Asymmetric allenophanes: Synthesis of a tris-meta-allenophane and tetrakis-meta-allenophane by sequential cross-coupling

Article abstract of DOI:10.1002/anie.200703003

Around and around: A strategy based on sequential Pd-catalyzed cross-coupling reactions was applied for the asymmetric synthesis of new macrocyclic meta-allenophanes composed of 18- or 24-membered rings (see structures). The chiral components, tertiary propargyl alcohols and allene bridges, were assembled by a general enantioselective protocol, which involved a Sharpless epoxidation, an oxidation, and Sonogashira cross-coupling. (Chemical Equation Presented).

Full text of DOI:10.1002/anie.200703003

Communications
DOI: 10.1002/anie.200703003
Asymmetric Synthesis
Asymmetric Allenophanes: Synthesis of a Tris-meta-allenophane and
Tetrakis-meta-allenophane by Sequential Cross-Coupling**
Mathieu Lecl›re and Alex G. Fallis*
The synthesis and design of new cyclophanes and assorted
cage compounds with novel shapes and supramolecular
geometries continue to be topics of current interest.[1] Our
research group has an expanding interest in the design and
synthesis of new carbocyclic[2] and heterocyclic[3] cyclophanes,
many of which contain a twisted conformation that imparts
helical chirality[4] to the assembled molecule. The extent of
helical chirality is a consequence of the number, combination,
and type of unsaturated linkages in these molecules.[1g,5]
Allenes are a unique family of organic molecules that
participate in a diverse range of reactions. Other attributes
include their axial chirality and synthetic versatility.[6] The
study of allenic natural products and their total synthesis
continues to expand,[7] while medical applications include the
inactivation of monoamine oxidase.[8] Recently we described
the asymmetric synthesis of a bis(alleno)bis(butadiynyl)meta-
Figure 1. Allenophane macrocycles 1, 2, and 3 containing 12-, 18-, and
24-membered rings, respectively.
cyclophane.[9] This compound represented a new structural
family of cyclophanes composed of both allene and acetylene
bridges. These macrocycles possess both axial and helical
chirality and thus are doubly chiral.[10] True allenophanes
(containing only allene bridges; 1,2-propadiene components)
are of great interest owing to their potential structural and
chiroptical properties. In addition, with the appropriate
substitution patterns, these molecules should be useful as
chiral ligands or as hosts for metal ions and small guest
molecules.[11]
A limited number of general synthetic methods are
available for constructing allenophanes. Previous examples
include para-allenophanes prepared as a mixture of diaste-
reomers,[12] and more recently para-acetylenic allenophanes
were prepared in which a mixture of enantiomers was
resolved and purified by HPLC.[13] The generic structure 1
(Figure 1) represents the simplest member of the symmetrical
meta-allenophane family (n = 1), in which R represents
various functional groups (hydrogen, alkyl, aryl), R1 is a
solubilizing group, and the number of allene bridges may also
be varied.
Our research group[14a,b] and others[14c] have previously
observed that the termini separation of the reactive compo-
nents for the cyclization of cyclophanes dictates whether the
intra- or intermolecular product dominates. For alkynes,
molecular modeling studies indicated that the intermolecular
product is favored when the termini separation is greater than
7 Š. Our initial synthetic strategy was based on the expect-
ation that the relatively rigid and restricted geometry
provided by a substituted allene with adjacent reactive
substituents (13, Scheme 3) should facilitate cyclization. The
calculated distance between the bromobenzene and ethynyl
substituents (termini) was 4.3 Š and thus supported this
view.[14d] Herein we describe the synthesis of two higher
homologues of 1, namely the allenophanes 2 (n = 2) and 3
(n = 3).
The difficulties encountered with literature methods for
the synthesis of the allene units in our substrates have been
previously described.[9] Consequently, we have devised a
versatile new protocol to synthesize allenophanes using
Sharpless epoxidation to give asymmetric tertiary propargyl
alcohol intermediates (Scheme 1). We then applied the
reliable SN2’ addition of an organocuprate to a propargyl
acetate to prepare the fully substituted allenes.[15] The
[*] Dr. M. Lecl›re, Prof. Dr. A. G. Fallis
Department of Chemistry
University of Ottawa
10 Marie Curie, Ottawa, ON, K1N 6N5 (Canada)
Fax: (1)613-562-5170
E-mail: afallis@uottawa.ca
Homepage: http://www.science.uottawa.ca/~afallis/
[**] We are grateful to the Natural Sciences and Engineering Research
Council of Canada (NSERC) for financial support. We thank L.-L. Tay
(NRC, Ottawa) for the Raman spectra, also K. Fagnou, T. Moon, and
S. Gorelsky for the DFTcalculations, and M. Frenette for CD spectral
measurements.
Scheme 1. An enantioselective route to synthesize tertiary propargyl
alcohols.
568
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 568 –572

Angewandte
Chemie
requisite bromoalkynyl epoxide 8 was prepared by this
method. L-Selectride was the most selective hydride source
(of the several examined) to open the epoxide and obtain
propargyl alcohol 9 without racemization (Scheme 2). Pro-
tection of the reactive hydroxy groups with silyl groups and
conversion of the bromine substituent into iodide afforded
the requisite building blocks 9, 10a, and 10b for conversion to
the designed allenophanes.
Scheme 2. Synthesis of building blocks 9, 10a, and 10b. Conditions
and reagents: a) L-Selectride (1.2 equiv), THF, 08C, 20 min, 96%;
b) TBSOTf (2.0 equiv), 2,6-lutidine, CH2Cl2, 0!228C, 12 h, 99%;
c) LDA (1.1 equiv), THF, 08C, 5 min, then TMSCl (2.0 equiv), 0!
228C, 15 min; d) tBuLi (2.5 equiv), THF, ꢀ788C, 3 min, then I2
(1.5 equiv), 5 min, ꢀ78!228C, 91%. LDA=lithium diisopropylamide,
L-Selectride=lithium tri-sec-butylborohydride, TBS=tert-butyldimethyl-
silyl, Tf=trifluoromethanesulfonyl, TMS=trimethylsilyl.
Scheme 3. Synthetic route to allenophane precursor 13. Conditions
and reagents: a) [PdCl2(PPh3)2] (0.05 equiv), CuI (0.1 equiv), Et3N,
THF, 668C, 16 h, 80%, d.r. 6:1; Path A: b) Ac2O, DMAP (0.1 equiv),
pyridine, 228C, 12 h, 99%; c) MeMgBr (18 equiv), CuI (18 equiv), LiBr
(18 equiv), 08C, 2 h, 76%; d) tBuLi (2.5 equiv), THF, I2 (1.5 equiv),
ꢀ78!228C, 10 min, 92%; or Path B: e) TMSCl (2.0 equiv), Et3N,
DMAP (0.1 equiv), 228C, 6 h, 93%; f) tBuLi (2.5 equiv), ꢀ788C, THF,
I2 (1.5 equiv), ꢀ78!228C, 10 min; g) Dowex 50X-400 H+, 228C, 2 h,
86% (over 2 steps); h) Ac2O, DMAP (0.1 equiv), pyridine 228C, 12 h,
99%; i) MeMgBr (18 equiv), CuI (18 equiv), LiBr (18 equiv), 08C, 2 h,
76%; j) tBuLi (2.5 equiv), THF, I2 (1.5 equiv), ꢀ78!228C, 10 min,
92%; k) 9 (1.2 equiv), [PdCl2(PPh3)2] (0.05 equiv), CuI (0.1 equiv), Et3N,
THF, reflux, 12 h, 51%. DMAP=4-dimethylaminopyridine.
Considerable experimentation was required to efficiently
effect the palladium(0)-mediated Sonogashira cross-coupling
of 9 and 10b to give 11. The best results were obtained for the
reaction with a slight excess of the silyl ether 10b (1.2 equiv)
in refluxing THF for 16 h (80% yield). Allene 12 was initially
generated from the acetate derivative of alcohol 11, which
was then exposed to an organocopper reagent. Unfortunately,
the basic conditions required for the bromine–iodine
exchange resulted in some racemization because of deproto-
nation of an allenic methyl group. An alternative route
(Path A, Scheme 3) involved protection of the alcohol with
acetic anhydride in pyridine and subsequent cuprate addition
then exposure to tBuLi and iodine to afford 12. Another route
(Path B, Scheme 3), employed silylation, halogen exchange,
ether cleavage with acidic resin, and allene formation to also
give 12.
Palladium(0) coupling of 12 with the alcohol 9 generated
the penultimate cyclization precursor 13 in 51% yield.
However, an isomeric mixture resulted from the competitive,
reversible formation of a palladium p-allyl intermediate
which destroyed the integrity of the allene chirality. Thus an
alternative route was selected to assemble the three chiral
units before ring closure (Scheme 4). Owing to the potential
flexibility of 17 (termini separation: 7.1 Š), the intramolec-
ular product would be expected. However, this approach
would require the allene-formation reactions to be conducted
three times on the same molecule.
High-dilution conditions (0.003m) were necessary for the
intramolecular coupling to give the desired triyne macrocycle
18 in 45% yield. Sequential deprotection and acetylation
generated a triacetate derivative, which was subjected to the
cuprate allene protocol described above. This sequence
proceeded smoothly in 83% yield to give the novel tris-
meta-allenophane 2.
Significantly, the intermediate TMS-protected acetylene
14 represents a synthetic precursor that comprises half of the
tetrakis-meta-allenophane target 3 (Figure 1) and the bis-
acetylene 15 could supply the second half of the tetra-
acetylene component required to construct the marcocycle.
With access to readily available precursors and the successful
synthesis of 2, we investigated the preparation of the higher
homologue 3. Cleavage of the TMS group of 14 was carried
out by exposure to base to afford the free acetylene 19, which
underwent palladium(0)-mediated cross-coupling when com-
bined with the bis-acetylene 15 to generate 20 in 73% yield
(Scheme 5). The iodoacetylene 21 was generated by halogen
exchange with 14 as described above. Intramolecular Sono-
gashira cyclization of 21 under high-dilution conditions
(0.003m) afforded the desired allene precursor 22 in 42%
yield. Removal of the silyl protecting groups and subsequent
introduction of four acetate groups proceeded in a direct
manner to give the asymmetric macrocyclic tetraalkyne 23.
The final stage of the synthesis used the cuprate allene
protocol to give tetrakis-meta-allenophane 3. The direct
The Sonogashira cross-coupling of 10a and 10b pro-
ceeded to give the diyne 14 in excellent yield (91%) but
required 16 h at reflux to reach completion. Base-mediated
halogen exchange and additional Sonogashira cross-coupling
with 10a gave the triyne 16. Final halogen exchange and
removal of the trimethylsilyl protecting group afforded 17.
Angew. Chem. Int. Ed. 2008, 47, 568 –572
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
569

Communications
Scheme 4. Final steps for the synthesis of (ꢀ)-allenophane 2. Condi-
tions and reagents: a) [PdCl2(PPh3)2] (0.05 equiv), CuI (0.1 equiv),
Et3N, THF, 668C, 16 h, 91%, d.r. 6:1; b) tBuLi (2.5 equiv), THF,
ꢀ788C, I2 (1.5 equiv), ꢀ78!228C 10 min, 85%; c) 10a (1.2 equiv),
[PdCl2(PPh3)2] (0.05 equiv), CuI (0.1 equiv), Et3N, THF, 668C, 16 h,
70%; d) tBuLi (2.5 equiv), THF, ꢀ788C, I2 (1.5 equiv), ꢀ78!228C,
10 min, 91%; e) KOH (2.0 equiv), MeOH/iPrOH/Et2O (2:2:1), 228C,
2 h, 71%; f) [PdCl2(PPh3)2] (0.05 equiv), CuI (0.1 equiv), Et3N, THF,
0.003m with respect to 17, 668C, 24h, 45%; g) TBAF, THF, 22 8C, 2 h,
83%; h) Ac2O, pyridine, DMAP (0.1 equiv), 12 h, 85%; i) MeMgBr
(18 equiv), LiBr (18 equiv), CuI (18 equiv), THF, 0!228C, 1.5 h, 83%.
TBAF=tetrabutylammonium fluoride.
Scheme 5. Final steps for the assembly of (+)-allenophane 3. Condi-
tions and reagents: a) KOH (2.0 equiv), MeOH/iPrOH/Et2O (2:2:1),
228C, 2 h, 86%; b) 15 (1.0 equiv), 19 (1.2 equiv), [PdCl2(PPh3)2]
(0.05 equiv), CuI (0.1 equiv), Et3N, THF, 668C, 16 h, 73%; c) tBuLi
(2.5 equiv), THF, ꢀ788C; I2, ꢀ78!228C, 10 min; d) KOH (2.0 equiv),
MeOH/iPrOH/Et2O (1:1:2), temp, 2 h, 80% (over 2 steps); e) [PdCl2-
(PPh3)2] (0.05 equiv), CuI (0.1 equiv), Et3N, THF, 0.003m with respect
to 21, 668C, 16 h, 42%; f) TBAF, THF, 228C, 1 h; g) Ac2O, pyridine,
DMAP (0.1 equiv), 228C, 12 h, 65%; h) MeMgBr (30 equiv), LiBr
(30 equiv), CuI (30 equiv), THF, 08C, 3.5 h, 228C, 15 min, 85%.
manner in which these components are assembled demon-
strates the utility and versatility of the building blocks for the
construction of acetylenic macrocycles by our sequential
cross-coupling strategy.
Both macrocycles (2 and 3) are very sensitive to mild acid
(e.g. silica gel or CDCl3) and were therefore purified by
chromatography on basic alumina. Upon storage (neat) they
continued to deteriorate, and this degradation was also
observed during 1H NMR spectroscopic analysis in
[D6]benzene. Exposure to air and moisture is also deleterious.
Fortunately, these macrocycles were stable when handled in
hydrocarbons or ethers and stored in a freezer (ꢀ208C).
Modeling studies[16] indicated very little twisting across
the ring, and the resulting structure diagrams were very
similar to the structural formulae drawn for macrocycles 2 and
3. Instead it appears that the allene bridges are considerably
strained and are likely distorted from their preferred geom-
etry. However, the modeling calculations may not be at a high
enough level of theory to confirm this observation (X-ray
analysis was not possible for these compounds). In our
previous studies with cyclic butadiynes the strain was not
evenly distributed throughout the ring but instead was
concentrated in a single bent triple bond (153.58).[17] Macro-
cycles 2 and 3 are likely to be similar examples, and this
accounts for their acid sensitivity. Consequently one or more
of the allene components behaves like a strained olefin with
increased reactivity and it is therefore prone to decomposi-
tion.
Allenic compounds usually display a unique antisym-
= =
metrical C C C IR stretching band in the region of 1950–
2150 cmꢀ1.[18] This characteristic signal is observed as a sharp
band at 2150 cmꢀ1 for allene 13. We were initially surprised to
observe no distinct corresponding band for macrocycle 3.
However, previous IR studies have shown the intensity of
these signals to be very sensitive to their environment. Key
570
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 568 –572

Angewandte
Chemie
(500 MHz, C6D6): d = 7.70 (t, J = 1.5 Hz, 3H), 7.45 (d, J = 1.5 Hz,
criteria, including strain, symmetry, and alkyl and aryl
substitution, collectively reduce the intensity of this n1 band.
This is illustrated by the IR spectra of a triphenylallene
6H), 2.13 (s, 18H), 1.33 ppm (s, 27H); 13C NMR (125 MHz, C6D6):
d = 205.9(”), 145.4(”), 123.0(+), 121.3(+), 102.7(”), 34.5(x), 31.1(+),
17.5(+) ppm; HRMS (EI) calcd for C45H54: 594.4225, found:
594.4214; IR: n˜ = 2593, 2924, 2854, 1759, 1688, 1684, 1593, 1456,
= =
(PhHC C CPh2), in which the band is very weak, and the
= =
tetraphenyl homologue (Ph2C C CPh2), in which the n1 band
1365, 1242, 879, 700 cmꢀ1
;
[a]2D0 = ꢀ211 degcm3 gꢀ1 dmꢀ1 (c =
is absent.[18] Macrocycles 2 and 3 also display IR bands of
reduced intensity. Given this situation, it is not surprising to
observe an extremely weak band at 1956 cmꢀ1 for 2; owing to
its higher symmetry, 3 should exhibit no absorption, and this is
indeed the case. In principle the Raman spectrum should
display a band corresponding to the allene vibration. The
three allene bridges in compound 2 give rise to a strong signal
at 1941 cmꢀ1, while for 3 the four allene bridges exhibit a
weaker, but still visible, band at 1930 cmꢀ1. The circular
dichroism (CD) spectra of macrocycles 2 and 3 (Figure 2)
were recorded in hexanes. These are the first such reported
spectra of asymmetric meta-allenophanes.
0.15 gcmꢀ3, hexanes). [(+) and (” ) denote the signal phase in the
DEPT NMR spectrum.]
Received: July 5, 2007
Published online: November 21, 2007
Keywords: allenes · allenophanes · asymmetric synthesis ·
.
enantioselectivity · macrocycles
[1] a) F. Diederich, Cyclophanes, Royal Society of Chemistry,
Cambridge, UK, 1991; b) F. Vögtle, Cyclophane Chemistry,
Wiley, New York, 1993; c) Y. Tobe, Top. Curr. Chem. 1994, 172,
1; d) J. Schulz, F. Vögtle, Top. Curr. Chem. 1994, 172, 4 1;
e) Modern Acetylene Chemistry (Eds.: P. J. Stang, F. Diederich),
Wiley-VCH, Weinheim, 1995; f) B. König, Top. Curr. Chem.
1998, 196, 92; g) T. Tsuji in Advances in Strained Interesting
Organic Molecules, Vol. 7 (Ed.: B. Halton), JAI Press, Green-
wich, CT, 1999, pp. 103 – 152; h) G. J. Bodwell, T. Satou, Angew.
Chem. 2002, 114, 4175; Angew. Chem. Int. Ed. 2002, 41, 4003;
i) Modern Cyclophane Chemistry (Eds.: R. Gleiter, H. Hopf),
Wiley-VCH, Weinheim, 2004; j) Acetylene Chemistry (Eds.: F.
Diederich, P. J. Stang, R. R. Tykwiniski), Wiley-VCH, Wein-
heim, 2005.
[2] a) M. A. Romero, A. G. Fallis, Tetrahedron Lett. 1994, 35, 4711;
b) S. K. Collins, G. P. A. Yap, A. G. Fallis, Angew. Chem. 2000,
112, 393; Angew. Chem. Int. Ed. 2000, 39, 385; c) S. K. Collins,
G. P. A. Yap, A. G. Fallis, Org. Lett. 2000, 2, 3185; d) S. K.
Collins, G. P. A. Yap, A. G. Fallis, Org. Lett. 2002, 4, 11; e) M. A.
Heuft, S. K. Collins, A. G. Fallis, Org. Lett. 2003, 5, 1911.
[3] M. A. Heuft, A. G. Fallis, Angew. Chem. 2002, 114, 4702; Angew.
Chem. Int. Ed. 2002, 41, 4520.
Figure 2. CD spectra of allenophanes (ꢀ)-2 and (ꢀ)-3.
[4] For related examples of helical chirality in cyclophanes, see a) R.
Boese, A. J. Matzger, K. P. C. Vollhardt, J. Am. Chem. Soc. 1997,
119, 2052; b) M. M. Haley, M. L. Bell, S. C. Brand, D. B.
Kimball, J. J. Pak, W. B. Wan, Tetrahedron Lett. 1997, 38, 7483;
c) M. J. Marsella, I. T. Kim, F. Tham, J. Am. Chem. Soc. 2000,
122, 974.
[5] a) G. J. Bodwell, T. Satou, Angew. Chem. 2002, 114, 4175;
Angew. Chem. Int. Ed. 2002, 41, 4003; b) C. Grave, A. D.
Schlüter, Eur. J. Org. Chem. 2002, 3075.
[6] a) J. A. Marshall, Chem. Rev. 1996, 96, 3067; b) R. Zimmer, C. U.
Dinesh, E. Nandanan, F. A. Khan, Chem. Rev. 2000, 100, 3067;
c) A. Hoffmann-Röder, N. Krause, Angew. Chem. 2002, 114,
3057; Angew. Chem. Int. Ed. 2002, 41, 2933; d) Modern Allene
Chemistry, (Eds.: N. Krause, A. S. K. Hashmi), Wiley-VCH,
Weinheim, 2004.
In summary, we have developed a general synthetic
strategy for the rapid assembly of both acetylenic and allenic
macrocycles from asymmetric tertiary propargyl alcohols. The
sequential application of Sonogashira cross-coupling reac-
tions[2,19] generated the core structures in a direct manner.
This protocol facilitated the synthesis of two new alleno-
phanes, the tris-meta-allenophane 2 and the higher homo-
logue, tetrakis-meta-allenophane 3. These transformations
were accomplished in nine and seven steps, respectively, from
propargyl ethers.
Experimental Section
[7] A. Hoffmann-Röder, N. Krause, Angew. Chem. 2004, 116, 1216;
Angew. Chem. Int. Ed. 2004, 43, 1196.
[8] R. A. Smith, R. L. White, A. Krantz, J. Med. Chem. 1988, 31,
1558.
[9] M. D. Clay, A. G. Fallis, Angew. Chem. 2005, 117, 4107; Angew.
Chem. Int. Ed. 2005, 44, 4039.
[10] D. L. An, T. Nakano, A. Orita, J. Otera, Angew. Chem. 2002, 114,
179; Angew. Chem. Int. Ed. 2002, 41, 171.
[11] T. Kawase, Angew. Chem. 2005, 117, 7500; Angew. Chem. Int.
Ed. 2005, 44, 7334.
[12] S. Thorand, F. Vögtle, N. Krause, Angew. Chem. 1999, 111, 3929;
Angew. Chem. Int. Ed. 1999, 38, 3721.
2: Anhydrous LiBr (95 mg, 1.1 mmol, 18 equiv) and CuI (210 mg,
1.1 mmol, 18 equiv) were suspended in THF (5 mL). The stirred
suspension was cooled to 08C, and MeMgBr (3m in diethyl ether,
0.37 mL, 1.1 mmol, 18 equiv) was added. The resulting yellow mixture
was stirred for 10 min at 08C before triacetate 18 (45 mg, 62 mmol) in
THF (2 mL) was added. The reaction mixture was then stirred for a
further 1.5 h at 08C. Pentane (30 mL) was added, and the insoluble
material was removed by filtration through a pad of Celite. The
solvents were removed in vacuo, and the crude product was purified
by flash chromatography (pentane) on basic alumina to give
1
allenophane 2 (30 mg, 83%) as a pale yellow, waxy solid. H NMR
Angew. Chem. Int. Ed. 2008, 47, 568 –572
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
571

Communications
[13] S. Odermatt, J. L. Alonso-Gomez, P. Seiler, M. M. Cid, F.
Suite from Molecular Simulations Inc. (San Diego, CA, 1999)
and by using the methods in: M. Dewar, W. Thiel, J. Am. Chem.
Soc. 1977, 99, 4499. The reviewer suggested applying: B3LYP/6-
31G** density functional theory calculations (a) C. Kind, M.
Reiher, J. Roder, B. A. Hess, Phys. Chem. Chem. Phys. 2000, 2,
2205). Unfortunately owing to the large number of orbitals (1290
for compound 3) computation required one month and resulted
in unsatisfactory data.
Diederich, Angew. Chem. 2005, 117, 5203; Angew. Chem. Int.
Ed. 2005, 44, 5074.
[14] a) S. K. Collins, DPhil Thesis, University of Ottawa (Canada),
2001; b) M. A. Heuft, DPhil Thesis, University of Ottawa
(Canada), 2003; c) Y. Tobe, J. Kishi, I. Ohki, M. Sonoda, J.
Org. Chem. 2003, 68, 3330; d) A. G. Fallis, Synlett 2004, 2249.
[15] T. L. Macdonald, D. R. Reagan, J. Org. Chem. 1980, 45, 4740.
[16] S. K. Collins, G. P. Yapp, A. G. Fallis, Org. Lett. 2002, 4, 11.
[17] Initial calculations: Semiempirical AM1 calculations were
performed by using the Cerius2-Dmol3 Molecular Modeling
[18] J. H. Wotiz, D. E. Mancuso, J. Org. Chem. 1957, 22, 207.
[19] R. Chinchilla, C. Najera, Chem. Rev. 2007, 107, 874.
572
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 568 –572