I’m often asked whether HPLC eluent additives should be added to one or both of the constituent mobile phase reservoirs - both organic and aqueous, when mixing gradients online. This is really a matter of reproducibility and robustness within the method.
We should start with a brief overview of the types of additive and their purpose:
Buffers – which consist of a weak acid or base and it’s salt in co-solution (phosphoric acid and disodium hydrogen orthophosphate for example) which are used to control retention behavior (by resisting pH changes) when the sample diluent and eluent mix at the head of the analytical column. This is important as, for ionisable analytes, the pH will control the degree of analyte ionsation and hence analyte retention and the selectivity of the separation. We should note here that the ionic strength of the eluent system is also controlled by the buffer strength.
Acids or bases – very often simple acids or bases are added to eluent systems to control pH. These are used for several reasons – primarily to shift the pH of the eluent to a place well away from the analyte pKa – to improve method robustness and resist changes to retention and selectivity with small changes in eluent pH. Addition of an acid to adjust the eluent to pH 2.5, for example, will cause basic species to be fully ionised and (most) acidic species to be non-ionised (ion suppressed) and their relative retention times will be shortened and lengthened respectively. Acid addition also helps to suppress the ionization of acidic surface silanol species of the silica support and so will also reduce the degree of peak tailing observed when chromatographic basic compounds.
Ion pair reagents – used much less frequently these days, these reagents act to either form ion pairs in solution with charged analytes, rendering them non ionic and therefore capable of retention in reversed phase mode, or adsorb to the stationary phase surface and undergo a type of ion exchange reaction with charged analytes – in practice a little of each mechanism occurs. The charge, nature of the hydrophobic moiety of the reagent and concentration in solution are all important factors in controlling retention and selectivity.
Usually one adds the reagent to the aqueous portion of the mobile phase system – solubility of reagents is usually better (especially buffers at higher concentration), and pH measurements will yield a more sensible figure when measured in aqueous solution……
However – one should consider the effects of altering the concentration of the additive (during a gradient analysis where only the aqueous component has had reagent added) against its function in the experiment.
For example – if we are using a combination of ammonium formate and formic acid as a buffer system, a 0.1% w/w solution in the aqueous phase will result in a 0.05% solution in 50:50 water:methanol and 0.01% in 10:90 water:methanol and the concentration of the hydroxonium ions of the each of these solutions will be very different. Will this be enough to adequately control pH and therefore the retention and selectivity of the separation?
Similarly – if the ionic strength of the solution is important having the reagent controlling this parameter added to only one eluent reservoir will mean that the ionic strength of the eluent is not constant. Ionic strength is important when there is any electrostatic element to the separation (you’d be amazed how often this is true with standard reversed phase separations), but also in separations designed for this purpose – such as HILIC or more properly perhaps, mixed mode separations. Situations in which TFA or other acids or bases are acting as an ion pair reagent (i.e. or every situation when you have a weak acid or base present without a salt with a common ion) are also relevant in this context – especially with LC-MS systems. Having a constant concentration of the species present in solution, no matter the combination of organic and aqueous, can often make a big improvement in the reproducibility of analyte retention times and separation selectivity.
Ever wondered why the peak shape of early eluting polar compounds is a little better than those later in the gradient?? If you are only adding acid to the aqueous portion, the eluent will be more acidic (actually will contain a higher concentration of hydroxonium ions) at the start of the gradient in reversed phase mode, than at the end, if only adjusting pH of the aqueous component. Therefore the silanol groups will be slightly more suppressed and this, amongst several other factors, will be why the peak shape is slightly better due to the increased suppression of the acidic surface silanol groups.
OK – so it takes a little time to work up the solutions and it may mean that you can’t re-use the organic with a different analysis – but in the situations described above, it could solve a lot of issues that you may not even realize you had!
For more tutorials on LC, GC, or MS, or to try a free LC or GC troubleshooting tool, please visit ChromAcademy at www.chromacademy.com