Catalytic asymmetric formation of δ-lactones by [4+2] cycloaddition of zwitterionic dienolates generated from α,β-unsaturated acid chlorides

Article abstract of DOI:10.1002/anie.200700859

Smooth elaboration: Versatile δ-lactone building blocks are provided by tertiary amine-catalyzed asymmetric [4+2] cycloadditions of α,β-unsaturated acid chlorides and the electron-poor aldehyde chloral (see scheme). Silyl-substituted acid chlorides (R¹ = R ₃Si) can be used for the diastereoselective synthesis of β-hydroxy-δ-lactones possessing quaternary stereocenters. (Chemical Equation Presented).

Full text of DOI:10.1002/anie.200700859

Angewandte
Chemie
DOI: 10.1002/anie.200700859
Asymmetric Cycloaddition
Catalytic Asymmetric Formation of d-Lactones by [4+2] Cycloaddition
of Zwitterionic Dienolates Generated from a,b-Unsaturated Acid
Chlorides**
Paolo S. Tiseni and RenØ Peters*
d-Lactones are subunits of numerous natural and unnatural
products that display a wide range of biological activity. In
many cases they show high efficacy as antibacterial,[1] antiviral
[2]
(HIVprotease inhibitors),
anticancer,[3] immunosuppres-
sive,[4] or cholesterol-lowering agents (HMGR inhibitors).[5]
For example, the majority of statin drugs such as Lipitor and
Zocor, the worldꢀs best-selling drugs in 2004, contain a b-
hydroxy-d-lactone moiety or the corresponding open-chain d-
hydroxy carboxylate form.[6] In addition, d-lactones are very
useful building blocks for the synthesis of bioactive com-
pounds.[7]
Many approaches, most often multistep procedures, have
been elaborated for the preparation of d-lactones.[8] The most
direct route to a,b-unsaturated d-lactones is based on hetero-
Diels–Alder (HDA) reactions.[9] In order to generate directly
the desired oxidation state, a vinylketene equivalent such as a
vinylketene acetal is required as the diene component. It has
been reasoned though that twofold substitution at the
butadiene terminus has a deleterious effect upon the enan-
tioselectivity of HDA reactions, thus explaining why these
dienes have been less frequently used in asymmetric HDA
reactions than Danishefsky-type dienes.[10] Up to now, only
three highly enantioselective, catalytic HDA-based method-
ologies using a vinylketene acetal have been reported,[11–13]
which are all restricted to the use of Brassard-type dienes (1,3-
dialkoxy-1-(trimethylsiloxy)butadienes)[14] and aromatic
aldehydes[15] and which all require long reaction times (48–
72 h) to provide useful yields.
We present herein a new concept for the synthesis of a,b-
unsaturated d-lactones 6 which circumvents the preformation,
isolation, and purification of moisture- and acid-sensitive
vinylketene acetals. Our work was based on the hypothesis
that substituted vinylketenes 2 should be formed in situ by
dehydrohalogenation of a,b-unsaturated acid chlorides 1
(Scheme 1).[16] Vinylketenes are known to be inherently
unstable species,[17] but they might be trapped and at the
Scheme 1. Proposed [4+2] cycloaddition of a,b-unsaturated acid chlor-
ides 1 with aldehydes 5 providing lactones 6.
same time activated as a diene component of a Diels–Alder
reaction by an enantiopure tertiary amine, thus forming an
enantiomerically pure zwitterionic dienolate 4. This type of
species was supposed to be reactive enough, in an s-cis
conformation, to undergo [4+2] cycloadditions with alde-
hydes, by either a stepwise or a concerted mechanism. Our
investigations were inspired by the tertiary-amine-catalyzed
asymmetric synthesis of b-lactones from ketene via zwitter-
ionic enolate intermediates.[18] Because of the considerable
homology of 4 to vinylketene acetals, the intermediate
formation of these dienolates was anticipated to overcome
the pronounced tendency of vinylketenes to preferentially
undergo [2+2] cycloaddition reactions.[19]
3,4-Dimethylpent-2-enoyl chloride (1a; R1 = iPr) and
trichloroacetaldehyde (chloral, 5a; R2 = CCl3) were chosen
as model substrates for the development of this process.
Initial experiments in acetonitrile at À158C using stoichio-
metric amounts of quinuclidine or triethylamine, both acting
as a base as well as an achiral nucleophilic catalyst, provided
the racemic d-lactone 6a (R1 = iPr, R2 = CCl3) in moderate
yield (45% and 44%, respectively). Having demonstrated
that the intended transformation is in principle possible, we
sought an asymmetric version. After an initial screening of
various combinations of cinchona alkaloid derivatives, stoi-
chiometric non-nucleophilic tertiary amine auxiliary bases,
and solvents, trimethylsilylquinidine (TMSQd, 3a, 1.0 equiv),
iPr2NEt (Hünigꢀs base, 2.0 equiv), and toluene were selected
for further studies. At À158C the product was formed in
82% ee, albeit in very low yield (18%). One of the reasons for
the poor performance was the incomplete conversion of 1a
[*] P. S. Tiseni,Prof. Dr. R. Peters
Laboratory of Organic Chemistry,ETH Zürich
Wolfgang-Pauli-Strasse 10,
Hönggerberg HCI E 111,8093 Zürich (Switzerland)
Fax: (+41)44-633-1226
E-mail: peters@org.chem.ethz.ch
Homepage: http://www.peters.ethz.ch
[**] This work was financially supported by Novartis (PhD fellowship for
P.S.T.) and by F. Hoffmann-La Roche.
Supporting information for this article (including experimental
details) is available on the WWW under http://www.angewandte.org
or from the author.
Angew. Chem. Int. Ed. 2007, 46, 5325 –5328
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5325

Communications
(NMR monitoring). To facilitate the deprotonation process
unbranched Et substituent the ee was only moderate
and to activate the aldehyde substrate, various metal triflate
salts M(OTf)n were investigated as Lewis acid co-catalysts.
The deprotonation process[20] was significantly improved with
all investigated Lewis acids,[21] and the yield of 6a increased to
> 70% in the presence of 20 mol% of Er(OTf)3 (76%),
Sn(OTf)2 (75%), Sc(OTf)3 (74%), or Nd(OTf)3 (72%) when
the acid chloride was added slowly over 30 min with a syringe
pump. The enantiomeric excess was determined to be 82% in
all experiments regardless of which Lewis acid co-catalyst was
used. As the enantioselectivity does not depend upon the
presence or absence of the metal triflate salt, we assume that
the Lewis acid is not involved in the cycloaddition step itself
and simply facilitates the dehydrochlorination of 1. Increasing
the temperature to 08C gave similar results (70% yield with
Er(OTf)3, 80% ee), while decreasing the temperature to
À408C resulted in a dramatic decrease in yield (22% with
Er(OTf)3, 84% ee).
The investigation of the co-catalyst loadings revealed that
reducing the amount of the Lewis acid to 10 mol% had
almost no negative effect in the case of Sn(OTf)2 (yield 72%,
82% ee), and even with as little as 1.4 mol% the reaction was
still relatively efficient (60% yield, 82% ee). Decreasing the
amount of TMSQd 3a required a slower addition of the acid
chloride to avoid massive polymerization. When 1a was
added over 120 min to a mixture of 20 mol% of 3a and
10 mol% of Sn(OTf)2, 6a was obtained in 78% yield
(82% ee, Table 1, entry 1).
(Table 1, entry 2), the values were significantly higher and
preparatively useful with branched or aromatic substituents
1
À
(entries 3–7). Notably, even in the case of 1b (R = Et) C C
bond formation occurred with complete regioselectivity. With
the bulky iBu and tBu substituents, higher catalyst and co-
catalyst loadings were required to obtain high conversions
(Table 1, entries 3 and 6).
To extend the synthetic value of the methodology, we
turned our attention to substrates 1g–j bearing a trialkylsilyl
moiety, since the silyl groups would eventually permit further
useful modifications of the d-lactones after the cycloaddition
step. The substrates were easily prepared by Ru-catalyzed
hydrosilylation of tetrolic acid according to an efficient
procedure developed by Trost et al.[23] While the silyl-
containing substrates 1g–j provided only moderate yields
for the cycloaddition event due to a less efficient conversion
of these acid chlorides, the obtained enantioselectivities were
excellent (Table 1, entries 8–12).
As a general trend, the remote control of the enantiose-
lectivity depends largely on the steric bulk of R1.[24] Based on
MMFF calculations performed to determine the preferred
conformation of the cisoid zwitterionic dienolate, we have
developed a working hypothesis to rationalize the enantiose-
lectivity for the formation of d-lactones 6a–j (Scheme 2).
Table 1: Preparation of a,b-unsaturated d-lactones 6.
Entry
1
R1
X
Y
Yield [%][a]
ee [%][b]
1
2
1a
1b
1c
1d
1e
1e
1 f
1g
1g
1h
1i
iPr
Et
iBu
cHex
tBu
20
20
40
20
20
10
10
20
10
10
20
10
30
20
30
30
30
78
60
73
75
58
80
73
54
43
47
51
61
82
54
70
83
95
95
81
96
96
92
97
97
Scheme 2. Rationalization of the remote stereocontrol.
3[c]
4
5
6
7
8
Since the re face (relative to the C1 enolate atom) is shielded
by the quinoline and the OTMS moieties, the aldehyde will
attack from the better accessible si face. In the preferred
orientation of the aldehyde in the transition state the residue
R1 and the CCl3 group should point away from each other to
avoid unfavorable steric interactions, thus explaining the
large influence of R1. Since no acyclic vinylogous Mukaiyama
aldol products could be detected, this might indicate that the
cycloaddition event proceeds by a concerted mechanism.
The trichloromethyl moiety in lactones 6 is a synthetically
versatile functional group that can be readily converted into
several valuable functionalities. Basic hydrolysis with LiOH
in water at 608C gave carboxylic acid 7 in 62% yield and
without significant racemization (Scheme 3).[25] Selective
partial reductive removal of the chlorine atoms of the
trichloromethyl group was achieved using tributyltin hy-
dride.[26] Performing the reduction in THF at 608C afforded
the dichloro derivative 8,[27] whereas in refluxing toluene, the
tBu
Ph
40
20
Et3Si
Et3Si
BnMe2Si
nPr3Si
nBu3Si
100
40
100
100
100
9[d]
10
11
12
1j
[a] Yields of isolated products. [b] Determined by HPLC on a chiral
column (Daicel OD-H). [c] Catalyst 3b was used (see Scheme 1). [d] Acid
chloride 1g was added over 220 min.
Similar reaction conditions were applied to several
alternative substrates 1b–f bearing different branched or
unbranched aliphatic, alicyclic, or aromatic groups R1
(Table 1). d-Lactones 6a–f were formed in good yield and
with up to 95% ee (Table 1, entries 1–7).[22] Both the enantio-
selectivity and the conversion of the acid chloride depended
primarily upon the steric bulk of R1. While with the
5326
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 5325 –5328

Angewandte
Chemie
TBAF.[28] To the best of our knowledge, the migration of
simple alkyl substituents has not been reported before. The
silanes 12 were subsequently oxidized under Tamao condi-
tions to provide the tertiary alcohols 13.[29,30] To accomplish
this oxidation, the lactone moiety in 11[31] had to be hydro-
lyzed first, presumably since the free carboxylate moiety
intramolecularly activates the silyl group. In the case of the
benzyl migration, the product is already ring-opened without
a basic hydrolysis step.
In conclusion, we have developed a tertiary-amine-
catalyzed enantioselective [4+2] cycloaddition of a,b-unsatu-
rated acid chlorides and the electron-poor aldehyde chloral to
provide synthetically versatile d-lactone building blocks. The
substituent at the b position can be varied to a large degree,
and the trichloromethyl group made it possible to install
several useful functional groups at the d position. b-Hydroxy-
d-lactones possessing a quaternary stereocenter at the b
position were diastereoselectively synthesized starting from
the a,b-unsaturated d-lactone carrying a trialkylsilyl substitu-
ent at the b position. Studies to further improve the scope of
this novel approach are ongoing.
Scheme 3. Synthetic modification of the trichloromethyl group.
Received: February 26, 2007
Published online: June 1, 2007
monochloro derivative 9 was obtained. The latter was
smoothly hydrolyzed to give the primary alcohol 10.
The silyl-substituted a,b-unsaturated d-lactones 6g–j
were utilized to prepare saturated d-lactones 13 carrying a
tertiary hydroxy group at the b position (Table 2). Treatment
Keywords: d-lactones · cinchona alkaloids · cycloaddition ·
hetero-Diels–Alder reaction · organocatalysis
.
Table 2: Preparation of b-hydroxy d-lactones 13 by alkyl migration/
oxidation.[a]
[1] Y. Kato, Y. Ogawa, T. Imada, S. Iwasaki, N. Shimazaki, T.
Kobayashi, T. Komai, J. Antibiot. 1991, 44, 66.
[2] F. E. Boyer, J. V. N. V. Prasad, J. M. Domagala, E. L. Ellsworth,
C. Gajda, S. E. Hagen, L. J. Markoski, B. D. Tait, E. A. Lunney,
A. Palovsky, D. Ferguson, N. Graham, T. Holler, D. Hupe, C.
Nouhan, P. J. Tummino, A. Urumov, E. Zeikus, G. Zeikus, S. J.
Gracheck, J. M. Sanders, S. VanderRoest, J. Brodfuehrer, K.
Iyer, M. Sinz, S. V. Gulnik, J. W. Erickson, J. Med. Chem. 2000,
43, 843.
[3] a) X.-P. Fang, J. E. Andersen, C.-J. Chang, J. L. McLaughlin, J.
Nat. Prod. 1991, 54, 1034; b) D. S. Lewy, C.-M. Gauss, D. R.
Soenen, D. L. Boger, Curr. Med. Chem. 2002, 9, 2005; c) S. J.
Shaw, K. F. Sundermann, M. A. Burlingame, D. C. Myles, B. S.
Freeze, M. Xian, I. Brouard, A. B. Smith III, J. Am. Chem. Soc.
2005, 127, 6532; d) M. Kobayashi, K. Higuchi, N. Murakami, H.
Tajima, S. Aoki, Tetrahedron Lett. 1997, 38, 2859.
[4] K. Yasui, Y. Tamura, T. Nakatani, K. Kawada, M. Ohtani, J. Org.
Chem. 1995, 60, 7567.
[5] A. Endo, J. Med. Chem. 1985, 28, 401.
Entry
Substrate
R’
R’’
Yield
Yield
d.r. 13[c]
12 [%][b]
13 [%][b]
1
2
3
4
6g
6h
6i
Et
Bn
nPr
nBu
Et
44
92
38
41
70
85
65
61
>99:1
25:1
>99:1
>99:1
Me
nPr
nBu
[6] E. S. Istvan, J. Deisenhofer, Science 2001, 292, 1160.
[7] J. Mulzer, E. Öhler, Chem. Rev. 2003, 103, 3753.
6j
[8] Reviews: a) I. Collins, J. Chem. Soc. Perkin Trans. 1 1999, 1377;
b) V. Boucard, G. Broustal, J. M. Campagne, Eur. J. Org. Chem.
2007, 225; selected leading references: c) G. R. Weihe, T. C.
McMorris, J. Org. Chem. 1978, 43, 3942; d) M. Hirayama, K.
Gamoh, N. Ikekawa, J. Am. Chem. Soc. 1982, 104, 3735; e) M.
Tori, N. Toyoda, M. Sono, J. Org. Chem. 1998, 63, 306; f) G. E.
Keck, X.-Y. Li, C. E. Knutson, Org. Lett. 1999, 1, 411; g) P. P.
Reddy, K.-F. Yen, B.-J. Uang, J. Org. Chem. 2002, 67, 1034; h) E.
Yoneda, S.-W. Zhang, D.-Y. Zhou, K. Onitsuka, S. Takahashi, J.
Org. Chem. 2003, 68, 8571.
[a] DCM=dichloromethane,MCPBA =meta-chloroperbenzoic acid,
TFA= trifluoroacetic acid. [b] Yield of isolated product. [b] Determined
from the 1H NMR spectrum of isolated 13.
of 6g–j with 2.0 equiv of tetrabutylammonium fluoride
(TBAF) in THF lead to diastereoselective migration of the
benzyl or alkyl substituents by an intramolecular 1,4-addition.
This reaction outcome is consistent with recent reports about
the migration of allyl, benzyl, and phenyl groups within
acyclic b-silyl enones or enoates upon treatment with
[9] General review on HDA reactions: K. A. Jørgensen, Angew.
Chem. 2000, 112, 3702; Angew. Chem. Int. Ed. 2000, 39, 3558.
[10] A. Togni, Organometallics 1990, 9, 3106.
Angew. Chem. Int. Ed. 2007, 46, 5325 –5328
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5327

Communications
[11] a) Q. Fan, L. Lin, J. Liu, Y. Huang, X. Feng, G. Zhang, Org. Lett.
[23] B. M. Trost, Z. T. Ball, J. Am. Chem. Soc. 2004, 126, 13942.
[24] Theoretical studies of HAD reactions have revealed that the
mechanism can change from a concerted nonsynchronous to a
stepwise mechanism depending on the substituents on the
reacting species and on the reaction conditions. See for example:
a) L. F. Tietze, J. Fennen, E. Anders, Angew. Chem. 1989, 101,
1420; Angew. Chem. Int. Ed. Engl. 1989, 28, 1371; b) M. A.
McCarrick, Y.-D. Wu, K. N. Houk, J. Am. Chem. Soc. 1992, 114,
1499; c) M. A. McCarrick, Y.-D. Wu, K. N. Houk, J. Org. Chem.
1993, 58, 3330; d) B. S. Jursic, Z. Zdravkovski, J. Phys. Org.
Chem. 1994, 7, 641; e) L. F. Tietze, A. Schuffenhauser, P. R.
Achreiner, J. Am. Chem. Soc. 1998, 120, 7952.
[25] Based on literature precedent for the hydrolysis of b-trichlor-
omethyl-b-lactones (H. Wynberg, E. G. J. Staring, J. Chem. Soc.
Chem. Commun. 1984, 1181) or b-trichloromethyl carbinols (B.
Jiang, Y.-G. Si, Adv. Synth. Catal. 2004, 346, 669) the stereocen-
ter suffers an inversion of configuration owing to the intermedi-
ate formation of a dichloroepoxide, which undergoes an SN2 ring
opening. Oxidation of primary alcohol 10 providing 7 revealed
the same configuration for 10 (see the Supporting Information).
[26] C. E. Song, J. K. Lee, S. H. Lee, S. Lee, Tetrahedron: Asymmetry
1995, 6, 1063.
[27] For the synthetic usefulness of dichloromethyl groups see e.g.:
a) K.-i. Sato, S. Akai, N. Sugita, T. Ohsawa, T. Kogure, H. Shoji,
J. Yoshimura, J. Org. Chem. 2005, 70, 7496; b) H. Arasaki, M.
Iwata, D. Nishimura, A. Itoh, Y. Masaki, Synlett 2004, 546; c) M.
Shimizu, T. Fujimoto, X. Liu, H. Minezaki, T. Hata, T. Hiyama,
Tetrahedron 2003, 59, 9811.
[28] See Ref. [23] and a) M. E. Jung, G. Piizzi, J. Org. Chem. 2002, 67,
3911; b) L. A. Aronica, F. Morini, A. M. Caporusso, P. Salvadori,
Tetrahedron Lett. 2002, 43, 5813.
2004, 6, 2185; b) Q. Fan, L. Lin, J. Liu, Y. Huang, X. Feng, G.
Zhang, Eur. J. Org. Chem. 2005, 3542.
[12] L. Lin, Q. Fan, B. Qin, X. Feng, J. Org. Chem. 2006, 71, 4141.
[13] H. Du, D. Zhao, K. Ding, Chem. Eur. J. 2004, 10, 5964.
[14] J. Savard, P. Brassard, Tetrahedron Lett. 1979, 20, 4911.
[15] A single example using an a,b-unsaturated aldehyde was
reported, see Ref. [12].
[16] G. B. Payne, J. Org. Chem. 1969, 34, 1341.
[17] 2-(Trimethylsilyl)vinylketene has been reported to be a remark-
able stable vinylketene and has been used for non-enantiose-
lective HDA reactions: D. M. Bennett, I. Okamoto, R. L.
Danheiser, Org. Lett. 1999, 1, 641.
[18] a) D. Borrmann, R. Wegler, Chem. Ber. 1967, 100, 1575; b) H.
Wynberg, E. G. J. Staring, J. Am. Chem. Soc. 1982, 104, 166; c) R.
Tennyson, D. Romo, J. Org. Chem. 2000, 65, 7248; d) C. Zhu, X.
Shen, S. G. Nelson, J. Am. Chem. Soc. 2004, 126, 5352; e) M. A.
Calter, O. A. Tretyak, C. Flaschenriem, Org. Lett. 2005, 7, 1809;
f) review on the use of ketenes in asymmetric synthesis: R. K.
Orr, M. A. Calter, Tetrahedron 2003, 59, 3545; reviews on the
history of ketene chemistry: g) T. T. Tidwell, Eur. J. Org. Chem.
2006, 563; h) T. T. Tidwell, Angew. Chem. 2005, 117, 5926;
Angew. Chem. Int. Ed. 2005, 44, 5778.
[19] a) T. T. Tidwell, Ketenes, Wiley, Hoboken, NJ, 2006; selected
leading references: b) D. A. Jackson, M. Rey, A. S. Dreiding,
Tetrahedron Lett. 1983, 24, 4817; c) S. Y. Lee, S. Kulkarni, B. W.
Burbaum, M. I. Johnston, B. B. Snider, J. Org. Chem. 1988, 53,
1848; d) D. A. Jackson, M. Rey, A. S. Dreiding, Helv. Chim. Acta
1983, 66, 2330.
[20] During the course of the reaction in toluene a precipitate of
[iPr2NHEt]Cl is formed.
[21] Yield of 6a: 55–76%; decreasing values in this order: M = ErIII
,
[29] K. Tamao, T. Kakui, M. Akita, T. Iwahara, R. Kanatani, J.
Yoshida, M. Kumada, Tetrahedron 1983, 39, 983.
[30] The relative configuration of 13 was assigned by NOE experi-
ments: see the Supporting Information for details.
[31] Product 11 is a mixture of three compounds with different
groups X, which can be F, OH, or an O bridge in a dimeric
species.
SnII, ScIII, NdIII, InIII = CuII = YbIII, LiI, SmIII, ZrIV, AlIII = MgII,
GdIII = BiIII, EuIII, ZnII, YIII. The combination of cinchona
alkaloid derivatives and metal triflate salts was previously
successfully applied to [2+2] cycloadditions of ketenes. See: S.
France, M. H. Shah, A. Weatherwax, H. Wack, J. P. Roth, T.
Lectka, J. Am. Chem. Soc. 2005, 127, 1206 and Ref. [18e].
[22] The absolute configuration of d-lactone 6 f was determined by
chemical correlation. For details see the Supporting Informa-
tion.
5328
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 5325 –5328