Chromatography Forum

Apologies for flogging the ion pairing horse again, but I would like to initiate some discussion on, well, simply whether the acetate anion can act as an ion pairing agent under reversed phase conditions.

I am being assured by a colleague that it is "because it's ionic", however I found that I quickly developed a debilitating headache during our conversation and was thus unable to continue to reason my side in greater depth.

Due to said colleague's strength of belief, I figured I would try to search for references, or gain other opinions on the matter.

Further evidence presented to me was given by the following observation made by my colleague:

- A method to analyse a surfactant molecule and side-products comprising the following functionality; quaternary alkylammonium ion, terminal or bridging phosphate, a small number of hydroxyls in an alkyl sidechain. Chain lengths are C6-C10 and the quaternary ammonium is formed with two methyl groups.
- Luna C8 column.
- Mobile phase consists of 10 mM ammonium acetate in water, methanol, THF and is run as a gradient to a maximum of 70% organic (30:40:30 aqueous:MeOH:THF). The aqueous pH is unspecified and unmeasured.

Now, because "peaks move around" when including or excluding ammonium acetate in the aqueous part of the mobile phase, this is interpreted as evidence of ion pairing.

Depending on the model considered, my initial thoughts on the subject are that 1) acetate ion is not hydrophobic enough to be retained on the column surface and 2) the mobile phase isn't of low enough dielectric constant to allow formation of a paired-ion.

Comments appreciated.

Fluorinated acids are used as ion-pairing agents because of their increased hydrophobicity relative to non-fluorinated organic acids. They also have the benefit of a very low pKa resulting in a negative ion at any pH. Of course, acetate itself can form ion pairs but is not as effective as TFA because of lower hydrophobicity.

In the method that you have cited, the ammonium ion is probably also critical to cover active sites on the column which should result in improved peak shape. Of course, ammonium acetate will also change the pH which could alter retention but probably not so much in this case since aqueous ammonium acetate pH is around 6.5-7 while the pKa of the surfactant is much higher.

I think that the question should be broken down in two parts...

1) Is acetic acid an ion-pairing agent?
2) Does acetic acid operates as an ion-pairing agent under certain chromatographic conditions (i.e. other additives, pH, buffer capacity, organic solvent % etc).

About the first question, I would establish a "baseline" such as an anion (very hydrophilic and without a side/hydrophobic chain) that does not operate as an ion-pair reagent. In order to keep all things equal your mobile phase pH needs to be buffered and be preferably 100% aqueous. Any anion that would offer increased retention compared to your baseline should be considered an ion-pairing agent, how weak or strong should be another issue...

About the second question, everyone understands that you can always find chromatographic conditions where you won't see any ion-pairing effect. In the case of acetic acid you can either decrease the mobile phase pH to very acidic (i.e. by the addition of sulfuric acid) where acetic acid (which has a pka around 4.75) can be completely uncharged or let's say operate at 100% ACN and you will eliminate all posibilities for ion pairing.

"Ion-pairing" is probably a misnomer. I would call it ion-exchange with adsorbed ions. If you search, you will get a lot of mechanistic information on the technique in the late 1970s and 1980s.

If you follow my model, the acetate ion will contribute to "ion-pairing" when adsorbed on the surface, i.e. in mobile phases with a high water content. At high organic concentrations, this is not likely, and ammonium acetate will work like an ammonium salt.

I liked the experiments by Bidlingmeyer and Deming in J. Chrom. 186 as a reasonable explanation of the mechanism.

I'm really unconvinced by the "ion-pairing" effect of acetate in any model under the conditions I have so far given.

I'm not sure of the appropriateness of this association, but I know that acetate isn't very retentive w.r.t fluoride under full ion exchange conditions - this therefore makes me doubtful of the formation of ion pairs in the aqueous stationary phase (hydrated shell).

The organic content of the method ranges from 30-70%. I will be very surprised if acetate ion is retained on a reversed phase column full stop. Further to this, if we could confer some retention of acetic acid under low pH conditions would we even be left with an electrostatic surface?

I don't expect acetate to be ion-pairing reagent as well as I don't 100% agree that TFA is ion-pairing reagent, you can argue this point but Hydrophobicity of both TFA and acetate are not that great. Fluorinated acid (C4) are used as ion-pairing reagents so I would expect unsubstituted carboxylic acids to show some IP properties at certain conditions, which are:
- IP needs to be at least C6 (butyric acid)
- low organic concentration
- pH of the mobile phase needs to be above 5 to have these acids ionized, this pH will reduce Hydrophobicity of the acid, so you might want to go to C8 at least
- analyte needs to be basic in nature (primary, secondary, tertiary and Q-amines), pyridines will not be retained due to IP mechanism because at pH above 5 most of them become neutral.

You need certain difference in pKa of IP reagent and analyte to observe effect of ion-pairing. This is a perfect case for fluorinated acids which show pKa close to 0.

JA,

I tend to agree with you that under the conditions you describe, acetic acid will not work as ion-pairing reagent (unless something atypical happens with the quaternary amines) and that "peaks move around" due to different reasons...

Kostas,

As per your post, assuming pKa of Hex.Sulfonate ionpair as 2.1 ( I do not know the exact number) - it is recommended to use pH > 2.1 for a optimal ion-pairing interaction to occur.

Is this because - the ion pair will be charged fully at this juncture to amplify the ion-pair mechanism.

Mohan_2008,

You want to keep your ion-pairing agent charged so that it can interact with ionic compounds of the opposite charge. Alkylsulfonic acids and perfluorcarboxylic acids have very low pka and this is one of the reasons they are so popular as ion pair agents (in addition to the hydrophobic chain).

Where does this idea come from that acetic acid is less hydrophobic than TFA?? Who can get acetic acid to elute before TFA on a RP column via pure hydrophobic interaction? (We just discussed this).
To JA´s original post: It is strong tobacco to suggest that acetic acid is an ion pairing agent, using data in which the pH is not even known.
We discussed, some time ago, whether TFA can be replaced by other substances. The only thing I remeber about this is that it is extremely difficult to get conditions in which comparisons are fair (scientific). Bidlingmeyer´s paper on this is a very good example of the difficulties involved. As far as I can ascertain the evidence for any ion pairing mechanism is still equivocal.
So, I can only think of one thing that may be useful for JA in case his colleagues do any more experiments "proving" the ion pairing prowess of acetic acid: Make sure that there is enough salt in the mobile phase to prevent any ion exchange with SiO-.

When in the same ionization state, TFA has a higher log Kow relative to acetic acid which means it is more hydrophobic. Obviously, acetic acid will be more hydrophobic at pH < 4 since it is not ionized while TFA is ionic. However, that isn't relevant if we are talking about the potential for acting as an ion pairing reagent since it must be ionized.

I'll chime in with the observation that nothing we do in LC is ever a "pure" mechanism. Furthermore, concepts like "ion-pairing", "ion-exchange with adsorbed ions" (or, for that matter, even "reversed-phase") represent simplified models which we apply to help us understand a complex reality.

I'll sidestep some of the semantics by re-phrasing the question: "Can acetic acid behave like an ion-pairing reagent?". The test of whether it *is* behaving like an ion-pairing reagent in the case you describe was described by Kostas.

All of that said, in my experience, acetic acid is not used primarily for it's ion-pairing-like behavior, but I would not be surprised to see related selectivity changes as a secondary effect.

sassman, here are the logPow that I found:

Acetic acid (VWR international MSDS) = -0.31/-0.17

TFA (Burdick & Jackson MSDS) = -2.1

I wouldn't be so sure about those numbers for TFA. This is from the SRC PhysProp Database:

CAS Number : 000064-19-7
Chem Name : ACETIC ACID
Log P (octanol-water):
Value : -0.17
Type : EXP
Ref : HANSCH,C ET AL. (1995)
pKa Dissociation Constant:
Value : 4.76
Temp : 25 deg C
Type : EXP
Ref : SERJEANT,EP & DEMPSEY,B (1979)

CAS Number : 000076-05-1
Chem Name : TRIFLUOROACETIC ACID
Log P (octanol-water):
Value : 0.50
Type : EST
Ref : MEYLAN,WM & HOWARD,PH (1995)
pKa Dissociation Constant:
Value : 0.52
Temp : 25 deg C
Type : EXP
Ref : KORTUM,G ET AL (1961)

I wouldn't be so sure about those numbers for TFA. This is from the SRC PhysProp Database:

CAS Number : 000064-19-7
Chem Name : ACETIC ACID
Log P (octanol-water):
Value : -0.17
Ref : HANSCH,C ET AL. (1995)

CAS Number : 000076-05-1
Chem Name : TRIFLUOROACETIC ACID
Log P (octanol-water):
Value : 0.50
Ref : MEYLAN,WM & HOWARD,PH (1995)

Both the SPARC online calculator and Marvinsketch also show that TFA has a higher log P; although, SPARC can only do the neutral species. Measurement of Kow is subject to a variety of experimental artifacts, but it seems that the majority of the evidence says that TFA is more hydrophobic.