clearly_not_atara
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 2725
Registered: 3-11-2013
Member Is Offline
Mood: Big
|
|
Birch reduction of tribenzylaluminium?
Just wondering what this would produce. The natural assumption is dibenzylcyclohexadienylmethylaluminium, but either this compound or the intermediate
preceding it can eliminate toluene to leave, presumably, dibenzylaluminium hydride. This would be the same as the Birch reduction but the affinity of
Al for hydrogen is much weaker than that of O.
If Bn2AlH is produced this would be the simplest (IMO) practical OTC preparation of hydrides. The increased molecular weight of benzyl vis-a-vis
isobutyl may decrease the pyrophoricity but I wouldn't bet on it.
Tribenzylaluminium is conventionally prepared by the reaction of aluminium with dibenzylmercury, but it might also be prepared by the reduction of
benzylaluminium sesquichloride with sodium analogous to the preparation of trimethylaluminium. I haven't been able to find literature confirming or
disconfirming that this works
EDIT: On second thought, BnCl + Al metal probably gives a Wurtz coupling. Mg is the way to go here most likely (Li also gives a Wurtz).
However, instead of tribenzylaluminum, maybe better to use dibenzylzinc and two Birch equivalents to go all the way to bis(cyclohexadienylmethyl)zinc
and thus to zinc hydride. ZnH2 is usually not pyrophoric but old samples can sometimes combust.
[Edited on 21-5-2017 by clearly_not_atara]
|
|
Alice
Hazard to Others
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 111
Registered: 11-5-2015
Member Is Offline
Mood: No Mood
|
|
Not sure about the Birch reduction of tribenzyl aluminium. May it be possible aluminium being reduced to the metal? But I really don't know it. Would
be neccessary to search the literature in order to find something just roughly similar.
About the preparation of tribenzyl aluminium, my intuition says a benzylmagnesium halogenide reacts with AlCl3 to yield the product.
I was checking out too how it may be possible making an aluminium hydride species and the most direct way I think (not the least dangerous one!) would
be making NaH from Na in a stream of hydrogen and reacting the obtained NaH with AlCl3 giving NaAlH4 which equals four reducing equivalents. (Link) But NaAlH4 is more reactive than LiAlH4 so it's even more critical to handle. However it can be converted to LAH with LiCl. Not a dream
solution for the problem.
[Edited on 22-5-2017 by Alice]
[Edited on 22-5-2017 by Alice]
|
|
halogen
Hazard to Others
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 372
Registered: 18-4-2004
Member Is Offline
Mood: No Mood
|
|
I too thought there'd be immediate reduction to benzyl sodium and aluminum. If two of us imagined it, it must be true.
F. de Lalande and M. Prud'homme showed that a mixture of boric oxide and sodium chloride is decomposed in a stream of dry air or oxygen at a red heat
with the evolution of chlorine.
|
|
Alice
Hazard to Others
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 111
Registered: 11-5-2015
Member Is Offline
Mood: No Mood
|
|
Found a patent which might be of interest here:
Manufacture of trialkylaluminium compounds
@halogen ![:D](./images/smilies/biggrin.gif)
[Edited on 23-5-2017 by Alice]
I wonder if tribenzyl aluminium would react with ammonia directly. Found this:
http://pubs.rsc.org/en/Content/ArticleLanding/1965/JR/jr9650...
Thinking about DIBALH and how it is made from triisobutyl aluminium by beta-hydride elimination, I wonder if a mixed compound like dibenzylisobutyl
aluminium would work too. A statistical distribution of substituents would lead to a fairly high percentage of the mixed form with just one isobutyl
subtituent.
[Edited on 24-5-2017 by Alice]
|
|
clearly_not_atara
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 2725
Registered: 3-11-2013
Member Is Offline
Mood: Big
|
|
I share your concern regarding the target of e- in the Birch reduction. I suspect tribenzylaluminium will indeed complex with amines but this could be
a good thing; the N lone pair coordinated to the Al center will prevent it from reacting with e- directly, will it not?
Another version of this that I liked was reacting ethylaluminium dichloride (EtAlCl2) with pinacol and then heating to eliminate ethylene, leaving
pinacolatoaluminium monohydride, but I don't know if pinacolatoaluminium hydride is even stable.
The use then of ethyldibenzylaluminium and its pyrolysis is a possibility, needing only slightly higher temperatures, and I am of course hoping, that
the increased mass and stability of the benzyl ligands vs isobutyl will reduce the tendency of the product to spontaneously catch fire.
EDIT: Dialkoxyaluminium monohydrides have indeed been reported and may even be kinetically stable:
http://link.springer.com/article/10.1007/BF02503792
Quote: | I wonder if a mixed compound like dibenzylisobutyl aluminium would work too. A statistical distribution of substituents would lead to a fairly high
percentage of the mixed form with just one isobutyl subtituent. |
Another possibility is the reaction of ethylaluminium dichloride with benzylsodium or a benzyl Grignard. I mention benzylsodium because there is a
patent relating to its formation from toluene and (dry) NaOH in DMSO, a solvent with the unusual property that toluene is more acidic than water
(!!!):
https://www.google.com/patents/US20060170118
This even allows you to get to dibenzylethylaluminium without using any pure metals other than aluminium, which is exciting.
![](https://s15.postimg.org/c9dqkzt0b/dibnal.png)
[Edited on 24-5-2017 by clearly_not_atara]
|
|
Alice
Hazard to Others
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 111
Registered: 11-5-2015
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by clearly_not_atara ![](images/xpblue/lastpost.gif) | I share your concern regarding the target of e- in the Birch reduction. I suspect tribenzylaluminium will indeed complex with amines but this could be
a good thing; the N lone pair coordinated to the Al center will prevent it from reacting with e- directly, will it not? |
In the paper I cited before, a paper from Ziegler and Gellert is cited, where they state trialkylaluminium and dialkylaluminiumhydride compounds form an adduct with ammonia which
decomposes according to the following scheme:
AIR3 x NH3 ---> R2AINH2 + HR
and
R2AIH x NH3 ---> R2AlNH2 + H2
Quote: | Another version of this that I liked was reacting ethylaluminium dichloride (EtAlCl2) with pinacol and then heating to eliminate ethylene, leaving
pinacolatoaluminium monohydride, but I don't know if pinacolatoaluminium hydride is even stable. |
I'm afraid the first thing that happens might be the ethyl group grabs a proton from the alcohol and is eliminated as ethane. But even if chloride
gets substituted there is HCl formed. What I can imagine is reacting trialkylaluminium with pinacol in order to eliminate two alkanes.
Quote: | The use then of ethyldibenzylaluminium and its pyrolysis is a possibility, needing only slightly higher temperatures, and I am of course hoping, that
the increased mass and stability of the benzyl ligands vs isobutyl will reduce the tendency of the product to spontaneously catch fire.
|
I think we both agree any of this should be done in small amounts and in an environment where a fire is no big deal. ![;)](./images/smilies/wink.gif)
Electron withdrawing groups reduce the electron density on Al and the adjacent carbons, so I guess this makes it less critical. It can be seen by
comparing MeLi, BuLi, and tBuLi. Red Al is less critical than DIBALH, I guess this has also something to do with the alkoxy substituents on Al,
stabilizing the negative charge. But vapor pressure follows the same trend of course.
EDIT: Your last suggestion looks very interesting. I will read and think about that in depth later.
[Edited on 25-5-2017 by Alice]
|
|
clearly_not_atara
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 2725
Registered: 3-11-2013
Member Is Offline
Mood: Big
|
|
Quote: | I'm afraid the first thing that happens might be the ethyl group grabs a proton from the alcohol and is eliminated as ethane. But even if chloride
gets substituted there is HCl formed. What I can imagine is reacting trialkylaluminium with pinacol in order to eliminate two alkanes.
|
I miscommunicated. I really meant to say the pinacolate dianion, which will result in the elimination of Cl- (in preference to ethylide), rather than
ethane (in preference to HCl). However, the pinacol rearrangement is an issue (owing to the acidity of Al) and may sink the idea. The diethoxy version
might be synthesized (from NaOEt), but I think this would be inferior to the dibenzyl, and loses the acidic property of DIBAL.
|
|
Alice
Hazard to Others
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 111
Registered: 11-5-2015
Member Is Offline
Mood: No Mood
|
|
Quote: | I miscommunicated. I really meant to say the pinacolate dianion |
That makes sense now.
DIBALH is a very nice reagent as it allows for chemo-selectively reducing many functional groups. Carboxylic acids as well as esters can be reduced to
aldehydes or alcohols, nitriles can be reduced either to aldehydes or amines. (Link)
I'd still like to find some literature sources about how good it works for reducing 2° amides.
Quote: | The diethoxy version might be synthesized (from NaOEt), but I think this would be inferior to the dibenzyl, and loses the acidic property of DIBAL.
|
The scope seems limited compared to DIBAL/LAH/NAH, but still interesting if it turns out they are better to handle and/or make. Basically M(RO)3AlH is
an adduct of (RO)2AlH and ROM.
Scope:
https://en.wikipedia.org/wiki/Reductions_with_metal_alkoxyal...
http://www.massey.ac.nz/~gjrowlan/oxid/meta.pdf (p. 10)
I have two questions about your last suggestion. First, how do you plan making monoalkylaluminium dichloride? Second, what are the conditions of
beta-hydride elimination for the ethyl group? The reason why isobutyl eliminates so well is that the beta-position is substituted by two methyl groups
pushing electrons and the stable isobutene is formed. So if isobutyl is unavailable, propyl or butyl may be better alternatives.
EDIT: Found a source for elimination temperatures: Link (p. 3)
Triethylaluminium: 120 °C
Triisobutylaluminium: 60 °C
Tributylaluminium: 100 °C
Trioctylaluminium: 80 °C
The patent about benzyl sodium compounds looks indeed appealing. A backdraw is, removing DMSO by reduced pressure distillation may be a bigger hassle
than it sounds.
[Edited on 25-5-2017 by Alice]
[Edited on 25-5-2017 by Alice]
EDIT: Old SM thread: http://www.sciencemadness.org/talk/viewthread.php?tid=12022
[Edited on 25-5-2017 by Alice]
|
|
Alice
Hazard to Others
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 111
Registered: 11-5-2015
Member Is Offline
Mood: No Mood
|
|
Great paper by Grosse and Mavity cited in the Ziegler patent (US2,691,668):
ORGANOALUMINUM COMPOUNDS: I. Methods of Preparation.
They managed to separate ethylaluminium dihalide from diethylaluminium halide by fractionated distillation, although state it's difficult and
inefficient. Instead they recommend the following:
R2AlX + AlX3 ---> 2 RAlX2
For the synthesis of alkoxyaluminium compounds:
2 R3Al + (R'O)3Al ---> 3 R2(R'O)Al
or
R3Al + 2 (R'O)3Al ---> 3 R(R'O)2Al
For the trialkylaluminium synthesis I wonder if distilling products is really neccessary or if Schlenk filtration would be satisfactory as well.
EDIT: DIY Magnalium
[Edited on 27-5-2017 by Alice]
|
|
clearly_not_atara
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 2725
Registered: 3-11-2013
Member Is Offline
Mood: Big
|
|
I think it's best to avoid making any triethylaluminium, but it looks like you'll get good results from combining ethylaluminium sesquichloride
(relatively easy, EtCl + Al) with AlCl3. So that solves that.
However I haven't found a ref confirming that benzylsodium synthesis. The patent has no literature citations. It's a little dubious because toluene is
less acidic than DMSO itself so you might get dimsyl sodium instead which is less useful here. If you use a different hydrocarbon, the theoretical
product switches back to the alkylsodium.
See the Bordwell pKa table easily found on Google; it seems like 1-naphthylmethyl sodium could work, or maybe 2-methylbenzofuran. There are no stable
aluminocenes in the literature so I'm guessing C-H acids based on cyclopentadiene are out.
[Edited on 27-5-2017 by clearly_not_atara]
|
|
Alice
Hazard to Others
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 111
Registered: 11-5-2015
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by clearly_not_atara ![](images/xpblue/lastpost.gif) | I think it's best to avoid making any triethylaluminium, but it looks like you'll get good results from combining ethylaluminium sesquichloride
(relatively easy, EtCl + Al) with AlCl3. So that solves that. |
Maybe this is wishful thinking now, but my impression is the substituents can be transfered between aluminium centers freely, so 1 eq. EtX and 2 eq.
BnCl may lead to the product EtBn2Al directly. I suspect the exchange to happen, as looking at the structure of the Et3Al dimer, there are two ethyl
groups shared between both Al centers. The distribution would depend on thermodynamic equilibrium then of course.
A problem with EtCl is that it's a gas and the reaction with aluminium is exothermic. Therefor Grosse and Mavity used an autoclave for this reacion.
On the other hand it seems not to be neccessary for higher boiling alkyl halides.
Quote: | However I haven't found a ref confirming that benzylsodium synthesis. The patent has no literature citations. It's a little dubious because toluene is
less acidic than DMSO itself so you might get dimsyl sodium instead which is less useful here. If you use a different hydrocarbon, the theoretical
product switches back to the alkylsodium.
See the Bordwell pKa table easily found on Google; it seems like 1-naphthylmethyl sodium could work, or maybe 2-methylbenzofuran. There are no stable
aluminocenes in the literature so I'm guessing C-H acids based on cyclopentadiene are out.
[Edited on 27-5-2017 by clearly_not_atara] |
Maybe I'm missing something, but isn't NaOH/DMSO a special superbase system? The equilibrium depends on the stabilization of all reactands, so the
whole picture needs to be examined.
EDIT: Maybe you're right, so direct comparison of pKa values may be legit. I just thought benzyl sodium might be favorable as it is not charged and
the carbon-sodium bond has a covalent character. The pKa measurement considers the following equilibrium:
toluene + DMSO <===> benzyl anion + DMSO-H⁺
not
toluene + dimsyl sodium <===> benzyl sodium + DMSO
Of course the patent may be scam, but wouldn't be too much effort to uncover it - given that a good vacuum pump is at hand.
[Edited on 28-5-2017 by Alice]
[Edited on 28-5-2017 by Alice]
|
|
Alice
Hazard to Others
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 111
Registered: 11-5-2015
Member Is Offline
Mood: No Mood
|
|
http://www.sciencemadness.org/talk/viewthread.php?tid=51199#...
The patent was already discussed here.
|
|
clearly_not_atara
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 2725
Registered: 3-11-2013
Member Is Offline
Mood: Big
|
|
:p
I suppose we shouldn't find that all too surprising.
I have always kind of wondered, though, about making diisopropylamides -- usually made from organometallics, but shouldn't you be able to make sodium
diisopropylamide by dissolving Na in iPr2NH? Is the reaction really too slow to be practical? In this paper, isoprene is used as a promotor -- why
isoprene?
http://pubs.rsc.org/en/content/articlelanding/1992/c310.1039...
According to this:
https://owl.oit.umass.edu/departments/OrganicChemistry/appen...
toluene is slightly (by 1 point) less acidic than diisopropylamine in water (although the pKa is theoretical), but it is also less volatile. It might
be possible to make benzylsodium from sodium diisopropylamide, or failing this, to make e.g. naphthylmethylsodium the same way. Possibly in some
ethereal solvents or in toluene itself we might find that toluene is more acidic than diisopropylamine, which would be convenient.
|
|
Alice
Hazard to Others
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 111
Registered: 11-5-2015
Member Is Offline
Mood: No Mood
|
|
Quote: | :p
I suppose we shouldn't find that all too surprising. |
Nope. A funny detail about the author is, he seemingly used 'the right to forget' on his google patents record. Probably getting rid of even more
unfavorable stuff.
Quote: | According to this:
https://owl.oit.umass.edu/departments/OrganicChemistry/appen...
toluene is slightly (by 1 point) less acidic than diisopropylamine in water (although the pKa is theoretical), but it is also less volatile. It might
be possible to make benzylsodium from sodium diisopropylamide, or failing this, to make e.g. naphthylmethylsodium the same way. Possibly in some
ethereal solvents or in toluene itself we might find that toluene is more acidic than diisopropylamine, which would be convenient.
|
I checked out pKa values for diisopropylamine and toluene in THF, which are reported being 36 (source) and 37.4 (source) respectively. As mentioned before, the pKa refers to the acid/anion equilibrium and is therefor only a valid description if ions were at
least solvent-separated, not for covalently bound metal organics. However, I'd give it a chance if temperature is kept low and DiPA removed under
reduced pressure in order to avoid decomposition.
http://pubs.rsc.org/en/content/articlelanding/1992/c310.1039...
Quote: | An alternative route to LDA is the reaction of lithium with
diisopropylamine with the aid of an electron carrier such as
styrene or isoprene.4 |
Electron carrier! There are further references given for the catalyst's action.
EDIT:
Found an interesting book chapter about alkylaluminium compounds. It says tribenzylaluminium doesn't form a dimer but a chain structure: Link
[Edited on 29-5-2017 by Alice]
|
|