Precipitation in HPLC with phosphate buffer?

[Bill Tindall]

Only 20 % acetonitrile will not have a very large effect on changing the actual pH of the eluant . Ask the column manufacturer about why the column failed. If the column is supposed to survive at pH 10 then I don't see that you did anything that should have bothered the column- caused it to crush or dissolve.

[moino]

@ Mattias:

Interesting, I've never thought of the pump itself. We have an Agilent 1100 quaternary pump (up to 400 bar), it is usually running with 1 or 0.5 ml/min with a size exclusion polymer gel column (no solvent gradients), so no high back pressure occours (only about 10 bar in total).
My guess was that precipitation of the phosphate buffer might happen in the degasser, I was thinking of including an inline filter just before the pump, after the solvent mixer. When it happens next time, I will give it a try.

[HW Mueller]

A lot of questions are raised here:
How long would it take for something like a pH 10 NaOH solution to produce a ~3cm void?
How can a plug ahead of the pump (in a degasser, etc.) produce a buildup of pressure in the pump?
How could a phosphate precipitate, unless it formed an insoluble metal complex, resist washing by water if the column still allows a flow of this water?
How can a pump produce a plugging of the column when phosphate is involved? The pump releases metal that causes insoluble phosphate complexes to precipitate in the column? (Or is this pump just not capble to handle certain viscosities [backpressures] with which the others have no problem?)

[moino]

> How can a plug ahead of the pump (in a degasser, etc.) produce a buildup of pressure in the pump?

In my case (after several hours with a low flow rate), the *teflon frit* in the purge valve attached to the pump head got blocked. Through this filter all solvents coming out ot the pump head are filtered, before entering the tubing to the other hardware modules.
For the symptoms, I only see the degasser (and now the pump) that may cause the blocking, but no idea about the actual mechanism how the phosphate (if it really is the phosphate) precipitates, and why no rinsing helps afterwards.

[Mattias]

It is the mixing of phosphate and acetonitrile that doesn´t work on Agilent quarternary pumps. If we premix the mobile phase, there is no problem.

I also tried to backflush the column after one of the column breakdowns, and didn't succeed. The only thing that helped was if the frit on the top of the column was replaced. I guess it is not that simple that it is phosphate salt that is precipitated on the top of the column. It could be anything from metal complexes (as suggested) to particles from the pump seals... I have no idea. Anyway, we never use the Agilent Quat pumps for phosphate methods anymore!

[HW Mueller]

Mattias, sounds like unfortunate coincidences, allright.

[Kostas Petritis]

Although quite strange, it is good that you narrow it down to the Agilent quarternary pumps. Just curious, you maybe know that Agilent has an option for continue pump seal wash. Does your pump had this option as well or not?

[Mattias]

No, it has just the passive seal wash (but that is on the otherhand very continuous)

This problem appeared at one of our QC labs when they bought new equipment (they had binary pumps before). We have two systems with Agilents quarternary pumps, and the method failed on both when I set up the method. In total, four different quat pumps all over the world have been investigated and none could handle the method, so the case is clear in my opinion.

Also, one of the blocked columns was transferred to a Agilent binary pump system and worked for many hundreds of injections! (with a new frit)

I have discussed this matter with Agilent and they also sent an service technician who spent two days without fixing the problem. Their official standpoint is still that it should work (but they must know...)

Just for information: we used 20 mM Phosphate buffer pH 6.5 in channel A and acetonitrile in channel B. Column: Zorbax Eclipse C18.

[moino]

We also don't have the seal wash option. However, if the seal became weared out, there should be dark particles in the blocked filter frit, that was not the case when we had the problem.
We had been running the HPLC for over 1 year with the same seals (since I'm in the lab), with 50 mM phophate + 100 or 150 mM NaCl. My impression is that the blocking only occoured after running at a low flow rate for some time or not flushing the HPLC before a longer pause, at least some days.