Full Paper
within Pfizer corroborated this observation, with this motif
being recognized as difficult to access.
stituted phenol derivative, in a ratio of around 2.5:1 with
a minor product. Intriguingly, the minor product was identified
as 2,3-dihydrobenzofuran 3a resulting from reaction at the
ortho-position and oxetane ring opening. The two heterocyclic
products 2a and 3a were readily separated by flash column
chromatography to give structurally distinct compounds. This
suggested the potential for an unusual substrate controlled
structurally divergent reaction based on the phenol regioselec-
tivity.
Herein we report the synthesis of 3,3-diaryloxetanes, pre-
pared by a lithium-catalyzed Friedel–Crafts reaction of phenols
with readily accessible oxetan-3-ols. We also disclose an effi-
cient synthesis of 3-aryl-3-hydroxymethyl-dihydrobenzofuran
derivatives using the same conditions, the reaction outcome
being dependent on the regioselectivity of the Friedel–Crafts
reaction. This presents a rare example of a divergent reaction
whereby a regiochemical difference leads to structurally dis-
To investigate and optimize the parameters of the reaction,
a design-of-experiments (DOE) approach was adopted, and
[
17]
tinct products, so avoiding closely related isomers.
[22]
was conducted for both Ca and Li catalysts. Temperature,
concentration, catalyst and additive loading, equivalents of nu-
cleophile and reaction time were investigated (see Table 1 for
Results and Discussion
We envisaged that the synthesis of 3,3-diaryloxetanes might
be achieved through a catalytic Friedel–Crafts reaction, directly
displacing a hydroxy group from 3-aryloxetan-3-ols. These sub-
strates would be readily generated by the addition of aryl-
Table 1. Selected optimization for the reaction of 1 with phenol.
[
a]
Entry Catalyst Cat./
Bu NPF
Equiv.
Yield 2a
Yield 3a
[11a]
metal reagents to commercially available oxetan-3-one.
[b]
[b]
4
6
phenol
[%]
[%]
However, this strategy posed interesting questions on the
nature and feasibility of the required carbocationic intermedi-
ate due to the increased p-character of the strained ring
bonds, and the electron-withdrawing nature of the oxygen
atom. To date there were only two transformations to displace
the hydroxy group of tert-oxetanol derivatives, both on aryl de-
rivatives; i) direct OH displacement with diethylaminosulfur tri-
fluoride (DAST) to generate tertiary fluorides and ii) conversion
[mol%]
1
2
3
4
5
6
7
Ca(NTf
2
)
2
5/5
3
3
5
5
5
10
5
5
44
47
63 (61)
19
19
20 (19)
Li(NTf2) 10/5
[
c]
[c]
Li(NTf
2
)
11/5.5
11/-
11/11
11/5.5
Li(NTf
Li(NTf
Li(NTf
2
)
)
)
0
0
6
[
d]
2
2
22
65
51
21
20
(21)
Li(NTf2) 20/10
Li(NTf 11/5.5
[
e]
[c]
[c]
8
2
)
(57)
to the tosylate and reduction with LiAlH to form the 3-aryl ox-
[a] Conditions: 1 (0.25 mmol), CHCl , 0.5m, 408C, 1 h. [b] Yield determined
by H NMR using 1,3,5-trimethoxybenzene as internal standard. [c] Yield
of isolated product in parenthesis. [d] 1 recovered in 46%. [e] 7.5 mmol
4
3
[
11a]
1
etane.
Furthermore, achieving the desired transformation
under catalytic conditions (see Scheme 1), would require the
1
used.
representative results). Comparable results were obtained with
Ca(NTf ) at 5 mol% and Li(NTf ) at 10 mol% loadings (entries 1
2
2
2
and 2). On optimization the best results were achieved using
the Li(NTf ) catalyst (11 mol%) and five equivalents phenol af-
2
fording a combined yield of 83% for the two products after
Scheme 1. Friedel–Crafts alkylation of oxetane 1 with phenol.
1
h (entry 3). The Bu NPF additive was needed, presumably to
4 6
solubilize the catalyst, and a 2:1 ratio of Li(NTf ) to Bu NPF
6
2
4
gave the best yield (entries 3–5). Increasing the amount of
phenol to 10 equivalents did not result in an improvement. A
decrease in the yield of oxetane product was observed on fur-
ther increasing the catalyst loading to 20 mol% (entry 7). Di-
chloroethane, chlorobenzene and hexane were also suitable
solvents but delivered slightly reduced yields of oxetane (52–
58%). Both Li(NTf ) and Bu NPF are inexpensive and easily
activation of the hydroxyl group selectively, over coordination
[
18]
to the Lewis basic oxetane lone pairs and other oxygen con-
taining species, and must prevent ring opening polymerization
[
19]
of the oxetanes under acidic conditions.
We selected aryl-oxetanol 1 for investigation being likely to
stabilize a carbocationic intermediate, and with the potential
to reveal phenol functionality for further derivatization on
benzyl deprotection (Scheme 1). Inspired by recent develop-
ments of Friedel–Crafts reactions displacing hydroxy groups at
2
4
6
handled powders. Increasing the scale of the reaction to
7.5 mmol gave comparable results, providing >1 g of oxetane
2a (57%) along with dihydrobenzofuran 3a (21%, entry 8). Ox-
etane 2a was stored at room temperature in air for
>6 months without any noticeable degradation.
[
20,21]
tertiary centers,
potential catalysts with electron rich aromatic nucleophiles.
We were delighted to observe that both calcium triflimide
we examined a variety of Lewis acids as
[22]
With optimized conditions in hand, the scope of the reaction
to selectively form oxetane products was investigated. Pleas-
ingly, the use of ortho-substituted phenols gave excellent se-
lectivity for the diaryloxetane products, which were formed ex-
clusively (Scheme 2, <3% dihydrobenzofuran for all examples
[
21]
with Bu NPF additive, as reported by Niggemann, and lithi-
4
6
um triflimide/Bu NPF6 successfully formed the 3,3-diaryloxe-
4
tane 2a using excess phenol on heating at 408C in chloroform.
This oxetane product was formed exclusively as the para-sub-
&
&
Chem. Eur. J. 2016, 22, 1 – 7
2
ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!