About TFA

[HW Mueller]

An article appeared in LC.GC Europe (vol. 18, #5, May 2005, p. 290) , recently, which touches on this. The problem appears to be that hardly anybody has the time to rigorously optimize things, so guessing/hearsay/feelings seem to be a large part of TFA in HPLC.

[Kostas Petritis]

I do not quite agree Hans,

In the past I came accross papers that compare different mobile phases additives including HCOOH, TFA and HFBA for basic analytes (some of them were pretty old).

One recent example is the paper by Walcher et al. J. Chromatography A, 1053 (2004) 107-117 which investigates peptide and protein separation at different tempertures and at different concentrations of TFA, HFBA and HCOOH; stationary phase is polymeric monoliths. The authors also compare the sensitivity achieved with ESI-MS at with the above additives.

The authors conclude: "From a mass spectrometric point of view, 0.1% FA and 0.050% TFA were found to be essentially equivalent in terms of detectability and spectral quality for peptides. TFA should be preferred as additive in RP-HPLC-ESI-MS systems due to its significantly better separation efficiency and selectivity. Nevertheless, the variations in detectabilities between the different additives were generally rather small, contrasting the results of previous investigations (the publication cited here is from the same group).

Finally I have seen several applications notes trying to prove that the new low silanol acitivity stationary phases do not need TFA or TEA to achieve good peak shapes.

So I think the favorism towards TFA is a result of previous studies that compared the additives or from people that compared these to their lab once or twice and saw that TFA provides better results, but never published them. Maybe the new chromatographic columns do not need really need TFA in terms of peak shape but TFA always shows to provide better separations than FA... for our proteomic studies we use a mixture of TFA and acetic acid as high peak capacity is a priority for us...

[adam]

Following Up on an earlier post from Mark Tracy...

Why would TFA give better peak shapes for organic bases than something like phosphoric acid (if the additional ion pairing effect is only limited to peptides/proteins).

If it were true that TFA always works better for organic bases, this would suggest that - in the pharmaceutical industry - we should never use anything else (almost all pharmaceuticals have basic nitrogen functionalities). So I think this is an important issue to clarify.

Thanks Adam

[Victor]

In my experience TFA and phosphate buffers give a similar result for basic compounds in terms of chromatographic performance. Retention times are longer in TFA due to the ion pairing effect referred to.

No-one has stated the obvious yet; phosphate is not volatile and thus not used (hardly ever) by the mass spec people who are doing peptide analysis. They don't much like TFA either because of suppression effects compared with formic acid but they will use the stuff if they have to for the sake of better peak shapes. If you are not using mass spec, why bother with TFA since phosphate will do and will not give you the problems referred to above, nor will it give you baseline instability if you are doing gradients.

[Rob Burgess]

Excuse my ignorance on this matter, but can we make a distinction here please and answer Q1) below:

1) Is phosphoric acid (H3PO4 volatile) and therefore compatible with LC-MS detection?

I think we all know that phosphates salts aren't MS compatible. It appears there have been some conflicting statements during this discussion thread that have interchnaged the words phosphoric and phosphate...

Cheers

[Kostas Petritis]

Hi Rob,

Phosphoric acid is not volatile so it is better not to be used with MS. Actually MS instrumentation companies have made their sources more tolerable to non volatile buffers (i.e. Z-spray etc) but still eventually you are going to experience decreased sensitivities.

For a comprehensive table of volatile acids, bases and their corresponding salts as well as different concentrations of the above you may see at: Petritis et al. Volatility Evaluation of Mobile Phase/Electrolyte Additives for Mass Spectrometry LCGC Europe, Feb issue, 2002.

You can find the pdf file directly at:
http://www.lcgceurope.com/lcgceurope/da ... rticle.pdf

[adam]

Good reference!

Anybody want to suggest any others?

Thanks

[Mark Tracy]

Here is a comparison of TFA versus Formic Acid for the analysis of a peptide standard, Sigma H2016. Note that peaks 3 & 4 coelute in FA, and that all the peaks are broader and less symmetric. Still, the results are not all that bad for FA. Without any acid, peak 4 is too distorted to use.



This is supposed to be on the Dionex website, but [rant deleted].

Maybe I should get ambitious and do a full comparison of phosphoric, methanesulfonic, perchloric, etc for this same test set.

[Kostas Petritis]

That would be great Mark.

If I may suggest it would be great if you could also include HFBA, NFBA and acetic acid (it's easy to make suggestion when you are not the one doing the work )

From what you show, not only the FA does not resolve all the peptides, the peaks are broader and the separation window is smaller (at least it seems like). That would have given significally lower peak capacities...

[jitender]

This may be the useful link about TFA for the forum members.

http://www.westernanalytical.com/vydac/ ... tm#tfatips

What if sombody moves or changes the above linking url?

[tom jupille]

jitender: if Western Analytical moves that web page, the link will be "broken". Happens all the time (unfortunately).

[Victor]

Kostas,

Is HFBA o.k. for MS use? I have seen some opinions that this stuff sticks to some MS components and gives memory effects-presumably because it is less volatile than TFA and more hydrophobic-pehrpas it could adsorb/dissolve in seals or something. Have you found the same problem?

[Kostas Petritis]

Victor,

I have used up to PDFOA with triple quadrupole MS (I can cite you some references if you want) without any problems. You see the respective ion pairing reagentions if you switch polarity but these goes away if you run your instrument overnight with the heated nebulizer interface at 400 C (and water:methanol).

I heard (in this forum) that people with ion traps had some problems with HFBA, which is strange as I have seen several proteomics articles where people use routinely HFBA for peptide analysis with ion traps.

What I personally found to be terrible with MS, is triethylamine and there are papers out there saying the same thing... I had a collegue that was using TEA with his triphosphates as it was declastering his Na adducts and everytime I would take the MS after him I could do nothing without thorough cleaning!

So, I would suggest to try it. If you need some supportive references or people using it, to show to your supervisor (if this is a case) I can provide you with some.

[syx]

I found some abbreviations those I do not understand here…
Please let me know the long forms of HFBA, NFBA and PDFOA.

[Kostas Petritis]

Syx,

Heptafluoro butyric acid
Nonafluoro pentanoic acid (should have written NFPA and not NFBA)
Pentadecafluoro octanoic acid...