2
When conducted at 50 oC (Entry 2), the conversion was
assumption, methyl benzoate was selected as an additive.
Unexpectedly, the conversion was rather slow in the presence of
excess methyl benzoate (5 eq.), and methyl protocatechuate (2)
was isolated in 78% yield after stirring for 18 h in hot acetonitrile
(Fig. 2).
typically complete within 1 h as indicated by TLC analysis.
However, it was still crucial to quench the reaction in time, since
prolonged stirring gave 5 in 11% isolated yield when the reaction
was quenched after 150 min (Entry 3). Upon quenching the
reaction after stirring at room temperature for 2 h, 2 was isolated
in 82% yield without the formation of 5 (Entry 4). The reaction
o
was sluggish at a lower temperature (15 C) and the conversion
was incomplete after stirring for 18 h, as observed by TLC (Entry
5). Unexpectedly, no further conversion was observed during
prolonged stirring, affording 2 in 68% isolated yield, possibly
because the aluminum triiodide was deactivated by moisture.
When the reaction was scaled up to 33 mmol, the transformation
became sluggish, and took 4 d to reach completion (Entry 6).
Finally, methyl acetate was also used, in which the reaction at 80
oC for 2 d proceeded without noticeable ester cleavage side-
reaction (Entry 7).
Figure 2. Demethylation of methyl vanillate using methyl benzoate
as a sacrificial ester
The unnecessary complication in product isolation made the
use of methyl benzoate less appealing. We reasoned that if the
aluminum triiodide promoted ether cleavage was faster than ester
cleavage, it would be feasible to run the demethylation
transformation in an ester solvent. Common solvents for
aluminum triiodide mediated reactions include carbon disulfide,
hexanes, benzene and acetonitrile.13 The use of ester solvents,
however, should be handled with care because of the known ester
cleavage reaction by aluminum triiodide.13 A survey of the
literature showed that ethyl acetate was once employed as the
solvent in the aluminum trichloride-pyridine mediated
demethylation of curcumin.14 Coincidentally, we previously
noted that the cleavage of methyl o-hydroxyphenyl ethers by
ATAS is rather fast when facilitated by neighboring o-hydroxyl
group participation.11, 12 Thus, eugenol (3) was selected as the
model compound for the evaluation of ethyl acetate as a solvent
in the aluminum triiodide mediated demethylation (Table 1).
Under optimized conditions for the demethylation of o-
hydroxyphenyl methyl ethers using aluminum triiodide (1.1 eq.)
and DIC (0.6 eq.) in hot acetonitrile,12 hydroxychavicol (4) was
isolated in 96% yield when quenched 3 min after the start of the
reaction, with the recovery of 2% unreacted eugenol (Entry 1).
The conversion was complete in 25 min (Entry 2). Next,
acetonitrile was removed under reduced pressure after
completion of the in situ preparation of aluminum triiodide,12 and
to the residue was added a solution of eugenol and DIC in ethyl
acetate. Pleasingly, 4 was isolated in 76% yield after stirring
overnight at reflux (Entry 3). The low isolated yield of 4 in ethyl
acetate is consistent with the assumption that aluminum triiodide
was partially consumed by competing ester cleavage.
Table 2. Demethylation of methyl vanillate using aluminum
triiodidea
Entry
Temp.
80 oC
50 oC
50 oC
20 oC
15 oC
20 oC
80 oC
Time
5 h
Yield 2 (%)
1
72b
2
2 h
83
3
2.5 h
2 h
81c
82
4
5
10 d
4 d
68d
88e
85f
6
7
2 d
aIsolated yield;
bCompound 5 was isolated in 14% yield;
cCompound 5 was isolated in 11% yield;
dNo further conversion was observed after 1 d, and thus the reaction was
halted after 10 d.
eReaction was carried on a 33 mmol scale.
fEthyl acetate was replaced by methyl acetate.
The method was then extended to methyl isovanillate (6a),
ethyl vanillate (6b) and methyl ferulate (6c), affording the
corresponding demethylated products in good to excellent yields
(Fig. 3). The demethylation of 6c was slower than 3 or 6a, and
took 4 h to reach completion.
Table 1. Demethylation of eugenol by aluminum triiodide
Entry
Solvent
CH3CN
CH3CN
EtOAc
Time
3 min
25 min
18 h
Isolated yield 4 (%)
1
2
3
96a
98
76b
a ~2% unreacted eugenol (3) was recovered.
bConversion was incomplete.
Next, we reasoned that it might be feasible to demethylate
methyl vanillate (1) in the presence of excess ethyl acetate. After
the preparation of aluminum triiodide, acetonitrile was not
removed for the convenience of operation. The selected
conditions are as follows: to a suspension of aluminum triiodide
(5.5 mmol, 1.1 eq.) in acetonitrile (20 mL) was added a solution
of DIC (379 mg, 3.0 mmol, 0.6 eq.) and methyl vanillate (5
mmol) in ethyl acetate (20 mL). The results are summarized in
Table 2. Partial ester cleavage was observed after stirring the
Figure 2. Demethylation of methyl isovanillate 6a, ethyl vanillate 6b
and methyl ferulate 6c
reaction mixture for
protocatechuate (2) as the major product in 72% yield along with
protocatechuic acid (5) as a by-product in 14% yield (Entry 1).
5 h
at 80 oC, affording methyl
A plausible mechanism is depicted in Figure 4. The
deprotonation of phenol 1 proceeds faster than ester cleavage,
and the resultant aluminum phenolate (9) undergoes
Tian, Juan
