Modification: Br/H2O2 oxidation THF -> GBL

[Beeswax]

I imagined a scale-up of the Br/H2O2 oxidation of THF away from the microscale quantities reported in the literature.  In answer to a previous question, the DCM is rather necessary since doing this procedure in water alone creates a run-away exothermic reaction after the 2nd addition of H2O2, and who knows what the hell is created.  Don't try this in water alone, I have a rash on my forearm from a fine mist of HBr while trying to immerse the flask in an icebath to cool it down.  Add this one to the stupid lab tricks thread.    

    Bicarb was considered as a wash to remove bromine then rejected because of its low solubility and thus large volume necessary to get the red out.  GBL is miscible in water, so any wash will theoretically remove product from the organic phase.  Reducing the wash volumes should increase yield, yah?  No?   Read on:

     300 ml DCM, 300 ml H2O, 100 ml THF and 20 ml Br2 were added to a 2-liter Erlenmeyer flask.  Setting up for reflux (only 37oC btw), the high walls of the flask and breeze from the hood served as a fine reflux mechanism.  Parafilm over the mouth of the flask finished the set-up.  After one hour of gentle stirring to keep the contents in motion (is this necessary in a two-phase system?) 50 ml 35% H2O2 was added.  After another hour an additional 50 ml H2O2 was added.  Reflux continued for 2 more hours, the flask cooled and contents separated in a 1000 ml funnel (whew, I'll do this one in the hood next time!).  The aqueous phase was discarded and the organic phase washed with 50 mls of a saturated solution of Sodium Sulfite which was discarded (on the bottom!).  The clear organic phase was washed again with 50 ml distilled H2O to remove residual salts, dried with silica gel and the DCM evaporated to yield 15 ml lemon yellow GBL identical in bioassay to an authentic sample.

     So, yield sucks.  What should be done to improve this protocol?  Shoulda saved the washes and back extracted, I know, maybe next time.  Would the larger volume have required a longer reaction time, maybe overnight?  More stirring or less?  Maybe some kind bee could look at the protocol, check my numbers and offer suggestions.  Thanks all!

[Rhodium]

Maybe extract with 5-10% NaCl - it will dissolve inorganics, but not as much GBL - and definitely perform those back-extractions...

[Chromic]

Your #s look right. 1.2mol THF (100ml), .48mol Br2 with net of .96 mol H2O2. You've reduced the water/DCM by half of what the ref recommends, which is good... try going even further down.

To increase the yield I would highly recommend reducing the amount of water you use to a minimum. (only use enough DCM to match the volume) Also extract 3-5x with DCM from the aqueous phase, instead of just once as you're doing. I've written about this in other threads.

Oh yeah... do NOT wash the DCM layer with 50ml of water as you're doing. Do NOT dry with silica gel (don't you know that silica gel is quite polar? whoops! infact, do not dry it at all! distilling the DCM removes the water, trust me... reject what you learned in undergrad here) The GBL, if it's yellow, is not pure. You need to distill it (or at least react with NaHCO3 or NaOH and treat with activated charcoal).

GREAT work btw! Keep it up.

[dioulasso]

Hi,

Experimentes were recently conducted to improve the yield described in

Post 474939

(dioulasso: "THF to GBL w/  2-phase Br2 /H2O2", Methods Discourse)

Into a 2l RB flask was charged w/ 300ml H2O, 300ml DCM, and 0,25mol of in situ generated Br2. 50ml THF (0,6mol) was added and the mixture was heated to reflux (bath temp. ~65 C) w/ stirring. The usual ox. procedure was conducted: 1h reflux, 35ml 30% H2O2, 1h reflux, 35ml 30% H2O2, 2h reflux.

Workup: Organic phase was separated, aq. phase extracted w/ 2x 150ml DCM, combined DCM layers throughly washed w/ 50ml 15% NaHSO3 and 50ml H2O, dryed, treated w/ some FeSO4, and evaporated. The crude dark residue was distilled to yield ~20ml substantially pure GBL.
yield: ~45% Better but should be improved!


Remarks:

The major differences compared to Beeswaxes run:

1.   Double the amount of solvents were used (according to the paper).
2.   Was conducted under reflux.
3.   Stirring was less efficent    (in an RB flask mixing of the 2 layers was much worse than in an E flask). The reflux compensates somewhat though.
4.   The aq. phase was extensively extracted w/ DCM (necesseary IMO).

-----

Also the amount of Br2 slightly waries:

% (mol) of Br2 compared to THF:

paper    40%
Beeswax    33%
dioulasso    42%

It is not clear to me wether this has a significant inpact on the yield. Not likely…

-----

IMHO the yield may be further improved by:
1.   Further reducing the loss to the aq. phases (more DCM ext. + Rhodiums suggestion).
2.   Enhancing the mixing of the 2 phases during reflux.

------

It would also be inportant to find out if the amount of solvents can be reduced upon scaleup.
Beeswax used ½ the amount of solvents, but it is not clear wether this inpacted the yield or not. (We should orchestrate our experimentes!   )

[Beeswax]

Thanks for all the advice, wiser bees.  I entered the laboratory armed with new information and put it to the test:

     250 ml DCM, 150 ml H2O, 100 ml THF and 20 ml Br2 were vigorously stirred for 30 minutes at reflux, 50 mls H2O2 added, stirred 1 hour, 50 ml more H2O2, stirred for two hours.  The mildly exothermic reaction kept the temperature at 40oC (reflux temperature for this reaction!).  The temperature dropped at the end, the cooled mixture was separated and the aqueous phase extracted with 2x50 ml DCM.  The combined organic extracts were washed with 30 mls saturated sulfite solution and the DCM evaporated.  Yield, 15 mls rather salty GBL, SAME AS BEFORE!!!  Son of a bitch.

     The reserved aqueous phase and washes were analyzed for the presence of GBL by the crude technique of rubbing some of it on the black benchtop, quickly evaporating the solvent with a hairdryer and sniffing to see if there was any oily residue of GBL.  Nothing.  Nothing at all.   Apparently all the GBL that was formed was extracted in the DCM!

     So what the hell happened to it all?  I have a couple of theories, the foremost of which is that the THF, the precursor was distilled off.  Somehow my makeshift reflux system allowed the THF to escape unbeknownst to me.  The only problem with this is that the THF boils at 66oC and the DCM boils at 39oC so I would expect the DCM to disappear first which it doesn't, there is plenty left in the reaction.  Perhaps the DCM is protected by the overlayer of water?  Which layer is the THF in, aqueous or organic?  (btw I didn't detect any THF in any of the fractions by nose, I assume it's either used up, evaporated or my sense of smell is really bad).   Thus only a portion of the THF reacted before it all disappeared and I was left with very little product.  But the seal seemed intact as far as I could see.  And why didn't the DCM distill off first?  Again, which layer is the THF in?  If it were the aqueous layer, it's possible the slow reflux allowed it to disperse.

     The second possibility is that only a portion of the THF actually reacted and the rest managed to escape when I evaporated the organic solvents.  If this were true it would suggest that I couldn't detect the THF over the smell of bromine and DCM. 

     There may be as yet other, more plausible reasons, I don't know.  The big questions are,

1) Where is the GBL (I'm positive I got it all out)
2) Where is the THF (I'm sure it was no longer present in the reaction) 
3) What the hell happened?

Is it entirely coincidental that I got similar low yeilds with the bromate oxidation?   Btw, the reagents are high grade, even if the chemist is a little low brow.

     I'm annoyed as all get out.   Your input, gentlebees, is, as always, welcome. 

[Chromic]

Keep the DCM:water ratio 1:1... so use less DCM at first (that would be 150ml). But also keep your extractions equal in size, so up that to 2x150ml instead of 2x50ml. Before you extract with the DCM, saturate it with sodium chloride, perhaps acidify it with sulfuric acid as well. Btw, you can't tell if the water contains GBL by evaporating it. The GBL will evaporate with the water (unless it's rich in GBL). Btw, can you get away with less than 150ml of water/DCM? What's the minimum?

I think your yields are low because of the workup, which is about the same in the bromate and bromine case. The methods are similar, but the Br2/H2O2 method is definitely the one to pursue research with.

I can't believe you're just getting 15%. I could understand if you were just getting a yield like 50% (that would be reasonable, IMHO)

I believe the THF stays in both the aqueous and the dcm layers... just like GBL does.