[synthesissynthia]

Okay so I am trying to create this para-substituted trifluoromethyl phenylacetylene, and so far my attempts have not bee successful. My first attempt was following a procedure (the name of the reaction escapes me now) but my starting material was trifluoromethylbenzaldehyde and I used LDA and TMS-diazomethane. However upon isolation, the NMR showed that I had reduced my aldehyde instead of turning it into the alkyne.

So I decided to do a Seyferth-Gilbert Homologation, using Bestmann-Ohira reagent. The reagent was clean and it worked in converting the 4-bromobenzaldehyde and 4-nitrobenzaldehyde to the cooresponding alkynes, however it didn't work for the trifluoromethylbenzaldehyde. TLC showed a new spot, and absence of the aldhyde spot so I quenched and boom I have nothing in my NMR! I'm thinking I didn't let it go to completion? But I left it for over 24 hrs and nitro group is more electron withdrawing and that worked overnight.

After doing the homologation 3 times and failing, I decided to to a Sonogashira coupling trifluoromethyliodobenzene with TMS-acetylene. A bullet proof reaction that also didn't work. Upon addition of  last few microliters of my TMS-acetlyene at the last step, the reaction mixture started to form a percipitate and then turned from yellow to black. I still let it mix over night for about 16 hrs and the TLC looked promising however I recovered my TMS-acetylene with my TEA! No trace of my iodobenzene anywhere, but it was there via TLC. I'm concluding that the iodobenzene is inthe oxidative insertion phase with the Pd and is "stuck" there. That is the only logical explanation for it otherwise I would've recovered it when I was extracting. But how do I get it to be "unstuck"?

[opsomath]

I haven't done that homologation so I can't comment on it from experience, but on the Seyferth-Gilbert if the TLC went as you said you must have lost it in workup. Fluorous materials are often weird in their solubility, what was the exact procedure including extractions that you used for workup?

I would actually say that you may be similarly "lost" on the Snog coupling. I've had a lot of 'em turn black on me, and although you should always degas everything thoroughly before starting it up it doesn't necessarily mean the thing is doomed. The fact that you lost your starting material suggests once again a workup problem. You couldn't possibly be using enough Pd catalyst for it to all be stuck in the Ar-Pd oxidative addition form, so I think you're overthinking that a bit.

[OC pro]

Regarding Sonogashira: they always worked well for me. 3mol% Cu(I)-iodide, 5mol% Pd(PPh3)2Cl2 in DMF with 1.0equiv. of the aryl iodide. 1.2 equiv. of TEA added (everything at rt, the mixture was degassed thoroughly) and then 1.1equiv. acetylene was added. 2h at rt did the job in most cases. Prepared >20 examples with yields from 80 to 95%.

[synthesissynthia]

Regarding Sonogashira: they always worked well for me. 3mol% Cu(I)-iodide, 5mol% Pd(PPh3)2Cl2 in DMF with 1.0equiv. of the aryl iodide. 1.2 equiv. of TEA added (everything at rt, the mixture was degassed thoroughly) and then 1.1equiv. acetylene was added. 2h at rt did the job in most cases. Prepared >20 examples with yields from 80 to 95%.

The Sonogashira coupling is what is giving me the problem, and I don't know why. Everyone else in my group is able to do it, and literature has a 99% yield for the substrate I'm trying to make but I get nothing! The literature I used the first time was JACS 2013 135 1264 from the supporting info. In the glove box I added my Pd(PPh3)2Cl2 (2 mol%) and my CuI (4 mol%), made sure no air got in and then added my iodobenzene derivative (1 eq) and in 2 ml of dried THF. Then I added my TEA (2 eq). I let that mix for a minute and then I add my TMS acetylene (1.5eq) dropwise over 3 minutes. Then added the rest of THF (0.4 ml) and let it stir under N2 over night. I use KMnO4 stain so when I saw something stain at a high Rf with hexane as my solvent front for TLC, I filtered the reaction over celite, and then removed the THF via rotavap. Then I extracted what was left with ethyl ether and water/brime 3x and dried with MgSO4. When I take and NMR of it looks like I have my amine and my TMS acetylene. No aromatic peaks revealing where my other starting material is.

I did the reaction again using a difference catalyst and a different procedure that showed the reaction yielding a 99% yield for my substrate. This is from Tetrahedron Letters 43 (2002) 7091. In my schlank tube I added Pd(PPh3)4(2 mol%), CuI (4 mol%) and TEA(1.5 eq) in THF (5ml) all under nitrogen. Then I let that stir for about a minute, added the iodobenzene derivative and then the reaction mitxure was cooled to 0 degrees. Let that stir for 10 minutes then added the TMS-acetylene (1.05 eq) dropwise over 30 minutes. The let it stir at room temp over night. Filtered with Celite and rotovapped the filtrate and just took an NMR of that because I thought my extraction methods were causing me to lose my stuff. However, I check the NMR and again I have the same result as yesterday - no aromatic peaks! I honestly want to pull my hair out, but I won't because I love my hair.

I talked to my labmates about it and they have no clue what is going on. So I decided to just do a test run with iodobenzene with Pd(PPh3)2Cl2 as my catalyst (since we are running out of the Pd(PPh3)4) and see how a test run goes. If it makes the alkyne then I have no idea what is going on with this substrate. If it doesn't than it is just me and I will have to just have my lab mates babysit me as I do this reaction because this is seriously ridiculous!

But whatever - chemistry, oh chemistry.

[BobfromNC]

Can you try another acetylene derivative to test this?  Perhaps TMSacetylene is too hindered or unreactive.   Get someone who tells you they can get 99% yield to run it to see if they are really that good.   I have had lots of people tell me they got 99% yield on this, but I rarely get them that good.   Same with Susuki reaction, usually works, but rarely in the yields reported, maybe due to older reagents and lousy catalyst.

[clarkstill]

Why not try the Corey-Fuchs?

http://www.sciencedirect.com/science/article/pii/S0040403901941577

It's always worked pretty reliably for me...