关键词: 记忆效应, 时间漂移, 信号强度依赖性, 数据标准化, 碳稳定同位素

Abstract: Stable isotopic composition of carbon (δ¹³C) in plants and soil can indicate, trace, and integrate key processes and functions of carbon cycling. Dual-inlet and continuous-flow isotope ratio mass spectrometry (IRMS) are two techniques for determining δ¹³C of carbon-containing compounds. The precision and accuracy of analyzed data depends on analytical error and data normalization method. Analytical errors include memory effect, time drift and signal intensity-dependence effect, etc. The easiest method to eliminate or minimize the memory effect is to set up a measurement sequence ranging from low to high isotopic composition of the actual samples, or increasing the flushing time of the tube. A sufficient number of identical standards should be included with the measurement sequence, and the function of measurement time and analysis error of identical standards are used to correct time drift. Signal intensity dependence can be corrected by considering the impacts of blank effect and nonlinear response of instrument separately, with the blank effect considered first, and the nonlinear response of the instrument second. Data normalization strategy includes the selection of standards and normalization methods. The δ¹³C of selected standards should bracket the range of δ¹³C of samples. Based on the principle of identical treatment, the selected standards should share similar physical and chemical properties with the samples, and undergo the same analytical procedure. The number of selected standards should be ≥4, or the number of repeated determinations for each standard should be ≥4. The same set of standard materials should always be used for a particular element and be specific to an analytical technique. The measurement sequence should add quality control standard material to monitor the long-term accuracy of the analytical results.

Key words: memory effect, time drift, signal intensity dependence, data normalization, carbon stable isotopic composition.