Organic & Biomolecular Chemistry
Communication
ation of the tetrahydrofuro[3,2-d]oxazole motif was confirmed due to the over reduction of other functional groups (Table 1,
by an X-ray crystal structure of 6 (CCDC 978589, Fig. 2). The entries 4–6).
X-ray analysis data demonstrate the exclusive formation of cis-
Encouraged by this promising result, a series of arene-1,4-
fused bicycles. The ring conjunction methyl and hydroxy- dione substrates, prepared by the method described in
methyl substituents were syn, as expected, and they were anti Scheme 3, were used to investigate the scope of the method for
to the benzyl hydrogen. So we can draw the conclusion that tetrahydrofuro[3,2-d]oxazole skeleton formation. Intermediate
compound 2k also has the same relative configuration. This 8 was prepared from ethyl acetoacetate/propionylacetate 7 via
interesting result encouraged us to explore the conditions, nitrosation, reduction and acylation. However, the reaction of
scope and mechanism of this novel reaction.
aroylation did not occur when R1 groups were aryl or aromatic
We initiated our studies by choosing 1k as the model sub- heterocyclic substituents. Subsequent replacement of the
strate. First, we screened various Lewis acids, Brønsted acids, bromide group of 9 with 8 gave substrate 1.
and solvents that might promote the heterocyclization. The
To our delight, the arene-1,4-dione substrate 1 smoothly
results indicated that only TiCl4 could promote the reaction to underwent heterocyclization to afford tetrahydrofuro[3,2-d]-
give racemate 2k. The choice of solvents was also crucial for oxazole racemate 2 in good yields and a series of functional
the reaction. The use of 1,2-dichloroethane resulted in a mode- groups including fluoro, chloro, bromo, methoxyl, trifluoro-
rate yield of products (Table 1, entry 7) and toluene gave a methyl, and phenoxyl on the aryl ring were well tolerated
much lower yield (Table 1, entry 8). Dichloromethane was under the optimal reaction conditions (2a–2t in Table 2). It
proved to be a good solvent for this reaction (Table 1, entry 4). was noticed that the substrates containing electron-donating
To further optimize the reaction parameters, the combination groups, such as alkyl, methoxy, and phenoxy, in general gave
of catalyst and reducing agent loading was screened. It was higher yields (2b–2d, 2h, 2k–2o, 2q–2t), while electron
found that the increased ratio of TiCl4 to Et3SiH resulted in deficient substrates with trifluoromethyl and halogen substitu-
higher yields of the product (Table 1, entries 1–4) while more ents gave lower yields (2e–2g, 2i). Moreover, heterocyclic sub-
than one equivalent of Et3SiH decreased the yield presumably strates could also react with TiCl4 and Et3SiH to smoothly
afford the expected products in good yields (2j, 2p). However,
for substrates bearing strong electron-withdrawing groups,
such as nitro, cyano, and pyridyl, no reaction occurred in the
presence of TiCl4 and Et3SiH under the optimized conditions.
Furthermore, we examined the heterocyclization of arene-1,4-
dione containing substitutions on alkyl counterparts (R1, R2).
Various substrates containing an ethyl group reacted smoothly
to provide the desired tetrahydrofuro[3,2-d]oxazoles (2n–2t,
respectively) in good yields. The above results indicate the gen-
Fig. 2 X-ray crystal structure of 6.
erality of the method in preparing tetrahydrofuro[3,2-d]oxa-
zoles in this new and highly efficient way.
To further study the substrate-controlled stereoselectivity of
this reaction, chiral separation of racemate 1k on a Chiralcel
AD-H column was employed to provide the chiral starting
Table 1 Optimization of reaction conditionsa
material (S)- and (R)-1k. The absolute configuration was con-
firmed by circular dichroism spectra. Both (S)-1k and (R)-1k
could be smoothly converted to the corresponding chiral 2k,
respectively, with over 99% ee and dr > 99 : 1 (Scheme 4).
A plausible mechanism to rationalize this transformation is
illustrated in Scheme 5. There are four carbonyls in the sub-
Catalyst
(equiv.)
Reductant
(equiv.)
Entry
Solvent
Yieldb (%)
strate (S)-1a. The real catalyst–reactant complex is not easy to
be captured. The complexation of benzyl carbonyl and acetyl-
1
2
3
4
5
6
7
8
9
TiCl4 (1.1)
TiCl4 (2.1)
TiCl4 (3.1)
TiCl4 (3.8)
TiCl4 (3.8)
TiCl4 (3.8)
TiCl4 (3.8)
TiCl4 (3.8)
TiCl4 (3.8)
TiCl4 (3.8)
Et3SiH (1.1)
Et3SiH (1.1)
Et3SiH (1.1)
Et3SiH (1.1)
Et3SiH (2.1)
Et3SiH (3.8)
Et3SiH (1.1)
Et3SiH (1.1)
Et3SiH (1.1)
Et3SiH (1.1)
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
ClCH2CH2Cl
Toluene
CH3CN
THF
NRc
65
74
81
52
45
68
15
NR
NR
10
a All reactions were performed at room temperature under an argon
atmosphere for 4 h with 0.5 mmol of 1k, the indicated amount of
TiCl4 and Et3SiH in 5.0 mL of the indicated solvent. b Isolated yields.
c No reaction.
Scheme 3 Synthesis of arene-1,4-dione substrates 1.
Org. Biomol. Chem., 2015, 13, 4418–4421 | 4419
This journal is © The Royal Society of Chemistry 2015