2.3. Sample analysis

Neodymium and Sr analyses were performed on a Thermo Scientific Triton Plus Thermal Ionization Mass Spectrometry (TIMS). Standards and samples were loaded on outgassed Re filaments in 1–2 µL 10% HNO₃ with 1 µL H₃PO₄ for Nd and 50% of the Sr fraction in 1 µL 10% HNO₃, with 1.5 µL TaCl₅ for Sr. Neodymium analyses were performed using 10¹³ Ω resistors fitted to the amplifier system and 10¹¹ Ω resistors if enough sample was available [2]. The 10¹³ Ω resistors measurements result in a 100 fold higher output voltage compared to default 10¹¹ Ω resistors (while the signal to noise ratio is improved by a factor of 10). This allows for more precise measurements of small data using the 10¹³ Ω resistors.

Strontium analyses were performed using 10¹¹Ω resistors. Mass-fractionation corrections were performed to ¹⁴⁶Nd/¹⁴⁴Nd = 0.7219 and ⁸⁶Sr/⁸⁸Sr = 0.1194. Standards were measured to check for accuracy and reproducibility (Table 2). Blank corrections were not necessary as total procedural blanks yielded negligible amount of neodymium (1.1 ± 1.7 pg, n = 56) and strontium (24.7 pg ± 38.9, n= 26). Blank data for the samples by Font et al. [11] was not reported but is assumed to be similarly low as the lead blanks reported in study (≤50 pg Pb, n = 27).

Results of standards measured during the data collection period. JNdi-1 and NBS987 are international standards. CIGO and TSTD are inhouse standards.