Tetraiodobutatriene: A new cumulenic carbon iodide

Article abstract of DOI:10.1002/1521-3773(20020816)41:16<3011::AID-ANIE3011>3.0.CO;2-0

In how many different ways can carbon and iodine combine to form a stable molecule? Here is a new possibility: the first reported iodocumulene. Tetraiodobutatriene (C₄I₄; see picture) is formed spontaneously and in good yield from diiodobutadiyne (C₄I₂). In solution, however, C₄I₄ disproportionates to give hexaiodobutadiene (C₄I₆) and unknown carbon-rich by-products.

Full text of DOI:10.1002/1521-3773(20020816)41:16<3011::AID-ANIE3011>3.0.CO;2-0

COMMUNICATIONS
DeGrado, J. Am. Chem. Soc. 1996, 118, 10337 10338.
[14] G. Cardillo, L. Gentilucci, A. Tolomelli, C. Tomasini, J. Org. Chem.
1998, 63, 2351 2353.
[15] G. A. G. Sulyok, C. Gibson, S. L. Goodman, G. Hˆlzemann, M.
Wiesner, H. Kessler, J. Med. Chem. 2001, 44, 1938 1950.
[16] J. C. Pelletier, M. P. Cava, Synthesis 1987, 474 477.
[17] C. A. Lipinski, F. Lombardo, B. W. Dominy , P. J. Feeney, Adv. Drug
Delivery Rev. 1997, 23, 3 25.
We recently synthesized two longer diiodopolyynes, C6I2
and C8I2,[3] and began exploring methods for the ordered
polymerization of such iodine-capped polyynes. As part of our
studies, we collected X-ray data on a crystal from a previously
pure sample of 1 which had been sitting on a benchtopfor
several months. Instead of the expected butadiyne 1, the
sample contained novel butatriene 3. Evidently, 1 (C4I2) can
disproportionate over time to yield 3 (C4I4), with unknown
carbon-rich by-products. However, the majority of crystals in
the sample still contained 1 instead of 3; the decomposition
reaction had occurred in low overall yield.
[18] J. Perregaard, E. K. Moltzen, E. Meier, C. Sµnchez, J. Med. Chem.
1995, 38, 1998 2008.
[19] Z. Zheng, S. P. Adams, C. L. Ensinger, WO9804247 1998.
Compound 3 is one of only a handful of halogenated
cumulenes that have been characterized. Perfluorobutatriene
5 (Scheme 2) is extremely unstable; it reportedly explodes
above its boiling point of ꢀ58C and decomposes even at
ꢀ808C.[4] Perchlorobutatriene 6 is more stable (m.p. 59
608C), but like many cumulenes it dimerizes to form radialene
8 when heated.[5] Perbromobutatriene 7 has not been report-
ed, and compound 3 is, to the best of our knowledge, the first
known iodine-substituted cumulene of any sort.[6]
Tetraiodobutatriene: A New Cumulenic
Carbon Iodide**
Jeffrey A. Webb, Pei-Hua Liu, Olga L. Malkina, and
Nancy S. Goroff*
Diiodopolyynes such as 1 have attracted our attention as
potential precursors to all-carbon molecules and materials.
Lespieau and Prÿvost reported 75 years ago that diiodobuta-
diyne (1) reacts with molecular iodine to form hexaiodobu-
tadiene (2).[1] Others repeated this experiment with similar
results, each time obtaining a yellow solid that melted at about
1658.[2] We now present evidence that, under appropriate
conditions, this same reaction can instead give cumulene 3,
which is stable, isolable, and fully characterized (Scheme 1).
Cl
Cl
Cl
Cl
Br
Br
Br
Br
F
F
F
•
•
•
•
•
•
F
5
6
7
Cl
Cl
Cl
Cl
Cl
Cl
100 °C
6
Cl
Cl
I
I
8
I
I
I
I
I
I
Scheme 2. Cumulenes 5 and 6 are known; 7 is not.
I
1
3
2
We have now found conditions which lead cleanly to 3 in
good yield. Iodination of 1 in concentrated solution in hexanes
produces a yellow precipitate after only a few minutes.[7]
Filtration separates the solid from the solvent and any
remaining I2, leaving a yellow powder (m.p. 142 1438C). As
discussed below, mass spectrometry, IR and NMR spectro-
scopy, and X-ray diffraction studies all confirm that this
material is cumulene 3.
I
I
I
I
•
•
I
I
I
4
Scheme 1. A family of carbon iodides C4Ix.
[*] Prof. Dr. N. S. Goroff, J. A. Webb, Dr. P.-H. Liu
Department of Chemistry, State University of New York
Stony Brook, NY 11794-3400 (USA)
Fax : (þ 1)631-632-8356
Without careful handling, however, compound 3 rapidly
decomposes in solution to give 2. The decomposition reaction
occurs in pure solutions of 3, suggesting a disproportionation
mechanism, much like the observed decomposition of 1. The
simplest disproportionation of 3, to give equal parts of 1 and 2,
would be endothermic by 9.1 kcalmolꢀ1, according to density-
functional (B3LYP/LanL2DZ) calculations.[8] Furthermore,
we found no evidence for the formation of 1 during the
decomposition reaction. Thus far, we have been unable to
identify the by-product(s) of this reaction.
The decomposition of 3 can be monitored by thin-layer
chromatography (TLC; AlO3/hexanes).[9] When 3 is dissolved
in a variety of solvents (e.g., THF, CS2), the diyne disappears
over several minutes, while a new TLC spot appears simulta-
neously, corresponding to 2. In some solvents (acetone,
benzene, CCl4), TLC indicates that 3 is stable unless exposed
E-mail: nancy.goroff@sunysb.edu
Dr. O. L. Malkina
Institute of Inorganic Chemistry and Computing Center
Slovak Academy of Sciences
Bratislava (Slovakia)
[**] We thank J. Lauher for crystallography assistance, D. Salahub for use
of the DeMon program at Stony Brook, and V. Malkin and M. Kaupp
for helpful discussions. We also thank the Computing Center of the
Slovak Academy of Sciences for computational resources. We
acknowledge the National Science Foundation (CHE-9984937), the
donors of the Petroleum Research Fund, administered by the
American Chemical Society, and the Slovak Grant Agency VEGA
(O.L.M., Grant No.2/7203/20) for support of this research.
Supporting information for this article is available on the WWW under
http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2002, 41, No. 16
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4116-3011 $ 20.00+.50/0
3011

COMMUNICATIONS
to light. In pyridine, compound 3 appears to form a stable
solvent solute complex and does not decompose.
The instability of 3 in many solvents makes spectroscopic
characterization difficult. GC/MS experiments on the crude
yellow precipitate indicate that it contains only one com-
pound, with molecular formula C4I4, consistent with either
ꢀ48.9 ppm. The discrepancy most likely results, at least in
part, from a Lewis acid base interaction between 3 and the
solvent.[3,16] The ꢁ 12-ppm difference in the low-frequency
peak in the spectra when in pyridine or in weakly basic
cyclohexanone supports this hypothesis.
The NMR spectroscopic data contain no evidence of enyne
4. Furthermore, vibrational spectroscopy confirms the struc-
ture of 3. The extremely simple experimental IR spectrum
matches calculations well: 668 and 1678 cmꢀ1 (calculated)
compared to 603 and 1616 cmꢀ1 (experimental). We have been
unable to obtain Raman or UV spectra of 3 because of its
swift decomposition upon irradiation in solution.
As final proof of the structure, we obtained X-ray
diffraction data on a new crystal grown from the yellow
powder. These data confirm that the yellow solid is tetraio-
dobutatriene (3), just like the initial material from decom-
position of 1. The two samples, prepared in very different
ways, have indistinguishable crystal structures. The crystal
packing is shown in Figure 2a. Figure 2b shows a crystal
structure of 2, prepared directly from 1 and then cocrystal-
lized with 1,7-phenanthroline.[17]
cumulene
3 or enyne 4. Density-functional (B3LYP/
LanL2DZ) calculations place compound 4 higher in energy
than 3 by 4.5 kcalmolꢀ1, but the B3LYP functional is known to
overestimate the stability of cumulenes relative to acetylenic
isomers.[10] Previous literature reports contain examples of
both 1,4- and 1,2-addition to diynes, and often mixtures of the
two.[11,12]
13C NMR spectroscopy should distinguish between 3 and 4
easily, but initial NMR spectroscopic experiments were
difficult to interpret. Figure 1a shows the 13C NMR spectrum
of starting diyne 1, whereas Figures 1b d show the NMR
spectra of the yellow precipitate dissolved in various solvents.
As a solid, cumulene 3 is more stable than diyne 1. Upon
heating, compound 1 decomposes at 908C, but 3 melts at 142
Figure 1. 13C NMR spectra; fl denotes sample signal, denotes solvent or
*
deuterium lock. a) 1 in [D6]DMSO; b) 3, decomposed, in [D6]DMSO; c) 3
in [D5]pyridine; d) 3 in cyclohexanone with internal [D12]cyclohexane lock.
The spectrum in Figure 1b was obtained first, with dimethyl
sulfoxide (DMSO) as solvent. Initially, based on symmetry, we
assigned the two peaks in this spectrum at d ¼ 37.7 and
121.2 ppm to compound 3.
To confirm the assignment, we calculated the 13C NMR
spectrum of 3, including the spin-orbit coupling effect of the
15]
iodine atoms.[13 These calculations predict chemical shifts
for C1 and C2 at d ¼ ꢀ48.9 and 163.2 ppm, respectively, far
from the observed spectrum in DMSO. This difference
provided our first evidence for the decomposition of 3 in
solution.
Figures 1c and d show the NMR spectra of freshly prepared
samples of the yellow precipitate, dissolved in pyridine and in
cyclohexanone, respectively. Each spectrum includes two
peaks: d ¼ 163.5 and ꢀ14.9 ppm in pyridine and d ¼ 162.9
and ꢀ27.5 ppm in cyclohexanone. We assign both these
spectra to compound 3, despite the difference between the
lower-frequency shifts and the calculated value of d ¼
Figure 2. a) X-ray crystal structure of 3; b) X-ray structure of 1:1 cocrystal
of 2 and 1,7-phenanthroline.
3012
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4116-3012 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 16

COMMUNICATIONS
[1] M. M. Lespieau, C. Prÿvost, Mem. Soc. Frib. Sci. Nat. Chim. 1925, 704.
[2] a) F. Straus, L. Kollek, Ber. Dtsch. Chem. Ges. B 1926, 59, 1664; b) F.
Straus, L. Kollek, H. Hauptmann, Ber. Dtsch. Chem. Ges. B 1930, 63,
1886.
[3] K. Gao, N. S. Goroff, J. Am. Chem. Soc. 2000, 122, 9320.
[4] a) E. L. Martin, W. H. Sharkey, J. Am. Chem. Soc. 1959, 81, 5256;
b) A. Bach, D. Lentz, P. Luger, M. Messerschmidt, C. Olesch, M.
Patzschke, Angew. Chem. 2002, 114, 311; Angew. Chem. Int. Ed. 2002,
41, 296.
1438C. We have produced 3 in gram quantities without
difficulty and can keepsolid samples in the freezer indef-
initely. Nonetheless, compound 3 decomposes much more
readily in solution than 1.
The weakness of carbon iodine bonds makes compound 3 a
promising precursor to all-carbon and carbon-rich molecules
and materials. Compound 3 also offers another Lewis acid for
the crystal engineer©s toolbox. Carbon iodides have previously
found use as components in designed materials by acting as
Lewis acid templates. For example, diiodoacetylene and other
carbon iodides can determine the spacing of ions in multi-
component conducting solids.[18] The stability of 3 in pyridine
solution indicates that it can survive Lewis acid base inter-
actions without decomposition, making it a suitable candidate
for such multicomponent systems.
[5] B. Heinrich, A. Roedig, Angew. Chem. 1968, 80, 367; Angew. Chem.
Int. Ed. Engl. 1968, 7, 375.
[6] Although no iodocumulenes are reported in the literature, several
iodoallenes are known, including tetraiodoallene, C3I4; see: a) F. Kai,
S. Seki, Chem. Pharm. Bull. 1965, 13, 1374; b) F. Kai, S. Seki, Chem.
Pharm. Bull. 1966, 14, 1122; c) A. M. Snider, P. F. Krause, F. A. Miller,
J. Phys. Chem. 1976, 80, 1262.
[7] No precipitate is formed in solvents in which 3 is moderately soluble,
and the isolated product is 2.
[8] a) A. D. Becke, J. Chem. Phys. 1996, 104, 1040; b) Gaussian98
(RevisionA.5), M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A.
Montgomery, R. E. Stratmann, J. C. Burant, S. Dapprich, J. M.
Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi,
V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo,
S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q. Cui, K.
Morokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B.
Foresman, J. Cioslowski, J. V. Ortiz, B. B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin, D. J.
Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, C.
Gonzalez, M. Challacombe, P. M. W. Gill, B. G. Johnson, W. Chen,
M. W. Wong, J. L. Andres, M. Head-Gordon, E. S. Replogle, J. A.
Pople, Gaussian, Inc., Pittsburgh, PA, 1998.
[9] Detailed TLC data are available in the Supporting Information.
[10] D. A. Plattner, K. N. Houk, J. Am. Chem. Soc. 1995, 117, 4405.
[11] 1,2-Addition of halogens: a) Y. I. Porfir©eva, A. A. Petrov, L. B.
Sokolov, Russ. J. Gen. Chem. 1964, 34, 1884; Y. I. Porfir©eva, A. A.
Petrov, L. B. Sokolov, Zh. Obshch. Khim. 1964, 35, 1714; b) Y. I.
Porfir©eva, E. S. Turbanova, A. A. Petrov, Russ. J. Gen. Chem. 1964,
34, 4026; Y. I. Porfir©eva, E. S. Turbanova, A. A. Petrov, Zh. Obshch.
Khim. 1964, 34, 3966; c) L. A. Dorofeeva, Y. I. Porfir©eva, A. A.
Petrov, Russ. J. Org. Chem. 1972, 8, 2049; L. A. Dorofeeva, Y. I.
Porfir©eva, A. A. Petrov, Zh. Org. Khim. 1972, 8, 2002; d) L. A.
Dorofeeva, Y. I. Porfir©eva, A. A. Petrov, Russ. J. Org. Chem. 1973, 9,
454; L. A. Dorofeeva, Y. I. Porfir©eva, A. A. Petrov, Zh. Org. Khim.
1973, 9, 449; e) J. J. Wright, M. S. Puar, B. Pramanik, A. Fishman, J.
Chem. Soc. Chem. Commun. 1988, 413; 1,4-addition or mixtures: f) V.
Grignard, Tchÿoufaki, Recl. Trav. Chim. Pays-Bas 1929, 48, 899;
g) B. G. Shakhovskoi, M. D. Stadnichuk, A. A. Petrov, Russ. J. Gen.
Chem. 1964, 35, 1715; B. G. Shakhovskoi, M. D. Stadnichuk, A. A.
Petrov, Zh. Obshch. Khim. 1964, 35, 1714.
Cumulene 3 is surprisingly easy to prepare and to isolate in
a pure form, is surprisingly stable in the solid state, and
disproportionates surprisingly readily in solution. It is one
more example of the unusual chemistry of carbon and
iodine.
ExperimentalSection
3: I2 (340 mg, 1.32 mmol) was added to a solution of 1 (400 mg, 1.32 mmol)
in hexanes (5 mL). Aluminum foil was used to wrapthe reaction vessel.
The reaction mixture was stirred vigorously at room temperature for
15 min, forming a yellow precipitate. The yellow powder was filtered and
washed with hexanes to yield crude product (604 mg, 82%). The crude
products from three iterations were combined and recrystallized from
cyclohexanone/hexanes to yield pure 3 as fine yellow needles (1.47 g, 67%
total). M.p. 143 1448C; MS (EI): m/z (%): 556 (24) [Mþ (C4I4)], 429 (45)
[MþꢀI], 302 (100) [Mþꢀ2(I)], 175 (28) [Mþꢀ3(I)], 127 (27) [Iþ].
2: Iodine (2.52 g, 9.93 mmol) was added to a solution of 1 (1.0 g, 3.3 mmol)
in methanol (30 mL). After stirring at room temperature for 2 h, the
precipitate was filtered and washed with hexanes to yield 2 as a yellow
powder. The material left in solution was not recovered, and no yield was
determined. M.p. 165 1668C; MS (EI)[19]: m/z (%): 682 (22) [MþꢀI], 556
(34) [Mþꢀ2(I)], 429 (34) [Mþꢀ3(I)], 302 (100) [Mþꢀ4(I)], 175 (38)
[Mþꢀ5(I)], 127 (50) [Iþ]. To prepare cocrystals of 2 and 1,7-phenanthroline,
compound 2 (0.14 g, 0.25 mmol) was dissolved in acetone (3 mL). A
solution of 1,7-phenanthroline (0.045 g, 0.25 mmol) in methanol (2 mL)
was added dropwise to this solution with stirring. The solution, in a small
vial, was cooled to ꢀ108C, forming small crystals.
X-ray structure analysis: X-ray intensity data were measured on
a
[12] Other 1,4-additions, without halogens: a) Y. Morimoto, Y. Higuchi, K.
Wakamatsu, K. Oshima, K. Utimoto, N. Yasuoka, Bull. Chem. Soc.
Jpn. 1989, 62, 639; b) T. Kusumoto, T. Hiyama, Bull. Chem. Soc. Jpn.
1990, 63, 3103; c) G. Dyker, S. Borowski, G. Henkel, A. Kellner, I.
Dix, P. G. Jones, Tetrahedron Lett. 2000, 41, 8259.
[13] NMR spectroscopic calculations carried out in the deMon-NMR
program: a) V. G. Malkin, O. L. Malkina, D. R. Salahub, Chem. Phys.
Lett. 1996, 261, 335; b) A. Stamant, D. R. Salahub, Chem. Phys. Lett.
1990, 169, 387; c) D. R. Salahub, R. Fournier, P. Mlynarski, I. Papai, A.
St-Amant, J. Ushio in Density FunctionalMethods in Chemistry (Eds.:
J. Labanowski, J. Andzelm), Springer, New York, 1991.
[14] Spin-orbit coupling contribution, by sum-over-states density-func-
tional perturbation theory (SOS-DFPT), AMFI(1e þ 2e)/F64/IGLO-
II: a) M. Kaupp, O. L. Malkina, V. G. Malkin, P. Pyykkˆ, Chem. Eur. J.
1998, 4, 118; b) M. Kaupp, O. L. Malkina, V. G. Malkin, J. Comput.
Chem. 1999, 20, 1304; c) O. L. Malkina, B. Schimmelpfennig, K.
Kaupp, B. A. Hess, P. Chandra, U. Wahlgren, V. G. Malkin, Chem.
Phys. Lett. 1998, 296, 93.
[15] Scalar contribution to chemical shift: Perdew-86/IGLO-II þ ECP;
a) J. P. Perdew, Phys. Rev. B 1986, 33, 8822; b) J. P. Perdew, Y. Wang,
Phys. Rev. B 1986, 33, 8800; c) W. Kutzelnigg, U. Fleischer, M.
Schindler, NMR: Basic Princ. Prog. 1990, 23, 165; d) M. Dolg, Ph.D.
Bruker AXS diffractometer. The structures were solved by direct methods
and refined by using full-matrix least-squares methods (SHELX97).[20]
Empirical absorption corrections were applied. Compound 3 crystallizes
in space group P21/n, with a ¼ 4.4362(4) ä, b ¼ 17.1459(14) ä, c ¼
12.5217(10) ä,
b ¼ 96.716(2)8,
V¼ 945.90(14) ä3,
Z ¼ 2,
1calcd
¼
4.877 gcmꢀ3, T¼ 293(2) K, 4223 reflections measured, 1352 crystallograph-
ically independent (Rint ¼ 0.068), and 1154 reflections with I > 2s(I),
MoKa ¼ 0.71073 ä, 2qmax ¼ 538, R(Fobs) ¼ 0.034, wR(F2)all ¼ 0.113. Com-
ꢀ
pound 2 forms cocrystals with [1,7]-phenanthroline in the space group P1,
with a ¼ 7.4475(17) ä, b ¼ 10.609(2) ä, c ¼ 14.845(3) ä, a ¼ 103.128(4)8,
b ¼ 103.840(4)8,
g ¼ 91.510(4)8,
V¼ 1105.0(4) ä3,
Z ¼ 2,
1calcd
¼
2.974 gcmꢀ3, T¼ 293(2) K, 6417 reflections measured, 4390 crystallograph-
ically independent (Rint ¼ 0.174), and 3591 reflections with I > 2s(I),
MoKa ¼ 0.71073 ä, 2qmax ¼ 538, R(Fobs) ¼ 0.120, wR(F2)all ¼ 0.379. Unit cells
and ORTEP representations are available as supplementary material.
CCDC-181226 and 181227 contain the supplementary crystallographic
data for this paper. These data can be obtained free of charge via
www.ccdc.cam.ac.uk/conts/retrieving.html (or from the Cambridge Crys-
tallographic Data Centre, 12, Union Road, Cambridge CB21EZ, UK; fax:
(þ 44)1223-336-033; or deposit@ccdc.cam.ac.uk).
Received: March 12, 2002 [Z18877]
Angew. Chem. Int. Ed. 2002, 41, No. 16
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4116-3013 $ 20.00+.50/0
3013

COMMUNICATIONS
thesis, Universit‰t Stuttgart, 1989; e) D. Andrae, U. H‰ussermann, M.
Dolg, H. Stoll, H. Preuss, Theor. Chim. Acta 1990, 77, 123.
[16] P. D. Rege, O. L. Malkina, N. S. Goroff, J. Am. Chem. Soc. 2002, 124,
371.
[17] Our experience with Lewis acidic iodides has shown that 1,7-
phenanthroline is an effective cocrystallizing agent, which forms
robust, easily analyzed crystals. We have been unable to isolate
crystals of 2 that are suitable for X-ray crystallography.
[18] a) S. Uji, C. Terakura, T. Terashima, H. Aoki, H. M. Yamamoto, J.
Yamaura, R. Kato, Synth. Met. 1999, 103, 1978; b) H. M. Yamamoto,
J. I. Yamaura, R. Kato, Synth. Met. 1999, 102, 1448; c) H. M.
Yamamoto, J. Yamaura, R. Kato, Synth. Met. 1999, 102, 1515;
d) H. M. Yamamoto, J. I. Yamaura, R. Kato, J. Am. Chem. Soc.
1998, 120, 5905; e) H. M. Yamamoto, J. I. Yamaura, R. Kato, J. Mater.
Chem. 1998, 8, 15.
¼
two C C bonds in the same carbon chain exclusively under-
went 6-endo and then 5-exo cyclization for 1a and 6-exo
cyclization for 1b in the presence of a Lewis acid to give the
fused ring products 6,5-cis-2a and 6,6-trans-2b, respectively
(Table 1, entries 1 4). In each case, four stereocenters of the
cyclization products 2a/2b were set upin one step. Further-
more, as little as 0.3 equivalents of Yb(OTf)3 was sufficient to
catalyze the cyclization of 1a to give 2a in 69% yield (Table 1,
entry 2), and a decrease in the amount of Mg(ClO4)2 (to
0.5 equiv) did not decrease the yield of 2b in the cyclization of
1b (Table 1, entry 4). For substrates 1c–e,[5] which have an
allyl substituent in the a-position, 6-endo (6-exo for substrate
1d) followed by 5-exo cyclization gave the corresponding
bicyclic products 2c e in moderate yields in the presence of a
stoichiometric amount of Mg(ClO4)2 (Table 1, entries 5 7).
Furthermore, triple cyclization of 1 f gave product 2 f as one of
16 possible stereoisomers in 26% yield (Table 1, entry 8).
When Yb(OTf)3 was used as the Lewis acid catalyst, no more
than 16% yield can be obtained in the cyclization of
substrates 1b f. For all the substrates investigated, no atom-
transfer radical cyclization was observed in the absence of
Lewis acids. These results demonstrate that the Lewis acids
not only promote tandem radical cyclization reactions, but
also control the stereoselectivities of the cyclization prod-
ucts.[6] The major side products of those tandem cyclization
reactions were the corresponding reductive debromination
products (< 12% yield), monocyclization products (< 30%
yield), and even a bicyclization product (for 1 f only; 12%
yield).[7] Especially for substrate 1b, the 1,3-diaxial interaction
of two methyl groups made the second ring cyclization
unfavorable, and as a result the radical chain process was
partially terminated in the monocyclization stage with the
isolation of a monocyclization product in up to 30% yield.
Then we investigated the enantioselective tandem cycliza-
tion of 1a and 1b by using Mg(ClO4)2 and bisoxazoline ligand
L1 as a chiral catalyst (Table 2). Poor ee values were observed
in the cyclization of 1a (Table 2, entry 1). Although the
addition of activated 4-ä molecular sieves[4,8] improved the ee
value, it decreased the yield dramatically (Table 2, entry 2).
High ee values (82 84%) were observed in the cyclization of
1b at higher temperature, despite lower yields (Table 2,
entries 3 and 4). With Yb(OTf)3 as the catalyst,[9] the effects of
chiral ligands,[10] additives,[11] and solvents on the enantiose-
lective tandem cyclization of 1a were evaluated. As shown in
Table 3, the L3/Yb complex gave the best results (60% yield
and 66% ee; Table 3, entry 4). Dichloromethane was found to
be a better solvent than toluene in this tandem cyclization
system (Table 3, entry 4 vs. 6). The addition of Et2O or water
had negative effects on these reactions (Table 3, entries 2 and
3). Surprisingly, the addition of molecular sieves led to
reversed enantiofacial selectivity of the cyclization[11e] (Ta-
ble 3, entry 5).
[19] Mass spectra measured on an Agilent 5973 spectrometer. The mass
limit of the instrument (m/zmax ¼ 800) does not allow the observation
of the parent ion of 2 with MW ¼ 810.
[20] SHELX97–Programs for Crystal Structure Analysis (Release 97-2):
G. M. Sheldrick, SHELXS-98, Program for the Solution of Crystal
Structures, University of Gˆttingen, Gˆttingen (Germany), 1998.
Atom-Transfer Tandem Radical Cyclization
Reactions Promoted by Lewis Acids**
Dan Yang,* Shen Gu, Yi-Long Yan, Hong-Wu Zhao,
and Nian-Yong Zhu
Tandem radical cyclization reactions are powerful methods
for the synthesis of a wide range of highly functionalized
polycyclic natural products as multiple stereocenters can be
[1,2]
constructed in one stepunder mild conditions.
However,
no enantioselective tandem radical cyclization reactions have
been reported.[3] Herein we describe a Lewis acid catalyzed
atom-transfer tandem radical cyclization method for the
formation of various bicyclic and tricylic ring skeletons.
Furthermore, the first enantioselective tandem radical cycli-
zation was achieved by using chiral Lewis acids as catalysts.
We recently reported an enantioselective atom-transfer
radical monocyclization catalyzed by a chiral Lewis acid to
prepare cyclic 2,3-disubstituted ketones.[4] Yb(OTf)3 (OTf ¼
trifluoromethanesulfonate) and Mg(ClO4)2 were found to be
the best Lewis acids in these monocyclization reactions.
Therefore, the two Lewis acids were further used to promote
tandem cyclization of a-bromo b-keto esters 1a f with Et3B/
O2 as the radical initiator (Table 1). Substrates 1a and 1b with
[*] Prof. D. Yang, S. Gu, Y.-L. Yan, H.-W. Zhao, N.-Y. Zhu
Department of Chemistry, The University of Hong Kong
Pokfulam Road, Hong Kong (P. R. China)
Fax : (þ 852)2859-2159
E-mail: yangdan@hku.hk
The stereochemistry of tandem radical cyclization reactions
in the presence of the chiral complex [Yb(Ph-pybox)(OTf)3]
(pybox ¼ 2,6-bis(2-oxazolin-2-yl)pyridine) can be explained
using the transition-state model shown in Scheme 1. Similar to
the chiral complex [Sc(Ph-pybox)(ethyl glyoxylate)(OTf)]2þ
proposed by Evans et al.,[12] the ytterbium center adopts a
square-pyramidal geometry with the ester carbonyl group
[**] This work was supported by The University of Hong Kong and the
Hong Kong Research Grants Council. D.Y. acknowledges the Bristol-
Myers Squibb Foundation for an Unrestricted Grant in Synthetic
Organic Chemistry and the Croucher Foundation for a Croucher
Senior Research Fellowship.
Supporting information for this article is available on the WWW under
http://www.angewandte.org or from the author.
3014
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4116-3014 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 16