Enantioselective Cu-catalyzed 1,4-addition of Grignard reagents to cyclohexenone using Taddol-derived phosphine-phosphite ligands and 2-methyl-THF as a solvent

Article abstract of DOI:10.1002/anie.200803247

A small library of modular P,P ligands was screened and a potent Cu-based catalyst system was identified for highly enantioselective 1,4-additions to cyclohexenone with an unprecedented broad spectrum of Grignard reagents. Surprisingly, the highest selectivities were achieved in most cases in 2-methyl-THF, a green solvent underestimated so far.

Full text of DOI:10.1002/anie.200803247

Communications
DOI: 10.1002/anie.200803247
Grignard Reagents
Enantioselective Cu-Catalyzed 1,4-Addition of Grignard Reagents to
Cyclohexenone Using Taddol-Derived Phosphine–Phosphite Ligands
and 2-Methyl-THF as a Solvent**
Tobias Robert, Janna Velder, and Hans-Günther Schmalz*
Professor Helmut Schwarz zum 65. Geburtstag gewidmet
The 1,4-addition (conjugate addition) of C nucleophiles to
a,b-unsaturated carbonyl compounds and related activated
give the 1,4-addition products 4 with excellent enantioselec-
tivity (up to 96% ee) and good regioselectivity (4/5 ꢁ 4:1)
(Scheme 1).[4,7] Furthermore, Feringa and co-workers were
ꢀ
olefins belongs to the most powerful and reliable tools for C
C bond formation. Accordingly, it is frequently exploited in
the synthesis of natural products and other complex organic
molecules.[1] Three decades of research on asymmetric[2] and,
in particular, asymmetric catalytic[3] versions of conjugate
addition reactions have resulted in the development of a
broad variety of methods.[1–3] However, the important task of
performing the 1,4-addition of simple Grignard reagents, the
most common type of organometallic reagents, in an enan-
tioselective fashion still remains a particular challenge.
As recently reviewed,[4a] several Cu-based catalyst systems
have been suggested in the past for the catalytic asymmetric
1,4-addition of Grignard reagents.[4] Nevertheless, high enan-
tioselectivities (ꢁ 90% ee) were achieved only in a few special
cases, and the reported methods did not find much application
owing to limited substrate scope, operational convenience,
and accessibility of the chiral ligands required. In 2004, an
important advance was made by Feringa and co-workers who,
by screening a set of commercially available chiral P,P ligands,
identified ferrocene-based diphosphines, in particular Tania-
phos (1)[5] and Josiphos (2),[6] as promising ligands for such
transformations.[7]
Scheme 1. Cu-catalyzed 1,4-addition of Grignard reagents to cyclohex-
[7]
enone accordingto Feringa
.
able to apply this catalytic system to other substrates and in
the synthesis of natural products.[8] However, the reaction of 3
with other relevant Grignard reagents such as iPrMgBr or
PhMgBr proceeded only with low selectivity under these
conditions, and also with an alternative catalyst (formed from
2 and CuBr–SMe2) ee values did not exceed 54% (R = iPr)
and 40% (R = Ph). Thus, the challenge remained open.
Considering that chiral diphosphine ligands possess an
obvious potential for the Cu-catalyzed 1,4-addition of
Grignard reagents[7,8] we wondered whether phosphine–
phosphite ligands of type 7[9] (developed in our laboratory)
would also be suited for this purpose.[10] These ligands are
efficiently prepared from o-bromophenols 6 and chiral diols
(such as Binol (2,2’-dihydroxy-1,1’-binaphthyl)[11] and Taddol
(a,a,a’-tetraaryl-2,2-dimethyl-1,3-dioxolan-4,5-dimetha-
nol)[12]), and the modular synthesis facilitates structural
variation and optimization (Scheme 2). We herein report
that compounds of this class indeed represent highly useful
ligands for the Cu-catalyzed 1,4-addition of various Grignard
For instance, the reaction of n-alkyl Grignard reagents
with cyclohexenone (3) proceeded smoothly in the presence
of 5 mol% of a catalyst generated in situ from 1 and CuCl to
[*] Dipl.-Chem. T. Robert, Dr. J. Velder, Prof. Dr. H.-G. Schmalz
Institut für Organische Chemie, Universität zu Köln
Greinstrasse 4, 50939 Köln (Germany)
Fax: (+49)221-470-3064
E-mail: schmalz@uni-koeln.de
[**] This work was carried out in the context of Cost D40 and supported
by the European Commission (Ligbank), the Chemetall GmbH, and
the Fonds der Chemischen Industrie.
Supportinginformation for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200803247.
Scheme 2. Modular phosphine-phosphite ligands of type 7 derived
from o-bromophenols 6 and chiral diols.
7718
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 7718 –7721

Angewandte
Chemie
reagents (including PhMgBr and iPrMgBr) to cyclohexenone
giving enantioselectivities of up to 92% ee.
compounds and to vary the reaction parameters such as
copper source, solvent, and reaction temperature (Table 2).
Much to our surprise, the choice of the solvent had an
extraordinary influence on the enantioselectivity. While
In a first series of experiments a library of ten ligands of
type 7 was screened in the Cu-catalyzed addition of ethyl
magnesium bromide to cyclohexenone under conditions
similar to those used by Feringa et al. (4 mol% CuCl,
6 mol% ligand, Et2O, 0 8C), which also in our hands gave
complete conversion within 20 min. As the results summar-
ized in Table 1 show, the Taddol-derived ligands 7e and 7 f
Table 2: Influence of solvent and temperature on the Cu-catalyzed
addition of EtMgBr to cyclohexenone according to Scheme 1 (R=Et)[aj
usingthe Taddol-derived chiral ligand 7 f.
Entry
Solvent
T [8C]
Cu source
4a/5a[b]
ee[c] [%]
1
2
3
4
5
6
7
8
9
Et2O
THF
MTBE
0
0
0
0
0
CuCl
CuCl
CuCl
CuCl
CuCl
CuCl
CuCl
CuCl
95:5
95:5
96:4
86:14
96:4
95:5
96:4
95:5
90:10
89:11
89:11
68
<5
56
58
60
84
86
82
86
88
90
Table 1: Results of the screeningof ligands of type 7 in the Cu-catalyzed
reaction of EtMgBr with cyclohexenone according to Scheme 1 (R=Et).[a]
Ligand
Ligand specification[b]
4a/5a[c]
ee[d] [%]
Config.
n.d.
toluene
toluene/Et2O
Me-THF
Me-THF
Me-THF
Me-THF
Me-THF
Me-THF
7a
R1 =R2 =R3 =R4 =H
95:5
<5
0
R*(OH)2 =(R,R)-Taddol
R1 =R2 =Me, R3 =R4 =H
R*(OH)2 =(R,R)-Taddol
R1 =Ph, R2 =R3 =R4 =H
R*(OH)2 =(R,R)-Taddol
R1 =iPr, R2 =R3 =R4 =H
R*(OH)2 =(R,R)-Taddol
R1 =tBu, R2 =R3 =R4 =H
R*(OH)2 =(R,R)-Taddol
R1 =R3 =tBu, R2 =R4 =H
R*(OH)2 =(R,R)-Taddol
R1 =R4 =Me, R2 =R3 =H
R*(OH)2 =(R,R)-Taddol
R3 =R4 =H, R1/R2 =(CH)4
R*(OH)2 =(R,R)-Taddol
R1 =R2 =Me, R3 =R4 =H
R*(OH)2 =(S)-Binol
ꢀ20
ꢀ40
ꢀ40
ꢀ60
ꢀ78
7b
7c
7d
7e
7 f
7g
7h
7i
99:1
99:1
99:1
95:5
95:5
92:8
99:1
21:79
23:77
8
10
20
70
68[e]
16
16
2
S
S
S
S
S
R
S
S
S
CuBr–SMe2
CuBr–SMe2
CuBr–SMe2
10
11
[a] Reaction conditions: 6 mol% 7 f, 4 mol% Cu source. [b] Determined
by GC-MS. [c] Determined by GC on a chiral stationary phase (BGB 176
SE).
replacing Et2O by tert-butyl methyl ether (MTBE) did not
have a major effect on the ee value, the use of THF as a
stronger coordinating solvent led to almost a complete loss of
enantioselectivity. However, when the solvent was changed to
2-methyltetrahydrofuran (2-Me-THF),[14]
a
significant
increase of the enantioselectivity by almost 20% ee was
observed. Lowering the reaction temperature to ꢀ788C led to
a further enhancement of the enantioselectivity. The choice of
the copper source did not have much effect at 08C; however,
the use of CuBr–SMe2 resulted in better ee values at low
temperatures, probably because it is better soluble than CuCl.
Under optimized conditions (Table 2, entry 11) the 1,4-
addition product 4a (R = Et) was obtained with a pleasing
enantioselectivity of 90% ee. This result could be reproduced
also on a multi-mmol scale (60% yield of 4a after distillation
and flash chromatography) when the diluted Grignard
solution (0.1–0.5m) was added very slowly to the stirred
reaction mixture by means of a syringe pump.
Encouraged by this result (Table 2, entry 11) we decided
to probe the scope of our methodology by applying other
Grignard reagents. Besides PhMgBr and iPrMgBr, which (as
mentioned above) gave only unsatisfactory enantioselectiv-
ities with ligands 1 and 2,[7] we considered 3-butenyl and
isopropenyl magnesium bromide as particularly relevant,
because the corresponding 1,4-addition products would
represent interesting building blocks for natural product
synthesis. To our satisfaction, the reaction proceeded well
with all of these Grignard reagents under the standard
conditions (using ligands 7e and 7 f) to give the 1,4-addition
products 4a–e with high to excellent enantioselectivities (82–
92% ee; see Table 3).
7j
R1 =R3 =tBu, R2 =R4 =H
R*(OH)2 =(S)-Binol
6
[a] Reaction conditions: 6 mol% L*, 4 mol% CuCl, Et2O, 08C, full
conversion after 20 min. [b] Only ligands with R5 =Ph were used.
[c] Determined by GC-MS. [d] Determined by GC on a chiral stationary
phase. [e] Conducted on a 3 mmol scale; the pure 1,4-product was
obtained in 68% yield after distillation.
gave the highest enantioselectivities of 68 and 70% ee,
respectively.[13] Apparently the enantioselectivity strongly
depends on the size of R1 (cf. Scheme 2) as the ligands with
a bulky tert-butyl substituent in the position ortho to the
phosphite group performed much better than ligands with
sterically less demanding substituents in this position. Also
noteworthy is the fact that the Binol-derived ligands (7i and
7j) not only exhibited low enantioselectivities but also mainly
gave rise to the regioisomeric products resulting from 1,2-
addition.
Having identified 7e and 7 f (Figure 1) as promising
ligands, we decided to continue our examinations with these
Again, 2-Me-THF was found to be a superior solvent in
most cases; however, there are exceptions. For instance, the
highest ee for 3-butenyl-MgBr was achieved in neat Et2O. In
the case of iPrMgBr optimal results were obtained when Et2O
Figure 1. Taddol-derived phosphine–phosphite ligands 7e and 7 f.
Angew. Chem. Int. Ed. 2008, 47, 7718 –7721
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7719

Communications
Table 3: Cu-catalyzed asymmetric 1,4-addition of various Grignard
reagents to cyclohexenone according to Scheme 1.[a]
Entry RMgBr
L*
Solvent
4/5[b]
ee [%][b] Config.[c]
1
2
3
4
EtMgBr
7 f
7 f
7 f
Me-THF 89:11 90
(ꢀ)-S
(ꢀ)-R
(+)-R
(+)-R
Et2O
91:9
98:2
99:1
91
67
82
iPrMgBr
iPrMgBr
Et2O[d]
7e Et2O[d]
5
7e Me-THF 91:9
7 f
7e Me-THF 91:9
92
(+)-R
6
7
PhMgBr
PhMgBr
Me-THF 60:40 74
92
(+)-R
(+)-R
[a] Reaction conditions: 6 mol% ligand, 5 mol% CuBr–SMe2, ꢀ788C,
full conversion after 2 h. [b] Determined by GC on a chiral stationary
phase. [c] Sign of the optical rotation at 589 nm in CH2Cl2 and absolute
configuration as determined by CD spectroscopy. [d] A solution of the
Grignard reagent in Me-THF was used.
Figure 3. Structure and enantiomeric purities (and yields) of 3-substi-
tuted cyclohexanones synthesized in this study.
high enantioselectivities (82–92% ee) and preparative yields
of up to 88%.[15]
was used to dissolve the substrate (3) and the catalyst, while
the Grignard reagent was prepared and used as a solution in
2-Me-THF. Performing the reaction with iPrMgBr in either
neat 2-Me-THF or Et2O resulted in low enantioselectivities
and, moreover, slow conversion in the latter case. Also
noteworthy is the fact that at ꢀ788C ligand 7e proved to be
superior for branched Grignard reagents, while the 1,4-
addition of both linear Grignard reagents proceeded most
selectively when ligand 7 f was employed. The other ligands of
type 7 (Table 1) were found to be much less selective.
In conclusion, we have identified phosphine–phosphite
ligands of type 7 as a novel (second) class of chiral P,P ligands
suitable for the Cu-catalyzed asymmetric 1,4-addition of
Grignard reagents to cyclohexenone.[16,17] These readily
accessible ligands proved to be compatible with an unsur-
passed range of Grignard reagents, and owing to their
modular nature it should be possible to achieve further
(individual) ligand tuning for specific reaction systems.[9b,10]
Furthermore, we came across a very interesting solvent
effect with (racemic) 2-methyl-THF, which has only recently
been recognized as an environmentally benign solvent
because of its low water miscibility and its origin from
renewable resources.[18] Our results suggest that this solvent
should be generally considered as an alternative to diethyl
ether and THF whenever organometallic reactions are
optimized.
The absolute configuration of the products (as given in
Table 3) was determined by means of CD spectroscopy
(Figure 2) applying the octant rule.[14] As Figures 2 and 3
illustrate, the stereochemical outcome of the reactions
Experimental Section
Typical procedure for the 1,4-addition on a preparative scale: Under
an atmosphere of argon, CuBr–SMe2 (0.05 equiv, 0.45 mmol) and the
ligand (0.06 equiv, 0.54 mmol) were dissolved in 15 mL of solvent and
stirred for 15 min at RT. After addition of enone 3 (1.0 equiv, 9 mmol)
the reaction mixture was cooled to ꢀ788C and a dilute (0.1–
0.5 molLꢀ1) solution of the Grignard reagent (1.2 equiv, 10.8 mmol)
was slowly added over 2 h by means of a syringe pump. The mixture
was then stirred for another 30 min at ꢀ788C and quenched by
addition of MeOH (5 mL) and 1m aqueous NH4Cl solution (10 mL).
The layers were separated, and the aqueous phase was extracted with
tert-butyl methyl ether. The combined organic solutions were washed
with brine and dried over MgSO4, and the solvent was evaporated in
vacuo. The crude product (yellowish oil) was purified by distillation
(0.2 mbar) and subsequent flash chromatography (cyclohexane/
EtOAc 10:1) to give the pure 1,4-addition product 4 as a colorless oil.
Figure 2. CD spectra of compounds 4a–e in CH3CN.
depends on the Grignard reagent used. In the case of the
unbranched reagents (EtMgBr and 3-butenylmagesium bro-
mide) the main enantiomers of the products (4a, 4b) result
from a Si-face attack at the position b to the ketone. However,
using the same (R,R)-Taddol-derived catalysts, the 1,4-
addition of the branched reagents afforded products 4c–4e,
resulting from a Re-face attack. We consider this switch of the
enantiofacial selectivity with the type of Grignard reagent as a
remarkable phenomenon which challenges any future
attempts to rationalize the stereochemical outcome of these
reactions based on detailed transition-state models. As
Figure 3 summarizes, the methodology developed opens
access to a variety of 3-substituted cyclohexanones with
Received: July 4, 2008
Published online: September 4, 2008
Keywords: asymmetric catalysis · conjugate addition ·
.
Grignard reagents · phosphine ligands · solvent effects
[1] Reviews: a) P. Perlmutter, Tetrahedron Organic Chemistry Series
9; Pergamon, Oxford, UK, 1992; b) V. J. Lee in Comprehensive
7720 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 7718 –7721

Angewandte
Chemie
Organic Synthesis, Vol. 4 (Eds.: B. M. Trost, I. Fleming),
Pergamon, Oxford, 1991, pp. 69 – 137; c) V. J. Lee in Compre-
hensive Organic Synthesis, Vol. 4 (Eds.: B. M. Trost, I. Fleming),
Pergamon, Oxford, 1991, pp. 139 – 168; d) J. A. Kozlowski in
Comprehensive Organic Synthesis, Vol. 4 (Eds.: B. M. Trost, I.
Fleming), Pergamon, Oxford, 1991, pp. 169 – 198.
[9] a) J. Velder, T. Robert, I. Weidner, J.-M. Neudörfl, J. Lex, H.-G.
Schmalz, Adv. Synth. Catal. 2008, 350, 1309 – 1315; b) F. Blume,
S. Zemolka, T. Fey, R. Kranich, H.-G. Schmalz, Adv. Synth.
Catal. 2002, 344, 868 – 883; c) R. Kranich, K. Eis, O. Geis, S.
Mꢀhle, J. W. Bats, H.-G. Schmalz, Chem. Eur. J. 2000, 6, 2874 –
2894.
[2] Reviews: a) H.-G. Schmalz in Comprehensive Organic Synthesis,
Vol. 4 (Eds.: B. M. Trost, I. Fleming), Pergamon, Oxford, 1991,
pp. 199 – 236; b) B. E. Rossiter, N. M. Swingle, Chem. Rev. 1992,
92, 771 – 806.
[10] For the application of ligands of type 7 in the highly enantio-
selective Rh-catalyzed hydroboration of styrenes, see Ref. [9b]
as well as: S. Werle, T. Fey, J. M. Neudörfl, H.-G. Schmalz, Org.
Lett. 2007, 9, 3555 – 3558; Erratum: S. Werle, T. Fey, J. M.
Neudörfl, H.-G. Schmalz, Org. Lett. 2007, 9, 4085.
[11] For a recent review, see: a) J. M. Brunel, Chem. Rev. 2005, 105,
857 – 898; Erratum: J. M. Brunel, Chem. Rev. 2005, 105, 857 –
898; for an update, see b) J. M. Brunel, Chem. Rev. 2007, 107, 1 –
45.
[12] a) D. Seebach, A. K. Beck, R. Imwinkelried, S. Roggo, A.
Wonnacott, Helv. Chim. Acta 1987, 70, 954. For a review, see:
b) D. Seebach, A. K. Beck, A. Heckel, Angew. Chem. 2001, 113,
96 – 142; Angew. Chem. Int. Ed. 2001, 40, 92 – 138.
[13] In a control experiment under exactly the same conditions,
ligand 1 (Taniaphos) afforded the product (4, R = Et) with 92%
ee, which is in acceptable agreement to the results of Feringa and
coworkers (Ref. [7]) who reported 96% ee with a slightly higher
catalyst loading.
[14] For reviews, see: a) E. L. Eliel, S. H. Wilen, L. N. Mander in
Stereochemistry of Carbon Compounds, VCH, New York, 1994,
pp. 1007 – 1043; b) D. N. Kirk, Tetrahedron 1986, 42, 777 – 818;
c) J. K. Gawronski, Tetrahedron 1982, 38, 3 – 26; d) R. D. Bur-
nett, D. N. Kirk, J. Chem. Soc. Perkin Trans. 1 1981, 1460 – 1468.
[15] Yields (obtained on a 9 mmol scale) are in the same range as
reported by Feringa et al. (Ref. [7]); they are sometimes
impaired through the formation of by-products resulting from
aldol-type processes.
[16] For asymmetric Cu-catalyzed 1,4-additions of Grignard reagents
to 3-alkylcyclohexenones using chiral N-heterocyclic carbene
ligands, see: a) D. Martin, S. Kehrli, M. dꢂAugustin, H. Clavier,
M. Mauduit, A. Alexakis, J. Am. Chem. Soc. 2006, 128, 8416 –
8417; b) Y. Matsumoto, K.-i. Yamada, K. Tomioka, J. Org. Chem.
2008, 73, 4578 – 4581.
[3] a) N. Krause, A. Hoffmann-Röder, Synthesis 2001, 171 – 196;
b) J. Christoffers, G. Koripelly, A. Rosiak, M. Rössle, Synthesis
2007, 1279 – 1300; c) K. Tomioka, Y. Nagaoka in Comprehensive
Asymmetric Catalysis, Vol. III (Eds.: E. N. Jacobsen, A. Pfaltz,
H. Yamamoto), Springer, Berlin, 1999, pp. 1105 – 1120.
[4] a) Review: F. López, A. J. Minaard, B. L. Feringa, Acc. Chem.
Res. 2007, 40, 179 – 188; for key publications in this field, see:
b) K.-H. Ahn, R. B. Klassen, S. J. Lippard, Organometallics
1990, 9, 3178 – 3181; c) D. M. Knotter, D. M. Grove, W. J. J.
Smeets, A. L. Spek, G. van Koten, J. Am. Chem. Soc. 1992, 114,
3400 – 3410 d) D. Spescha, G. Rihs, Helv. Chim. Acta 1993, 76,
1219 – 1230; e) Q.-L. Zhou, A. Pfaltz, Tetrahedron 1994, 50,
4467 – 4478; f) A. L. Braga, S. J. N. Silva, D. S. Lꢀdtke, R. L.
Drekener, C. C. Silveira, J. B. T. Rocha, L. A. Wessjohann,
Tetrahedron Lett. 2002, 43, 7329 – 7331; g) D. Seebach, G.
Jaeschke, A. Pichota, L. Audergon, Helv. Chim. Acta 1997, 80,
2515 – 2519; h) M. Kanai, Y. Nakagawa, K. Tomioka, Tetrahe-
dron 1999, 55, 3843 – 3854; i) E. L. Stangeland, T. Sammakia,
Tetrahedron 1997, 53, 16503 – 16510.
[5] T. Ireland, G. Grossheimann, C. Wieser-Jeunesse, P. Knochel,
Angew. Chem. 1999, 111, 3397 – 3400; Angew. Chem. Int. Ed.
1999, 38, 3212 – 3215.
[6] A. Togni, C. Breutel, A. Schnyder, F. Spindler, H. Landert, A.
Tijani, J. Am. Chem. Soc. 1994, 116, 4062 – 4066.
[7] B. L. Feringa, R. Badorrey, D. Peꢁa, S. R. Harutyunyan, A. J.
Minaard, Proc. Natl. Acad. Sci. USA 2004, 101, 5834 – 5838.
[8] a) F. López, S. R. Harutyunyan, A. J. Minaard, B. L. Feringa, J.
Am. Chem. Soc. 2004, 126, 12784 – 12785; b) F. López, S. R.
Harutyunyan, A. Meetsma, A. J. Minaard, B. L. Feringa, Angew.
Chem. 2005, 117, 2812 – 2816; Angew. Chem. Int. Ed. 2005, 44,
2752 – 2756; c) S. R. Harutyunyan, F. López, W. R. Browne, A.
Correa, D. Peꢁa, R. Badorrey, A. Meetsma, A. J. Minaard, B. L.
Feringa, J. Am. Chem. Soc. 2006, 128, 9103 – 9118; d) R.
Des Mazery, M. Pullez, F. López, S. R. Harutyunyan, A. J.
Minaard, B. L. Feringa, J. Am. Chem. Soc. 2005, 127, 9966 – 9967;
e) S.-Y. Wang, S.-J. Ji, T.-P. Loh, J. Am. Chem. Soc. 2007, 129,
276 – 277; f) T.-K. Lum, S.-Y. Wang, T.-P. Loh, Org. Lett. 2008, 10,
761 – 764.
[17] To the best of our knowledge, only one example for the use of a
chiral phosphine–phosphite ligand in the Cu-catalyzed asym-
metric conjugate addition (of Et2Zn and Et3Al) has been
reported; see: M. Diꢃguez, S. Deerenberg, O. Pꢄmies, C.
Claver, P. W. N. M. van Leeuwen, P. C. J. Kamer, Tetrahedron:
Asymmetry 2000, 11, 3161 – 3166.
[18] D. F. Aycock, Org. Process Res. Dev. 2007, 11, 156 – 159.
Angew. Chem. Int. Ed. 2008, 47, 7718 –7721
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7721