I have a question about injecting poly(ethylene oxide)[PEO, or PEG] using a tosohbiosep TSK-2000SW, which is
a silica based aqueous size exclusion column. Upon receipt of the column we injected PEO's of various molecular weights to measure the calibration curve. Additionally we injected various proteins(lysozyme, BSA), Dextran's of various molecular weights and some small molecules. After a number of months of running only protein we returned to injecting PEO and found that the PEO was selectively retained longer than the "total" volume, i.e. that the PEO was not behaving in the size exclusion branch, but rather in an adsorption branch. The retention for all the other molecules(proteins, dextrans, etc.) remained the same. This is the second time that this same strange phenomenon has happened to us with the tosoh TSK-2000SW column.
Has this happened to anyone else?
What happened to the column?
Is there any way to return the column to its original state, i.e. size exclusion behavior for PEO?

What you are experiencing is pretty typical of G2000SW-G4000SW, and G2000SWxL-G4000SWxL materials when using PEG. These packings have a load of SI and OH groups just waiting for something like PEG to pass by so it can latch on. I first saw this problem back in 1988 at a company using PEG as part of an enzyme reaction. They got about 20 injections in before the column lost all resolution ability and was subsequently replaced. We did try to clean-up the columns but always failed in bringing the column back to the original conditions required for the job. You may want to consider another SI sizing column if the column life is suffering.

Why does PEG interact with the column?

TSK Gel Super SW 3000 columns also have done strange things, as mentioned before. The separation of an antibody from it´s Fab fragment was almost to base line when the columns were new, with phosphor buffered saline (PBS)as MP. When TFA, or even more so pentafluoropropionic, was used in the MP the resolution was lowered drastically. After that the resolution was bad even with PBS as MP, but improved somewhat SLOWLY.
Addition of alcohols like iso-propanol also deteriorated resolution including that of subsequent PBS runs, though this appears to be ~ reversible. This seems to indicate that strong attachement of the fluorinated acids or alcohol changed the column´s performance (via a changed surface interaction with the protein). In addition the acids changed the surface permanently, I think this must be more than simple adsorption....some chemical reacion must be behind that.
Contacts with Tosoh yielded only the information that the silica has received a hydrophilic surface. Hplcman, do you have any more info, for instance: what kind of OH groups?
If PEG is merely adsorbed the effect should also be reversible.
The recommendation to use another column is not as easy as it seems: There appear to be no columns that even approach the mentioned resolution.

If something has happened to the column causing PEO to be adsorbed to the column, what would happen if you added a small amount of a low MW PEO(or ethylene glycol or different MW PEO than you are expecting in your samples) to the mobile phase? Would the PEO in the mobile phase "neutralize" the "active" sites where the PEO injected is being adsorbed? This is assuming this would not cause a problem with your detector (RI for example). Sorry if these are dumb questions.

TSKgels have been around forever and still use much of the older technogies for preparing silica. Unfortunately, I do believe that's where the origin of the problem still is with regards to PEG products and the coating issue. Tosoh is not about to change its procedures to new preparation schemes as long as the old scheme works for 98% of the successful applications.

Again, cleaning never solved the problem. We were never sucessful in reversing the PEG off.

Using small amounts in the mobile phase I don't believe would be advisible in this case. I see it more as an adsorbing or total coating issue. Whether using it in the mobile phase or in an injected sample did not seem to make a difference.

Those having troubles should contact the resident expert at Tosoh - Roy Eckstein. He has been working with TSKgels for some time now.

I rarely defend the strategy of a competitor, but the criticism of hplcman of TSK gels and older technology is unfounded. A manufacturer should not ever change the technology by which an adsorbent or a SEC gel is prepared, since there may be repercussions in the use of the product by an established user. If one wants to change the technology or upgrade the technology, one should introduce a new product with a new name.

Dear all,
yes, Tosoh has upgraded its technology, to improve performance of its SEC columns. And the new product is called Super SW 4µm. We at Tosoh Biosep are more than open to take any criticism about our products, and we communicate and discuss all "problems" with our colleagues in Japan. And even we can influence what direction our people in product development are going. So, for the US please contact Roy Eksteen, for Europe please send your "problems" and comments to me, g.kratz@tosohbiosep.de! Thanks and regards, Gerhard.

Dear Joshua,
SEC of Pegylated Proteins was first presented as a poster at HPLC ´96 in San Francisco, CA. Ratto, O´Conner, Distler, Wu, Hummel Treuheit, Herman and Davis from Analytical Research and Development Department, Pharmaceutics and Drug Delivery Department, Amgen Inc. Thousand Oaks, CA 91320-1789 published their work also in the J. of Chrom.! 10% ethanol added to the phosphate buffer mobile phase solved the problem. In the paper it is explained, what happens to the silica surface etc.! If urgently needed, I can fax you some pages of this paper. Please let me know. We also have some new applications available for pegylated proteins. Hope this helps. Gerhard