degradation of phosphites on RP-HPLC columns

[Brian1706]

Hello

I am facing a problem with 3 diphosphites on a RP18 HPLC column.

Working with the same HPLC system and the same eluents (H2O/Isoprop), I used 2 columns which are filled with the same RP 18 material.
On column 1 there is a heavy degradation of the diphosphites and I can see 2 additional peaks (with intense tailing, maybe the (di)phosphates?!) in high amounts.
On column 2 I am not seeing degradation above 1-2 % percent in comparison to the diphosphite peak.

To check the RP18 material we ordered a new column with the same batch of column 1. The degradation became much smaller and there was only 1 additional peak, but this peak still had 10 % of the diphosphite peak. So it is somehow in between column 1 and 2, but still not acceptable.

I talked to the column manufacturer and he said that beside poor RP18 material the frits made of stainless steel could be a possible reason.

Does anyone have more ideas how such degradation of phosphites can occur on a RP-HPLC column and how to prevent it?

Thank you in advance for your help.

[danko]

Who says it's on-column degradation?
I would sure be considering selectivity (and potentially efficiency) differences between the columns you're testing.

Best Regards

[Brian1706]

Who says it's on-column degradation?
I would sure be considering selectivity (and potentially efficiency) differences between the columns you're testing.

The area of the phosphite peak decreases as the peak areas of the other 2 substances increase. Retention time differences between each peak are about 7 min.
All 3 columns are filled with the same material from one manufacturer (coloumn 1 and 3 are the from the same batch) and work comparable with many other substances.

How can selectivity play a role here?

[danko]

How can selectivity play a role here?

If your sample material is degraded to some extend (before even injecting/loading it) the one column might separate these degradants due to relevant/viable selectivity, while the other might not - due to inferior selectivity. That would be the legitimate reason.
Best Regards

[Brian1706]

How can selectivity play a role here?

If your sample material is degraded to some extend (before even injecting/loading it) the one column might separate these degradants due to relevant/viable selectivity

Ok, I get your point!

But I am still kind of skeptical, because I would guess that there should not be something in between like column 3 showed. Because I injected from the same standard solution I would have expected that the 2 additional peaks are either present or not.

But in contrast I get a significant change in peak areas for the additional peaks on column 1 and 3. Or could there be something like an equilibrium for the separation?

The question is how can I test if it is selectivity or degradation?

Maybe another interesting or confusing fact:
If I take off the pre-column of column 1 the areas for the 2 additional peaks are going down significantly, but they are still there...

[danko]

I think it’s quite important to investigate whether you’re dealing with on-column degradation or separation discrepancy. I’m even more convinced now that you’re dealing with the latter.
OK here is my suggestion for a viable investigation. Divide some of your standard solution into 3 portions.
The first portion would be a reference (no challenges, potentially kept cool in a fridge). The second portion could be heated and potentially exposed to UV light for an hour or so. The third portion could be the one you try to expose to the same conditions as it would be during elution (same pH and temperature as you chromatographic conditions and for the same time as the time for elution). Her you should add a piece of stainless steel – a screw or a column fitting).
If you’re dealing with on-column degradation you’ll see some of your peaks in the first portion when running it and it would be the same amount as you’ve seen previously. The third portion should result in something like the double amounts. The second portion should show or whether or not the HPLC conditions are the source of the degradation or it would be happening under all circumstances.
Finally run all 3 sample solutions on the column that doesn’t “degrade” your compound and see what happen.

Best Regards