Merrow Scientific Blog

Just drifting along: Removing the influence of balance drift on your measurement

 digg it |   delicious |   StumbleUpon |  reddit 

One fact of life is that all balances used for measuring mass (and therefore mass change) drift over time. 

Drift occurs gradually and randomly over time and can be due to a number of factors (including changes in temperature, static, radio frequency interference) and happens to all balances.  Typically the drift could be something between 5ug (microgram) and 10 ug (maybe higher) on a 0.1ug resolution balance. 

This drift doesn't really effect one off measurements (e.g. measuring the mass of an item) as the user can tare (re-zero) the balance and even re-calibrate before measurement, so maximising the accuracy of the balance.

But in experiments where the user is measuring continual mass change over time (such as gas sorption, material degradation and other thermal gravimetric type measurements) the drift can have a large effect on the measurement.

Let's look at this a bit further, say a researcher is investigating mass change of a solid over a certain time period (e.g hydrogen gas storage into a solid or thermal degradation) and they're using a system that is listed on the brochure as having a resolution of 0.1ug, and as part of their studies are happy to report mass changes of 20ug. (200 times the reported resolution) Sounds ok? Well maybe not...

As mentioned all balances drift over time,  some companies take steps to limit this drift by trying to control the temperature around the balance which is certainly possibly for ambient to moderate temperature range experiments, but even then they quote for a 0.1ug balance drift of <5ug.  So using a maximum drift of 5ug when the researcher reports mass changes of 20ug it could be in fact 15ug or 25ug (drift can be upwards or downwards) so 20ug +/- 25%.  Sound ok still?  Probably not.... But lets look further at a balance that is not temperature controlled or maybe is in an environment such that temperature control around the balance is very difficult, so lets give that a conservative drift of 10ug.  So same results would be in fact 10ug or 30ug so 20ug +/- 50%....  Probably not worth reporting the result until you're able to get to mass changes of perhaps a reading +/- 1% maybe so with that balance it'd be 1000ug (even though the brochure says 0.1ug resolution) as that would be 1000ug +/- the max balance drift of 10ug. (1000ug = 1mg)

So I think from the above we can safely assume for continual mass change measurement the resolution of 0.1ug is pretty irrelevant and is probably no more than a marketing spec on the brochure.  For one off measurements though resolution is relevant.

Hopefully you're still with me at this point, but what I wanted to highlight is something Rubotherm have been doing for a number of years, in fact this year they celebrate 20years as a company supplying their patented magnetic suspension balance for gravimetric analysis.

As mentioned elsewhere on our site the Rubotherm system uses a patented technique where the precision balance is magnetically coupled to where the sample is located unlike all other gravimetric system where the sample is directly connecteed to the sample.

So what does that mean?  Well normally we highlight the fact that we can subject the sample to a wide range of conditions - gases and vapours, corrosive environments, pressures up to 2000bar, temperature to 1600degC all while measuring mass changes - without any risk to the balance as it's in a separate chamber completely untouched by (or lets say "blissfully unaware of") whats happening to the sample.

But another major benefit of being magnetically coupled between the sample and the balance is that we can re-zero/tare and even re-calibrate the balance at any point during our analysis!

So the high resolution version that Rubotherm supply has a resolution of 1ug.  But because we can re-zero the balance (typically we advise re-zeroing every 10mins) that means we have absolutely no effect of balance drift on the measurements.  The balance does drift still of course, but we can re-zero whenever we wish to (a bit like re-zeroing before a one off mass measurement on a normal balance) so the mass change is precisely measured.

Hopefully the above pictures shows how easily we can do this zero point measurement to remove the influence of balance drift (the user can easily set this in the software)

So our 1ug balance is able to measure to +/-1ug.

Going back to the previous example of reporting 20ug mass changes during a continuous gravimetric experiment, with the Rubotherm system we're able to say that this mass change is 19ug up to 21ug so 20ug +/-5% - does that sound ok - probably better than the previous examples even though we have a "lower resolution balance"!.

Sorry for the long post but it's an area very often overlooked.  Next time looking at a gravimetric system don't just look at the resolution on the brochure ask about the balance drift likely......