Short monolithic disks for large molecules

[adam]

Has anyone experimented with the monolithic disks made by BIA Separations. I just came across this in the June 2004 supplement to LC-GC.

It seems pretty clever. The argument is that, given the all-or-nothing retention mechanism of large molecules in reverse phase columns (see below if you want an explanation) the length of the column is essentially irrelevant. Therefore, they manufacture very short - 3 mm - disks. Apparently it is easier to make such short columns out of monolith material than particles (t's ironic that one of the main selling points of monolithic columns is the lesser back-pressure, and here we have an application where backpressure is not the issue at all).

I am curious why I don't see very short column used more often, for large molecules. In fact, I've seen many papers that talk about long narrow ID columns. The narrow ID makes sense only if there is a limited volume of sample. The long length - it seems to me - would never make sense with a large molecule separation?

Thanks in advance for the feedback.
Adam

PS
"All or nothing retention mechansim" refers to the fact that the peptide will just sit on the surface until a critical organic conent is reached. It will then be desorbed. So it doesn't partition back and forth between the 2 phases as a small molecule would).

[tom jupille]

Actually the "all or nothing" model is an oversimplification, and protein molecules do distribute back and forth between the phases as they move down the column.

To still oversimplify: in reversed-phase, retention (k') is a logarithmic function of the fraction strong solvent in the mobile phase {log(k') is a linear function of (%B expressed as a decimal)}. The slope of that function (the S value) can be interpreted as the number of strong solvent molecules required to displace one analyte molecule from the stationary phase. {if you have an ion exchange background, you will see an analogy with the z-value, which is the ratio of analyte charge to salt charge}.

As molecular size increases, so does S (roughly as the square root of molecular weight). That means that large molecules have very steep slopes, and that a very small change in %B can result in a big change in k'. For infinitely large molecules, that would indeed imply "all or nothing". For proteins and peptides, however, S values only get up to around 50 or so for anything that you could do by reversed-phase.

If the preceding two paragraphs make your eyes glaze over (that's that they do to me, and I wrote them! ), he bottom line is that distribution between phases does occur, and column length does matter.

[adam]

Thanks for the feedback Tom.

Could you please give me a book title or a paper that covers what you've stated. I thought I had read my way through enough books over the years, but here you have presented something I've missed. Now I have some homework to do.

[tom jupille]

That whole spiel is a condensed version of "linear solvent strength theory". There's a good treatment of it in the Snyder, Glajch, and Kirkland Practical HPLC Method Development book (I'm out of the office this week, so I can't give you specific page numbers). There's info on that book (and a link to Amazon) on the LC Resources web site (check the sponsors list at the upper right).
If you want to wallow in the math, Snyder and Dolan wrote a review in 1998 in the "Advances in Chromatography" series (Marcel Dekker). If memory serves, it was vol 38 of the series.

[adam]

Tom (or anyone else who can help):

It would be helpfull to have some idea to what extent a large molecule will follow the all-or-nothing rule (discusssed above). It would be very usefull to have some knowledge of this when developing a method. Because then you know how relevant - or irrelevant - column length is.

The more irrelevant length is, the shorter the column you can use, and in turn the smaller particles you can use. So it would be very usefull to know this.

Can anyone suggest a reference where I can find a listing of S values for a variety of compounds. Or can someone suggest a rule of thumb that may help.

Thanks

[tom jupille]

Roughly speaking:

MW ......... S
100 ......... 3
500 ......... 7
1000 ....... 10
5000 ....... 20
10000 ..... 30
20000 ..... 40
50000 ..... 50

I think the empirical relationship is something like:
S = 0.48*MW^(0.44)
but be aware that it's only a one significant-figure calculation.

There's a good readable discussion in the book Basic HPLC and CE of Biomolecules by Cunico, Gooding, and Wehr.