Study on the Difference Between the Triple-Point Temperatures of $^{20}$Ne and $^{22}$Ne Using Sealed Cells

Abstract

At the National Metrology Institute of Japan (NMIJ), the triple points of $^{20}$Ne and $^{22}$Ne were realized using modular sealed cells, manufactured by the Istituto Nazionale di Ricerca Metrologica (INRiM) in Italy to measure the difference of the triple-point temperatures of $^{20}$Ne and $^{22}$Ne. Standard platinum resistance thermometers (SPRTs) were used that were calibrated by NMIJ on the International Temperature Scale of 1990 (ITS-90). In previous reports, sealed cells of $^{20}$Ne and $^{22}$Ne were mounted one at a time in a cryostat and their triple points were realized in separate cool-downs (the single-cell measurement). In this study, first, the triple point was realized using the single-cell measurement for $^{20}$Ne and $^{22}$Ne cells. Second, the $^{20}$Ne and $^{22}$Ne cells were mounted together on the same copper block and their triple points were realized subsequently one after the other in the same cool-down of the cryostat (the double-cell measurement). The melting curves observed by the single-cell and the double-cell measurements were almost identical for each cell. The difference of the triple-point temperatures between the two cells, $^{22}T -^{20}\!T$, was estimated, not only using the subrange of SPRTs defined in the ITS-90 from 13.8033 K to 273.16 K (subrange 1) but also that defined from 24.5561 K to 273.16 K (subrange 2). The difference in $(^{22}T-^{20}\!\!T)$ between the subranges 1 and 2 is within 0.06 mK, which is caused by the subrange inconsistency in the ITS-90. The standard uncertainty in $(^{22}T-^{20}\!T)$ due to the subrange inconsistency is estimated to be 0.017 mK. After correction for the effects of impurities and other isotopes in the $^{20}$Ne and $^{22}$Ne cells, the difference in the triple-point temperatures between pure $^{20}$Ne and pure $^{22}$Ne is estimated to be 0.146 64 (5) K on subrange 1, which is consistent within the uncertainty with the former studies. When $^{22}T-^{20}\!T$ for pure $^{20}$Ne and pure $^{22}$Ne is estimated on subrange 2, $^{22}T-^{20}\!\!T$ becomes 0.146 60 (5), which agrees very well with the former reports of INRiM evaluating $^{22}T-^{20}\!T$ on subrange 2.