TFA vs HFBA

[grumbledook]

I am considering using HFBA instead of TFA for the separation some variants of a very basic protein.
Is there any downside? What about the absorption in the far-UV (detection limit, baseline drift)?
I have only found absorbance data for neat HFBA but not for dilutions in water/ACN. And can I easily switch
between TFA and HFBA containing solvents on the same system or do I have to flush system
and columns (C4) extensively in between?
Thanks in advance!

[annaand]

HFBA is a stronger ion pairing reagent than TFA. They both are strong acids (pKas=~0.2).
Switching to HFBA, I suspect, will change retention times of your proteins. They will be longer retained since HFBA is more hydrophobic than TFA. I would definitely recommend first cleaning the column and than equilibrating with a new mobile phase in this case.

[grumbledook]

Thanks for your comments, Anna. I think I know the benefits I can expect from using HFBA,
in particular improved peak shape for basic proteins. But apart from cleaning the system,
I was wondering if the UV absorbance at 214 nm is very different comparing HFBA and TFA, and
if would have to expect stronger baseline fluctuations with HFBA using a C4 column?

[tom jupille]

214 nm is used with TFA/acetonitrile mobile phases because is represents an isosbestic point as the UV spectrum of the TFA ACN charge-transfer complex shifts as a function of the ACN concentration. I would not expect that to be the optimum wavelength for HFBA. Since HFBA is not that commonly used, I don't know if it behaves the same way, but even if it does, I would not expect exactly the same wavelength.

You could probably find out in a couple of hours by preparing a set of HFBA solutions using different ACN/water ratios (e.g., 11 solutions at 5% intervals from 0% to 50%) and running UV spectra on them. The overlay the spectra and look for the isosbestic point (pivot point where the absorbance remains constant).

[grumbledook]

I know that 214.5 is the quasi-isosbestic point of TFA in water/ACN. But it's absolute absorbance is nonnegligible
at this wavelength already. I was just wondering if this is much higher with HFBA. Then I wouldn't even buy & try it.
(And we use detection at 214 also for quantification. Switching to a different wavelength is therefore not an option.)
And I observed pronounced baseline fluctuations not with C4 but rather with C18 columns when using TFA. Will this
be worse with HFBA, meaning do I have to expect strong baseline fluctuations with C4 columns already?

[tom jupille]

I understand the question, but I don't know the answer! As a guess, I would say that HFBA / ACN would not have the same isosbestic wavelength as TFA / ACN, and that the baseline issues would likely be worse. But this is only a guess; unless someone else can chime in with experience, the way to tell is to do the experiment .

[DJ]

I've found HFBA to be useful for separating synthetic peptides with Arg deletions. Most of the time TFA is sufficient to resolve Arg deletions, but there were instances when TFA system could not separate the deletions peptides no matter how shallow the gradient. In these cases, HFBA did the trick. More recently I've used mixture of 0.08% TFA + 0.02% HFBA, and this was also effective.

I think someone mentioned- i've experienced baseline fluctuations with HFBA. I have even used freshly distilled stuff without much better results. This is a problem when monitoring below 220 nm at high-sensitivity detector attenuation.. Depending on the sample (Y, W content), i've had luck monitoring at 280 nm, where, Unlike at 220 nm, the attenuation can be cranked all the way up without a great deal of baseline noise.

If this prep hplc, consider if hfba counterion is ok for subsequent application

[DJ]

Not sure if my imagination, but it seems longer initial equilibration with hfba was needed for reproducible results on C4 columns.

HFBA will increase the retention factor for peptides roughly according to the basic residue content. I wonder if this negatively impacts recovery?