[Babcock_Hall]

I am having problems with a reaction that I have run many times previously.  It is similar to a transformation that I posted about last March.  Starting with (tert-butylO)₂P(O)H, I first added LiN(TMS), then added chlorotrimethylsilane, then I took a P-31 NMR and observed a major product near 132 ppm, suggesting the formation of (tert-butylO)₂POTMS as the major product, with a small amount of starting material.  Then I added 2 equivalents of para-formaldehyde, and stirred for three hours, then overnight at 47 °C.   Usually the reaction is done in three hours.  However, even overnight, all I am seeing is apparent partial reversion of (tert-butylO)₂POTMS back to the starting material (the H-phosphonate), plus two small peaks between 10-20 ppm, one of which may be the desired compound.  I would estimate that slightly more than half of the phosphite remains.

One difference is that I used new para-formaldehyde (PFA), and it was more granular, as opposed to powdery.  Possibly the sluggishness of the reaction is a result of the particles of PFA being too large.  I ran a related reaction at Dan's suggesting several months ago with the same bottle of PFA, and it was slower than expected, based on the literature.  BTW, I realize that the method I used is longer and clunkier compared to the paper Dan suggested, but having run this reaction successfully in the past, I thought it was best to try it again.  I was also concerned about the difference in R groups in the ester:  today I am working with tert-butylesters versus benzyl esters several months ago.

What are some of the things that might have gone wrong?  I was thinking of tossing in some powdered PFA now, and seeing if I can convert the rest of my material to the desired product.  That way I might get some useful product, at least.  However, I also need to think about next time.  Link to related thread:  http://www.chemicalforums.com/index.php?topic=66416.0

[discodermolide]

Could it be that the new PFA is wet? May be worth measuring the water content.
Certainly using the powder is better than granules. Grind some up and chuck them in.

[Babcock_Hall]

I dried in vacuo in the presence of P₂O₅ overnight before using it.  Lost about 0.1 gram of mass (IIRC) out of 6 grams.  How would one check the water content?

[discodermolide]

Karl-Fischer titration.

[Babcock_Hall]

I had a leaky hose overnight that put a good deal of water into the hood.  I have no reason to think it got into my reaction, but obviously not desirable, either.  I just added some PFA that was more powdery and that I had dried for a few hours in vacuo.  I am now going to monitor by P-31.

[discodermolide]

Good luck

[Babcock_Hall]

After three hours, the peak at 132 ppm disappeared, and the major peak is a triplet at 15.8 ppm.  That matches a previous synthesis, but the coupling constant to hydrogen is much smaller as the one that was reported to me via personal communication by a very good chemist (there always has to be one fly in the ointment).  At this point I plan to remove the silyl group and recharacterize the alcohol product.

[discodermolide]

Well that's good news.
We measured the coupling constants in our series as,  3.98ppm (AB qd, J = 18 and 5 Hz, 2H, PCH₂OH), but we had a phosphinate  not a phosphonate. In a substituted phosphinic acid we saw 3.95ppm (d, J = 6 Hz, 2H, CH₂OH. The former in CDCl₃, the latter in D₂O.

[Babcock_Hall]

I filtered away the unreacted para-formaldehyde twice, rinsing the filter cake with THF (THF was also the solvent in the reaction).  I then removed solvent and gently capped the flask.  As I was preparing to weigh the contents of the flask, the flask grew quite hot, and it evolved a gas.  The odor of the gas reminded me of an odor I have noticed when tert-butylesters of phosphorus decompose slowly with time.  The material solidified and turned orange.  My tentative interpretation is that something catalyzed the loss of the tert-butyl groups (I have not yet attempted to characterize the orange solid).  The only thing I can think of that I did differently is that I don't recall having rinsed with THF before.  This THF was old, and I did not check for peroxides before I used it.  However, I don't know whether or not tert-butylesters are labile to oxidants (they are quite labile to acid).

[discodermolide]

Some of our phosphonites did that!
The seem to undergo some sort of redox process? This generates phosphines and a yellow-orange-red solid.
I once tried to isolate this material but it was pyrophoric.
During the distillation of methyldichlorophosphine for some reason I didn't get a decent vacuum and increased the temperature, suddenly the lab began to smell of phosphines. The distillation apparatus was full os a pyrophoric gas, I noticed this when I tried to remove the temperature. Left over was this solid, which was also pyrophoric.
There is a series of books by Kosalopoff and Maier, quite old now, where this type of solid is described, I can't remember the exact volume.
Perhaps if you can get your hands on the series it may help. You may also like to look at the Houben-Wehyl series of books
Just watch out for pyrophoric material.

[Babcock_Hall]

What's odd is that I have been doing this reaction for many years, and this is the first time I have seen the reaction behave in this manner.  Of course, this reaction spent much more time at 45 or so degrees than it usually does, owing to the problem with the para-formaldehyde. 

[discodermolide]

That was also my observation in our case. We dis the preps. lots of times with no problems, but then one time it would mis-behave as I described. Not only with methyldichlorophosphine but with other phosphinites. No pattern or reason, just this funny behaviour with a orange-red solid precipitate which was pyrophoric.
Redox reactions, traces of radicals, I don't know why this happens. But when it does the only course is to incinerate the whole thing. Which I did a few times with the help of the company fire brigade.

[Babcock_Hall]

I will be making some fresh H-phosphonate, di-tert-butylester on Monday, and I'll try the synthesis again later in the week.  BTW the two ²J_HP coupling constants that were reported to me were 17 Hz for the P-CH₂-OSiMe₃ compound, and 7 Hz for the corresponding alcohol after removal of the TMS group.