Chlorotrimethylsilane as an activating reagent in the samarium-promoted cyclopropanation of allylic and α-allenic alcohols

Full text of DOI:10.1021/jo951365i

2210
J . Org. Chem. 1996, 61, 2210-2214
found that chlorosilanes activate titanium(III) chloride
Ch lor otr im eth ylsila n e a s a n Activa tin g
and even commercial titanium powder in McMurry-type
Rea gen t in th e Sa m a r iu m -P r om oted
Cyclop r op a n a tion of Allylic a n d r-Allen ic
Alcoh ols
carbonyl coupling reactions.8
Resu lts a n d Discu ssion
The cyclopropanation of allylic and R-allenic alcohols
has been investigated using the Sm/CH2I2/TMSCl system.
As is demonstrated in Table 1, TMSCl is a good substi-
tute for HgCl2 for virtually every R-allenic alcohol exam-
ined in the cyclopropanation reaction. However, we were
intrigued by the observation that the diastereoselectivity
changed in the presence of TMSCl. In all the substrates
tested thus far, the presence of TMSCl led to an increase
in the amount of diastereomer 2. In some cases the
change was dramatic while in others a modest effect was
observed. When R is a primary alkyl chain like n-heptyl,
the HgCl2 system slightly favors the formation of dia-
stereomer 3 (Table 1, entries 2 and 11). With the TMSCl
system, however, the selectivity either inverted (Table
1, entry 11) or was approximately 1:1 (Table 1, entries 2
and 3). In substrates where HgCl2 slightly favored the
formation of diastereomer 2, TMSCl increased the selec-
tivity to the extent that the selectivity was synthetically
useful (Table 1, entry 12). For substrates where HgCl2
showed moderate to good selectivity, TMSCl doubled
(Table 1, entry 5) or even tripled (Table 1, entry 10) the
diastereoselectivity. In all cases, complete chemoselec-
tivity toward the allylic olefin was retained.
The reactivity of the TMSCl-modified reagent system
depends on the substitution pattern of the substrate.
When R′ is H, the yield of the reaction is normally
equivalent to or up to 10% higher than the corresponding
HgCl2 reaction series (see Table 1, entries 2, 5, and 12).
However, it appears that reactions using TMSCl are more
sensitive to the substituent R′. When R′ is a methyl
group, the yields dropped 10-15% compared to reactions
using HgCl2 (Table 1, entries 10 and 11), although the
selectivity was improved.
A catalytic amount of TMSCl was found to give the best
results, and cyclopropanation has been performed on up
to a 5-g scale in good yield (Table 1, entry 3) under the
optimized conditions. Doubling the amount of TMSCl or
the use of a large excess did not improve the diastereo-
selectivity or the yield (Table 1, entries 1, 2, 5, and 6).
Unfortunately, 7-10 equiv of Sm and CH2I2 are still
necessary to achieve satisfactory yields, and further work
to overcome this problem is warranted.
It is known that ClCH2I often improves yields in the
cyclopropanation of both allylic4 and R-allenic alcohols,5
presumably due to a longer lifetime of the carbenoid.
Thus, a combination of TMSCl and ClCH2I was examined
in the presence of TMSCl with the expectation that a
better yield and excellent selectivity could be achieved
for hindered substrates. However, the selectivity was
reduced and the temperature at which the reaction
initiated was higher (Table 1, entry 7)!
Mark Lautens* and Yi Ren
Department of Chemistry, University of Toronto,
Toronto, Ontario M5S 1A1, Canada
Received J uly 26, 1995
Hydroxyl-directed transformations are a well-estab-
lished strategy for controlling cyclic and acyclic stereo-
chemistry in organic chemistry.1 Directed cyclopropan-
ations have been known for many years, and both cyclic
and acyclic allylic alcohols undergo diastereoselective
cyclopropanation in the presence of zinc carbenoids.2,3
Samarium carbenoids react with olefins to give cyclopro-
panes at -60 °C which suggests they are significantly
more reactive than those based on zinc.4 The absolute
requirement of a hydroxy group can be exploited in
achieving chemoselective and stereoselective cyclopro-
panation. We recently illustrated the advantages of this
requirement in the selective cyclopropanation of allenic
alcohols which react selectively at the proximal position
to provide methylenecyclopropanes.5
There are several problems which need to be solved in
order to increase the utility of samarium carbene re-
agents. For example, 1 equiv of HgCl2 is typically added
in order to “activate” the Sm. In the absence of this
additive, the reaction is sluggish and often fails to
initiate. The toxicity of HgCl2 suggests that an alterna-
tive method to activate Sm would be very useful. Fur-
thermore, samarium carbenes are often used in moderate
to large excess, and the source of Sm is also known to be
a crucial factor in successful formation of an active
reagent. In this report, we demonstrate that the replace-
ment of HgCl2 by TMSCl addresses both of these short-
comings and in addition influences the diastereoselec-
tivity.
Zinc is known to be activated by TMSCl in a variety of
reactions including the Simmons-Smith cyclopropana-
tion.6,7 Its role appears to be to clean the oxide layer on
the surface of the zinc and also react with trace impuri-
ties (particularly lead) which are present due to the
method of preparing the zinc. Very recently, Fu¨rstner
(1) (a) Hoveyda, A. H.; Evans, D. A.; Fu, G. C. Chem. Rev. 1993, 93,
1307. (b) Hoffmann, R. W. Chem. Rev. 1989, 89, 1841.
(2) (a) Simmons, H. E.; Smith, R. D. J . Am. Chem. Soc. 1958, 80,
5323. (b) Simmons, H. E.; Smith, R. D. J . Am. Chem. Soc. 1959, 81,
4256. (c) Simmons, H. E. Org. React. 1973, 20, 1. For the Furukawa
modification of the Simmons-Smith reaction, see: (d) Furukawa, J .;
Kawabata, N.; Nishimura, J . Tetrahedron Lett. 1966, 28, 3353. (e)
Furukawa, J .; Kawabata, N.; Nishimura, J . Tetrahedron 1968, 24, 53.
(f) Denmark, S. E.; Edwards, J . P.; Wilson, S. R. J . Am. Chem. Soc.
1991, 113, 723. (g) Denmark, S. E.; Edwards, J . P.; Wilson, S. R. J .
Am. Chem. Soc. 1992, 114, 2592.
(3) (a) Winstein, S.; Sonnenberg, J .; DeVries, L. J . Am. Chem. Soc.
1959, 81, 6523. (b) Winstein, S.; Sonnenberg, J . J . Am. Chem. Soc.
1961, 83, 3235.
Cyclopropanation of some simple allylic alcohols was
also briefly studied in the presence of TMSCl to deter-
mine the generality of the reaction. Substrates were
chosen in order to allow a comparison with previous data
obtained using HgCl2.4,9 Once again, TMSCl was a good
substituent for HgCl2 as an activating agent for Sm,
Table 2.
(4) (a) Molander, G. A.; Etter, J . B. J . Org. Chem. 1987, 52, 3942.
(b) Molander, G. A.; Harring, L. S. J . Org. Chem. 1989, 54, 3525.
(5) (a) Lautens, M.; Delanghe, P. H. M. J . Org. Chem. 1993, 58, 5037.
(b) Lautens, M.; Delanghe, P. H. M. J . Am. Chem. Soc. 1994, 116, 8526.
(6) (a) Picotin, G. and Miginiac, P. J . Org. Chem. 1987, 52, 4796.
(b) Picotin, G. and Miginiac, Ph. Tetrahedron Lett. 1987, 28, 4551. (c)
Gawronski, J . K. Tetrahedron Lett. 1984, 25, 2605. (d) Erdik, E.
Tetrahedron 1987, 43, 2203. (e) Knochel, P.; Chang, M.; Yeh, P.; Berk,
S. C.; Talbert, J . J . Org. Chem. 1988, 53, 2392.
(7) Takai, K.; Kakiuchi, T.; Utimoto, K. J . Org. Chem. 1994, 59,
2671.
(8) Fu¨rstner, A.; Hupperts, A. J . Am. Chem. Soc. 1995, 117, 4468.
(9) Lautens, M.; Delanghe, P. H. M. J . Org. Chem. 1992, 57, 798.
0022-3263/96/1961-2210$12.00/0 © 1996 American Chemical Society

Notes
J . Org. Chem., Vol. 61, No. 6, 1996 2211
Ta ble 1. Cyclop r op a n a tion of r-Allen ic Alcoh ols Usin g th e Sm /CH2I2/TMSCl System
2/3c
yieldd (%)
substratea
Smb
CH2I2
TMSCl
TMSCl
HgCl2
TMSCl
HgCl2
e
e
entry
1
2
1a
1a
1a
1a
1b
1b
1b
1b
1b
1c
1d
1e
R ) C7H15, R′ )H, R′′ ) H
R ) C7H15, R′ )H, R′′ ) H
R ) C7H15, R′ )H, R′′ ) H
R ) C7H15, R′ )H, R′′ ) H
R ) C6H11, R′ )H, R′′ ) H
R ) C6H11, R′ )H, R′′ ) H
R ) C6H11, R′ )H, R′′ ) H
R ) C6H11, R′ )H, R′′ ) H
R ) C6H11, R′ )H, R′′ ) H
R ) C6H11, R′ ) CH3, R′′ ) H
R ) C7H15, R′ ) CH3, R′′ ) H
R ) C6H11, R′ ) H, R′′ ) n-Pr
10
10
8
10
10
8
8.8
0.22
0.4
1.29
0.2
0.4
0.2
0.3
0.2
1:2.2
1.1:1
1:1.6
2.5:1
17:1
14.5:1
7:1
30:1
11:1
25:1
1.2:1
4.4:1
1:2.1
1:2.1
81
82
67
62
65
65
80
44
50
52
40
52
73
73
3f
4g
5
10
10
10
10
12
10
10
10
10
10
10
10
10
12
10
10
10
10
9.5:1
9.5:1
9.5:1
64
64
82
6
7h
8g
9i
9.8:1
7.9:1
1:2.1
1.7:1
37
66
52
43
10
11
12
0.28
0.24
0.3
a
b
C7H15 and C6H11 stand for n-heptyl and cyclohexyl, respectively. Unless stated otherwise samarium from Rhoˆne-Poulenc was used.
c Ratios are measured by GC. Yields refer to isolated yield of pure product. e Data are from ref 5 except for entry 9. f Five g scale reaction.
TMSCl as activating reagent, but TMSCl was then removed by decanting TMSCl containing THF solution under N2 followed by washing
samarium metal three times with fresh THF before adding the allenic alcohol and CH2I2. Using ClCH2I instead of CH2I2. Samarium
from Aldrich was used.
d
g
h
i
Ta ble 2. Cyclop r op a n a tion of Allylic Alcoh ols Usin g th e Sm /CH2I2/TMSCl System
a
Unless stated otherwise samarium from Rhoˆne-Poulenc was used. b Ratios are measured by GC except for entry 4 which was determined
by 1H NMR. c Yields refer to isolated yield of pure product(s). Data are from ref 4 except for entry 4 which is from ref 9. e Samarium
d
from Aldrich was used.
Geraniol underwent smooth cyclopropanation using the
TMSCl system to give a single product in excellent yield
with no byproduct resulting from cyclopropanation of the
isolated olefin detected by GC analysis of the crude
reaction mixture (Table 2, entry 1). Cyclohexenol and
bimetallic olefins also gave an excellent yield of cyclo-
propanes (Table 2, entries 2 and 4). An (E)-allylic alcohol
showed a modest improvement in selectivity (Table 2,
entry 3) in accord with the observations with allenic
alcohols, although the stereoselectivity was inferior to
that recently reported by Charette using Zn-based re-
agents.10
We and others have found that the source of the Sm
has a dramatic effect on the efficiency of the cyclopro-
panation. We have investigated two different sources of
Sm and found that TMSCl improves the yields in both
cases (entry 9 vs. 5, Table 1, and entry 1, Table 2). The
data sheets provided by Aldrich and Rhoˆne-Poulenc
(10) (a) Charette, A. B.; Lebel, H. J . Org. Chem. 1995, 60, 2966.
(b) For a recent example of the effect of Lewis acids on cyclopropana-
tions, see: Charette, A. B.; Brochu, C. J . Am. Chem. Soc. 1995, 117,
11367.
(11) (a) Girard, P.; Namy, J . L.; Kagan, H. B. J . Am. Chem. Soc.
1980, 102, 2693. (b) Soderquist, J . A. Aldrichim. Acta, 1991, 24, 15.

2212 J . Org. Chem., Vol. 61, No. 6, 1996
Notes
Ta ble 3. Effects of Differ en t Activa tin g Rea gen ts on th e
Cyclop r op a n a tion
indicate that Sm from Rhoˆne-Poulenc has fewer impuri-
ties than that available from Aldrich. Yttrium, lantha-
num, sodium, calcium, iron, and magnesium are present
in significantly higher amounts in the Aldrich sample,
and we cannot be certain if this is a factor in reducing
the reactivity or if the oxide layer is simply thicker.
While the yields are 15-30% lower when the Sm from
Aldrich was used, TMSCl did lead to a significant
improvement compared to HgCl2 (Table 1, entry 9) or no
activating agent.
The nature of the activating effect of TMSCl was briefly
studied. TMSCl is known to remove the zinc oxide layer
and activate the reaction of Zn with CH2I2 in Simmons-
Smith cyclopropanation.7 In a similar fashion, the
titanium oxide layer could also be removed by TMSCl
treatment.8 29Si NMR of the supernatant of the reaction
mixture of Sm and TMSCl in THF showed two new
signals at 11.7 and 7.8 ppm in addition to a major signal
at 31.5 ppm (TMSCl). The 7.8 ppm signal was assigned
to TMSOTMS which was confirmed by preparation of a
authentic sample formed by mixing TMSCl with water
in THF. The peak at 11.7 ppm may be a TMS-O-X-
containing structure (due to the ring opening of THF in
the presence of Sm?), but we have been unable to
conclusively establish its structure. A control experiment
with Zn in THF in the presence of TMSCl showed signals
at 7.8 and 31.5 ppm. Also, when Sm and TMSCl were
mixed in CDCl3 instead of THF, peaks at 7.8 and 31.5
ppm appeared. We conclude that in analogy to the
results using Zn, TMSCl activates the Sm by reaction
with the surface layer of samarium oxide.
If the role of TMSCl was simply to activate the metal
toward reaction with CH2I2, then Sm pretreated with
TMSCl should initiate the cyclopropanation even if the
TMSCl was removed prior to addition of the substrate.
A portion of Sm was treated with TMSCl in THF for 30
min under a N2 atmosphere. The supernatant was
removed by cannula under a N2 atmosphere, and the
metal was washed three times with fresh THF. A 29Si
NMR spectrum of the final wash solution was free of any
silicon signals. When an R-allenic alcohol and CH2I2 were
added to the Sm and fresh THF at -78 °C, the cyclopro-
panation occurred readily. The yield was lower but
increased selectivity favoring 2 was observed (Table 1,
entries 4 and 8).
The effect on the diastereoselectivity associated with
addition of TMSCl suggests that TMSCl also participates
in the reactive complex. The Houk model was used by
Molander4 to explain the observed diastereoselectivity of
the cyclopropanation of allylic alcohols, and our results
with 1,1-bimetallic allylic alcohols and allenic alcohols
provide further support for this model.5,9 The addition
of TMSCl results in an increase in the product predicted
from the Houk model which suggests that coordination
of a silicon species to the oxygen and/or Sm leads to an
increase in the steric bulk of the metal carbenoid.
Other silyl chlorides were also examined to determine
their effect on the diastereomeric ratio, Table 3. When
TBDMSCl was used instead of TMSCl, a significant
improvement in selectivity was observed for R equals
cyclohexyl (Table 3, entry 9), and the opposite diastere-
omer was favored when R equals n-heptyl (Table 3, entry
3 vs. 4). These results are expected if the silyl group is
involved in the reactive complex. (CH3O)3SiCl was a less
effective activating agent (Table 3, entry 5). Association
of the silyl chloride with the oxygen also suggests that
Lewis acids may influence the diastereoselectivity. Ti-
entry
substrateb
methods
2:3c yieldd (%)
1
2
3
4
5
6
7
8
9
1a R ) C7H15 no activating reagent 2.1:1
34
73
82
78
66
79
64
65
60
59
e
1a R ) C7H15 HgCl2
1:2.1
1:1.1
1.7:1
1:1.1
1.5:1
9.5:1
17:1
1a R ) C7H15 TMSClf
1a R ) C7H15 TBDMSClg
1a R ) C7H15 (CH3O)3SiClg
g
1a R ) C7H15 Ti(O-i-P r )4
e
1b R ) C6H11 HgCl2
1b R ) C6H11 TMSClf
1b R ) C6H11 TBDMSClg
44:1
>50:1
g
10
1b R ) C6H11 Ti(O-i-P r )4
a
b
Samarium from Rhone-Poulenc was used. C7H15 and C6H11
stand for n-heptyl and cyclohexyl, respectively. c Ratios are meas-
d
ured by GC and 400 MHz NMR. Yields refer to isolated yield of
pure product. e Data is from ref 5. f Reaction was carried out with
10 equiv of Sm, 10 equiv of CH2I2, 0.2 equiv of TMSCl in THF at
g
-78 °C to rt. Reaction was carried out with the same procedure
described above except using the activating reagent indicated.
(O-i-Pr)4 was less successful at activating the metal but
gave the highest selectivity toward 2b observed in this
study (>50:1, Table 1, entry 10).
Finally, we note that addition of TMSCl to Sm in the
presence of CH2I2 is an effective procedure for the
preparation of SmI2.11 In a typical procedure Sm was
treated with 2 mol % of TMSCl in THF at room temper-
ature for 30 min prior to addition of CH2I2 at -78 °C.
The typical blue color of SmI2 formed as the reaction was
warmed to -50 °C.
Con clu sion s
We have found that a catalytic amount of TMSCl can
substitute for toxic HgCl2 as an excellent activating agent
in Sm-promoted cyclopropanation of both allylic and
R-allenic alcohols. The yield of the reaction is generally
higher or comparable to the corresponding HgCl2 reac-
tion. The reaction was found to universally favor the
formation of the major diastereomer predicted by the
Houk model. Diastereoselectivity of the reaction is
comparable and in many cases significantly higher than
the corresponding HgCl2 reaction. This activating agent
renders the reaction much more convenient to handle,
work up, and purify. The high reactivity, diastereo-
selectivity and convenience associated with this modifica-
tion may represent a useful improvement in the area of
Sm-promoted cyclopropanation of both allylic and R-
allenic alcohols. In addition, it was found that TMSCl
also activated Sm toward formation of SmI2.
Exp er im en ta l Section
Sta r tin g Ma ter ia ls. The R-allenic alcohols 1a -1e were
prepared according to the literature procedures.5 Geraniol and
2-cyclohexen-1-ol were purchased from Aldrich and used without
further purification. (E)-1-Phenyl-1-hepten-3-ol (8) was pre-
pared by addition of n-BuLi to cinnamaldehyde. 1,1-Bimetallic
allylic alcohol 10 was prepared according to the literature
procedure.9
Gen er a l P r oced u r e for t h e Cyclop r op a n a t ion of r-
Allen ic a lcoh ol w ith Sm /CH2I2 Activa ted by TMSCl. To a
flame-dried round-bottomed flask was added 10 equiv of Sm
metal. The flask was flame dried while being flushed with N2.
After the flask was allowed to cool to rt, THF was added followed

Notes
J . Org. Chem., Vol. 61, No. 6, 1996 2213
by TMSCl (0.2 equiv). The suspension was stirred for 30 min
at rt. The R-allenic alcohol (1 equiv) was dissolved in a small
volume of THF and transferred via cannula to the flask. The
suspension was stirred for 20 min at rt. The flask was cooled
to -78 °C, and CH2I2 was added dropwise. The mixture was
allowed to warm to rt over 3 h and stirred overnight. The
Cyclop r op a n a tion of Ger a n iol (4) Usin g Rh oˆn e-P ou len c
Sm (Ta ble 2, En tr y 1). Allylic alcohol 4 (300 mg, 1.94 mmol)
was cyclopropanated by treatment with Sm (1.46 g, 9.70 mmol),
TMSCl (0.05 mL, 0.39 mmol), and CH2I2 (0.78 mL, 9.7 mmol) in
THF (150 mL) to provide 5 (315 mg, 97%) after flash chroma-
tography using 10% EtOAc in hexanes.4
viscous dark blue reaction mixture was quenched with
a
Cyclop r op a n a tion of Ger a n iol (4) Usin g Ald r ich Sm
(Ta ble 2, En tr y 1). Allylic alcohol 4 (544 mg, 3.53 mmol) was
cyclopropanated by treatment with Sm (Aldrich, 2.65 g, 17.6
mmol), TMSCl (0.09 mL, 0.71 mmol), and CH2I2 (1.28 mL, 17.6
mmol) in THF (150 mL) to provide 5 (442 mg, 75%) after flash
chromatography using 10% EtOAc in hexanes.4
Cyclop r op a n a tion of 2-Cycloh exen -1-ol (6) (Ta ble 2,
En tr y 2). Allylic alcohol 6 (300 mg, 3.05 mmol) was cyclopro-
panated by treatment with Sm (2.30 g, 15.3 mmol), TMSCl
(0.077 mL, 0.61 mmol), and CH2I2 (1.2 mL, 15.3 mmol) in THF
(150 mL) to provide 7 (297 mg, 87%) after flash chromatography
using 10-20% EtOAc in hexanes.4
Cyclopr opan ation of (E)-1-P h en yl-1-h epten -3-ol (8) (Table
2, En tr y 3). Allylic alcohol 8 (300 mg, 1.58 mmol) was
cyclopropanated by treatment with Sm (1.20 g, 7.89 mmol),
TMSCl (0.05 mL, 0.39 mmol), and CH2I2 (0.64 mL, 7.95 mmol)
in THF (150 mL) to provide a 1.9:1 mixture of two diastereomers
as determined by GC analysis. After flash chromatography
using 5% EtOAc in hexanes, 9a (200 mg) and 9b (93 mg) were
isolated in a combined yield of 90%.4
Cyclop r op a n a tion of 10 (Ta ble 2, En tr y 4). Allylic alcohol
10 (500 mg, 1 mmol) was cyclopropanated by treatment with
Sm (1.50 g, 10 mmol), TMSCl (0.025 mL, 0.2 mmol), and CH2I2
(0.8 mL, 10 mmol) in THF (150 mL) to provide 11 as the only
detectable diastereomer by 400 MHz NMR analysis. Flash
chromatography using 5% EtOAc in hexanes provided 11 (417
mg, 81%).9
saturated aqueous K2CO3 solution and extracted three times
with Et2O. The organic layers were collected, washed three
times with brine, dried over anhydrous MgSO4, and filtered.
After concentration of the crude product purification was
executed by flash chromatography on silica gel using 5-7%
(volume) EtOAc in hexanes. Caution: Large-scale reactions (>1
g of allenic alcohol) can be very exothermic. A cooling bath
should be available to control the temperature.
Cyclop r op a n a tion of 1,2-Un d eca d ien -4-ol (1a ) (Ta ble 1,
En tr y 3). Allenic alcohol 1a (5.00 g, 29.8 mmol) was cyclopro-
panated by treatment with Sm (36.0 g, 239 mmol), TMSCl (1.50
mL, 11.8 mmol), and CH2I2 (19.0 mL, 236 mmol) in THF (500
mL). The crude mixture was found to be a 1:1.66 mixture of
two diastereomers as determined by GC analysis. After flash
chromatography, 2a (1.34 g) and 3a (2.33 g) were isolated in a
combined yield of 67%.5
Cyclop r op a n a tion of 1-Cycloh exyl-2,3-bu ta d ien ol (1b)
(Ta ble 1, En tr y 5). Allenic alcohol 1b (300 mg, 1.97 mmol)
was cyclopropanated by treatment with Sm (2.97 g, 19.7 mmol),
TMSCl (0.05 mL, 0.39 mmol), and CH2I2 (1.60 mL, 19.9 mmol)
in THF (150 mL) to provide a 17:1 mixture of two diastereomers
as determined by GC analysis. After flash chromatography, 2b
(200 mg) and 3b (12 mg) were isolated in a combined yield of
65%.5
Cyclop r op a n a tion of 1-Cycloh exyl-2,3-bu ta d ien ol (1b)
Usin g ClCH2I (Ta ble 1, En tr y 7). Allenic alcohol 1b (300 mg,
1.97 mmol) was cyclopropanated by treatment with Sm (3.00 g,
20.0 mmol), TMSCl (0.05 mL, 0.39 mmol), and ClCH2I (1.45 mL,
20.0 mmol) in THF (150 mL). The crude mixture thus obtained
was found to be a 7:1 mixture of two diastereomers as deter-
mined by GC analysis. After flash chromatography, 2b (231 mg)
and 3b (30 mg) were isolated in a combined yield of 80%.5
Cyclop r op a n a tion of 1-Cycloh exyl-2-m eth yl-2,3-bu ta d i-
en ol (1c) (Ta ble 1, En tr y 10). Allenic alcohol 1c (230 mg, 1.38
mmol) was cyclopropanated by treatment with Sm (2.08 g, 13.9
mmol), TMSCl (0.05 mL, 0.39 mmol), and CH2I2 (1.1 mL, 13.9
mmol) in THF (150 mL) to provide a 25:1 mixture of two
diastereomers as determined by GC analysis. After flash
chromatography, 2c (127 mg) and 3c (3 mg) were isolated in a
combined yield of 52%.5
Cyclop r op a n a tion of 1-Cycloh exyl-4-p r op yl-2,3-h ep ta -
d ien ol (1e) (Ta ble 1, En tr y 12). Allenic alcohol 1e (300 mg,
1.27 mmol) was cyclopropanated by treatment with Sm (1.90 g,
12.7 mmol), TMSCl (0.05 mL, 0.39 mmol), and CH2I2 (1.0 mL,
12.7 mmol) in THF (150 mL) to provide a 4.4:1 mixture of two
diastereomers as determined by GC analysis. After flash
chromatography, 2e (135 mg) and 3e (30 mg) were isolated in a
combined yield of 52%.5
Cyclop r op a n a tion of 1-Cycloh exyl-2,3-bu ta d ien ol (1b)
w ith Ald r ich Sm Activa ted by TMSCl (Ta ble 1, En tr y 9).
Allenic alcohol 1b (500 mg, 3.29 mmol) was cyclopropanated by
treatment with Sm (Aldrich, 5.00 g, 32.9 mmol), TMSCl (0.083
mL, 0.658 mmol), and CH2I2 (2.6 mL, 32.9 mmol) in THF (250
mL) to provide an 11:1 mixture of two diastereomers as
determined by GC analysis. After flash chromatography, 2b
(243 mg) and 3b (25 mg) were isolated in a combined yield of
50%.5
Cyclop r op a n a tion of 1-Cycloh exyl-2,3-bu ta d ien ol (1b)
w ith Ald r ich Sm Activa ted by HgCl2 (Ta ble 1, En tr y 9).
Allenic alcohol 1b (300 mg, 1.97 mmol) was cyclopropanated by
treatment with Sm (Aldrich, 2.97 g, 19.74 mmol), HgCl2 (536
mg, 1.97 mmol), and CH2I2 (1.6 mL, 19.7 mmol) in THF (200
mL) to provide a 9.8:1 mixture of two diastereomers determined
by GC analysis. After flash chromatography, 2b (113 mg) and
3b (10 mg) were isolated in a combined yield of 37%.5
Gen er a l P r oced u r e for t h e Cyclop r op a n a t ion of r-
Allen ic Alcoh ol w ith “TMSCl + Wa sh in g” Tech n iqu e.
According to the general procedure, Sm was treated with TMSCl
for 30 min at rt. The clear THF solvent was removed by a
cannula from the flask, under a N2 atmosphere. Fresh THF (20
mL) was then added via
a cannula to the flask and the
suspension was stirred for 5 min at rt before the THF was
removed. This “washing” procedure was repeated three times.
Fresh THF was added. A 29Si NMR of the clear solution showed
no signals. The R-allenic alcohol (1 equiv) was dissolved in a
small volume of THF and transferred via cannula to the flask.
The reaction was then carried out in the usual way.
Cyclop r op a n a tion of 1-Cycloh exyl-2,3-bu ta d ien ol (1b)
Usin g “TMSCl + Wa sh in g” Tech n iqu e (Ta ble 1, En tr y 8).
Allenic alcohol 1b (160 mg, 1.05 mmol) was cyclopropanated by
treatment with Sm (2.00 g, 13.16 mmol), TMSCl (0.04 mL, 0.315
mmol), and CH2I2 (1 mL, 13.16 mmol) in THF (150 mL) to
provide a 30:1 mixture of two diastereomers as determined by
GC analysis. After flash chromatography, 2b (77.6 mg) was
isolated in 44% yield.5
Gen er a l P r oced u r e for t h e Cyclop r op a n a t ion of r-
Allen ic Alcoh ol Usin g Oth er Activa tin g Rea gen ts. General
procedure for the cyclopropanation of R-allenic alcohol with Sm/
CH2I2 activated by TMSCl was followed except that TMSCl was
substituted by other activating reagents.
Cyclop r op a n a tion of 1,2-Un d eca d ien -4-ol (1a ) w ith Sm /
CH2I2 Activa ted by (CH3O)3SiCl (Ta ble 3, En tr y 5). Allenic
alcohol 1a (300 mg, 1.78 mmol) was cyclopropanated by treat-
ment with Sm (2.68 g, 17.86 mmol), (CH3O)3SiCl (0.05 mL, 0.356
mmol), and CH2I2 (1.43 mL, 17.9 mmol) in THF (150 mL) to
provide a 1:1.1 mixture of two diastereomers as determined by
1H NMR (400 MHz) analysis. After flash chromatography, 2a
(105.5 mg) and 3a (107 mg) were isolated in a combined yield of
66%.5
Cyclop r op a n a tion of 1-Cycloh exyl-2,3-bu ta d ien ol (1b)
w ith Sm /CH2I2 Activa ted by TBDMSCl (Ta ble 3, En tr y 9).
Allenic alcohol 1b (300 mg, 1.97 mmol) was cyclopropanated by
treatment with Sm (2.96 g, 19.7 mmol), TBDMSCl (65 mg, 0.43
mmol), and CH2I2 (1.6 mL, 19.7 mmol) in THF (150 mL) to
provide a 44:1 mixture of two diastereomers as determined by
1H NMR (400 MHz) analysis. After flash chromatography, 2b
(195 mg) was isolated in 60% yield.5
Gen er a l P r oced u r e for th e Cyclop r op a n a tion of Allylic
Alcoh ols w ith Sm /CH2I2 Activa ted by TMSCl. Identical to
the general procedure for the cyclopropanation of R-allenic
alcohols except that Sm and CH2I2 were used in 5 equiv excess
to allylic alcohols.
Cyclop r op a n a tion of 1-Cycloh exyl-2,3-bu ta d ien ol (1b)
w ith Sm /CH2I2 Activa ted by Ti(OiP r )4 (Ta ble 3, En tr y 10).
Allenic alcohol 1b (210 mg, 1.38 mmol) was cyclopropanated by

2214 J . Org. Chem., Vol. 61, No. 6, 1996
Notes
treatment with Sm (2.00 g, 13.8 mmol), Ti(OiPr)4 (0.081 mL,
0.28 mmol), and CH2I2 (1.1 mL, 13.8 mmol) in THF (150 mL) to
provide a greater than 50:1 mixture of two diastereomers as
mL) was added followed by TMSCl (0.02 mL, 0.16 mmol). The
suspension was stirred for 30 min at rt. The flask was cooled
to -78 °C, and CH2I2 (0.48 mL, 6.59 mmol) was added dropwise.
The mixture was stirred at -78 to -50 °C for 1 h. The typical
dark blue solution of SmI2 formed.
1
determined by H NMR (400 MHz) analysis. Flash chromatog-
raphy provided 2b (135 mg) in 59% yield.5
Cyclop r op a n a tion of 1,2-Un d eca d ien -4-ol (1a ) Usin g Sm /
CH2I2 w ith ou t Activa tin g Rea gen t (Ta ble 3, En tr y 1). 1a
(130 mg, 0.77 mmol) was cyclopropanated by treatment with Sm
(1.16 g, 7.74 mmol) and CH2I2 (1.60 mL, 19.7 mmol) in THF (150
mL) to provide a 2.1:1 mixture of two diastereomers determined
by 1H NMR (400 MHz) analysis. Flash chromatography pro-
vided 2a (32 mg) and 3a (16.4 mg) in a combined yield of 34%.5
Sm I2 P r ep a r a tion by Rea ctin g Sm w ith CH2I2 Activa ted
by TMSCl. Sm (1.0 g, 6.65 mmol) was flame dried. THF (100
Ack n ow led gm en t. The A. P. Sloan Foundation, the
Natural Science and Engineering Research Council
(NSERC) of Canada, the E. W. R. Steacie Foundation,
the Merck Frosst Center for Therapeutic Research, and
Eli Lilly (USA) are thanked for financial support.
J O951365I