Design and development of bioinspired guanine-based organic catalyst for asymmetric catalysis

Article abstract of DOI:10.1002/ejoc.201200118

Design, preparation, and studies of a family of new organic catalysts are presented. The design of the catalysts is inspired by the ability of DNA nucleobases to develop precise and explicit hydrogen bonds. We have shown that this phenomenon can be used to create a useful organic catalyst that demonstrates a recognition pattern similar to those of common organic substrates. A selected bifunctional catalyst based on a guanine structure has been shown to catalyze the conjugate addition of 1,3-dicarbonyl compounds to various nitroalkenes, providing the products in good yields and enantioselectivities.

Full text of DOI:10.1002/ejoc.201200118

SHORT COMMUNICATION
DOI: 10.1002/ejoc.201200118
Design and Development of Bioinspired Guanine-Based Organic Catalyst for
Asymmetric Catalysis
Gal Suez,[a] Victoria Bloch,[a] Gennady Nisnevich,[a] and Mark Gandelman*[a]
Keywords: Hydrogen bonds / Catalyst design / Organocatalysis / Asymmetric catalysis / Michael addition
Design, preparation, and studies of a family of new organic
catalysts are presented. The design of the catalysts is in-
spired by the ability of DNA nucleobases to develop precise
and explicit hydrogen bonds. We have shown that this phe-
nomenon can be used to create a useful organic catalyst that
demonstrates a recognition pattern similar to those of com-
mon organic substrates. A selected bifunctional catalyst
based on a guanine structure has been shown to catalyze the
conjugate addition of 1,3-dicarbonyl compounds to various
nitroalkenes, providing the products in good yields and
enantioselectivities.
dition of 1,3-dicarbonyl compounds to nitroolefins is dem-
onstrated.
Introduction
The use of small organic molecules to catalyze organic
reactions (organocatalysis), with particular appeal to asym-
metric catalysis, has gained much attention within the
chemical community in recent years.[1] Although the area
is relatively new, organocatalysis has become an attractive
strategy for the catalytic preparation of chiral molecules
complementary to valuable organometallic and enzyme-
based approaches. The field can be classified according to
generic modes of catalyst activation and induction,[2]
whereas the activation of substrates by hydrogen-bond-do-
nor organic molecules represents one of the major direc-
tions.[3] It has been demonstrated that, as in complex enzy-
matic systems, certain relatively simple organic scaffolds can
efficiently activate the substrate and control stereoselectivity
through well-defined hydrogen-bonding interactions. Un-
doubtedly, progress in the field relies on the discovery and
design of both new modes of activation and new catalyst
architectures. Development of new efficient and tunable or-
ganic catalysts is of conceptual and practical importance,
as it might lead to the discovery of novel organic transfor-
mations and guide to a deeper understanding of mechan-
istic features of the studied activation mode. Here, we pres-
ent the design and development of a new organic catalyst,
inspired by the guanine nucleobase structure and its ability
to develop explicit hydrogen bonds in biological systems.
The capability of this robust and structurally simple catalyst
to enantioselectively mediate the benchmark conjugate ad-
Results and Discussion
Our approach to organic catalyst design was inspired by
the phenomenon of specific molecular recognition of DNA
bases. It is well recognized that guanine–cytosine (G–C)
and adenine–thymine (A–T) exhibit relatively strong and se-
lective pairing by formation of reciprocal hydrogen bonds
(Figure 1a).[4] We assumed that if a nucleobase selectively
recognizes its complementary counterpart by hydrogen
bonding interactions, then it may also recognize certain or-
ganic molecules that possess a structural hydrogen-bonding
motif similar to that of its complementary base. For exam-
ple, precursors bearing ester or nitro functional groups are
attractive candidates (Figure 1b). The notion that nucleo-
bases such as guanine could recognize and activate electro-
philes through a multiple-hydrogen-bonding mechanism
finds solid support in the scientific literature: artificial re-
ceptors possessing electrophilic acceptor patterns (N–H
units) similar to guanine and capable of selectively binding
to carboxylic acids and carboxylates via multiple hydrogen
bonds to their oxygen atoms (donor unit) have been re-
ported.[5] In addition, the relatively strong and directional
hydrogen bonds formed by guanine and its structurally
closely related species in noncovalent assemblies represent
an important facet of supramolecular chemistry.[6] It should
be mentioned that structurally related, mainly bicyclic,
guanidines have been shown to be efficient organocatalysts[7]
(Figure 1c). It is suggested, however, that these molecules
provide a single hydrogen bond (N–H) for electrophile
binding, whereas the adjacent nitrogen atom serves as an
internal base, thus making guanidine a bifunctional cata-
lyst.[7a,8] Guanidinium salts have been shown to act as cata-
[a] Schulich Faculty of Chemistry, Technion- Israel Institute of
Technology
Technion City, Haifa 32000, Israel
Fax: +972-4-829-5703
E-mail: chmark@tx.technion.ac.il
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201200118.
2118
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2012, 2118–2122

Bioinspired Guanine-Based Organic Catalyst for Asymmetric Catalysis
lysts operating by presumably dual hydrogen-bond do- stalled for solubility reasons. We examined the catalytic
nation (Figure 1d).[9a] Such guanidinium ions are also sug- properties of 3 in the conjugate addition of malonates to
gested to be active catalysts formed by protonation of neu- nitroalkenes. This carbon–carbon-bond-forming reaction
tral guanidine during the course of the catalyzed reac- leads to valuable building blocks in synthetic chemistry.[12]
tion.[9b]
Recently, several hydrogen-donor organic molecules have
proven to be efficient catalysts in this reaction,[13] and as
such it might serve as a benchmark for the evaluation of
the activity of new organocatalytic systems. When diethyl-
malonate was added to nitrostyrene in the presence of
20 mol-% of 3 and 1 equiv. of triethylamine in toluene at
room temperature, the formation of addition product 4a
was observed in 60% yield after 24 hours (Table 1, entry 1).
No product was observed when the reaction was performed
without triethylamine, despite the addition of 3. This con-
trol experiment indicates that molecule 3 has no suitable
basic moiety for the activation of malonate. On the other
hand, when the reaction was run under identical conditions
in the presence of 1 equiv. of triethylamine and without 3,
product 4a was formed in 17% yield only, which clearly
indicates the rate enhancement with catalyst 3. Disappoint-
ingly, analysis of the product exhibited no enantiomeric ex-
cess when enantiopure catalyst 3 was applied.
Figure 1. (a) The guanine–cytosine pair. (b) The hypothesized
mode of substrate–catalyst binding. (c) A representative guanidine-
based catalyst. (d) A representative guanidinium catalyst. (e) Cata-
lyst 1.
Table 1. Enantioselective Michael addition of diethylmalonate to
nitrostyrene in the presence of 3 and 5a–n.
The generic structure of the proposed neutral guanine-
based catalyst 1 is presented in Figure 1e. Several features
make this scaffold attractive as a potential organocatalyst.
Compound 1 can donate two hydrogen bonds simulta-
neously for use in electrophile binding and activation. This
two-point binding, which might be important for a rigid
transition state and hence improved stereoselectivity, has
proven to be a highly successful strategy for electrophile
activation, both in enzymes and in synthetic catalytic sys-
tems.[1,3] N–H protons, which are supposed to be responsi-
ble for the hydrogen bonding to the substrate, are consider-
ably acidic.[10] Recently, it was demonstrated that the maxi-
mum possible rate acceleration is observed in some organo-
catalyst-mediated reactions when a catalyst with more
acidic hydrogen bonding is utilized.[11] The ability to pre-
pare various chiral catalysts by a single reaction of 2
(Scheme 1) with chiral amines offers tunability to the basic
structure. The fused aromatic ring (installed instead of the
imidazole unit of the original guanine) in catalyst 1 allows
the introduction of various functionalities in its backbone
and provides better solubility in common organic solvents.
Entry Catalyst Time (days)
Yield (%)[a]
ee (%)[b]
1
2
3
4
5
6
7
8
3
1
1
4
3
1
1
1
1
7
6
6
6
6
1
1
60
94
54
78
23
61
84
15
30
10
11
33
n.r.[c]
74
27
0
84
68
90
80
77
82
92
5a
5b
5c
5d
5e
5f
5g
5h
5i
5j
5k
5l
9
n.d.[c]
n.d.[c]
11
10
11
12
13
14
15
45
n.d.[c]
73
5m
5n
0
[a] Isolated yield. [b] Enantiomeric excess was determined by HPLC
analysis of 4a using a chiral column. [c] Abbreviations: n.d. (not
determined); n.r. (no reaction).
In view of these results, we decided to covalently connect
a base, which is required for the reaction, to the hydrogen-
donor guanine-like backbone in order to form a bifunc-
tional catalyst (Figure 2). Bifunctional organocatalysts cap-
able of simultaneous activation of nucleophile and electro-
phile have proven very efficient for both enantioselectivity
Scheme 1. Synthesis of molecule 3.
To test our ideas, molecule 3 was prepared from readily and rate enhancement.[14] As such, a series of bifunctional
available chloride 2 (see Supporting Information) and phen- potential catalysts was conveniently prepared by starting
ylethylamine (Scheme 1). The latter was used in order to from the same precursor 2 and utilizing various chiral di-
introduce both an N–H donor unit and asymmetry in the amines instead of phenylethylamine (Scheme 1 and Fig-
potential catalyst 3. The trifluoromethyl group in 3 was in- ure 2, 5a–l; see Supporting Information for synthetic details
Eur. J. Org. Chem. 2012, 2118–2122
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2119

G. Suez, V. Bloch, G. Nisnevich, M. Gandelman
SHORT COMMUNICATION
and X-ray crystallographic confirmation of the molecular comes. Thus, installation of an additional trifluoromethyl
structure for selected catalysts). These include basic func- group on the catalyst backbone did not improve the results
tions bearing primary, secondary, and cyclic/acyclic tertiary (entry 7); however, a removal of the parent CF3 greatly re-
amines, as well as chiral units based on trans-cyclohexane duces the rate of the process (entry 9). The rate was also
(5a–h),[15] diphenylethane (5i,j), and indane (5k,l) scaffolds. significantly reduced when the trans-cyclohexane diamine
Structurally related molecules 5m,n were also prepared for unit was replaced by a conformationally unrestricted, acy-
comparative studies. The behavior of compounds 5a–n was clic diphenylethylene diamine (allowing a stable anti confor-
examined in the above-mentioned reaction of nitrostyrene mation; entry 10). Interestingly, isocytosine-based catalyst
and diethylmalonate, and the results are summarized in 5m, possessing no additional aromatic ring, brings notable
Table 1. As can be seen from this data, the bifunctional ap- but not dramatic changes (entry 14); however, installing an
proach proved distinctively superior in terms of enantio- amide instead of the key guanine unit (5n; entry 15) signifi-
selectivity. Gratifyingly, when catalyst 5a was used, product cantly diminishes both yield and ee. The latter experiment
4a was isolated in 94% yield and 84% ee (enantiomeric ex- indicated that the two-point hydrogen-binding fragment is
cess) after 24 hours (entry 2). Modifications of the amine crucial for activation of the substrate in order to provide
base gave usually lower yield and ee (entries 3–6), except good yield and ee.
for catalyst 5c, for which a somewhat higher enantiomeric
Compound 5a was selected as an optimal catalyst with
excess (90%) of product 4a was obtained (entry 4). How- respect to the reaction outcome, and it was used to study
ever, the reaction was slower and only 78% yield was ob- the reaction parameters (Table 2). We found that toluene
tained after three days. Other modifications involving sub- proved to be superior for our model reaction over all other
stituents on the aromatic ring (entries 7–9) or in the chiral tested solvents. Additives such as water or molecular sieves
scaffold (entries 10–13) did not provide better reaction out- did not improve the results (entries 6 and 9). The enantio-
selectivity can be slightly improved if the reaction is carried
out at low temperatures, yet at the expense of a substan-
tially longer reaction time (entry 8). Therefore, we decided
to examine the scope of the conjugate addition of 1,3-di-
carbonyl compounds to nitroalkenes by applying 20 mol-%
of catalyst 5a in toluene at room temperature.
Table 2. Enantioselective Michael addition of diethylmalonate to
nitrostyrene in the presence of 5a under various conditions.
Entry
Solvent
Yield (%)[a]
ee (%)[b]
1
2
3
4
toluene
hexane
CH2Cl2
THF
THF/H2O
toluene
toluene
toluene
toluene/H2O
94
75
62
n.d.
22
32
60
94
45
84
80
78
49
45
48
86
87
84
5[c]
6[d]
7[e]
8[f]
9[c]
[a] Isolated yield. [b] The enantiomeric excess was determined by
HPLC analysis of 4a using a chiral column. [c] The reaction was
conducted with 20 mol-% of H2O. [d] The reaction was conducted
with 4 Å molecular sieves. [e] The reaction was conducted with
10 mol-% of 5a. [f] The reaction was carried out at 0 °C for 4 d.
We were pleased to find that various nitroalkenes are
suitable substrates for this catalytic reaction: substituted
aromatic, heterocyclic, and aliphatic nitroalkenes afforded
moderate to excellent yields and very good ee values
(Table 3).
Moreover, a number of 1,3-dicarbonyl compounds, in-
cluding malonates and diketones, provided the products in
high yields and very good levels of enantioselectivity upon
addition to nitrostyrene (Table 4). Quite a rare example of
Figure 2. Structures of bifunctional catalysts.
2120
www.eurjoc.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2012, 2118–2122

Bioinspired Guanine-Based Organic Catalyst for Asymmetric Catalysis
Table 3. Enantioselective Michael addition of diethylmalonate to no enantioselectivity in the model reaction of nitrostyrene
various nitroalkenes 6a–h in the presence of 5a.
and diethylmalonate.
A molecular structure of 8b obtained by X-ray analysis
of its single crystals demonstrates two well-defined hydro-
gen bonds between N–H groups and oxygen atoms of the
triflate counterion.[16] This observation might indicate a ge-
neral ability of our catalysts to make a two-point binding
via N–H groups as hydrogen-bond donors to the substrates
Entry
6
R
4
Yield
(%)[a]
ee (%)
possessing hydrogen-bond acceptors.
[b]
1
2
6a
6b
6c
6d
6e
6f
C6H5
4-ClC6H4
4-OMeC6H4
4a
4b
4c
94
41
82
57
76
60
82
88
84
88
84
88
89
88
80
88
Conclusions
3
4[c]
5
2,6-(OMe)2C6H3 4d
In conclusion, we have designed, prepared, and studied
a series of new organic catalysts inspired by the phenome-
non of specific molecular recognition of DNA bases. A se-
lected bifunctional catalyst based on a guanine structure
has been shown to catalyze the conjugate addition of 1,3-
dicarbonyl compounds to various nitroalkenes, providing
the products in good yields and enantioselectivities. Further
studies on the catalyst structure and its application to chal-
lenging asymmetric transformations as well as studies con-
cerning the mechanism of substrate activation are underway
in our laboratories.
2-BrC6H4
2-thienyl
2-naphthyl
pentyl
4e
4f
4g
4h
6
7
6g
6h
8[d]
[a] Isolated yield. [b] The enantiomeric excess was determined by
HPLC analysis of 4a using a chiral column. [c] The reaction was
carried out for 3 d. [d] The reaction was carried out for 4 d.
diethylbromomalonate, compound 7d (entry 4), offers an
additional facet in a possible further functionalization of
product 4k.[13e]
Supporting Information (see footnote on the first page of this arti-
cle): Experimental procedures, characterization of all new com-
pounds, X-ray structures of 5c, 8a, and 8b, copies of HPLC chro-
matograms and NMR spectra.
Table 4. Enantioselective Michael addition of various 1,3-dicarb-
onyl compounds 7a–d to nitrostyrene in the presence of 5a.
Acknowledgments
We are thankful to Prof. Amir H. Hoveyda for valuable discussions.
Financial support from the United States–Israel Binational Science
Foundation (BSF-grant no. 2008391) is acknowledged.
Entry
7
R
RЈ
4
Yield (%)[a] ee (%)[b]
1
2
3
4
7a
7b
7c
7d
OEt
OMe
Me
H
H
H
Br
4a
4i
94
90
98
62
84
86
83
85
[1] For some representative reviews, see: a) A. Dondoni, A. Massi,
Angew. Chem. 2008, 120, 4716; Angew. Chem. Int. Ed. 2008,
47, 4638–4660; b) S. Bertelsen, K. A. Jorgensen, Chem. Soc.
Rev. 2009, 38, 2178–2189; c) P. H.-Y. Cheong, C. Y. Legault,
J. M. Hm, N. Celebi-Olcum, K. N. Houk, Chem. Rev. 2011,
111, 5042–5139; d) A. Berkessel, H. Gröger in Asymmetric Or-
ganocatalysis: From Biomimetic Concepts to Applications in
Asymmetric Synthesis (Eds.: A. Berkessel, H. Gröger, David
MacMillan), Wiley-VCH, Weinheim, 2005; e) P. R. Schreiner,
Chem. Soc. Rev. 2003, 32, 289–296; f) C. Palomo, M. Oiarbide,
R. Lopez, Chem. Soc. Rev. 2009, 38, 632–653.
4j
OEt
4k
[a] Isolated yield. [b] The enantiomeric excess was determined by
HPLC analysis of 4a using a chiral column.
Interestingly, the related guanidinium species 8, prepared
by protonation of compound 3 with triflic or tetrafluoro-
boric acids (Figure 3), exhibited low catalytic activity and
[2] D. W. C. Macmillan, Nature 2008, 455, 304–308.
[3] a) A. G. Doyle, E. N. Jacobsen, Chem. Rev. 2007, 107, 5713–
5743; b) M. S. Taylor, E. N. Jacobsen, Angew. Chem. 2006, 118,
1550; Angew. Chem. Int. Ed. 2006, 45, 1520–1543.
[4] W. Saenger, Principles of Nucleic Acid Structure, Springer-
Velag, New York, 1984.
[5] a) C. Schmuck, L. Geiger, J. Am. Chem. Soc. 2005, 127, 10486–
10487; b) A. Ragusa, S. Rossi, J. Hayes, M. Stein, J. D. Kilburn,
Chem. Eur. J. 2005, 11, 5674–5688.
[6] For selected reviews, see: a) J. L. Sessler, J. Jayawickramarajah,
Chem. Commun. 2005, 1939–1949; b) L. J. Prins, D. N. Rein-
houdt, P. Timmerman, Angew. Chem. 2001, 113, 2446; Angew.
Chem. Int. Ed. 2001, 40, 2382–2426; For recent paper, see:
H. C. Ong, S. C. Zimmerman, Org. Lett. 2006, 8, 1589–1592.
[7] a) While preparing this paper, a similar study appeared: T. In-
okuma, M. Furukawa, T. Uno, Y. Suzuki, K. Yoshida, Y. Yano,
K. Matsuzaki, Y. Takemoto, Chem. Eur. J. 2011, 17, 10470; b)
Figure 3. (a) Structure of 8. (b) X-ray structure of 8b (ORTEP
drawing). Hydrogen bonds between N–H units and triflate groups
are indicated.
Eur. J. Org. Chem. 2012, 2118–2122
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2121

G. Suez, V. Bloch, G. Nisnevich, M. Gandelman
SHORT COMMUNICATION
E. J. Corey, M. J. Grogan, Org. Lett. 1999, 1, 157–160; c) For
a review, see: M. P. Coles, Chem. Commun. 2009, 3659–3676.
[8] For selected examples, see: a) C. Ghobril, P. Hammar, S. Kode-
pelly, B. Spiess, A. Wagner, F. Himo, R. Baati, ChemCatChem
2010, 2, 1573–1581; b) X. Chen, S. Dong, Z. Qiao, Y. Zhu, M.
Xie, L. Lin, X. Liu, X. Feng, Chem. Eur. J. 2011, 17, 2583–
2586; c) M. K. Kiesewetter, M. D. Scholten, N. Kirn, R. L. We-
ber, J. L. Hedrick, R. M. Waymouth, J. Org. Chem. 2009, 74,
9490–9496; d) J. Li, W.-Y. Jiang, K.-L. Han, G.-Z. He, C. Li,
J. Org. Chem. 2003, 68, 8786–8789.
hedron Lett. 2007, 48, 8456–8459; d) D. Almasi, D. A. Alonso,
E. Gomez-Bengoa, C. Najera, J. Org. Chem. 2009, 74, 6163–
6168; e) T. Okino, Y. Hoashi, T. Furukawa, X. Xu, Y. Takem-
oto, J. Am. Chem. Soc. 2005, 127, 119–125; f) J. M. Andres, R.
Manzano, R. Pedrosa, Chem. Eur. J. 2008, 14, 5116–5119; g)
H. Li, Y. Wang, L. Tang, L. Deng, J. Am. Chem. Soc. 2004,
126, 9906–9907; h) M. Terada, H. Ube, Y. Yagushi, J. Am.
Chem. Soc. 2006, 128, 1454–1455; i) H. Huang, E. N. Jacobsen,
J. Am. Chem. Soc. 2006, 128, 7170–7171.
For some reviews, see: a) G. Della Sala, A. Russo, A. Lattanzi,
Curr. Org. Chem. 2011, 15, 2147–2183; b) Y. Wang, L. Deng in
Catalytic Asymmetric Synthesis (Ed.: I. Ojima), John Wiley &
Sons, Hoboken, 2010, 3rd ed., pp. 59–94; c) J. Aleman, A.
Parra, H. Jiang, K. A. Jorgensen, Chem. Eur. J. 2011, 17, 6890–
6899; d) P. S. Bhadury, B.-A. Song, S. Yang, D.-Y. Hu, W. Xue,
Curr. Org. Synth. 2009, 6, 380–399; e) X. Liu, L. Lin, X. Feng,
Chem. Commun. 2009, 6145–6158; f) S. J. Connon, Chem.
Commun. 2008, 2499–2510.
[14]
[15]
[9] a) C. Uyeda, E. N. Jacobsen, J. Am. Chem. Soc. 2008, 130,
9228–9229 and references cited therein b) T. Kita, A. Georgi-
eva, Y. Hashimoto, T. Nakata, K. Ngasawa, Angew. Chem.
2002, 114, 2956; Angew. Chem. Int. Ed. 2002, 41, 2832–2834;
c) T. Schuster, M. Kurz, M. W. Gobel, J. Org. Chem. 2000, 65,
1697–1701; d) T. Schuster, M. Bauch, G. Durner, M. W. Gobel,
Org. Lett. 2000, 2, 179–181.
[10] The pKa for NH1 is about 9–10 and the pKa for NH2 is about
26–27: a) CRC Handbook of Chemistry and Physics, 90th ed.,
CRC Press, Boca Raton, N. J., 2009; b) Y. H. Jang, W. A. God-
dard III, K. T. Noyes, L. C. Sowers, S. Hwang, D. S. Chung, J.
Phys. Chem. B 2003, 107, 344–357.
Chiral N,N-diaminocyclohexane diamines, required for the
synthesis of 5a–h, were prepared in a facile and efficient way
by modification of the procedure published in: J. M. Mitchell,
N. S. Finney, Tetrahedron Lett. 2000, 41, 8431–8434.
[11] a) P. R. Schreiner, A. Wittkopp, Org. Lett. 2002, 4, 217–220; b)
A. Wittkopp, P. R. Schreiner, Chem. Eur. J. 2003, 9, 407–414.
[12] N. Ono, The Nitro Group in Organic Synthesis, Wiley-VCH,
New York, 2001.
[13] For representative examples, see: a) H. Li, S. Zhang, C. Yu, X.
Song, W. Wang, Chem. Commun. 2009, 2136–2138; b) S. H.
McCooey, S. J. Connon, Angew. Chem. 2005, 117, 6525; Angew.
Chem. Int. Ed. 2005, 44, 6367–6370; c) F.-X. Chen, C. Shao,
Q. Wang, P. Gong, D.-Y. Zhang, B.-Z. Zhang, R. Wang, Tetra-
[16] CCDC-855444 (for 8b), -855445 (for 8a), and -855446 (for 5c)
contain the supplementary 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. See ORTEP drawings in the Supporting Infor-
mation.
Received: February 1, 2012
Published Online: March 2, 2012
2122
www.eurjoc.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2012, 2118–2122