Sciencemadness Discussion Board

"Novel Method of Reducing Ketones Using Sodium Hydroxide in Isopropanol"

notoxicshit - 23-2-2017 at 12:01

I'd like to discuss a reaction I stumbled upon. It's from 2010, from an american university.

I attached the paper at the bottom of this post.

My amateurish mind found the claim of "Ketones are readily reduced to secondary alcohols with sodium hydroxide in refluxing isopropanol" pretty unbelievable, because it seemed to lack a reduction reagent at first glance.
Looking into the paper shows that just like in the Meerwein-Ponndorf-Verley reduction, the IPA solvent is oxidised to acetone.
With the authors not caring about putting forth any mechanism, I suspect in situ sodium propoxide formation in sufficient traces to be the cause of the reaction.

A problem are the long reaction times at IPA reflux temperature. Not one of the examples needs much less than a whole day.
They cure this by using a microwave and cut the reaction time down to minutes.

I thought, with this falling in the "too good to be true" category, this needs to be veryfied.
Reading about microwaves and their thermal advantages over normal heating (local superheating) plus their non-thermal magic properties, I've come to the conclusion that the latter don't exist and it's just a hype.

Thus, a modified kitchen microwave is not of much use, especially considering the hassle of controlling the hazard.
They're just fast autoclaves.
Speaking of autoclaves and kitchen, I have a kitchen autoclave a.k.a. pressure cooker.

So I put .25 moles of propiophenone as a substrate and .25 moles of sodium hydroxide into the demanded half litre of 99.9% IPA, which I dried with a handful of mol sieves, just to be sure.

My IR thermometer initially read 35*C, just when the valve on top had closed. I figured, there's something wrong and fixed the notoriously low reading of the device by attaching a thin piece of tape outside just below, where I knew that the level of the fluid inside was.
The reflection of the chrome surface seemed to fuck with my temperature reading.
So now, the piece of tape shows 91*C, when the security valve opens slightly. Turned it down to 87*C on the tape, because that eliminated the IPA smell.

Now, when will my reaction be finished?
Got the pot on setting "2", which is maximum pressure.
The manufacturer says, using water, the first position of the security valve means 110*C and the second one means 116*C.
116*C mean 1,75 bar, as I calculated.
I could see, how hot IPA is at 1,75*C, but then again I'm unsure if that maximum is reached without considerable loss through the sec valve.

So maybe I can guess how hot the 87*C on the tape are on the inside. Any educated guesses?
What rule of thumb should I use to calculate?
Van't Hoff says every 10 K more give 2-4 times the reaction speed.

Acetophenone is one their examples. It needs 22.5 hours to completion. Didn't have that at hand, so propiophenone was used. Will it be faster, slower or alike?

Feel free to discuss anything about this reaction.





Attachment: Novel isopropanol NaOH ketone to alcohols reduction.pdf (94kB)
This file has been downloaded 751 times

[Edited on 23-2-2017 by notoxicshit]

battoussai114 - 23-2-2017 at 12:16

Fudge! This forum has an hability to eat my texts!

I guess it's similar to a Meerwein-Ponndorf-Verley Reduction.

diggafromdover - 23-2-2017 at 13:04

This gives me the notion of turning MEK into sec butanol. I wonder how hard it would be to separate the product from unused reactants.

notoxicshit - 23-2-2017 at 13:42

Would you be so kind to retype your answer? I'm interested in anything people have to say.

I missed out on checking the outer temperature when pressure cooking water at this setting, that might give me a clue, as to how hot the IPA might momentarily be.

Do you people even think, the pressure cooker is a good idea? It's a very high quality one, however.


And yes, I wanted to discuss the workup, too.

That's what they do:
Per 50 mmol of substrate ,they partition between 200 ml of diethyl ether and 100 ml of water. The organic phase is washed three times, with 100 ml water, each, once with brine, NaSO4 drying and rotovaling of the ether and distilling the residue in vacuum.

I'm a green minimalist, so is it possible to just wash out the hydroxide and maybe some of the IPA with salty water washes?
The alcohol I may get isn't water soluble and propiophenone neither. Would spare me the ether or anything alike.

notoxicshit - 23-2-2017 at 13:48

Other points I'd like to hear opinions about : how moisture sensitive should the reaction be?

Theoretically, quite, but practically they use open systems? I'm asking because alkali hydroxides often are far from dry.

Would you expect KOH to work, too, besides being more hygroscopic?

Does someone have experience, in how much less hot an IR thermometer reads a 2 mm stainless steel vessel at those ~100*C?

I have a hard time estimating reaction time.

Corrosive Joeseph - 23-2-2017 at 23:01

Initially it does seem a bit too good to be true.

And how exactly does one 'reflux NaOH in isopropanol for 18.5hrs' in a pressure cooker.................?

Just for comparison..................

REDUCTION WITH ALUMINUM ALKOXIDES
The Meerwein-Ponndorf-Verley Reduction - A.L. Wilds

In-depth review; pages 178-223

http://www.sciencemadness.org/library/books/organic_reaction...


/CJ



[Edited on 24-2-2017 by Corrosive Joeseph]

notoxicshit - 24-2-2017 at 09:05

Well the reflux doesn't serve any purpose other than maximizing the reaction temperature to the BP of the solvent. Thus 82*C.
So I used an enclosed vessel and it apparently became 10 K hotter, what van't Hoff would have read as a speed 2-4 times higher.
I let it steam for 8 hours yesterday. That may already mean a completed reaction, whilst I don't know if Propiophenone needs more or less time than Acetophenone.

I'll work it up and see if the reaction oil produces any water when heated with KHSO4.

And I do know the MPV, I mentioned it, too.

DraconicAcid - 24-2-2017 at 14:35

I suspect that it will only work for high-boiling ketones. RCOR + CH3CH(OH)CH3 = RCH(OH)R + CH3COCH3 isn't going have a significant delta H or delta S, but if you boil off the acetone as it's formed, you'll eventually get near-complete reduction of the RCOR. MEK is going to boil off nearly as fast as the acetone (or more quickly, since you start with more of it), so you can't push the reaction to products so well. (Propio/aceto)phenone seem much better candidates.

stoichiometric_steve - 25-2-2017 at 16:55

Am i the only one doubting this procedure? I just don't see why some alkoxide would act as a reducing agent all by itself. Even so, any acetone produced would undergo aldol condensation with the product ketone.

Sounds like BS.

Corrosive Joeseph - 25-2-2017 at 17:01

Quote: Originally posted by stoichiometric_steve  
Am i the only one doubting this procedure?


No................ You are not


/CJ

Ozone - 25-2-2017 at 19:53

This looks like a transfer hydrogenation. Ish. Typically, the reductant is a secondary alcohol, tetralin, or formic acid (or the like). In my experience, these are conducted using a stoichiometric excess of reductant and a catalyst such as Ru/C or Cu chromite in a stirred Parr reactor (inconel/hastelloy @ 120-180°C). Pd/C can also be used, but was avoided due to expense.

This can work as-is, negating the complications associated with highly alkaline conditions (aldolization, etc.).

https://en.wikipedia.org/wiki/Transfer_hydrogenation

http://nopr.niscair.res.in/bitstream/123456789/16740/1/IJCB%2039B(9)%20709-711.pdf

et al.

O3

CuReUS - 26-2-2017 at 00:16

Quote: Originally posted by stoichiometric_steve  
Am i the only one doubting this procedure? I just don't see why some alkoxide would act as a reducing agent all by itself.Sounds like BS.

this reaction has been known since the 60's,so its more fact than fiction.
Quote:
Even so, any acetone produced would undergo aldol condensation with the product ketone.

No,it wouldn't because
1.the acetone is being distilled off as soon as it forms
2.the aldol reaction between ketones is quite slow
here is a recent paper -http://www.sciencedirect.com/science/article/pii/S1381116915...
I remember seeing a reference on the organic chemistry portal which claimed that carbonyl compounds could be reduced to alcohols using just IPA and pressure.But I have been searching for the past few days and I still can't find that paper:mad::(

clearly_not_atara - 26-2-2017 at 01:31

Acetone is a poor aldol acceptor, but a good donor. If the conditions will deprotonate acetone it will react with electrophilic carbonyl compounds. This limits the reaction's scope, but there are still many compounds on which it may be applicable, phenylacetone e.g. might work okay although benzaldehyde probably wouldn't

stoichiometric_steve - 26-2-2017 at 06:48

Quote: Originally posted by CuReUS  

this reaction has been known since the 60's,so its more fact than fiction.


If you wouldn't mind providing a reference, that would be great. Otherwise, i still call this reaction BS.

Quote: Originally posted by Ozone  
This looks like a transfer hydrogenation. […]
This can work as-is, negating the complications associated with highly alkaline conditions (aldolization, etc.).


This would be a transfer hydrogenation by definition, but since no catalyst is involved, it is IMHO highly unlikely to work at all.

Quote: Originally posted by CuReUS  

No,it wouldn't because
1.the acetone is being distilled off as soon as it forms



The mixture is refluxed, so any formed acetone will be refluxed and limit the mixtures' boiling point. Nothing is distilled off here, which proves my point.
Even so, i don't even believe that any acetone would be formed, and by what mechanism? Suddenly because of a little NaOH, IPA will become some magic hydrogen donor? Sure.

I'm glad Meerwein, Ponndorf and Verley aren't around to witness this simple procedure totally making their method useless. I mean, how could they have been so stupid not to try it?

In the MPV reduction, acetone MUST be distilled off to shift the equilibrium.

Quote: Originally posted by clearly_not_atara  
phenylacetone e.g. might work okay although benzaldehyde probably wouldn't


Of course not, it would just go Cannizzaro.

[Edited on 26-2-2017 by stoichiometric_steve]

[Edited on 26-2-2017 by stoichiometric_steve]

notoxicshit - 26-2-2017 at 09:50

Works, gives a product that can be dehydrated with KHSO4.
Did this because the bps of product and substrate are alike.
Got 7 g of a fraction boiling from 175-180*C, what seems to be the product of the dehydration. That's from 13 g of substrate and sloppy technique doing this the first time, plus I only let it run for 8 hours at approximately 92*C.

Currently set up a bigger batch, but without pressure because the IPA killed my pressure cooker.

Will report back.

All the questions asked are answered in the paper. How about you people read it and try to actually contribute :)

What I wonder is if the yield could be raised by removing acetone. Then, however, there are 4 l of IPA per mole substrate, so that might be negligible.
Another thing might be that acetone may lower the reflux temperature and thus greatly lengthen reaction temperature.

Help me optimize this procedure. Unusual ideas especially welcome.

[Edited on 26-2-2017 by notoxicshit]

notoxicshit - 26-2-2017 at 10:22

Excuse the double post, but since all of my questions have been ignored, here are the most important ones:

In which direction is the reaction time of propiophenone expected to be off, regarding the 22,5 h acetophenone needs for this reaction?
Important question RIGHT NOW, as it's 22,5 h into the second batch, now.

Do you believe acetone removal to be significantly beneficial and do you have ideas of how to make that feasible?
How much Acetone does cheap IPA contain, from experience? It's made from it, AFAIK.

Please discuss water content of the reactant and need for drying, as this is not mentioned in the paper. Commercial hydroxides of alkali metals often contain not exactly little water as they're hygroscopic.
If yes, how to bind/remove water (do mol sieves work in refluxing IPA?)?
If not important, as the open system used, suggests, please discuss why this is. This is paradoxical, as alkoxides don't exactly like water.

DraconicAcid - 26-2-2017 at 10:28

I don't know about the water, but the acetone will probably evaporate during reflux. Its boiling point is sufficiently low that its condensation in the condensor will be insignificant (since all the IPA vapour will keep it warm).

macckone - 26-2-2017 at 10:42

The acetone needs to be removed for the reaction to
progress. Otherwise acetone would be a competing
reaction back to isopropanol. In theory this could
work with MEK to 2-butanol. The acetone has a
20C differential, so should evaporate preferentially.
With good reflux and keeping the column top
below 70C, there should be good resolution of the
acetone. Obviously higher boiling point ketones
are going to work better. And obviously the
column top needs to stay below 80C to avoid
boiling off the isopropanol as this is long reflux.

The final workup is going to be a pain.
The source ketone remaining is going to
be soluble in ethyl ether. The good thing
is that there is likely to be a decent differential
between the boiling point of the alcohol and the
ketone. So the final distillation should have
decent separation. However, distilling more than
once is probably a good idea.

Without a mechanism established, it is impossible
to say rather water will help or hinder the reaction.
My guess is that it will hinder it but some minimum
amount is ok and possibly even necessary.

stoichiometric_steve - 26-2-2017 at 13:56

Quote: Originally posted by notoxicshit  

All the questions asked are answered in the paper.


Actually, the paper raises even more questions:

"After cooling to room temperature, the reaction mixture (200ml IPA)
was partitioned between diethyl ether (200 mL) and water (100 mL). The layers were separated […]"

I hardly believe that the mixture given above will even separate into layers.

Again, BS.
Quote: Originally posted by notoxicshit  
but since all of my questions have been ignored


That's because you're expecting to be spoonfed without giving details of your experiments.

Also, as stated before, how do you reflux this stuff and thus remove acetone in a pressure cooker?

I don't know what you expect from this forum, since you even seem to be unable to stick to the given procedure.


[Edited on 26-2-2017 by stoichiometric_steve]

CuReUS - 26-2-2017 at 21:01

Quote: Originally posted by stoichiometric_steve  

Even so, i don't even believe that any acetone would be formed, and by what mechanism? Suddenly because of a little NaOH, IPA will become some magic hydrogen donor? Sure.

read the 2015 paper in my previous post.They give a mechanism for ethanol.A similar mechanism should work for IPA
Quote: Originally posted by clearly_not_atara  
phenylacetone e.g. might work okay although benzaldehyde probably wouldn't
Quote: Originally posted by stoichiometric_steve  
Of course not, it would just go Cannizzaro.

Cannizaro requires a high concentration of base(>5 moles of NaOH).The 0.25 moles of NaOH used in this reaction isn't enough.I think atara was refering to the claisen schmidt reaction between benzaldehyde and acetone.
Quote: Originally posted by notoxicshit  

In which direction is the reaction time of propiophenone expected to be off, regarding the 22,5 h acetophenone needs for this reaction?

I think propiophenone will take more time than acetophenone because since the reduction happens by hydride shift,the ethyl group of propiophenone will reduce the positive charge on the carbonyl carbon more than the methyl group of acetophenone.So propiophenone will be less reactive and hence take longer time to reduce.
Quote:
Do you believe acetone removal to be significantly beneficial and do you have ideas of how to make that feasible?

from common sense,removal of acetone should shift the equilibrium forward and speeden up the reaction.As draconic acid mentioned,the acetone probably evaporates during reflux.Othewise you can remove it by following the MPV procedure:-
http://www.sciencedirect.com/science/article/pii/00404020588... (pg 37)
Quote:
Please discuss water content of the reactant and need for drying

I think it would be better if anhydrous reagents are used

[Edited on 27-2-2017 by CuReUS]

byko3y - 26-2-2017 at 21:52

At room temperature mixture of DEether:IPA:H2O=2:2:1 is surely biphasic, it's close to 7:6:12 in molar parts with DE ether and water being main constituents: http://dx.doi.org/10.1016/j.fluid.2011.07.011
Alkoxides are know to catalyze MPV/openauer reaction, there's a lot of confirmations, including this one http://www.synarchive.com/named-reactions/Oppenauer_Oxidatio... and that one https://commons.wikimedia.org/wiki/File:Woodward_modificatio... , just google it and you'll find much more.

[Edited on 27-2-2017 by byko3y]

byko3y - 4-3-2017 at 04:39

50 g (64 ml) of 99% isopropanol, 6.5 g (160 mmol) of NaOH, 4.2 g of cyclopentanone (50 mmol). Almost a day practically at reflux, despite the fact the device was for distillation - no significant amount of acetone detected, maybe because i screw something up and didn't notice the acetone coming of. Reaction mixture quickly turns yellow and in an hour becomes completely dark. Those are water-insoluble dimeric-polymeric compounds i'm not going to analyse in any way. After two distillations, removal of ketones with bisulfite (that's why cyclopentanone is good - it reacts well with bisulfite), drying, 0.41 g of compound boiling at 139-141 C was obtained, 10% yield of a pure colorless cyclopentanol. I believe the true yield is at least 20% and significant amount of cyclopentanol was lost during manipulations because such a small amount of the product is not sutable for my 20 ml scale devices.
Obviously, cyclopentanone quickly condenses with itself and acetone, forming some products having quite a good freshy odor, kinda similar to camphora. I was not going for the recovered ketone, so god only knows how much of it left unchanged. Some high boiling alcohols are present as products, probably as a result of reducing the aldol addition products.
I have a microwave oven, but I really doubt I can get same results as the original researchers. The method definitely works, but I would like to see somebody obtaining same resealts as the researchers.

ziqquratu - 5-3-2017 at 16:14

Unless you've got a pressure vessel for the microwave, don't bother - notoxicshit's original assessment was reasonable, in that the evidence for a specific microwave effect (beyond that achieved through higher temperatures under pressure) is very limited (and heavily questioned these days), except in a few very specific circumstances.

The relative rates of reaction for acetophenone and propiophenone should be fairly similar (propiophenone, as mentioned by CeRuS, likely a little slower if anything - but I wouldn't expect a dramatic difference). Do you have any ability to monitor the reaction by t.l.c. to check for completion? That'd be the best way to figure out when it's done.

If there's a concern about water (I'm not sure how sensitive the MPV reduction is - particularly under these conditions!), you could add the NaOH to some excess 100% isopropanol then distil off the extra solvent - the azeotrope should remove much of the water that the base may have introduced.
Quote: Originally posted by stoichiometric_steve  
Quote: Originally posted by CuReUS  

this reaction has been known since the 60's,so its more fact than fiction.


If you wouldn't mind providing a reference, that would be great. Otherwise, i still call this reaction BS.


Steve, check out the Meerwein-Ponndorf-Verley Reduction and the corresponding Oppenauer Oxidation. Obviously the classical reactions use aluminium alkoxides, and mechanistically involve bonding between the aluminium and the ketone being reduced, but it's not impossible to envision the reaction working with simple alkali metal alkoxides. Indeed, the linked page for the MPV reduction cites a reference from Organic Letters in 2007 which describes the MPV reduction of aromatic aldehydes (plus cinnamaldehyde) using stoichiometric LiBr with excess triethylamine in isopropanol, rendering this paper at least plausible.

byko3y - 5-3-2017 at 16:58

ziqquratu, unlike sodium and potassium alkoxides, aluminium alkoxides are hydrolyzed by water almost irreversibly. In the mean time, sodium and potassium alkoxides catalyze aldol condensation much more than those of aluminium, thus removal of water might lead to dehydration of aldol condensation products.

CuReUS - 5-3-2017 at 22:13

Quote: Originally posted by byko3y  
50 g (64 ml) of 99% isopropanol, 6.5 g (160 mmol) of NaOH, 4.2 g of cyclopentanone (50 mmol).Reaction mixture quickly turns yellow and in an hour becomes completely dark. Those are water-insoluble dimeric-polymeric compounds

you have used 10 times the conc of NaOH required for the reaction.No wonder there was so much aldol:o
Quote: Originally posted by ziqquratu  
Unless you've got a pressure vessel for the microwave, don't bother - notoxicshit's original assessment was reasonable, in that the evidence for a specific microwave effect (beyond that achieved through higher temperatures under pressure) is very limited (and heavily questioned these days), except in a few very specific circumstances.

I don't understand why everyone is behind microwave chemistry.So many books have been written on it and there are few reactions which have not been tried in MW.Obviously these chemists are not fools to follow something if it was completely fake.
Quote:
If there's a concern about water (I'm not sure how sensitive the MPV reduction is - particularly under these conditions!), you could add the NaOH to some excess 100% isopropanol then distil off the extra solvent - the azeotrope should remove much of the water that the base may have introduced.

I had said earlier that anhydrous conditions should be used,basing my reasoning on MPV,but later I came across a paper where HCOOH was used as a catalyst for reduction of carbonyl group and water was essential in that reaction.So unless both anhydrous and wet conditions are tried and compared,nothing can be said for this reaction.
Quote:
Quote: Originally posted by stoichiometric_steve  
Quote: Originally posted by CuReUS  

this reaction has been known since the 60's,so its more fact than fiction.


If you wouldn't mind providing a reference, that would be great. Otherwise, i still call this reaction BS.


Steve, check out the Meerwein-Ponndorf-Verley Reduction and the corresponding Oppenauer Oxidation.

I made a mistake there.The paper I saw from the 60's also used alumina ,so its not applicable in this reaction.But the earliest papers for reactions exactly like this are from 2000's
Quote: Originally posted by byko3y  
ziqquratu, unlike sodium and potassium alkoxides, aluminium alkoxides are hydrolyzed by water almost irreversibly.

but I thought that even Na/K alkoxides were extremely sensitive to water?
Quote:
thus removal of water might lead to dehydration of aldol condensation products.

I am not understanding what you are trying to say.Are you saying that water should not be removed from this reaction because then you would get predominantly aldol products ?

[Edited on 6-3-2017 by CuReUS]

byko3y - 5-3-2017 at 23:48

CuReUS, yes, I used higher concentration, because the 0.25 M carbonyl and NaOH concentration used by original researchers is some really dillute solution, you gonna need 1 L of isopropanol to reduce 21 g of cyclopentanone, which is unacceptable, unless you are reducing some really precious ketone. I was using 0.78 M cyclopentanone and 2.5 M NaOH concentration (exactly 10x original), but sodium hydroxide was not completely dissolved, so the real concentration was below 2 M.
Quote: Originally posted by CuReUS  
I don't understand why everyone is behind microwave chemistry.So many books have been written on it and there are few reactions which have not been tried in MW.
It's easy to declare novelity of your research just by repeating a well known reaction using MW. Especially indian and iranian researchers love it.
Unraveling the Mysteries of Microwave Chemistry Using Silicon Carbide Reactor Technology
Quote: Originally posted by CuReUS  
I thought that even Na/K alkoxides were extremely sensitive to water?
They do. Small amount of water can heavily shift the equilibrium of IPA+NaOH<->NaIPA + H2O towards left side.
Quote: Originally posted by CuReUS  
Are you saying that water should not be removed from this reaction because then you would get predominantly aldol products ?
It might be true, it might be not though. Only practice will tell us, and my trial yielded some mediocre results.
I have a data for cyclohexanone self condensation only: the equilibrium is shifted to the monomer at room temperature and saturation with water, while at higher tempeartures solubility of water rizes and monomer becomes the sole compound. Above 100 C ketol (dimer) is dehydrated, irreversibly yielding an unsaturated ketone.
But in our case there's also acetone and reduction of condensation products, so the resulting mixture is very complex, and I have cyclopentanone instead of cyclohexanone, so god only knows what are the products.

ziqquratu - 6-3-2017 at 16:13

Quote: Originally posted by CuReUS  
I don't understand why everyone is behind microwave chemistry.So many books have been written on it and there are few reactions which have not been tried in MW.Obviously these chemists are not fools to follow something if it was completely fake.


It's not about being stupid, it's about it being very easy to fool yourself into thinking there's an effect when there may not actually be one. For microwave chemistry, much of the initially reported effect seems to have resulted from poor measurement of the reaction temperature (meaning the rate was faster than predicted for the temperature measured - but the temperature measured was lower than the actual temperature, accounting for the increased rate!). That was compounded by the "lab" microwaves which had dodgy IR sensors, because it's *really* easy to just trust what the shiny, expensive new machine tells you...

There can be advantages. With the lab instruments, it's an easy way to do a reaction under pressure, without needing a special autoclave setup. You can also generally heat the reaction very quickly and more evenly than a heat bath, which can have its advantages (less hot-spots on the walls can lead to less charring, say). You can get reactors to improve the rate of heating, too - like the SiC ones byko3y mentioned. And there do seem to be a few examples where there might be actual microwave-specific (albeit still thermal) rate increases - if you have a solid reactant which strongly absorbs microwaves, for example, you can get a high surface temperature (which means high temperature where the reaction is actually taking place) but a low bulk temperature, leading to fast reaction with less thermal decomposition. But much of the hype seems to be unjustified when people have gone back and studied it more carefully. There was a reasonably amusing (for the scientific literature) argument in Angewante Chemie back in 2012 or so between Oliver Kappe (one of the microwave chemistry gurus, who holds that there's no microwave specific effect for reactions in homogeneous solutions) and another author on the topic.

byko3y is right, though, that using a microwave is a good way to get a paper published!
Quote: Originally posted by CuReUS  
I thought that even Na/K alkoxides were extremely sensitive to water?


I think the point was that water will irreversibly destroy aluminium alkoxides (because aluminium hydroxide will precipitate from solution), whilst sodium and potassium alkoxides exist in a reversible equilibrium - which the reaction in the original reference seems to rely upon.
Quote: Originally posted by byko3y  
sodium and potassium alkoxides catalyze aldol condensation much more than those of aluminium, thus removal of water might lead to dehydration of aldol condensation products.


It's possible, certainly. I don't know how big a deal it's going to be, given the low concentration of base and the *very* low concentration of alkoxide. I'd think the hydroxide itself (which is entirely capable of promoting aldol condensation, and present in a much higher concentration) would be the bigger concern - and taking the paper at face value, it doesn't seem to be an actual concern. But also, they seem to have taken no particular precautions about the presence of water - perhaps you'd see a rate increase if you removed the water (due to an increase in alkoxide concentration), but whether that's worth the hassle is another matter, too.

zed - 6-3-2017 at 18:31

Well, somewhere, I read an account of quantitative yields, reducing benzaldehydes w excess isopropyl-alcohol via heat and pressure. No catalyst.

Naturally, I can no longer find the reference.

The suggestion that aldehydes are preferentially reduced vs ketones, does suggest that once benzylalcohol has been formed, it likes to stay benzylalcohol.

The Vespiary has a reference.


"reducing benzaldehyde with isopropyl" Type the foregoing in, and search via Google. Should come right up.


[Edited on 7-3-2017 by zed]

[Edited on 7-3-2017 by zed]

[Edited on 7-3-2017 by zed]

byko3y - 6-3-2017 at 22:05

Quote: Originally posted by ziqquratu  
given the low concentration of base and the *very* low concentration of alkoxide.
Equilibrium constant for IPA+NaOH<-> NaIPA+H2O, K=[NaIPA][H2O]/[IPA][NaOH] is close to 1, although i have no idea how close it is: it might be 3, it might be 0.3. For initial 13 M IPA and 0.25 M NaOH with relatively anhydrous reagents the equilibrium is K=[NaIPA]*[NaIPA]/(15-[NaIPA])(0.25-[NaIPA]), from which we get something like [NaIPA] = [H2O] = 0.246 M, [NaOH] = 0.004 M.
For my case I can calculate [NaIPA] = [H2O] = 2.14 M, [NaOH] = 0.36 M. I believe it was finally dissolved completely, despite the fact I was not able to verify it due to some difficulties.
As you can see, the ratio [NaIPA]/[NaOH] in original conditions is 300-1000, while in my conditions it is 5-10. AFAIK, alkoxides are just as potent aldol catalysts, as hydroxides. But MPV is not catalyzed by hydroxides. I believe the reason I got so much by-products is high concentration of carbonyls, and not a high concentration of the base.

ziqquratu - 7-3-2017 at 15:27

I've never even bothered to look up those equilibria - you're somewhere around the right number (for EtOH + NaOH <--> NaOEt + H2O, K ~ 0.7). Colour me surprised, for sure!

I think my analysis holds, though, just changing the focus to the still low (but not "very" low) alkoxide concentration, and effectively neglecting the hydroxide now.

I agree with your analysis, too - higher ketone concentration will likely favour addition/condensation for a given concentration of base, and you've got a much higher concentration of base as well (the paper reported 0.25 M with respect to both ketone and base).

CuReUS - 11-3-2017 at 11:57

Unless someone repeats the synthesis,following the author's instructions to the letter(byko3y:D) and reports their findings,there is no use talking about it.The whole point of the thread was to find out whether the paper was correct or not.

byko3y - 11-3-2017 at 12:54

This insane level of dilution makes the method of a low practical value, not just because of the need for a large reaction vessel, but also because of tedious workup to separate few grams of the target product from a bunch of byproducts in the excessive amount of solvent. In fact, it took me more time to extract the pure product than to perform the actual reaction.
If I had to do the reduction, I would prefer a good old covalent Al(Oi-Pr)3 catalyst with high concentration of reagents, from which you can easily distill acetone, and finally do a liquid-liquid extraction right from the reaction mixture without the need to concentrate it.