Stability of reference masses: VII. Cleaning methods in air and vacuum applied to a platinum mass standard similar to the international and national kilogram prototypes

Mercury contamination and the build-up of carbonaceous contamination are two contributing factors to the instability observed in kilogram prototype masses. The kilogram prototypes that lie at the core of the dissemination of the SI base unit were manufactured in the late 19th century, and have polished surfaces. In papers IV and V of this series we developed a method for cleaning noble metal mass standards in air to remove carbonaceous contamination. At the core of this 'UVOPS' protocol is the application of UV light and ozone gas generated in situ in air. The precise nature of the carbonaceous contamination that builds up on such surfaces is difficult to mimic demonstrably or quickly on new test surfaces, yet data from such tests are needed to provide the final confidence to allow UVOPS to be applied to a real 19th century kilogram prototype. Therefore, in the present work we have applied the UVOPS method to clean a platinum avoirdupois pound mass standard, 'RS2', manufactured in the mid-19th century. This is thought to have been polished in a similar manner to the kilogram prototypes. To our knowledge this platinum surface has not previously been cleaned by any method. We used x-ray photoelectron spectroscopy to identify organic contamination, and weighing to quantify the mass lost at each application of the UVOPS procedure. The UVOPS procedure is shown to be very effective. It is likely that the redefinition of the kilogram will require mass comparisons in vacuum in the years to come. Therefore, in addition to UVOPS a cleaning method for use in vacuum will also be needed. We introduce and evaluate gas cluster ion-beam (GCIB) treatment as a potential method for cleaning reference masses in vacuum. Again, application of this GCIB cleaning to a real artefact, RS2, allows us to make a realistic evaluation of its performance. While it has some attractive features, we cannot recommend it for cleaning mass standards in its present form.