28.09.2010
The mass of the 28Si sphere AVO28-S8 could be determined in vacuum with a relative standard uncertainty of 6.5 × 10-9.
Within the scope of an international project for the redetermination of the Avogadro constant, and as primary standards of the solid density, 1 kg silicon spheres are used. However, due to their significantly smaller density - compared to Pt-Ir prototypes and steel standards - their mass can at present not be determined with standard uncertainties smaller than 10 µg (relative < 1 × 10-8) in air - as is being aimed at within the international Avogadro project. Mass determinations in vacuum (< 0.1 Pa) have so far been impaired by a relatively large uncertainty contribution of the sorption correction. By using sorption artefacts of a platinum-iridium alloy (PtIr 10), this influence could be reduced by more than 80%, compared to the values obtained so far. Here, the sorption artefacts serve as transfer standards for the transfer between air and vacuum.
The reference mass for the calibration of the sorption artefacts and for the mass determination of the 28Si sphere AVO28-S8 in air was the prototype of the kilogram no. 70 (Figure 1). In addition, air buoyancy artefacts were used to determine the air density during the mass comparisons in air. Figure 2 shows the measurement sequence consisting of four air cyles and two vacuum cycles. The figure shows the mean values calculated from at least three weighing series of six ABBA weighing cycles in each case. The standard uncertainty (k = 1) of the mass determination in air amounts to 12 µg. Thereby, the largest contributions to the overall uncertainty result from the air density determination (9 µg) and from the mass of the kilogram prototype (6 µg). For the mass determination under vacuum conditions, a standard measurement uncertainty of 6.5 µg could be achieved which is mainly composed of the uncertainty contributions of the kilogram prototype (6 µg) and of the sorption correction (1 µg) determined by means of the Pt-Ir sorption artefacts. The mass difference between the first and the second vacuum measurement amounts to -0.8 µg, and the difference between the first and the last measurement in air to -1.7 µg. The mass differences between the sphere masses determined in air and in vacuum are smaller than 2.5 µg. This value speaks for very small sorption effects on the sphere surface and for the high surface quality of the AVO28-S8.
Figure 1: View into the weighing chamber during the mass determination of the 28Si sphere AVO28-S8 (1) with prototype of the kilogram no. 70 (2), platinum-iridium sorption artefacts (3, 5) and air buoyancy artefacts (4, 6)
Figure 2: Mass determination of the 28Si sphere AVO28-S8: Measurement results in air and vacuum (the indicated uncertainty ranges correspond to the respective standard uncertainties u(k = 1); the mass values in air are without correction of the sorption layers on the sphere surface)
Contact person:
Michael Borys, Dept. 1.1, WG 1.11, E-mail: michael.borys@ptb.de