Acetic Acid Method Development

[Alchemist5]

Hi Guys,

I am having to develop a GC method for residual solvents for a particular API. Of the six solvents, one is acetic acid. The client is adamant about having one method.

I am steering clear of headspace and am using liquid injection. The material is not water soluble to a great extent, so I am using DMF as a diluent.

This is a 1µL injection onto a 624 phase 30m column on an Agilent 6890. I am having to play with solvent concentration levels so as to be able to detect 1,2-dimethoxyethane (one of the other solvents) and have a reasonable test sample concentration. In adjusting the concentration levels, I am having to adjust my split ratio to maintain a reasonable response for the DME.

Now the interesting part: while doing all this, I have noticed acetic acid elute later as I lower the split ratio. The ratio went from 15:1 to 2:1 and I have seen almost a 2.5 minute shift in retention time. None of the other five solvents moved.

Now, not having much experience with acetic acid by GC, but knowing the 624 column is not the best for acetic acid, has anyone had this issue before? Let's just say for arguments sake that the 624 column is my only choice, what is the explanation for the retention time shifts?

I have tried a wax column and a G27 column and both gave me partial co-elution of 1,2-DME and IPA. This is no good. The 624 gave the best separation.

Burt

For those who are wondering, the solvents are: methanol, ethanol, IPA, acetone, 1,2-DME, acetic acid. API is marginally soluble in water, soluble in acid (but is unstable), soluble in DMF, DMSO.

[chromatographer1]

Burt,

When you say your sample is unstable in acid, does it produce artifacts which are volatile, and which might co-elute with any of your solvents?

If your answer is YES, and your client is adamant that he wants a single method, I would ask him if he is willing to pay for method development time even if no method is ultimately formulated.

If yes, then proceed. If development time is on your nickel, then I might say something else.

Waiting to hear your answers,

Rod

[gcguy]

If you are looking at acetic acid on a 624 can I suggest that you try THF as the solvent. This massively improves the peak shape for acetic acid. Just make sure that you run some blanks to confirm what peaks are part of the THF makeup. There can be a lot of stabilisers in THF, I have found the best one to use is the HPLC grade from Acros.
We use this set up for several sample types and have never seen a change in RT with a change in split ratio.

GCguy

[Alchemist5]

Burt,

When you say your sample is unstable in acid, does it produce artifacts which are volatile, and which might co-elute with any of your solvents?

Rod

Hi Rod,

The material is an antacid. When dissolved in an acidic medium, the material eventually crystallizes out. The higher the acid concentration, the faster it crystallizes out. Thus, the stability of the test solutions is low. Method development is on them, but its limited and they have a deadline for validation (I think they are focusing on the July 1st <467> revision).

This material is also not soluble in THF. I actually tried that yesterday after reading some archived posts on this forum. The shifting of acetic acid is not so much a problem as a question to why the retention time shifts with varying split ratios. Seeing 1,2dimethoxyethane at 100ppm is really my main issue.

Burt

[chromatographer1]

The amount of carrier solvent placed on column can affect the retention of GC analytes with a very strong retention, and, by golly, a free acid like acetic acid is a classic example.

This has been called a 'solvent effect' and affects analytes which most commonly elute AFTER the carrier solvent. As the solvent plug on column gets smaller (you change the split ratio) the amount of this effect can change.

The retention time for the analyte can also change, an effect which you have noticed.

If your drug matrix does not have a large amount of available protons, (water is an example) use of aprotonic solvents like DMSO and THF will often improve the peak shape of acetic acid (if present in non-disassociated form) in your drug solution.

DME is very volatile and is difficult to keep in a liquid solution if a direct injection method is attempted. Careful headspace analysis might be advised.

I believe you will save time and money if a separate method for measuring acetic acid is used. I have seen many hours of effort wasted in attempting to validate GC methods for acetic acid in API or final drug matrices rigorously (ie, with full confidence).

The particular matrix must be just right or you can be caught in an embarrassing situation later. (reputations or positions can be lost)

NMR is recommended. If N/A then ion chromatography is the next best choice.

Remember, if it were easy, anyone could do it.

Good luck and best wishes,

Rod

[Willy]

Hi Burt,

Ive some experience doing that type of analysis. Last time we tried to validate a GC method for acetic acid we did a derivatization (methylation) with Boron Trifluoride in Methanol, basified with ammonium hydroxide and water, adding the water at the final step. One thing is when adding water the reaction takes place and you need to analyse the samples and standards within 3 hours because it is a reversable reaction. I personally did the quality control analysis about 10 time and it was working pretty well. If you try or if you have any questions just reply in here and we could discuss about it if it isnt working.

Have a nice day!

Willy the GC

[Alchemist5]

We were able to get acetic acid to be reproducible on a G27 column. However, peak shape was not very good (fronting). We finally convinced our client to move acetic acid to HPLC! Problem solved....or at least averted. Thanks for all your input guys.