Amonium Acetate and Formate Peaks in Gradient Profiles

[Yorkie31]

Hi,

I just wondered if anyone had done any work or had any idea with this problem.

On numerous occasions now, when using either amonium acetate or formate as the aqueous buffer in a gradient system (typically with acetonitrile as the organic) a peak appears that always elutes at the exact RT of the main band?

This is not just coinsidence, as this occurs with numerous methods and numerous different drug compounds and we only see this in gradient elution, not isocratic.

We have contacted the supplierss and looked at purity of the salts. We cannot use phosphate as we quite often need an MS friendly method. The only temperary fix we have found so far is to filter the aqueous phase through a LC column (even though it is vacuum filtered through a 0.45µ membrane prior to use).

Any ideas, theories, answers??

Thanks for any help
Anthony

[JA]

For clarification, you are stating that you see interfering peaks (ie. at the same RT as your main band will be) in blanks run before samples? I want to know how this interference knows what your analyte's RT is going to be, especially in different methods.

If not, this could be indicative of a problem in the application and if more details are provided the board may be better equipped to offer suggestions.

Regards

[Mark Tracy]

Try making your buffer from ammonium hydroxide and acetic acid. Since the problem can be cured by passing it through a C18 column, most likely it is some minor impurity in the mobile phase. The source could be your buffer salt, your glassware, your filter apparatus, and so on.

[syx]

The only temperary fix we have found so far is to filter the aqueous phase through a LC column

Same as my problem last time, I still found that it is the only way to purify so far. You may read “Haunted Peak”.
http://www.sepsci.com/chromforum/viewtopic.php?t=3184

[Yorkie31]

I want to know how this interference knows what your analyte's RT is going to be, especially in different methods.

This is unfortunately what I am trying to find the answer to. We can adjust the buffer concentration, the gradient time etc.. the peak in blank injections will always move the same amount as the main band and so remain underneath.
It is not carryover either, as the peak can occur on a brand new column that has had no drug down it.

Try making your buffer from ammonium hydroxide and acetic acid.

We have tried this to no affect.

It does seem that the only way around is to filter through a separate analytical column.
One thing that has just been suggested to me is to insert an on-line C18 filter prior to the pump. Has anyone tried this?

[tom jupille]

If you haven't already done so, run a series of three "dummy" gradients (no injection) with different pre-equilibration times. Check if the size of the interfering peaks increases with increasing equilibration time. If it does, then you have confirmed that what you are seeing is coming from the "A" reservoir and building up on the head of the column.

If you prepare your standards and mobile phases in the same hood (or even in the same room), there is a finite chance of cross-contamination. In fact, that's the only explanation I can think of compatible with the observation that the impurity somehow always manages to match the retention time of your analyte. If that is indeed the problem, the cure is to prepare the mobile phases elsewhere.

"Band-aid" fixes (which address the symptoms rather than the cause) include installing a C18 guard cartridge between the "A" pump and the mixer (this only works with high-pressure-mixing systems) or filtering the "A" solvent through a C18-impregnated membrane (for example:
http://www.3m.com/Empore/pdfs/lcgc2-72-05e.pdf )

[Yorkie31]

We have already tried running "zero volume" injections to simply look at the gradient profile. The peak in question still appears.

We have not looked at SPE disks however and maybe something we can have a look at, thanks!

[HW Mueller]

It seems to me that one needs an extraordinry talent to always contaminate solvent A with sample. Could it be that your peak is a system peak that has to do with to (tm) and none of your analytes are retained?
How about Tom´s suggestion of varying pre-equilibrium? Or collecting the peaks and analyzing them with another method?

[AA]

Try This:

This is how I purify buffers when I run into this problem. Make your buffer as usual. Filter your buffer through 6 grams of Oasis SPE material. You can buy 6 gram cartridges, break one open and dump the material into filter appuratus (I use a sandwich of whatman #1 on top and a 0.45 um membrane underneath for the filter). You must wash the Oasis material first with methanol (200 mL) and then 100 mL of water which you then discard. Then filter your buffer through the bed. I do this with a vacuum filter appuratus (with a 1 liter funnel) so it does not take too long. For very severe contaminations you might have to do this process a couple of times. The good news is you can re-use the Oasis bed, just wash with methanol and rinse with the water.