IEX Theory Question

[Eric Moore]

I think I understand that IEX separates by charge. Species with different charges will elute at different times.

But what effect does SIZE of the species have? Will two molecules with the same net charge, but with different sizes (say, for instance, one is 2x bigger by MW) elute at the same time? If they have similar charges, but different sizes, will they elute near each other, but one will have a higher response?

And what effect does shape play?

Thanks for your help, and feel free to make snide comments. I'm new to IEX.

Eric Moore

[IC_Gal]

I am not an expert but I have done my share of IC on anions and cations and I have been where your are and asking these same questions. There are several nice reference books on IC out there. I'll leave it to the Metrohm, Dionex, etc folks to point these out to you.

The way I understand it is most of the sepn mechanism in IEX is based on charge but the size of the ion can limit it's "mobility". For instance fluoride is small ion with one charge and usually elutes very quickly on most of the IEX columns I have used with carbonate and hydroxide eluants. However another ion in the halide series, iodide, will elute even after the divalent anion sulfate. The size of the iodide ion makes it somewhat poarizable and it tends to interact with the stationary phase for a much longer time than fluoride even though both are monovalent anions. So yes size does have something to do with the sepn mechanism as well as charge.

[danko]

Size does matter (even in IEX chromatography) – often anyway. The effect is due to the molecule’s dissociation constant. The best example is (one I use to illustrate the case anyway) separation of benzoic acid and methyl- benzoic acid. The first is more willing to be ionized and thus easier to be released from the stationary phase. So if you create the right chromatographic conditions, you can easily separate those two compounds by Anion Exchange Chromatography despite the fact that they both carry a single charge.

Other examples would be acetic and higher linear carboxylic acids. And then there come the peptides/proteins where the molecule conformation plays a huge role in addition to the mechanism mentioned above.

So, don’t hold back. Experiment and you will be awarded.

Good luck.

[tom jupille]

All of the above.

The first factor for single-charged-site molecules is the valence of that charged site. Other things being equal, the higher the charge, the more tightly the molecule "sticks" to the opposite charge site. Partially ionized weak acids or weak bases can be considered as fractional charges for this purpose.

Given the same valence, charge density plays a role. The denser the charge (i.e., the smaller the site), the more tightly the molecule sticks.

Given the same valence and charge density, molecular size will affect retention (large molecules may not be able to nestle as close to active sites on the ion exchanger compared with smaller molecules).

For multiple-charged-site molecules, in general, the more charged sites available to interact with the ion exchanger, the stronger the retention. This is a function of the number of charged sites and the geometry of the molecule. Eg., two charged sites on the same side of a big molecule (such that both can "see" the ion exchanger) will give stronger retention compared to two charged sites on opposite sides (such that only one at a time can interact).

That's the general hierarchy, but as you can imagine, the number of possible combinations is immense (do two low-density charged sites on the same side of one molecule trump one high-density site on another?), which is why we have to do experiments!

[HW Mueller]

Because of the complexity of this subject I always take a look at brochures of manufacturers before I start experimenting with ion exchange chromatography. For instance, Bio-Rad´s AG 1 (a trimethylammonium column) has the following selectivity:
Benzene sulfonate 500
I- 175
NO3- 65
Br- 50
Cl- 22
HCO3- 5.5
Formate 4.6
Acetate 3.2
F- 1.6

The ions with the higher number (selectivity) attach most strongly to this column, which means you can easily raplace (exchange) F- with I- , but you need a very high concentration of F- to get rid of attached iodide. Apparently, the interaction of ions with water (mobile phase) plays a very large role here.

There have been some good discussion on related matters.

[Chris Pohl]

Eric,

While Tom's answer covers your question, let me elaborate.

For two ions of equal charge but different hydrated ionic radii, the smaller one will elute last since it experiences the highest electrostatic attraction. For example, of the halides, iodide elutes last because its hydrated ionic radius is smallest (exactly opposite the case for the "naked" ionic radius where iodide is the largest). Fluoride elutes first in this series because it has, conversely, the largest hydrated ionic radius. When you add to this consideration the question of what happens when the ion is increased in size through incorporation of nonionic components, the answer is that smallest ion should still have the longest retention time assuming there is no secondary retention mechanism. If you take the case of monovalent carboxylic acids, formate should have the longest retention time if no other retention mechanism is playing a role. Typically, however, there is a retention mechanism involving hydrophobic interaction with the ion exchange phase, especially if it is of styrenic architecture so under 100% aqueous condidtions there is nomally a retention minimum for either the C2 or C3 acid and further increases in chain length result in increased retention relative to shorter chain homologs. This effect can be eliminated by adding solvent to the mobile phase, in which case the theoretical selectivity should be observed.

[danko]

Chris,

This is exactly the opposite of my experience. As I mentioned earlier the longer the carboxylic acid’s chain (just to use it as an example) the longer is the retention time (everything else equal) on an anion exchanger. This seams to be your experience as well, but you explain it with hydrophobic interactions. And this is where I’ll have to disagree with you - hope you don’t mind. The hydrophobic interaction between acetic acid (for instance) and anion exchanger support is very weak indeed. Much weaker than the electrostatic forces involved in ion exchanging. On the other hand C3-chain-carb. acid is much harder to dissociate/ionize than C2-chain (the electro-negativity thing) and when both C3 and C2 are in basic environment they’ll be attracted to the positive sites of the anion exchanger, but C2 will be easier to elute than C3 carb. acid – with or without addition of organic solvent. You can make pretty certain predictions by comparing/examining the pKa(s) in the case of acids.

Best Regards

[Chris Pohl]

Danko,

You are welcome to express your opinion but I'm afraid your description doesn't fit with the facts. The facts are:

1). Most ion exchange phases do exhibit significant hydrophobic character.
2). More polar ion exchange materials shift the retention minimum to longer chain length homologs.
3). If you add enough solvent to the eluent you will observe the theoretical elution order (smallest elutes last).

One concrete example: LAS (linear alkylbenzenesulfonte with an average alkyl chain length of 12 carbons) never elutes in 100% aqueous conditions on an IonPac AS4A column but with 50% acetonitrile in the eluent it elutes before fluoride.

Chris

[Chris Pohl]

Danko,

Two other points relative to your earlier post:

1) The pKa of acetic and propionic acid has essentially no bearing on their retention unless you are operating within 2 pH units of their pKa. Since the pKa of acetic is 4.75 and that of propionic is 4.87, then there is essentially no effect of the pKa when the stationary phase pH is greater than 7. So unless you are referring to anion exhange separations under acidic contions, the pKa of acetic and propionic acid have no effect on retention. Both are essentially 100% ionized when the stationary phase pH is greater than 7.

2) One way to better understand the role of hydrophobic retention in ion exchange is to examine the effect of the ion exchange site size, structure and hydrophobicity on ion exchange selectivity. For example, switching from a trimethylammonium anion exchange site to a triethyl ammonium ion exchange site has several effects on selectivity. For highly hydrated anions with essencially no hydrophobic properties such as fluoride, sulfate or phosphate, the effect of switching from a trimethylammonium anion exchange site to a triethyl ammonium ion exchange site is minimal (at least when the eluent system uses carbonate or hydroxide). If anything, this shift results in a slight decrease in retention for these ions with the most pronounced shift being observed for the more highly charged ions. However, for ions with hydrophobic substituents the effect is the opposite: retention increases. Furthermore, the more hydrophobic the substituents associated with the ion the greater the enhancement in retention. Clearly this wouldn't be observed if hydrophobic character in the proximity of the ion exchange site weren't highly relevant to the retention process under aqueous conditions.

Of course, as I mentioned before, you can mask this added hydrophobic retention by adding solvent to the mobile phase. I should add, however, that it isn't possible to quantitatively suppressed hydrophobic interactions between analytes and the stationary phase in all cases. Some stationary phases, particularly phases with ion exchange sites not constrained to the surface such as was common for phases developed in the 70s and 80s exhibit so much hydrophobic character that it's not possible to quantitatively suppress the hydrophobic character with solvent. In such a case, it's more efficient to mask the hydrophobic character by adding a low concentration of a co-ion which exhibits even higher levels of hydrophobic interaction with the stationary phase. For example, micromolar levels of p-cyanophenol is particularly effective in masking such sites in the case of anion exchange materials derived from styrene monomers.

[danko]

Chris,

I recognize the existence/influence of hydrophobic interaction in ion exchange chromatography, but as mentioned earlier, the effect is negligible compared to the electrostatic forces the technique is based on. The hydrophobic effect manifests it self through peak tailing and general broadening and is eliminated by addition of 5 – 10 % organic modifier (e.g. ethanol, IPA or ACN). I use it my self almost every time, but I’ve never seen retention time shifting upon addition of organic comp. The peaks just get slimmer and elute marginally earlier. Maybe you’re using one of these mix-mode columns?
Regarding your pKa remarks, I never referred to acidic conditions – on the contrary. As I wrote, under basic conditions (i.e. both C2 and C3 protonized, say at pH 7 - 8 ) the anions will exhibit slightly different attraction to the stationary phase (e.g. the positive amine) and thus be susceptible to selective retention/separation. The reason for me mentioning pKa is that the effect follows the pKa values in the case of organic acids for instance.
Yes I realize that the difference between pH 4.75 and 4.87 seems negligible but it indicates that these two compounds exhibit a different degree of affinity to a positive ion. Besides - one should remember that the pH scale is logarithmic. Anyway if in doubt, you should try to separate acetic and propionic acid by anion exchange chrom. with and without organic modifier and you’ll be convinced.
Experimenting can be quite rewarding.

Best Regards

[danko]

Sorry, I meant deprotonized/ionized and not protonized.

[HW Mueller]

danko, your acetic and propionic will both be essentially totally ionized at pH 7 (as Chris already indicated). If one were close to the pKa of the acids there should be inddeed a very small diff. in ionization, with the acetic very slightly more dissociated (more ionized). One would expect it (acetic) to attach more strongly to the stationary + ions, not less as you suggested?

[danko]

Yesterday evening I went in the lab and performed a quick and dirty experiment.
Unfortunately I couldn’t find the same column brand as the one I’ve used earlier (a couple of years ago).
No propionic acid either, but I found formic and acetic acid. The following conditions were set up:

Column: Anion exchanger (no brands shall be mentioned)
Eluent A: 20 mM TRIS pH 8
Eluent B: 20 mM TRIS pH 8 plus 0.5 NaCl.
No organic modifier in either of the eluents
Gradient from 0 to 50% B in 15 min.
Flow rate 0.6 mL/min
Detection: 210 nm

Results:

Retention time acetic acid: 10 min
Retention time formic acid: 11.5 min

The outcome confirms Chris’ prediction!
Obviously my earlier observations have been jammed by non specific/hydrophobic interactions that haven’t been eliminated by approx. 10 % organic.
I would’ve loved to recreate the “oldâ€

[HW Mueller]

Great Danko, looking in during my vacation has payed off: I can celebrate the new year without the fear that the acids are changing character.