Importance of kPa in determining pH of mobile phase

[CP]

what is the importance of kPa in determining pH of mobile phase in RP HPLC?

i'm analysing amlodipine using HPLC with UV detection. I'm using column Zorbax SB-C8 5um.

buffer that i'm using is KH2PO4 0.01M.
Mobile phase is 37% ACN and currently i'm setting the pH at 3.0.

Somebody told me that i should review back the kPa of amlodipine in order to get the right pH for the mobile phase to get good result.

*the kPa of amlodipine that i know is 8.7. (correct me pls if i'm wrong)

since i'm still new in this thing, can someone help me on how to calculate the right pH for my mobile phase for amlodipine analysis?

[sassman]

If the pH of your mobile phase is close (within 2 units) of your analytes pKa, you will see poor peak shape. This is because some of your analyte will be in one ionization state while the rest of your analyte is in another ionization state (in reality it is an equilibrium) causing the peak shape to be broadened. For this reason, it is best to adjust the pH of you mobile phase to be at least 2 units away from your compounds pKa. For basic drugs, like amlopidine, this usually means using a low pH to fully protonate the molecule. On some columns with some drugs, it may be necessary to use additional mobile phase additives to prevent strong interaction of the protonated base with the column. It is also possible to analyze the basic drug at a pH above the pKa of the analyte (in the neutral state), but this requires a column that is designed for analysis of basic drugs at high pH.

[CP]

the column that i'm using now can be use for pH 1-8.

so does that means that i have to change the pH of the buffer at about 6?

[Uwe Neue]

CP: your method is perfectly fine. You are far away from the pKa of the analyte, and you should get reproducible chromatography. The only concern is the buffer capacity of your buffer. It is worse at pH 3 than say at pH 2. If you inject large amounts of your analyte, you could overload the buffer and get some peak distortion.

Sassman: It is a persistent rumor that the peak will widen if you are working with a buffer close to the pKa of an analyte. This is actually incorrect. You will get peak widening and distortions if your buffer is too weak, but not if the pKa of the analyte is at the pKa of a well buffered mobile phase. I can send you an example that demonstrates this (HPLC Troubleshooting in American Laboratory).
The only concern about working at a pH near the pK of your analyte is that retention might change, if you are not able to control the pH of your mobile phase well enough. But there are ways in which this can be controlled, specifically using weighing procedures instead of pH measurements.

[HW Mueller]

The "persistent rumor" seems to stem from the persistent belief that the column separates the acid/base from its conjugates. Acid base equilibria are much too fast for competition from a chromatographic separation.

One can think of a situation where broad peaks result if a base already has been partially converted to cations, but there are still some SiO-. In that case the acid base equilibrium could be disturbed severely

[lmh]

Dear HW Mueller and Uwe Neue,

I think another reason for the persistent rumour stems from those of us who tend to be a bit lazy and use weak acids without checking pH. When peak shape is bad, we make it better by adding more acid, reasoning that we are taking the pH another unit in the right direction if we increase the acid 10-fold. But of course we are also strengthening the buffer capacity too!

Mechanistically, I can see that peak broadening could just about, in certain circumstances, result from using a pH close to a pKa, but I've never tried to recreate the conditions, and can't prove this would ever happen in real life: Imagine a situation where you inject a sample in a bad choice of sample solvent, for instance 90% methanol for a typical reverse phase method. Instead of binding at the top of the column, the sample now binds over the first few mm, as it takes a while for the column's solvent to dilute the eluting injection solvent. Now imagine the typical scenario that the column oven is imperfect, or we have a very good oven but pump cold solvent straight in without pre-warming. There is now a temperature gradient over a region of column where sample is present. Since pH and acid-base dissociations are both temperature dependent, it is quite possible that parts of this region will be in slightly different mixes of ionic forms, and will start to elute at a different times, leading to broadening... of course this is all rather artificial, and the extent of the problem depends on how bad a chromatographer is in charge, and what the rest of the column does to improve things.

[HW Mueller]

Imh, you are describing a slow or non-equilibrium situation, we were talking about fast equilibria, except for my example of "silanol" interaction. That broadening could also be due to slow or non-equilibrium.

[lmh]

Oh, agreed entirely. I agree absolutely with your point about speed. Fast processes like acid-base equilibria will obviously average out over even the shortest chromatography run. It's only slow processes we have to worry about. Thanks for taking the trouble to read and reply!