Competing cyclopropane over epoxide formation from γ-halogeno-δ-hydroxy-ketones

Article abstract of DOI:10.1016/j.tetlet.2010.02.003

Carbocyclization has been selectively achieved over epoxide formation from a γ-chloro-δ-hydroxy-ketone in the presence of a lithiumamide or using a different strategy in which the related silyloxyenol ether bearing an iodine atom at gamma-position and a silyloxy group in delta-position is reacted with tetrabutylammonium fluoride. These approaches take advantage of (i) the poor reactivity of the intermediate β-halogeno lithiumalkoxide first formed in the former case and (ii) the poorer ability of the fluoride ion to desilylate a silyl ether over a silylenol ether.

Full text of DOI:10.1016/j.tetlet.2010.02.003

Tetrahedron Letters 51 (2010) 1942–1944
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Competing cyclopropane over epoxide formation
from
c-halogeno-d-hydroxy-ketones
*
Alain Krief , Adrian Kremer
Department of Chemistry, Facultés Universitaires N.-D. de la Paix, 61 Rue de Bruxelles, Namur B-5000, Belgium
a r t i c l e i n f o
a b s t r a c t
Article history:
Carbocyclization has been selectively achieved over epoxide formation from a
c-chloro-d-hydroxy-
Received 22 December 2009
Revised 29 January 2010
Accepted 1 February 2010
Available online 4 February 2010
ketone in the presence of a lithiumamide or using a different strategy in which the related silyloxyenol
ether bearing an iodine atom at gamma-position and a silyloxy group in delta-position is reacted with
tetrabutylammonium fluoride. These approaches take advantage of (i) the poor reactivity of the interme-
diate b-halogeno lithiumalkoxide first formed in the former case and (ii) the poorer ability of the fluoride
ion to desilylate a silyl ether over a silylenol ether.
Keywords:
Pyrethroids
Ó 2010 Elsevier Ltd. All rights reserved.
Carbocyclization
O-Alkylation versus C-alkylation
Silyl enolates
Epoxide synthesis
We report an original approach to scalemic (1S)-cis-chrysanthe-
mic acid 1a, precursor of (1R)-trans-chrysanthemic acid 1b,1,2 from
scalemic (1S,6R)-2,2,5,5-tetramethyl-7-oxabicyclo[4.1.0]heptan-3-
one 2 (Scheme 1).2 It involves as a key-step intramolecular annela-
tion of (3S,4S)-4-chloro-3-hydroxy-2,2,5,5-tetramethylcyclohexa-
none 3a00, whose synthesis has been achieved by high yielding
epoxide ring opening using beryllium dichloride (Scheme 1).
However the latter reaction occurs with modest selectivity
delivering also the regioisomeric chlorohydrin 3a0 as a by-product.
For this purpose we have devised an original and simple strategy
which allows to recycle 3a0 to the epoxide 2 and at the same time
favors the carbocyclization of the major isomer 3a00 to the bicyclic
[3.1.0] cyclopentanone 4 bearing an exo-hydroxyl group and pre-
cursor of (1S)-cis-chrysanthemic acid 1a (Scheme 1).3 It also takes
advantage of the difference of polarity between the epoxide 2 and
the keto-alcohol 4 which allows to achieve their efficient separa-
tion by chromatographic techniques (SiO2).
(iii) the high propensity of the alcoholate 6a00, to generate 4 either
through an equilibrated process leading to the enolate 7a00
(Scheme 2, entry b; Table 1, entry b) or even better after a further
metallation involving the dianion 8a00 (Scheme 2, entry a; Table 1,
entry b).
We have finally achieved the in situ transformation of the lith-
ium-alcoholate 8a0 to the epoxide 2 by taking advantage of the
lithium–potassium exchange which occurs on reacting potassium
t-butoxide (2 equiv, Scheme 1, entry d, compare to entry c).
The transformation reported above is far from obvious since we
have been unable to apply it to the related bromo- and iodohydrin
mixtures which arise from the ring opening of epoxide 2 by tita-
nium tetrabromide (3b00/3b0: 43/57) or trimethylsilyl iodide4,5 fol-
lowed by hydrolysis of the resulting b-iodotrimethylsilyloxy
ketone since in both cases the epoxide 2 is mainly formed in low
yield indistinctly from both regioisomers.2
We have nevertheless achieved the same goal using a related
strategy (Scheme 3), in which the crucial steps that involve epox-
ide ring opening and enolate formation have been carried out in
a different order. For that purpose we have first synthesized the
‘protected’ silyl enolate 10 from the epoxy-ketone 2 ((i) 1 equiv
LiHMDS, THF, ꢀ78 °C (ii) 2 equiv Me3SiCl, ꢀ78 to 20 °C) and
achieved the epoxide ring opening using trimethylsilyl iodide4,5
(1 equiv Me3SiI, THF, ꢀ78 to 20 °C). Although the latter reaction
is high yielding, it provides an almost one to one mixture of the
two regioisomeric b-silyloxyalkyl iodides 11c (Scheme 3, entry a)
from which the regioisomer 11c00 has been selectively transformed
to the bicyclic [3.1.0] cyclopentanone 4 and its regioisomer 11c0
recycled via the epoxide 2.
It implies the apparent contradictory requirement to find the
proper conditions which favor the transformation of 3a0 to 2 and,
at the same time, avoid its synthesis from 3a00 (Scheme 2).
It relies on (i) the expertise we previously gained on producing
4, resulting from a C-alkylation reaction, from LiTMP and the re-
2a
lated
c
-bromo-d-hydroxy-cyclohexanone 3b00 at the expense of
the epoxide 2 which should have resulted from a competing O-
alkylation reaction, (ii) the extremely poor capacity of b-chloro
lithium-alcoholates 6a to generate 2 (Table 1, entries a, b) and
* Corresponding author. Tel.: +32 474 318181; fax: +32 81 724 536.
E-mail address: alain.krief@fundp.ac.be (A. Krief).
0040-4039/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2010.02.003

A. Krief, A. Kremer / Tetrahedron Letters 51 (2010) 1942–1944
1943
O
OH
1a
S
R
O
S
R
R
R
Cl
O
O
2
ee 94 %
(v)
3a'
20 %
80 %
HO
S
R
O
19 %, ee 94 %
92%
(ii, iii)
(i)
+
O
OH
OMs
86 %
93 %
S
S
HO
Cl
O
2
R
R
(iv)
ee 94 %
O
4
O
5
3a''
84 %
74 %, ee 94 %
ee 94 %
Scheme 1. Reagents and conditions: (i) 5 equiv BeCl2, CH2Cl2, 20 °C; (ii) (a) 2 equiv LiTMP; (b) 2 equiv t-BuOK; (c) H3O+ (see entry d of the table); (iii) separation on
chromatography on SiO2; (iv) 1.2 equiv MsCl CH2Cl2, ꢀ10 °C, 2 h; (v) 6 equiv t-BuOK-3 equiv H2O, THF, 20 °C, 1 h.
OH
HO
Cl
OLi
O
4
7a''
H3O+
OLi
HO
Cl
LiO
Cl
LiO
Cl
O
O
O
O
OLi
OLi
1 eq. LiTMP
THF
1 eq. LiTMP
THF
O
8a''
3a''
6a''
9
+
+
+
Cl
Cl
Cl
Cl
O
H3O+
1 eq. LiTMP
THF
1 eq. LiTMP
THF
HO
LiO
LiO
HO
3a'
8a'
6a'
3a'
(i) t-BuOK
(ii) H3O+
O
O
2
Scheme 2.
This has been efficiently achieved on reacting the 11c mixture
with tetrabutylammonium fluoride (2.2 equiv, THF, ꢀ25 to 20 °C,
0.8 h, then acidic workup, Scheme 3, entry d) taking advantage of
the faster desilylation of the silylenolate moiety over that of the sil-
ylether. It allows the exclusive carbocyclization of 11c00 leading to
4 at the expense of epoxide formation which would have left the
epoxy-ketone 2 instead and the quantitative formation of the latter
from the regiosiomeric 11c0 (Scheme 3, entry d). This proposal has
been assessed by the fact that bicyclic [3.1.0] cyclopentanone 4 has
been produced along with the b-iodotrimethylsilyloxy ketone
(transformed to 3c0 after acidic workup) when using a single equiv-
alent of tetrabutylammonium fluoride instead (Scheme 3, entry c).
In a related approach, the whole process has been carried out in
a single pot using trimethylsilyl iodide to achieve sequentially both
Table 1
3a00/3a0 Reagent
ratio
T (°C) t (h) 4/2/3a00/3a0 4/2 isol.
(equiv/addit.a)
ratio
a
b
c
80/20 LiTMP (1/N)
80/20 LiTMP (1/N)
80/20 LiTMP (2/R)
80/20 LiTMP (2/R) then t-BuOK (2/N) ꢀ25 1.5 80/20/0/0 74/19
ꢀ25
20 0.5 67/8/8/16 66/—
ꢀ25 80/0/0/20 68/—
1
21/0/58/21 20/—
1
d
a
N = addition of the base to 3a; R = addition of 3a to the base.

1944
A. Krief, A. Kremer / Tetrahedron Letters 51 (2010) 1942–1944
O
OSiMe3
(i) 1 eq. LiHMDS, THF, -
1 eq. Me3SiI, THF,
-78 to 20 °C, 1.5 h
O
O
78 °C, 1h (ii) 2 eq.
Me3SiCl, THF, -78 to
20 °C, 1.5h
71 %
2
10
79 %
58
42
a
b
OSiMe3
OSiMe3
+
Me3SiO
I
I
Me3SiO
11c''
37
11c'
63
(i) 1 eq. LiTMP, THF, -78 °C, 1h
(ii) 2 eq. Me3SiI, THF, -78 to 20 °C, 1.5h
OH
O
I
O
11c'' + 11c'
(i) n eq. Bu4NF, THF,
+
O
+
-25 °C, 0.8h (ii) H3O+
(58/42)
HO
O
2
3c'
4
n= 1
52 %
52 %
28 %
00 %
00 %
35 %
c
d
n= 2.2
Scheme 3.
Berlin, 1990; (c) Krief, A. Stereocontrolled Organic Synthesis. In ‘Chemistry for the
21th Century’ Monograph; Trost, B. M., Ed.; International Union of Pure and
Applied Chemistry; Blackwell Scientific, 1994, pp 337–397 and references cited
therein; (d) Krief, A.; Jeanmart, S.; Kremer, A. Bioorg. Med. Chem. 2009, 17, 2555–
2575.
the in situ silylation of the lithium enolate resulting from metalla-
tion of the epoxy-ketone 2 with LiTMP, and the epoxide ring open-
ing (Scheme 3, entry b).
Although this approach is very short, especially when the next
step involving ammonium fluoride is carried in the same pot and
it uses a single equivalent of base as compared to that using the
‘beryllium dichloride’ instead (Scheme 1), the poor 11c00/11c0 ratio
of the b-silyloxyalkyl iodides formed precludes its synthetic use
unless that ratio is greatly improved.
2. (a) Krief, A.; Gondal, H.; Kremer, A. Chem. Commun. 2008, 4753–4755; (b) This
was not expected on the ground of the results previously disclosed.2a
3. Krief, A.; Kremer, A. Synlett 2007, 607–610.
4. (a) Denis, J. N.; Krief, A. Tetrahedron Lett. 1981, 22, 1429–1430; (b) Detty, M. R.;
Seidler, M. D. J. Org. Chem. 1981, 46, 1283–1292; (c) Sakurai, H.; Sasaki, K.;
Hosomi, A. Tetrahedron Lett. 1980, 21, 2329–2332.
5. Trimethylsilyl chloride and bromide are unable to perform the opening of the
epoxide ring of 2 and 9 even when used in excess and under harsher conditions.
References and notes
1. (a) Tessier, J. Chem. Ind. (London) 1984, 199–204.; (b) Naumann, F. In Chemistry
of Plant Protection; Bowers, W. S., Ebing, W., Martin, D., Wegler, R., Eds.; Spriner: