Logo of the Physikalisch-Technische Bundesanstalt

Successful testing of platinum-iridium sorption artefacts for precision mass determinations under vacuum conditions

27.07.2010

The influence of sorption effects on the accuracy of mass determinations under vacuum conditions was significantly reduced by means of sorption artefacts made of a platinum-iridium alloy.

Mass determinations under vacuum conditions (< 0.1 Pa) are of particular interest for mass metrology, as in vacuum the influence of the air buoyancy and convection effects can be neglected. This advantage is especially important when the mass of objects is determined which show a large density difference to the reference mass. This is, for example, the case when the mass of kilogram prototypes of platinum-iridium (density: 21.5 g/cm³) is compared with standards made of steel (density: 8.0 g/cm³) or silicon (density: 2.3 g/cm³).

Here it must, however, be kept in mind that - in the course of the change between air and vacuum - the mass standards are subject to mass changes which must be taken into account by sorption corrections. For their gravimetric determination, special mass standards – so-called sorption artefacts – are used. These artefacts are designed in such a way that they have the same mass, the same volume, the same surface properties, but a large surface area difference. In most cases, one of the artefacts has the form of a cylinder, whereas the other artefact is composed of several discs (Figure 1). From the knowledge of the surface difference and the change of the mass difference between the two artefacts measured during the air-vacuum-transfer, the mass change of the sorption layer per unit area - i.e. the sorption coefficient in ng/cm² - can be determined.

Platinum-iridium sorption artefacts during the test in a vacuum mass comparator. Left: Cylinder, right: Stack of discs, composed of 8 discs and 21 distance rods

Figure 1: Platinum-iridium sorption artefacts during the test in a vacuum mass comparator. Left: Cylinder, right: Stack of discs, composed of 8 discs and 21 distance rods

For highest requirements, sorption artefacts made of a platinum-iridium alloy (PtIr 10) - the same material from which also the prototypes of the kilogram are made - are used [1, 2]. Due to the same density, the influence of the buoyancy correction on the measurement uncertainty can be neglected when the mass of the platinum-iridium cylinder is compared in air with a prototype of the kilogram. This cylinder, whose mass can be determined in air with the accuracy of a prototype, serves as the reference mass for mass comparisons under vacuum conditions. It thus represents a transfer standard between air and vacuum and is - during the transfer between air and vacuum - subject to mass changes, which can be determined from the change of the mass difference between both artefacts during the measurements and taken into account. The respective sorption coefficient depends on the environmental conditions (air pressure, air humidity) and the surface characteristics (material, roughness, state of cleaning).

The raw material for the platinum-iridium sorption artefacts of PTB was verified by the Bureau International des Poids et Mesures (BIPM) in accordance with the requirements for prototypes of the kilogram and delivered to PTB. PTB's department 5.5 "Scientific Instrumentation” performed the surfaces treatment and the mass adjustment. In the last manufacturing step, a mean roughness of 2 nm < Ra < 4 nm was achieved which corresponds to the surface quality of the kilogram prototypes of the BIPM [3]. The mass of the platinum-iridium cylinder was adjusted to approx. 1 mg to the mass of a kilogram prototype. The surface difference between the stack of discs and the cylinder amounts to approximately 180 cm². Figure 1 shows the completed sorption artefacts during their testing in a vacuum mass comparator. Repeat measurements in air and vacuum showed that the mass difference between the stack of discs and the cylinder changes during an air-vacuum-transfer on average by approximately 7 µg (cf. Figure 2). This corresponds to a sorption coefficient of 38 ng/cm² or a mass equivalent of 1.2 monomolecular water layers. Due to the relatively small surface of the platinum-iridium cylinder (72 cm²), this means for its mass a sorption correction of only -2.7 µg, i.e. the mass of the cylinder is in vacuum by 2.7 µg smaller than in air. The standard uncertainty of the sorption correction lies here in a range of only 1 µg – a reduction of more than 80% compared to values obtained so far.

Mass differences measured between the stack of discs and the cylinder in air and in vacuum

Figure 2: Mass differences measured between the stack of discs and the cylinder in air and in vacuum

After successful completion of the manufacture and testing, the platinum-iridium sorption artefacts can now be used as transfer standards for precision mass determinations under vacuum conditions. First measurements are, for example, planned for the mass determination of silicon spheres within the scope of an international project for the redetermination of the Avogadro constant and preparatory investigations for the practical realisation and dissemination of the mass unit after a possible redefinition of the kilogram based on a physical constant.

[1] Picard, A.; Fang, H.: Methods to determine water vapour sorption on mass standards. Metrologia 41 (2004) 333–339

[2] Davidson, S.; Brown, S.; Berry, J.: A report on the potential reduction in uncertainty from traceable comparisons of platinum-iridium and stainless steel kilogram mass standards in vacuum. NPL Report CMAM 88 (2004), Teddington: National Physical Laboratory, 1–24

[3] Haidar, Y.; Tollens, E.; Silvestri, Z; de Fornel1, F.; Zerrouki, C.; Picard, A.; Pinot, P.: Study and comparison of two polishing methods for platinum–iridium surfaces, by means of three characterization techniques. Metrologia 42 (2005) 115–128

Contact person:

Michael Borys, Department 1.1, WG 1.11, E-mail: michael.borys@ptb.de