Performance optimization of isotope ratio measurements in transient signals by FI-ICP-TOFMS

Isotopic ratios for Pb (NIST 981) and Rb (NIST 984) and their precisions measured in transient and steady state signals are presented and discussed in this work. The transient signals were generated using a flow injection system (FI) using four different loop volumes coupled to an Inductively Coupled Plasma Time-of-Flight Mass Spectrometer (ICP-TOFMS). Pneumatic nebulization (PN) with a Meinhard nebulizer or an ultrasonic nebulizer (USN) was used for sample introduction. The isotopic reference materials of Pb and Rb were dissolved and diluted to between 0.5 and 100 ng ml(-1) in 4% HNO3 in Milli-Q water. Another set of mixtures was prepared with the same concentrations but fortified with 110 mg l(-1) of Na. This set was introduced to produce a matrix load to the plasma in order to investigate possible effects on the isotope ratio with respect to precision and mass bias. No matrix effect on isotopic ratio and precision was observed. Analytical data for steady state signals are included for comparison to establish the best possible precision attainable on the present instrumentation given the above mentioned concentrations employing 30 s integration time and n>5. Under those conditions, the best possible precision for isotopic ratios was 0.08% (as RSD) for Rb-85/Rb-87, 0.04% for Pb-207/Pb-206, 0.49% for Pb-204/(206) Pb and 0.07% for Pb-208/Pb-206, respectively, using the analog detection mode. In transient mode the corresponding precisions were 0.2, 0.3, 1.0 and 0.3% RSD, respectively. It was found that calculation of ratio and precision performed on the whole peak profile for non-unity ratios resulted in deteriorated precision compared with calculations performed at peak apex on a minimum of five data points. As the peak intensity increases or decreases over the transient peak profile a slight change in the ratios close to the rims of the transient peak could be observed. The quality of the measurements improved for longer integration times (as long as the transient peak profile can be properly defined) and higher concentrations and with transient peak profiles that reached a steady state level. The data obtained for the two smallest loops was of poorer quality and especially for the 10 ng ml(-1) solution since the analog signals obtained were rather low. For these transients results using the ion counting mode are also reported. The ion counting data obtained for concentrations greater than or equal to10 ng ml(-1) and the two larger loops had, however, to be dead-time corrected. The use of USN resulted in improved precision thanks to a 10-fold increase in signal intensity.