Impurity makes a split peak, Risedronate

[Susanne]

Hello
I have a question about a "split peak-situation".

When I spike my risedronate with orto-risedronic acid (about 1% of main peak), the peak for orto-risedronic acid splits. Other imputity-peaks do not split. When I analyse the orto-risedronate alone the peak does not split.

pH in my spiked sample is about 6,3. pH in my orto-risedronic acid solution i about 7,5 but even when I ajust the sample til pH 6,3 the peak is not splitting up. pH in my mobile fase i 7,5 and I use a Luna (2) C18 column.

Please help me find an explaination for this. Thank you:)

[Rob Burgess]

Is it just an occurence of your diluent solvent strength being much weaker than your mobile phase. I presume your acid has very low retention at this pH. Can you lower the pH below the pKa of the acid.

Also try lowering your injection volume, does the peak still split.

This is a common problem for chromatographers, your not alone... I'm having a nightmare with this phenomena myself at the moment!!!

[Susanne]

Dear Rob

Thank you very much for your reply. It is nice to know that I'm not alone

What do you mean when you say that my peak split i because of the solvent strengh in my sample?
I use mobile phase as sample solvent and my mobile phase consists of Phosphate buffer pH 7,5, MeoH, EDTA and tetrabutylammoniumphosphate.

It is not possible for me to lower my pH to below pKa.


Thanks

[Roy]

It may be possible that two peaks you are seeing are partially protonated form and non protonated form in the mixture.

[tom jupille]

Roy, I'd be very surprised at seeing separate peaks for the protonated/unprotonated. This type of equilibration is orders of magnitude faster than the residence time on the column; all you will ever see is a single peak representing the "average" degree of protonation.

Susanne, Rob was hypothesizing about what is probably the most common cause of peak splittin (as well as bad tailing, shoulders, etc.): injecting a sample which is dissolved in a solvent which is "stronger" (i.e., provides a lower k' value for the analytes) than the mobile phase. What can happen is that the sample hits the head of the column in that solvent, and then slows down as it mixes with the mobile phase. In essence the peak shape problem represents a "skid mark".

I can speculate that you might get a similar effect due to an overload of the risedronate. To check, you could inject a smaller amount of a sample spiked with a higher % of the ortho-risedronate. That might help to establish at what point the problem kicks in.

[Roy]

Tom,

I agree with you about equilibration and solvent sterngth. However we have been able to resolve a lot of peak splitting issue by shifting the equilibrium. If its a dynamic equilibrium then in your sample you will see two distinct peaks as any given time % of protonated and non protonated entity are fixed.

Susan, To answer your solvent strength question, one of the rule of RP Chromatography is your injection solvent strength should be lower than your initial gradient. Thats pretty good rule however there are plenty of examples where sample is dissolved in ACN and the gradient starts with 10% organic. So it doesnt hold always.

Thanks.

[tom jupille]

Roy, I think we may be saying the same thing but in a different way.

If you have an equilibrium between two compounds (or two forms of the same compound), and the equilibrium is pH dependent, then you will see two peaks, assuming two things:
1. the system can separate the two forms
2. the equilibration between the forms is slow compared to the residence time on the column (i.e., slow enough that no significant conversion occurs during the run time).

If the equilibration is reasonably fast (i.e., fast enough that a significant interconversion occurs), the result can be anything ranging from a split peak, through a flat-top peak, to a shapeless "blob". In that situation, changing the pH to force the equilibrium one way of the other can result in a dramatic improvement in peak shape. The mutarotation of glucose falls in this category (there have been a couple of threads dealing with that topic in the past few weeks), although there the equilibrium is controlled more by temperature than by pH.

If the equilibration is very fast, then you will get a sharp peak reprenting the "average" of the forms. Ionization/deionization (i.e. simple pH equilibration) usually falls in this category unless there is something else (ring formation, steric blocking, etc.) going on to slow things down.

[Susanne]

Thank you very much for your replies.
Since I have already tried ajusting the pH in my orto-risedronic acid sample to the same pH as my risedonate sample, I do not think that the split peak is because of pH alone.
I think it is somehow connected to the heavy load of risedronate I inject into my column? I can not change this load though so the way I see it I will have to accept my spit peak for now.
Again thank you very much for your relpies.

[HW Mueller]

Roy,
of course, at equilibrium the concentrations of the two entities are fixed, on the average (that´s a definition of equilibrium). But an equilibrium is not static, in the case of protonation and deprotonation of a carboxylate it´s extremely fast. The life time of these is on the order of nanoseconds if I remember corretly. Even in the much faster nmr there is not even a remote chance to see attached and solvated protons in such equilibria at room temp.
(Remember: Kequ = kforward/kfreverse, small k being the rate constants)

What´s a dynamic equilibrium? What´s a non-dynamic equilibrium?

[HW Mueller]

The carelessness devil struck again, the carboxylate above should be oxygen species.

[Roy]

HW,

A dynamic equilibrium is one where conversion to both the form are still going on, but it appears to be static b/c the rate of forward and backward reaction are equal.

I have to agree with Tom with my experience. The species we are talking about are Protonated (acid form) and deprotonated (carboxylate form). Its not a matter of equilibrium of one molecule its matter of the system and at any given time certain % are ionized and certain % are not and its very possible to see both the form as close eluting peaks.

Its just a hypothesis and may not be the actual case, as Sus mentioned that she didnt think it was related to pH. However from the scientific point of view it should be related to relative capacity of providing H+ ion of the solution.

[HW Mueller]

Roy,
the thermodynamic definition of a chemical equilibrium is via the difference of free energy of the species:
G = -RTlnK
K is the equilibrium constant (mass action product at equilibrium).
This only says what the equilibrium concentrations of the species are, it says nothing about how long individual molecules stay as such. That depends on kinetics, or, the free energy of activation, which determines the rate constant. The rate constant, being a constant, doesn´t disappear when equilibrium is achieved, the reaction keeps going with the same rate constant as when it is far from equilibrium. The rate (rate constant x the respective species´ concentrations) is the same for the backwards and forward rea. at equilibrium, but it´s reacting extremely fast if the k are large (difference of activation energy is small). As stated, the differences in activation energies involved in the dissociation of -OH are small, thus very fast reactions obtain at equilibrium. To me an equilibrium is dynamic, otherwise there is no equilibrium.

[ym3142]

excuse me, Tom,
do I understand correctly?

1) the rate/speed of protanation/depro has nothing to do with equilibrium(the ratio of species);

2) the rate, but not equilibrium, does determine if the peak splits;

3) more likely, changing pH shifts equilibrium, but not rate;

4) conclusion is that changing pH usually does not solve spliting problem.
Further, when Roy claimed that he solved the problem by changing pH indicating his systems happened to be those rare, where the rates change significantly under different pH.

Thanks,

[KarenJ]

My experience with analysis of species in equilibrium is limited, but I thought that I'd offer this comment. I have used an HPLC method for estimating the pKa of a compound. In this method, the pH of the mobile phase is varied linearly with each successive injection of the same sample. So for example injection 1 is run at pH 3, injection 2 is run at pH 3.5, injection 3 is run at pH 4, etc. The result is a single analyte peak whose retention time shifts slightly with each run. As the pH approaches the pKa of the compound the shift is more dramatic with the retention time becoming shorter as the equilibrium favors the ionized species. Retention time decreases for acids and increases for bases as pH increases.

The reason that I bring this up is that the method works because the peak shape doesn't change as the pH changes, only the retention time changes. When I have run this method this has held true and in all of the papers that I've read on this method it has held true as well. Therefore, I would think that a change in pH that affects the ionization equilibrium should not result in a split peak but should result in varying retention times. This would suggest that the peak splitting problem would not be caused by a pH problem.

I would be interested in hearing about examples where this isn't true.

KarenJ

[ym3142]

KaronJ, thanks, it's a great input.