Thread: Advanced Elite Labs ** Test C Results Posted**

You never get exactly the same result for multiple tests. Even if you put the same sample into the machine, the result will come out a tiny bit different each time.

First, you pipette an aliquote (a small part) of the oil into the container that can be used in the machine used for testing.

I personally use 10 microliters and Gilson Microman pipette. The margin of error is about 1% I believe. Microman is special pipette for oils and other viscous liquids.
If a different pipette (a classic one some might have seen or used) were to be used, the margin of error would be much higher (I've seen 10% and more).

Then 990 microliters of chlorinated solvent is added (methylene chloride or chloroform) to dilute the sample. This also carries some tiny error..

After that, the vial is put into the machine, where 0,5 microliters (margin of error is about 0.5%) is automatically injected into pressurized liquid and the analysis is carried out.

There are also many other sources of very tiny errors and it's impossible to avoid them. I try to keep the margin of error within 2% and I believe that I'm managing to do so.

It's not really possible within reasonable means to provide more precise results.



I usually tell people that they ought to expect the margin of error of maximum 5% with my analyses.

No need to apologize at all. I'm happy to share the knowledge.

When pipette is used to transfer 10 microliters of solution, which is then put in a different container, it doesn't transfer exactly 10 microliters, but anywhere from 9.85-10.15 ul (my exact pipette). That's accuracy, which, funny enough doesn't really matter at all.

We are interested in precision, as we are not doing absolute quantitation, but relative, to the standard. The pipette that I use is precise within 0.05 ul, or 0.5% ( looked it up, less than I expected, https://www.pipettesupplies.com/asse...2_(Gilson).pdf page 15, model M10, 10 microliters ). So with one sample you can get 1% variation maximum, only via this. This error can't be avoided in a reasonable manner, whatever you do.

The said pipette works on a principle called positive displacement. If you used normal pipette, even expensive one, on oil, it would yield MUCH higher error. This shows just one of examples why normal pipettes can't be used to very viscous liquids, such as oils - Air Displacement vs Positive Displacement Pipetting - YouTube


Regarding reference standards, they are trouble all by themselves, however directly answering to your question - if you measure AAS oil sample, you expect concentration between 10 mg and 500 mg, with very rare outliers. So you make reference standard of 10 mg/ml concentration, 100 mg/ml concentration, 250 mg/ml and 500 mg/ml concentration. If the plot of peak area of the studied compound vs concentration passes through all points in a linear fashion, then you can safely assume, that you can put the unknown on the plot and that'll determine the concentration.

If you for example used only 1000 mg/ml standard and tried to analyse methyltren 1 mg/ml oil, you'd get skewed results.


Margin of error can be decreased a lot by weighting everything on microscales - it's called gravimetric analysis. Measuring stuff by mass is much more accurate than any volumetric (measuring by volume) based measurement. However, for this we don't really need THAT much accurate analysis and it would take amounts of time so massive, that I wouldn't be able to provide the service at a reasonable cost. Also, nobody is really interested in mg/gram of oil, but rather mg/ml )


I apologize if I'm not making much sense, I'm not really used to explain this stuff even in my mother tongue, much less English.

Originally Posted by janoshik

No need to apologize at all. I'm happy to share the knowledge.

When pipette is used to transfer 10 microliters of solution, which is then put in a different container, it doesn't transfer exactly 10 microliters, but anywhere from 9.85-10.15 ul (my exact pipette). That's accuracy, which, funny enough doesn't really matter at all.

We are interested in precision, as we are not doing absolute quantitation, but relative, to the standard. The pipette that I use is precise within 0.05 ul, or 0.5% ( looked it up, less than I expected, https://www.pipettesupplies.com/asse...2_(Gilson).pdf page 15, model M10, 10 microliters ). So with one sample you can get 1% variation maximum, only via this. This error can't be avoided in a reasonable manner, whatever you do.

The said pipette works on a principle called positive displacement. If you used normal pipette, even expensive one, on oil, it would yield MUCH higher error. This shows just one of examples why normal pipettes can't be used to very viscous liquids, such as oils - Air Displacement vs Positive Displacement Pipetting - YouTube


Regarding reference standards, they are trouble all by themselves, however directly answering to your question - if you measure AAS oil sample, you expect concentration between 10 mg and 500 mg, with very rare outliers. So you make reference standard of 10 mg/ml concentration, 100 mg/ml concentration, 250 mg/ml and 500 mg/ml concentration. If the plot of peak area of the studied compound vs concentration passes through all points in a linear fashion, then you can safely assume, that you can put the unknown on the plot and that'll determine the concentration.

If you for example used only 1000 mg/ml standard and tried to analyse methyltren 1 mg/ml oil, you'd get skewed results.


Margin of error can be decreased a lot by weighting everything on microscales - it's called gravimetric analysis. Measuring stuff by mass is much more accurate than any volumetric (measuring by volume) based measurement. However, for this we don't really need THAT much accurate analysis and it would take amounts of time so massive, that I wouldn't be able to provide the service at a reasonable cost. Also, nobody is really interested in mg/gram of oil, but rather mg/ml )


I apologize if I'm not making much sense, I'm not really used to explain this stuff even in my mother tongue, much less English.

No need for the apology at all sir. I completely understand.
Your essentially plotting at said mg/ml and using trend line to map that concentration to the listed plots.

It makes sense actually, I do something similar with non oil based materials using a similar process using a pre-mapped trend line plotted by percentage.

I appreciate that level of clarity.

M