I'm currently running analyses with these conditions:
C18 column, phosphate 5mM, TBAHSO4 5mM, pH 6.5, 15% methanol.
My problem is that the retention time of the analyte is continuosly decreasing. For example: with a sequence of 40 runs, 1 hour each, retention time decreased from 18.3 to 16.1 minutes. If I perform a gradient the decrease is less evident.
Has anybody else had this problem? Many thanks for any help!

How do you equilibrate the column with the ion pair reagent? How do you store the column? What is the quality of your TBA reagent? When do you measure the pH, before adding TBAHSO4 or after?

Hello OSM, every ion pair reagent will modify your stationary phase in an irreversible way. But if the column is "modified", it should show a reproducible performance. At first, why 1 hour runtime? What are your compounds and the matrix? Do you use a column thermostate at a fixed temperature? Anonymous asked when do you measure pH, and I will ask you, how do you measure pH? And why pH 6,5? Do you have a low pressure gradient HPLC system, or a high pressure gradient system? How do you mix buffer and methanol? Gerhard

Are you running a gradient over 1 hour?

If it is so, you have to re-equilibrate your system with the starting ratio of your mobile fase.

To Anonymous:
1) with 25 column volumes of mobile phase
2) generally the column is stored in H2O/MeOH 60/40, but I was wandering if storing in mobile phase could improve the situation
3) TBAHSO4 is puriss. >99% (Fluka), NOT for IPC
4) pH is measured after dissolving TBA and phosphate, immediately before filtering the phase

To Gerhard:
1) 1 hour run time is simply to be sure that late-eluting impurities are eluted, because 2) the sample is quite "dirty", as it is a reaction mixture (but I also had it with a clean standard)
3) the column is termostated at 40°C
4) I measure pH with a pHmeter which is calibrated at least once a day, using an electrode dedicated to mobile phases measurements
5) pH 6.5 because the method used to be performed at that pH, only it was a gradient analysis (trying to simplify it I got into troubles!): as I mentioned before, using the gradient the retention time decrease is less important. This is the reason why I thought to an equilibrium problem and I ran the analysis for 40 hours: I hoped that at it end I should have reached it!
6) low pressure. I tried a Jasco low pressure mixing system, a 1100 low pressure, a 1090, it is always there: I can't think it is a mixing problem, also because 7) I used both separate and premixed solvents.

To add puzzling matters:
1) I also tried to change pH at 7.0: RT changed but still drifted, so it shouldn't be pH
2) I run a similar 40 hours sequence with 20mM TBAHSO4: RT decreased (I expected the opposite), separation was slightly worst but drifting had EXACTLY the same trend, only smoother.
I am presently trying TBAOH, and it looks a bit better, but the sequence is still running

Thank to all for your suggestions!

osm,

Your method seems to have a very low ion pair agent conc., perhaps raising the conc. of the pairing agent might help with the drift (might be overwhelming the pairing agent conc. Just my 2 cents worth.

regards,
Mark

25 column volumes of mobile phase are insufficient to reequilibrate the column with the ion pair reagent, especially after you have washed some of it off with the 40% methanol. Store the column in the mobile phase, and the problem will disappear.

To the last anonymous: I can admit that 25 column volumes are insufficient, but I ran a sequence which lasted 40 hours (1 hour = about 50 column volumes), and I thought that at a certain point the column should have been stabilized, and this didn't happen...
To Mark: I have already tried to increase to 20 mM...

40 hours is plenty. What is your C18 column?

Is the analyte which is drifting in retention time an anion, a cation or a neutral species under the conditions of operation? I realize this might seem like a stupid question but it's not clear from your earlier posting whether or not the analyte in question is specifically the motivation behind adding the ion pair reagent in the first place. If the analyte is the same charge as the ion pair reagent then retention would be expected to decline during initial equilibration as well as when using a higher concentration of ion pair reagent. Otherwise, the most likely cause of the problem is a highly retained anionic species present as an impurity in one of the components being used to make up your eluent (i.e. the water, the methanol, the phosphate buffer or the ion pair reagent). Are you sure you don't have something growing in one of your stock solutions (assuming you are using stock solutions to prepare this eluent)? Also, are you making up your ion pair solution using Fluka tetrabutylammonium hydroxide or are you using the phosphate from Fluka. The former is generally of higher purity although if the source is old it may have partially degraded (to produce tributyl amine and butene).

The column is a standard (not base-deactivated) Nucleosil.
The analyte has a pKa of 4.7, so at pH 7 it should be an anion.
Water is freshly withdrawn from Milli-Q; methanol and phosphate are HPLC grade. The TBAOH which I used as second option (the first one was TBAHSO4) was quite old, so yesterday I received a brand new one. After a very long initial conditioning I had about 5 hours of stable retention time (0.2 minutes range), then it began again to decrease. Could it be a degradation of the mobile phase into the reservoir? I can't believe it...

osm:

Since everything else seems to be in order, I think we have to go after the less likely scenarios. Considering the fact that the old tetrabutylammonium hydroxide looks like it may have been contaminated, there is now a possibility that other system components are likewise contaminated as a consequence of this. You may want to consider replacement of components exposed to this solution, especially the eluent container, the endline filter if you have one and possibly also the eluent line leading from the bottle in question to the pump. I'm not saying this is the case in your situation but I once visited a customer who had an ongoing "situation" and I found to my surprise that the eluent lines on his instrument were all green from contamination with what I assume was algae. Anyway, needless to say in this case we had to do some major disinfectant work to get the system back in proper working order. I know this is a long shot but I thought it worth mentioning just in case.

Well, there is little left to chase. Keep the column equilibrated with the mobile phase, and let us see if the problem repeats. Incidentally, you did not say anything about temperature. Are you controlling the temperature or working at room temperature?

Hi OSM,
Chris Pohl is right, it seems that you have a major contamination in your system. Under your conditions even a standard Nucleosil column should work properly. I also had such cases at customers labs, and many years ago, when I worked myself in HPLC labs I had such a problem. After cleaning every part of the system, "plumbing" new capillaries and with a fresh column the problem was gone. Anyway, your HPLC system will thank you, getting from time to time such a maintenance. Good luck. Gerhard

You probably have carbonate contamination.

You cannot get "carbonate" contamination at pH 6.5 although you can certainly get carbon dioxide contamination at that pH. However, this would not be expected to result in the described symptoms since the since the small fraction of bicarbonate present at that pH would be expected to have been extremely minor effect on retention.

In relation with retention time changing in ion pair chromatography, have anyone any explanation for the following:
I have been using OSA 1.5 mM in aqueous mobile phase with a new C18 column with reproducible retention times for three days. After that, I washed the column with 25% acetonitrile in water and then I stored it in 100% acetonitrile. 20 days later I reused the column, flushing it first with acetonitrile 25 % and then with the same mobile phase; retention times for all peaks doubled the initial ones and never got back to the original values.

I would check first the flow rate between your first runs and the second run.

Of course I have checked the flow rate between both runs as well as reagents and mobile phase preparation.