[Babcock_Hall]

We are running a Horner Wadsworth Emmons reaction in which the product has a trifluoromethyl group versus a methyl group in another product.  Would we expect the same R_f in TLC or a higher one for the trifluoromethyl compound.  In previous HWE reactions the product is typically slower (meaning having a lower R_f value) than the starting aldehyde.  In the present reaction the starting aldehyde and the presumed product co-migrate in two different solvent systems.  Qualitatively the product lane, a vinyl sulfone, gives a stronger reaction with permanganate than the starting aldehyde.  We can detect the presence of the other starting material (a phosphonate) in the product lane, although it is weak.

The two interpretations are that no reaction occurred and that the product and starting material have the same R_f.  I seem to recall reading that the trifluoromethyl group is more lipophilic than the methyl group, but I do not know how general this is.
https://onlinelibrary.wiley.com/doi/abs/10.1002/jps.2600751016

[rolnor]

They are to similar to ad any significant difference I think. If you use neutral permanganate the alkene should stain much faster than aldehyde.

[Babcock_Hall]

I took a P-31 NMR spectrum of the reaction mixture, and I see one peak that I tentatively attribute to starting material and one to side-product, roughly in equal proportions (there is a third peak whose identity is unclear to me).  The F-19 spectrum is more difficult to interpret, but it is consistent with such an interpretation.

[rolnor]

Its becoming a little to complex to give any advise without knowing more about the structure. Is the trifluoromethyl group present in the aldehyde?

[Babcock_Hall]

The trifluoromethyl group is on the sulfur of the sulfonyl group.  The expected side product is diethyl phosphate.  We are currently searching for TLC conditions to separate the aldehyde from the Horner product.  We are also considering refluxing the reaction mixture.
EDT
Our third trial on TLC solvents is giving us a little bit of separation, and the aldehyde has a slightly lower R_f vs. the product, which is the opposite of what happens when a methyl group is present.  It is a struggle to get much separation between the two, however.

[rolnor]

I guess this sulphonyl group will be very electron withdrawing then?
Heat is maybe what you need as you suggest.
You could try alumina TLC plates just to change something drastic, it can give better separation.

[Babcock_Hall]

Yes, the organic literature suggests that the sulfonyl group will be more electron-withdrawing.  I should have thought of alumina, but we have not tried that yet.
EDT
So far the only solvent that provided good separation was 50:50 toluene/hexanes.  That would be an unusual choice for a silica column.

[rolnor]

Yes, its a bit unusual and very lipophilic.

[Babcock_Hall]

Also toluene is a little bit higher in boiling point than most other solvents.

[Guitarmaniac86]

Yes, the organic literature suggests that the sulfonyl group will be more electron-withdrawing.  I should have thought of alumina, but we have not tried that yet.
EDT
So far the only solvent that provided good separation was 50:50 toluene/hexanes.  That would be an unusual choice for a silica column.

I've run toluene heptane columns, sometimes you gotta do what you gotta do. Toluene can give great separation especially where you have multiple aromatic species present. For example, I had a benzothiophenepyrimidine structure (thiophene with benzene ring on one side and pyrimidine on the other side) and I used Pd2DBA3 as my catalyst. I could only separate the DBA from my product using toluene heptane. Every other solvent system the DBA co-eluted with my product.

Sorry that this is an aside to your main question. I have no answer on that front.

[Babcock_Hall]

On the contrary, your answer is very helpful to the problem that we may have to face, which is to separate the product from the unreacted starting material.  We are letting the reaction run a bit further, but I doubt that it will go to 100% completion.  Was there some reason that you chose heptane over hexanes?

I was a little surprised that the change from methyl to trifluoromethyl has the effect that it does on R_f values, but chemistry is full of surprises.

[Guitarmaniac86]

On the contrary, your answer is very helpful to the problem that we may have to face, which is to separate the product from the unreacted starting material.  We are letting the reaction run a bit further, but I doubt that it will go to 100% completion.  Was there some reason that you chose heptane over hexanes?

I was a little surprised that the change from methyl to trifluoromethyl has the effect that it does on R_f values, but chemistry is full of surprises.

Company policy. Hexane damages fertility and heptane doesn't.

Chemistry being full of surprises just keeps us on our toes and keeps in interesting!

[rolnor]

On the contrary, your answer is very helpful to the problem that we may have to face, which is to separate the product from the unreacted starting material.  We are letting the reaction run a bit further, but I doubt that it will go to 100% completion.  Was there some reason that you chose heptane over hexanes?

I was a little surprised that the change from methyl to trifluoromethyl has the effect that it does on R_f values, but chemistry is full of surprises.

Its possible that the trifluoromethyl reduces electron density on the sulphonyl-oxygen atoms, making them poorer acceptors for hydrogen bonding, thereby reducing polarity and increase Rf on TLC.

[Babcock_Hall]

The column using toluene and hexanes provided us with pure product.

[Guitarmaniac86]

The column using toluene and hexanes provided us with pure product.

Awesome!