0.03152m 5-OH-Vanillin (5.3g re-x from toluene), 0.09456m K2CO3 (13.07g, finely powdered), 0.03152M TBAB (10.16g), and 1.00864M Dimethyl Carbonate
(90.96g) was placed in a 500ml RBF and stirred in an oil bath under reflux at 88-93C for 5 hours.
Workup: diluted room-temp reaction mixture with water, based with a few grams KOH, extracted reaction mixture with ethyl acetate until satisfied
(don't know how many mls), then acidified reaction mixture with dilute H2SO4 and again extracted with ethyl acetate until satisfied. Each ethyl
acetate solution was washed with a bit of water, dried with Na2SO4, and vapped off under vacuum to yield a residue. Each residue was then extracted
with several changes of cyclohexane, then each cyclohexane solution was simply evaporated rather than crystallizing. Yields were poor, in the 3-5%
range for the trimethoxy compound.
The reason I wanted to post this at all was that in previous experiments using methyl bromide on the same substrate, it appeared to be very effective
with step-wise methylation. Poor yields (5-10%) of the fully methylated compound on the first round, but decent yields (50-60%) of mono-methylated
product (syringaldehyde structural isomers). Then, taking the mono-methylated product and running it through again resulted in about 50% yield of
trimethoxy compound. I noticed working this reaction up resulted in a lot of unchanged product, the dihydroxy compound but almost none of the
monomethylated compound. Also ethyl acetate will extract 5-oh-vanillin from aqueous solution and DCM will not, so it could be that sequential
extraction of the acidified reaction mixture, first with DCM to remove syringaldehyde/isomers then with ethyl acetate to remove unchanged starting
material would be effective for separating the 3 compounds.
I know that the temp could be much higher and the TBAB use was not economical. Just wanted to report that the monomethyl compound is produced in much
less amount relative to the dimethyl compound compared to other tried methylating agents.
|
