Chromatography Forum

Many people are involved in the analysis of anionic species (e.g. organic acids) or cationic species (e.g. basic drugs) using C18 or C8 columns in which the retention mechanism is largely hydrophobic interactions. Quite often, potassium phosphate buffers are used at low pH because this suppresses the ionisation of column sites which can contribute additional retention mechanisms. Neither potassium nor phosphate are recognised as efficient ion pairing species. My question is this: what opinions are there of the form of the solute when held on the C18 surface? Due to the requirements of electrical neutrality, are e.g. organic acids held on the surface as neutral ion pairs with buffer cations e.g. K+? Clearly, with potassium ions (or phosphate with protonated basic drugs) this would not be an efficient process, but this could explain the low retention of many ionised species on C18 phases. Or is the neutrality of the system maintained by a loose atmosphere of buffer ions which surround the discrete ionised solute species held on the surface? I have heard both opinions expressed but I will not declare my own in order not to bias views.

I don't think that one should look at this as an ion pair. In an aqueous system, there is a cloud of cations surrounded by a cloud of anions. The C18 layer is so thin (one molecule thick) that the layer thickness makes little difference. Thus there is no reason to think that the environment will change.
In addition, it can be shown that in acetonitrile the adsorbed ions are on the very surface of the layer formed by acetonitrile and the stationary phase. This is apparently a bit different in methanol, where analyte ions can penetrate into the methanol - stationary phase layer. But if analyte ions can do so, counter ions can do as well, and the situation is no different than in the pure mobile phase.
Bottom line: a monomolecular layer is not a bulk solvent.

Am I misunderstanding your question Victor?
Normally one would use an acid buffer to keep the acidic analyte´s equilibrium on the nonionized side (for instance, RCO2H which would be expected to have much higher retention than RCO2-) and a basic buffer for basic analytes for the same reason (for instance, R3N rather than R3NH+). Now if you don´t have a column which can withstand a strong basic buffer you hope that the pH you use gives you enough nonionized basic species to get the desired retention...... Or simply, one selects a buffer which shifts the equilibrium sufficiently to the nonionized side for the desired retention.

Thanks for your replies. HWM- I am assuming that the solutes are either strong acids or bases and that it is not possible to suppress their ionisation by working at very low pH for acids or very high pH for bases.t is actually very common to analyse bases under conditions where they are entirely protonated because only for the weakest bases is it possible to get sufficiently above the pKa value without damaging the column (apart from the few phases that are genuinely stable above pH .

Mr Neue, what evidence do you have that in methanol the solutes are able to penetrate into the stationary phase layer?

There have been publication by others, I believe at least by Kazakevich, which I still need to dig out.

My own evidence is as follows: I measure the gradient retention for a huge number of compounds in acetonitrile and methanol on a C18 and a C8 column. I calculate the difference in retention between both packings. I plot the retention difference measured in acetonitrile vs the same measured in methanol. For neutral compounds, this gives a scatterplot with a 45 degree angle. For all ionized compounds (protonated bases and deprotonated bases), it gives a scatterplot with a very flat angle. There is clearly a difference between ionized and non-ionized compounds, independent of the structure.

I just showed this at the HPLC conference.

Almost forgot about this.
I may still be off track, but there are diffuse ions, or such that have low heat of dissolution in H2O, or are termed chaotropic, or produce low surface tension in aqu. solution (what else?). These have a certain affinity for relatively low polar media. TcO4- is one of these. On the industrial scale this ion is extracted from aqu. wastes via organic solvents. On a C-18 column it is fairly well retained, independent of pH (shows no tendency to have a proton attached).
So, Victor, isn´t it time to disclose your view?

HWM- I think this is a very difficult question. With ions like the one you have described (i.e. very chaotropic ions) , I have no problem with envisaging them being retained in the stationary phase and interacting with solute ions of opposite charge. Such retention is facilitated by the presence of acetonitrile in the stationary phase layer. Some people call this ion pairing via a dynamic ion exchange mechanism. However, I still have a problem with this with less chaotropic ions, and for these I would tend towards Mr Neue's idea.

Victor, I don´t see any discrepancy between what Uwe said and my elucidation. Ions which go into methanol are not the most lyotropic. So what ions do you have in mind?