Agilent 1100 detector linearity test

[bgiles]

Does anyone know the performance qualification test recommended by Agilent for the Agilent 1100 G1365A MWD detector. We are currently using the waters test that uses propylparaben from 5mg/L to 30mg/L. Curious if this is anything similar to what agilent recommends.

thanks

THE DUDE

[shaun78]

I thought they used various caffeine standards and a restriction capillary. Of course, that tests both injector and detector ... but it should work for you.

[bgiles]

We have a test, we are using the waters detector linearity test. It is a propylparaben solutions 5-30mg/L. The problem is that it is failing this test. At the high level we are getting 2.3 Au. This is above the agilent stated maximum linear range of 2.0AU. It is failing at the high end. The waters specs are also a maximum AU of 2.0 but it does not fail. Just wondering if the detector is failing due to malfunction or failing due to inherent limitations or wrong PQ test.

Do you know the Au of the highest caffine standard used in the Agilent test?

[Consumer Products Guy]

Either should work fine; write your own SOP and follow it. As bgiles stated, this can serve as your reproducibility (injector, detector, retention time) as well. We've been outsourcing our yearly checks to Agilent. On our DAD units they use restriction capillary instead of a column, caffeine standards kit #8500-6917, Method OQRESP.M, flow 1.000 ml/min., solvent A is HPLC grade water, 273 nm detection, injection size 20.0 ul, 40 C, sequence OQRESP.S, which is one injection of each of the five vials. Specification is 0.5 ug/ml +/- 5%, 1.0 +/- 3%, 5.0 +/- 2%, 25 +/- 2%, 50 +/- 2% .

[rnelson]

The previous post states the exact Agilent procedure. I have some Agilent 1100's and if I'm interpreting the acceptance limits properly, I feel this test is insufficient.

The only acceptance criteria I see is correlation coefficient, and I don't think that calculation alone tells you enough about linearity. I'd prefer to see a response factor rsd calculation similar to ASTM. I can show you data that would come nowhere close to meeting the ASTM definition of linearity, but would still meet the Agilent limits of 0.9990. USP requires more than corr coeff when performing linearity studies as well.

I previously tried to get a discussion started on this topic but didn't get many replies. I'm curious how different companies actually perform detector linearity. I'm trying to move toward something similar to ASTM, but it's going to require a lot of time and effort, so it's been a slow process.