Author Topic: 1,4-butanediol dehydrogenation with Cu(II)-oxide  (Read 4198 times)

0 Members and 1 Guest are viewing this topic.


  • Guest
1,4-butanediol dehydrogenation with Cu(II)-oxide
« on: September 10, 2004, 06:15:00 PM »

Recently SWIA wondered about why the method outlined in

  (and translated by GC_MS,

Post 441898 (missing)

(GC_MS: "1,4-Butanediol to gamma-Butyrolactone in English", Methods Discourse)
) wasn't tried on 1,4-BDO here at the hive - and decided to give the dehydrogenation a try.

SWIA first measured out 87ml 1,4-butanediol (1 mol) and placed it in a clean, dry 500ml 2-necked BRF together with a mag.stirbar and 0.79g black copper(II)-oxide (CuO; 0.01 molar eq.). A thermometer and dean-stark trap were attached and the trap fitted with a big vigreux column as air-cooled conderser, then everything was heated to >200°C with good stirring.

As soon as 180°C rxn temp. were reached, gas evolution started (bubbles evolved). After ~0.75 hours of stirring/heating, the black CuO suspension had cleared up almost completely, and a tiny copper "mirror" could be seen at the side walls (elemental copper precipitated). At this point, the internal temperature was noted to be 204°C.

During the next hours, some few ml's of distillate were collected in the trap, they were dried over some 4A sieves and stored to be added to the workup later.

After 12 hours of reflux, the mixture was cooled down, then the dean-stark was replaced with a distillation setup and the reaction mixture was distilled at atmospherical pressure, collecting distillate until the temp. neared 230°C, then dist. was stopped.

The crude product and unreacted diol were then diluted with d.H2O (1/10th of the total volume), heated to reflux in other clean RBF and a solution of 32g NaOH in 80ml d.H2O was added dropwise until pH 8 was reached. After addition was complete, reflux was maintained for 5 minutes, then everything was poured into a sep. funnel, cooled down to room temp. and the upper BDO layer was separated and further extracted with Et2O. The remaining NaGHB/H2O solution was then concentrated on a hot plate (160°C) until no more water boiled off, then everything was cooled very slowly to -7°C and the precipitated GHB*Na collected, dried on hotplate (110°C) and finally over anhydrous 4A sieves in a desiccator.

The yield of dried NaGHB amounted to ~27 grams, corresponding to a molar 35% yield calculated from 1,4-butanediol.

After redissolving the Na salt in H2O and adding 1.01 eq. H2SO4, followed by steam dist. and extracting the distillate with EtOAc and solvent removal, the yield was 17ml (or ~19g) of purified gamma-butyrolactone, which resulted in very clean product when reacted with ACS grade KOH to pH 7.5 (no soap taste whatsoever, slight licorice taste/smell, quite agreeable).

Nice procedure, dare I say. And with longer rxn times, yields could still be improved a bit I think (the original paper recommends 15h rxn time for Bu-1,4-OH)

As the reaction proceeds through removal of formed water: does anybody know if 4A mol. sieves will react with CuO at these high temperatures? If not, adding enuff sieves to absorb all formed water would increase yield (or at least rxn speed). But I'm afraid that the catalyst may react with the sieves and become unreactive... (?)

Greetz, A


  • Guest
No need for molecular sieves
« Reply #1 on: September 10, 2004, 08:14:00 PM »
C4H10O2 (1,4-BDO) --> C4H6O2 (GBL) + 2H2

No water evolved, only hydrogen. That's why it is a dehydrogenation and not a dehydration.


  • Guest
strange - the reference tells H2O is evolved
« Reply #2 on: September 10, 2004, 08:53:00 PM »
there are polyester sideproducts formed, I think this is where the water originates from.
I am quite sure that there was water present in the trap, and the original article too says that the trap is used to remove the minute amount of water formed at the beginning of the reaction...

Maybe it would be favourable to not remove water to supress ester formation (whatever they meant specifically)?
Or maybe this H2O removal drives the reaction forward and is therefore mandatory, dunno.

Experiment will tell - as soon as I know whether sieves and catalyst would interact in a not desired way or not (I don't want to waste time, so it would be nice to know about it before starting further experiments)  ;) ...

Or maybe the water simply originates from the CuO becoming reduced? Elemental copper was observed, so there's oxygen somewhere - and as you said, there's the dehydrogenation, giving hydrogen...



  • Guest
« Reply #3 on: September 11, 2004, 12:12:00 AM »
The question of the copper precipitation is indeed a good one. CuO getting reduced to copper and water while something else gets oxidised would at least be a simple explanation. Does the water evolution happen at the same time the metallic copper is formed?

Edit: Another possibility is the hydrogenolysis of an alcohol by another alcohol giving rise to a hydrocarbon, a carbonyl compound and water the article discusses, but they noticed it only with benzylic alcohols, not aliphatic so I think it is unlikely. If it happened, either butanal (by dehydrogenation of 1-butanol, resulting from hydrogenolysis of 1,4-BDO) or tetrahydrofuran (from 2-hydroxytetrahydrofuran, the cyclic hemiacetal of 4-hydroxybutanal) could be formed, and in theory butanal could be the starting point for a polymer chain CH3-CH2-CH2-CO-(O-(CH2)3-CO-)n or then it might just evaporate and leave.

I can't come up with a single mechanism that would lead to polyester formation producing water, such esterification is a reaction between a free acid and an alcohol.

By the way, where are the hydrogen bubbles evolving? They don't happen to be coming from the copper mirror?

Of course, it's hard to say what could be going on at such high temperatures.


  • Guest
« Reply #4 on: September 11, 2004, 04:19:00 AM »
I think it's plainly hydrogen reducing the highly dispersed and thus, higly active, CuO (remember, copper stands to the right of hydrogen in the electrochemical metal hierarchy).

This of course gives us a molecule of water.

SWIA has also seen this reduction phenomena when using copper chromite contaminated with CuO. Browning of the catalyst doesn't happen though if it is prepared correctly.


  • Guest
hydrogen and copper..
« Reply #5 on: September 11, 2004, 06:38:00 AM »
Antoncho: yep, IMO CuO+H2->Cu+H2O is the simplest explanation, and the amount of water is indeed very small - according to this explanation, it should theoretically amount to 0.18ml, and this approximately matches with the observations (the amount of H2O was indeed small - so small SWIA wasn't able to determine the amount  ;) )...

Moo: strangely the elemental copper forms 0.5 - 1 hours after starting with heating, but the water is evolved through the whole rxn, at least it looked like it did so... (water looks really different from BDO or GBL when condensing - it doesn't "smear" that much and gives many small droplets not those bigger viscous ones).
(EDIT: well, THF does so, too - maybe SWIA identified THF  as being H2O, don't know.)

And even stranger: the hydrogen evolved from the stirbar (of course), but when stirring was stopped for a while, it became visible that lots of small bubbling evolved from the finely divided Cu having settled to the bottom (not all Cu deposits on the walls) - I would say it is almost impossible to determine where the gas originated from, as bubbles always will develop on "impurities" like boiling stones, solid substances and stirbars if there are any present, but if asked while someone was holding a gun to my head I would maybe say the hydrogen evolved from the copper..

About polyester formation: the alcohol gives off -OH, the acid splits off H+ - and there is water and ester formed...

X-OH + Y-H -> XY + HOH

Or is this "poly" case different from normal esterifications? I mentioned it because the article has it in the table under "side products", not only for BDO but also for other diols - so I made a guess...



  • Guest
more strange observations
« Reply #6 on: September 11, 2004, 07:03:00 AM »
Another rxn was performed, w/same conditions, but at the 4.5 hour point, some additional CuO (~200mg) was added - result: no change, and the additional CuO just didn't react (rxn still in progress though), but the initial amount formed the nice copper mirror as usual - in contrast to the second amount, which still is floating around and pretty black (i.e. no rxn). Hydrogen is still evolved (t=12h), SWIA will let it run three hours longer this time.

Wierd, wierd....

(at least we know that the amount of catalyst isn't really critical, but the addition time/temp. influences its activity somehow  ::) )

More variations SWIA will try in future runs:

1. use mol. sieves

2. don't use stirbar

3. try separating 1,4-BDO from GBL through freezing the distillate, instead of saponifying/extracting it

Any further suggestions for changing parameters? I mean more ways to modify the rxn, so one can try to interpet the different results in a useful way?



  • Guest
« Reply #7 on: September 11, 2004, 10:45:00 AM »
Cu is also a well known dehydogenation catalyst. It is possible that during the reaction CuO actually gets reduced to Cu by the evolving hydrogen, like Antoncho said, producing water and eventually metallic Cu being the species that catalyzes the dehydrogenation. It could be tested by making precipitated copper in similar way used to make dehydrogenation catalysts and using it instead of CuO. I don't remember the precipitation temperature from top of my head though. This would answer the question of what kind of an effect does the Cu formation have on the reaction, it is likely it would not have any other merits as the CuO procedure is good because there is practically no catalyst preparation needed.

It could well be that the water doesn't affect the reaction at all, at least as long as it is removed from the mixture in some way.

In the esterification equation Y would have to be R-COO for the product to be an ester, the equation would become X-OH + R-CO-OH --> R-CO-O-X + HOH, but there is no free acid present unless it is produced by water hydrolysing an ester, which would not give net increase in water... well, I think it is irrelevant regarding the water formation anyway so I'll just shut up for now. ;D


  • Guest
no success :~(
« Reply #8 on: September 11, 2004, 06:00:00 PM »
The run with additional catalyst addition after 4.5h was allowed to react for 15h and let sit for another 7h before SWIA recommenced with distillation (the poor guy suddenly couldn't fight sleep any longer  ;) ), this time he distilled under 150torr vacuum. After ~70ml had been collected, he diluted the distillate with H2O, heated to 110°C and added NaOH soln. - but after having added only approx. 4g NaOH in solution, the pH already had turned to 11  :( . In other words: 0.1mol NaGHB from 1mol BDO. Bad, bad...

I would guess that either the air entering the rxn vessel during second addition of catalyst, or the maybe too high amount of catalyst, or maybe even the 7h waiting before distilling rxn mixture are responsible for this disappointing 10% yield... :P

But another advantage of this rxn besides its ease: unreacted diol can easily be recycled by extracting from the GHB salt solution, stripping the solvent and simply using the remainder in another CuO dehydrogenation.



  • Guest
Re: The yield of dried NaGHB amounted to ~27...
« Reply #9 on: September 12, 2004, 03:03:00 PM »

The yield of dried NaGHB amounted to ~27 grams, corresponding to a molar 35% yield calculated from 1,4-butanediol.

35%...And you reckon this is better then copper chromite (yeild 98%) with the same reaction conditions?


  • Guest
« Reply #10 on: September 12, 2004, 08:50:00 PM »


  • Guest
Please explain how a 35% yeild while using the
« Reply #11 on: September 13, 2004, 11:05:00 AM »
Please explain how a 35% yeild while using the same reaction time and conditions except a different catalyst, is better for any reason other then throwing away 60% of your precursor.


  • Guest
« Reply #12 on: September 13, 2004, 01:21:00 PM »
CuO hasn't to be prepared - and you don't have to throw away the BDO you isolated, you can reuse it. Yes, copper chromite gives higher yields, but is a lot more work.

The advantage of above procedure is its easiness. I case you didn't realize it already.


  • Guest
Most likely the same yields could be reached...
« Reply #13 on: September 13, 2004, 09:39:00 PM »
Most likely the same yields could be reached given more time to react.