SAX method optimization

[MestizoJoe]

Hi everyone,

I recently developed a method in which four components of a mixture were resolved. The system I used was 0.5mM NH3 and 0.24 mM TFA and the pH was near 9.5 or so. All my sample components come off the column within 35 minutes.

The column is an Alltech Novosep A-1 Anion which utilizes tetra alkyl ammonium.

The problem I am having is the peaks become very broad when I increas the sample load and I lose resolution.

I believe the NH3 is important to keep the pH above 8 and keep my sample ionized. I think TFA is my ion exchanger.

I was wondering what I can do to increase sample load without losing resolution and retention. Previously I increased my buffer concentration and each component eluted faster and I lost resolution. For example at 6 mM TFA and 12 mM NH3 everything in my sample elutes in ten minutes where at above mentioned conditions they come out within 35 min.

No organic was used. All aqueous.

Thanks,

[tom jupille]

It's a bit hard to tell without seeing the chromatograms, but I would speculate that you are overloading the buffer capacity of your mobile phase (i.e. there is enough analyte present that it's changing the local pH). At 9.5 it's the ammonium that's doing the buffering, so you could try increasing that if your column will tolerate the higher pH.

Another possibility is to check with Alltech to find out the ion exchange capacity of that column and then look for one of higher capacity in order to get reasonable retention with a buffer in the millimolar concentration range (up by a factor of 10 or so).

Yet another is to look for a higher-capacity weak exchanger or a mixed-mode column which might let you get away with a more concentrated buffer.

All of the above assumes that it's not simply an injection volume issue.

[MestizoJoe]

Thanks, Tom. I don't believe it's an injection volume issue and agree it's more than likely the sample changing the local pH. It is a weak acid after all and my buffer is very weak. The injection volumes are between 1 and 10 uL of a relatively concentrated solution (~ 10 mg/mL).

here is before and after pics. before when i had a good chromatogram and after when i increase inj vol.

before



after



Today I am switching from ammonia TFA to ammonia HCl with hopes that the Cl ion will need to be present in higher concentration to elute the sample and therefore buffer conc will be higher.

I doubt I'll be able to get a new column because I'm sure my column isn't the best for this work. I should have a semipreparative column.

[tom jupille]

As you up the load, your peaks are fronting, which is consistent with overwhelming the buffer.

The "classical" order ( F- < acetate < formate < Cl- < Br- < I- < NO3- <sulfate ) basically tracks the charge density of the hydrated ion. Trifluoracetate elutes between Br- and NO3- in published chromatograms from Dionex, which would put it in the same ballpark as I-.

That sequence suggests that acetic acid might be a better choice. ??

[danko]

I've never seen decent isocratic elution in ion exchange mode.
If I were to optimize/re-develope this method I'd certainly go for a weak initial conditions and then apply a stronger eluting gradient.
You'll be impressed by the results if you make the effort to get it right.

Best Regards

[tom jupille]

Danko, you haven't looked hard enough! Check the Dionex or Metrohm catalogs for many examples. The isocratic / gradient issues are the same in *any* chromatographic mode: if the range of retention is sufficiently narrow, isocratic works; if the range of retention is too wide, you need a gradient.

[danko]

Hi Tom,

My point is much more related to a kind of "on column concentration" than compensation for too wide range of retention. Especially in protein separations the gradient mode is as good as indispensable.
I think we discussed a similar topic once and as far as I remember nobody could show data demonstrating "nice" protein separations in isocratic IEx mode.
So, maybe I don't need to check Dionex and Metrohm?

Best Regards

P.S. I got this feeling that the data here is an example of attemt to separate some peptides (looks like this to me) and that's the reason I suggested a viable path to follow.

[tom jupille]

No question that gradients are the norm for peptide / protein separations by ion exchange, but they are also the norm for those separations by reversed-phase. Even if the intent is to develop an isocratic separation, a gradient is a good place to start (in fact, that's the way we teach it in our Method Development course).

In the event, however, the ionic strengths in question are lower than what would typically be used in a gradient.

[MestizoJoe]

Actually, this is a small molecule.

I've had some improvements since the chromatogram I showed you before. When I switched to HCl / Ammonia I was able to use a higher buffer concentration than with TFA / Ammonia. This gave me better retention, resolution, and peak shape.

One problem I had was that the sample pH was around 2.5 and my mobile phase pH is around 9 (I was a bit embarrassed about that fact). I fixed it and my peak shapes improved greatly.

Today I'm trying HOAc / Ammonia with the hope that increasing my buffer conc. and decreasing my eluent strength (OAc < Cl) will give me better retention and separation and better peak shape due to a stronger buffer.

This is my first IEC method. I don't think the next one will give me as much trouble.

Thanks for your suggestions.