Re-estimation of argon isotope ratios leading to a revised estimate of the Boltzmann constant. (10th August 2017)

• Record Type: Journal Article
• Authors: de Podesta, Michael; Mark, Darren F; Dymock, Ross C; Underwood, Robin; Bacquart, Thomas; Sutton, Gavin; Davidson, Stuart; Machin, Graham
• Is Part Of: Metrologia. Volume 54:Number 5(2017:Oct.)
• Page Start: 683
• Page End: 692
• Publication Date: 2017-08-10
• Subjects: Boltzmann constant -- argon isotope ratio -- argon molar mass -- acoustic gas thermometry; Weights and measures -- Periodicals; Weights and Measures -- Periodicals; 530.805
• DOI: 10.1088/1681-7575/aa7880
• Languages: English
• ISSNs: 0026-1394

Abstract

Abstract: In 2013, NPL, SUERC and Cranfield University published an estimate for the Boltzmann constant (de Podesta et al 2013 Metrologia50 354–76) based on a measurement of the limiting low-pressure speed of sound in argon gas. Subsequently, an extensive investigation by Yang et al (2015 Metrologia52 S394–409) revealed that there was likely to have been an error in the estimate of the molar mass of the argon used in the experiment. Responding to Yang et al (2015 Metrologia 52 S394–409), de Podesta et al revised their estimate of the molar mass (de Podesta et al 2015 Metrologia52 S353–63). The shift in the estimated molar mass, and of the estimate of k B, was large: −2.7 parts in 10 6, nearly four times the original uncertainty estimate. The work described here was undertaken to understand the cause of this shift and our conclusion is that the original samples were probably contaminated with argon from atmospheric air. In this work we have repeated the measurement reported in de Podesta et al (2013 Metrologia 50 354–76) on the same gas sample that was examined in Yang et al (2015 Metrologia 52 S394–409) and de Podesta et al (2015 Metrologia 52 S353–63). However in this work we have used a different technique for sampling the gas that has allowed us to eliminate the possibility of contamination of the argon samples. We have repeated the sampling procedure three times, and examined samples on two mass spectrometers. This procedure confirms the isotopic ratio estimates of Yang et al (2015 Metrologia 52 S394–409) but with lower uncertainty, particularly in the relative abundance ratio R 38:36 . Our new estimate of the molar mass of the argon used in Isotherm 5 in de Podesta et al (2013 Metrologia 50 354–76) is 39.947 727(15) g mol −1 which differs by +0.50 parts in 10 6 from the estimate 39.947 707(28) g mol −1 made in de Podesta et al (2015 Metrologia 52 S353–63). This new estimate of the molar mass leads to a revised estimate of the Boltzmann constant of k B = 1.380 648 60 (97) × 10 −23 J K −1 which differs from the 2014 CODATA value by +0.05 parts in 10 6 .