V. D. V. Bodduri et al. / Tetrahedron Letters 56 (2015) 7089–7093
7091
esters and chloro-esters via cleavage of acyclic and cyclic ethers
using various acid chlorides in the presence of iron powder
(325 mesh) as an environmentally benign catalyst. Firstly, we
examined the reaction between tetrahydrofuran (1a) and benzoyl
chloride (2a) to give corresponding 4-chlorobutyl benzoate (3a)
(Table 1). A screen of common solvents established 1,2-dichlor-
oethane (DCE) for the best yields (78–80%) compared to THF, CH3-
CN, and CH2Cl2 (Table 1). It is important to note that 0.2 equiv
amounts of iron gave considerable yields in 7–8 h, which demon-
strates the catalytic action of this metal. Thus, a typical experimen-
tal procedure for this reaction has been standardized to stir acid
chloride, 1.1 equiv of ether and 0.2 equiv of iron powder at 45–
80 °C for 7–8 h in DCE. After the reaction iron powder was easily
separated by simple filtration, and the recovered iron (about 94%
each time)12 was reused for five cycles (yields with recovered iron
from Table 1, entry 6 in %: 78, 72, 76, 77, 74) to ensure the catalytic
activity.
These admirable preliminary results led us to investigate the
scope of our methodology utilizing different ethers 1b–1l and ben-
zoyl chloride (2a) in DCE, and the results are summarized in
Table 2. Most of the acyclic ethers underwent smooth cleavage to
avail the corresponding benzoyl esters (3b, 3d and 3e) in excellent
yields (Table 2, entries 1 and 3–5). tert-Butyl methyl ether (1c) and
diphenyl ether (1g) did not give any product with benzoyl chloride,
presumably due to steric hindrance in the former example and the
lack of orbital approach onto the sp2-hybridized aromatic carbons
in the latter example (Table 2, entries 2 and 6, respectively).
Regioselectivity from the reaction between 2-methyl tetrahydrofu-
ran (1h) and 2a proved instructive (Table 2, entry 7); to preferen-
tially afford 4-chloropentyl benzoate (3h) in 91% yield. After
separation of iron by filtration, the filtrate was again added 1h
and 2a in equivalent amounts and heated at 80 °C to afford the pro-
duct 3h. This process was repeated for four times with the filtrate
of every sequential reaction mixture to afford a consolidated yield
of 83% for 3h. The above result completely contradicted our
assumption for a heterogeneous catalysis, and suggesting homoge-
neous one with dissolved residual iron ion (Fe3+) in the solvent. To
confirm the same we have performed salicylic acid treatment in
aqueous ethanol solution to give Fe(H-sal)3 complex from the reac-
tion mixture. The complex has shown an intensive violet color and
possessed an absorption maximum at 535 nm confirming the Fe3+
ion and was determined by spectrophotometry.12,13 In addition,
iron content in the filtrate of a reaction mixture from 35.55 mmol
of 2a was determined by ICP-OES to realize 415 ppm of iron ion.12
Furthermore to evaluate the reactivity, we have investigated the
reaction between 1h and 2a in presence of catalytic FeCl3 and also
in combination with iron powder (Fe/FeCl3) to afford 3h in 30 min
with similar yields but comparably lower than with iron powder
alone in 8 h (Table 2, entry 7). The short reaction time in presence
of FeCl3 reveals no participation of iron powder when used in com-
bination with it and also showed no improvement in yields, unlike
Zn/ZnCl2 mixture.5a These observations elucidate the efficiency of
our new procedure over past ones.
products from epichlorohydrin may be due to steric effects from
the chlorine atom. Among cyclic ethers, styrene oxide (1l) also
failed to give chloro-esters leaving the starting materials intact
may also be due to steric hindrance (Table 2, entry 11).
To understand the effects of substituent on the reactivity of the
acid chlorides and subsequent product distribution, reactions of 1a
and 1h with various aroyl chlorides 2b–g were explored and the
results are presented in Table 3. Methyl substitution at para-posi-
tion of the aroyl chloride gave almost similar yields for the corre-
sponding adducts (Table 3, entries 1 and 4) compared with those
without methyl substitution. Along with the regioselectivities
observed in the products from 1h, it can also be seen that aroyl
chlorides having electron donating groups (Table 3, entries 4, 5
and 7) gave slightly better yields versus those with electron with-
drawing groups (Table 3, entries 6, 8, and 9), where 4-nitrobenzoyl
chloride (2g) gave 3hg in only 59% yield. Epichlorohydrin on treat-
ment with 4-methoxybenzoyl chloride (2c) afforded a mixture of
symmetric and un-symmetric chloro-benzoates (3kc and 3kc0) in
the ratio 3:2, respectively (Table 3, entry 10). We have also noticed
the formation of 3–6% of byproducts only from 1h, which will be
Table 3
Substituent effects of aroyl chlorides (2b–g) on acylative cleavage of 1a, 1h and 1k
Entry Ether
Aroyl chloridea
Productb
Yieldc
(%)
O
O
O
Cl
2b
O
O
1
75
77
1a
Cl
3ab
O
Cl
O
2
1a
1a
Cl
2c
3ac
O
O
F
O
O
Cl
O
3ad
3
4
5
68
92
93
Cl
2d
F
O
O
2b
O
3hb
O
1h
Cl
O
1h
2c
Cl
3hc
O
F
O
O
3hd
6
7
8
9
1h
1h
1h
1h
2d
88
91
74
59
Cl
O
O
O
O
O
O
Cl
Cl
3he
O
2e
O
O
Cl
O
3hf
O
Cl
2f
O
Cl
Cl
However, 2-(hydroxymethyl)tetrahydrofuran (1i) produced the
simple benzoyate 3i in 97% yield under almost identical reaction
conditions (Table 2, entry 8). Tetrahydropyran (THP, 1j) failed to
react with benzoyl chloride instead only dibenzoyl was formed,
which may be attributed to ring stability (Table 2, entry 9). A com-
petitive reaction between 2a and a mixture of 1a and 1j realized
only 3a and confirmed the profound influence of ring size on reac-
tivity. The only result may not be sufficient to generalize the state-
ment for all acyl chlorides. However, epichlorohydrin (1k) when
subjected to these reaction conditions with benzoyl chloride, gave
a mixture of symmetric and un-symmetric dichlorobenzoates 3k
and 3k0 in the ratio 3:1, respectively, in 95% combined yield
(Table 2, entry 10).14 The low regioselectivity in the formation of
O
Cl
Cl
3hg
2g
O2N
O2N
OBz-pOMe
56d
29d
Cl
Cl
Cl
3kc
10
1k
2c
Cl
OBz-pOMe
3kc'
a
1 equiv of 2, 1.1 equiv of 1 and 0.2 equiv of iron powder were heated at 70 °C in
DCE for 7–8 h.
b
All products were characterized by NMR spectrometry and GC–MS analysis.
Isolated yields based on 2 (except 3–5% byproducts from h).
Heated with 1.0 equiv of iron powder for 3 h.
c
d