Generations and Columns

[fstabile]

Hi,
I've seens a few times things about new generation columns.
I know from articles on the internet and in catalogs that these new generations are more stable, reproductible and everything else. Just that I cannot find them anymore.
We are having different problems on columns that I believe are from older generations.
Does anyone know where I can find information on the subject (small articles) and/or lists of the different generations with the columns rattached to it.

[Uwe Neue]

“Newâ€

[tom jupille]

"Generations" of packings is largely marketing hype (in my not-so-humble opinion ). The technology has continuously improved so that newer columns are likely to be better (more reproducible, less tailing, etc.) than older columns.

To the extent that there was a "generational" break, it came in the early 90's, as indicated by Uwe. Older, "lower purity" columns are often designated as "Type A" ("acidic") and newer, "higher purity" columns are designated as "Type B" ("basic"). If you need to find out which is which, the most general place is the PQRI column selectivity database on the USP web site:
http://www.usp.org/USPNF/columnsDB.html
(the USP database is first; you have to scroll down the page to get to the PQRI database). There are over 300 columns, each identified as A or B.

[fstabile]

Can the new techniques also mean better reproducibility between batches?
The column we are questionning is the (ever so popular it seems) ubondapak from waters.
I don't know where, but I think I've heard that sometime, when they change the sorbent packing, things like retention time or resolution between two peaks can change, greatly.

Maybe going more in specifics could clear up the question.
The situation that we enconter is an analysis of impurity. We inject a very concentrated sample to 'see' small peaks. A important peak appears just before the massive peak.
Before, the peak was a massive tailing peak. Now it is a fronting, with 'stairs'.
So either, there's a unknown impurity that we can now detect. Or something new happens.

[Uwe Neue]

Well, if you want to get the best reproducibility that is possible, I suggest to convert your method to a Symmetry column. microBondapak was developed in 1973 with the best ideas possible at that time. Process improvements were made between 1985 and 1990, which improved its reproducibility roughly 2-fold. However, there is a limit how far one can go with 1973 technology. Symmetry was designed from the beginning with high reproducibility in mind, and the process variation has been reduced further over its now 14 years of history.

Now, to your specific problem: when you overload a column, it is expected that you get tailing peaks. This is completely normal. You should NOT get a fronting peaks with stairs from overload. If your column is otherwise OK, i.e. non-overloaded peaks look fine, my suspicion is that something has changed with the sample that is injected. Anything like it is now a salt form and in the old times it was not, or the injection solvent has changed, or there is a matrix components that is not the same as it used to be or other related things. While one needs a more detailed description about the event that you are observing to do further troubleshooting, it appears on first glance that your problem may have NOTHING to do with the reproducibility of the packing.

[fstabile]

Well, thank you for all this information.
We'll be looking for a new method of separation.
I still don't know what happened, but we'll certainly go look in the reagents we use.

Thanks again.

[Bryan Evans]

You should also consider Imtakt (for new generation columns).

Triton X-100 on Cadenza CD-C18, 500x4.6mm, 3um (100,000 plate column):
http://www.imtakt.com/TecInfo/TI150E.pdf