Tandem cross-dimerisation/oxonia-cope reaction of carbonyl compounds to homoallylic esters and lactones

Article abstract of DOI:10.1002/anie.201200425

Conceptually different: This allyltransfer reaction is catalyzed by Lewis acids (LAs) and proceeds atom-economically by disproportionation of the carbonyl groups through organized oxonia-Cope transition states (see scheme). A stereoselective [n+4] ring enlargement leads to a variety of macrolides with 9- to 16-membered rings. Copyright

Full text of DOI:10.1002/anie.201200425

Angewandte
Chemie
DOI: 10.1002/anie.201200425
Synthetic Methods
Tandem Cross-Dimerisation/Oxonia-Cope Reaction of Carbonyl
Compounds to Homoallylic Esters and Lactones**
Yue Zou, Changming Ding, Lijun Zhou, Zhiming Li, Quanrui Wang,* Franziska Schoenebeck,
and Andreas Goeke*
Dedicated to Professor Georg Frꢀter
Homoallylic subunits constitute ubiquitous structural motifs
in organic chemistry, and their stereoselective introduction is
of fundamental academic and industrial importance. In
addition to carbonyl ene reactions,[1] allylations of carbonyl
compounds with allylic organometal species represent the
main route to these subunits.[2] The reactivity of diverse
reagents can be tuned to effect high levels of diastereo- and
enantioselectivity, which usually originates from the chairlike
6-membered-ring transition state A (Scheme 1).[2b] A charac-
teristic feature of this approach is the reaction with allylic
group via an oxonia-Cope transition state B (Scheme 1); one
equivalent of a carbonyl compound is sacrificed, thus assuring
an irreversible course of the rearrangement.
Herein, we describe a stereoselective and conceptually
new allylation method of broad scope (Scheme 2). The allyl
Scheme 2. Tandem cross-dimerisation/oxonia-Cope sequence of two
different aldehydes.
unit is not delivered by an organometallic reagent to generate
a homoallylic alcoholate, but through a cross-disproportiona-
tion of a b,g-unsaturated carbonyl compound with another
aldehyde to result in a homoallylic ester. Unlike in the
crossed-Tishchenko reaction,[6] this disproportionation of two
different aldehydes is not the result of a hydride transfer,
instead it is brought about by an oxonia-Cope rearrangement
to form homoallylic esters and lactones.
Scheme 1. Allylation of aldehydes and allyl transfer to a adducts.
L=ligand, LA=Lewis acid, M=metal.
During investigation of this unique stereoselective trans-
formation, the concept was further developed into a general,
convergent, and atom-economical [n + 4] ring enlargement to
homoallylic macrolides with 9- to 16-membered rings. With
regard to the latent functionality embedded in the homoallyl
substructure, this novel macrolactonization illustrates a pow-
erful synthetic strategy toward a large variety of important
bioactive macrolides.[7] In addition to presenting the scope of
this tandem cross-dimerization/oxonia-Cope sequence, we
also elucidate its mechanistic and stereochemical principles.
The transformation was discovered during an attempt to
synthesize compound 4 as a derivative of oxaspiro odorants
inversion of allyl metal species to form g products. Changing
this regioselectivity toward the homoallylic a adducts repre-
sents a significant challenge.[3] The Lewis acid catalyzed allyl
transfer of secondary and tertiary alcohols to carbonyl
compounds[4] is an established method and has been used
successfully in natural product synthesis.[5] This reaction
proceeds with inversion of configuration at the hydroxy
[*] Dr. Y. Zou, L. Zhou, Dr. Z. Li, Prof. Dr. Q. Wang
Department of Chemistry, Fudan University
220 Handan Road, Shanghai 200433 (P.R. China)
E-mail: qrwang@fudan.edu.cn
that
have
interesting
cassis/grapefruit-like
odors
(Scheme 3).[8] However, instead of undergoing the expected
Dr. Y. Zou, C. Ding, L. Zhou, Dr. A. Goeke
Givaudan Fragrances (Shanghai) Ltd.
298 Li Shi Zhen Road, Shanghai 201203 (P.R. China)
E-mail: andreas.goeke@givaudan.com
Prof. Dr. F. Schoenebeck
Laboratorium fꢀr Organische Chemie, ETH Zꢀrich
Wolfgang-Pauli-Str.10, 8093 Zꢀrich (Switzerland)
[**] We thank Dr. G. Brunner for NMR experiments, Dr. Z. Chen at Fudan
University for X-ray measurements, and G. Tang at Fudan University
for the high-resolution MS data.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201200425.
Scheme 3. 2-Oxonia-Cope rearrangement of 1a versus expected Prins
reaction.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
Table 1: Scope of the oxonia-Cope rearrangement.[a]
Prins cyclization in the presence of acetaldehyde (2a) to
hemiacetal 4, b,g-unsaturated aldehyde 1a was smoothly
converted into homoallylic ester 3a in good yield.[9,10] This
rearrangement is remarkable from two perspectives: the
transformation is clean, and the formation of homo- or cross-
aldol products was not observed. The reaction obviously
involves an intermolecular disproportionation of the different
carbonyl groups to form a homoallylic ester, and the high E/Z
and syn/anti selectivities of the product clearly indicate a well-
organized transition state (TS). In addition, the transforma-
tion is significantly accelerated by a wide variety of Lewis and
Brønsted acid catalysts (examples are listed in the Supporting
Information); the conversion of 1a with acetaldehyde to
product 3a was not observed under thermal conditions at
temperatures of < 3508C. Thus, this reaction could be
described as an oxy-oxonia-Cope rearrangement.
Entry
Substrate
Aldehyde
Product
Yield [%][b]
(E/Z)
{anti/syn}
1
2
3
70
(56:44)
1
2
1b
1c
2c
2a
3c
3d
40
(57:43)
90
{>99:1}
3
4
1d
1e
2d
2e
3e[c]
75
{>99:1}
The broad scope of the rearrangement is shown in Table 1.
The b,g-unsaturated substrates were prepared by Diels–Alder
reactions[8] or by Friedel–Crafts acylations of olefins.[11] Other
options to access such deconjugated b,g-unsaturated carbonyl
compounds are, for example, hydroacylation reactions of
dienes[12] or carbonyl ene reactions followed by oxidation.[1]
Several conclusions can be drawn from the results that are
summarized in Table 1. Substitution at the vinyl group results
in a smooth outcome of the reaction, thus indicating that
stabilization of the positive charge at this position in the
intermediate is important. For example, compound 3d was
obtained with lower yield than product 3c (Table 1, entries 1
and 2). Diverse structures are accepted by the transformation
and steric bulk around the carbonyl group of compound 1 is
generally tolerated. For example, a-cyclocitral 1d is diaste-
reoselectively converted to 3e in high yield; isomerisation of
the double bond of a-cyclocitral to obtain conjugated b-
cyclocitral was not observed under the acidic reaction
conditions (Table 1, entry 3; see also entries 4, 5, and 8). A
variety of functional groups, such as esters (Table 1, entry 8)
or additional double bonds (entries 1, 2, 5, and 7), in
compounds 1 and 2 are also permitted in this reaction.
Excellent diastereoselectivities can be achieved when the
double bond at the b,g position in compounds 1 is part of
a ring system (Table 1, entries 3 and 4), an observation that
led us to analyze the essential question regarding the extent of
chirality transfer in this rearrangement (Scheme 4). For this
purpose, we chose to examine the conversion of compound
(1R,3R,6S)-(+)-1i, which was obtained from readily available
D-3-carene with known absolute configuration,[11] with acet-
aldehyde 2a. The reaction proceeded under complete transfer
of stereochemical information to give ester 3n (Scheme 4).
The relative and absolute configurations of compound 3n
were determined by electronic circular dichroism (ECD)
measurement and X-ray crystallographic analysis after trans-
formation of 3n into its 4-nitrobenzoic acid ester (Supporting
Information). In parallel, we studied this transformation
computationally,[13] applying B3LYP,[14] M062X,[15] and SCS-
MP2[16] levels of theory. Chairlike TSs are generally preferred
over boatlike TSs in [3,3]-sigmatropic shifts;[17] however, in
cyclic systems, the preference for chairlike TSs may be smaller
as a result of larger steric hindrance, which may even cause
the chairlike TS to be disfavored.[18]
3 f[c]
85[d]
(75:25)
{50:50}
5
1 f
2a
2 f
3g
3h
6
7
91
51
1g
1g
2g
3i
94[d]
(72:28)
{50:50}
8
1h
2a
3j
[a] Conditions: a solution of 1,2-dichloroethane (20 mL), aldehyde
1 (10 mmol), aldehyde 2 (12 mmol), and BF3OEt2 (1.0 mmol) was stirred
under Ar for 2 h at 08C!RT. [b] Yields of isolated products after flash
chromatography or Kugelrohr distillation. [c] Relative configuration
shown. [d] SnCl4 used instead of BF3OEt2.
Scheme 4. Favored reaction pathway for the oxy-oxonia-Cope rear-
rangement and illustration of transition states.
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
Table 3: Oxonia-Cope [n+4] ring enlargement to macrolides.
Our calculations of the oxy-
oxonia-Cope reaction of substrate
1i to 3n predict a chairlike TS
(chair-TS-S, Scheme 4) to be
favored (by DDH° = 7.6 kcalmolꢀ1
over the minimum-energy boatlike
TS at B3LYP/6-31G(d)). The chair-
like TS is asynchronous but con-
certed, and is formed with an acti-
vation barrier of DG° = 25.5 kcal
molꢀ1 at SCS-MP2 (DCM)/6-
31G(d)//B3LYP/6-31G(d) level rel-
ative to the aldehyde and the BF3-
complexed reactant 1i.[19] The reac-
tion path via chair-TS-S (5, see
Scheme 4) gives rise to the exper-
imentally observed S selectivity.
The lowest energy pathway that
leads to R configured product 8
would proceed via chair-TS-R (see
Scheme 4 and also Scheme S5 in the
Supporting Information for compi-
lation of all possible TS conform-
ers). Chair-TS-R is substantially
higher in energy than chair-TS-S
(DDG° = 3.5 to 4.8 kcalmolꢀ1,
depending on the level of theory
applied, see Table 2), an observa-
tion that is in agreement with the
high enantioselectivity observed
experimentally. An exergonicity of
DGrxn = ꢀ7.9 kcalmolꢀ1 at M062X,
but only DGrxn = ꢀ1 kcalmolꢀ1 at
SCS-MP2 level of theory was calcu-
lated, thus suggesting that the oxy-
oxonia-Cope rearrangement might
be reversible at higher tempera-
tures,[20] although we did not see
any indication of such reversibility
at low temperature. Overall, the
model predicts a concerted asyn-
chronous reaction pathway for the
oxy-oxonia-Cope rearrangement,
but this may vary with the substitu-
tion pattern.
Entry Substrate
Aldehyde
2
Lewis
T
Product
11
Yield [%][b]
10
acid [20%] [oC]
(E/Z)[c]
83
1
SnCl4
0
(<1:99)[d]
10a
2a
2a
11a
11b
35[f]
(<1:99)
2
SnCl4
0
10b
97
(96:4)
3
BF3OEt2
ꢀ20
10c
2b
2j
11c
11d
90
(82:18)
4
BF3OEt2
BF3OEt2
ꢀ20
10c
93
ꢀ20
(90:10)[d]
88[e]
5
ꢀ70[e]
(85:15)[d,e]
10d
2a
11e
84
6
BF3OEt2
ꢀ20
(86:14)[d]
10d
2k
2k
2a
11 f
11g
11h
74
7
BF3OEt2
BF3OEt2
0!RT
(15:85)[d]
10d
72
(58:42)
8
0
10e
52
(62:38)
9
SnCl4
0
10 f
2a
11i
94
(92:8)
10
BF3OEt2
ꢀ20
10g
2b
11j
[a] Conditions: a solution of aldehyde 10 (10 mmol), aldehyde 2 (12 mmol), and BF3OEt2 or SnCl4
(2.0 mmol) in CH2Cl2 (20 mL) was stirred under Ar at the indicated temperature for 6 h. [b] Yields of
isolated products after flash chromatography or Kugelrohr distillation. [c] E/Z ratios were determined by
GC and NMR analysis. [d] syn/anti<1:99. [e] The reaction was carried out on a 100 g scale (see
Experimental Section). [f] The yield was low because of double bond isomerization of 10b into
conjugation.
With these considerations in
mind, we intended to further
amplify the scope of this reaction
through a novel and general macro-
lactonization method that involves
linking R1 together with R2 or R3 by a carbon chain of a given
Table 2: Free energy difference (kcalmolꢀ1) between the lowest-energy
chairlike transition states, which lead to the (R)-8 or (S)-3n.
length (Scheme 2, Table 3). It should be noted that this
sequence does not represent a macrolactonization, which
proceeds primarily through an intramolecular ring closure via
activated esters, in the classical sense.[7] The described macro-
lactonization is rather an [n + 4] ring enlargement of 2-vinyl-
substituted cyclic ketones, and does not require application of
high-dilution techniques. Lactones with 9- to 16-membered
Method
DDG°
[TS-R(7)ꢀTS-S(5)]
SCS-MP2/6-31G(d)//B3LYP/6-31G(d)
M062X/6-31+G(d,p)
B3LYP/6-31G(d)
4.8
3.5
4.6
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
rings were prepared (Table 3); yields and selectivities of the
lactone formation depend primarily on the ring size of the
product and the reaction temperature. This behavior was
expected, because the conformations of 6-membered-ring TSs
can deviate very much from chairlike geometries with
equatorial substituents, depending on the strain exerted by
the annulated ring system. High Z selectivity was observed in
lactones with 9-membered rings (Table 3, entries 1 and 2)
even at 08C. In addition, the methyl group in bicyclic
compound 11a is exclusively in the anti-position. This
excellent diastereoselectivity was also determined in higher
homologues 11e–g, irrespective of the double bond config-
uration in the product (Table 3, entries 5–7). The double
bonds in lactones with larger ring sizes preferentially occupy
E configurations, although the selectivity also depends on the
reaction temperature. This influence is particularly obvious in
the reaction of substrate 10d with benzaldehyde: while
E configuration prevails at ꢀ208C (Table 3, entry 6), it
switches to Z selectivity at higher temperatures (entry 7).
The sensitive influence of the reaction temperature on the
configuration of the double bond is also evident from
a reaction of compound 10d with acetaldehyde on a large
scale (100 g) and with a high concentration (2.5m; Table 3,
entry 5; see also Experimental Section). Although this
reaction was carried out at ꢀ708C, the E/Z ratio of 85:15
was lower than that obtained in a reaction on a small scale and
with a lower concentration (1m; E/Z = 9:1 at ꢀ208C). This
result can be explained by limited heat transfer occurring in
the exothermal reaction at high concentration.
ion[21] 13, which would have generated product (E)-(R)-11k.
But the question remained how (Z)-(R)-11k arose. We can
exclude the possibility of isomerization of both the substrate
(R)-10c and the product (E)-(S)-11k, because neither of the
compounds showed a sign of isomerization, when treated with
BF3OEt2 or SnCl4 under the reaction conditions and reaction
time. Clearly, the Z configured isomer must come from the
cis-decalin-like conformer 14, which leads to the observed
R configured isomer 11k with opposite absolute configura-
tion. Again, the occurrence of a Z oxocarbenium ion like
species 15 can be excluded,[22] because it would have led to
partial racemization through formation of isomer (Z)-(S)-
11k. Expectedly, this experiment also confirms that the
stereoselectivity of the oxonia-Cope reaction is mainly
controlled by the population of different chairlike conformers
at a given temperature, rather than by the configuration of
intermediate oxocarbenium ions.
We have described an unprecedented stereoselective
allylation method, which does not include an allylic organo-
metallic reagent, but a cross-dimerization of an b,g-unsatu-
rated carbonyl compound with another aldehyde to produce
homoallylic esters through disproportionation of the carbonyl
functionalities in an atom-economical way. The proposed
irreversible and well-organized oxy-oxonia-Cope rearrange-
ment enables high diastereoselectivities and transfer of
chirality. The broad scope of the approach was further
substantiated by the introduction of a novel [n + 4] ring
enlargement toward macrolides with 9- to 16-membered
rings.[23] The reactions are highly productive and work on
multigram scale.
The stereochemistry was investigated in more detail in
lactones with 10-membered rings (Scheme 5). Compound
(R)-10c (99% ee)[9] was reacted with acetaldehyde 2a in the
presence of 10 mol% BF3OEt2 to give a mixture of (E)-(S)- Experimental Section
A solution of BF3OEt2 (7.84 g, 0.055 mol) in 50 mL CH2Cl2 was added
11k and (Z)-(R)-11k in a ratio of 4:1. The fact that the optical
information was completely transferred, presumably via
trans-decalin-like species 12, to the major (E)-(S)-11k (99%
ee) isomer, excludes the occurrence of a Z oxocarbenium
dropwise to a cooled (ꢀ788C) mixture of compound 10d (107 g, 92%
purity, 0.55 mol) and acetaldehyde (2a, 29.2 g, 0.66 mol) in CH2Cl2
(200 mL) over a period of 60 min while the temperature was kept
below ꢀ708C. The temperature was allowed to increase to ꢀ208C
during 10 h. Saturated aqueous NaHCO3 (50 mL) was added and the
organic phase was separated. The aqueous layer was extracted once
with CH2Cl2 (100 mL). The combined organic layers were washed
twice with saturated aqueous NaHCO3 solution (100 mL), dried over
MgSO4 (10 g), and concentrated in vacuo to give a light-yellow oil,
which was distilled through a short-path distillation head, thus
affording lactone 11e (109 g, 88%) as a colorless oil. B.p. 105–1108C
(0.12 mbar). Mixture of E/Z isomers in a ratio of 85:15 (d.r. > 99:1).[9]
Received: January 16, 2012
Published online: && &&, &&&&
Keywords: allylation · homoallylic esters · rearrangement ·
.
ring enlargement · Tishchenko reaction
[1] a) M. Terada, Sci. Synth. 2011, 3, 309 – 346; b) M. L. Clarke,
M. B. France, Tetrahedron 2008, 64, 9003 – 9031.
[2] a) M. Yus, J. C. Gonzalez-Gomez, F. Foubelo, Chem. Rev. 2011,
111, 7774 – 7854; b) S. E. Denmark, J. Fu, Chem. Rev. 2003, 103,
2763 – 2793; c) M. Lombardo, C. Trombini, Chem. Rev. 2007,
107, 3843 – 3879.
Scheme 5. Chirality transfer in the series of lactones with 10-mem-
bered rings.
[3] a) A. Yanagisawa, S. Habaue, H. Yamamoto, J. Am. Chem. Soc.
1991, 113, 8955 – 8956; b) A. Yanagisawa, Sci. Synth. 2004, 7,
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
555 – 571; c) L. M. Zhao, H. S. Jin, L. J. Wan, L. M. Zhang,
J. Org. Chem. 2011, 76, 1831 – 1837.
[12] F. Shibahara, J. F. Bower, M. J. Krische, J. Am. Chem. Soc. 2008,
130, 14120 – 14122.
[4] a) J. Nokami, K. Yoshizane, H. Matsuura, S. I. Sumida, J. Am.
Chem. Soc. 1998, 120, 6609 – 6610; b) A. V. Malkov, M. A.
Kabeshov, M. Barlog, P. Kocovsky, Chem. Eur. J. 2009, 15,
1570 – 1573; c) Y. H. Chen, F. E. McDonald, J. Am. Chem. Soc.
2006, 128, 4568 – 4569.
[5] F. E. McDonald, C. L. Pereira, Y. H. Chen, Pure Appl. Chem.
2011, 83, 445 – 459.
[6] a) Y. Hoshimoto, M. Ohashi, S. Ogoshi, J. Am. Chem. Soc. 2011,
133, 4668 – 4671; b) W. I. Dzik, L. J. Goossen, Angew. Chem.
2011, 123, 11241 – 11243; Angew. Chem. Int. Ed. 2011, 50, 11047 –
11049.
[7] a) A. Parenty, X. Moreau, J. M. Campagne, Chem. Rev. 2006,
106, 911 – 939; b) E. J. Kang, E. Lee, Chem. Rev. 2005, 105, 4348 –
4378; c) I. Shiina, Chem. Rev. 2007, 107, 239 – 273; d) H. M. C.
Ferraz, F. I. Bombonato, L. S. Longo, Jr., Synthesis 2007, 3261 –
3285; e) E. A. Crane, K. A. Scheidt, Angew. Chem. 2010, 122,
8494 – 8505; Angew. Chem. Int. Ed. 2010, 49, 8316 – 8326; Angew.
Chem. 2010, 122, 8494 – 8505.
[13] M. J. Frisch et al., Gaussian 09, Revision A. 01, Gaussian, Inc.
Wallingford CT, 2009. See the Supporting Information for the
full reference.
[14] For adequateness of B3LYP methodology: O. Wiest, K. A.
Black, K. N. Houk, J. Am. Chem. Soc. 1994, 116, 10336 – 10337.
[15] a) Y. Zhao, D. G. Truhlar, Theor. Chem. Acc. 2008, 120, 215 –
241; b) S. N. Pieniazek, F. R. Clemente, K. N. Houk, Angew.
Chem. 2008, 120, 7860 – 7863; Angew. Chem. Int. Ed. 2008, 47,
7746 – 7749.
[16] T. P. M. Goumans, A. W. Ehlers, K. Lammertsma, E. U.
Wuerthwein, S. Grimme, Chem. Eur. J. 2004, 10, 6468 – 6475.
[17] a) For the acyclic Cope rearrangmenet, a 6 kcalmolꢀ1 preference
for the chairlike TS has been determined: W. v. E. Doering,
W. R. Roth, Tetrahedron 1962, 18, 67 – 74; b) D. A. Hrovat, W. T.
Borden, J. Am. Chem. Soc. 2001, 123, 4069 – 4072.
[18] a) F. E. Ziegler, Chem. Rev. 1988, 88, 1423 – 1452; b) S. Gꢀl, F.
Schoenebeck, V. Aviyente, K. N. Houk, J. Org. Chem. 2010, 75,
2115.
[8] a) Y. Zou, Q. Wang, A. Goeke, Chem. Eur. J. 2008, 14, 5335 –
5345; b) A. Goeke, Y. Zou, PCT Int. Appl. WO 2010125100 A1
20101104, 2010.
[9] See the Supporting Information for characterization of com-
pounds.
[10] Derivative 3b is not crystalline. The relative configuration was
determined by crystal structure analysis of derivative 26, see the
Supporting Information. CCDC a11130b contains the supple-
mentary crystallographic data for this paper. These data can be
obtained free of charge from The Cambridge Crystallographic
Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
[11] P. J. Kropp, D. C. Heckert, T. J. Flautt, Tetrahedron 1968, 24,
1385 – 1395.
[19] The reactant complex is higher in energy than the isolated
reactants.
[20] For SCS-MP2/6-31G(d), both in gas-phase and DCM a reaction
free energy of 1 kcalmolꢀ1 is predicted. For M062X/6-31 + G-
(d,p), the gas-phase value (ꢀ7.9 kcalmolꢀ1) is given.
[21] D. Cremer, J. Gauss, R. F. Childs, C. Blackburn, J. Am. Chem.
Soc. 1985, 107, 2435 – 2441.
[22] Z-Configured oxocarbenium ions were shown to participate in
the racemization during Prins cyclizations: R. Jasti, S. D.
Rychnovsky, J. Am. Chem. Soc. 2006, 128, 13640 – 13648.
[23] For new applications in musk synthesis: Y. Zou, H. Mouhib, W.
Stahl, A. Goeke, Q. Wang, P. Kraft, Chem. Eur. J., DOI: 10.1002/
chem.201200882.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
Communications
Synthetic Methods
Y. Zou, C. Ding, L. Zhou, Z. Li, Q. Wang,*
F. Schoenebeck,
A. Goeke*
&&&& — &&&&
Conceptually different: This allyltransfer
reaction is catalyzed by Lewis acids (LAs)
and proceeds atom-economically by dis-
proportionation of the carbonyl groups
through organized oxonia-Cope transi-
tion states (see scheme). A stereoselec-
tive [n+4] ring enlargement leads to
a variety of macrolides with 9- to 16-
membered rings.
Tandem Cross-Dimerisation/Oxonia-
Cope Reaction of Carbonyl Compounds
to Homoallylic Esters and Lactones
6
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!