The total synthesis of the annonaceous acetogenin, muricatetrocin C

Full text of DOI:10.1002/1521-3773(20001016)39:20<3622::AID-ANIE3622>3.0.CO;2-H

COMMUNICATIONS
[10] K. Nagayama, D. Ohmori, Y. Imai, T. Oshima, FEBS Lett. 1983, 158,
208 ± 212.
[11] A. L. Macedo, I. Moura, J. J. G. Moura, J. LeGall, B. H. Huynh, Inorg.
Chem. 1993, 32, 1101 ± 1105.
[12] C. M. Gorst, Y.-H. Yeh, Q. Teng, L. Calzolai, Z.-H. Zhou, M. W. W.
Adams, G. N. La Mar, Biochemistry 1995, 34, 600 ± 610.
[13] I. Bertini, C. Luchinat, G. Mincione, A. Soriano, Inorg. Chem. 1998, 37,
969 ± 972.
[14] J. L. C. Duff, J. L. J. Breton, J. N. Butt, F. A. Armstrong, A. J.
Thomson, J. Am. Chem. Soc. 1996, 118, 8593 ± 8603.
[15] B. Shen, L. L. Martin, J. N. Butt, F. A. Armstrong, C. D. Stout, G. M.
Jensen, P. J. Stephens, G. N. La Mar, C. M. Gorst, B. K. Burgess, J.
Biol. Chem. 1993, 268, 25928 ± 25939.
[16] M. K. Johnson, D. E. Bennett, J. A. Fee, W. V. Sweeney, Biochim.
Biophys. Acta 1987, 911, 81 ± 94.
[17] P. J. Stephen, G. M. Jensen, F. Delvin, T. V. Morgan, C. D. Stout, B. K.
Burgess, Biochemistry 1991, 30, 3200 ± 3209.
[18] C. D. Stout, J. Biol. Chem. 1993, 268, 25920 ± 25927.
[19] J. Hirst, J. L. C. Duff, G. N. L. Jameson, M. A. Kemper, B. K. Burgess,
F. A. Armstrong, J. Am. Chem. Soc. 1998, 120, 7085 ± 7094.
[20] M. Gueron, J. Magn. Reson. 1975, 19, 58 ± 66.
[21] H. A. Azab, L. Banci, M. Borsari, C. Luchinat, M. Sola, M. S. Viezzoli,
Inorg. Chem. 1992, 31, 4649 ± 4655.
of NADH oxidase found in the plasma membranes of cancer
cells.[2, 3] These actions lead to ATP deprivation and subse-
quent apoptosis.[4] More recently the annonaceous acetoge-
nins have also been shown to overcome resistance in multi-
drug resistant (MDR) tumors.[5] Thus, for the above reasons
and by virtue of their limited availability from natural sources,
these compounds have been targeted for total synthesis by a
number of research groups.[6]
In 1996 McLaughlin and co-workers reported the isolation
of muricatetrocin C (1; see Scheme 1) from the leaves of
Rollina mucosa.[7] The molecule exhibits potent inhibitory
action against PC-3 prostatic adenocarcinoma, PACA-2
pancreatic carcinoma, and A-549 lung carcinoma. Herein we
report the first total synthesis of 1, which was achieved by
using a highly convergent synthetic strategy.[8]
The synthetic plan to 1 hinged on the enantioselective
preparation of fragments 2, 3, and 4 and their subsequent
coupling reactions (Scheme 1). It was believed that in
[22] I. Bertini, A. Dikiy, C. Luchinat, R. Macinai, M. S. Viezzoli, M.
Vincenzini, Biochemistry 1997, 36, 3570 ± 3579.
[23] S. E. Iismaa, A. E. Vazquez, G. M. Jensen, P. J. Stephen, J. N. Butt,
F. A. Armstrong, B. K. Burgess, J. Biol. Chem. 1991, 266, 21563 ±
21571.
OH
OH
O
O
OH
OH
O
[24] H. Cheng, K. Grohmann, W. V. Sweeney, J. Biol. Chem. 1990, 265,
12388 ± 12392.
Sonogashira
coupling
chelated
addition
Muricatetrocin C 1
[25] S. Aono, S. Nakamura, R. Aono, I. Okura, Biochem. Biophys. Res.
Commun. 1994, 201, 938 ± 942.
[26] T. Inubushi, E. D. Becker, J. Magn. Reson. 1983, 51, 128 ± 133.
34
OMe
TBSO
OBn
16
I
O
O
O
O
4
O
O
OMe
2
3
4
The Total Synthesis of the Annonaceous
Acetogenin, Muricatetrocin C**
TBSO
OMe
Darren J. Dixon, Steven V. Ley,* and
Dominic J. Reynolds
O
O
O
O
O
BnO
O
TBDPSO
SnBu3
OMe
O
5
6
7
The annonaceous acetogenins comprise a rapidly growing
class of natural products exhibiting a broad spectrum of
biological properties. These include antibacterial, antimalar-
ial, in vivo antitumor, parasiticidal, and pesticidal effects.[1]
Perhaps most exciting is their novel and selective mode of
action as inhibitors of oxidative phosphorylation, which offers
a unique potential for these compounds as anticancer agents.
The acetogenins are known to be powerful inhibitors of
complex I (NADH:ubiquinone oxidoreductase) in mamma-
lian and insect mitochondrial electron transport systems and
Scheme 1. Synthetic plan for 1. Bn ˆ benzyl, TBS ˆ tert-butyldimethylsilyl,
TBDPS ˆ tert-butyldiphenylsilyl.
addition to allowing an efficient synthesis of 1, these three
components would provide an excellent platform for both the
evolution of existing group methodology and the potential
development of new synthetic tools. The key features of our
approach are: the application of the recently reported
(R',R',R,S)-2,3-butanediacetal-protected butane tetrol 8 (see
Scheme 2) as a building block for the anti-1,2-diol component
2 through selective chemical differentiation of the incongru-
ous hydroxy termini;[9] the use of the recently developed
anomeric O ± C rearrangement protocol for the stereoselec-
tive construction of the 2,5-trans-disubstituted THF ring
component 3;[10] and finally the implementation of a new
approach to the (S)-hydroxy-butenolide terminus 4, using a
hetero-Diels ± Alder (HDA) reaction to simultaneously in-
stall the 1,5-stereochemical relationship and mask the bute-
nolide double bond.
[*] Prof. Dr. S. V. Ley, Dr. D. J. Dixon, D. J. Reynolds
Department of Chemistry
University of Cambridge
Lensfield Road
Cambridge, CB2 1EW (UK)
Fax : (‡44)1223-336-442
E-mail: svl1000@cam.ac.uk
[**] We thank the EPSRC (D.J.D. and D.J.R.), the Novartis Research
Fellowship (S.V.L.), and the BP Research Endowment (S.V.L.) for
generous financial support for this work.
3622
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3920-3622 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2000, 39, No. 20

COMMUNICATIONS
lation of the mixture allowed chromatographic separation to
give diastereomerically pure material, 17 (80% yield) at this
stage. Treatment of 17 with TBAF followed by selective
hydrogenation of the alkyne over Raney nickel, gave 18, a
direct precursor to the desired coupling fragment 3, in 89%
yield.
HO
TBSO
O
O
OH
OMe
OMe
b,c
a
OMe
O
O
MeO
O
O
HO
OH
O
OMe
OMe
8
5
d-e
With the two units 10 and 18 completed, the remaining
butenolide fragment 4 was then investigated. It was envisaged
that the 1,5-stereochemical relationship could be installed by
a HDA reaction between diene 7 and nitrosobenzene. The
required regioselectivity was predicted by frontier orbital
calculations,[13] whilst it was believed that unfavorable steric
interaction between the aryl ring of the dienophile and the
methyl substituent of 7 would deliver the necessary diaster-
Br
TBSO
OMe
Br
OMe
f-h
O
O
O
O
OMe
OMe
10
9
Scheme 2. Synthesis of fragment 10, a fragment for 2. a) Reference [9];
b) NaH, THF, 30 min, 208C then TBSCl, 1 h, 208C (72%); c) DMSO,
(COCl)2, CH2Cl2, À788C; NEt3, À788C !08C; d) CH3(CH2)10PPh3I,
nBuLi, THF, 30 min, À788C (75%; two steps); e) RaNi, H2, EtOH,
30 min, 208C; f) TBAF, THF, 2 h, 208C (93%, two steps); g) DMSO,
(COCl)2, CH2Cl2, À788C; NEt3, À788C !08C; h) PPh3, CBr4, CH2Cl2,
30 min, 08C (91%; two steps). TBSCl ˆ tert-butyldimethylsilyl chloride,
DMSO ˆ dimethyl sulfoxide, RaNi ˆ Raney nickel, TBAF ˆ tetrabutyl-
ammonium fluoride.
À
eoselectivity. Subsequent N O bond cleavage followed by
elimination of the aryl amine portion to reintroduce the
butenolide unsaturation would permit elaboration to 4.
Thus, starting with 1,4-butanediol 19, monobenzylation by
trapping of the sodium anion precipitate of 19 with benzyl
bromide, followed by tandem Swern ± Wittig reaction afford-
ed the a,b-unsaturated ester 20 in excellent yield (79%) over
three steps (Scheme 4). Homo-enolate formation by depro-
tonation with LDA in the presence of HMPA,[14] followed by
quenching with (S)-(2)-(tert-butyldimethylsilyloxy)propa-
nal,[15] then subjection of the resulting crude oil to acidic
methanol allowed cyclization to form a b-hydroxy-g-lactone.
Subsequent treatment with methanesulfonyl chloride and
triethylamine afforded diene 21 in 60% yield, with a 6:1 (Z/E)
selectivity at the external double bond. The desired E isomer
With regard to the final steps of the synthesis, addition of
the alkynyllithium reagent derived from 2 to the anomerically
disposed aldehyde 3 was envisaged to introduce the remaining
stereogenic center at C-16. Subsequent manipulation to
unmask a terminal alkyne, would then allow a Sonogashira
coupling with 4 to complete the carbon skeleton; upon
selective hydrogenation and global deprotection this would
afford the natural product 1.
O
The synthesis of fragment 2 began with the easily accessible
diol 8, which was readily converted to the equatorial aldehyde
5 in two steps according to the previously established
procedure (Scheme 2).[9b] Subsequent Wittig reaction allowed
introduction of an eleven-carbon side chain with moderate
(1:4, E/Z) selectivityÐfor ease of characterization the un-
saturation was removed at this stage by hydrogenation over
Raney nickel to afford 9 in good yield (50%, four steps).
Conversion to the vinyl dibromide required a further three
steps; thus desilylation, Swern oxidation, then olefination
according to the Corey ± Fuchs procedure furnished 10 in
46% yield starting from 8.[11]
a
b
TBDPSO
OH
O
OH
TBDPSO
TBDPSO
11
12
13
c
O
OH
OH
TBDPSO
15
d,e
O
O
TBDPSO
14
O
TBDPSO
The 2,5-trans-disubstituted tetrahydrofuran unit 3 was then
constructed by addition of allylmagnesium bromide to the
least hindered end of O-silylated-(R)-glycidol 11, which
proceededÐin the presence of catalytic Li2CuCl4Ðin excel-
lent yield (93%) (Scheme 3).[12] Ozonolysis of the alkenol 12
to give the lactol 13, followed by treatment of this crude
material with an excess of propargyl alcohol and a catalytic
amount of Amberlyst A15 in refluxing benzene afforded the
rearrangement precursor 14 in 93% yield as a 1:1 mixture of
anomers. With multigram quantities of the propargylic ether
available, the anomeric O ± C rearrangement was attempted
following our original protocol.[10a] Thus standard stannyla-
tion, followed by treatment of the resulting crude material
with boron trifluoride diethyl etherate produced the rear-
ranged species 16 and 15 in a combined yield of 85%, and in a
5.5:1 ratio favoring the trans product 16. Subsequent benzy-
16
f
g,h
OBn
O
O
OBn
TBDPSO
HO
17
18
Scheme 3. Synthesis of fragment 18, a precursor for 3. a) allylMgBr,
CuLi2Cl4, (10 mol%), THF, 2 h, À308C (93%); b) O3, CH2Cl2, 15 min,
À788C; PPh3, 14 h, À788C !208C (93%); c) propargyl alcohol, A15,
benzene, 15 min, reflux (quant); d) nBuLi, THF, 30 min, À788C; Bu3SnCl,
30 min, À788C !208C; e) BF3 ´ OEt2, CH2Cl2, 5 min, À108C (85%; two
steps (5.5:1, 16:15)); f) KHMDS, 30 min, À788C, BnBr, THF,
À788C !208C (80%); g) TBAF, THF, 2 h, 208C; h) RaNi, H2, EtOH,
30 min, 208C (89%; two steps). A15 ˆ Amberlyst 15, KHMDS ˆ potassium
hexamethyldisilazide.
Angew. Chem. Int. Ed. 2000, 39, No. 20
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3920-3623 $ 17.50+.50/0
3623

COMMUNICATIONS
08C afforded a mixture of regioisomers (7:3, 23:22), favoring
the desired adduct 23 in overall 89% yield. Pleasingly,
inspection of the crude H NMR spectrum showed that the
BnO
O
OH OH
a-c
d-f
O
BnO
1
O
O
major regioisomer 23 had been formed with a diastereomeric
ratio of over 20:1Ðthus the observed diastereoselection
appeared to be limited only by the original geometry of the
external olefin in the diene precursor. The required diaster-
eomerically pure regioisomer 23 was obtained in 55% yield
following separation by HPLC; the slightly diminished yield
reflecting the propensity for retro-Diels ± Alder reaction on
silica that has been previously documented for these types of
compounds.[17]
19
20
21
g
O
h
N
O
O
O
BnO
BnO
O
22
7
The next synthetic challenge required selective cleavage of
À
the N O bond in the presence of the a,b-unsaturation. This
was crucial in order to avoid translactonization to the other
possible g-lactone species.[18] After screening a variety of
reagents we found that freshly prepared [Mo(CO)3(MeCN)3]
in the presence of water could effect this transformation at
ambient temperature, affording 24 in 70% yield.[19] Subse-
quent silylation of the resultant secondary alcohol followed by
hydrogenation of the a,b-unsaturation over platinum dioxide
proceeded to give 25 as one diastereoisomer in 74% yield
over two steps. The aryl amine was activated as a suitable
leaving group to unmask the butenolide double bond by
treatment with trifluoroacetic anhydride and Hünigꢁs base,
leading to the formation the trifluoroacetamide 26. Prepara-
tion of the vinyl iodide required a further three steps:
hydrogenation of the benzyl ether over Pd(OH)2, then
Swern oxidation followed by one-carbon homologation
according to the Takai procedure.[20] This provided 27 as
a 4:1 ratio of isomers in overall 64% yield (four steps),
favoring the E geometry. Subsequent elimination of the
trifluoroacetamide group was effected by using DBU to
afford 4 in 95% yield with no epimerization of the C-34
methyl substituent.
With all three fragments successfully prepared, the initial
coupling between 2 and 3 was investigated. Swern oxidation of
18 produced aldehyde 2, which was found to be unstable and
hence was not purified, but used crude in the subsequent
coupling step (Scheme 5). Thus, treatment of 10 with butyl-
lithium produced the lithiated alkyne species 27, which was
then quenched by the addition of crude 2 to afford propargylic
alcohol 28 in good yield (72%). However, the reaction
proceeded with poor diastereoselection, which was subse-
quently shown to favor the undesired R epimer at the newly
formed C-16 stereocenter. Attempts to optimize this coupling
proved unsuccessful and ultimately an oxidation ± reduction
protocol was adopted. Consequently, hydrogenation of the
alkyne 28 over Raney nickel allowed complete separation of
the desired S epimer 29. The remaining material was then
recycled by TPAP oxidation,[21] followed by reduction with l-
Selectride; this route gave a 4:1 ratio of products favoring 29,
in 64% yield over three steps. With the final stereogenic
center installed, 29 was prepared for coupling with 4. Thus
silylation of the secondary alcohol then hydrogenation over
Pd(OH)2, allowed Dess ± Martin oxidation to the aldehyde,[22]
followed by one-carbon homologation using the Colvin ±
Gilbert ± Seyferth reagent to afford the terminal alkyne 30
in 62% yield.[23]
i
HN
N
HO
O
O
O
O
BnO
BnO
O
23
24
j,k
O
F3C
N
H
HN
l
TBSO
TBSO
O
O
BnO
BnO
H
O
O
26
25
m-o
O
TBSO
F3C
N
H
p
I
O
TBSO
I
O
O
O
27
4
Scheme 4. Synthesis of fragment 4. a) NaH, BnBr, DMF, 208C (82%);
b) DMSO, (COCl)2, CH2Cl2, À788C; NEt3, À788C !08C; c) (tert-butoxy-
carbonylmethylene)triphenylphosphorane, CH2Cl2, 14h, 08C !208C
(96%; two steps); d) LDA/HMPA/THF, 30 min, À788C, then (S)-2-(tert-
butyldimethylsilyloxy)propanal, À788C !08C; e) MeOH/HCl (sat.),
5 min, (repeat), 208C; f) MsCl, NEt3, CH2Cl2, 30 min, 08C; g) I2
(5 mol%), 4 h, irradiation (60%; four steps); h) PhNO, MeOH/CH2Cl2
(1/1), 14 h, 08C (89% (7:3, 23:22)); i) [Mo(CO)6], MeCN, reflux 4 h; then
23, H2O, 15 min, 208C (70%); j) TBSCl, imidazole, DMF, 14 h, 208C
(80%); k) PtO2 (20 mol%), H2, EtOH, 30 min, 208C (92%); l) TFAA,
Hünigꢁs base, CH2Cl2, 30 min, 08C; m) Pd(OH)2 (20 mol%), H2, MeOH,
2 h, 208C (80%; two steps); n) DMSO, (COCl)2, CH2Cl2, À788C; NEt3,
À788C !08C; o) CrCl2, CHI3, 14 h, THF 08C !208C (80%; two steps);
p) DBU, CH2Cl2, 30 min, 08C (95%). DMF ˆ N,N-dimethylformamide,
LDA ˆ lithium diisopropylamide, HMPA ˆ hexamethyl phosphoramide,
MsCl ˆ methanesulfonyl chloride, TFAA ˆ trifluoroacetic anhydride,
DBU ˆ 1,8-diazobicyclo[5.4.0]undec-7-ene.
7 was then accessed with a ratio of greater than 20:1 by
sunlamp irradiation of 21 in the presence of a catalytic
quantity of iodine.[16]
With 7 in hand, the key HDA reaction was then inves-
tigated. It was found that overnight stirring of a methanol/
dichloromethane (1/1) solution of 7 with nitrosobenzene at
3624
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3920-3624 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2000, 39, No. 20

COMMUNICATIONS
rotation)[25] were in excellent agreement with those reported
for naturally occurring Muricatetrocin C.[7]
Li
HO
OBn
OMe
O
In summary the first stereoselective synthesis of 1 has been
achieved by implementing new methods for each of the three
coupling fragments. Thus our anti-diol building block,
(R',R',R,S)-2,3-BDA-protected butane tetrol 8 has found
further utility in total synthesis, the recently developed
anomeric O ± C rearrangement methodology has been used
to install the 2,5-trans-disubstituted THF moiety 3 and a new
approach to the (S)-hydroxy-butenolide terminus 4 that
employs a highly diastereoselective HDA reaction has been
developed. Further studies aimed at expanding our synthetic
approach to other members of this family of bioactive natural
products are currently underway.
a
O
10
O
OMe
OMe
O
O
27
OMe
28
b
OBn
18
O
2
O
c-e
HO
OBn
TBSO
Received: July 6, 2000 [Z15399]
O
O
OMe
OMe
f-i
O
O
O
O
[1] a) F. Q. Alali, X.-X. Liu, J. L. McLaughlin, J. Nat. Prod. 1999, 62, 504;
b) L. Zeng, Q. Ye, N. H. Oberlies, G. Shi, Z. Gu, K. He, J. L.
McLaughlin, Nat. Prod. Rep. 1996, 13, 275, and references therein.
[2] M. C. Gonzalez, J. R. Tormo, A. Bermejo, M. C. Zafra-polo, E.
Estornell. D. Cortes, Bioorg. Med. Chem. Lett. 1997, 7, 1113, and
references therein.
OMe
OMe
30
29
j
Â
[3] D. J. Morre, R. deCabo, C. Farley, N. H. Oberlies, J. L. McLaughlin,
TBSO
TBSO
Life Sci. 1995, 56, 343.
[4] D. Decaudin, I. Marzo, C. Brenner, G. Kroemer, Int. J. Oncol. 1998,
12, 141.
[5] N. H. Oberlies, V. L. Croy, M. L. Harrison, J. L. McLaughlin, Cancer
Lett. 1997, 15, 7 3.
[6] For recent reviews see: a) G. Cassiraghi, F. Zanardi, L. Battistini, G.
Rassu, G. Aappendino, Chemtracts: Org. Chem. 1998, 11, 803; b) R.
O
O
OMe
O
O
O
OMe
31
Á
Hoppe, H. D. Scharf, Synthesis 1995, 1447; c) B. Figadere, Acc. Chem.
Res. 1995, 28, 359.
k,l
[7] G. Shi, Z. Gu, K. He, K. Wood, L. Zeng, Q. Ye, J. MacDougal, J. L.
McLaughlin, Biorg. Med. Chem. 1996, 4, 1281.
OH
OH
[8] Independently, Koert and co-workers recently adopted a similar
disconnection strategy in their synthesis of the related compounds
Muricatetrocin A and B and Mucocin. a) S. Bäurle, U. Peters, T.
Friedrich, U. Koert, Eur. J. Org. Chem. 2000, 2207; b) S. Bäurle, S.
Hoppen, U. Koert, Angew. Chem. 1999, 111, 1341; Angew. Chem. Int.
Ed. 1999, 38, 1263.
[9] a) D. J. Dixon, A. C. Foster, S. V. Ley, D. J. Reynolds, J. Chem. Soc.
Perkin Trans. 1 1999, 1631. b) D. J. Dixon, A. C. Foster, S. V. Ley, D. J.
Reynolds, J. Chem. Soc. Perkin Trans. 1 1999, 1635.
[10] a) M. F. Buffet, D. J. Dixon, S. V. Ley, E. W. Tate, Synlett 1998, 1091;
b) M. Chorghade, D. J. Dixon, S. V. Ley, D. J. Reynolds, Synth.
Commun. 2000, 30, 1955.
[11] E. J. Corey, P. L. Fuchs, Tetrahedron Lett. 1972, 3769.
[12] The various applications of this useful reagent are reviewed by A. S.
Thompson in Encyclopedia of Reagents for Organic Synthesis, Vol. 3
(Ed.: L. A. Paquette), Wiley, New York, 1995, pp. 1957± 1960. For a
O
O
OH
OH
O
Muricatetrocin C 1
Scheme 5. Coupling reactions and synthesis of 1. a) nBuLi, THF, 30 min,
À788C !08C; À788C then 2; b) DMSO, (COCl)2, CH2Cl2, À788C; NEt3,
À788C !08C (72%; two steps); c) RaNi, H2, EtOH, 30 min, 208C (91%);
d) TPAP (7mol%), NMO, NEt 3, CH2Cl2, 45 min, 08C; e) L-Selectride,
THF, À1008C (70%; two steps); f) TBSCl, DMF, imidazole, 14 h, 458C;
g) Pd(OH)2 (20 mol%), H2, EtOH, 10 h, 208C (87%; two steps); h) DMP,
CH2Cl2, 45 min, 08C !208C; i) diethyl methyldiazophosphonate, tBuOK,
THF, À788C, 16 h; 208C 4 h, (71%; two steps); j) [Pd(PPh3)2Cl2]
(10 mol%), CuI (30 mol%), NEt3, 2 h, 208C (86%); k) [Rh(PPh3)3Cl],
H2, THF, 10 h, 208C (76%); l) TFA/H2O 9/1, 1 min; repeat (82%).
TPAP ˆ tetrapropyl ammonium perruthenate, NMO ˆ 4-methyl-morpho-
line-N-oxide, DMP ˆ Dess ± Martin periodinanne, TFA ˆ trifluoroacetic
acid.
Â
discussion of the reaction mechanism see J.-E. Backväll, M. Sellen, J.
Chem. Soc. Chem. Commun. 1987, 827.
[13] Geometries were obtained in Macromodel using MM2 forcefield.
Calculations of eigenvalues carried out in CADPAC using 3-21‡‡G
basis set. Calculations kindly carried out by A. G. Leach.
[14] M. W. Rathke, D. Sullivan, Tetrahedron Lett. 1972, 4249.
[15] Y. Ito, Y. Kobayashi, T. Kawabata, M. Takosa, S. Terashima,
Tetrahedron 1980, 36, 557.
[16] P. E. Sonnet, Tetrahedron 1980, 36, 557, and references therein.
[17] E. R. Mùller, K. A. Jùrgensen, J. Org. Chem. 1996, 61 5770, and
references therein.
[18] T. R. Hoye, P. R. Hanson, L. E. Hasenwinkel, E. A Ramiraez, Z.
Zhaung, Tetrahedron Lett. 1994, 35, 8525.
[19] a) M. Nitta, T. Kobayashi, J. Chem. Soc. Perkin Trans. 1 1985, 1401.
b) S. Cicchi, A. Goti, A. Brondi, A. Guarna, F. De Sarlo, Tetrahedron
Lett. 1990, 31, 3351.
Sonogashira cross coupling of 30 with the vinyl iodide 4
proceeded smoothly in the presence of catalytic
[Pd(PPh3)2Cl2] and CuI in triethylamine, in 86% yield.[24]
Selective hydrogenation of the resultant enyne 31 was
achieved in 76% yield by using Wilkinsonꢁs catalyst. Final
desilylation and removal of the butane diacetal (BDA) group
could be carried out in one step using aqueous trifluoroacetic
acid. Purification by silica gel chromatography afforded 1 as a
white amorphous solid in 82% yield. The spectroscopic data
for synthetic 1 (1H NMR, 13C NMR, IR, MS, m.p. and specific
Angew. Chem. Int. Ed. 2000, 39, No. 20
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3920-3625 $ 17.50+.50/0
3625

COMMUNICATIONS
blocks to the pharmaceutical industry[1] as well as stereo-
regular macromolecules to the polymer industry.[2] Due to the
complicated mechanistic nature of many transition metal
based catalysts, structure ± activity relationships are often
unpredictable leaving empirical exploration and serendipity
the most common routes to discovery. Although impressive
catalyst breakthroughs have been made, more efficient
strategies clearly must be implemented to aid the pursuit of
new catalysts. Perhaps the most widely heralded approach
that is proposed to influence the discovery and optimization
of new catalysts is combinatorial chemistry.[3±5] Combinatorial
methods have significantly hastened the discovery of new
drugs through the rapid synthesis and efficient screening of
diverse sets (libraries) of organic molecules.[6±8] It seems
reasonable that a similar strategy might impact catalyst
discovery and optimization if metal complex libraries can be
rapidly synthesized and their desired properties tested.[9]
Although the combinatorial approach is often viewed by
some with skepticism, it should be stressed that the more
rapidly new classes of highly selective catalysts are discovered,
the faster traditional chemists can initiate studies to elucidate
their detailed mechanisms of operation. Large collections of
structure ± activity data will not only provide a solid informa-
tion base upon which mechanistic hypotheses can be proposed
and supported, but will also facilitate the development of new
catalyst systems. Herein we report a combinatorial approach
for the discovery of stereoselective polymerization catalysts.
Using this method, we identified a new catalyst system for the
syndiospecific polymerization of propylene.
[20] K. Takai, K. Nitta, K. Utimoto, J. Am. Chem. Soc. 1986, 108, 7408.
[21] S. V. Ley, J. Norman, W. P. Griffith, S. P. Marsden, Synthesis 1994, 639.
[22] a) D. B. Dess, J. C. Martin, J. Org. Chem. 1983, 48, 4155; b) D. B. Dess,
J. C. Martin, J. Am. Chem. Soc. 1991, 113, 7277.
[23] For preparation of the reagent see: R. T. Lewis, W. B. Motherwell,
Tetrahedron 1992, 48, 1465; for a review see: F. Eymery, B. Iorga, P.
Savignac, Synthesis 2000, 185.
[24] a) T. R. Hoye, P. R. Hanson, A. C. Kovelesky, T. D. Ocain, Z. P.
Zhuang, J. Am. Chem. Soc. 1991, 113, 9369; b) K. Sonogashira, Y.
Thoda, N. Magihara, Tetrahedron Lett. 1975, 12, 4467.
[25] Spectroscopic data for synthetic 1, a white amorphous solid: M.p. 65 ±
678C (lit. [7] 65 ± 668C); [a]D29 ˆ ‡5.8 (c ˆ 0.38 in CH2Cl2, lit. [7] ‡6.3
in CH2Cl2); 1H NMR (600 MHz; CDCl3): d ˆ 7.18 (d, 1H, J ˆ 1.1 Hz;
H-33), 5.06 (qd, 1H, J ˆ 6.8, 1.1 Hz; H-34), 3.89 ± 3.87(m, 1H, H-12),
3.87± 3.83 (m, 1H, H-4), 3.82 (q, 1H, J ˆ 7.2 Hz, H-15), 3.63 ± 3.60 (m,
2H; H-19, H-20), 3.47± 3.43 (m, 1H; H-16), 2.90 (s (br), 2H; OH),
2.53 (dt, 1H, J ˆ 15.1, 1.6 Hz; H-3), 2.40 (dd, 1H, J ˆ 15.1, 8.3 Hz;
H-3), 2.04 ± 2.01 (m, 1H; H-13), 2.00 ± 1.97(m, 1H; H-14), 1.70 ± 1.40
(m, 6H; H2-21, H2-18, H2-17), 1.60 ± 1.40 (m, 3H; H-13, H2-11), 1.62 ±
1.55 (m, 1H; H-14) 1.50 ± 1.40 (m, 2H; H2-5), 1.43 (d, 3H, J ˆ 6.8 Hz;
H3-35), 1.40 ± 1.20 (m, 32H; H2-6 !H2-10, H2-22 !H2-31, 2 Â OH),
0.88 (t, 3H; J ˆ 6.9 Hz, H3-32); 13C NMR (150 MHz; CDCl3): d ˆ
174.5 (C-1), 151.7 (C-33), 131.2 (C-2), 81.7 (C-15), 79.3 (C-12), 77.9
(C-34), 74.7, 74.4 (C-20, C-19), 74.3 (C-16), 70.0 (C-4), 37.4 (C-5),
33.5 ± 22.7(C-6 !C-11, C-17, C-18, C-21!C-31), 33.4 (C-3), 32.4 (C-
13), 28.4 (C-14), 19.1 (C-35), 14.1 (C-32); IR (CDCl3): nÄmax ˆ 3422 (br
OH), 2928, 2855 (C H), 1733 (C O) and 1675 (C C) cmÀ1; HRMS
[M‡Na] found: m/z: 619.4522, C35H64O7Na required: m/z: 619.4544.
À
ˆ
ˆ
‡
The development of new polymerization catalysts can be
subdivided into three main steps: ligand preparation, complex
synthesis, and screening of the behavior of these complexes
for a specific reaction. Depending on the catalyst system
under investigation, any one of these can be the rate-
determining step that limits improvement. The synthesis and
testing of catalyst libraries can occur primarily in two formats,
parallel (spatially separate reaction vessels) and pooled
(combined in one reaction vessel) libraries.[10, 11] Although
each format has its advantages, the use of combinatorial
methods for developing enantioselective catalysts for small-
molecule transformations has thus far relied on the parallel
synthesis of ligands and complexes, followed by the serial
screening for enantioselectivity using chiral chromatogra-
phy.[12, 13] Due to the time-consuming nature of sequentially
screening the enantioselectivities of the products of a parallel
library, one might wonder why the screening of a pooled,
bead-bound stereoselective catalyst library has not been
reported. Exchange of products between different beads
and/or the reaction solution occurs, therefore only an average
stereoselectivity of the library can be determined
(Scheme 1).[14]
Development of a Diversity-Based Approach
for the Discovery of Stereoselective
Polymerization Catalysts: Identification of a
Catalyst for the Synthesis of Syndiotactic
Polypropylene**
Jun Tian and Geoffrey W. Coates*
The discovery of efficient and selective catalysts for organic
and polymer synthesis will be a crucial requirement for the
sustained growth of the chemical industry as economic and
environmental constraints become more restrictive in the new
millennium. Increasingly important will be the stereoselective
catalysts that provide key enantiomerically pure building
[*] Prof. G. W. Coates, Dr. J. Tian
Department of Chemistry and Chemical Biology
Baker Laboratory
Cornell University
Ithaca, NY 14853-1301 (USA)
Fax : (‡1)607-255-4137
E-mail: gc39@cornell.edu
[**] This work was supported by the Cornell Center for Materials
Research (CCMR), a Materials Research Science and Engineering
Center of the National Science Foundation (DMR-9632275), and the
Exxon Chemical Corporation. G.W.C. gratefully acknowledges an
NSF Career Award (CHE-9875261), a Camille and Henry Dreyfus
New Faculty Award, a Research Corporation Research Innovation
Award, an Alfred P. Sloan Research Fellowship, an Arnold and Mabel
Beckman Foundation Young Investigator Award, a Camille Dreyfus
Teacher-Scholar Award, a 3M Untenured Faculty Grant, an IBM
Partnership Award, and a Union Carbide Innovation Recognition
Award.
Interestingly, the situation changes in the case of stereo-
selective polymerization catalysts (Scheme 1). Unlike the
asymmetric transformation of small molecules where the
stereochemical events of the reaction are unconnected, the
polymer itself serves as a stereochemical recording of the
events of the polymerization catalyst. Assuming that the
catalyst species do not interact with one another, then a group
of complexes for stereoselective polymerization can be
3626
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3920-3626 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2000, 39, No. 20