Secondary Ion Mass Spectrometry Bias on Isotope Ratios in Dolomite–Ankerite,Part II: δ13C Matrix Effects

This study is Part II of a series that documents the development of a suite of calibration reference materials for in situ SIMS analysis of stable isotope ratios in Ca‐Mg‐Fe carbonates. Part I explored the effects of Fe²⁺ substitution on SIMS δ¹⁸O bias measured from the dolomite–ankerite solid solution series [CaMg(CO₃)₂–CaFe(CO₃)₂], whereas this complementary work explores the compositional dependence of SIMS δ¹³C bias (calibrated range: Fe# = 0.004–0.789, where Fe# = molar Fe/(Mg+Fe)). Under routine operating conditions for carbonate δ¹³C analysis at WiscSIMS (CAMECA IMS 1280), the magnitude of instrumental bias increased exponentially by 2.5–5.5‰ (session‐specific) with increasing Fe‐content in the dolomite structure, but appeared insensitive to minor Mn substitution [< 2.6 mole % Mn/(Ca+Mg+Fe+Mn)]. The compositional dependence of bias (i.e., the matrix effect) was expressed using the Hill equation, yielding calibration residual values ≤ 0.3‰ relative to CRM NBS‐19 for eleven carbonate reference materials (6‐μm‐diameter spot size measurements). Based on the spot‐to‐spot repeatability of a drift monitor material that ‘bracketed’ each set of ten sample‐spot analyses, the analytical precision was ± 0.6–1.2‰ (2s, standard deviations). The analytical uncertainty for individual sample analyses was approximated by combining the precision and calibration residual values (propagated in quadrature), suggesting an uncertainty of ± 1.0–1.5‰ (2s).     

USGS46 Greenland Ice Core Water – A New Isotopic Reference Material for δ2H and δ18O Measurements of Water

Ice core from Greenland was melted, filtered, homogenised, loaded into glass ampoules, sealed, autoclaved to eliminate biological activity, and calibrated by dual‐inlet isotope‐ratio mass spectrometry. This isotopic reference material (RM), USGS46, is intended as one of two secondary isotopic reference waters for daily normalisation of stable hydrogen (δ²H) and stable oxygen (δ¹⁸O) isotopic analysis of water with a mass spectrometer or a laser absorption spectrometer. The measured δ²H and δ¹⁸O values of this reference water were ?235.8 ± 0.7‰ and ?29.80 ± 0.03‰, respectively, relative to VSMOW on scales normalised such that the δ²H and δ¹⁸O values of SLAP reference water are, respectively, ?428 and ?55.5‰. Each uncertainty is an estimated expanded uncertainty (U = 2u_c) about the reference value that provides an interval that has about a 95‐percent probability of encompassing the true value. This reference water is available in cases containing 144 glass ampoules that are filled with either 4 ml or 5 ml of water per ampoule.     

SIMS Bias on Isotope Ratios in Ca‐Mg‐Fe Carbonates (Part III): δ18O and δ13C Matrix Effects Along the Magnesite–Siderite Solid‐Solution Series

This study explores the effects of cation composition on mass bias (i.e., the matrix effect), which is a major component of instrumental mass fractionation (IMF) in the microanalyses of δ¹³C and δ¹⁸O by SIMS in carbonates of the magnesite–siderite solid‐solution series (MgCO₃–FeCO₃). A suite of twelve calibration reference materials (RMs) was developed and documented (calibrated range: Fe# = 0.002–0.997, where Fe# = molar Fe/[Mg + Fe]), along with empirical expressions for regressing calibration data (affording residuals < 0.5‰ relative to certified reference material NIST‐19). The calibration curves of both isotope systems are non‐linear and have, over a 2‐year period, fallen into one of two distinct but largely self‐consistent shape categories (data from ten measurement sessions), despite adherence to well‐established analytical protocols for carbonate δ¹³C and δ¹⁸O analyses at WiscSIMS (CAMECA IMS 1280). Mass bias was consistently most sensitive to changes in composition near the magnesite end‐member (Fe# 0–0.2), deviating by up to 4.5‰ (δ¹³C) and 14‰ (δ¹⁸O) with increasing Fe content. The cause of variability in calibration curve shapes is not well understood at present and demonstrates the importance of having available a sufficient number of well‐characterised RMs so that potential complexities of curvature can be adequately delineated and accounted for on a session‐by‐session basis.     

Matrix Corrections and Error Analysis in High‐Precision SIMS 18O/16O Measurements of Ca–Mg–Fe Garnet

We report technical and data treatment methods for making accurate, high‐precision measurements of ¹⁸O/¹⁶O in Ca–Mg–Fe garnet utilising the Cameca IMS 1280 multi‐collector ion microprobe. Matrix effects were similar to those shown by previous work, whereby Ca abundance is correlated with instrumental mass fractionation (IMF). After correction for this effect, there appeared to be no significant secondary effect associated with Mg/Fe²⁺ for routine operational conditions. In contrast, investigation of the IMF associated with Mn‐ or Cr‐rich garnet showed that these substitutions are significant and require a more complex calibration scheme. The Ca‐related calibration applied to low‐Cr, low‐Mn garnet was reproducible across different sample mounts and under a range of instrument settings and therefore should be applicable to similar instruments of this type. The repeatability of the measurements was often better than ± 0.2‰ (2s), a precision that is similar to the repeatability of bulk techniques. At this precision, the uncertainties due to spot‐to‐spot repeatability were at the same magnitude as those associated with matrix corrections (± 0.1–0.3‰) and the uncertainties in reference materials (± 0.1–0.2‰). Therefore, it is necessary to accurately estimate and propagate uncertainties associated with these parameters – in some cases, uncertainties in reference materials or matrix corrections dominate the uncertainty budget.     

Zircon M127 – A Homogeneous Reference Material for SIMS U–Pb Geochronology Combined with Hafnium,Oxygen and,Potentially, Lithium Isotope Analysis

In this article, we document a detailed analytical characterisation of zircon M127, a homogeneous 12.7 carat gemstone from Ratnapura, Sri Lanka. Zircon M127 has TIMS‐determined mean U–Pb radiogenic isotopic ratios of 0.084743 ± 0.000027 for ²⁰⁶Pb/²³⁸U and 0.67676 ± 0.00023 for ²⁰⁷Pb/²³⁵U (weighted means, 2s uncertainties). Its ²⁰⁶Pb/²³⁸U age of 524.36 ± 0.16 Ma (95% confidence uncertainty) is concordant within the uncertainties of decay constants. The δ¹⁸O value (determined by laser fluorination) is 8.26 ± 0.06‰ VSMOW (2s), and the mean ¹⁷⁶Hf/¹⁷⁷Hf ratio (determined by solution ICP‐MS) is 0.282396 ± 0.000004 (2s). The SIMS‐determined δ⁷Li value is ?0.6 ± 0.9‰ (2s), with a mean mass fraction of 1.0 ± 0.1 μg g^?1 Li (2s). Zircon M127 contains ~ 923 μg g^?1 U. The moderate degree of radiation damage corresponds well with the time‐integrated self‐irradiation dose of 1.82 × 10¹⁸ alpha events per gram. This observation, and the (U–Th)/He age of 426 ± 7 Ma (2s), which is typical of unheated Sri Lankan zircon, enable us to exclude any thermal treatment. Zircon M127 is proposed as a reference material for the determination of zircon U–Pb ages by means of SIMS in combination with hafnium and stable isotope (oxygen and potentially also lithium) determination.     

Ion Implants as Matrix‐Appropriate Calibrators for Geochemical Ion Probe Analyses

Ion microprobe elemental and isotopic determinations can be precise but difficult to quantify. Error is introduced when the reference material and the sample to be analysed have different compositions. Mitigation of such ‘matrix effects’ is possible using ion implants. If a compositionally homogeneous reference material is available which is ‘matrix‐appropriate’ (i.e., close in major element composition to the sample to be analysed, but having an unknown concentration of the element, E, to be determined) then ion implantation can be used to introduce a known amount of an E isotope, calibrating the E concentration and producing a matrix‐appropriate calibrator. Nominal implant fluences (ions cm^?2) are inaccurate by amounts up to approximately 30%. However, ion implantation gives uniform fluences over large areas; thus, it is possible to ‘co‐implant’ an additional reference material of any bulk composition having known amounts of E, independently calibrating the implant fluence. Isotope ratio measurement standards can be produced by implanting two different isotopes, but permil level precision requires postimplant calibration of the implant isotopic ratio. Examples discussed include (a) standardising Li in melilite; (b) calibrating a ²⁵Mg implant fluence using NIST SRM 617 glass and (c) using Si co‐implanted with ²⁵Mg alongside NIST SRM 617 to produce a calibrated measurement of Mg in Si.     

Development and Re‐Evaluation of Tourmaline Reference Materials for In Situ Measurement of Boron δ Values by Secondary Ion Mass Spectrometry

Six tourmaline samples were investigated as potential reference materials (RMs) for boron isotope measurement by secondary ion mass spectrometry (SIMS). The tourmaline samples are chemically homogeneous and cover a compositional range of tourmaline supergroup minerals (primarily Fe, Mg and Li end‐members). Additionally, they have homogeneous boron delta values with intermediate precision values during SIMS analyses of less than 0.6‰ (2s). These samples were compared with four established tourmaline RMs, that is, schorl IAEA‐B‐4 and three Harvard tourmalines (schorl HS#112566, dravite HS#108796 and elbaite HS#98144). They were re‐evaluated for their major element and boron delta values using the same measurement procedure as the new tourmaline samples investigated. A discrepancy of about 1.5‰ in δ¹¹B was found between the previously published reference values for established RMs and the values determined in this study. Significant instrumental mass fractionation (IMF) of up to 8‰ in δ¹¹B was observed for schorl–dravite–elbaite solid solutions during SIMS analysis. Using the new reference values determined in this study, the IMF of the ten tourmaline samples can be modelled by a linear combination of the chemical parameters FeO + MnO, SiO₂ and F. The new tourmaline RMs, together with the four established RMs, extend the boron isotope analysis of tourmaline towards the Mg‐ and Al‐rich compositional range. Consequently, the in situ boron isotope ratio of many natural tourmalines can now be determined with an uncertainty of less than 0.8‰ (2s).     

Simultaneous Determination of Fluorine,Chlorine, Bromine and Iodine in Six Geochemical Reference Materials Using Pyrohydrolysis,Ion Chromatography and Inductively Coupled Plasma‐Mass Spectrometry

Concentrations of halogens (fluorine, chlorine, bromine and iodine) were determined in six geochemical reference materials (BHVO‐2, GS‐N, JG‐1, JR‐1, JB‐1b, JB‐2). Halogens were first extracted from powdered samples using a pyrohydrolysis technique, then hydrolysis solutions were analysed by ion chromatography for F and Cl and inductively coupled plasma‐mass spectrometry for Br and I. The detection limits in solutions were 100 μg l^?1 for both F and Cl and 10 ng l^?1 for Br and I. Considering the extraction procedure, performed on a maximum of 500 mg of sample and producing 100 ml of pyrohydrolysis solution, detection limits in rock samples were 20 mg kg^?1 for F and Cl and 2 μg kg^?1 for Br and I. The mean analytical errors on the studied composition ranges were estimated at 10 mg kg^?1 for F and Cl, 100 μg kg^?1 for Br and 25 μg kg^?1 for I. The concentration values, based on repeated (generally > 10) sample analysis, were in good agreement generally with published values and narrowed the mean dispersion around mean values. Large dispersions are discussed in terms of samples heterogeneity and contaminations during sample preparation. Basaltic RMs were found to be more suitable for studies of halogen compositions than differentiated rock material, especially granites – the powders of which were heterogeneous in halogens at the 500 mg level.     

A Case Study of Spectral and Non‐Spectral Interferences on Copper Isotope Measurements by Multi‐Collector ICP‐MS (Wet Plasma)

Mathematical modelling was combined with experimental Cu isotope measurements to demonstrate the effect of the sample matrix in changing the absolute and relative abundances of spectral interferences from Ti and Cr species. This unforeseen non‐spectral effect, evidenced by variable inaccuracies of the different Zn‐normalised Cu isotope ratios, was investigated by comparing real sedimentary samples and artificial solutions intended to match the Cu:Ti:Cr ratios of the real samples after (one or two step) chromatographic processing. Artificial solutions showed positive bias in δ⁶⁵Cu_X/Y with the magnitude depending on (a) the ^6XZn/^6YZn ratio used for normalisation, (b) the Ti/Cu ratio and (c) the transmission coefficient of the TiO species. In contrast, real samples showed different δ⁶⁵Cu_X/Y patterns and displayed a more complex population of Ti and Cr oxides and hydroxides, giving rise to positive and negative inaccuracies that were two to six times higher compared with the artificial samples. The results evidence contrasting behaviour of Ti and Cr when forming polyatomic species in the plasma and stress that artificial solutions may fail to predict how residual elements interact with the analyte/dopant pair during MC‐ICP‐MS analyses. More importantly, the study shows that all Zn isotope ratios do not have the same merit in correcting for mass bias in the presence of matrix elements and should all be monitored to verify the absence of spectral interferences for Cu isotope measurements. In this respect, accurate Cu data could be generally obtained by a two‐step chromatographic purification providing a minimum reduction of ~ 21000 and ~ 3000 times the initial amounts of Ti and Cr, respectively.     

Multiple Sulfur Isotope Determination by SIMS: Evaluation of Reference Sulfides for Δ33S with Observations and a Case Study on the Determination of Δ36S

High spatial resolution multiple sulfur isotope studies undertaken by multi‐collector secondary ion mass spectrometry (SIMS) commonly use well‐characterised sulfide reference materials that do not (or are assumed not to) exhibit mass‐independent fractionation in ³³S and ³⁶S, taking advantage of the three‐isotope plot to evaluate the extent of such fractionation in unknown targets. As a result, few studies to date have used a mass independently fractionated reference sulfide to demonstrate accuracy of measurement and/or data reduction procedures. This article evaluates two mass independently fractionated sulfides, a pyrite from the 3.7 Ga Isua greenstone belt and a pyrrhotite from a 2.7 Ga gold deposit in Minas Gerais, Brazil, which may be used to provide additional confidence in the obtained multiple sulfur isotope data. Additionally, the article presents a method for measuring quadruple sulfur isotopes by SIMS at a comparable spatial and volume resolution to that typically employed for triple sulfur isotopes. This method has been applied to the Isua pyrite as well as to a sample of 2.5 Ga pyrite from the Campbellrand, Transvaal, South Africa, previously investigated using SIMS for triple sulfur isotopes, illustrating its potential for quadruple sulfur investigations.     

In Situ Determination of Au and Cu in Natural Pyrite by Near‐Infrared Femtosecond Laser Ablation‐Inductively Coupled Plasma‐Quadrupole Mass Spectrometry: No Evidence for Matrix Effects

Gold and copper concentrations were determined in natural pyrite by near‐infrared femtosecond LA‐ICP‐QMS, using both sulfide reference materials (pyrrhotite Po‐726 and in‐house natural chalcopyrite Cpy‐RM) and NIST SRM 610 as external calibrators. Firstly, using NIST SRM 610 as the external calibrator, we calculated the Au concentration in Po‐726 and the Cu concentration in Cpy‐RM. The calculated concentration averages for Au and Cu were similar to the values published for Po‐726 and Cpy‐RM, respectively. Secondly, we calculated Au and Cu concentrations taking NIST SRM 610 as an unknown sample and using Po‐726 and Cpy‐RM as external calibrators. Again, the average values obtained closely reflected the preferred concentrations for NIST SRM 610. Finally, we calculated Au and Cu concentrations in natural pyrite using sulfide and silicate reference materials as external calibrators. In both cases, calculated concentrations were very similar, independent of the external calibrator used. The aforementioned data, plus the fact that we obtained very small differences in relative sensitivity values (percentage differences are between 5% and 17% for ⁵⁷Fe, ⁶³Cu and ¹⁹⁷Au) on analyses of silicate and sulfide RMs, indicate that there were no matrix effects related to the differences in material composition. Thus, it is possible to determine Au and Cu in natural sulfides using NIST silicate glasses as an external calibrator.