Challenging separation of moenomycins by ion-pair HPLC

[cyanogen78]

Hi to all !
I have to optimize an HPLC separation of Flavomycine Complex which mainly consists of 5 isomers of moenomycin:



The original method was taken from somewhere a long time ago - it uses Hypersil BDS-C18, 150 x 4.6 mm, 5um column and mobile phase consisting of 15 mM Na-heptane sulphonate as ion-pair reagent, and 50 mM phospate buffer with pH=7.0. Flow rate was 0.9 ml/min, Vinj=20ul, UV=256 nm. These conditions give the following chomatogram with 30 min run time:



Because this routine analysis takes a lot of time we want to optimize the method in terms of analysis time mainly, and Rs of min = 1.8 is required. We have to use conventional HPLC system.

Because I work for an Agilent distributor we initially chose Eclipse XDB C-18, 75 x 2.1 mm ID, 3.5 um. I didn't manage to achieve analysis time below 15 min with this column. Now I'm working with Poroshell 120 EC-18, 100 x 3.0 mm ID, 2.7 um, the best result received is below (with run time 10 min) :



0.7 mL/min mobile phase of 10 mM AmAcetate + 25 mM Na-heptane sulphonate (53%), AcN (35%) and MeOH (12%).

And then began problems with method stability - shifts in the retentions and, as a consequence loss of Rs. All temperatures are stabilized except that of solvents before entering the pump.

I'm open for all opinions and I continue to look for points influencing separation process reproducibility.

[danko]

Hi Alexander,

What is the pH of the new buffer (the acetate)?
Under all circumstances, increase it's concentration (f. x. to 25 mM).
Why did you change from phosphate to acetate btw?

Best Regards

[cyanogen78]

Hi Dancho,

I'm so glad to see you are still very active in this forum. I'll contact you personally through e-mail.
The pH of the new buffer is 7.0. Because I considered such lack of reproducibility with one colleague who is much more experienced from me, he adviced me to use phosphate buffer for pH 7.0. Initially I chose acetate because it is more friendly for the HPLC system.
Yesterday I tried with 20 mM phosphate buffer an 10 mM IP-reagent but again I was not able to achieve the separation showed on 2-nd chromatogram. I tried also pH 6.0 with citrate buffer and pH 3.0 with phosphate. I noted that lowering the pH increases retention and affects selectivity.

[HW Mueller]

For what is the detergent ion pairing agent (heptane sulfonate)?
You are trying to reduce rt and then add the detergent to increase rt?
Anyway we have seen some other examples recently of using ion pairing agents in the mobile phase on detergent drugs (analytes), what is the purpose?
The increase of salt concentration should have the effect to obliterate that of the ion pairing agent, so why not take it out?
(Incidentally, ammonium acetate at pH 7 is hardly a buffer).

[danko]

Hi again,

Yeah I’m still here – as long as there is something called chromatography I’ll be there.
OK, without seeing the latest chromatograms my suggestion would be to further increase the buffer (phosphate pH 7) concentration to say 30 – 40 mM. The ion pair concentration should return to the original (i.e. 15 mM). And then let’s see what happens. Do you control the separation/column temperature and if yes at what level?
What about the flow rate, did you change it when you changed the column diameter from 4.6 to 2.1 mm?

Best Regards

[cyanogen78]

Hi again! Thanks for comments.

The ion-pairing is necessary because the molecules are quite soluble in water (please note that they have carbohydrate rings) - w/o IP the rettention is too poor even at high water content in the MPh. I tried initally on C18 column w/o IP but w/o reasonable retention as result too.

I agree about the choice of AmAc as buffer for this separation - it was a mistake. Currently I can't upload the results with 25 mM phosphate buffer at pH 3.0 and 25 mM citrate at pH 6.0. I can note that change in selectivity happened with pH 3.0. I tried such less and more acidic conditions because the molecules acts with their N atoms with the IP anion and I hoped that such process will be enhanced with more active N-s in acidic pH. And I can confirm that it works because the retention increases in more acidic MPh and selectivity changes too. But it didn't influence resolution - unfortunately - the band spacing is the same, and the change in elution order doesn't enjoy the customer .

I promise that I'll upload the chromatograms in the beginning of next week. The problem is that I'm working on different HPLCs and the file formats are different. I recently thought that it will be fantastic software to exist to be able to import and working with all chromatographic and/or LC-MS data files - only for data analysis ... But I haven't such tool.

I can also try to change the IP reagent - we are looking for minimum analysis time with min Rs of 1.8 at 250 bar max pressure, and possibly using of hexanesulphonate will decrease retention without loss of Rs.

To danko - I read somewhere that in IP-LC the ion strenght of the buffer should be as lower as possible because the buffer ions compete with IP ions in the interactions with the analytes molecule. Thats why I used so low concentration (but too low) buffer.

What you think about that?

[MestizoJoe]

Is that IPR best? It seems to me you shoud be using a anion exchanger not a cation exchanger.

At pH around 7 your carboxylic acid and phosphate should be negatively ionized. I think you should be using a positively charged IPR.

I could be wrong. If so, please correct me.

[cyanogen78]

It is not an IEX mechanism desired here but an ion-pairing - we want to ionize our molecules's active site and to catch to them an ion of an opposite charge with large hydrophobic part, and to retain the complex via hidrophobic interactions on RP conditions - that's ion-pair LC.

In our case we have a molecule with N-atoms which are protonated in acidic pH and have positive charge. To them a heptanesulphonate ion is paired because it is anion, with negative charge at all LC pH scale.

[MestizoJoe]

I don't see any basic nitrogens. Amides have a pKa of about 0 or 1. At pH greater than 5 you most definitely will have negative ions for your api. The pKa of phosphoric acid is around 2 or 3 and you have a phosphate in your molecule.

Ion pairing is a combination of hydrophic interactions and ion exchange. That's ion pair LC. Look it up in any HPLC method development book. I recommend Snyder or Meyer.

I'm still not sure if your IPR is best. Anyone else want to chime in?

[cyanogen78]

It's right we have something like phophate group in the molecule, but it is too steric protected in my opinion to act as an acid center. Probably I'm wrong ...
We have secondary amine center in the molecule ... About the amides' pKa:
Wikipedia - "Compared to amines, amides are very weak bases. While the conjugate acid of an amine has a pKa of about 9.5, the conjugate acid of an amide has a pKa around -0.5. Therefore amides don't have as clearly noticeable acid-base properties in water. This lack of basicity is explained by the electron-withdrawing nature of the carbonyl group where the lone pair of electrons on the nitrogen is delocalized by resonance. On the other hand, amides are much stronger bases than carboxylic acids, esters, aldehydes, and ketones (conjugated acid pKa between -6 and -10). It is estimated in silico that acetamide is represented by resonance structure A for 62% and by B for 28%.[1] "

[MestizoJoe]

There are no amines. Those are all amides or derivatives of an amide.

All I'm saying is you have a negatively charged API at pH >5 and you have a cation exchanger (CnHnSO4 -) which is used for weak bases, because they are positively charged. You might get better results using an anion exchanger like CnHnNH4 + or even tetrabutyl ammonium salts.

Good luck.

[carls]

There are no amines. Those are all amides or derivatives of an amide.

I agree with Joe. The N atoms are very weakly basic and will not be protonated in aqueous solution. There are 2 acidic groups on this molecule (phosphate and carboxylic acid) and therefore it will be negatively charged in solution at pH >3.

Have you tried the separation without heptane sulfonate? In the last chromatogram you have significant retention with almost 50% organic so maybe the sulphonate is not needed. Ion pair chromatography (IPC) is notorious for poor reproducibility and long equilibration time.

Perhaps something as simple as 0.1% phosphoric acid will give you a reasonable separation.

Since this column has been exposed to an IP reagent you should use a new column for any method development that does not include this same IPR

[cyanogen78]

Probably you are right about the acidity/basicity of molecule's active groups ... I'll check it next week.

[HW Mueller]

Ì have searched the structure again for a secondary amine until my eyes watered, can´t find it. Forget the steric hindrance, these moenomycins appear to be negatively charged detergents. So it should be advantageous to go to a acidic mobile phase, not only for the reasons just given, but also to prevent charge exclusion. I suspect that addition of the sulfonate increased the rt only by obliterating this exclusion.
Also, it might help you to review IP-HPLC to see what the most likely mechanism is.

[danko]

Hi Alexander,

The guys are right.
Dump the IP, prepare a 0.05 M Potassium Phosphate buffer, pH 2.5 and give it a try.

Best Regards