How to get rid of mobile phase salts?

[labcat]

I am performing semi-preparative HPLC using a tetrabutylammonium/ phosphate mobile phase. Following first purification step, sample are dried by lyophilization, and I obtain a small amount of the analyte and a lot of tetrabutylammonium /phosphate. Then I need to perform a second purification step, but I have to keep injection volumes very low, otherwise chromatography is so poor that I cannot purify at all.
Is there a simple way to get rid, or at least to reduce, of the mobile phase salts?
Thanks for any advice!

[Yury Zelechonok]

You can use flush cation-exchange to get rid of tetrabutylammonia. But phosphate is not easy to remove since your analyte most likely is acidic and hydrophilic. If you compound of interest are peptide you probably can use dialysis, but I am not sure if it is available for lab scale separation.
If you can go with different technique from the beginning you will not face this problem now. SIELC Technologies offering columns with embedded ion-pairing reagents basic and acidic. In your case I would use Primesep B2 column for first separation step, using water/MeCN/ TFA or water/MeCN/Ammonium Acetate as your mobile phase. Both MPs are totally volatile and can be evaporated or lyophilized before the next step. See this example from our application web site:
http://allsep.com/makeChr.php?chr=Chr_067

[HW Mueller]

Labcat can probably do this first HPLC with volatile buffers on his column as well.
What is your stuff?

There are all kinds of small dialysis devices on the market (check google, just saw a tiny mini, maybe Pierce?). If you really have a macro-molecule I would try ultra-filtration first (Millipore, etc.). Also a SPE inbetween might do, or even an extraction with an organic spolvent. Also prep-TLC is nice.

[Uwe Neue]

Since you do not need a high-preformance separation, a simple desalting step on a hydrophobic packing with a large particle size might do. If your sample is a base, you load the sample dissolved in ammonia, ammonium phosphate will break through, and your sample and some of the TBA/phosphate may stay behind. If your sample is retained much more than the TBA, then the addition of some organic to the elution solvent will do the job.
If your sample is an acid, you dissolve and acidify the sample with a solution containing formic acid or similar. You load it onto the device, and the TBA/phosphate and the TBA/formate are likely to elute early, while your sample is retained.
You may be able to play similar games with large particle size ion-exchangers.

[labcat]

Thank for your answers. I realize that some more details are needed.
My samples are impurities from an API, molecular weight around 500 (no dialysis), 1st pKa about 4.5, 2nd pKa about 5, 3rd pKa about 10, does not tolerate extreme pH (so I cannot lower pH to 3 and extract in organic solvent, or use SPE), amount about 0.1-0.4% of the main peak
Second purification step is performed in ammonium acetate, but as the original analytical method requires TBAHSO4/phosphate I must firstly separate the impurities with this mobile phase to individuate them (I have an UV detector, no MS).
Then I inject them with ammonium acetate, but injecting 50-60 µl (on a 4.6mm column) is quite the limit, otherwise peak shapes are very poor. If it were possible to get rid at least of a part of the salts in a simple way it could be useful, but I’m afraid that patience will be the best tool…
I’ve seen some desalting devices, like small cartridges that can be installed on a Rheodyne valve, but I guess they fundamentally are chromatographic columns, so I wonder if they can retain polar molecules, dissolved in water, without lowering pH.

[Nagy]

I am so happy to join this great forum and great discussion.
I usually use very small ODS open colum to get rid of the Buffer salts. but frist dry the sample and add water. then all the salt will dissove and then apply to ODS column and wash several times with water. then use Methanol to elute the organic compound. I don't know if it is restricted to my case only or it can be a general one. any how it works with me.

[labcat]

Thank you. I think the small columns you use are similar to the ones I have seen. I will try them.[/quote]

[Uwe Neue]

Labcat,

The same materials that I recommended earlier are also avaible in small column formats (Oasis columns from Waters, 2 mm x 20 mm). They are essentially nicely hydrophobic columns that might to the job for you. The reason that I had suggested off-line methods is that it appeared to me to be a simpler approach (no valves required), but either way is fine.

[Yury Zelechonok]

To labcat

This recommended desalting procedure will work if your compound is more hydrophobic in formic MP than TBA. But if you use TBA because your compound is very hydrophilic in a first place, then the short ODS column or cartridge will not work for you. All salts and your compound will elute together in column void. According to your pKa values you have an acidic compound. And as Uwe suggested I would use formic acid to reduce acidity of your compound but use short amino-RP-column instead of short RP column. On such columns TBA, which has positive charge as well as the column surface, will exclude from interaction with SP and will elute in prevoid volume. Sulfate or phosphate will retain strongly by ion-exchange and since formic acid is very poor electrolyte you may expect their very long retention. In opposite, your compound will elute soon after the void due to small hydrophobicity + small ion-exchange. If retention of the compound is too long you can always increase the amount of the organic and make it faster.
By the way what is API (atm. pressure ionization)?

[tom jupille]

API = "Active Pharmaceutical Ingredient"

[Yury Zelechonok]

Thank you Tom