[Babcock_Hall]

We have an amino acid in the form of a salt with trifluoroacetate, and the side chain is not charged.  We do not presently known the identity of the impurities.  We are planning to try chromatography on Dowex-1 in the acetate form, but for a related molecule that we are also working on, we think that Dowex-based resins would be a poor choice.  Therefore, we are looking at alternatives.  We have some C₁₈-silica with a 40 µm particle size.  We also have a few C₁₈-TLC plates that could be used for method development, but I have never run a gravity or flash column with this solid phase before.  Would it be reasonable to try a column-based approach?  Does anyone have a recommendation for a book which discusses this technique?  I checked Advanced Practical Organic Chemistry (Leonard) and Practical Organic Synthesis (Keese), but I did not see anything. 

If not, what should we be using, Dowex-50, and eluting with HCl?  We want to avoid the classical method of elution from Dowex-50, in which the amino acid is eluted with ammonia.  Some years ago we worked out a way to purify amino acid derivatives with charged side-chains using a volatile buffer and Sephadex-based matrices near pH 8, but I don't see that method being useful in this instance, owing to the lack of a charged side chain.

[rolnor]

RP-chromatography is similar to straight-phase. Its best to pack the column in pure MeOH, not H2O/MeOH, then elute with some of the mobile-phase you want to use before applying the sample on the column. The RP-TLC-plates may need to be heated to around 120°C before use. This can be done by putting them on a heating-plate. In general, separation on RP is a little less efficient then on straight-phase silica. A gradient can be good, then the mobile phase composition is not so critical.

[Babcock_Hall]

We may need to use a step-gradient, owing to instrumental limitations.  Is the loading on RP-silica lower, more like 100 to 1 by mass?  That is the impression that I received in reading some literature from Teledyne/ISCO.  What is the reason for heating the plates, removal of water?  Thanks for the tip.  Yesterday I was leaning toward Dowex-1 in the acetate form, but today I am leaning toward Dowex-50.

[rolnor]

The loading is maybe 1:40 but the separation is not so good, tailing. If you use buffer its better just like HPLC. I think 0.1% TFA can be used, more acid will hydrolyze the RP-gel. I had problem with the gel on the RP-plates, it came off when developing the plates, heating them removed this problem. Maybe it was old plates.

[Babcock_Hall]

We may eventually try reverse phase chromatography on a different product, an aromatic glycoside, to remove a small amount of residual Bu₄NX, where X is probably fluoride ion.  We will need to scale up our synthesis.

[rolnor]

I think RP-gel is expensive so that could be a problem depending on the scale. I think it should completely remove the very lipophilic Q-salts.

[Babcock_Hall]

We obtained a 20-gram bottle (40-µm silica bonded to C₁₈)from a lab where the PI was retiring.  I am tempted to try a small-scale purification on a commercial glycoside first, as practice.  Does anyone know of a "methods" article on this technique?

[rolnor]

I dont have that. If you pack the column with MeOH and elute with a little 10%MeOH in H2O, then add the sample and elute with a step-gradient 10%MeOH inH2O to 90% MeOH in H2O it could be a start. Its really not difficult. Its a good idea to test the gel with a commercial sample, not wasting precious product.