Practical enantioselective synthesis of β-lactones catalyzed by aluminum bissulfonamide complexes

Article abstract of DOI:10.1002/adsc.200700084

The development of an efficient and practical aluminum-bissulfonamide complex catalyzed enantioselective formation of β-lactones by [2+2] cycloaddition of ketene (generated in situ from acetyl bromide by dehydrobromination) with various α-unbranched and -

Full text of DOI:10.1002/adsc.200700084

COMMUNICATIONS
DOI: 10.1002/adsc.200700084
Practical Enantioselective Synthesis of b-Lactones Catalyzed by
Aluminum Bissulfonamide Complexes
Thomas Kulla and RenØ Petersa,*
a
ETH Zürich, Laboratory of Organic Chemistry, Wolfgang-Pauli-Str. 10, Hçnggerberg HCI E 111, 8093 Zürich,
Switzerland
Fax : (+41)-44-633-1226; e-mail: peters@org.chem.ethz.ch
Received: February 5, 2007
Supporting information for this article is available on the WWW under http://asc.wiley-vch.de/home/.
Abstract: The development of an efficient and prac-
tical aluminum-bissulfonamide complex catalyzed
enantioselective formation of b-lactones by [2+2]
cycloaddition of ketene (generated in situ from
acetyl bromide by dehydrobromination) with vari-
ous a-unbranched and -branched aliphatic alde-
hydes is presented. The methodology offers the ad-
vantage of operational simplicity not only as the
ligand synthesis requires just a single sulfonylation
step from commercially available enantiomerically
pure diamines. The products are formed in high to
excellent yields with ee values typically ranging
from 78 to 90% using 10 mol% of the bissulfona-
mide ligand. The key finding of this work was a re-
markable rate acceleration by using an aluminum/
ligand ratio of 1.5:1.
The main goal of the work presented herein was to
develop a widely applicable and practical catalyst
system for b-lactone formation by [2+2] cycloaddi-
tion of ketene and aldehydes.[4] Previous work by
other groups had shown that the asymmetric forma-
tion of b-lactones can be catalyzed either nucleophili-
cally, for example, by action of chiral tertiary amines
such as brucine[5] or Cinchona alkaloids,[6] or by Lewis
acid catalysts.[7] The most promising results with
regard to the latter strategy were obtained by alumi-
num-based systems. In the pioneering work by
Miyano et al.[8] and Kocienski et al.[9] bissulfonamide
ligands derived from chiral C2 symmetric 1,2-diamino-
1,2-diphenylethane (DiPh) were utilized employing
either the preformed gaseous parent ketene[8] or the
commercially available, yet expensive trimethylsilyl-
ketene.[9] However, the induced enantioselectivities
were in general moderate. Recently, Nelson et al.
could significantly improve both the scope and enan-
tioselectivity by employing Al catalysts prepared from
tridentate aminobissulfonamide ligands and trimethy-
laluminum.[10] Moreover, the ketene substrates could
be generated in situ from acyl bromides by treatment
Keywords: aluminum; catalysis; cycloaddition; ke-
tenes; b-lactones; sulfonamides
b-Lactones can be regarded as activated aldol equiva- with ethyldiisopropylamine. Based on these prece-
lents, since they readily undergo ring opening reac- dents our overall goal was to develop a system with
tions due to their intrinsic ring strain.[1] Various hard enhanced practicality which should be as simple as
nucleophiles are able to regioselectively cleave the possible, but still should provide high enantioselectivi-
acyl-oxygen bond thus providing the corresponding b- ties and yields. For that reason we decided to reinves-
hydroxy carbonyl derivatives.[2] Accordingly, the de- tigate bissulfonamide-derived Al complexes
2
velopment of catalytic asymmetric [2+2] cycloaddi- (Scheme 1) which although simple to prepare still
tions of ketenes[3] and aldehydes offers the possibility
to replace catalytic asymmetric ester aldol reactions
which in most cases require the preformation, isola-
tion and purification of moisture sensitive silyl ketene
acetals. From both a technical and economical point
of view, the use of silyl protecting groups is an issue
on production scale, not only because SiO2 being
formed during waste combustion processes has the
tendency to block the combustorsꢀ chimneys.
Scheme 1. Formation of bissulfonamide aluminum com-
plexes 2.
Adv. Synth. Catal. 2007, 349, 1647 – 1652
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1647

Thomas Kull and RenØ Peters
COMMUNICATIONS
allow a high structural diversity: along these lines the reasonable reaction rates (entries 1 and 2), while in
influence of different Al substituents R’’, sulfonamide the case of isopropyl residues, the conversion was
residues R’ and C2 symmetric diamino backbones can very slow even at ꢀ508C(entry 4). The bis- ortho-sub-
be investigated. Surprisingly, only a single application stituted aromatic residues R’ allowed the formation of
of this type of Al catalyst possessing a bissulfonylated lactone 6a with ee values >70%, the best results
cyclohexane-1,2-diamino (Cy) ligand backbone has being realized with the Cy-Dmtb ligand 5a (ee=80%,
been reported in literature, namely the investigation half conversion after 2.5 h, entry 1). Entries 1 and 2
of Al-catalyzed cyclopropanations,[11] whereas the cor- also revealed, that the para-substituents have a sub-
responding 1,2-diphenylaminoethane-derived alumi- stantial influence upon the reaction rate.
num catalysts have been frequently applied in cataly-
The investigation of the influence of the Al sub-
sis after the pioneering studies by Corey et al. on cat- stituent R’’ showed that the bulky isobutyl moiety
alytic enantioselective Diels–Alder reactions.[12]
permitted a significantly higher enantioselectivity
The cyclocondensation of acetyl bromide (3)[13] and (Cy-Trim 5b, T=ꢀ608C, ee=80%) than the ethyl
dihydrocinnamaldehyde (4a) was selected as model (ee=72%) or methyl (ee=53%) residues, but again
reaction. The initial ligand screening using the (R,R)- the higher selectivity was at the expense of a reduced
cyclohexane-1,2-diamino (Cy) backbone revealed that reaction rate.[14]
two ortho substituents on an aromatic sulfonyl residue
With a lower catalyst loading of 10 mol% the reac-
R’ are essential to achieve acceptable enantioselectivi- tion was not only further decelerated, but proceeded
ties with 20 mol% of the catalyst, but that the ortho also less enantioselectively (Cy-Trip 5d, T=ꢀ708C,
substituents also slow down the reaction to a large ee=65%). The catalysts with ortho-disubstituted aro-
degree (Table 1). Al-complexes missing the ortho-sub- matic residues R’’ being selected for further investiga-
stituents on R’ generally catalyzed the model reaction tions were generally prepared in situ by stirring a 1:1
smoothly at ꢀ788C(half conversion after 0.1 to 0.7 h
mixture of the bulky ligands 5a, 5b or 5d and Dibal at
1
as determined by H NMR monitoring with the ex- room temperature for 1 h followed by heating the
[15]
ception of entry 9), but the enantioselectivities were mixture to 808Cfor 4 h.
The complex 1H NMR
far from being preparatively useful (entries 5–8, 10). spectra of these mixtures showed that about one third
With methyl or ethyl ortho-substituents, the reaction of the ligand was not consumed under these condi-
temperature had to be increased to ꢀ608Cto obtain
tions.[16] By rising the Dibal amount to 1.5 equivs. per
Table 1. Initial screening of bissulfonamide ligands 5 derived from (R,R)-cyclohexyl-1,2-diamine.
Entry[a]
Ligand
R’
T [8C]
t1/2[b] [h]
ee[c] [%]
1
2
3
4
5
6
7
8
Cy-Dmtb 5a
Cy-Trim 5b
Cy-Trie 5c
Cy-Trip 5d
Cy-1-Naph 5e
Cy-2-Naph 5f
Cy-BTFM 5g
Cy-Ts 5h
2,6-(CH3)2-4-t-BuC6H2
2,4,6-(CH3)3C6H2
2,4,6-(C2H5)3C6H2
2,4,6-(i-Pr)3C6H2
1-naphthyl
2-naphthyl
3,5-(CF3)2C6H3
4-CH3C6H4
ꢀ60
ꢀ60
ꢀ60
ꢀ50
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
2.5
8
3.3
21
0.2
0.25
0.2
0.1
35
80
80
73
71
43
37
35
30
19
17
2
9
10
11
Cy-2-NO2 5i
Cy-PFP 5j
Cy-Tf 5k
2-NO2C6H4
C6F5
CF3
0.7
<0.5
[a]
All reactions were performed at a concentration c=0.13M. The ketene was preformed in a separate flask by treatment
of 3 (3 equivs.) with i-Pr2NEt (2.5 equivs.) in toluene at ꢀ788Cfor 4 h. The ketene solution was subsequently transferred
via canula into the reaction flask.
[b]
[c]
1
Determined by H NMR monitoring.
ee determined by chiral column HPLC(Daicel OD-H, see Supporting Information).
1648
asc.wiley-vch.de
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 1647 – 1652

Practical Enantioselective Synthesis of b-Lactones
COMMUNICATIONS
equiv. of ligand, the signals of the free ligand com-
pletely disappeared. With 10 mol% of the ligand and
15 mol% of Dibal the reaction is not only dramatical-
ly accelerated by the excess of Dibal, but surprisingly
also more selective than with a 1:1 stoichiometry of
ligand and Al source (Table 2). As demonstrated for
Table 2. Investigation of the effect of the ligand/Al-source
ratio for Cy-Trip 5d and Dibal.
Scheme 2.
Entry 5d/DIBAL Reaction Conversion[b] Yield[c] ee[d]
[mol%]
time [h]
[%]
[%]
[%]
1
2
3
4
5
10:10
10:12.5
10:15
10:17.5
10:20
18
18
15
15
15
37
57
100
100
98
32
55
87
62
66
65
70
78
77
72
AlEt3). The stoichiometry effect for the catalyst for-
mation was also observed in this case, albeit less pro-
nounced: with 10 mol% DiPh-Dmtb 7a and 10 mol%
AlEt3 the reaction gave 87% yield after 45 h with a
slightly diminished ee of 87%.[17]
At present we have no direct experimental evi-
dence about the origin of the stoichiometry effect.
The 1H NMR spectra of complexes which were
formed from various combinations of 1.0 equiv. of the
bissulfonamide ligands 5 or 7 and 1.5 equivs. of the Al
sources show like in the case of a 1:1 stoichiometry
[a]
All reactions were performed at ꢀ708Cat a concentra-
tion c=0.25M. The ketene was formed in situ.
[b]
[c]
1
Determined by H NMR.
Yield after aqueous workup determined by 1H NMR
using acetophenone as an internal standard.
ee determined by chiral column HPLC(Daicel OD-H,
see Supporting Information).
[d]
C1-symmetric dimeric complexes
2
(n=2 in
Scheme 1) as the main species.[18] However, in addi-
5d, the reaction which was very sluggish at ꢀ508C tion to these dimers, the formation of small amounts
with 20 mol% of catalyst (half conversion of the alde- of unidentified complexes has been detected which
hyde after 21 h), proceeded within 15 h at ꢀ708Cto
might differ from the 1:1 stoichiometry and which
completion using 10 mol% of the chiral ligand and 15 might possess a significantly higher activity. The
mol% of Dibal. A further increase of the amount of higher activity could finally result from a Lewis acid-
Dibal is detrimental to both yield and enantioselectiv- assisted Lewis acid activation (LLA concept).[19]
ity.
The scope of the methodology is summarized in
Due to the markedly enhanced activity the reaction Scheme 3 and Table 3: in general both a-branched
temperature was further reduced. At ꢀ858C, product and -unbranched aldehydes are well tolerated and fur-
6a was obtained in 93% yield after 48 h and with nish the corresponding b-lactones 6 in high yield and
88% ee (Table 3, entry 1, compare to the reactions at with ee values ranging from 78 to 90% using the best
ꢀ608C: Cy-Trip 5d: 99% yield, 75% ee; Cy-Dmtb 5a: combinations (Table 3) of sulfonamide and Al
72% yield, 73% ee, Cy-Trim 5b: 62% yield, 70% ee). source.[20,21] The results obtained for the two model
When the same conditions were applied to cyclo- substrates 4a and 4b can be generalized: the (R,R)-cy-
hexylcarbaldehyde 4b chosen as a model substrate for
a-branched aldehydes, the product was obtained in
disappointing 17% yield after 40 h at ꢀ858C(20%
conversion) and with just 72% ee. However, with
ligand Cy-Dmtb 5a, which performed inferiorly for di-
hydrocinnamaldehyde 4a, the product was obtained in
87% yield and with 86% ee after 40 h at ꢀ858C
(Table 3, entry 12).
Next the influence of the diamino backbone was in-
vestigated. While the cyclohexane-1,2-diamino (Cy)
backbone is superior for dihydrocinnamaldehyde (4a),
ligands derived from 1,2-diphenyl-1,2-diamine (DiPh)
are more efficient for the a-branched cyclohexanecar-
baldehyde (4b, Scheme 2). In the latter case, AlEt3
was superior as Al source. Under these conditions the
product was obtained in 88% yield after 25 h and
with 90% ee (10 mol% DiPh-Dmtb 7a, 15 mol% mation of b-lactones 6.
Scheme 3. Al-bissulfonamide-catalyzed enantioselective for-
Adv. Synth. Catal. 2007, 349, 1647 – 1652
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1649

Thomas Kull and RenØ Peters
COMMUNICATIONS
Table 3. Scope of the Al-bissulfonamide-catalyzed enantioselective formation of b-lactones 6.
Entry[a]
4
Ligand
Reaction time Conversion[b]
Al
source
Yield[c]
[%]
ee
[%]
Configuration[g]
[h]
[%]
1
2
Cy-Trip 5d
DiPh-Dmtb 48
7a
Cy-Trip 5d
DiPh-Trip 7b 144
48
98
78
Dibal
Dibal
93
66
88[d]
82[d]
(S)
(R)
4a
4c
4d
3
4
63
100
75
Dibal
Dibal
86
44
84[e]
73[e]
(S)
(R)
5
6
Cy-Trip 5d 62
DiPh-Dmtb 88
100
85
Dibal
Dibal
82
71
84[f]
79[f]
(R)
(S)
7a
7
8
Cy-Trip 5d
DiPh-Dmtb 140
7a
140
100
40
Dibal
Dibal
92
34
88[f]
74[f]
(S)
(R)
4e
4f
9
10
DiPh-Trip 7b 26
Cy-Trip 5d
100
100
Dibal
Dibal
98
84
85[f]
84[f]
(R)
(S)
49
DiPh-Dmtb
7a
Cy-Dmtb 5a 23
11
12
25
100
99
Et3Al
Et3Al
88[h]
87
90[f]
86[f]
(S)
4b
(R)
DiPh-Dmtb
7a
Cy-Dmtb 5a 84
13
14
84
100
100
Et3Al
Et3Al
90
81
80[f]
80[f]
(S)
4g
4h
4i
(R)
DiPh-Dmtb
7a
Cy-Dmtb 5a 113
15
16
136
95
98
Et3Al
Et3Al
94
87
80[f]
68[f]
(S)
(R)
DiPh-Dmtb
7a
Cy-Dmtb 5a 135
17
18
135
100
100
Et3Al
Et3Al
83
89
78[e]
75[e]
(S)
(R)
[a]
All reactions were performed at ꢀ858Cat a concentration c=0.25M. The ketene was formed in situ.
Yield determined by 1H NMR using acetophenone as internal standard.
[b]
[c]
[d]
[e]
1
Determined by H NMR.
ee determined by chiral column HPLC(Daicel OD-H, see Supporting Information).
ee determined by chiral column HPLC(Daicel OD-H) after nucleophilic ring opening of the product with ( S)-1-methyl-
benzyl amine (see Supporting Information).
[f]
ee determined by chiral column GC(Supelco Gamma Dex TM, see Supporting Information).
The configuration was determined by comparison of the [a]D values of compounds 6a, 6b, 6f and 6i with literature data
(see Supporting Information). Since a uniform reaction pathway can be assumed, the absolute configuration can be as-
signed to all cycloaddition products 6.
[g]
[h]
The isolated yield on larger scale was 90% (see Experimental Section).
1
clohexane-1,2-diamino (Cy) backbone is superior for mined after aqueous work-up by H NMR using ace-
a-unbranched aldehydes, whereas ligands derived tophenone as internal standard. However, similarly
from (S,S)-1,2-diphenyl-1,2-diamine (DiPh) are more high yields were obtained after isolation by column
efficient for the a-branched substrates. In Table 3 chromatography using a low-boiling eluent as exem-
both standard conditions for a-branched and -un- plified for 6a (87%), 6b (90%) and 6f (85% isolated
branched systems, respectively, are given for compari- yield).
son for each substrate. For the b-branched isovaleral-
In summary, we have developed an efficient and
dehyde 4f enantioselectivities were almost identical practical methodology for the aluminum-catalyzed
using the different ligand backbones, however, DiPh- enantioselective formation of b-lactones. In contrast
Trip 7b led to a considerably higher reactivity to Nelsonꢀs Al catalyst, the system described herein
(entry 9). Due to the volatility of most of the pre- tolerates also a-branched aldehydes and offers the ad-
pared b-lactones, the yields in Table 3 were deter- vantage that it is remarkably simple as the ligand syn-
1650
asc.wiley-vch.de
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 1647 – 1652

Practical Enantioselective Synthesis of b-Lactones
COMMUNICATIONS
Lett. 2005, 7, 1809; e) application in total synthesis: X.
Shen, A. S. Wasmuth, J. Zhao, C. Zhu, S. G. Nelson, J.
Am. Chem. Soc. 2006, 128, 7438; f) use of a quinine-
based bifunctional catalyst: Y.-M. Lin, J. Boucau, Z. Li,
V. Casarotto, J. Lin, A. N. Nguyen, J. Ehrmantraut,
Org. Lett. 2007, 9, 567; g) development of an intramo-
lecular version: G. S. Cortez, R. L. Tennyson, D. Romo,
J. Am. Chem. Soc. 2001, 123, 7945; h) S. H. Oh, G. S.
Cortez, D. Romo, J. Org. Chem. 2005, 70, 2835; i) H.
Henry-Riyad, C. Lee, V. C. Purohit, D. Romo, Org.
Lett. 2006, 8 4363;.
thesis requires only a simple sulfonylation step of
commercially available enantiomerically pure dia-
mines. We believe that due to the simplicity of the re-
action system it should also be interesting for techni-
cal applications.
Experimental Section
Typical Procedure
[7] a) Review: R. K. Orr, M. A. Calter, Tetrahedron 2003,
59, 3545; b) use of Cu(II) complexes: D. A. Evans,
J. M. Janey, Org. Lett. 2001, 3, 2125; c) use of Rh(II)
complexes: R. E. Forslund, J. Cain, J. Colyer, M. P.
Doyle, Adv. Synth. Catal. 2005, 347, 87; d) use of a
boron catalyst: V. Gnanadesikan, E. J. Corey, Org. Lett.
2006, 8, 4943.
[8] Y. Tamai, H. Yoshiwara, M. Someya, J. Fukumoto, S.
Miyano, J. Chem. Soc., Chem. Commun. 1994, 2281.
[9] B. W. Dymock, P. J. Kocienski, J.-M. Pons, Chem.
Commun. 1996, 1053.
[10] a) S. G. Nelson, T. J. Peelen, Z. Wan, J. Am. Chem. Soc.
1999, 121, 9742; b) S. G. Nelson, C. Zhu, X. Shen, J.
Am. Chem. Soc. 2004, 126, 14; c) application in total
synthesis: S. G. Nelson, W. S. Cheung, A. J. Kassick,
M. A. Hilfiker, J. Am. Chem. Soc. 2002, 124, 13654.
[11] N. Imai, H. Takahashi, S. Kobayashi, Chem. Lett. 1994,
177.
To a mixture of ligand 7a (0.15 mmol, 0.1 equiv.) in absolute
toluene (6.0 mL) was slowly added at ambient temperature
a
solution of AlEt3 (1.0M in hexane, 0.225 mmol,
0.15 equivs.). The mixture was heated to 808Cand stirred
for 4 h. Subsequently, the solution was stirred for 1 h at am-
bient temperature. The catalyst solution was then cooled to
ꢀ858Cand cyclohexylcarbaldehyde ( 4b, 1.5 mmol), acetyl
bromide (3, 4.5 mmol, 3 equivs.) and diisopropylethylamine
(3.75 mmol, 2.5 equivs.) were successively added. The result-
ing heterogeneous mixture was stirred at ꢀ858Cfor 25 h
until complete conversion as monitored by 1H NMR. The
reaction mixture was poured into aqueous 1M HCl (60 mL)
and extracted with diethyl ether (3”45 mL). The combined
organic phase was dried over MgSO4, filtered and diethyl
ether was removed under vacuum. The solution of the crude
product was directly used for column chromatography (pen-
tane ! pentane/diethyl ether, 8:1) without prior removal of
toluene giving 6a as colorless oil; yield: 209 mg (1.35 mmol,
90%), ee=90%.
[12] E. J. Corey, R. Imwinkelried, S. Pikul, X. B. Xiang, J.
Am. Chem. Soc. 1989, 111, 5493.
[13] Attempts with acetyl chloride instead failed. This might
be ascribed to a blocking of the hard Al ion by coordi-
nation of the hard chloride anion which is generated
during the ketene formation. The coordination of the
softer bromide anion is anticipated to be more labile.
[14] The reactions were less enantioselective with R’’=Cl
(e.g., Cy-Trip 5d, Al source Et2AlCl, ꢀ788C, ee=50%,
very slow reaction) or R’’=F (e.g., Cy-Trip 5d, Al
source i-Bu2AlF, ꢀ788C, ee=28%, very slow reaction).
[15] Both longer (6 h) and shorter (2 h) heating periods re-
sulted in less enantioselective reactions.
Acknowledgements
This work was financially supported by ETH Research
Grant TH-30/04–2. We thank Reuter Chemische Apparatebau
KG (RCA) in Freiburg, Germany for the generous donation
of enantiomerically pure 1,2-diaminocyclohexane.
References
[16] Note that with AlMe3 the signals for free ligand were
already disappeared after 30 min.
[1] Review on b-lactones: H. W. Yang, D. Romo, Tetrahe-
dron 1999, 55, 6403.
[17] Similarly, DiPh-Trip 7b/Dibal gave the following results
with (a) cyclohexanecarbaldehyde 4b (0.25M, toluene,
ꢀ858C, 45 h): 10 mol% 7b/10 mol% Dibal: 70% yield,
70% ee; 10 mol% 7b/15 mol% Dibal: 75% yield, 75%
ee; (b) dihydrocinnamaldehyde 4a (0.25M, toluene,
ꢀ858C, 44 h): 10 mol% 7b/10 mol% Dibal: 42% yield,
74% ee; 10 mol% 7b/15 mol% Dibal: 79% yield, 78%
ee.
[18] Dimeric complexes of this type have been character-
ized by X-ray structure analysis in earlier studies: E. J.
Corey, S. Sarshar, J. Am. Chem. Soc. 1992, 114, 7938.
[19] H. Yamamoto, K. Futatsugi, Angew. Chem. Int. Ed.
2005, 44, 1924.
[2] Soft nucleophiles are known to cleave preferentially
ꢀ
the alkyl O bond. b-Lactones are therefore also
masked b-substituted carboxylic acids, see ref.[1]
[3] Comprehensive reviews on ketene chemistry: a) Ke-
tenes, (Ed.: T. T. Tidwell), Wiley, New York, 1995;
b) T. T. Tidwell, Eur. J. Org. Chem. 2006, 563; c) ketene
history: T. T. Tidwell, Angew. Chem. 2005, 117, 5926;
Angew. Chem. Int. Ed. 2005, 44, 5778.
[4] C. Schneider, Angew. Chem. 2002, 114, 771; Angew.
Chem. Int. Ed. 2002, 41, 744.
[5] D. Borrmann, R. Wegler, Chem. Ber. 1967, 100, 1575.
[6] a) H. Wynberg, E. G. J. Staring, J. Am. Chem. Soc.
1982, 104, 166; b) H. Wynberg, E. G. J. Staring, J. Org.
Chem. 1985, 50, 1977; c) with LiClO4 as co-catalyst: C.
Zhu, X. Shen, S. G. Nelson, J. Am. Chem. Soc. 2004,
126, 5352; d) with metal triflate salts as co-catalysts:
M. A. Calter, O. A. Tretyak, C. Flaschenriem, Org.
[20] Similar to previous catalytic asymmetric [2+2] cycload-
ditions of the parent ketene and aldehydes, the catalyst
systems reported herein are not applicable to benzalde-
hyde and a,b-unsaturated aldehydes, since the products
decomposed during the isolation/purification proce-
Adv. Synth. Catal. 2007, 349, 1647 – 1652
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1651

Thomas Kull and RenØ Peters
COMMUNICATIONS
dures presumably due to elimination reactions. The
same problem was faced with conjugated ynals.
[21] No clear picture is evident with regard to the reaction
times: while in some cases, the conversion is complete
after one day even with sterically hindered aldehydes
such as cyclohexanecarbaldehyde 4b or isovaleralde-
hyde 4f, for other substrates up to 6 days at ꢀ858Care
necessary to achieve full conversion.
1652
asc.wiley-vch.de
ꢁ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 1647 – 1652