Fig. 1 The schematics of TC/EA-IRMS configuration.
(A) Autosampler, (B) TC/EA reactor, (C) Gas trap and GC column, (D) Con Flo IV, (E) IRMS, (F) pyrolysis chromatograph

2.2.3. Sample encapsulation procedure

Except for IAEA-601 and -602, all samples and silver capsules were heated at 220˚C in an oven for more than 4 h before being encapsulated into the silver capsule to remove adsorbed water molecules. Additives such as fluorine compounds and graphite or additives homogenized with the samples in advance were dried under the same conditions. Immediately after retrieval from the oven, each sample was weighed to obtain an oxygen mass of 100 µg. The measured weight was ranged to within ±5% of the intended weight (100 μg of oxygen) in order to ensure quick weighing. The samples and additives were sealed by folding the silver capsule while pushing out air, and the capsules were again kept dry at 220°C for at least 4 h before being placed in an autosampler. Considering the melting point of benzoic acid (122°C), IAEA-601 and -602 were not heated prior to, during, or following the weighing and sealing procedures. Upon completion of the TC/EA-IRMS, the silver capsules were promptly transferred to an autosampler. Prior to initiating the measurement, the autosampler underwent an overnight purge using He gas to eliminate any remaining gas.

2.2.4 experimental runs

Several recipes for the sample – additive mixture and mixing procedures were compared to identify the most promising method. This can be divided into four series of measurements (Table 2).
Run1: Analysis for comparing the reactivities of the sample-PTFE mixtures with different mixing procedures and different C sources, graphite, or Ni/C. Three different procedures were tested. (1) Sample–PTFE–graphite enclosed in a silver capsule without homogenization. (2) Sample–PTFE–Ni/C was enclosed in a silver capsule without homogenization. (3) Sample–PTFE–Ni/C mixture was homogenized using an agate mortar and pestle before weighing and encapsulation. For these three recipes, the F/O and C/O molar ratios were >2.3 and >1.5, respectively.
Run2: Analysis of reactivities with different F sources. Six inorganic fluorides (LiF, NaF, KF, CaF2, BaF2, and AlF3) and graphite powder were added and homogenized before weighing. The F/O and C/O molar ratios were set to 2 and 1.5, respectively.
Run3: Analysis of reactivities with different F/O molar ratios. NaF was added to the samples at F/O molar ratios of 2, 3, 4, and 6, and homogenized with graphite or Ni/C before encapsulation.
Run4: Analysis for confirming O yields and δ18O values using recipes that showed high reactivity in Run1–3. NBS28, SWy-1, and other montmorillonite samples (TKN-01, 29, 31) were analyzed using (1) NaF and graphite with F/O and C/O molar ratios of 6 and 1.5, respectively, and (2) NaF and Ni/C with F/O and C/O molar ratios of 2 and 1.5, respectively.
Table 2. Mixing recipe of each analytical run