Loss in Retention Time

[Parsival]

For an HPLC Method which uses an {(Ammonium Bicarbonate - Tetrabutyl ammonium hydroxide Buffer (pH 9.8)}/ Methanol mobile phase we observe a continous decline in retention times (both for the main compound as well as any other impurities).

Retention times are restored when washing the column with water/methanol and re-equilibration with mobile phase.
So the column - an Xterra RP18 - should be happy with this pH.

As the run-time is quite long (approx 12h for 19 injections) and RT loss significant (>5% during this time), I wonder whether someone would know how to prevent this RT loss?

Thank you very much.

[Consumer Products Guy]

One cause could be components from your injections building up in your column, effectively blocking some of the active phase from doing its job, which is retaining components of interest selectively. Your washout procedure seems to remove those, restoring performance.

[wanda50]

There is obviously a buildup (as consumer guy said), but is it the buffer components or the analytes/matrix?

I assume this is an isocratic separation? What is the percent of buffer to methanol? What percent of water/methanol is the wash? Do you need the water in the washout, or is the buildup removed with a high methanol concentration?

You could set up the isocratic separation on 2 pumps, then at the end of each injection jump the methanol up to see if this will wash off the culprit.

[JGK]

Is it possible to add a column flushing step into each injection (flushing with water/ methanol then equilibration with the MP)?

This may add time to your sequences, but would (possibly) solve the RT issues

[tom jupille]

A washout after each injections (or perhaps after every "n" injections) should work, but has drawbacks, as has been pointed out.

The other approach is to get rid of the contaminants before they get to the column. That requires that you figure out if they are coming from the sample or the mobile phase. That requires investment of a day or so: run a series of samples, wait for a long while (with the pump running), then run another series of samples. Plot retention versus sample number and then plot retention versus total elapsed time.

If the contamination is coming from the mobile phase, the 'versus sample number" plot should have a jump in the middle (corresponding to the waiting period), while the 'versus time" plot should be more linear.

If the contamination is coming from the sample, the "versus sample number' plot should be reasonably linear, while the 'versus time' plot should have a plateau corresponding to the waiting period.

We have this slightly better documented in the LC Troubleshooting Wizard on our web site:
http://www.lcresources.com/resources/TSWiz/hs390.htm

[Parsival]

Hi,

Thanks a million for your recommendations so far!

The method is a kind of gradient: after 30 min the mobile phase changes from A: 80% (Buffer + 0.2 %TBAH) : 20% Methanol to B: 55% (Buffer + 0.3 %TBAH) : 45% Methanol, to wash out a component in the samples. However, the decrease in RT is observed also when only A is used, for example to run standards.
The column/needle wash is Methanol : Water = 70 : 30.

The decrease is independent of the number of injections, but approx. linear to the duration of the time, mobile phase is running through the column. This might indicate that, if there is an impurity to be blamed, it must come from the mobile phase, isn't it?

As we use the ammonium bicarbonate and methanol components in many other methods without observing this effect, it should come from the TBAH component.
Would have anyone already observed such a contamination of TBAH, we use Fluka technical approx 40% in water, product no. 86880?

Thanks again from Ireland.

[HW Mueller]

Could it be that your column is simply not in equilibrium when you start injection? That is, the initial retention times are nonequilibrium times?

[Uwe Neue]

I agree with Hans... The equilibration with TBAH may take longer than planned.

[Bruce Hamilton]

Regardless of the actual cause of your problem, I would strongly recommend obtaining a better quality of TBAH.

My experience is that technical grades can contain impurities that can interfere in many analyses.
That technical grade is ~40% in H2O (~1.5 M) with:-
Total impurities = ≤3% halides (as bromide).
Anion traces = sulfate (SO42-) ≤3000 mg/kg.

I use Romil 40% in water product, but I'm sure there are many other brands that are superior to what you are using.

Please keep having fun,

Bruce Hamilton

[seaclimb2001]

My knowledge may be dated, but we used to have a hard time getting pure TBAH. I believe it is also somewhat unstable after opening the bottle. Tetrabutylammonium phosphate is available in high purity and is stable.