UPLC/HPLC Method Development and High pH buffer

[Karen Lee]

I have been developing a high pH gradient method for UPLC and HPLC using the Waters XBridge and Acquity C8 columns. I am seeing many peaks which elute part way through the run even when a 0 uL blank injection is performed. I think these peaks must be related to compounds from the buffer solution building up on the column. The starting condition is 98% Ammonium Bicarbonate buffer/ 2% ACN and the ending condition is 50/50 Buffer/ACN. I have ruled out contamination by preparing buffer with previously unopened reagents. The bicarbonate buffer is also an extra pure grade. The buffer is pH'd with Ammonium hydroxide. The API's I am working with are both acidic and basic. Hence I need the low organic for retention of the acidic compound and the high pH of the buffer and higher organic concentration for the elution and better peak shape of the basic compound. These peaks are problematic since one of them elutes at the same RT of the latest eluting compound.

My questions are:

1. Has anyone seen this same phenomenon?
2. Is there anything I could try to eliminate these peaks?
3. Does anyone have experience using high pH pyrrolidine buffer and if so could you provide me with a supplier name and or part number?

[PJ8]

Could it be contamination in your injection loop? I've seen that before. If you inject a sample diluted in a weak diluent you can lose some on the walls of your injection loop, then when the mobile phase flow through you elute a little and get small impurity peaks eluting regularly in each chromatogram. You can easily diagnose this by running a gradient but bypassing the injection loop and seeing if peaks disappear (if this is possible on your system.)

To clean the loop use high injections of a solvent wihch will dissolve the contaminent (if you don't know then use a few things, nPrOH, DMSO, water, try to cover the full solvent range)

Of course it may be nothing to do with this, but this is a problem I have encountered in the past.

[Uwe Neue]

I think the first thing to do is to prove that the peaks are coming from your solvents. Do the 0 volume injection exercise as you have done, but vary the initial equilibration time: as is, + 10 minutes, + 20 minutes... If the peaks increase, they come from the solvents. If not, the cause of the problem is somewhere else.

[domino1]

I have had the same problem. THat is the issue with high pH buffer systems. Borate is even worse. I wish I had words of wisdom but I do not. I have gone so far and passivate my system to see if it would help - but no such luck. I just wanted you to know that you are not alone!

I am going to keep my eye on this thread to see if anyone can come up with a good answer!

[HW Mueller]

PJ8, here comes my conceptual problem again with some sources of carryover. I can see deposition between rotor and stator doing this, but loop? Did you run the chromatogram with the injection valve on LOAD after INJECT, and then switch back and forth regularily? Or did you have a proteins deposit in the loop and the gradient swept different proteins out at different gradient positions (just learned that some people had managed to do something like that)?

[bartjoosen]

Can you record baselines without injecting anything? As Uwe stated, make some runs, with various equilibration times: 0,5,10,15 minutes.
You can do a ramp from 100% buffer till 50% ACN. If the peaks increase with equilibration time, the problem must be related to the mobile phase, otherwise the source is elsewere.
Do the test, and let us know, then the problem is a bit better located.

Bart

[PJ8]

Wasn't working with proteins was a fairly hydrophobic API. We had to use as low an organic content in the diluent as possible to prevent horrible peak shape in an early eluting impurity, hence the solubility issues.

I guess it doesn't have to have been the loop thinking about it, perhaps it was more likely to be in the rheodyne? Wherever they were they disappeared after a few large injections of neat MeCN which we do periodically when running this method.

[Rob Burgess]

Karen,
It seems that high pH grdaient separations are notorious for baseline difficulties... I've never seen a good baseline when the pH is > 9.5 even with ultra pure ammonium bicarbonate. What pH do you need to operate at. Perhaps you could try phosphate at pH 13. The X-bridge columns are purported to be stable at this pH.. you could take a try it and see approach. Also try some form of in-line mobile phase clean up cartridges. Our department are recommending oasis cartridges if I'm not mistaken and they do seem to in certain situations. Good luck!

[Karen Lee]

I have tried increasing the equilibration time between injections and the impurity peaks do increase in height and area. Hence, I beleive it is coming form the aqueous phase of the mobile phase.

[bartjoosen]

Alright, you scaled down the problem to the mobile phase.

Now you can check where the impurities come from:
You can check a gradient with water without the buffer additives (but don't change the habit of filtrating, stirring, ...). Now step by step add the other components until you found the cause of the disturbance.

(And please let us know what the answer is)

Bart

[Karen Lee]

I have tried running the analysis using pH 10.0 water (pH'd with ammonium hydroxide) and ACN. The peak profile is the same. Conclusion: impurities are either due to water or the ammonium hydroxide or a combination of both. I have ordered a bottle of HPLC grade water to try.

[bartjoosen]

Do you filter the water?
Maybe its due to the filter?

[AA]

I have run into this issue many times with ammonium bicarb buffers using ammonium hydroxide to adjust the pH against ACN gradients. For reasons unknown to me I often see what I have identified (using PDA data, so not 100% confirmed) as varoius phalates (spelling right, not sure). It seems to matter little about the source of water, bicarb, or hydroxide. I have found ways to deal with it by pre cleaning my buffer using a solid phase technique which I have posted on this board in the past. If you cant find the method in the archives I will post it again, out of time right now.

[Uwe Neue]

All pre-instrument components and tools are suspect, when you get the described profile. Tubing, containers, stirbars, lines for helium degassing (if you use that), pH meter, the tubing of your Milli-Kuh, plastic graduates...

[Karen Lee]

I have been filtering the buffer through a 0.45 u Nylon membrane filter. I have ordered 0.2 u Nylon filters to try for UPLC analysis. I have also checked with the supplier and they have indicated that this filter type is good for aqueous phases with a pH range of 3-12.

My thanks to everyone who has posted a reply,