Abstract
The kilogram is the last of the seven base SI units that is defined as an artefact rather than a naturally occurring physical phenomenon and as such presents a unique set of problems in both its maintenance and dissemination. The accuracy of the International Prototype Kilogram, a cylinder of platinum-iridium alloy, and of its 90 National Standard copies is limited by the surface stability of the artefacts. Various algorithms are used by the holders of these copies, based on empirical mass measurements, to predict the mass gain of platinum-iridium mass standards but their accuracy is severely limited by the amount and reliability of the weighing data. Work is also underway to replace the current definition of the kilogram with a more fundamental derivation. At present there are three major methods being investigated all of which require a mass to be realised in vacuum, presenting the problem of discontinuity between weights realised in vacuum and those used in air. The work described investigates the effect on the surfaces of mass standards of storage in and transfer to and from vacuum both by mass measurement and by surface analysis using X-ray photoelectron spectroscopy. Standards of both stainless steel and platinum-iridium were investigated and values for the step change in mass on exposure to vacuum were calculated for both materials, allowing, for the first time, traceability between weighings made in air and in vacuum. The measurements of the contamination on the surfaces of the weights made gravimetrically and by XPS showed excellent correlation and models for the mass gain processes of artefacts exposed to ambient and vacuum conditions were derived. Additionally analysis of the results allowed a value for the average density of the overlayer to be calculated based on surface studies and weighing data.