DirkDinkel - 30-6-2014 at 13:55
I just stumbled over an interesting summary of different uses of glycerol as a green chemical:
Glycerol: A promising Green Solvent and Reducing Agent for Metal-Catalyzed Transfer Hydrogenation Reactions and Nanoparticles Formation
Personally, I am very interested in the procedure of the reduction of carboxylic acids to alcohols with glycerol, KOH and CoCl2
6H2O since it sounds very cheap in comparison to LAH. I would love to try this reaction on different molecules to find out which functional
groups are tolerated.
Has anyone done this reaction before? What where the substrates used?
And has anyone an English version of the Korean Patent KR 2012006276?
I'm a little bit "afraid" of the workup, I guess it's going to be very messy due to the high viscosity of glycerol and dihydroxyacetone. 
forgottenpassword - 30-6-2014 at 14:09
Very interesting. The paper that you linked to lists the substrates that were used. If you can provide the patent I'm sure that the experimental will
be easy enough to decipher.
DirkDinkel - 30-6-2014 at 15:13
I'm aware of the substrates used and it seems the method tolerates alkanes and benzene (which is trivial to me). After trying to reproduce the results
I'm going to find out if other functional groups are tolerated. I'm going to try the method with some Aldehyde, Ketone and Amino group containing
compounds. I'm especially interested in alpha functional groups. Well, how many experiments I may perform depends on how easy the work up and how much
money I can spend for my research.
mnick12 - 30-6-2014 at 23:21
It is a cool reaction if it actually works, they claim reduction of benzoic acid to benzyl alcohol, which Im not sure I believe. Anyone know if they
propose a mechanism for the transfer hydrogenation?- I didn't see one.
forgottenpassword - 1-7-2014 at 01:30
You could try reducing beta-alanine to ethanolamine. Both are easily available and you could check the success of the reaction using comparative TLC
with an authentic sample.
Glycine, rather. Possibly amines would interfere with the catalytic action of cobalt, but it would be easy enough to try it out and see.
[Edited on 1-7-2014 by forgottenpassword]
solo - 1-7-2014 at 02:08
Reference Information
Transfer hydrogenations of benzaldehyde using glycerol as solvent and hydrogen source
Dorith tavor, Oxana Sheviev
Can. J.Chem.
2010, vol 88, pg 305
Attachment: Transfer hydrogenations of benzaldehyde using glycerol as solvent and hydrogen source.pdf (84kB)
This file has been downloaded 716 times
DirkDinkel - 1-7-2014 at 02:28
@solo: Nice! Thank you!
So, in the paper solo posted, they just extracted the product with diethylether and also tried to recycle the catalyst. I guess I'll try this too.
@forgottenpassword:
Beta-alanine is on my list! Beta-alanine is also a good precursor to halocarboxylic acids via diazotation. And 1-Amino-3-propanol is a good precursor
for the reduction products. Sounds easy and cheap.
[Edited on 1-7-2014 by DirkDinkel]
I suspect that the use of amino acids will result in imine formation between dihydroxyacetone and the amine due to the basic conditions, which could
end in reductive alkylation of the amino group, if imines are also hydrogenated under these conditions. So perhaps it's better to try an reductive
amination before. With two educts which can be easily identified by TLC.
[Edited on 1-7-2014 by DirkDinkel]
zed - 2-7-2014 at 12:24
Yeah, but it's shooting fish in a barrel. At higher temperatures, excess isopropyl alcohol will reduce benzaldehyde to benzyl-alcohol in ~99%
yield. It is a variation of the Meerwin Pondorff Verley reduction.
Now, use that glycerol to reduce something that isn't already a slam-dunk, and I'll be impressed.
Whoops! I stand corrected. I just finished reading the first article. I'm impressed. Reducing benzoic acids to alcohols, isn't usually easy.
Likewise, the convenient reduction of nitro groups to amino, would be handy.
[Edited on 2-7-2014 by zed]
mnick12 - 3-7-2014 at 16:07
Thanks for the paper solo.
Anyway I still don't believe they reduced benzoic acid to benzyl alcohol with a 90-95% yield.
Crowfjord - 3-7-2014 at 16:16
I'm skeptical as well, but some experimentation will show the truth. I might be able to give it a shot in a month or two, once I get other projects
and personal crap out of the way.
Burner - 3-7-2014 at 19:25
I healthy skepticism is a good thing in science! 
Having used glycerol a number of times in my career I must say that it is not my favorite material when it comes to work-ups. It has a high viscosity
(making removal of solids extremely difficult), it has a very high boiling point (forcing you to do very high vacuum distillations), and it loves to
form esters (making its use with organic acids problematical).
Having said that, if this works this is a great tool for the amateur chemist's arsenal since reagents such as LAH are nearly impossible to acquire (at
least in the US).
PHILOU Zrealone - 4-7-2014 at 03:25
Also under certain circumstances dihydroxyaceton (artifical sun-tanner) polycrotonises to dark polyphenolic martials...
It is also in equilibrium with an aldehyd moeity:
HOCH2-C(=O)-CH2OH <==> HOCH2-C(-OH)=CH-OH <==> HOCH2-CHOH-CH=O
Pumukli - 24-7-2014 at 10:57
Hello,
I found the korean patent (the cobalt-chloride - glycerol reduction) but unfortunately it is written in, ummm, korean. :-)
Fortunately though there is a table of summary of the reactions claimed in the patent and it describes the ammounts of reagents , the product(s) and
yield (probably).
It is suspicious, because it gives the ammount of starting benzoic acid in millilitres as if it was a liquid too. So they may not even run this
reaction just lumped it in to relate to an aromatic acid too. The other (aliphatic) acids seems to be OK because they are liquids at room temp so
their ammounts in millilitres is not suspicious at least.
Anyway, if someone wants to try the reactions, here are the ammounts what I could decipher from the patent:
Benzoic acid: 2 ml
CoCl2.6H2O: 3 g
KOH:0.22 g
Glycerol: 25 ml
Temperature: 140-150 Celsius
Yield: 95% benzyl alcohol
Butyric acid: 3ml
Cobalt: 6g
KOH: 0.44g
Glycerol: 50ml
T: 140-150 Celsius
Yield: 90% n-butanol
Valeric acid: 3ml
Cobalt: 6.2 g
KOH 0.45 g
Glycerol: 50ml
T: 140-150 Celsius
Yield: 91% n-pentanol
n-C6-acid: 3.5 ml
others as with valeric acid
Yield: 94% n-hexanol
n-C7-acid: 3.8 ml
others as with valeric acid (but KOH is just 0.42 g)
Yield: 92% n-heptanol
n-C8-acid: 4.2 ml
others as with valeric acid (but cobalt is 6.25 g)
Yield: 90% n-octanol
n-C9-acid: 4.6 ml
others as with valeric acid
Yield: 93% n-nonanol
(Cobalt means the Co-chloride-hydrate but it was easier to type just that.)
Unfortunately the reaction time is not given so it may be several days. :-)
A quick smell-test could be used at least with benzoic acid to tell if there is any benzyl alcohol in the flask.
Pumukli - 25-7-2014 at 03:31
Update:
I could machine-translate the patent, it says things like this:
The reaction condition was optimized in 140 through 150℃ (the Entry 1, and the table 1) by using as the benzoic acid 2mL, the CoCl 2 6 H 2 0 3.0g,
and the rain of the glycerol 25mL and KOH 0.22g. The reaction progress was observed with TLC. It was observed that the conversion to the benzyl
alcohol of the benzoic acid was successfully finished within 1 hour. But the reaction time was more lengthened with 4 hours. In RT, the reaction
mixture was made a quick cooling with the brine liquid. The reaction mixture was extracted with the ethyl acetate (EtOAc). The generated EtOAc
solution was extracted in the saturation K 2 CO 3 solution with addition. Subsequently it was dry with the anhydrous Na 2 SO 4 . It concentrated and
the solvent was removed. It confirmed with the spectrum analysis with addition that the product was the benzyl alcohol. It showed that the wide
(strong and broad) bond standing against 003e#3300 cm - 1 area the carbonyl binding of the carboxylic group is nonexistent according to IR data of the
product in 1710 cm - 1 existed and -OH existed in the product (figure 1reference).
And another one:
The electricity stir bar, the reflux condenser, and part set up the funnel and thermometer in 2 neck round bottom flask dried in the oven. The butyric
acid 3mL, the CoCl 2 6 H 2 0 6g, and the KOH 0.44g and glycerol 50mL were injected in the flask in RT. In 140 through 150℃ the reaction mixture, it
agitated for 1 hour. The reaction mixture was extracted using the ethyl acetate 50mL. The generated EtOAc solution was extracted with the saturation K
2 CO 3 solution with addition. Subsequently it dried with the anhydrous Na 2 SO 4 . It concentrated and the solvent was removed. It confirmed with the
spectrum analysis with addition that the product was the butanol.
It showed that the wide bond standing against 003e#3300cm-1 area the carbonyl binding of the carboxylic group is nonexistent according to IR data of
the product in 1710cm-1 existed and -OH existed in the product (figure 2reference).
Moreover, according to fig. 2, 2972-2885 cm - 1 , and the C-O stretch of about 3440 cm - 1 , and the aliphatic chain C-H stretch of the O-H stretch
(broad) hydrogen-bonded between the molecule can know that 1049 cm - 1 , and the C-O-H bending show up in 1465 - 1261 cm - 1 .
forgottenpassword - 25-7-2014 at 08:56
Very helpful, thanks.