Azane - 26-3-2015 at 16:40
My idea is to sequentially react silicon tetrachloride with alkali hydrides and ethoxides in ether (both ethoxyethane and cyclic ethers should work)
to yield the final product (a hydride donor for a selective zinc-catalyzed amide reduction I plan to perform later on). All three of these reactants are rather straightforward in their
preparation, but I have questions regarding the hydrides and ethoxides themselves, as well as in what order they should be allowed to react with the
SiCl4.
The hydrides and ethoxides can be produced most easily and directly by the reaction of an alkali metal with, respectively, hydrogen gas and ethanol
(the hydrogen gas from the ethoxide production even being recycled for the hydride production); lithium metal would seem to be the most practical
choice for these reactions, given that it is both readily available online and considerably less expensive than sodium and potassium. However, from
what I've read, ionic lithium compounds (not counting alkyllithium reagents, which aren't even completely ionic) are generally insoluble or poorly
soluble in ether. No doubt that this is especially true for LiH, but does lithium ethoxide behave similarly, or does the ethoxide anion's ethyl group
aid the compound's ether solubility? If the latter is true, the question would then be to what extent the ions actually dissociate in solution, if the
compound's solubility is mainly due to non-polar interactions.
While I'm sure that SiCl4 would initially be able to react with a particle of LiH via surface reactions, I'd be worried about the formation
of a passivation layer of lithium chloride that might reduce final yield, as well as potentially throw off stoichiometry. Yes, I am aware that using a
suspension of finely-divided LiH in ether would help with this problem, but a suspension is, nonetheless, made up of solid particles, and there's a
limit to how small I, personally, can make those particles.
So, I ask, are there any alkali metal hydrides with appreciable solubility in ether? If there is at least one, the first practical benefit it would
confer would be a greater degree of control over stoichiometry, and a better chance of the reaction being able to proceed to completion. Another
benefit would be that, upon reacting in solution, the alkali cation and the chloride anion (donated by the substituted silane) would form an
ether-insoluble alkali chloride that would precipitate out of solution upon formation, thereby acting as an indicator that the reaction has occurred.
An final benefit would be that the alkali hydride could be safely formed in a glass container (allowing reaction process to be easily monitored),
unlike lithium hydride, due to the reactivity of molten lithium towards glass.
My last question: should I make the triethoxysilane from the ethoxide and trichlorosilane, itself formed by the reaction of SiCl4 with the
hydride? Or should I take the opposite approach, reacting the SiCl4 with the ethoxide to form triethoxychlorosilane (if it exists), and
then react that with the hydride to form the triethoxysilane? Does it even make a difference? Is any of this even viable?
Triethoxysilane from TEOS and an Alkali Hydride
Azane - 26-3-2015 at 17:30
Alright so I know I just posted a lengthy discussion of how to make triethoxysilane from silicon tetrachloride and an alkali hydride and ethoxide, so
I apologize for this in advance. After posting, I decided on a whim to search online for tetraethyl orthosilicate (tetraethoxysilane) for the first
time, and much to my surprise, it's available in significant quantities of supposedly high purity at reasonable prices. I can't decide whether to feel
stupid at not having done this earlier, or disappointed at no longer having a practical reason to exercise my ability in chemical synthesis.
Nevertheless, I'm interested in the reaction now. It's a nucleophilic substitution reaction between tetraethoxysilane and a hydride ion, forming an
alkali ethoxide as a byproduct.
Si(OC2H5)4 + MH --> MOC2H5 + (OC2H5)3SiH
It looks relatively straightforward on paper, but is it actually viable? Needless to say, ethoxide probably isn't as good of a leaving group as
something like chloride (from a chloro-substituted silane), but hydride is a stronger nucleophile and base than ethoxide, so I'd think that the
reaction would favor the formation of the triethoxysilane, if there isn't steric hindrance. Will the properties of ethoxide as a leaving group have an
impact on yield?
Either way, triethoxysilane is a liquid, meaning that it can be distilled to separate it from the solid ethoxide, even if as a solution in ether
(which shouldn't affect its viability as a reagent). Also, I am again wondering which alkali hydride would work best for this reaction.