magma recharge， con-tamination and residence times revealed by in situ laser ablation isotopic analysis offeldspar in volcanic rocks. Earth Planet.Sci. Lett. 184(2)， 427-442. Davidson， J.P.， Morgan， D.J.， Charlier， B.L.A.， Harlou， R.， Hora， J.M.， 2007. Microsampling and isotopic analysis of igneous rocks： implications for the study ofmagmatic systems. Annu. Rev. Earth Pl. Sci. 35 (1)，273-311. Durrant， S.F.， 1994. Feasibility of improvement in analytical performance in laser abla-tion inductively coupled plasma-mass spectrometry (LA-ICP-MS) by addition of ni-trogen to the argon plasma. Fresenius J. Anal. Chem. 349 (10-11)， 768-771. Fernandez， B.， Claverie， F.， Pecheyran， C.， Donard，O.F.X.， Claverie， F.， 2007. Directanalysis of solid samples by fs-LA-ICP-MS. TrAC-trend. Anal. Chem. 26 (10)，951-966. Fliegel， D.， Frei， C.， Fontaine， G.H.， Hu， Z.C.， Gao， S.， Gunther， D.， 2011. Sensitivity im-provement in laser ablation inductively coupled plasma mass spectrometry achievedusing a methane/argon and methanol/water/argon mixed gas plasma. Analyst 136(23)，4925-4934. Freydier， R.， Candaudap， F.， Poitrasson， F.， Arbouet， A.， Chatel， B.， Dupre， B.， 2008.Evaluation of infrared femtosecond laser ablation for the analysis of geomaterials byICP-MS. J. Anal. At.Spectrom. 23 (5)， 702-710. Fu， J.L.， Hu， Z.C.， Zhang， W.， Yang， L.， Liu， Y.S.， Li， M.， Zong， K.Q.，Gao， S.， Hu， S.H.，2016. In situ sulfur isotopes (834S and 833S) analyses in sulfides and elemental sulfurusing high sensitivity cones combined with the addition of nitrogen by laser ablationMC-ICP-MS. Anal. Chim. Acta 911， 14-26. Gagnevin， D.， Daly， J.S.， Poli， G.， Morgan， D.， 2005. Microchemical and Sr isotopic in-vestigation of zoned K-feldspar megacrysts： insights into the petrogenesis ofa graniticsystem and disequilibrium crystal growth. J. Petrol. 46 (8)， 1689-1724. Galler， P.， Limbeck， A.，Boulyga， S.F.， Stingeder， G.， Hirata， T.， Prohaska， T.， 2007.Development of an on-line flow injection Sr/matrix separation method for accurate，high-throughput determination of Sr isotope ratios by multiple collector-inductivelycoupled plasma-mass spectrometry. Anal. Chem. 79 (13)， 5023-5029. Guillong， M.， Heinrich， C.A.， 2007. Sensitivity enhancement in laser ablation ICP-MSusing small amounts of hydrogen in the carrier gas. J. Anal. At. Spectrom. 22 (12)，1488-1494. Giinther， D.， Jackson， S.E.， Longerich， H.P.， 1999. Laser ablation and arc/spark solidsample introduction into inductively coupled plasma mass spectrometers.Spectrochim. Acta B At. Spectrosc. 54 (3)， 381-409. He，D.T.， Liu， Y.S.， Tong， X.R.， Zong， K.Q.， Hu， Z.C.， Gao， S.， 2013. Multiple exsolutions ina rare clinopyroxene megacryst from the Hannuoba basalt， North China： implicationsfor subducted slab-related crustal thickening and recycling. Lithos 177 (3)， 136-147. Hergenroder， R.， Samek， O.， Hommes， V.， 2006. Femtosecond laser ablation elementalmass spectrometry. Mass Spectrom. Rev. 25 (4)， 551-572. Hirata， T.， Kon， Y.， 2008. Evaluation of the analytical capability of NIR femtosecond laserablation-inductively coupled plasma mass spectrometry. Anal. Sci. 24 (3)， 345-353. Horn， I.， von Blanckenburg， F.， 2007. Investigation on elemental and isotopic fractiona-tion during 196 nm femtosecond laser ablation multiple collector inductively coupledplasma mass spectrometry. Spectrochim. Acta B At. Spectrosc. 62 (4)， 410-422. Horn， I.， von Blanckenburg， F.， Schoenberg， R.， Steinhoefel， G.， Markl， G.， 2006. In situiron isotope ratio determination using UV-femtosecond laser ablation with applica-tion to hydrothermal ore formation processes. Geochim. Cosmochim.Acta 70 (14)，3677-3688. Horstwood， M.S.A.， Evans， J.A.， Montgomery， J.， 2008. Determination of Sr isotopes incalcium phosphates using laser ablation inductively coupled plasma mass spectro-metry and their application to archaeological tooth enamel. Geochim. Cosmochim.Acta 72(23)，5659-5674. Hu， Z.C.， Gao， S.， Liu， Y.S.， Hu， S.H.， Chen， H.H.， Yuan， H.L.， 2008. Signal enhancement inlaser ablation ICP-MS by addition of nitrogen in the central channel gas. J. Anal. At.Spectrom. 23 (8)， 1093-1101.cCl Hu， Z.C.， Liu， Y.S.， Gao， S.， Liu， W.G.， Zhang， W.， Tong， X.R.， Lin， L.， Zong， K.Q.， Li， M.，Chen， H.H.， Zhou， L.， Yang， L.， 2012a. Improved in situ Hf isotope ratio analysis ofzircon using newly designed X skimmer cone and jet sample cone in combinationwith the addition of nitrogen by laser ablation multiple collector ICP-MS. J. Anal. At.Spectrom. 27(9)， 1391-1399. Hu， Z.C.， Liu， Y.S.， Gao， S.， Xiao， S.Q.， Zhao， L.S.， Gunther， D.， Li， M.， Zhang， W.， Zong，K.Q.， 2012b.A“wire”signal smoothing device for laser ablation inductively coupledplasma mass spectrometry analysis. Spectrochim. Acta B At. Spectrosc.78 (0)， 50-57. Ikehata， K.， Hirata， T.， 2013. Evaluation of UV-fs-LA-MC-ICP-MS for precise in situ copperisotopic microanalysis of cubanite. Anal. Sci. 29 (12)， 1213-1217. Ikehata， K.， Notsu， K.，Hirata， T.， 2008. In situ determination of Cu isotope ratios incopper-rich materials by NIR femtosecond LA-MC-ICP-MS.J. Anal. At. Spectrom. 23(7)，1003-1008. Jackson， M.G.， Hart， S.R.， 2006. Strontium isotopes in melt inclusions from Samoan ba-salts： implications for heterogeneity in the Samoan plume. Earth Planet. Sci. Lett. 245(1-2)，260-277. Jochum， K.P.， Stoll， B.， Weis， U.， Kuzmin， D.V.， Sobolev， A.V.， 2009. In situ Sr isotopicanalysis of low Sr silicates using LA-ICP-MS. J. Anal. At. Spectrom. 24 (9)，1237-1243. Jochum， K.P.， Stoll， B.， Weis， U.，Jacob， D.E.， Mertz-Kraus， R.， Andreae， M.O.， 2014. Non-matrix-matched calibration for the multi-element analysis of geological and en-vironmental samples using 200 nm femtosecond LA-ICP-MS： a comparison with na-nosecond lasers. Geostand. Geoanal. Res. 38 (3)， 265-292. Kimura， J.， Takahashi， T.， Chang， Q.， 2013. A new analytical bias correction for in situ Srisotope analysis of plagioclase crystals using laser-ablation multiple-collector in-ductively coupled plasma mass spectrometry. J. Anal. At. Spectrom. 28 (6)， 945-957. Kimura， J.， Chang， Q.， Itano， K.， Iizuka， T.， Vaglarov， B.S.， Tani， K.， 2014. An improved U-Pb age dating method for zircon and monazite using 200/266 nm femtosecond laserablation and enhanced sensitivity multiple-Faraday collector inductively-coupledplasma mass spectrometry. J. Anal. At. Spectrom. 30 (2)， 494-505. Koch， J.， Walle， M.， Pisonero， J.， Giinther， D.， 2006. Performance characteristics of ultra-violet femtosecond laser ablation inductively coupled plasma mass spectrometry at265 and 200 nm. J. Anal. At. Spectrom. 21 (9)， 932-940. Koornneef， J.M.， Nikogosian， I.， van Bergen，M.J.， Smeets， R.， Bouman， C.， Davies， G.R.，2015.TIMS analysis of Sr and Nd isotopes in melt inclusions from Italian potassium-rich lavas using prototype 1013 Q amplifiers. Chem. Geol. 397， 14-23. Li， Z.， Hu， Z.C.， Liu， Y.S.， Gao， S.， Li，M.， Zong， K.Q.， Chen， H.H.， Hu， S.H.， 2015. Accuratedetermination of elements in silicate glass by nanosecond and femtosecond laserablation ICP-MS at high spatial resolution. Chem. Geol. 400， 11-23. Li， Z.， Hu， Z.C.， Giinther， D.， Zong， K.Q.， Liu， Y.S.， Luo， T.， Zhang， W.， Gao， S.， Hu， S.H.，2016. Ablation characteristic of ilmenite using UV nanosecond and femtosecond la-sers： implications for non-matrix-matched quantification. Geostand. Geoanal. Res. 40(4)， 477-491. Lin， L.， Hu， Z.C.， Yang， L.，Zhang， W.， Liu， Y.S.， Gao， S.， Hu， S.H.， 2014. Determination ofboron isotope compositions of geological materials by laser ablation MC-ICP-MSusing newly designed high sensitivity skimmer and sample cones. Chem. Geol. 386(0)， 22-30. von der Linde， D.，Sokolowski-Tinten， K.， Bialkowski， J.， 1997. Laser-solid interaction inthe femtosecond time regime. Appl. Surf. Sci. 109-110， 1-10. Liu， X.， Du， D.， Mourou， G.， 1997. Laser ablation and micromachining with ultrashortlaser pulses. IEEE J. Quantum Elect. 33 (10)，1706-1716. Liu， S.H.， Hu， Z.C.， Gunther， D.， Ye， Y.H.， Liu， Y.S.， Gao， S.， Hu， S.H.， 2014. Signal en-hancement in laser ablation inductively coupled plasma-mass spectrometry usingwater and/or ethanol vapor in combination with a shielded torch. J. Anal. At.Spectrom. 29 (3)， 536-544. Mao， S.S.， Quere， F.， Guizard， S.， Mao， X.， Russo， R.E.， Petite， G.， Martin， P.， 2004.Dynamics of femtosecond laser interactions with dielectrics. Appl. Phys.A-Mater 79(7)，1695-1709. Oeser，M.， Weyer， S.， Horn， I.， Schuth，S.， 2014. High-precision Fe and Mg isotope ratios ofsilicate reference glasses determined in situ by femtosecond LA-MC-ICP-MS and bysolution nebulisation MC-ICP-MS. Geostand. Geoanal. Res. 38 (3)， 311-328. Ohata， M.， Nonose， N.， Dorta， L.， Giinther， D.， 2015. Comparison of 265 nm femtosecondand 213 nm nanosecond laser ablation inductively coupled plasma mass spectro-metry for Pb isotope ratio measurements. Anal. Sci. 31 (12)， 1309-1315. Outridge， P.， Chenery，S.， Baba3lauk， J.， Reist， J.， 20pa02. Analysis of geological Sr isotopemarkers in fish otoliths with subannual resolution using laser ablation-multicollector-ICP-mass spectrometry. Environ. Geol. 42 (8)， 891-899. Pisonero， J.， Giinther， D.， 2008. Femtosecond laser ablation inductively coupled plasmamass spectrometry： fundamentals and capabilities for depth profiling analysis. MassSpectrom. Rev. 27 (6)，609-623. Poitrasson， F.， Mao， X.， Mao， S.S.， Freydier， R.， Russo， R.E.，2003. Comparison of ultra-violet femtosecond and nanosecond laser ablation inductively coupled plasma massspectrometry analysis in glass， monazite， and zircon. Anal. Chem. 75 (22)，6184-6190. Ramos， F.C.， Tepley， F.J.， 2008. Inter- and intracrystalline isotopic disequilibria： techni-ques and applications. Rev. Mineral. Geochem. 69 (1)， 403-443. Ramos， F.C.， Wolff， J.A.， Tollstrup， D.L.， 2004. Measuring 87Sr/86Sr variations in mi-nerals and groundmass from basalts using LA-MC-ICPMS. Chem. Geol. 211 (1-2)，135-158. Russell， W.A.， Papanastassiou， D.A.， Tombrello， T.A.， 1978. Ca isotope fractionation onthe earth and other solar system materials. Geochim. Cosmochim. Acta 42 (8)，1075-1090. Russo， R.E.， Mao， X.L.， Gonzalez， J.J.， Mao， S.S.， 2002. Femtosecond laser ablation ICP-MS. J. Anal. At. Spectrom. 17 (17)， 1072-1075. Schmidberger， S.S.， Simonetti， A.， Francis， D.， 2003. Small-scale Sr isotope investigationof clinopyroxenes from peridotite xenoliths by laser ablation MC-ICP-MS-implications for mantle metasomatism. Chem. Geol. 199 (3-4)，317-329. Schuessler， J.A.， von Blanckenburg， F.， 2014. Testing the limits of micro-scale analyses ofSi stable isotopes by femtosecond laser ablation multicollector inductively coupledplasma mass spectrometry with application to rock weathering.Spectrochim. Acta BAt. Spectrosc. 98， 1-18. ( S h a h een ， M . ， Fr y e r， B . J . ， 2 0 10. I mp rov i ng t he ana ly tica l ca pab i li t ies o f fem t o s eco nd l as e r a blation m ul tico ll ect o r I CP - M S fo r hi g h p r e c isi o n P b iso t o p ic an a l y si s： th e r o l e of hy dro g en a n d ni tr o g en . J . An al . At . S p ect r o m . 2 5 ( 7) ， 1 0 06-1 0 1 3. ) Shaheen， M.E.， Gagnon， J.E.， Fryer， B.J.， 2012. Femtosecond (fs) lasers coupled withmodern ICP-MS instruments provide new and improved potential for in situ ele-mental and isotopic analyses in the geosciences. Chem. Geol. 330， 260-273. Steinhoefel， G.， Horn， I.， von Blanckenburg， F.， 2009a. Matrix-independent Fe isotoperatio determination in silicates using UV femtosecond laser ablation. Chem. Geol. 268(1-2)， 67-73. Steinhoefel， G.， Horn， I.， von Blanckenburg， F.， 2009b. Micro-scale tracing of Fe and Siisotope signatures in banded iron formation using femtosecond laser ablation.Geochim. Cosmochim. Acta 73 (18)， 5343-5360. ( S t e inh o e f e l ， G . ， B re u er ， J .， von B lanck e nb u rg， F. ， H o r n ， I . ， K a c z o r ek， D. ， Som mer， M . ， 2 01 1 . M ic ro mete r s i li c o n iso t o p e d i a g no s tic s of so i l s b y UV fe mt o second l aser a bl a t i on . C hem . G eol . 286( 3 -4)，280- 2 89. ) Tong， X.R.， Liu， Y.S.， Hu， Z.C.， Chen， H.H.， Zhou， L.， Hu， Q.H.， Xu， R.， Deng， L.X.， Chen，C.F.， Yang， L.， Gao， S.， 2016. Accurate determination of Sr isotopic compositions inclinopyroxene and silicate glasses by LA-MC-ICP-MS. Geostand. Geoanal. Res.40 (1)，85-99. Vroon， P.Z.， Wagt， B.V.D.， Koornneef， J.M.， Davies， G.R.， 2008. Problems in obtainingprecise and accurate Sr isotope analysis from geological materials using laser ablationMC-ICPMS. Anal. Bioanal. Chem. 390 (2)， 65-76. Waight， T.， Baker， J.， Peate， D.， 2002. Sr isotope ratio measurements by double-focusingMC-ICPMS： techniques， observations and pitfalls. Int. J. Mass Spectrom. 221 (3)，229-244. Woodhead， J.， Swearer， S.， Hergt， J.， Maas， R.， 2005. In situ Sr-isotope analysis of car-bonates by LA-MC-ICP-MS： interference corrections， high spatial resolution and anexample from otolith studies. J. Anal. At. Spectrom. 20 (1)， 22-27. Wu， T.， Xiao， L.， Wilde， S.A.， Ma， C.Q.， Li， Z.L.， Sun， Y.， Zhan， Q.Y.， 2016. Zircon U-Pb ageand Sr-Nd-Hf isotope geochemistry of the Ganluogou dioritic complex in thenorthern Triassic Yidun arc belt， Eastern Tibetan Plateau： implications for the closureof the Garze-Litang Ocean. Lithos 248-251， 94-108. Xu， L.， Hu， Z.C.， Zhang， W.， Yang， L.， Liu， Y.S.， Gao， S.， Luo， T.， Hu， S.H， 2015. In situ Ndisotope analyses in geological materials with signal enhancement and non-linearmass dependent fractionation reduction using laser ablation MC-ICP-MS. J. Anal. At.Spectrom. 30(1)， 232-244. Yang， Y.H.， Wu， F.Y.， Wilde， S.A.， Liu， X.M.， Zhang， Y.B.， Xie， L.W.， Yang， J.H.， 2009. Insitu perovskite Sr-Nd isotopic constraints on the petrogenesis of the OrdovicianMengyin kimberlites in the North China Craton. Chem. Geol. 264 (1-4)， 24-42. Yang， Y.H.，Zhang， H.F.， Chu， Z.Y.， Xie，L.W.， Wu， F.Y.， 2010. Combined chemical se-paration of Lu， Hf， Rb， Sr， Sm and Nd from a single rock digest and precise andaccurate isotope determinations of Lu-Hf， Rb-Sr and Sm-Nd isotope systems usingmulti-collector ICP-MS and TIMS. Int. J. Mass Spectrom. 290(2-3)， 120-126. Yang， Y.H.， Wu， F.Y.， Xie， L.W.， Yang， J.H.， Zhang， Y.B.， 2011a. High-precision directdetermination of the 87Sr/86 Sr isotope ratio of bottled Sr-rich natural mineraldrinking water using multiple collector inductively coupled plasma mass spectro-metry. Spectrochim. Acta B At. Spectrosc. 66 (8)， 656-660. Yang， Z.P.， Fryer， B.J.， Longerich， H.P.， Gagnon， J.E.， Samson， I.M.， 2011b. 785 nmfemtosecond laser ablation for improved precision and reduction of interferences inSr isotope analyses using MC-ICP-MS. J. Anal. At. Spectrom. 26(2)， 341-351. Yang， Y.H.， Wu， F.Y.， Xie， L.W.， Chu， Z.Y.， Yang， J.H.， 2014. Re-evaluation of inter-ferences of doubly charged ions of heavy rare earth elements on Sr isotopic analysisusing multi-collector inductively coupled plasma mass spectrometry. Spectrochim.Acta B At. Spectrosc. 97，118-123. Zhang， W.， Hu， Z.， Yang， L.， Liu， Y.， Zong， K.， Xu， H.， Chen， H.， Gao， S.， Xu， L.， 2015.Improved inter-calibration of faraday cup and ion counting for in situ Pb isotopemeasurements using LA-MC-ICP-MS： application to the study of the origin of theFangshan pluton， North China. Geostand. Geoanal. Res. 39 (4)， 467-487. Zhang， W.， Hu， Z.， Giinther， D.， Liu， Y.， Ling， W.， Zong， K.， Chen， H.， Gao， S.， 2016. Directlead isotope analysis in Hg-rich sulfides by LA-MC-ICP-MS with a gas exchange deviceand matrix-matched calibration. Anal. Chim. Acta 948， 9-18.    激光剥蚀-多接收电感耦合等离子体质谱法（LA-MC-ICP-MS）对地质矿物的n位Sr同位素分析对岩浆源组成和地质过程来说的是一种强大的追踪技术。然而，由于Sr浓度低、同重元素或复杂结构小颗粒干扰，因此在对天然矿物特别是对长石等透明矿物进行分析时87Sr/86Sr比值的准确度和精密度不能令人满意。在这项研究的分析结果表明，飞秒激光对各种样品的剥蚀率（每个脉冲0.08 -0.11μm）是一致的。但是使用纳秒激光剥蚀效率受地质材料影响相当明显，例如长石和黄铁矿剥蚀率分别为每个脉冲0.144μm和0.026μm。此外，由于飞秒激光的剥蚀效率较高，在相同的能量下分析长石中的Sr飞秒激光灵敏度是纳秒激光敏度的3.4倍。飞秒激光的这些优点不仅有利于消除激光剥蚀过程中的基体效应，而且有助于提高透明矿物的分析准确度。我们还证明了在6 - 12mLmin-1 N2条件下，同重元素钙二聚体（CaAr++CaCa+）和Kr+的干扰值分别降低了6.5-11.7和5-12.5。此外，随着N2 （12 mLmin-1）的加入，铷的灵敏度受到抑制，Rb/Sr信号比下降1.47倍。由于加入N2的抑制作用，尤其是对富含铷的长石87Sr/86Sr和84Sr/86Sr比值的准确度和精密度均有提高。结合飞秒激光系统的优点和氮气的加入，改进了原位微区Sr同位素的分析方法。对天然斜长石、高Rb/Sr（0.46）的K-长石和低Sr的斜长石进行分析，87Sr/86Sr比值的准确度和精密度结果令人满意，验证了该方法的可靠性。主要元素Sr和Rb含量不同的四种长石具有均匀的Sr同位素组成，因此可以推荐作为原位微区Sr同位素分析合适的参考材料。本文提出的方法可以为单一矿物提供高空间分辨率的地球化学信息。