MS Bleed with Polar Embedded Columns

[Koen Hollebekkers]

Hello all,

Can someone give me references which discus the occurrence of phase bleed with polar embedded phases with MS/ELSD detction.

I have been reading some topics, articles and the study of Alltech. But I haven’t found a satisfying answer.

Which species/mechanisms are causing this noise? Not the alkyl chains, in my opinion.

Are there groups who have been investigating this in detail?

Koen

[Mark Tracy]

There is a certain amount of controversy surrounding this subject, and vendors don't like to talk about it. There are two things going on: inherent hydrolytic stability of the bonded phase, and detectability of the bleed. It is hard to say whether a polar-embedded phase is going to have inferior hydrolytic stability compared to a straight alkyl phase, for that reason alone. It is clear that the typical amide/urea/urethane type polar groups have a much greater proton affinity than an alkyl group, therefore the degree of ionization in the MS source is much greater also.

I have observed that for best results, stay in the middle of the recommended pH range, and avoid the extremes. Also, keep the % organic low, because the bleed is not very soluble in water. Also, condition the column thoroughly before connecting it to the MS. If you must flush the column with high organic to remove late eluters, use your diverter valve to keep the junk out of the MS.

[Bryan Evans]

We have columns that can retain polar
analytes and give very little column bleeding.

If you email me, i'll be happy to send you some
information, thank you.

[Koen Hollebekkers]

Thanks Mark for the answer,

I’m not a specialist in ionization or evaporation techniques, so I hope I can learn something.
As you clearly can see in the chromatogram below, you indeed observe more bleed at a higher percentage modifier.
The strange thing is, if the phase is loosing organic material, I would expect loss of retention of toluene. But this is not the case.
Than I would expect the noise is caused by ion, but than I would expect the bump in the beginning.

I agree that column manufactures are not keen to be open concerned this topic.

I really haven’t found one scientific paper that covers this observation.
I hope this discussion is continued!!

Brian, I’m not interested in buying columns. I want to know the under laying mechanism.

[HW Mueller]

How do you know that this is due to bleeding?

[Bryan Evans]

I think there has been a lot written about this topic.
Under acidic environments, the ligands can be cleaved off
(especially for monomerically bonded phases). Therefore, you need
a column with difunctional or polyfunctional ligand bonding.

My guess is that this is column bleeding - and that you will see reduced
retention over more time. Try letting the column sit in 0.5% TFA overnight, test the column the next day and see what happens.

My apologies if it sounded like I was trying to push our product.
But I think you will need a more durable column if you are going to
be using ELS/MS detection for this application.

[Victor]

Koen,

I think what Mark has said is that he sees no particularly reason why an EPG phase should be less stable than a conventional C18 phase. However, he then points out that a MS is likely to be much more sensitive to the column bleed from an EPG column due to its chemical structure. I agree with this analysis of the situation (I hope I did not misinterpret Mark's post).

However, you are showing a chromatogram with an ELSD detector. I have never used such a thing, but I do not see why it should be more sensitive to bleed from an EPG column. In this case it is possible that you are showing (in contrary to the conventional wisdom, and to Mark's wisdom) that the EPG column is indeed losing stationary phase and it is not just a detection issue. However, I agree with HWM- you have not shown that what you are removing is column bleed because of the non-specificity of your detector. It could be junk that has accumulated from your column, maybe from the samples. Do you have acess to HPLC-MS? Then you could monitor likely bleed ions considering the structure of the EPG phase and tell for certain if what you are seeing is column bleed. It would be interesting to know which brand of EPG column this is, but I understand if you do not want to tell us. It is best not to point the accusing finger until one is certain of the data. I don't think ELSD is a particularly sensitive detector, so it would seem that if this were true, you are losing a lot of stationary phase...and yet you have a constant toluene retention time. Does not quite add up........

[Mark Tracy]

All columns bleed. It's a question of how much. Under the conditions you are using, even a good C18 column will eventually lose enough bonded phase to wear out. Polar-embedded phases are no different in that regard. The modern MS is a strikingly sensitive machine, and amides are highly detectable especially in acid media. Good, modern polar-embedded phases should be able to withstand the gradient conditions you showed for weeks of continuous exposure before the retention times are seriously affected.

Fortunately, you probably have only a very small number of masses that make up that bleed signal, and your analytes probably don't overlap them. If you have MS-MS capability, you can dodge most of the problem. The bad news is that the bleed can cause ion suppression, leading to poor or variable sensitivity for your analyte.

[Uwe Neue]

Regarding ELSD bleed, a column with an embedded polar group is no worse than a standard C18. In both cases, you will get some dissolution of the surface of the packing during storage of the column, but it should be much less in a second run. Another way to fix this is to store the column in acetonitrile, instead of mobile phase.

You also need to be aware that under acidic conditions, the stability of different surface bondings is different. Monofunctionally bonded C18s or EPG phases are less stable than difuncitonally or trifunctionally bonded phases. The other thing that you might see is the endcapping reagent, which is for the most part based on a monofunctional silane.

MS bleed is different. You will commonly see less bleed from a C18 phase than from a phase with an embedded polar group. The reason is that the MS is more blind to the C18 than to the EPG phase. However, you will "see" the EPG phase only at specific masses, and as long as your analyte does not have this particular mass in its spectrum, there is nothing to worry about either.

[Koen Hollebekkers]

Nice discussion,

First of all, this isn’t my chromatogram that I posted. This is taken from the Alltech presentation.

The columns that I use are all new, and no real life analysis are performed on them. Noise of matrixes is excluded.

The ELSD and MS type of noise differ slightly from each other, but it comes in both causes from the column.

I doubt if the noise is caused only by endcapping reagents, because polymeric (PLRS-S) columns should in that case give no bleed. Unfortunately ………….

Concluding:
• Polar embedded phases can show more bleed, especially under acid condition, because there somewhat lower stability compared to ODS
• Polymeric bonded phases are somewhat more stable than Monomeric bonded phase
• Store the column in 100% modifier, this reduces dissolving of ligands
• Washing with acid will “cleanâ€

[Victor]

Koen,

No- I think you are confused about the various issues.

3 posters here have suggested to you that the inherent chemical stability of a (monomeric?) EPG phase is likely to be no different to that of a conventional monomeric C18 phase. They have suggested to you, that on the contrary the mass spec is much more sensitive to n molecules of bleed from an EPG phase than n molecules from a C18 column. You have not presented any firm evidence to refute our suggestions. It is possible that you DO have evidence to support your assertion that an EPG phase is inherently less stable. However, we have not seen it. At least you need to show what is coming off your column is really phase decomposition products and not something else, which can be done with a MS, which you seem to possess. An ELSD detector, which has no particular sensitivity to the bleed from either EPG or normal C18 column could then be used to provide further information-but don't rely on data you see in catalogues, which are sometimes put there just to persuade you to buy that product. That is my comment on your first point.

I agree that polymeric C18 phases are likely to be more stable than monomeric C18 phases. However, columns like PLRP-S are made from completely different materials, and the bleed from them is likely to be completely different, from the silica based phases.

Storing the column in 100% ACN is a good idea. Water can cause hydrolyis of the ligands in storage, and these will bleed off when the column is next used.

I'm not sure where the idea of washing with acid came from. As Mr Neue says, the endcapping reagent is likely to be removed much more easily than the C18 ligand under acid conditions. It is quite possible that 0.1% TFA could cause this loss, because it has a low pH. In short, using with acid will cause a gradual loss of endcapping, followed by the main ligand, so is not beneficial to the column lifetime. This gradual loss may not be a significant problem in many cases.

[Koen Hollebekkers]

Thanks everyone,

Well, I don’t rely on the commercial brochures. But the Alltech presentation contained MS spectra and ELSD chromatograms, so they compared both detection method. Although it was still to commercial to me, I liked there approach.

It’s a nice discussion, but much speculation. Maybe nice to investigate and report about it.

I’m always interested in somebody’s opinion……

[Uwe Neue]

Koen,

No speculation on anything that I said, but facts.

[HW Mueller]

If one changes his "much speculation" to "big speculation" I could agree with Koen: His assumption that the above peak represents bleed still stands as speculation, at this point, as far as I can see.

[Koen Hollebekkers]

I agree with all the possible causes discussed here. No discussion about that……
So sorry for my poor vocabulary. Maybe instead of using “much speculationâ€