The Kilogram Is Putting On Weight

There's long been debate over the accuracy of the standardised kilogram. Now, scientists have shown once and for all that the lump of metal defining the unit of mass has been putting on some weight.

The original kilogram -- the International Prototype Kilogram, or IPK, made of platinum and iridium alloy -- is the standard against which all other measurements of mass are set. So, it's quite important. 40 replicas exist across the world, and the UK is in possession of replica 18. Researchers from Newcastle University, UK, thought they'd take a look at -- and they found something amiss.

Using a state-of-the-art X-ray photoelectron spectroscopy machine -- the only one of its kind in the world -- they were able to analyse the block with unprecedented accuracy. Turns out age hasn't been kind. The X-ray measurements show that there's a build-up of hydrocarbons on the surface -- weighing up to 100 micrograms.

While that would be fine if every block was ageing in the same way -- as we'd all still be working to the same standard, it'd just be subtly different than it used to be -- the scientists don't think that's the case. Peter Cumpson, Professor of MicroElectroMechanical Systems at Newcastle University, explains:

"Around the world, the IPK and its 40 replicas are all growing at different rates, diverging from the original... We're only talking about a very small change... but mass is such a fundamental unit that even this very small change is significant and the impact of a slight variation on a global scale is absolutely huge."

So what's to do? Well, the researchers have also shown that by exposing similar blocks to a mixture of UV and ozone it's possible to remove the surface contaminants and -- potentially -- restore them to their original weights. But nobody has attempted that task with the kilogram standards so far, so it remains to be seen if it will actually work in practice. Who's gonna try first? [Science Daily]

Picture: Wikimedia Commons

WATCH MORE: Science & Health News


    Or just make some more and replace them every so often?

      How would you make more exactly? What standard would you use to make sure your new one is actually a kilogram.
      Would you measure it against one of the existing ones? It's not as easy as it sounds.

        Couldn't be much harder than it sounds, if they can measure a current one to tell that it has 100micrograms of crap on it then surely they can measure a new one to make sure it doesn't.

          I think the X-ray photoelectron spectroscopy showed them there was extra weight there. Not that the overall thing weighed 1kg+100ug. As in they could weight the carbon deposits not the whole thing.

          Here is the fundamental problem everyone is missing. What is 1Kg? You can't just throw something on a scale and say it's 1Kg, because 1Kg is now not what it was 1 year ago (or whatever).
          This is more of a theoretical problem not a practical one. 1Kg changing isn't going to affect your cooking scale, and it probably won't affect how you calibrate your scales.
          But when you start doing very specific and accurate scientific or engineering calculations, the difference is a problem.
          Realistically, we don't physically calibrate our equipment to the IPK, and normally there is a tolerance for error anyway.
          However for those situations were you do need to calibrate exactly to the IPK, you can introduce an offset. And then that offset would carry through whatever process you were working on.

          Here is a practical example. Take company that manufactures calibration weights. They have an expensive high accuracy scale to measure the weights that they make. They get that scale (which will be their quality control gold standard) calibrated with the IPK.
          Now when they use that scale, all their manufactured weights will be off. And then when engineering firms and universities use these weights to calibrate their own equipment, they will still have that offset. And thus you start doing chemistry or physics experiments or building products with incorrect weights... I'm sure you can see the problem here.

          Even with the above example with using X-ray photoelectron spectroscopy, how do you know that it's actually 100ug? That measurement is still based around a standard 1Kg, so it's acutally going to be 100ug + however much the IPK has change, and since you are measuring the IPK, you now have an endless vicious cycle.

          I hope that clears some things up.

          Sounds much more like a management issue more then anything...

          a really small file, with a steady hand ?

    How could they lose track fo what 1kg is.

    its exactly 1 liter of pure water. volume is easier to keep constant. no problems.

      Not true. Approximately, yes. Exactly, no. And it depends on temperature and pressure.

        At 4 °C, measured at sea level and at latitude 0°00'00'' (the equator).

        Last edited 09/01/13 12:21 pm

          Wouldn't 4 °C at sea level on the equator be really difficult to achieve?

            No more difficult than storing food and drink at that temperature.

              But if you artificially generate one condition (temperature), why require the other conditions, particularly as there are so many other variables that could ...

              Oh wait, I see from comments below that I've just independently come to the same objection that resulted in that standard for the litre being changed half a century ago. On the plus side, it looks like I'm 7 years off coming up with the mol. Cool. Then I could look at what a mole of moles would ... no wait, that's been done too:

          even then, stuff going on in the mantle can also be enough to mess with the accuracy needed.

      And one liter of pure water is defined as 1 kg against the kilo-weight on a scale at a certain altitude at a certain location?

        1 kilogram is 1 litre of water, 1000 litres of water is 1 cubic meter
        It doesn't really matter what the temperature or altitude is. You can work your way back from 1 cubic meter and convert a distance type measurement into volume and then weight

        If all else fails we can use gravity to figure out what a given distance is to help measure meters etc.

        Last edited 09/01/13 9:48 am

          1 cubic metre is 1000L of water or 1 tonne. You mean 1 cubic decimetre.

            Sorry you're right about that.
            First day back to work without coffee =/

          "... 1 litre of water is 1 cubic meter"


          The metal kilo is based on the earlier idea of a litre of water at 4 deg C (the temp at which water is densest), but was used to provide a constant weight for measurement rather than, I assume, trying to maintain a liquid at a certain temperature with no evaporation losses, contamination etc. It's used for the sake of convenience, as most scientific conventions are.

            You also missed that 1 cubic metre of water is ten thousand litres.

            Just saying.

          No - 1 litre of water is based off the IPK 1Kg standard.
          In fact pretty much every mass measurement is based against a standard, 1Kg is only 1Kg because of the standard. Otherwise it's an arbitrary made up value.
          And volume is simply the length x width x height that is taken up by a known mass.
          Platinum iridium alloy is used as a standard because it's very inert. You can't just get 1m x 1m x 1m of pure H2O and know exactly how much it weighs.
          I hope that makes sense

            I would think 1 litre would be based of the metre - i.e. 1/1000 of a cubic metre. Regardless of what material is within that volume.

              Sorry you are right,
              But my point still stands.

              This is from the wiki page:
              From 1901 to 1964, the litre was defined as the volume of one kilogram of pure water at 4 °C and 760 millimetres of mercury pressure. The kilogram was in turn specified as the mass of a platinum/iridium cylinder held at Sèvres in France and was intended to be of the same mass as the 1 litre of water referred to above. It was subsequently discovered that the cylinder was around 28 parts per million too large and thus, during this time, a litre was about 1.000028 dm3. Additionally, the mass-volume relationship of water (as with any fluid) depends on temperature, pressure, purity, and isotopic uniformity. In 1964, the definition relating the litre to mass was abandoned in favour of the current one. Although the litre is not an official SI unit, it is accepted by the CGPM (the standards body that defines the SI) for use with the SI system. CGPM defines the litre and its acceptable abbreviations.

              The main problem with measuring mass from volume (or vice versa) is that there are too many other factors as stated above (pressure, temp, etc).

      Flicking through the wiki, this definition was apparently abandoned decades ago. Which makes sense, you need exact temperature and pressure, which introduces the same room for error as the platinum method.

      Not entirely.

      1 litre of water will not always be 1 kg. The closest this comes to being true is at 1 atmosphere at around 4 degrees Celsius.

      Increased pressure means 1 litre will have more water molecules and thus have more than 1kg of mass, and vice versa.

      Increased temperature above 4 degrees increases the vibrations of the molecules meaning they spread out more and thus 1 litre will have fewer molecules and thus less mass.

      Decreased temperature (in water anyway) below 4 degrees and the water begins to crystalise into ice, which takes up more space than liquid water, thus 1 litre will have fewer molecules and thus less mass.

      We're not talking about "good approximations" with this, we are talking "exact", so if there are variances that can't be controlled adequately, then it can't be used as the standardised model.

    Huh, a little heavier? I remember a year or so back and the news was it was shedding molecules. Damn you science, make up your mind!

      I remember this too! The block was wearing down... not fattening up...

      I remember this too. I think it was the master version in Paris that was losing molecules of the material it is made of, but this was a UK replica gaining hydrocarbons externally.

      Just shows how important it is to keep these as absolutely protected as possible. That or figure out a mathematical/theoretical instead of a relative definition of a kilo.

        ahhk, that explains it.

        I know that the second and the meter are defined by the passage of the speed of light through a vacuum, could do something like that - 'X amount of Y atoms, each with Z electrons'

          Well as i recall they were planning to try build a sphere of silicon with a specific number of atoms to define the kilo, but i think they ran into some issues, not sure what stage its at now.

    Time to go to WeightWatchers.

    You'd think by now they would have changed the kilogram to something less arbitrary to do with light, like they did with the metre.

    Can just hear all those Americans laughing, what with their pounds and imperial weights and measures laughing at our metric system failing! :(

    Last edited 09/01/13 2:08 pm

    Wouldn't they have enough super accurate scales for scientific purposes throughout the world by now to make this a non-issue? Just use one that was based on the correct kg to make more models

      And routinely calibrate it with what, exactly?

Join the discussion!

Trending Stories Right Now