phenol gc method reproducibility issues

[SamC489]

I am creating a gc method for 4-tert-amylphenol, 2-phenylphenol, and 2-benzyl-4-chlorophenol (disinfectant formulations). I am hamstringed a bit, as it has been strongly suggested that I use a Restek RTX-BAC1 column, so that the lab techs in out sister lab wont have to change out a column they use for ethanol id and quantification, as they have little in the way of gas chromatography training, and even less in gc troubleshooting skills. They mostly run physical tests like specific gravity, viscocity and RI.

Anyway, I have made a method to ID the three phenols of interest, and get good linearity from 10-1000 ug/mL. When I test the reproducibility of the middle of the curve calibration standard, I get <2% RSD for the injections. So, good to go, only when I inject actual samples the variability seems to increase up to around 5% RSD.

I am using an agilent 7890 with a split injection of 10:1 and a 2 uL injection. Any ideas on how to decrease my variance?

[Consumer Products Guy]

Personally, I like to assay phenolic compounds by GC after trimethylsilyl derivatization on a non-polar column.

Perhaps HPLC with UV detector should be used for the disinfectant determination...that's what I'd do if it is a regulated product, which it proabably is.

EPA or FDA won't want to hear that lab techs have little training and skills either....

[chromatographer1]

CPG has given you very good advice, Sam.

Your GC test will probably require frequent replacement of the injection liner and column in order to meet the tight specs you seem to prefer.

HPLC makes very good sense. If you require a lower limit test, TLC may suffice.

Of course, the matrix is a prime factor and without that information, any advice is conditional.

best wishes,

Rod

[Peter Apps]

If standards give good repeatbaility and samples poor repeatbaility then you need to look at the sample prep - and if you are injecting sample as is, and the matix is water, I suspect that you have excessive vapour volume.

Peter

[SamC489]

Personally, I like to assay phenolic compounds by GC after trimethylsilyl derivatization on a non-polar column.

Perhaps HPLC with UV detector should be used for the disinfectant determination...that's what I'd do if it is a regulated product, which it proabably is.

EPA or FDA won't want to hear that lab techs have little training and skills either....

Sadly, I cannot control who gets hired to work in our labs. We have just aquired another company (that had no methods in place) that makes disinfectants and pesticide formulations and never had them QC checked for actives. I have been hired to create methods on a new HPLC and GC. I will move the analysis over to HPLC, which is what i was planning on doing, I guess I was just hoping for some insight on the specific reproducibility problem.

I increased the flow rate by 50%, and it seems to have helped. I plan on going forward with the method on GC, and begin new method development on LC in the new year.

The matrix is ethanol/water, and I am using methanol as a diluent.

[chromatographer1]

Given the additional information you have provided, you can see how Peter's comment is most helpful to you.

Your increase of flow (pressure) reduces the size of the injection evaporation 'cloud' which is most likely the cause of your reproducibility issue. The other change is to inject your sample slowly over 2 to 5 seconds. This will also reduce any backflash of the sample and help to reduce your RSDs.

Still underivitized phenols can be a problem over time as the injector loses its clean pristine condition.

A simple ACN water and acetic acid HPLC method should work fine on a C18 column.

best wishes,

Rod

[Consumer Products Guy]

A simple ACN water and acetic acid HPLC method should work fine on a C18 column.
Rod

Yep.


As I understand EPA/GLP, the EPA is primarily concerned that your finished product for sale has the correct levels of active, but doesn't really care about your batch sheets, QC, etc., just the GLP studies themselves.

Whereas cGMP for pharmaceuticals not only cares about the finished product, but every step along the way must be done according to regulations and documented, no matter how "perfect" the finished product turns out to be.