Peak splitting (mulptiple peaks) in Ion-Pair RP mode

[2152812]

compound is aliphatic amine containing 4 primary amine groups and two tertiary amine group in structure of MW ~300Da
mobile phase: Water+0.1%TFA
Column: Synergy Hydro C18
Injection volume: 10uL

We start seeing two peaks when injected samples at concentration higher than 0.5g/L. The peak area ratio of these two peaks changed as concentration increased. (Compound is a reference std, >98% purity)

We learned that either higher TFA concentration in mobile phase or higher injected amount resulted in additional peak which is well separated from the "main" one.

Any idea about peak splitting phenomena in ion pair RP LC, especially when TFA is employed.

Your input is highly appreciated

SOS

[Rob Burgess]

TFA is a relatively weak ion pairing agent compared traditional reagents such as SDS or alkyl sulphonic acids.

I suspect your dissolving solvent is >> in strength than your starting mobile phase conditions.

[HW Mueller]

Besides knowing precisely what you are injecting it could also be instructive to know whether the "additional peak" is the same one for both described conditions. Could it simply be a system peak? Does the area of the main peak go down when the "additional peak appears or increases?

[2152812]

Rob: sample is dissolved in water

RW: main peak dcreased as second one increased. Second peak is NOT system peak. Both are 1 min away and 2 min after "void" volume

Column was 250x4.6mm

One more thing I learned that increasing column temp from 30 to 50oC totally suppressed the growth of second peak, i.e. only single peak was observed

Thanks,
SOS

[HW Mueller]

On this temp. thing... how do you know that the "additional" or "second"? peak didn´t disappear into the void system? Also, we don´t know what you kept constant and what you changed for your different experiments (for instance amt. of injection...).

[2152812]

when I got one peak that means second peak disappeared into first one.
RT for this only one peak is 1.5min greater than "void volume" RT.

Peak splitting happened when injected amount increased or TFA concentration is higher than 0.05%.

I wonder the "polyamines" type of molecules like this one behaves in Water 0.1%TFA mobile phase? Different conformers? Isomers are excluded in this case.
Thanks,
SOS

[Bruce Hamilton]

What happens when you inject sample in mobile phase, rather than water, and when you inject larger volumes of the dilute sample, both in water and mobile phase?.

I'd suspect that the acidity of TFA is responsible for the magic, and column temperature is just accelerating a slow interaction. In my limited experience, conformers appear with long retention times and aren't concentration dependent, but can be solvent dependent.

Bruce Hamilton

[ym3142]

I'd like agree with Bruce, meaning it seemed to me that the sample concentation was relatively high, the one or more amino groups in the molecule can be protonated inside the column, but the diluent was water instead of mobile phase, causing the differently protonated species came out of colmn at different times. Good Luck.

[2152812]

how can you explain that adding TFA to the sample make peak splitting even more (and peak tailing too).

I try to understand why the temperature effect (T increase) is similar to the one seen by lowering injected amount in terms of making two peaks becoming one which is what I would expect to see from pure compound...
Thanks,
SOS

[Bruce Hamilton]

how can you explain that adding TFA to the sample make peak splitting even more (and peak tailing too).

I try to understand why the temperature effect (T increase) is similar to the one seen by lowering injected amount in terms of making two peaks becoming one which is what I would expect to see from pure compound...
Thanks,
SOS

Explanations cost more, whether correct or incorrect, especially when I don't have specific experience of the molecule . Why not try the suggested experiments and see if they provide more insight?. Because I'm a simple person, I prefer, whenever possible, to inject samples disolved in the mobile phase.

Increasing temperature can speed up or even initiate interactions, increasing acidity can also change the chromatographic behaviour, especially of a molecule with some of the functional groups on yours.

We can apply all sorts of magic to make one peak become two, and two peaks become one, sometimes deliberately, sometimes accidently, even without having all the information about a molecule's solubility behaviour and pKa etc.

If you want to understand the process, I'd suggest performing controlled experiments, varying column temp, mass/concentration injected, TFA concentration, etc etc, but keeping other parameters the same. You would then understand the critical paremeters.

I would want to be assured that the mobile phase has enough buffering/ionic capacity to provide consistent chromatography before reporting any results, but YMMV.

Please keep having fun,

Bruce Hamilton

[HW Mueller]

This is another case of needing to make sure that the peaks in question are really due to the injected substance. You don´t tell us how you know that the second peak "disappears into first one", etc. etc.
The only way I can imagine that you separate an amine from its ammonium counterpart is if you have a breakthrough of parts of an injected amine, but then it would be at ~ tm.

[ym3142]

if you are not willing to perform all these investigations suggested by people here, this is another suggestion either keep low concetration(e.g. 10mg/100mL)low inject volume (e.g. 20ul) or use large column( 20mm ID). I am sure you should be able to solve the problem. Good luck.

[2152812]

thanks everyone for valuable inputs. I appologize for not responding to your questions sooner. We did conduct numerous experiments systematically (varying one parameter at the time, etc) before I posted the questions. What I shared with you are more or less summary of what I could observe from the data. However I couldn't understand the elution/splitting phenomena in this case.
e.g. how do I know second peak dissapeared into first one: by calculating total peak area since the response factor is likely the same (amine detected by postcolumn derivatization with ninhydrin). Total peak area remained proportionally with injected amount, the ratio changed as we increased injected amount and/or TFA.

We have been doing pretty much like YM3412 suggested: inject less into larger column at elevated T like 50oC.

YM3412: what is your estimate for max loading of column of size like 250x4.6mm.

HW: I did not get what you meant by "separate an amine from its ammonium counterpart is if you have a breakthrough of parts of an injected amine, but then it would be at ~ tm"

Thanks again,
SOS

[ym3142]

Did you try the conditions I gave in my last post?
the numbers I provided last post should be OK. It is better you figure out yourself such as inject 50mg/100ml, 1omg/100mlm, 5mg/100ml, 2mg/100ml, 1mg/100ml, 0.5 mg/ml etc. till you find the proper conc.
In the meantime, you can try different vol.
another effect should the sample preparation 1) as close as the mobile phase; 2) either fully protonated or non-protonated if possible(may impossible for your amines).
These are short simple experiment (two prep at most) .
Good luck!

[2152812]

what we found was up to 20micrograms injected amount was okay for 250x4.6mm in this water 0.1%TFA mobile phase. Is that consistent with what you observed, YM?

We tried to match matrix with mobile phase, etc but those did not help...

Thanks again for your input.
SOS