[Babcock_Hall]

We are trying to extract a carboxylic acid into bicarbonate, protonate it, then extract it back into ethyl acetate to free it from 4-nitrophenol, a by-product of the synthesis.  The last two times we attempted this, we found some 4-NP in the second ethyl acetate layer.  The pH of saturated bicarbonate appears to be about 8.5, and the pK_a of 4-NP is about 7; therefore, the 4-nitrophenolate should predominate.  I was thinking about adjusting the pH of the bicarbonate down to 6 and extracting with this, in the hope of bringing along less 4-NP.  Does this sound reasonable?  I have to admit, I am not sure that it does.  Thanks in advance.

[Babcock_Hall]

The problem with trying to acidify even to pH 6-7 is that some fraction of the bicarbonate is lost as CO₂.

[discodermolide]

Why not try adjusting the pH with a buffer solution of say between 5 and 6? Then extract. I assume you are not doing this on a large scale?

[Babcock_Hall]

We are working at the gram scale, approximately.  Do you mean some buffer other than bicarbonate?  We have a number of buffers available to us, such as succinic acid and phosphoric acid, as well as some more esoteric (at least to organic chemists) substances, that are also more expensive.  But they are cheap if they work...

[discodermolide]

We used to buy pre-made buffer solutions of defined pH. Aldrich may have them. I think they were hydrogen phosphate based.

[Doc Oc]

I'm assuming this is the same compound you're running your flash column on.  Can you confirm?  That compound will not behave terribly well in extractions unless you're pulling something very polarized away.

Many phenols that I've worked with don't behave well in extractions, even under strongly basic conditions, and 4-NO₂-phenol is no exception.  You should also be aware that the phenoxide is bright yellow and easy to spot, but the phenol is colorless.

[Babcock_Hall]

Doc Oc, Yes, we are trying both extraction and silica gel chromatography as purification methods to rid ourselves of the 4-NP.  In the extraction, much of the 4-NP ends up in the first organic layer.  This layer looks yellow (even though the 4-NP should be neutral).  The compound of interest goes into the aqueous layer, then after acidification, goes into the second organic layer.  In the past, we have seen variable amounts of 4-NP by NMR in the final product.  The extraction might have gone with higher recovery yesterday than previously, but I don't have hard numbers yet.  We don't know how much 4-NP came along for the ride yet.

Discodermolide, We make buffers all the time in our lab, so making them is not a problem.  I suppose the question is whether the acidification/second extraction would go smoothly or not with buffers other than bicarbonate.  I cannot think of a reason why it would fail, but then neither have I heard of people doing an extraction with a phosphate buffer before.

[discodermolide]

It's not so much the buffer it's the pH I was thinking about and the troubles you had with bicarbonate. I've used buffers for extraction before, especially if the pH was important for the survival of the compound.

[Babcock_Hall]

I would imagine that a high concentration (at least 1 M, maybe higher) of buffer is a good thing, both to keep the ionic strength high, and to make sure that there is enough of the basic form to deprotonate the carboxylic acid and pull it into the aqueous layer.

[Babcock_Hall]

I would imagine that a high concentration (at least 1 M, maybe higher) of buffer is a good thing, both to keep the ionic strength high, and to make sure that there is enough of the basic form to deprotonate the carboxylic acid and pull it into the aqueous layer.  We hope to give it a try this week or next.

[opsomath]

I am thinking that these compounds are not so different in acidity that it will be easy to do a differential extraction based on pH. I wonder if, perhaps, you could sublime off the nitrophenol instead?

For the extraction, my first attempt would be to use a weak bicarbonate solution to extract the acid, precipitate with HCl, then recrystallize the whole thing from some kind of alcohol or alcohol/halogenated solvent mixture. You won't get a perfect separation due to solubility of 4NP, so you do the best you can with the extraction and let crystallization do the rest. The nitrophenol should be highly soluble in, say, chloroform, so maybe something like a chloroform/methanol mix would work well for your crystallization depending on how polar the carboxylic acid product is.

[Babcock_Hall]

Sublimation...I had only vaguely thought of that before, so thank you.  We just tried the phosphate buffer idea today, but we don't have data on how well it worked yet.  The pK_a of 4-NP is 7 and the pK_a of our carboxylic acid could be 3-4.  We tried a phosphate buffer that was between pH 6 and 6.5, made by mixing two saturated solutions (NaH₂PO₄ and Na₂HPO₄).  I cannot recall with certainty, but I think that I measured the pH of a saturated bicarbonate solution on hand and found it was about 8.6.

[Babcock_Hall]

It seems as if the phosphate buffer leads to a smaller concentration of 4-nitrophenol in our amide product than bicarbonate did.  Our recovery is not great in either case (less than 50% if I am not mistaken).  However, the improvement in purity is very welcome.

[discodermolide]

That's good to know, especially the purity improvement.

[Babcock_Hall]

We performed two extractions using phosphate buffer again this week; in both cases the buffer had a pH of about 6.3.  The compound of interest are (iso)nicotinamide derivatives of BOC-lysine or BOC-ornithine, both of which have a free carboxylic acid group.  Again, the purification method is to extract the compound into aqueous phase, lower the pH to 3, and extract it back into ethyl acetate.

The recovery at the final step has never been great, but this time was almost nonexistent.  We checked the first EtOAc layer and found some of each product, along with the 4-nitrophenol that we expected to find.  The ratio of moles of product to moles of 4-NP was about 30% for one sample and 60% for the other.  We also noticed in both cases that an oil formed at the interface between the ethyl acetate and the phosphate buffer.  In one of the cases, we collected that oily substance separately and it is now an amorphous, light-yellow, sticky solid.  It is substantially our product by NMR, but it has large masses of white solid, presumably phosphate crystals.  It is unclear to me whether or not the product is in the carboxylic acid or in the carboxylate form in this fraction (phase).

There are several questions that naturally arise.  One, why did we not pull out more of our product into the aqueous buffer?  I can think of two explanations:  incomplete mixing and potentially poor choice of pH.  Two, why did we get an oil at the interface, and what should we have done with it?  Three, what should we do with the oil/salt crystal mixture?  This is where the question of the protonation state of the product becomes important.  The oil/salt mixture is much more pure than the ethyl acetate fraction.  However, it is an open question as to how to solubilize it.  I think I might try to add methanol to dissolve the oil but leave the salt crystals.  Then evaporate the methanol and ?