Fig. 5. The m /z 28 traces during analysis of
quartz and SWy-1 with the combination of Ni/C and NaF. Red arrows
represent the positions of CO peak ends.
3.4. Oxygen yields and δ18O values using a promising
recipe (Run4)
The study from Run1 to Run3 revealed two promising recipes: (1) a recipe
using graphite and NaF with F/O = 6 (hereafter, this recipe will be
referred to as ”FO6”), and (2) one using Ni/C and NaF with F/O = 2
(hereafter, ”NiC_FO2”). Using these two recipes, a series of
measurements were performed for NBS28, SWy-1, TKN-01, -29, and -31
samples (Run4, 74 samples in total, including 36 standards). The oxygen
yields monitored in this series of measurements are shown in Fig. 6. The
oxygen yields shown here were calculated based on the CO peak area
normalized to the oxygen mass and assuming that the average oxygen
yields of IAEA-601 and -602 were 92.8% 24.
First, the oxygen yields of NBS18 calcite were about 60%, indicating
that about 2/3 of the oxygen was converted to CO gas by the reaction
CaCO3 + Cexcess → CaO + 2CO↑, and the
remaining 1/3 became residue as CaO (melting point 2,572°C). For the
NaF-bearing samples, the oxygen yields of the first samples in a
sequence of five or six consecutive measurements tended to be
anomalously high. Especially in the first sequence (NBS28_FO6, run
order #16–20), the first sample showed 190% of yield and then
gradually declined to approximately 92.8%. The chromatographs of the
first sequence were shown in Fig. 7. In the subsequent sequences, only
the first sample showed a significantly higher yield, whereas the yields
of the second and subsequent samples tended to stabilize at
approximately 90–100% for the sample using FO6 and approximately 110%
for the samples using NiC_FO2. The high yield of the first sample in
each sequence was probably derived from oxygen released from the residue
of the standards analyzed in the just preceding sequence. NBS18 is
expected to leave approximate 40% of the oxygen in the reactor.
IAEA-601 and 602 were reported to have an approximate 93% of oxygen
yield in thermal decomposition13,24 , and 7% may
remain in the residue. Fluorine bearing samples just after the
measurement of standards also show longer tailings. The wider peaks for
the earlier samples in the same sequence (Fig. 7). Except for the first
sequence of fluorine bearing samples (run order #16–20), tailings were
rarely observed in the second or subsequent samples in each sequence. It
was ruled out that abnormal yields could have been caused by the order
in which the weigh-ins were done after the drying oven was removed. This
is because, regardless of the weighing procedure’s order, the latter
portion of the measurement series, which contained the FO6 and NiC_FO2
recipes in the same sequence (run order #48–53, #57–62, #66–71)
likewise demonstrated the same trend. The oxygen yields of the samples
with NiC_FO2 were higher than those of the samples with FO6, suggesting
the possible oxidation of Ni in Ni/C during storage.