Chromatography Forum

Hi all,

I'm working with purification of alkaloids by HPLC and LCMS. I have free usage of a standalone HPLC with UV detector, but have to pay for time on the LCMS.(LCMS also has quite a poor UV detector, so it is easier for me to monitor UV on the standalone) Due to this, I have been running my extracts on the HPLC initially using a standard water/ACN gradient with 0.1% TFA, with expected results and clean peaks on a c18 column, to visually assess the extract and position/quantity of peaks.

However, switching to the LCMS, running the same column and gradient, just with 0.1% formic acid, I was seeing broad and tailing peaks, with none of my anticipated peaks matching up in retention.
To verify what I was seeing, I made my mobile phases with formic acid and ran my standard on the standalone HPLC. Lo and behold, what was a beautiful sharp peak in the TFA gradient has become ugly and broad with FA, shifted almost 4 minutes in retention time. (This is a ~25 min method on a 250x4.6mm analytical c18)

Nowhere in my research have I found any mention of a shift this drastic when switching from TFA to FA, and anyone I have asked has said the retention times should be relatively the same. What could be causing such a huge shift? Would it be the ion-pairing effects of TFA? The alkaloid of interest is quite hydrophobic but quite standard in its composition as far as alkaloids go.

Much appreciated for any help!

TFA and Formic Acid are completely different acids with completely different properties and pkAs. *** A difference of 3.5 pH UNITS !!!! ! If used at the same concentration, then it would in fact be expected in most cases to see a change in peak shape and/or retention time if the acids were swapped.
"Common pKa Values for ACIDS & BASES used in HPLC and LC/MS Method Development >"; https://hplctips.blogspot.com/2015/12/c ... -used.html

Briefly: Formic acid is rather weak acid (esp compared to TFA), but helps many types of samples ionize well (which is VERY important) under LC-MS ESI conditions whereas TFA acid (ion pairing acid) often causes complete contamination of the HPLC and detector flow path, plus suppresses ionization for many (not all) samples types (so it not normally used on a MS system unless it is dedicated to just one sample type, because it will be contaminated afterwards).

yes, if you're working on alkaloids TFA will act as an ion-pair reagent and change the retention time quite substantially. If you're struggling to get decent chromatography with formic acid (which understandably you need to do, if you want to use LC-MS) consider using one of the modern columns that can cope with alkaline conditions (Waters BEH can, I believe, and Phenomenex's EVO and probably others besides) and trying mildly alkaline conditions. This will avoid protonating the amines/nitrogens in your alkaloids, which achieves something a little bit like screening them from interactions by sticking TFA all over them. It's not the same, so don't expect the same retention times, but you might (just might) get better retention and better peak shape. TFA is lovely for HPLC but a pain for MS, and in my experience, the famous post-column TFA-fix addition doesn't help one jot.

I think the effect of TFA as ion-pairing reagent for small molecules is non-existent, you need to have at least C4 acid to get any retention gain. Your compound needs to be basic in nature. I don't know if your alkaloid is basic ( I assume it is based on the discussion). If compound is basic indeed, I would suspect that you are seeing an interaction between your alkaloid and residual silanols. TFA is a much stronger acid and it creates low enough pH to suppress most of the residual silanols. Formic acid is weaker and difference in pH is probably 1.5-2, so some of the silanols might be ionized and you see the effect of overloading these residual silanols.

There are much better solutions for your analysis. Contact me through our website if you want to discuss how to address this issue.

I think it's worth remembering a few things here.
(1) The retention time depends on the analyte, and its state of ionisation, which depends on pH. How you achieve that pH is irrelevant. The basic groups on an alkaloid are going to be almost-fully protonated at the pH of 0.1% formic acid or 0.1% TFA, that won't make any difference.
(2) But formic acid has a relatively low ability to screen those positive charges, so they form very hydrophilic spots on the analyte, which will tend not to associate with a reverse phase column. TFA on the other hand definitely does. Just Google. Vlad, you're in a serious minority on that one: see for example https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2744697/, which compares TFA to phosphoric acid and to longer-chain ion-pair reagents; TFA already clearly has a large effect; see also Sigma's viewpoint https://www.sigmaaldrich.com/deepweb/as ... 10142h.pdf and many others. Of course the majority might be wrong, but for the moment I'm inclined to believe there's some truth in what they write.
(3) Yes, alkaloids interact with residual silanols, but that's why column manufacturers have spent the last couple of decades desperately trying to improve silica-based columns to avoid such interactions. The effect is fairly small in modern columns.