peak shifting

[niallg]

I am running an isocratic system with ammonium carbonate buffer/MeOH (25/75) pH 6. The analyte and internal standard elute consistently for about 5-6 injections ( Resolution = 6) and then the analyte peak begins to approach that of the internal standard to give a final resolution of approx 2, with the peak tailing gradually becoming worse. Any ideas?

Thanks
Niall

[Uwe Neue]

I have not been able to create a stable carbonate buffer in the neutral pH range, therefore I think you are running into the same problem. How do you adjust the pH, and do you measure if before the addition of the methanol or after?

[Mark Tracy]

This is one instance where you do not want to degas your buffer. CO2 will be removed and the pH will drift up. You can only use a binary high pressure gradient mixing pump in this situation. Even without degassing, this buffer system in unstable.

[KC]

I hope he is measuring it before the addition of the methanol. can it be done any other way? Never heard of measuring pH after the addition of organic fraction.

[DR]

We refer to this as the "apparent pH"

[Mark Tracy]

You can measure the pH after addition of the mobile phase, and it will be valid if you understand a few things. First, the pH scale is no longer 0-14 with 7 being neutral. Any given organic/water mixture will have its own pH scale, with a fixed relation to the familiar aqueous one. This means that you can calibrate with conventional buffers, and still have a repeatable measurement of your mobile phase pH. Second, the reference electrode will have a different liquid junction potential depending on the filling solution and the organic/water mixture; fortunately, this effect is usually minor and you can ignore it. Third, the pKa (and thermal coefficient) of your buffer salt will shift in the organic/water mix. This effect can be as much as one pH unit either way.

There are a couple practical issues also. The liquid junction of the reference electrode may suffer from crystallization if the mobile phase is mostly organic. There are reference electrodes with more generous junction area that will work better.

If you attempt to transfer to another lab a method that measures pH after organic addition, be sure to specify the type of electrode you use. This will minimize the hassles for the other lab. The exact brand is not (usually) important, just the type.

[niallg]

thanks for the responses but we are still having problems. Our client has provided us with a saturation column saying that this should sort out our problem. I have no experience with these columns. Has any one any suggestions / recommendations

[Uwe Neue]

I do not know what a saturation column is supposed to do.

Please explain:
How you prepare the buffer and the mobile phase?
When do you measure the pH?
Do you use a high-pressure gradient mixer?
Do you use a degasser?
What setup did your client use?

The more we know, the better we will be able to troubleshoot.

[Mark Tracy]

A saturation column is a sacrificial silica column placed between the pump and the injector. They are used when the pH is alkaline. The concept is to pre-saturate the mobile phase with dissolved silica to retard the hydrolysis of the silica in the analytical column. I have never personally used this approach.

[Uwe Neue]

Mark, I am aware what a saturation column is. I do not see how it would improve the problem described here. Unless pH 6 is not pH 6 but pH 9, or something....

[HW Mueller]

KC,
there have been extensive discussions on buffers in this forum, including measuring the pH in other than pure water solutions. Now it has been pointed out that if one developed a working method one does not really need to know the pH, only how you got there, and that others can follow what you did. Thus I prepare aqueous solutions of desired pH and/or measure the pH only as a rough check. There is no intention on my part to ever measure a pH in a solution containing organics (at least I presently see no need for that).

niallg,
in addition I would like to know whether your column has changed permanently.

[niallg]

HW M
The buffer in the mobile phase is pH adjusted to 6.0 then the organic is added.

The retention times of the peaks return to their original times when a freshly prepared mobile phase is used and then the shift is observed as the run progresses. It is a USP method we are trying to replicate.

Niall

[HW Mueller]

It seems obvious that Uwe pointed out your problem. Simplest solution: Use a more stable buffer. (Do you want to analyze your stuff or satisfy some official?)

[Uwe Neue]

It might be worth your while to communicate the problem to the USP.

[Victor]

Niallg,

Take note of the cautions posted concerning pH measurements in aqueous-organic solution. However, there is nothing wrong with trying it in your case.

1) Make up your aqueous buffer as usual.

2) Immediately, collect the mobile phase prepared by your HPLC (if it is mixed online) e.g. by collecting the mixture from the instrument purge valve (without the mobile phase going through the column). Measure the pH of this mobile phase (which contains methanol). It will not be pH 6.0. Leave the electrode in the solution for sufficient time for a stable reading to be obtained.

3) Purge the column thoroughly and collect the mobile phase, preferably at a time when your separation is working. Remeasure the pH of the effluent.

4) Collect the mobile phase from the end of the column when your separation is not working. Remeasure the pH of the effluent.

If you compare the readings, you will be able to see if the pH of the mobile phase is unstable. I am sure that this is your problem. I believe if you watch the prepared mobile phase carefully, you will even be able to observe bubbles of CO2 emerging from it, if you look carefully or perhaps drop in a bumping granule......