Development of a new and practical route to chiral 3,4-disubstituted cyclopentanones: Asymmetric alkylation and intramolecular cyclopropanation as key C-C bond-forming steps

Article abstract of DOI:10.1021/jo025890a

An efficient and practical asymmetric synthesis of (+)-trans-3-hydroxymethyl-4-(3-fluorophenyl)-cyclopentanone (1) is described. An asymmetric Mo-catalyzed alkylation reaction was used to establish the first stereocenter and a Cu-catalyzed intramolecular diastereoselective cyclopropanation reaction was used to set the second stereocenter. The last step involved a one-pot ring-opening/deprotection/hydrolysis/decarboxylation sequence that furnished the desired product in good yield.

Full text of DOI:10.1021/jo025890a

Develop m en t of a New a n d P r a ctica l Rou te to Ch ir a l
3,4-Disu bstitu ted Cyclop en ta n on es: Asym m etr ic Alk yla tion a n d
In tr a m olecu la r Cyclop r op a n a tion a s Key C-C Bon d -F or m in g
Step s
Michael Palucki,* J oann M. Um, Nobuyoshi Yasuda, David A. Conlon, Fuh-Rong Tsay,
Frederick W. Hartner, Yi Hsiao, Benjamin Marcune, Sandor Karady, David L. Hughes,
Peter G. Dormer, and Paul J . Reider
Department of Process Research, Merck Research Laboratories, P.O. Box 2000,
Rahway, New J ersey 07065-0900
michael_palucki@merck.com
Received May 1, 2002
An efficient and practical asymmetric synthesis of (+)-trans-3-hydroxymethyl-4-(3-fluorophenyl)-
cyclopentanone (1) is described. An asymmetric Mo-catalyzed alkylation reaction was used to
establish the first stereocenter and a Cu-catalyzed intramolecular diastereoselective cyclopro-
panation reaction was used to set the second stereocenter. The last step involved a one-pot ring-
opening/deprotection/hydrolysis/decarboxylation sequence that furnished the desired product in good
yield.
In tr od u ction
available cyclopentanone precursor is not straightforward
given that the carbons at the 2- and 5-position are
significantly more reactive than the carbons at the 3- and
4-positions. For this reason, successful synthetic routes
have typically employed either an intermolecular cyclo-
addition reaction of two functionalized components3 or
an intramolecular cyclization reaction of a properly
substituted acyclic precursor. In the latter approach,
methods used to effect intramolecular cyclization include
classical methods such as a Dieckman cyclization4 and
an aldol reaction,5 as well as newer methods such as a
rhodium-catalyzed cyclization of 3,4-disubstituted 4-pen-
tenals,6 a rhodium-catalyzed cyclization of carbenoid
precursors,7 a ruthenium-catalyzed three-component cou-
pling followed by a Nozaki-Kishi reaction,8 and a Pd-
catalzyed addition of 1,3-diones to unactivated olefins.9
Our approach in the synthesis of 1 was based on
identifying a properly substituted substrate that can
potentially undergo cyclization by various means with
minimal chemical steps.
Substituted cyclopentanones are found in numerous
pharmacologically active compounds,1 and are versatile
synthetic intermediates for the preparation of target
molecules containing a substituted cyclopentane subunit.2
In the latter context, cyclopentanone 1 was recently
identified as an advanced key intermediate in the syn-
thesis of an investigational new drug candidate at Merck
& Co. Accordingly, we required a rapid and efficient
asymmetric synthesis of 1 from readily available starting
materials in order to furnish necessary quantities of our
investigational drug for preclinical and clinical studies.
(3) For reviews of [3+2] cycloaddition reactions, see: Little, R. D.
In Comprehensive Organic Synthesis; Trost, B. M., Fleming, I., Eds.;
Pergamon: Oxford, UK, 1991; Vol. 5, Chapter 3.1, pp 239-270. (b)
Chan, D. M. T. In Comprehensive Organic Synthesis; Trost, B. M.,
Fleming, I., Eds.; Pergamon: Oxford, UK, 1991; Vol. 5, Chapter 3.2,
pp 271-314.
The asymmetric synthesis of optically active trans-3,4-
disubstituted cyclopentanones directly from a readily
(1) (a) Elks, J .; Ganellin, C. R. In Dictionary of Drugs: Chemical
Data, Structures and Bibliographies, 1st ed.; Chapman and Hall: New
York, 1990. For reviews see: (b) Paquette, L. A. Top. Curr. Chem. 1984,
119, 1. (c) Mehta, G.; Srikrishna, A. Chem. Rev. 1997, 97, 671-719.
(d) Hudlicky, T.; Rulin, F.; Lovelace, T. C.; Reed, J . W. In Studies in
Natural Products Chemistry; Atta-ur Rahman, Ed.; Elsevier: Oxford,
UK, 1989; Vol. 3, Part B, pp 3-72.
(2) For reviews of cyclopentane synthesis see: (a) Ramaiah, M.
Synthesis 1984, 529-570. (b) Hudlicky, T.; Price, J . D. Chem Rev. 1989,
89, 1467-1486. (c) Trost, B. M. Angew. Chem., Int. Ed. Engl. 1986,
25, 1. (d) Corey, E. J . Angew. Chem., Int. Ed. Engl. 1991, 30, 455-
612. For examples in prostaglandin synthesis see: (e) Noyori, R. In
Asymmetric Catalysis in Organic Synthesis; J ohn Wiley & Sons: New
York, 1994; pp 298-322.
(4) Boeckman, R. K., J r.; Napier. J . J .; Thomas, E. W.; Sato, R. I. J .
Org. Chem. 1983, 48, 4152-4154.
(5) (a) Wenkert, E.; Greenberg, R. S.; Raju, M. S. J . Org. Chem. 1985,
50, 4681-4685. (b) Pecuniosos, A.; Menicagli, R. J . Org. Chem. 1988,
53, 2614-2617.
(6) Wu, X.-M.; Funakoshi, K.; Sakai, K. Tetrahedron Lett. 1993, 34,
5927-5930.
(7) (a) Taber, D. F.; Ruckle, R. E., J r. Tetrahedron Lett. 1985, 26,
3059-3062. (b) Taber, D. F.; Hoerrner, R. S. J . Org. Chem. 1992, 57,
441-447.
(8) Trost, B. M.; Pinkerton, A. B. J . Org. Chem. 2001, 66, 7714-
7722.
10.1021/jo025890a CCC: $22.00 © 2002 American Chemical Society
Published on Web 07/19/2002
5508
J . Org. Chem. 2002, 67, 5508-5516

New Route to 3,4-Disubstituted Cyclopentanones
SCHEME 1
SCHEME 2
Acyclic precursor 2 was identified as an initial target
because it contained the desired substituted functionality
and because a number of synthetic pathways were
envisioned that could be used to synthesize 1 from 2
(Scheme 1). For example, pathway A, shown in Scheme
1, utilizes a transition metal-catalyzed C-H insertion as
the key cyclization step. Pathway B utilizes an intra-
molecular epoxide ring opening to form the desired five-
membered ring, and pathway C utilizes an intramolec-
ular cyclopropanation reaction to provide the core
cyclopentanone ring system.
Preliminary results from our efforts to use either routes
A or B were disappointing.10 Our attempt to effect a Rh-
catalyzed intramolecular C-H insertion of intermediate
4 to give 5 provided, instead, a complex mixture of
byproducts. The intramolecular epoxide ring opening of
intermediate 6 afforded the O-alkylated ring-opened
product 11 as the major product, via alkylation of the
of several highly selective reactions including a Mo-
catalyzed asymmetric alkylation, a Cu-catalyzed in-
tramolecular cyclopropanation, and a one-pot, four-step
reaction sequence involving a cyclopropane ring-opening,
deprotection, hydrolysis, and decarboxylation. The results
and efforts used therein are discussed below.
Resu lts a n d Discu ssion
Our retrosynthetic approach toward the synthesis of
2 is shown in Scheme 2. We envisioned that 2 could be
obtained from the chiral acid 12 via chain extension using
the protocol of Masamune. Chiral acid 12 can be derived
from the malonate derivative 13. The synthesis of 13
would be the key stereodefining step and would be
accomplished via molybdenum-catalyzed asymmetric al-
lylic alkylation of carbonate 14 with dimethyl malonate.
Asym m etr ic Alk yla tion Rea ction . Recently, signifi-
cant efforts have been focused on the development of a
transition metal-catalyzed asymmetric allylic alkylation
of unsymmetrically substituted allylic derivatives for the
preparation of optically active branched regioisomers in
high enantioselectivity and regioselectivity.11 In this field,
Trost et al. have developed one of the most promising
methods involving the use of (EtCN)3Mo(CO)3/ligand 15
(Scheme 3). The results published by Trost using this
system suggested to us that the molybdenum-catalyzed
ketone oxygen. By comparison, we found that utilization
of route C eventually provided the desired cyclopen-
tanone product 1. The route involved the development
(11) (a) Lloyd-J ones, G. C.; Pflatz, A. Angew. Chem., Int. Ed. Engl.
1995, 34, 462-464. (b) J anssen, J . P.; Helmchen, G. Tetrahedron Lett.
1997, 38, 8025-8026. (c) Trost, B. M.; Hachiya, I. J . Am. Chem. Soc.
1998, 120, 1104-1105. (d) Evans, P. A.; Nelson, J . D. J . Am. Chem.
Soc. 1998, 120, 5581-5582. (e) Glorius, F.; Pfaltz, A. Org. Lett. 1999,
1, 141-144.
(9) Pei, T.; Widenhoefer, R. A. J . Am. Chem. Soc. 2001, 123, 11290-
11291.
(10) These preliminary results (routes A and B) will be published
elsewhere.
J . Org. Chem, Vol. 67, No. 16, 2002 5509

Palucki et al.
SCHEME 3
SCHEME 4
SCHEME 5
asymmetric alkylation methodology would be well-suited
for the preparation of the desired product 13 in good
enantio- and regioselectivity from an allyl carbonate
precursor.
The use of linear allyl carbonates in the molybdenum-
catalyzed asymmetric alkylation reaction has been shown
to produce slightly better enantioselectivity and regio-
selectivity than branched racemic carbonates.11c,e Con-
sequently, our initial efforts focused on the synthesis of
the linear carbonate. Starting with 3-fluorocinnamic acid
16, we sought to carry out a two-step procedure involving
the reduction of the acid to the alcohol 17, followed by
carbonate formation via reaction with methyl chlorofor-
mate to provide 19, Scheme 4. However, our attempts to
reduce the acid to the alcohol were accompanied by
reduction of the double bond to produce 18. With use of
either NaBH4/I2 or LiAlH4, a substantial amount of the
over-reduced product was observed along with multiple
unidentified byproducts as determined by HPLC.12 These
results coupled with the fact that the branched allyl
carbonate 21 can be made in one pot, vide infra, led us
to abandon the use of the linear carbonate as a possible
starting allyl carbonate.
For the synthesis of the racemic branched allyl carbon-
ate 21, we focused on the development of a one-pot
method consisting of vinyl Grignard addition to aldehyde
20, followed by trapping of the magnesium alkoxide
intermediate with methyl chloroformate, Scheme 5. With
use of vinylmagnesium bromide, benzylic carbonate 21
was formed in up to 15%, suggesting that reduction of
aldehyde 20 was competitive. Transmetalation of the
vinylmagnesium bromide from magnesium to zinc via
addition of ZnCl2 suppressed the reduction pathway to
10%. However, it was found that using vinylmagnesium
chloride in place of the bromide reduced the formation
of impurity 22 to <1%, and that transmetalation to zinc
was unnecessary. By using toluene as solvent for ease of
workup, carbonate 21 was formed in 76-85% yield in one
pot from aldehyde 20.
Results from preliminary studies on the asymmetric
alkylation of carbonate 21 with sodiodimethyl malonate
using the Trost catalyst system were promising: 75-90%
yield, 88-96% ee, regioselectivity of 6-20:1 branched:
linear. However, it became evident that direct imple-
mentation of the reaction protocol to large-scale produc-
tion would not be straightforward due to the limited
availability of the precatalyst (EtCN)3Mo(CO)3 and ligand
15, both of which were not commercially available during
this study.13 Consequently, our efforts turned to develop-
ing a simple procedure for the preparation of ligand 15,
and the use of a more readily available Mo precatalyst.
These studies, which have been communicated recently,
culminated in an improved and practical synthesis of
ligand 15, as well as the use of the readily available, cost-
effective, and air-stable Mo(CO)6 as the molybdenum
precatalyst.14 These improvements allowed us to perform
the reaction on multikilo scale numerous times, while
affording the desired branched product in 84-91% yield
and in 96-97% ee (Scheme 6).15
(13) Strem Chemical Company currently sells ligand 15.
(14) (a) The development of the Mo-catalyzed asymmetric reaction
employing Mo(CO)6 as precatalyst to prepare compound 13 including
optimization and other experimental results are described in the
following communication: Palucki, M.; Um, J . M.; Conlon, D. A.;
Yasuda, N.; Hughes, D. L.; Mao, B.; Wang, J .; Reider, P. J . Adv. Synth.
Catal. 2001, 343, 46-50. (b) Conlon, D. A.; Yasuda, N. Adv. Synth.
Catal. 2001, 343, 137-138.
(12) (a) Bhaskar Kanth, J . V.; Periasamy, M. J . Org. Chem. 1991,
56, 5964-5965. (b) Nystrom, R. F.; Brown, W. G. J . Am. Chem. Soc.
1947, 69, 2548-2549.
5510 J . Org. Chem., Vol. 67, No. 16, 2002

New Route to 3,4-Disubstituted Cyclopentanones
SCHEME 6
SCHEME 7
P r ep a r a tion of th e r-Dia zo-â-k etoester In ter m e-
d ia te 24. Solvolysis/decarboxylation of 13 to provide
monoacid 12 was first accomplished by heating a dioxane
solution of 13 at reflux in the presence of 6 N HCl (aq),
Scheme 7. However, the yields obtained with these
conditions varied to a large degree (48-80%), due to a
competitive acid-catalyzed lactonization pathway that
afforded lactone 28. Intermediates that were observed
during the course of the reaction include the diacid 25,
monoacid-monoester 26, and the monoester 27.
formation of 28, but starting material conversion was
very slow. In contrast, base hydrolysis with dilute NaOH
and no cosolvent was found to be fast and effective in
producing diacid 25. The subsequent decarboxylation was
carried out in the same reaction vessel by adjusting the
pH to 1 using concentrated HCl and heating for 12 h to
provide 12 in good yield. The major byproduct of the
reaction was methyl ester 27, which increased propor-
tionally at lower pH values. Formation of 27 was sup-
pressed by the removal of MeOH after the base hydrolysis
portion of the sequence. Isolation and purification of 12
was accomplished via salt formation with the amine base
(S)-sec-phenethylamine.16 The stereochemical integrity of
the stereogenic benzylic carbon was maintained through-
out these studies.
Chain extension of 12 using the protocol of Masamune
provided the â-ketoester 2 in excellent yield.17 Formation
of the diazo-compound 24 was accomplished in excellent
yield via reaction of â-ketoester 2 with 4-acetomidoben-
zenesulfonyl azide.18 Because of the general thermal
instability of R-diazo compounds the product was not
isolated, but rather used directly as a solution in the
subsequent diastereoselective intramolecular cyclopro-
panation reaction.
Because of the problems encountered in the one-pot
hydrolysis/decarboxylation of 13, efforts to develop a
stepwise protocol were initiated. Acid hydrolysis was first
studied; however, the results were discouraging. High
concentrations of aq HCl in the presence of an organic
cosolvent (THF, 1,4-dioxane, and DMF) resulted in fast
starting material conversion, but provided lactone 28 as
the major product. Low concentrations of acid minimized
Dia ster eoselective In tr a m olecu la r Cyclop r op a -
n a tion Rea ction . Stereoselective cyclopropanation reac-
(15) The workup of the asymmetric alkylation reaction was com-
prised of a single water wash followed by filtering the toluene solution
through a small silica gel pad to remove any molybdenum residues as
well as ligand 15. The resulting filtrate was concentrated and used
directly in the subsequent reaction. Crystallization of 13 from toluene
was accomplished with an upgrade in ee (>99% ee and >99% chemical
purity); however, a 20% loss of 13 in the filtrate was incurred.
(16) On laboratory scale (1 g), an upgrade in ee of 12 was observed
with salt formation/salt break of 17 with (S)-sec-phenethylamine;
however, on large scale (1 kg), no increase in ee was observed.
(17) Brooks, D. W.; Lu, L. D.-L.; Masamune, S. Angew. Chem., Int.
Ed. Engl. 1979, 18, 72-74.
(18) Davies, H. M. L.; Cantrell, W. R.; Romines, D. R.; Baum, J . S.
Org. Synth. 1993, 70, 93-100.
J . Org. Chem, Vol. 67, No. 16, 2002 5511

Palucki et al.
SCHEME 8
TABLE 1. In tr a m olecu la r Cyclop r op a n a tion Rea ction
of 24 w ith Ach ir a l Rh a n d Cu Ca ta lystsa
tions of R-diazocarbonyls are catalyzed by a large number
of transition metal complexes, of which Cu and Rh
complexes have emerged as the most prominent cata-
lysts.19,20 In our synthesis, we required a diastereoselec-
tive intramolecular cyclopropanation of 24 to provide the
desired exo cyclopropane product 29 in preference to the
endo cyclopropane product 30, Scheme 8. Although we
were unable to find literature precedent for achieving the
desired diastereoselectivity, we felt that through some
catalyst screening, we could identify a catalyst that would
effect the reaction with the required selectivity.21 At the
minimum, we hoped that the use of a chiral catalyst
would potentially provide match/mismatch interactions
between itself and the chiral substrate to effect the
desired reaction in high diastereoselectivity.
Our first attempt at screening transition metal cata-
lysts involved the use of achiral rhodium dimer com-
plexes. Intramolecular cyclopropanation reactions of 24
catalyzed by rhodium dimer complexes were typically
characterized by complete starting material conversion
with minimal byproduct formation. Although these reac-
tions were clean and fast, the undesired endo isomer 30
was obtained as the major product (Table 1, entries 1-3).
The use of achiral copper catalysts in the intramolecu-
lar cyclopropanation reaction provided better selectivities
overall compared to rhodium dimer complexes (Table 1,
entries 4-11). In contrast to reactions catalyzed by the
rhodium dimer complexes, cyclopropanation reactions
with copper catalyst were very sluggish at room temper-
ature and required heating to 75 °C in 1,2-dichloroethane
to achieve reasonable rates. A comparison of various
copper sources shows that copper catalysts containing
moderately coordinating ligands (Table 1, entries 4-8)
exhibited lower diastereoselectivty than copper catalysts
containing weakly or noncoordinating ligands (Table 1,
entries 9 and 10). Indeed, the most selective catalyst was
generated from reaction of CuCl with AgOTf to produce
CuOTf.
temp conversion
(°C)
entry
catalyst
Rh2(OAc)4
Rh2(Octa)4
Rh2(CAP)4
CuCl
Cu(acac)2
CuOAc
Cu2O
CuSCN
Cu(OTf)2
[(CH3CN)4Cu]PF6 1,2-DCE
CuCl/AgOTf 1,2-DCE
solvent
(%)c
29:30d
1
2
CD2Cl2
rt
100 (90)
100 (87)
95
43:57
33:67
33:67
50:50
54:46
46:54
42:58
70:30
77:23
83:17
85:15
CD2Cl2
rt
3
CD2Cl2
rt
4
1,2-DCE
1,2-DCE
1,2-DCE
1,2-DCE
1,2-DCE
1,2-DCE
75
75
75
75
75
75
75
75
99
5
37
6
88
7
3
8
10
9
100 (92)
100 (98)
100 (89)
10
11b
a
Entries 1-3, 10, and 11 were performed with 1 mol % of
catalyst and at 0.1 M 24. Entries 4-9 were performed with 5-15
b
mol % of catalyst and at 0.2 M 24. A ratio of 1:1.2 was used for
CuCl/AgOTf. c The numbers in parentheses are isolated yields of
d
combined 29/30. The ratio of 29/30 was determined by HPLC
by using a Zorbax SB--C18 column 4.6 × 250 mm.
possible. Both chiral rhodium and copper catalysts were
examined and the results of these studies are shown in
Table 2.
As shown in Table 2, the use of chiral catalysts did
not improve the diastereoselectivitiy for the intramolecu-
lar cyclopropanation of 24. Four commercially available
chiral Rh catalysts were examined with the idea of
improving the diastereoselectivity through match/mis-
match interactions with the chiral substrate.22 Unfortu-
nately, all of the Rh catalysts displayed only modest
diastereoselectivity. In contrast to the achiral Rh catalyst,
reactions with these catalysts were extremely slow at
room temperature and required heating to 60 °C to
achieve reasonable rates of reaction. Similarly, the use
of chiral copper complexes of bis-oxazolines provided only
modest diastereoselectivity.23 These results coupled with
the results shown in Table 2 with achiral copper com-
plexes suggest that a ligand free copper catalysts is best
for preferentially forming the desired product 29 over 30.
Transition metals other than copper or rhodium were
also examined, including Ni(acac)2, NiCl2, CoCl2, Co-
(acac)3, Yb(OTf)3, and [RuCl2(p-cymene)]2; however, ex-
cept for [RuCl2(p-cymene)]2, the starting material con-
version with these catalysts was less than 5%. The use
of [RuCl2(p-cymene)]2 gave complete conversion of the
starting material, but with a 29:30 ratio of 35:65.
The catalysts that provide the highest desired dia-
stereoselectivity for the intramolecular cyclopropanation
of 24 are [(CH3CN)4Cu]PF6 and CuCl/AgOTf. In both
cases, the reactions can be carried out at low catalyst
loading. In the presence of strongly coordinating ligands,
the diastereselectivity is lower and the rate of reaction
is slower.
Because 24 is chiral, the potential for identifying
match/mismatch interactions with chiral catalysts is
(19) For a review on asymmetric cyclopropanations with Rh cata-
lysts, see: (a) Lydon, K. M.; McKervey, M. A. In Comprehensive
Asymmetric Catalysis II; J acobsen, E. N., Pfaltz, A., Yamamoto, H.,
Eds.; Springer-Verlag: Berlin, Germany, 1999; Chapter 16.2, pp 539-
580. For a review on asymmetric cyclopropanations with Cu catalysts
see: (b) Pfaltz, A. In Comprehensive Asymmetric Catalysis II; J acobsen,
E. N., Pfaltz, A., Yamamoto, H., Eds.; Springer-Verlag: Berlin,
Germany, 1999; Chapter 16.1, pp 513-538. For a review on asymmetric
catalytic metal carbene transformation, see: (c) Doyle, M. P.; Forbes,
D. C. Chem. Rev. 1998, 98, 911-935.
(20) For reviews on asymmetric synthesis with diazo compounds,
see: (a) Calter, M. A. Current Org. Chem. 1997, 1, 37-70. (b) Ye, T.;
McKervey, M. A. Chem. Rev. 1994, 94, 1091-1160. (c) Adams, J .;
Spero, D. Tetrahedron 1991, 47, 1765-1808.
(21) For examples of diastereoselective intramolecular cyclopropa-
nation reactions, see: (a) Doyle, M. P.; Dyatkin, A. B.; Kalinin, A. V.;
Ruppar, D. A. J . Am. Chem. Soc. 1995, 117, 11022-11022. (b) Martin,
S. F.; Spaller, M. R.; Liras, S.; Hartmann, B. J . Am. Chem. Soc. 1994,
116, 4493-4494.
(22) Doyle, M. P. Aldrichim. Acta 1996, 29, 3-11.
(23) For a review on bis-oxazolines in asymmetric synthesis see:
Ghosh, A. K.; Mathivanan, P.; Cappiello, J . Tetrahedron Asymmetry.
1998, 9, 1-45.
5512 J . Org. Chem., Vol. 67, No. 16, 2002

New Route to 3,4-Disubstituted Cyclopentanones
TABLE 2. Cyclop r op a n a tion of 24 w ith Ch ir a l Ca ta lystsa
conversion
(%)
entry
catalyst
ligand
29:30b
1
2
3
4
5
6
7
8
Rh2(5S-MEPY)4
Rh2(5R-MEPY)4
Rh2(4S-MEOX)4
Rh2(4S-MPPIM)4
[(CH3CN)4Cu]PF6
[(CH3CN)4Cu]PF6
[(CH3CN)4Cu]PF6
[(CH3CN)4Cu]PF6
71
90
77
42:58
52:48
48:52
30:70
60:40
54:46
59:41
52:48
44
31
32
33
34
100
100
100
100
a
For entries 1-4, reactions were performed with 5 mol % of catalyst at 0.1 M 24 in 1,2-dichloroethane. For entries 5-8, reactions were
b
performed with 20 mol of Cu and 22 mol % of ligand at 0.1 M 24 in 1,2-dichloroethane. The ratio of 29/30 was determined by HPLC by
using a Zorbax SB-C18 column 4.6 × 250 mm.
In addition, based on the trans orientation of the
substituents on the cyclopentane ring product, one could
predict that 29 would be the major diastereomeric
product. However, Molecular Mechanics calculation show
a ground-state energy difference between 29 and 30 to
be only 1.3 kcal.24 Additionally, cyclopropanation reac-
tions are thought to occur through an early transition
state, thus minimizing the influence of product ground-
state energy differences on diastereoselectivity. Further
studies are clearly needed to determine if the resulting
diastereoselectivity with and without added ligands is
due to nonbonding interaction in the transition state or
to different mechanistic pathways.25
slowly added over 4 h to a catalyst solution at 75 °C. By
using this protocol, a minimal buildup of 24 occurs during
the course of the reaction. Upon reaction completion, the
mixture was concentrated, filtered through a small pad
of silica gel, and used directly in the next reaction.
R in g Op en in g/Dep r ot ect ion /H yd r olysis/Deca r -
boxyla tion . Our initial approach for forming 1 from a
mixture of 29/30 was to develop a two-step process. The
first step would involve ring opening the cyclopropane
ring with sodium or potassium acetate in a polar aprotic
solvent such as DMF or DMSO to yield 35 (Scheme 9).26
The second step would involve removal of the acetate and
hydrolysis/decarboxylation of the methyl ester to furnish
1. In our very first attempt at ring opening 29/30 (5 equiv
of NaOAc in DMF at 100 °C), we were pleased to find
that 1 was formed in 30% yield. Most interesting was
that the cis product derived from ring opening of 30 was
not observed. Although this result affirms that we can
directly obtain 1 from 29/30 in a one-pot procedure, we
quickly found that product yields were not always
reproducible.
Because of the general thermal instability of diazo
compounds, a solution of 24 in 1,2-dichloroethane was
(24) Molecular Mechanics calculations performed on PC SPARTAN
PRO 6.0.6.
(25) Two different mechanisms have been proposed for the formation
of the cyclopropane ring: (1) Mechanism A, which proceeds via a
concerted pathway, and (2) Mechanism B, which proceeds via a two-
step pathway involving a metallocyclobutane followed by reductive
elimination to the cyclopropane product. Mechansim B may be favored
by catalyst systems devoid of ligands, whereas mechanism A may be
relatively favored by the sterically demanding chiral ligands.
Further studies in the optimization of this reaction led
to two key advancements in the reaction protocol. The
first was the addition of 5 equiv. of AcOH, which
consistently effected complete conversion of the starting
cyclopropane. The second involved changing the workup
of the reaction. Although efforts to use anhydrous sol-
vents and reagents were undertaken, clearly the presence
of adventitious water facilitated the deprotection/hy-
(26) (a) Tanimori, S.; Tsubota, M.; He, M.; Nakayama, M. Synth.
Commun. 1997, 27, 2371-2378. (b) Tanimori, S.; Tsubota, M.; He, M.;
Nakayama, M. Biosci. Biotechnol. Biochem. 1995, 61, 2091-2093.
J . Org. Chem, Vol. 67, No. 16, 2002 5513

Palucki et al.
SCHEME 9
SCHEME 10
drolysis/decarboxylation sequence. Consequently, a two-
step, one-pot protocol was developed in which a solution
of 29/30, HOAc, and NaOAc in DMF was heated at 100
°C for 12-15 h followed sequentially by cooling to room
temperature, adding 5 N NaOH until pH >11, and
heating at 60 °C for 30 min to ensure complete decar-
boxylation. With use of this procedure, the desired
cyclopentanone 1 was reproducibly obtained in 50-60%
yield.
opening of the cyclopropane ring with acetate to provide
intermediate 35 (Scheme 10). A HPLC-MS analysis of a
reaction mixture sample prior to the addition of DMF and
NaOH (aq) gave a mass of m/z 308 for the major peak in
the chromatogram. The next two steps involve removal
of the acetate group and hydrolysis of the methyl ester
to give either 36 or lactone 37. We have synthesized 37
via a different route and have observed that 37 can be
converted to 1 under the workup conditions.27 In addition,
HPLC analysis of the reaction mixture shows that a peak
appears at the retention time where authentic 37 elutes
and subsequently disappears. The final step of the
sequence is the decarboxylation of 36 to provide 1.
Starting with a 82:18 mixture of 29/30, a mixture of
the trans cyclopentanone product 1 and its epimer is
obtained in a 94:6 ratio. We propose that ring opening of
the endo cyclopropane starting material 30 is slow
relative to ring opening of the trans cyclopropane starting
material 29 due to steric interaction between the aryl
group on the cyclopentane ring system and the attacking
nucleophile (Scheme 11). The unreacted cyclopropane 30
is hydrolyzed to acid 38 upon addition of DMF and NaOH
(aq). Fortunately, the decarboxylation of 38 to 40 does
not occur due to its inability to generate the immediate
enol product at the bridgehead carbon, and removal of
acid 38 is accomplished by a simple base extraction.
Overall, this simple protocol effects a one-pot ring-
opening/deprotection/hydrolysis/decarboxylation reaction
sequence, as well as providing a means of selectively
removing the undesired isomer 30.
Although the two-step, one-pot protocol provided the
desired ring-opened product in moderate yield, a signifi-
cant percentage of the starting material was still unac-
counted for. 1H NMR of the crude reaction mixture
revealed that a byproduct identified as deriving from ring
opening of the cyclopropane ring with dimethylamine was
formed in about 15-25%. This is not surprising as DMF
is known to decompose to dimethylamine and carbon
monoxide at high temperatures and that dimethylamine
is substantially more nucleophilic than sodium acetate.
To avoid the formation of the dimethylamine byprod-
uct, other polar, aprotic solvents were investigated
including DMAc, DMSO, and NMP. Reactions in these
solvents were found to be much slower and provided
multiple byproducts as determined by HPLC. Further
solvent studies led to the use of HOAc directly as solvent.
Reactions with HOAc were fast; however, the reaction
was accompanied by approximately 6% of the cis ring-
opened product. This byproduct posed minimal problems
as it was easily removed in subsequent reactions. Using
the new procedure, a one-pot ring-opening/deprotection/
hydrolysis/decarboxylation reaction sequence provided
the desired product 1 in 86-94% yield from 29.
Con clu sion
The remarkable transformation of a mixture of cyclo-
propanes 29 and 30 to cyclopentanone 1 involves multiple
bond-making and bond-breaking steps, as well as a
selective kinetic resolution of the diastereomeric cyclo-
propanes. The first step of the sequence involves the ring
In summary, an efficient and practical asymmetric
synthesis of (+)-trans-3-hydroxymethyl-4-(3-fluorophe-
nyl)cyclopentanone (1) was accomplished. An asymmetric
(27) The synthesis of 37 will be reported in due course.
5514 J . Org. Chem., Vol. 67, No. 16, 2002

New Route to 3,4-Disubstituted Cyclopentanones
SCHEME 11
mol). The mixture was cooled to 0-5 °C followed by the slow
addition of triethylamine (3.55 L, 25.5 mol) over 2-5 min. After
being stirred for 30 min at 0-5 °C, the solution was warmed
to 20-25 °C and subsequently stirred for 30 min. A 22-L,
round-bottom flask, equipped with mechanical stirrer, ther-
mocouple, N2 inlet, and addition funnel, was charged with 1,2-
dichloroethane (5.3 L) and 1,1-carbonyldiimidazole (1.52 kg,
9.37 mol). A 1,2-dichloroethane (2 L) solution of 12 (1.65 kg,
8.50 mol) was added to the CDI slurry over 20 min. After being
stirred for 20 min, the CDI/12 solution was added to the
monomethylmalonate solution over 5 min. The resulting
solution was heated at 45 °C for 1.5 h and then cooled to 0-5
°C. To the cold mixture was added cold 2 N HCl (25 L). The
aqueous phase was separated and extracted with 1,2-dichlo-
roethane (6 L). The 1,2-dichloroethane extracts were combined,
dried over anhydrous Na2SO4, and filtered and the cake was
washed with 1,2-dichloroethane (1 L). The filtrate and washes
were combined and concentrated to yield 2.10 kg of 2 (98%
yield). Silica gel chromatography on a small sample yielded
pure 2 as a colorless oil. Rf 0.44 (hex:EtOAc ) 3:1). Colorless
oil, 9:1 â-ketoester:enol. 1H NMR (400 MHz, CDCl3) δ 7.28 (m,
1H), 7.00 (d, J ) 7.7 Hz, 1H), 6.92 (m, 2H), 5.94 (ddd, J )
17.1, 10.3, 6.8 Hz, 1H), 3.96 (m, 1H), 3.72 (s, 3H), 3.40 (s, 2H),
2.99 (m, 2H). 13C NMR (100.6 MHz, CDCl3) δ 200.3, 167.2,
162.9 (d, J CF ) 246.1 Hz), 145.1 (d, J CF ) 7.2 Hz), 139.6, 130.0
(d, J CF ) 8.8 Hz), 123.3 (d, J CF ) 2.4 Hz), 115.4, 114.5 (d, J CF
) 20.1 Hz), 113.6 (d, J CF ) 20.1 Hz), 52.3, 49.5, 47.9, 43.7. IR
(neat, cm-1) 2955, 1746, 1717, 1589, 1438. Anal. Calcd for
Mo-catalyzed alkylation reaction was used to set the first
stereocenter. Optimization of this reaction led to signifi-
cant improvements, including the use of the readily
available, air-stable Mo(CO)6 as the molybdenum pre-
catalyst. A diastereoselective Cu-catalyzed intramolecu-
lar cyclopropanation reaction led to the establishment of
the second chiral center in a 85:15 ratio. Extensive
catalyst screening, including chiral catalysts, identified
(CH3CN)4CuPF6 and CuCl/AgOTf as the most selective
catalyst. Finally, a one-pot ring-opening/deprotection/
hydrolysis/decarboxylation sequence was used to gener-
ate 1 in good yield.
Exp er im en ta l Section
Gen er a l Meth od s. All reactions were performed under an
atmosphere of nitrogen unless otherwise specified. NMR
spectra were obtained in CDCl3 at 400 MHz for 1H and 100
MHz for 13C. Elemental analyses were performed by Quantita-
tive Technologies Inc. (Whitehouse, NJ ). Silica gel chroma-
tography was performed with 200-400 mesh SiO2 (EM
Science). All commercial reagents and solvents were used as
received unless otherwise noted.
(3S,4S)-3-(3-F lu or op h en yl)-4-(h yd r oxym et h yl)cyclo-
p en ta n on e (1). A 72-L round-bottom flask equipped with a
temperature probe, overhead stirrer, and nitrogen and vacuum
inlet was charged with the 29/30 (5:1 exo/endo, 1.60 kg trans,
6.45 mol trans), NaOAc (5.29 kg, 64.5 mol), and HOAc (32 L,
560 mol). The mixture was evacuated and back filled with N2
(3 cycles) and heated to 105 °C for 15 h. Acetic acid (22 L) was
distilled off with slight vacuum while maintaining the tem-
perature above 90 °C. The mixture was then cooled to 80 °C
followed by the addition of DMF (27 L) and then cooled to room
temperature. To this was added 3 N NaOH (17 L) and then 5
N NaOH (32 L) until pH 12 while maintaining the temperature
below 50 °C. The resulting solution was heated to 70 °C for
30 min and then cooled to room temperature. To the mixture
was added MTBE (60 L) and the aqueous layer was separated
and extracted with MTBE (60 L). The organic layers were
combined, washed with 30 L of 10% NaCl in water, dried over
MgSO4, filtered, and concentrated to give an assay yield of 1.46
kg of 1 in a 16:1 ratio of trans to cis (90% yield total). Silica
gel chromatography on a small sample yielded pure 1 as a clear
C
14H15FO3: C, 67.19; H, 6.04. Found: C, 67.14; H, 6.06.
(3S)-3-P h en yl-4-p en ten oic Acid (12). A toluene solution
of malonic ester intermediate 13 (5.50 kg, 20.66 mol) was
charged to a 50-L flask equipped with mechanical stirrer,
thermocouple, and condenser. Water (18.3 L) and 5 N NaOH
(12.2 L) were added to the flask and the resulting mixture was
heated to 90 °C for 1 h. The flask was fitted with a distillation
head and a downward condenser, and the mixture was distilled
to remove MeOH. The distillation was continued until the level
of MeOH byproduct was <3-4 mol % (determined by 1H
NMR). After the distillation was stopped, the mixture was
transferred to a 100-L reactor and the pH was adjusted to 1.0
with concentrated HCl (4.63 L). The mixture was refluxed for
15 h then cooled to 20-25 °C, and the organic layer was
separated. The aqueous layer was extracted with 18 L of
toluene. The first organic layer and the toluene extract were
combined and washed with 7.3 L of 5% NaCl (aq). The solution
was dried with Na2SO4 (2.44 kg), filtered, weighed (19.0 kg,
HPLC assay: 3.52 kg product, 90% yield), and used directly
in the next reaction. Silica gel chromatography on a small
sample yielded pure 12 as an oil. Rf 0.31 (hex:EtOAc ) 3:1).
1H NMR (400 MHz, CDCl3) δ 7.29 (m, 1H), 7.01 (d, J ) 7.8
Hz, 1H), 6.94 (m, 2H), 5.97 (ddd, J ) 17.2, 10.3, 6.8 Hz, 1H),
5.12 (m, 2H), 3.88 (m, 2H), 2.78 (m, 2H). 13C NMR (100.6 MHz,
CDCl3) δ 178.0, 163.0 (d, J CF ) 245.7 Hz), 144.7 (d, J CF ) 7.2
Hz), 139.2, 130.1 (d, J CF ) 8.0 Hz), 123.2 (d, J CF ) 3.2 Hz),
115.6, 114.5 (d, J CF ) 21.7 Hz), 113.7 (d, J CF ) 21.7 Hz), 44.8,
39.7. IR (neat, cm-1) 1713, 1614, 1590, 1488. Anal. Calcd for
1
oil. Rf 0.30 (hex:EtOAc ) 1:1). H NMR (400 MHz, CDCl3) δ
7.32 (m, 1H), 7.06 (d, J ) 7.7 Hz, 1H), 6.98 (m, 2H), 3.73 (dd,
J ) 10.7, 3.8 Hz, 1H), 3.61 (dd, J ) 10.7, 5.3 Hz, 1H), 3.28
(ddd, J ) 19.0, 10.7, 8.2 Hz, 1H), 2.75 (dd, J ) 18.5, 7.9 Hz,
1H), 2.61 (dd, J ) 17.7, 7.1 Hz, 1H), 2.31-2.51 (m, 3H). 13C
NMR (100.6 MHz, CDCl3) δ 216.1, 163.1 (d, J CF ) 251 Hz),
144.5 (d, J CF ) 7.2 Hz), 130.4 (d, J CF ) 8.0 Hz), 123.0 (d, J CF
) 2.4 Hz), 114.2 (d, J CF ) 20.9 Hz), 113.9 (d, J CF ) 20.9 Hz),
62.5, 47.1, 46.3, 43.9, 41.7. IR (neat, cm-1): 3445, 2914, 1737,
1588. GCMS: m/z 208. The ee was determined to be 98%.
HPLC ee analysis (ChiralPak AD-RH, 4.6 × 250 mm, flow rate
) 0.5 mL/min, detection at 210 nm, 0.1% H3PO4 in H2O:
acetonitrile 60:40): tR ) 9.5 min (minor), tR ) 10.3 min (major).
Meth yl (5S)-5-(3-F lu or op h en yl)-3-oxo-6-h ep ten oa te (2).
A 50-L, 4-necked round-bottom flask, equipped with mechan-
ical stirrer, thermocouple, reflux condenser with N2 inlet,
addition funnel, and steam pot, was charged sequentially with
1,2-dichloroethane (7.2 L), monomethylmalonate potassium
salt (1.99 kg, 12.17 mol), and magnesium chloride (830 g, 8.72
C
11H11FO2: C, 68.03; H, 5.71. Found: C, 67.52; H, 5.69. The
ee was determined to be 98%. HPLC ee analysis (ChiralPak
AD-RH, 4.6 × 250 mm, flow rate ) 0.5 mL/min, detection at
210 nm, 0.1% H3PO4 in H2O:acetonitrile 68:32): tR ) 15.0 min
(minor), tR ) 16.3 min (major).
Dim et h yl 2-[(1R)-1-(3-flu or op h en yl)-2-p r op en yl]p r o-
J . Org. Chem, Vol. 67, No. 16, 2002 5515

Palucki et al.
p a n ed ioa te (13).28 A 12-L, three-necked round-bottomed flask
was charged with Mo(CO)6 (219 g, 0.828 mol) and ligand 15
(402 g, 1.24 mol) and evacuated/back filled with argon (3
cycles). To this was added anhydrous toluene (4.36 L). The
flask was evacuated/back filled with argon (3 cycles), and the
resulting mixture was heated to 85-88 °C for 4 h. Separately,
a 100-L flask was charged with dimethyl sodiomalonate (2.36
kg, 12.42 mol), a toluene solution of carbonate 21 (2.0 kg, 8.28
mol, 3 L of anhydrous toluene), and toluene (30.6 L). The flask
was evacuated/back filled with argon and the resulting solution
was heated to 70 °C followed by the addition of the catalyst
solution via cannula. The resulting mixture was heated to 85
°C for 15 h and subsequently cooled to room temperature.
Water (36 L) was added and the organic layer was separated
and filtered through a small pad of silica gel. Product assay
yield was 91%, with a 97% ee and a 95:5 regioselectivity
(branched:linear ratio). The solution was concentrated in vacuo
and used directly in the next reaction. Crystallization of a
small sample resulted in large crystalline prisms. Mp 34-35
°C. Rf 0.35 (hex:EtOAc ) 9:1). 97% ee. 1H NMR (400 MHz,
CDCl3) δ 7.24-7.52 (m, 2H), 7.01 (d, J ) 7.7 Hz, 1H), 6.90-
6.95 (m, 2H), 5.96 (ddd, J ) 17.0, 10.2, 8.2 Hz, 1H), 5.11-5.16
(m, 2H), 4.09-4.14 (m, 1H), 3.84 (d, J ) 10.9 Hz, 1H), 3.75 (s,
3H), 3.53 (s, 3H). 13C NMR (100 MHz, CDCl3) δ 162.8 (d, J CF
) 241 Hz), 154.9, 140.8, 135.2, 130.2 (d, J CF ) 10.1 Hz), 122.6,
118.0, 115.3 (d, J CF ) 20.1 Hz), 113.9 (d, J CF ) 20.1 Hz), 79.3,
54.9. IR (neat, cm-1) 2955, 1732, 1614, 1590, 1489. Anal. Calcd
for C14H15O4F: C, 63.15; H, 5.68. Found: C, 62.94; H, 5.63.
The ee was determined to be 98%. HPLC ee analysis (Chiral-
Pak AD, 4.6 × 250 mm, flow rate ) 0.8 mL/min, detection at
over MgSO4, filtered, and used directly in the next reaction
(4.88 kg, HPLC assay yield: 95%). Silica gel chromatography
on a small sample yielded pure 24 as a colorless oil. Rf 0.52
(hex:EtOAc ) 3:1). 1H NMR (400 MHz, CDCl3) δ 7.27 (m, 1H),
7.04 (d, J ) 7.7 Hz, 1H), 6.97 (d, J ) 10.1 Hz, 1H), 6.90 (m,
1H), 5.98 (ddd, J ) 17.2, 10.4, 7.1 Hz), 5.08 (m, 2H), 4.02 (m,
1H), 3.85 (s, 3H), 3.67 (d, J ) 7.3 Hz, 2H). 13C NMR (100.6
MHz, CDCl3) δ 190.3, 162.9 (d, J CF ) 247.4 Hz), 145.4 (d, J CF
) 7.2 Hz), 139.8, 129.9, (d, J CF ) 8.0 Hz), 123.4 (d, J CF ) 2.4
Hz), 115.2, 114.6 (d, J CF ) 20.9 Hz), 113.4 (d, J CF ) 20.1 Hz),
52.2, 44.8, 44.5. IR (neat, cm-1) 2956, 2135, 1716, 1658, 1589.
Anal. Calcd for C14H13FN2O3: C, 60.87; H, 4.74; N, 10.14.
Found: C, 60.83; H, 4.72; N,10.10. The ee was determined to
be 98%. HPLC ee analysis (ChiralPak AD-RH, 4.6 × 250 mm,
flow rate ) 0.5 mL/min, detection at 210 nm, H2O:acetonitrile
20:80): tR ) 7.2 min (minor), tR ) 8.5 min (major).
Met h yl (1S,4S,5R)-4-(3-F lu or op h en yl)-2-oxob icyclo-
[3.1.0]h exa n e-1-ca r boxyla te (29). A 100-L flask equipped
with a 5-L addition funnel, reflux condenser, temperature
probe, stopper, and overhead stirrer was charged with 1,2-
dichloroethane (11 L) and (CH3CN)4CuPF6 (55.41 g, 0.149 mol)
and heated to 77 °C. To this was slowly added a 1,2-
dichloroethane solution of 24 (1.38 kg, 5.0 mol, 2.5 L) over a
period of 8 h, with the reaction temperature maintained
between 77 and 81 °C. At the end of the addition, the reaction
was stirred for an additional hour and then cooled to room
temperature. Brine was added (30 L) and the organic layer
was separated, dried over MgSO4, and filtered. The filtrate
was filtered through a silica gel pad and concentrated to give
1.20 kg of 29/30 (96.5% yield for the combined trans and cis
products) in a ratio of 4.47:1 trans:cis. Silica gel chromatog-
raphy on a small sample yielded pure 29 as a colorless oil. Rf
0.35(hex:EtOAc ) 3:1). 1H NMR (400 MHz, CDCl3) δ 7.28-
7.33 (m, 1H), 6.89-6.96 (m, 3H), 3.82 (s, 3H), 3.49 (d, J ) 8.5
Hz, 1H), 2.76 (dd, J ) 19.1, 8.5 Hz, 1H), 2.64 (m, 1H), 2.28 (d,
J ) 19.1 Hz, 1H), 2.13-2.17 (m, 1H), 1.54 (t, J ) 5.3 Hz, 1H).
210 nm, 2% MeOH in Hexane): tR ) 9.1 min (minor), tR
)
10.4 min (major).
1-(3-F lu or op h en yl)-2-p r op en yl Meth yl Ca r bon a te (21).
A 100-L flask equipped with mechanic stirrer, thermocouple,
N2 inlet, and addition funnel was charged with toluene (60 L)
and 3-fluorobenzaldehyde (4 kg, 97%, 31.2 mol). After the
solution was cooled to -5 °C, vinylmagnesium chloride (20.6
L, 1.6 M in THF, 32.8 mol) was added dropwise over a period
of 1.5 h while maintaining the temperature below 0 °C. The
reaction was stirred for an additional 60 min at -5 to 0 °C,
followed by the addition of methyl chloroformate (2.74 L, d )
1.223, 35.4 mol) over 30 min while maintaining the temper-
ature below 0 °C. After the mixture was stirred for 60 min,
0.5 N HCl (40 L) was added, while maintaining the temper-
ature <20 °C. The organic layer was separated and filtered
through silica gel (4 kg), and the filtrates containing the
product were concentrated to 8.88 kg. 1H NMR of the solution
showed the material was 56.7 wt %, yielding 5.01 kg of allyl
carbonate 21 (76% yield). Silica gel chromatography of a small
sample yielded pure 21 as a light yellow oil). Rf 0.37 (hex:
13C NMR (100.6 MHz, CDCl3) δ 205.5, 168.3, 163.1 (d, J CF
241 Hz), 147.1, 130.8 (d, J CF ) 10.1 Hz), 122.0, 114.1 (d, J CF
)
)
10.1 Hz), 113.6 (d, J CF ) 10.1 Hz), 52.7, 42.4, 39.0, 38.9, 38.0,
21.5. IR (neat, cm-1) 1732, 1589, 1438. Anal. Calcd for C14H13
-
FO3 (5.5:1 mixture of 29:30): C, 67.73; H, 5.28. Found: C
67.36; H, 5.26.
1
EtOAc ) 9:1). H NMR (400 MHz, CDCl3) δ 7.39 (dd, J ) 7.8,
5.9 Hz, 1H), 7.16 (d, J ) 7.8 Hz, 1H), 7.09 (d, J ) 9.6 Hz, 1H),
7.00-7.15 (m, 1H), 6.09-5.97 (m, 2H), 5.30-5.40 (m, 2H), 3.81
(s, 3H). 13C NMR (100 MHz, CDCl3) δ 162.8 (d, J CF ) 241 Hz),
154.9, 140.8, 135.2, 130.2 (d, J CF ) 10.1 Hz), 122.6, 118.0, 115.3
(d, J CF ) 20.1 Hz), 113.9 (d, J CF ) 20.1 Hz), 79.3, 54.9. IR
(neat, cm-1) 2959, 1748, 1593, 1489, 1442. Anal. Calcd for
Met h yl (1R,4S,5S)-4-(3-F lu or op h en yl)-2-oxob icyclo-
[3.1.0]h exa n e-1-ca r boxyla te (30). Colorless oil. Rf 0.40 (hex:
EtOAc ) 3:1). 1H NMR (400 MHz, CDCl3) δ 7.28-7.36 (m, 1H),
7.08 (d, J ) 7.7 Hz, 1H), 6.89-7.03 (m, 2H), 3.87 (m, 1H), 3.80
(s, 3H), 2.90-2.95 (m, 1H), 2.70 (dd, J ) 18.3, 9.0 Hz, 1H),
2.35 (dd, J ) 18.3, 10.5 Hz), 2.14-2.18 (m, 1H), 1.62 (t, J )
5.4 Hz, 1H). 13C NMR (100.6 MHz, CDCl3) δ 204.1, 168.2, 163.1
C
11H11O3F: C, 62.85; H, 5.27. Found: C, 62.78; H, 5.32.
Met h yl (5S)-2-Dia zo-5-(3-flu or op h en yl)-3-oxo-6-h ep -
ten oa te (24). A 100-L flask equipped with mechanical stirrer,
thermocouple, addition funnel, and nitrogen inlet was charged
with 2 (94.63 kg, 18.5 mol), 1,2-dichloroethane (20 L), and
4-acetomidobenzenesulfonyl azide (4.40 kg, 18.3 mol). To this
was added triethylamine (7.74 L, 55.5 mol) while maintaining
the temperature <25 °C. The reaction was stirred at 20-25
°C for 4 h and filtered. The collected solid was washed with
1,2-dichloroethane (20 L). The combined 1,2-dichloroethane
solutions was cooled to 5 °C and washed with 2 N HCl (14.4
L) and twice with H2O (12.5 L). The organic layer was dried
(d, J CF ) 246 Hz), 143.3 (d, J CF ) 10.1 Hz), 130.4 (d, J CF
)
10.1 Hz), 122.6, 114.1(d, J CF ) 10.1 Hz), 113.9 (d, J CF ) 10.1
Hz), 52.5, 39.8, 38.5, 37.3, 35.8, 20.6. IR (neat, cm-1) 1732,
1589, 1438. Anal. Calcd for C14H13FO3 (5.5:1 mixture of 29:
30): C, 67.73; H, 5.28. Found: C 67.36; H, 5.26
Ack n ow led gm en t . We thank Professor Barry M.
Trost for helpful discussions and Drs. Chris Welch and
Bing Mao for HPLC analyses. We also thank Ms. Lisa
DiMichele for helpful NMR analysis.
(28) The absolute stereochemistry was assigned by analogy to the
phenyl analogue. See ref 11c.
J O025890A
5516 J . Org. Chem., Vol. 67, No. 16, 2002