[Babcock_Hall]

Hi Rolnor,

Good to hear from you again!  Since I have two samples, I will try repeating the Kaburagi-Kishi method on one and using reverse phase methodology on the other.  On my shelf I have an old, probably unused 20-gram sample from JT Baker, of Bonded phase-octadecyl (C18), having an average particle diameter of 40-µm and described as being for flash chromatography.  I also have a product from Analtech (catalog #50010) that is RPS hydrocarbon impregnated silica gel, 35-75 micron.  I found a description at Amazon.  If I am reading this correctly, the hydrocarbon impregnated product is not covalently linked to the silica and would come off in a sufficiently non polar solvent (if one is not careful).

"Silica gel and bonded phases are based upon a 150A pore silica gel, 35-75um particle. SPICE Packings are available in bulk quantities of 100 grams, 500 grams, 5 kg, and 10 kg. For larger quantities call for a price quote. Analtech offers two types of reversed phase packings for column chromatography: A C18 octadecyl bonded silica gel and a hydrocarbon impregnated silica gel. The most common reversed phase adsorbent is the octadecyl (C18) bonded silica gel. The C18 hydrocarbon is chemically anchored to the silica gel support. This material is compatible with all aqueous and organic solvent combinations. The carbon content is about 14%. Pore size is 150 Å Particle size is 35-75 microns. The hydrocarbon impregnated silica gel is a column grade equivalent of Analtech's unique RPS UNIPLATE. The long chain hydrocarbon is not chemically bonded to the silican get but is adsorbed onto the surface. This material is compatible with classical reversed phase eluting solvents such as MeOH/Water and Acetonitrile/Water. Less polar organic solvents will solubilize the hydrocarbon. The Reversed Phase Separation (RPS) column material is ideal for prep scale separations established on RPS UNIPLATES. Carbon content is about 5%. Pore size is 60 Å. Particle size is 35-75 microns."

I seem to recall from a previous discussion that packing and running a column of reverse phase material is quite different from running a silica column, perhaps needing more pressure. 

Is one of these two products suitable for our application?  If not, what should I buy?

[Babcock_Hall]

I attempted a second Kaburagi-Kishi cleanup on one of the two samples, in which there were more moles of tetrabutylammonium ion than product.  Now there are fewer moles of tetrabutylammonium ion than product, but the impurity is still present.  I don't see the point in running this method  a third time.  Instead, I will try one of the two reverse phase methods outlined in my previous message.  If I get a satisfactory result on one sample, I will also purify the second sample.  Is either of these products suitable for use in a gravity or low pressure column format?  Does anyone have any sort of procedural reference?  Thanks for any ideas.

[Babcock_Hall]

Tetrabutylammonium will stick very hard to RP-gel, it will not move. Just 50%MeOH/H2O will do the trick I think. Load the column with pure MeOH, then wash a little with 50%MeOH/H2O before loading the sample.

A few weeks ago we repeated the Kaburagi-Kishi cleanup, and we see a lower integral of tetrabutylammonium ion than we saw before.  In round numbers we went from 0.6 mole fraction to 0.4 mole fraction TBA ion.  Today we tried two kinds of RP C-18 TLC plates.  In 50% aqueous methanol the R_f of the presumed product is close to 0.5.  On my shelf I have an old, probably unused 20-gram sample from JT Baker, of Bonded phase-octadecyl (C18), having an average particle diameter of 40-µm and described as being for flash chromatography."  I was planning to try a purification with this material as the stationary phase. 

One, will using pressure that is typical for flash chromatography produce a sufficient flow rate?

Two, I seem to recall that the capacity is lower for C18, but I don't remember where I heard this.  I was thinking about using C18 silica to product at a 100:1 ratio by mass.  Does this sound about right?

[rolnor]

Yes, I agree, the capacity is lower, att least for low-pressure RP-chromatoghraphy.
Thanx! I have had some serious problem with blackmoldtoxinepoisoning, it got in my livingroom via a pipe down to the basement, catastrophic. I am much better now, got some "progressive brainfog".
1:100 is probably good!

[rolnor]

I am not sure if the flow will be good enough, I dont recall the diameter of the RP-gel I have used, sorry.

[Babcock_Hall]

We will try dipping our plates in PMA to see if we can find where the TBA is.  Our long-term goal is an anti-fungal compound...

[rolnor]

Nice! Anti-fungal, is this a co-project with the industry? Is the compound meant to be oraly active? The TBA can be hard to spot, it does not react with just about anything, maybe molybdate-spray can visualize it.

[Babcock_Hall]

The PMA apparently stains the C-18 stationary phase, and we did not see any unambiguous spots above background.  Perhaps there is another stain that would work, but the more I think about it, the less likely it seems.

The most promising version of our compound would not be orally active, but it might be suitable in the form of an injection.  That is a long way off, however.

[Babcock_Hall]

Interim update: We tried using C-18 coated silica twice. The first time we isolated what appears to be the aromatic portion of the molecule minus the glycosidic portion. The second time we obtained two small fractions, each of which appear to include the glycosidic portion. But the two fractions are not identical by H-1 NMR. There are additional fractions that we have not analyzed yet, but by their R(f) values, these might contain the aromatic portion of the molecule. The good news is that I don't see evidence of the tetrabutylammonium ion. My tentative conclusion is that over the course of multiple purification attempts, some hydrolysis of the bond between glucose and the aromatic ring has occurred. I am looking into the possibility that this bond isomerized.

[rolnor]

Ninhydrine would work if you have a amide/amine I think.
Great that the TBA is gone! Maybe you need to control the pH, if you use reagents like TBDMSOTf and the use water in the workup you need NaHCO3 or TEA as acid scavenger, also if you have trace of triflate left after workup, the acid may form later and hydrolyse upon storage/evaporation.

[Babcock_Hall]

We checked the fractions with reverse phase TLC, and we mainly detected with UV light.  PMA stains the whole plate, if I recall correctly.  There is no amine, but there is a vinyl group present.  There is a chance that KMnO4 would work as a stain.

I think that the acid may come from Dowex-50 in the Kaburagi-Kishi technique of removing the TBA ions.  Perhaps adding a greater proportion of CaCO₃ relative to Dowex-50 would buffer this.

[rolnor]

KMnO4 will react with the carbon in the C-18 but initially the vinyl will stain, it reacts faster.

[rolnor]

There is anisaldehyde/H2SO4/MeOH i think, that stains almost anything, it could work.