Dr. Beaker - 24-5-2007 at 07:33
Hi all,
I'm interested in making the compound: H2N-(CH2)2[O(CH2)2]3-NH2 (i.e - 4 condensed units of ethylene glycole, with the terminii NH2 instead of OH.
I thought of 2 methods:
1. starting from ethylene glycol and ethanol amine:
condensing 2 units of ethylene glycol (how to control the degree of condensation?) to HO(CH2)2O(CH2)2OH and condens this diol with 2 equivalents of
ethanol amine - I assume that in the acidic condition needed the amine will be protonated.
that's the method I prefer.
2. starting from 15 crown 5 (:
1. bromination, to a monobromide (i.e very low conc. of Br2)
2. dehydrohalogenation (-HBr, KOH in EtOH)
now I have a double bond with 2 alkoxy substituents.
3. oxidation. since I'm getting ester of formic acid it will further oxidized and loose CO2.
now I have almost got my target molecule, but the last carbons have OH, or oxidized to aldehyde or carboxyl. so the next step would be to turn any of
those to amine...(what's the most efficient ways??)
any ideas to make it as easy as possible?
[Edited on 24-5-2007 by Dr. Beaker]
sparkgap - 24-5-2007 at 08:15
If you have the requisite equipment, oxirane might be better to use than glycol. 
sparky (~_~)
Nicodem - 24-5-2007 at 08:20
Not exactly what you asked for, but it might help you nevertheless:
http://www.syntheticpages.org/browse.php?&action=1&p...
http://www.syntheticpages.org/browse.php?&action=1&p...
Dr. Beaker - 24-5-2007 at 09:10
Thanks for the info.
However, I found that the ditosylate of Diethylene glycol is sold at sigma-aldrich. from there it is only 1 step by using H2N(CH2)2OLi/Na in a good
ol' williamson.
Nicodem - 24-5-2007 at 11:24
I don't think that would be particularly efficient and the separation of the side products would be a nightmare solvable only by chromatography which
would be a pain to follow due to the UV non-absorbance of these compounds.
I think the simplest solution is to buy tetraglycol, which by the way is very cheap (~25EUR/L), and either use the Mitsunobo reaction as outlined in
the Synthetical pages procedure linked above or to add another step by first ditosylating tetraglycol and then do the Gabriel amine synthesis step.
The alternative of using 2-phthaloylethanol and diglycol ditosylate with NaH/THF should work but I see no advantage in the amount of work involved (it
would actually be more tedious and expensive).
Dr. Beaker - 24-5-2007 at 12:37
I agree with you in regard to the last comment.
But about the first - what side products do you mean?
attack of the amine instead of the alkoxide? won't the later be a better nuc.?
or maybe elimination?
because the methodes you suggested will force me to order more chemicals like DEAD phtalimide etc., and add at least 1 more step.
Nicodem - 26-5-2007 at 00:49
I understand that intuitively the -O(-) site of the deprotonated aminoethanol seems the most nucleophilic side of this bidentate nucleophile. However
this ceases to be so apparent when one considers that the nucleophilicity is not directly proportional to the pKa (the pKa is a factor describing the
nucleophilic relationship toward the proton as electrophile, while nucleophilicity is formally defined by the relationship toward MeBr as electrophile
which behaves very differently than protons do). The negative charge on oxygen makes the -NH2 site even more nucleophilic by inductive effect while
the nucleophilicity of the -O(-) site is lower than that of the comparable ethoxide due to the involvement in the pentahedral H-bonding with the
amine.
Anyway, what I'm saying is that I have my serious doubts that the reaction of H2N-CH2CH2-O(-) with any alkyl tosylate would be nearly as selective as
you appear to expect it. To my knowledge there are no examples of such reactions with ethanolamine in the literature but I will check when time allows
since I'm curious about it too.
Mind also that the intermediate product H2N-(CH2CH2O)3-Ts (as well as end product) is also nucleophilic and might attack another electrophile as well
as cyclisize to crown ether. The other potential problem is the high basicity of the deprotonated ethanolamine promoting the E2 elimination, which I
would however expect to be only a very minor side reaction. In any case I would expect a complex mixture of products very difficult to separate.
The advantage of working with the tetraglycol ditosylate and phthalimide is obvious: the tosylation uses the cheap TsCl and Et3N (or Py) on even
cheaper tetraglycol; only one product obtained upon nucleophilic substitution; the excess phthalimide is easily removed; the N-protected intermediate
easily purified and TLC visible; the deprotection straightforward… The Mitsunobo method seems too much of a fuss since it requires work with DEADly
chemicals in dry solvents, besides being somewhat more expensive.
Nicodem - 28-5-2007 at 10:13
Well, I checked the literature and there are less then a dozen references for O-alkylation of ethanolamine. Most cases are of benzylation with benzyl
chlorides and the yields are usually from 25-50% for the O-alkylated product. By some strange coincidence one of the examples is exactly the reaction
you proposed, diglycol ditosylate with ethanolamine, but the yield is of only 25% (Liebigs Annalen der Chemie (1990) 129-143). That is also
the reference you need because it discuss other routes using the azide nucleophilic substitution as well. The other preparation method for your end
compound includes the phthalimide nucleophilic substitutions of the appropriate tetraglycol derivative as proposed in the earlier post (Organic
& Biomolecular Chemistry3 (2005) 2255-2261).
Dr. Beaker - 29-5-2007 at 08:12
Thank you for the very helpfull info.
"The advantage of working with the tetraglycol ditosylate and phthalimide is obvious".
It's good it's obvious for at least one of us.