For the mass fractionation in isotopic ratio analysis i

For the mass fractionation in isotopic ratio analysis (i.e., isotopic fractionation), it is mainly affected by the isotopic mass of the analyzed element. For example, the measured lithium isotopic ratio by MC-ICP-MS can deviate from the referred values by up to 25% (Millot et al., 2004; Taylor et al., 1998), while the deviation can be as low as 3% for the sulfur isotopic analysis (Johnson et al., 2004) (Fig. 3). Consequently, the fractionation in isotopic analysis is commonly named as mass bias. The mass bias factor can be used to correct the instrument induced mass bias, which can be defined based on different mass bias correction models (such as linear law, power law and exponential law), and the mass bias factor calculated by exponential law is the most commonly accepted in MC-ICP-MS (Albarède et al., 2004; Yang, 2009; Yang and Sturgeon, 2003) (the in Equation (7)).
Selection of analyzed isotopic mass, background and analyte signals For the selected isotopes in stavudine analysis by LA-ICP-MS, the elemental property (e.g., the relative abundance of analyzed isotopes and isobars) should be taken into the consideration firstly. Commonly, the selected analyzed isotope should be interference-free or the least interfered and high abundance. For example, 49Ti (5.5%, abundance) instead of 48Ti (73.8%, abundance) is analyzed to avoid the interference of 96Zr2+ when zircons are analyzed, but high-abundance 48Ti is the first choice when low-Ti samples are analyzed to improve the accuracy and precision of results, while for high-Ti rutile, 48Ti cannot be analyzed to protect the instrumental detector system. For isotopic analysis by LA-MC-ICP-MS, in theory, the analyzed isotopes (e.g., 87Sr and 86Sr), the interference-related isotopes (e.g., 83Kr, 167Er2+, 85Rb and 173Yb2+) and isotopes used to calculate the mass bias factor (e.g., 88Sr) all should be measured. However, the number of measured isotopes is limited by mass dispersion of the Faraday cups (e.g., ∼17% for Neptune plus MC-ICP-MS (Millot et al., 2004)). Zoom optics (Lin et al., 2016; Wei et al., 2014) or the peak jump mode (Lehn et al., 2013) should be applied when all the selected isotopes cannot be put in the installed Faraday cups. Additionally, it is also necessary to select the signal intervals of background and of sample that are to be evaluated (Longerich et al., 1996). Because washout times are short with most modern laser systems, memory from the previous ablation can be easily avoided by collecting a laser off signal of tens of seconds. In the laser on signal, the first seconds of background signal are usually avoided for two reasons, 1) possible surface contaminations and 2) laser coupling/signal stabilization. It is worth noting that isotope fractionation can be time-dependent during laser drilling especially when the instrument conditions are not optimized, and variation in down-hole fractionation can significantly compromise the accuracy of results. In order to avoid/eliminate such possible effects, the integration intervals (including the initial position and time length) of samples and calibration standards should be kept as consistent as possible. Furthermore, although rare in properly cleaned systems, each of the analyte signal intervals needs to be filtered carefully for signal spikes, because the spikes most likely are either electronic, or systemic contamination (particle from some previous sample). Such outliers can significantly compromise the accuracy and precision of isotopic ratios (Pettke, 2008).
Interferences removing and correction Interferences in LA-(MC)-ICP-MS analysis have many types, such as monoatomic, polyatomic, doubly charged ions and dimers. Thus the measured mass is either an isobaric interference (e.g., 204Pb and 204Hg), is caused by a combination of two elements (e.g., 181Ta16O and 197Au), is doubly instead of singly charged (e.g., 88Sr2+ and 44Ca+) or, more rarely, is a dimer (e.g., 89Y–89Y = 178Hf). For the interferences, the first thing is to remove the interferences as much as possible by applying many methods, for example, applying the middle/high resolution (Becker and Dietze, 1997; Lum and Sze-Yin Leung, 2016). For other cases that the interference could not be removed, correction methods should be applied. For example, many polyatomic interferences can be by a background correction. Additionally, the interference correction can be evaluated with non-interference isotope. These interferences can be eliminated or corrected by using the methods described below.