Measuring Autosampler Carryover

[johngibbons]

A lot of HPLC manufactures are now claiming ultra low carryovers of less then 0.005%. Yet they refuse to publish the method used and the actual data to back it up. Is there are standard method using a appropriate analyte for measuring carryover at this level?[/b]

[danko]

Carry-over is not a standard/universal test that is applicable to every situation. A system could exhibit zero carry-over with a certain compound/analyte, injector-wash-liquid, washing procedure (e.g. time or volume) eluent, injection volume etc. etc.
So…you’ll have to design your own test procedure that is relevant to your needs (e.g. relevant compound etc.)
Here is a kind of starting point that you can build up on:
1. Remove/shortcut the column.
2. Set the flow on.
3. Inject a relevant volume of the sample solvent that you use. Might be several times, until you see no peak at the sensitivity level you’re interested in.
4. Then inject your sample solution and record the peak area.
5. Inject again the sample solvent and record the peak area – if any.
6. Divide the peak area of the solvent/blank peak by the peak area of the sample peak and multiply by 100.
7. The resultant value is percentage carry-over that is relevant to this particular method of analysis.

Best Regards

[lmh]

danko's right. The obvious approach is to include a blank runs after your highest calibration points, and look to see whether the peaks in these blanks are significant compared to the lowest level samples with which you expect to work. If they aren't, you are probably fine.

[sam.pedraglio]

danko's right. The obvious approach is to include a blank runs after your highest calibration points, and look to see whether the peaks in these blanks are significant compared to the lowest level samples with which you expect to work. If they aren't, you are probably fine.

Completely agree. This is exactly the same procedure we apply in our validation protocols.

[Markus Laeubli, Metrohm]

You might inject the solvent a second time. This actually would show you any contamination in the solvent.

A contamination in the solvent could otherwise idicate a too high carry-over.

[shahis77]

I AGREE WITH IMH.....INJECT A BLANK AFTER YOUR HIGHEST CONCENTRATION ....