2% Formic Acid destroying my column (dead space)

[Anna01]

Hi

I am doing an online extraction on Oasis applying a washing protocoll with 2% FoAc in 20% MeOH(actually MeOH with 2% vv% FoAc and H2O with 2 % vv FoAc mixed by the pump). I measured the pH of my H2O to 2.0.

Now...all of a sudden I got very broad peaks...split peaks! (the last sample yesterday was nice, the first sample this morning was i disaster!) Changing the analytical column fixed this, so I opened the analytical ordinary C18 column. The inspection revealed that the packing material had "collapsed", leaving a large dead space. The column is quite new...and should be compatible with down to 2.

This is what I am doing:...the samples is applied to the extraction column in pure water, then a washing step of 30 sec with 20:2:78 MeOH:FoAc:H2o is applied. The extraction column is then reequilibrated back to 0,1% FoAc in H2O for another 30 sec (flow 2000uL/min), before mobile phase shifts direction and is eluted on the analytical column.

Now...can the formic acid explain this?

I doubt it, but I am not sure. Running samples without such a strong FoAc component has never given such problems.

To me it doesn't seem likely because the pH isn't low enough and it the formic acid is quite diluted(or washed out) before entering the analytical column.

Btw, this has happened now on 3 columns (of different brands, but all C18).

Does anyone have a clue what's going on? (I think i have ruled out pressure surges)

Any help would be appreciated!

Thanks,

Anna

[Uwe Neue]

I am sure that this has nothing to do with the formic acid.. An analytical column is chemically stable under these circumstances, plus when you elute onto your analytical column, your formic acid concentration is much lower.

I think that the problem is either the physical stability of your analytical columns, or a significant contamination that your sample preparation protocol does not remove. Did you see a large increase in backpressure, before the column failure ocurred? How was the analytical column stored just before it failed?

[Parsival]

What kind of sample solution are you working with?

[Parsival]

What kind of sample matrix are you working with?

[Anna01]

The matrix is serum.

There is definitely a dead volume in my column, and I can eliminate the problems by adding more C18 stationary phase to fill this gap at the inlet of my column.

I wonder why this is this occuring...3 times! With different columns?

[danko]

Too high backpressure? Especially pressure spikes under column shift.
Forget the pH – low pH won’t “consumeâ€

[Anna01]

OK...The column operate at about 20-80 BAR and flow rates about 200-300 mikroliters per minute.
I have looked for such pressure spikes, but haven't confirmed this.

Are you suggesting I should do a pressure gradient program every time I change column? And when I start the instrument every day?

[danko]

You’re right – the pressure is far from excessive but sudden and repeating changes from 1 to 80 Bar could cause void volume/headspace.

Are you suggesting I should do a pressure gradient program every time I change column? And when I start the instrument every day?

Not really. You’ll get too much chromatographic variation e.g. Retention time and peak shape if you do gradual flow rate increasing. But you might try a guard column – let it die once in a while, instead of the analytical column. I’m not the strongest believer in guard columns but it’s a better option than destroying analytical columns. It may cause slight efficiency reduction, so you’ll need to test and evaluate this option.

Best Regards

[Uwe Neue]

I am not sure if I understand what you are saying. Did the column backpressure increase over time from 20 to 80 bar, or is this increase in pressure something that is happening in a single run? My question was related to contamination that builds up on the column. If this is the problem, the sample prep procedure is not yet sufficiently optimized. Considering that the problem occurred on different C18 brands (which I assume also come from different suppliers), I would be more inclined to look at the sample prep procedure as the cause of the problem.

[Anna01]

No...I am running a gradient program, during which the pressure increase a litte (mobil phase viscosity, i presume). Normally in the range of 40-70 bar,