Column lifetime:- C8 Vs. C18???

[rick1112]

Hi

We recently observed that our C8 column goes bad (my peaks starts showing tailing etc..) -sooner than C18 columns???
Could anyone tell the reason for this??
Is life of a column depend on the ligand chain length, density of group attached ??

Thank you

[coquim]

[quote]We recently observed that our C8 column goes bad (my peaks starts showing tailing etc..) -sooner than C18 columns???

i suppose you are using a precolum....in this case i´d check it before to start thinking the column is not working properly...

[Victor]

Yes-this is very well known that the stability of alkyl bonded phases decreases with the chain length. Thus C18 columns are generally more stable than C8, which are in turn more stable than C4 or even C1.

The stability of columns will depend on the mobile phase conditions that are used. Low pH causes hydrolysis of the bonded ligands. High pH causes dissolution of the silica itself, but I believe also can cause some hydrolysis of bonded ligands (independent of the silica structure dissolving, and the ligands falling off with it....)

C1 is relevant as it is the endcapping agent on many columns with longer chains. The C1 is likely to be removed first, especially at low pH. This process may change the selectivity of the column, which is why some manufacturers offer non-endcapped columns (e.g. some Zorbax columns).

Probably, a longer chain hinders access of aggressive reagents to the surface of the column where ligands can be hydrolysed.


There has been a lot of debate about the effect of density of bonded ligands on the stability of the phase. I believe the Zorbax columns are made with diffent bonding densitities which they claim suit different mobile phase conditions.

[Kostas Petritis]

Is the chemistry between the two columns exactly the same (i.e. silica based, endcapping, etc...)? For example are we talking about Zorbax or XTerra or Hypersil (or whatever other TMs) of C18 vs. C8?

[rick1112]

to Kostas Petritis
..yes the chemistry of the two columns are same (they are both from ACE)

to Victor...

..dont you think as C8 can have better converage of all the silinol bonds and may offer better protection??

..also i would like to know the main reasons behind stationary phase/column going bad???

[Uwe Neue]

Tailing can be caused by many different things, from the buildup of debris at the column inlet to a dissolution of the stationary phase, which results in silanols, which results in tailing for analytes with amine functions. There is a section in my book on "HPLC Columns" that is dedicated to column troubleshooting, and it has nearly 40 pages...

Assuming that you are talking about analytes with amine functions that can interact with silanol goups, then the next question is what is the limiting element in surface coverage. For some phases, this has nothing to do with the chain length, and under these circumstances, it is possible that a C8 chain will be split faster than a C18 chain. In the general case of monofunctional silanes and full coverage, you will have a slightly higher coverage with a C8 than with a C18, and under these conditions, one can demonstrate that a C8 is better than a C18.

Otherwise, I agree with everything that Victor has said, especially the point that the C1 endcapping is the item that comes off first under acidic conditions, nearly independent of the other properties of the packing.

[Victor]

Rick,

You are right that C8 can give a higher coverage than C18, as Mr Neue has pointed out. A new C8 column will often give better peak shapes for difficult compounds than a C18 column, and a C4 column may be better still. However, this does not necessarily mean that C8 columns offer better protection and stability, as the reverse is observed experimentally-also by yourself. We then have to try to explain the experimental results. Maybe a forest of C18 chains hinders the access of aggressive reagents to the surface more than a slightly thicker forest of C8 chains. I do not know the data for the ACE columns, but my guess is that the surface coverage of C8 chains (in micromoles/square metre) will be somewhat higher than that of the equivalent C18 column, but that there will not be a huge difference in the figures. This factor may be of some relevance to the argument above.

[tom jupille]

An explanation that I heard a number of years ago from Fred Regnier went something like this:

The silyl ether linkage that holds the bonded phase to the substrate is continually being hydrolyzed at low pH. Something like a C18 group, however, is so hydrophobic that it's "pressed" in place by the water molecules in the mobile phase (or, to look at it another way, it's so insoluble in water that it stays sorbed to the stationary phase surface), giving the bond a chance to re-form. Every now and then, one gets far enough away for the cleavage to be permanent, and so the column packing does degrade.

All other things being equal, the less hydrophobic the bonded phase, the less tightly it's locked in place and the more likely it is to be permanently lost.