Secondary Ion Mass Spectrometry Bias on Isotope Ratios in Dolomite–Ankerite, Part I: δ¹⁸O Matrix Effects

Abstract

We document the development of a suite of carbonate mineral reference materials for calibrating SIMS determinations of δ18O in samples with compositions along the dolomite–ankerite solid solution series [CaMg(CO3)2–CaFe(CO3)2]. Under routine operating conditions for the analysis of carbonates for δ18O with a CAMECA IMS 1280 instrument (at WiscSIMS, University of Wisconsin‐Madison), the magnitude of instrumental bias along the dolomite–ankerite series decreased exponentially by ~ 10‰ with increasing Fe content in the dolomite structure, but appeared insensitive to minor Mn substitution [< 2.6 mol% Mn/(Ca+Mg+Fe+Mn)]. The compositional dependence of bias (i.e., the sample matrix effect) was calibrated using the Hill equation, which relates bias to the Fe# of dolomite–ankerite [i.e., molar Fe/(Mg+Fe)] for thirteen reference materials (Fe# = 0.004–0.789); for calibrations employing either 10 or 3 μm diameter spot size measurements, this yielded residual values ≤ 0.3–0.4‰ relative to CRM NBS 19 for most reference materials in the suite. Analytical precision was ± 0.3‰ (2s, standard deviations) for 10‐μm spots and ± 0.7‰ (2s) for 3‐μm spots, based on the spot‐to‐spot repeatability of a drift monitor material that ‘bracketed’ each set of ten sample‐spot analyses. Analytical uncertainty for individual sample analyses was approximated by a combination of precision and calibration residual values (propagated in quadrature), suggesting an uncertainty of ± 0.5‰ (2s) for 10‐μm spots and ± 1‰ (2s) for 3‐μm spots.

Predicting instrumental mass fractionation (IMF) of stable isotope SIMS analyses by response surface methodology (RSM)

“…15a, b, c). Notably, this characteristic behaviour of IMF vs. the electrostatic deflectors was also observed in the response surface models fitted using the previously published SIMS data of Śliwiński et al 15 on dolomite-ankerite and Pollington et al…”

“…14c, d), both using a Cameca 1280HR (WiscSims). 15,46 Unexpectedly, the variations observed along Y were slightly negative parabolic or linear with low slopes in all the cases, indicating that the variation of IMF along X is much stronger than along Y. In addition, this behaviour is independent of the value ranges of X, Y, and IMF (Fig.…”

“…14 Śliwiński et al correlated the Fe content with bias variation in 18 O/ 16 O and 13 C/ 12 C analyses of the dolomite-ankerite solid solution series using the Hill's equation for nonlinear effects. 15,16 Introduced by Box and Wilson, RSM has been applied to build up multivariate statistical models in a wide variety of industrial, engineering and experimental processes. 17 Successful RSM applications can be found in, e.g., Riley [18][19][20][21][22][23][24][25][26][27][28][29][30] In addition, the mathematical and statistical aspects of RSM and related experimental techniques are covered in e.g., Box, Box, and [31][32][33][34][35][36][37] Simultaneously, the statistical prediction models have been recognized as powerful tools for e.g., exploring the underlying causal relationships below the datasets, building and/or assessing new knowledge and improving previous models.…”

Quantification of oxygen isotope SIMS matrix effects in olivine samples: Correlation with sputter rate

“…For example, Page et al (2010), found more complicated O-isotope matrix effects in garnet where sputter rate was not correlated with the magnitude of instrumental mass fractionation. Matrix effects in carbonates are also complicated (e.g., Rollion-Bard and Marin-Carbonne, 2011;Śliwiński et al, 2016a, 2016b) although Śliwiński and colleagues were able to fit the IMF data as a function of Fe / (Fe + Mg) content of the carbonates along the dolomite -ankerite solid solution to a Hill equation, which has a similar functional form to the sigmoid that we used relating change in IMF to sputter rate in olivine.…”

Predicting instrumental mass fractionation (IMF) of stable isotope SIMS analyses by response surface methodology (RSM) [Dataset]

“…14 Śliwiński et al correlated the Fe content with bias variation in 18 O/ 16 O and 13 C/ 12 C analyses of the dolomite-ankerite solid solution series using the Hill's equation for nonlinear effects. 15,16 Introduced by Box and Wilson, RSM has been applied to build up multivariate statistical models in a wide variety of industrial, engineering and experimental processes. 17 Successful RSM applications can be found in, e.g., Riley [18][19][20][21][22][23][24][25][26][27][28][29][30] In addition, the mathematical and statistical aspects of RSM and related experimental techniques are covered in e.g., Box, Box, and [31][32][33][34][35][36][37] Simultaneously, the statistical prediction models have been recognized as powerful tools for e.g., exploring the underlying causal relationships below the datasets, building and/or assessing new knowledge and improving previous models.…”