Challenge for my HPLC system

[Karl Fieldman]

Dear chromatographers,

Probably this post has been on the forum many times. But then again….

I’m looking for a very critical mixture to test the performance of my HPLC system. The mix should not be sensitive to column material (I will just use a C18 column). So no amitryline or compounds like that. Important matters would be mixing performance, gradient performance, detection linearity /sensitivity. But also sticky compounds which could stick to the injection needle or rotor seals. It should be a challenge for the system.

Detection with UV.

I was thinking about a mixture containing both polar and non polar compounds.

Suggestions are welcome.

Karl

[Peter Skelton]

http://chromatographyonline.findanalyti ... rticle.pdf

That should be a good place to start with most of your questions

[Karl Fieldman]

Thanks for your comments. I know these tests, but I’m more looking for a mixture that can give an impression (or idea) of all the single tests as described in the article.

For clarification, I’m looking for 1 challenging mixture that can be analyzed under gradient conditions and will tell something about my system.

[unmgvar]

Do you mean something like this companies prodcuts?

http://www.v-kit.com/

[Karl Fieldman]

Dear unmgvar

These mixes are interesting, but more suited for complete validation. I'm more looking for 1 mix with polar and non polar compounds included. Just to test the system if you don't know the status of the machine.

For columns you also have test mixes like SRM 870. A few coumpounds that desribes the column already pretty good.

[unmgvar]

Dear Karl

HPLC systems are periodically validated to show that they function correctly,
it is done using PV/PQ protocols that use the kits shown in the website i gave.

when all is done the system passes the test, but more important those tests and kits are capable of showing if something is wrong and not to specs and thus to know what to fix,
in most cases that i know of those procedures for the HPLC tend not to use a column but a long piece of tubing that creates volume delay and back pressure
a column after all is a variable that could influence the test and like you state could influence results.

application wise it is always good to have a simpler system suitability test that is specific to your specific method and will allow you to check the system in combination with the column and method parameters
then you generally check for
RT
RSD
K'
resolution
tailing


PQ of a system will need of you to test for:
pump- flow and gradient. you could show RT precision using a compound but this does not show flow precision and reproducability

autosampler- you need to show RSD mainly, carry over, you do not need a column again here as well,
carry over test is actually a joke in a PQ, it could happen for several factors, some of them being system wide contamination, or old spare parts like rotor seal and injector seat seal needing replacement (they generally are in a preventing maintenance routine)

column oven, you need a thermometer here only. these days to show a good performance you should test at several points for a long period of time, at least 3 and also show repeatability. that takes a few hours to do

UV- linearity, drift & noise WL accuracy- of those the tricky one for you is the linearity. especially these days with all the new UV with new ranges. the old ones go to 1.5AU, other 2 AU or 2.5AU or 3 AU
or you could simply limit yourself to a range that works for all and limit your methods to it


So in general you do not need a column and you should actually avoid it when you test your HPLC