In our laboratory, we used to analyze a very chelating substance on a reversed-phase column with UV detection. We choose a Purospher RP18e colum from Merck (125 x 4.6 mm, 5 µm). The sensibility was not to great for our application. In order to improve the sensibility of our method and having in mind that this substance can form fluorescent complex with lanthanide as terbium, we tried to use this phenomenon.
We used a solution of terbium chloride in 0.1 M acetate buffer pH 5.6 as a reagent to form complex with our substance to analyze and injected it on our column with 0.1 M acetate pH 5.6 / acetonitrile 90/10 as mobile phase. The separation was made isocratically at 30°C with a flow of 1 mL/min.
We were very astounded to see that the peak was eluting before the dead time. Injecting decreasing concentrations of our complex showed decreasing peaks always eluting before dead time.
Up to now I thought that it was impossible that a product could elute before dead time. How can we explain this phenomenon ? Did anyboby share the same experience ? And how could we correct this problem ?

Thank you very much,

Gilles DELATTRE

The standard explanation for this phenomenon is that the compound eluting before the void volume is excluded from the pores of the stationary phase. Pores on silica make up a sizeable portion of the column's volume. Therefore, these excluded compounds have a much shorter mean path through the column and a much smaller volume to traverse. I'm not absolutely sure, but I believe Purospher is a porous silica vs a non-porous one.

This may be a very interesting result. What is your pore size and what the expected size of your uncomplexed and complexed molecules?
We have had results like this with proteins on RP columns. In comparison to GPC column results a simple explanation based ONLY on size exclusion seemed unlikely.

Is the complex anionic or cationic? At pH 5.6, some of the residual silanols are likely to be ionized. This raises the possibility of ion exclusion for anionic complexes.

-- Tom Jupille / LC Resources Inc.

Thank you for your answers,

Purospher from Merck is a porous phase with a porosity of 120 A. I think it is hardly probable that the complex formed may be sterically excluded by the column. The hypothesis from Tom Jupille is interesting. In our analytical conditions, the complex seems to be anionic. Adding an additive as a ion-pairing salt in our mobile phase can perhaps correct this problem ???

Gilles DELATTRE

Actually, I would have thought that 100 mM acetate (well, OK, its actually only 90 mM in the final mobile phase) would be enough to "swamp" the ion exclusion, so any suggestions I can make at this point would be pure guesswork. That seldom stops me, so:

An ion-pair reagent would certainly be one possibility (any chance that it might disrupt your complex?).

Another would be to increase the buffer or salt concentration (that is commonly done to combat ion exclusion effects in protein SEC separations).

A third would be to substitute a neutral salt (like NaCl) for part of your buffer (the idea here is that acetate itself is a weak ion-pairing reagent -- in this case, of the wrong charge -- and may be contributing to exclusion).

From your name, I would assume you are a Francophone. I'm not sure if there is a French equivalent to the (American) English acronym "SWAG" (Scientific Wild-Ass Guess), but that's all those suggestions are.

Bonne chance!

-- Tom Jupille

Thank you for your good advices,

You have well guessed, I'm french. We have not that kind of expression but I've understood its meaning.
I will try your suggestions but for the moment I've got some trouble with the pump.

Gilles DELATTRE

would it be possible to separate the compound of interest and then derivatize it with terbium post column? If I understood the problem, without the terbium, the separation was acceptable. Terbium was added to improve detection. If the terbium complex forms rapidly, terbium could be added after the column but before the detector.

Delattre-I’m not sure you should dismiss exclusion
as the source of the early elution you are
experiencing. Depending on the coordination number
of your complex (chelate?) and the ionic
(molecular?) volume (in cubic angstroms),
exclusion may well be occurring. For example, a
significant difference in retention volume exists
(1.85 vs. 1.64 ml) between sodium nitrate and
sodium nitroprusside on an ordinary Zorbax C-8
column. This exclusion from pores of the packing
and the static interstitial volume is entirely due
to the difference in ionic volumes of the salts.
To test this out you might try columns w
increasingly larger pore sizes and see if the
elution time of your complex increases.

While this may be an interesting exercise, in
order to do any useful chromatography you need to
get the analyte away the vicinity of the void
volume. So a suggestion;
As long as you have fluorescent capability can you
do a post column derivatization i.e. do the
separation first and the enhanced detection after
the sepn? It’s more involved but sometimes it’s
the only way - for example either amino acids or
carbamates at very low levels.