By Devin Powell Video: The almost perfectly round balls of pure silicon could be used to redefine the kilogram (Image: CSIRO) When asked by the Pope to demonstrate his artistic skill, 14th century Italian painter Giotto di Bondone supposedly drew a perfect circle freehand and said: “That’s more than enough.” Now, an international group of engineers and craftsmen has gone him one better and built a pair of nearly perfect spheres that are thought to be the roundest objects in the world. The unusual balls, discussed last week at the SPIE Astronomical Telescopes and Instrumentation conference in France, were created as an answer to the “kilogram problem”. The kilogram is the only remaining standard of measurement tied to a single physical object: a 120-year-old lump of platinum and iridium that sits in a vault outside of Paris, France. But the mass of this chunk of metal is slowly changing relative to the 40-odd copies kept by other countries, and no one knows why or by how much. So researchers charged with policing units and measures, called metrologists, have come up with several suggestions to redefine the kilogram. In 2011, the International Committee for Weights and Measures will try to decide the issue. One proposal, pushed by an international team called the Avogadro Project, aims to define the kilogram in terms of a specific number of silicon atoms. Just how many? That’s where the newly created silicon spheres come in. Over the next few years, groups in Italy, Belgium, Japan and the US will try to calculate the exact number of atoms in each one. It is no easy task. To determine the volume of each sphere, they will use optical interferometers to measure its width from 60,000 different points on its surface. Meanwhile, X-ray crystallographers will take pictures of the silicon crystal structure to determine the spacing and density of the atoms. By multiplying volume by density, each group should produce its own count of how many silicon atoms make up a kilogram. The important thing is for those numbers to agree with each other. But making the spheres, which were presented to the Avogadro Project in April by the Australian Commonwealth Scientific and Industrial Research Organization (CSIRO), was also a challenge (watch a video of the process). The silicon was purified in ex-Soviet centrifuges once used to refine uranium for nuclear weapons. The centrifuges separated the silicon by isotope, allowing researchers to create a remarkably pure batch of silicon-28. From Russia, the material travelled to Germany’s national metrology institute, PTB, to be grown into a giant crystal by ageing equipment that produced silicon for East Germany decades ago. After six failed attempts, a pure crystal was made and cut into two 5-kg blocks that were shipped to Australia. To shape the spheres, the Australian Center for Precision Optics pulled optical engineer Achim Leistner out of retirement. Leistner, who has been creating precision spheres for decades, considers these final two to be his masterpieces. The ACPO team used techniques similar to the way Isaac Newton ground lenses for his telescopes 300 years ago. Opticians manipulated two spinning rotors to grind the surface by hand. After months of sanding, the team produced two spheres with diameters of 93.75 millimetres. The mass of each sphere matches that of the Australian copy of the kilogram. The small-scale roughness of the balls varies by only 0.3 nanometres, and their curvature by 60 to 70 nanometres. “If you were to blow up our spheres to the size of the Earth, you would see a small ripple in the smoothness of about 12 to 15 mm, and a variation of only 3 to 5 metres in the roundness,” Leistner told New Scientist. This kind of roundness was not possible 20 years ago, according to Leistner, because we could not see variations on this scale. Also, advances in computer processing have improved the speed at which data can be decoded from measuring devices such as the talyrond, a spindle-like instrument that rotates around the ball and can detect deviations from perfect roundness down to about 5 nanometres. That allows researchers to quickly correct any imperfections they find. But even if all of the Avogadro Project’s research teams arrive at the same number of silicon atoms in each sphere, it’s far from clear that the International Committee for Weights and Measures will take up their definition. That’s because some metrologists believe the Avogadro Project’s precision spheres may simply replace one ailing physical standard with another. They support a competing approach called the watt balance, which would redefine the kilogram in terms of magnetic fields and electrical forces. Richard Steiner, an advocate of this proposal at the National Institute of Standards and Technology in Boulder, Colorado, US, argues that energy can be measured with less error than physical properties like the diameter of a sphere. There are problems with the watt balance technique – so far the two groups that have tried it have arrived at different numbers for the kilogram – but Steiner’s criticism raises a thorny question. These million-dollar spheres may be the roundest in the world, but will they be round enough?