Page 1 of 1

Best way to test column efficiency for new columns

Posted: Mon Feb 18, 2008 9:02 am
by LCGUY
Hi,
What is the best way to test the efficiency of the new columns. My manufacturer use Uracil, toluene, biphenyl, naphthalene as test mix but advice me that

"Generally manufacturers use most ideal conditions to get maximum possible number of theoretical plates. They have their system in best running conditions, keep the connectors shortest possible, use minimum volume flow cell, check the dead volume to minimum possible....etc. Hence many a times the efficiency check test conducted in users lab conditions don't match the plate numbers specified by manufacturers. Hence its better to go for system suitability test and check the column efficiency. If the column is passing SST for your application, it should be used irrespective of the results obtained by benchmarking the efficiency with that of provided in test chromatogram"

Can I have comments on aforesaid statement?
What is the purpose of each using Uracil, Toluene, Biphenyl, Naphthalene as test mixture in RP columns?


Thanks in advance...

Posted: Tue Feb 19, 2008 4:47 pm
by tom jupille
Can I have comments on aforesaid statement?
Basically, what is says is that if the column passes your system suitability test, it is, ipso facto, suitable for your analysis. Makes perfect sense to me.
What is the purpose of each using Uracil, Toluene, Biphenyl, Naphthalene as test mixture in RP columns?
There is nothing sacred about those particular compounds; other manufacturers may use different ones. That particular set is convenient because all are easily detectable by UV, cheap, available in good purity, not too toxic, and typically show minimal tailing.

Posted: Wed Feb 20, 2008 1:33 pm
by sfe-co2
Uracil is basically an unretained species, so it can be used as a void marker. Some people use thiourea for the same purpose, although I believe thiourea is still slightly more retained than uracil.

Tom, aren't the other compounds chosen for the range of structural differences? That is, they have different chemical properties, such as polarities.

Posted: Wed Feb 20, 2008 3:43 pm
by tom jupille
Tom, aren't the other compounds chosen for the range of structural differences? That is, they have different chemical properties, such as polarities.
I'm not the one who chose them! :lol:

Actually, they're not that different: aside from uracil, they're all aromatics with no hetero atoms (which is why they tend not to tail). They do differ in polarity (maybe "hydrophobicity" would be more appropriate in this context), so they should cover a reasonable range of k'. That information would let you compute the amount of extra-column band broadening and establish a "true" (whatever that means!) plate number for the column.

But again, I don't think there's anything magic about those particular compounds; many other sets would do just as well. These just happen to be convenient.

Posted: Wed Feb 20, 2008 9:23 pm
by sfe-co2
Yep, there are many compounds suitable for testing purposes. I recall including benzamide (void marker, although I still think uracil is better) and propylbenzene, but excluding biphenyl and napthalene. I also agreed that compounds are chosen by column manufacturers so tailing will be minimized....so no nitrogenous type compounds, for instance.

Posted: Thu Feb 21, 2008 6:50 am
by bookoon
Hi LCGUY

The following link might help you in understanding why those particular compunds are used.

http://www.lifescience.ca/pdf/870.pdf

Posted: Thu Feb 21, 2008 10:53 am
by AdrianF
This topic has been discussed before: The best test is one using the compounds you are going to test using the conditions for that compound.

If you keep a record of the results you can compare it to a new column of the same type. You can also keep a track of its performance through the life of the column.

I think you should trust the manufacturers test that comes with the column

Posted: Thu Feb 21, 2008 11:55 am
by AdrianF
An example of a recent discussion of this topic:

http://www.sepsci.com/chromforum/viewtopic.php?t=7366

Posted: Fri Feb 22, 2008 6:49 am
by LCGUY
Hi Guys,

Thanks to you all for your inputs and excellent insights. Now I know that its better to go for SST rather than spending my scarce time on testing columns as per manufacturer test report.

Manufacturers HAVE TO BE trusted and quality conscious or else they wouldn't be in this business for long. You have to trust your manufacturer by default.

Oops! Dropped your package.

Posted: Fri Feb 22, 2008 9:35 pm
by HUH?
The one reason I repeat the manufacturers test mix is to ensure that the column is not damaged in transit. This has become less of an issue for analytical colums than for Prep columns. The efficiency numbers we calculate have always been very close to the manufacturers efficiency report. Happy HPLCing