RATIONALE: Environmental and biological investigations may require samples that vary over a wide range of concentrations and isotope ratios, making measurements using continuous flow isotope ratio mass spectrometry (CF-IRMS) problematic due to nonlinear signal response. We therefore developed a mathematical approach for correcting nonlinearities over a wide range of sample concentrations and actual δ values. METHODS: Dilution series for two standards were prepared in septum-capped vials and introduced into the mass spectrometer via the standard sampling pathway. Parameters for a nonlinear signal correction were determined by regression on measured isotope ratio vs. both signal strength and actual isotope ratio. We further extended the dynamic range by adjusting the position of an open split based on analyte concentration. Effects of the open split setting required additional mathematical correction. RESULTS: The nonlinearities were corrected over a 100-fold range of sample concentrations and across a 600‰ change in isotope ratios (for δO(2) /N(2) values). The precision, measured as standard deviation, across the upper 90% of the concentration range was ±0.08‰, ±0.05‰, and ±2.6‰ for δ(18) O, δ(15) N, and δO(2) /N(2) values, respectively; the precision across the lower 10% of the range was ±0.22‰, ±0.07‰, and ±7.6‰, respectively. In all cases the linearity correction represented only a small fraction of these precision values. CONCLUSIONS: The empirical correction described here provides a relatively simple yet effective way to increase the usable signal range for CF-IRMS applications. This improvement in dynamic range should be especially helpful for environmental and biological field studies, where sampling methods may be constrained by external factors.
RATIONALE: Environmental and biological investigations may require samples that vary over a wide range of concentrations and isotope ratios, making measurements using continuous flow isotope ratio mass spectrometry (CF-IRMS) problematic due to nonlinear signal response. We therefore developed a mathematical approach for correcting nonlinearities over a wide range of sample concentrations and actual δ values. METHODS: Dilution series for two standards were prepared in septum-capped vials and introduced into the mass spectrometer via the standard sampling pathway. Parameters for a nonlinear signal correction were determined by regression on measured isotope ratio vs. both signal strength and actual isotope ratio. We further extended the dynamic range by adjusting the position of an open split based on analyte concentration. Effects of the open split setting required additional mathematical correction. RESULTS: The nonlinearities were corrected over a 100-fold range of sample concentrations and across a 600‰ change in isotope ratios (for δO(2) /N(2) values). The precision, measured as standard deviation, across the upper 90% of the concentration range was ±0.08‰, ±0.05‰, and ±2.6‰ for δ(18) O, δ(15) N, and δO(2) /N(2) values, respectively; the precision across the lower 10% of the range was ±0.22‰, ±0.07‰, and ±7.6‰, respectively. In all cases the linearity correction represented only a small fraction of these precision values. CONCLUSIONS: The empirical correction described here provides a relatively simple yet effective way to increase the usable signal range for CF-IRMS applications. This improvement in dynamic range should be especially helpful for environmental and biological field studies, where sampling methods may be constrained by external factors.