buffered pH to avoid peak tailing

[Sofia]

Hi,

can anybody explain why a buffered mobile phase reduces peak tailing for acid anaytes?

If I run a gradient with one component of buffered water and the other of methanol will the buffer capacity (and pH) be the same for the mixture of the two components as it was for the buffer alone?

[David Blais]

Sofia,

By buffering the pH of the mobile phase (preferrably to a pH 1.5 units less than the pKa of your acid analyte) you keep the acid analyte in a neutral (rather than ionized) state. By keeping the acid in a neutral state, you greatly reduce/eliminate the interaction of the charged acid analyte with the free silanols of your stationary phase. This interaction is a major cause of peak tailing.

As for your second question, I think that may depend on some other details such as steepness of gradient, how much methanol, etc. In my experience, setting up a gradient system such as this should not significantly change the buffering capacity and you should be okay. But, of course, the only way to be certain is to try it!

Good luck,

David

[Uwe Neue]

David and Sofia,

One does need a buffer or other means of pH control to make sure that the entire analyte peak is in the same state of ionization. If you do not do that (for example if you do not have a buffer), the analyte itself is the only determinant of the ionization. This means that the ionization changes with the concentration, and one gets distorted peaks.

There are several ways to control the ionization of a compound. One is to make up a mobile phase with a pH that is far away from the pK of the analyte. If you do that, the ionization does not change, and you get a nice peak (except for bases interacting with silanols, which is a different subject). The other way is to use a true buffer with a reasonable concentration (5 mM for MS, 20 mM for LC). Contrary to common perception (and to the post by David), this gives symmetrical peaks independent of the pK of the analyte. This is especially true for acids. One can demonstrate this by injecting an acidic sample with a pK of around 4.5 into a acetic acid buffer at pH 4.75. Symmetrical peaks result. If you do the same thing with phosphate at pH 4.75 (no buffering capability) you get tailing.

Bottom line: pH control controls ionization and thus retention and peak shape.

If you are interested, I'll sent you a collection of troubleshooting articles, one of which covers specifically this experiment.

[Basil]

Thanks Uwe for your easy undesrstand words,

And for basic compounds with interactions with silanol groups ??...

How important is to be near or far away pK of the compound..

[David Blais]

Uwe,

Yes, please forward me the list of articles to which you refer. I would be interested in reading them.

-Dave

[Newman768]

There are different silanol groups at the surface of any silanol stationary phase: SiOH in the neighborhood of SiOH, Si with 2 OH-groups, free standing SiOH, SiOH in the neighborhood of metal ions (did I forget one?).
SiOH in the neighborhood of metal ions reacts very acidic and can absorb basic compounds even at low pH. This absorbtion is under some circumstances irreversible. These silanols are responsible for peak tailing with basic compounds. Adding low conc of TEA for example can mask this silanols. Ultra pure silanol contains normally less than 10 ppm of metal ions. Search the LC/GC archive for more information.
An easy rule is that silanols ionize at pH above 3 and can absorb basic compounds if no buffer or acid is add to the mobile phase.
Silanol groups are responsible for the polar selectivity in rp chromatography and the reason why silica is the most popular type of hplc stationary phase.
Another easy rule is that only 50% of the silanols can be modified by the functional groups and endcapping.

[Uwe Neue]

Basil,

Newman768 has answered some of your question. If you got a high-purity silica, you can get good peak shape even at pH 7, if your base is still ionized. I do not have all the answers, but I know that the pK of silanols on a high-purity C18 is shifting into the alkaline pH, and this makes tailing less likely on the high-purity C18.

You can also obtain good and unproblematic peak shapes for bases if you go to the alkaline pH and make them not ionized any more.

[Bill Tindall]

I don't think your second question got answered. Yes, diluting the aqueous buffer with methanol will reduce the buffer capacity. To a first approximation the buffercapacity will be lessened by the dilution factor.

Second , adding methanol will affect the proton acitvity of the buffered solution as well as the pH as measured by a glass electrode.

[HW Mueller]

I am having a slight conceptual problem here: Above ~pH 3 the ionization of Si-OH will yield Si-O-, which should be able to interact with a base only if it´s protonized. If M/M+ raises the acidity then the Si-OH should ionize earlier (below pH = 3). One can imagine that M/M+ could be a Lewis acid and interact with the unprotonated base. At high pH one has, presumably, only base, Si-O- , and "Lewis" neutralized metal species, such that amines can behave "normally". Is this the correct explanation?