Bis(phosphoryl)-bridged biphenyls by radical phosphanylation: Synthesis and photophysical and electrochemical properties

Article abstract of DOI:10.1002/anie.201104114

The efficient fourfold radical phosphanylation of 2,2,2′,2′- tetrabromobiphenyl with a bis(stannyl)phosphane followed by oxidation produces a highly strained biphenyl with two phosphoryl bridges (see picture). Extended π-conjugated compounds consisting of this core unit can be also synthesized by using the same approach. The two phosphoryl bridges significantly alter the electronic structure, making this biphenyl a unique electron-accepting scaffold. Copyright

Full text of DOI:10.1002/anie.201104114

Communications
DOI: 10.1002/anie.201104114
p-Conjugated Compounds
Bis(phosphoryl)-Bridged Biphenyls by Radical Phosphanylation:
Synthesis and Photophysical and Electrochemical Properties**
Achim Bruch, Aiko Fukazawa, Eriko Yamaguchi, Shigehiro Yamaguchi,* and Armido Studer*
The design of new electron-accepting p-conjugated frame-
works is of particular significance for the development of
n-type semiconducting materials[1] and narrow-band gap
polymers,[2] which show great potential as components in
organic electronics, such as thin-film transistors and photo-
voltaic cells. Biphenyls containing electron-accepting moiet-
ies might be simple and viable scaffolds for the design of such
p-electron systems. However, by simply introducing electron-
withdrawing groups as substituents renders the biphenyl
framework to appear in a twist conformation, resulting in a
decrease of p-conjugation. This problem can be elegantly
Figure 1. Electron-accepting dibenzoheteroles 1 and doubly bridged
biphenyls 2 and 3.
solved by incorporating an electron-accepting unit as the
bridging moiety. In this regard, main group elements are
attractive as the bridging moieties, since they not only fix the
biaryl framework in a planar geometry, but also allow the
electronic structure to be modified through the choice of
element.[3] Accordingly, various electron-accepting dibenzo-
heteroles 1 featuring Si,[4] P, [5] and S[6] as bridging elements
have been reported and used in various applications. In
particular, P-containing p-electron systems have so far
attracted considerable attention, because of their rich
follow-up chemistry, which is attributed to easy transforma-
tions to oxides, sulfides, and metal/Lewis acid complexes.[7]
Among the various functionalities derived from phosphanes,
phosphine oxides or sulfides are of particular interest due to
their highly electron-accepting character.[8,9]
As a novel electron-accepting biphenyl, we designed
bis(phosphoryl)-bridged biphenyl (BPB) 2, which displayed
the following characteristics: a) compact and planar structure
enabling effective orbital overlap and b) high electron-
accepting ability owing to two phosphine oxide units. Related
biphenyls 3 containing Si,[10,11] S,[12] Se,[13] and C[14] as the
bridging moieties have already been reported (Figure 1).
Herein, we disclose the synthesis of this novel skeleton and
discuss its potential as the electron-accepting unit.
All our initial attempts to access BPB 2 by fourfold
lithiation of 2,2,2’,2’-tetrabromobiphenyl (4) followed by
trapping with dichlorophenylphosphane and P-oxidation
failed. The problem was associated with multiple lithiation
of 4.[15]
We have recently shown that radical phosphanylation of
reactive aryl radicals is a highly efficient approach for the
synthesis of arylphosphanes.[16] The reactions occur under
rather mild conditions, and expensive transition-metal cata-
lysts are not necessary. Stannylated and silylated phosphanes
have been used as reagents, and reactions occurred in high
yields. We envisioned that 2 should be accessible by multiple
radical phosphanylation of biphenyl 4 with bis(trimethylstan-
nyl)phenylphosphane and subsequent oxidation.
We first tried radical phosphanylation of 4 with readily
prepared (Me3Sn)2PPh using a,a’-azobisisobutyronitrile
(AIBN) as initiator at 808C in benzene. Disappointingly,
after 24 h little conversion of the starting material had
occurred and after H2O2 oxidation none of the targeted
BPB 2 was identified (Table 1, entry 1). The same result was
obtained by performing the reaction at 1258C (Table 1,
entry 2). To our delight, switching to 1,1’-azobis(cyclohex-
ane-1-carbonitrile) (V-40) as initiator afforded traces of the
desired BPB after 24 h (Table 1, entry 3). Prolonged reaction
time led to higher conversion, and after oxidation with H2O2
trans-2 and cis-2 were isolated in good yields (Table 1,
entry 4). As expected, oxidation did not occur diastereose-
lectively and both isomers were isolated in similar yields. By
running the reaction in benzotrifluoride the reaction time
could be shortened to two days without decreasing the yield
(Table 1, entry 5). The success of this transformation is an
impressive demonstration of the efficiency of the radical
phosphanylation: four highly reactive aryl radicals are
trapped sequentially and although severe ring strain is
generated in the formation of the second five-membered
[*] A. Bruch, Prof. Dr. A. Studer
Organisch-Chemisches Institut
Westfꢀlische Wilhelms-Universitꢀt
Corrensstrasse 40, 48149 Mꢁnster (Germany)
E-mail: studer@uni-muenster.de
Prof. Dr. A. Fukazawa, E. Yamaguchi, Prof. Dr. S. Yamaguchi
Department of Chemistry, Graduate School of Science
Nagoya University and JST-CREST
Furo, Chikusa, Nagoya 464-8602 (Japan)
E-mail: yamaguchi@chem.nagoya-u.ac.jp
[**] A.S. and A.B. thank the Deutsche Forschungsgemeinschaft (DFG)
for financial support.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201104114.
12094
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 12094 –12098

Table 1: Radical phosphanylation of 4 with (Me3Sn)2PPh.
We found that isomerically pure phosphane trans-5
precipitated in 40% yield by stirring the crude reaction
mixture at room temperature overnight (Scheme 1). This
Entry
Initiator
(equiv)
Solvent
T
[8C]
t
[h]
Yield
[%][a]
1
2
3
4
5
AIBN (0.2)
AIBN (0.2)
V-40 (0.2)
V-40 (0.5)
V-40 (0.5)
PhH
PhH
PhH
PhH
PhCF3
80
24
24
24
84
50
–
–
125
125
125
125
Scheme 1. Synthesis of trans-5.
trace
30/29
29/27
result shows that 5 can undergo trans/cis isomerization by
inversion at phosphorus at room temperature in solution. We
measured the isomerization barrier by dissolving trans-5 in
CDCl3 and monitoring the trans/cis isomerization by
31P NMR spectroscopy. The activation energy was determined
to be EA = (21.0 ꢁ 1.6) kcalmolꢀ1, which is considerably
smaller than those of dibenzophosphole derivatives (25–
27 kcalmolꢀ1).[20] In general, barriers for the pyramidal
inversion at the phosphorus atom in various phosphole
derivatives are suggested to be dependent on the extent of
aromaticity of the planar phosphole skeleton in the transition
state.[20a] To elucidate the relationship between the structure
and the aromaticity in the transition state, geometry optimi-
zations for 5 and dibenzophosphole 1b (R = Ph) were
performed at the B3LYP/6-31G(d) level of theory. The
activation barriers were estimated to be + 25.5 and
+ 27.5 kcalmolꢀ1 for 5 and 1b, respectively, which are semi-
quantitatively consistent with the experimental results. How-
ever, the nucleus-independent chemical shift (NICS) values in
the transition states at the HF/6-311 + G(d,p) level of theory
using the optimized geometries are comparable to each other
(NICS (0): ꢀ10.4 for 5, ꢀ9.8 for 1b; NICS(1): ꢀ8.0 for 5, ꢀ8.8
for 1b). These results indicate that not the aromatic character
of the phosphole ring in the P-planar transition state, but the
higher ring strain in 5 may be primarily responsible for the
decrease in the energy barrier for the inversion at the
phosphorus atom at least in this case.
[a] Yields of isolated trans/cis isomers of 2.
ring (see discussion below) the final intramolecular trapping
proceeds in acceptable yield.
Single crystals of both trans-2 and cis-2 were obtained
from CH2Cl2/n-hexane solution and analyzed by X-ray
diffraction (Figure 2).[17] As expected bis(phosphoryl)-
Figure 2. X-ray crystal structures of a) trans-2 and b) cis-2. Solvent
molecules are omitted for clarity for cis-2. For trans-2, two crystallo-
graphically independent molecules were found in the unit cell, one of
which is shown. Thermal ellipsoids are drawn at the 50% probability
level.
Next we focused on the structural modification of the
BPB core skeleton. Applying Ir-catalyzed C H borylation,
[21]
ꢀ
4 was converted to bis(pinacolatoboronate) 6, which was
subsequently coupled with different aryl iodides to give
diarylated biphenyls 7 (Scheme 2). Radical phosphanylation
of biphenyls 7 led to the desired BPBs 8 (trans-8a: 22%,
trans-8b: 13%).
bridged biphenyl skeletons in trans-2 and cis-2 adopt almost
planar geometries, indicating the effective extension of
p-conjugation over the framework. Notably, the endocyclic
ꢀ
P C bonds are 1.8369(16)–1.8431(17) ꢀ in trans-2 and
UV/Vis absorption and fluorescence spectra of the
bis(phosphoryl)-bridged biphenyls cis-2, trans-2, and the
p-extended derivatives trans-8a,b were evaluated. The photo-
physical data are summarized in Table 2, together with those
of mono-phosphoryl-bridged biphenyl (dibenzophosphole
oxide) 1c (R = Ph)[5e] and bis(silicon)-bridged biphenyl 3a[10]
for comparison. In the UV/Vis absorption spectra, the cis- and
trans-bis(phosphoryl)-bridged biphenyls 2 have an absorption
band with the maximum wavelength (labs) at 347 nm and a
relatively small molar absorption coefficient e = 337 and e =
367 for cis-2 and trans-2, respectively. The cis and trans
1.828(2)–1.839(2) ꢀ in cis-2. These are the longest among
the already reported triarylphosphine oxides without hydro-
gen-bonding interactions (1.801–1.833 ꢀ). Moreover, the
ꢀ
ꢀ
central C1 C1ꢁ bonds (trans-2: C1 C1* 1.442(3) and
1.443(3) ꢀ; cis-2: C1 C7 1.442(3) ꢀ) are significantly short-
ꢀ
[18]
ꢀ
ened in comparison to those in biphenyl (C1 C1’ 1.494 ꢀ)
and dibenzophosphole oxide (C1-C1’ 1.480 ꢀ).[19] These
results reflect the highly strained structure caused by the
annulation with two phosphoryl groups.
Angew. Chem. Int. Ed. 2011, 50, 12094 –12098
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12095

Communications
respectively, which are significantly red-shifted compared to
those of the BPBs 2, demonstrating the effective extension of
p conjugation. Whereas the biphenyls 2 only show a faint
fluorescence, the p-extended derivatives trans-8a and trans-
8b show more intense emissions with FF values of 0.20 and
0.34, respectively.
The notable feature of the bis(phosphoryl)-bridged
biphenyls is their electron-accepting character due to the
electron-withdrawing phosphoryl groups. Remarkably, the
lem of bis(diphenylaminophenyl)-substituted BPB trans-8b is
122 nm longer than that of bis(diphenylamino)quarterpheny-
lene (lem = 425 nm in CHCl3).[22] To compare the electron-
accepting properties, we also prepared mono(phosphoryl)-
bridged biphenyl 9 starting from 4,4’-dibromo-2,2’-diiodo-
1,1’-biphenyl (see the Supporting Information). The photo-
physical data of the two compounds in various solvents are
Scheme 2. Synthesis of 4,4’-diarylated BPBs. a) [{Ir(OMe)(cod)}2],
B2Pin2, TBME, 608C. b) ArI, [Pd(PPh3)4], K3PO4, dioxane, water, 1008C.
cod=cyclooctadiene, Pin=pinacol, TBME=tert-butylmethyl ether.
summarized in Table S7 in the Supporting Information. For
both compounds, whereas the absorption maxima show small
solvent dependence, the fluorescence spectra show substan-
tial bathochromic shifts as the solvent polarity increases.
These facts demonstrate that these molecules have more
polar structures in the excited state than in the ground state.
To compare the degree of the polarity of the structures in the
excited state, we employed the Lippert–Mataga equation
[Eq. (1)],[23]
Table 2: Photophysical data for the bridged biphenyls.[a]
UV/Vis Absorption
Fluorescence
labs [nm][b]
e [ꢂ103]
lem [nm][c]
FF
[d]
Cmpd
cis-2
trans-2
347
347
375 (sh)
399
332
0.337
0.367
3.50
37.6
0.794
12.1
43.5
387
387
422
547
366
361
499
0.03
0.03
0.20
0.34 (0.30)
0.042
trans-8a
trans-8b
1c[e]
3a[f]
9
285
396
0.08
0.89 (0.18)
2
[a] In CH2Cl2 unless stated otherwise. [b] Only the absorption maxima at
the longest wavelengths are shown. [c] Emission maxima upon excitation
at the absorption maximum wavelengths. [d] Absolute fluorescence
quantum yields determined by a calibrated integrating sphere system
within ꢁ3% errors. Fluorescence quantum yields in the solid state are
shown in parentheses. [e] Ref. [5e]. [f] Ref. [10]; in cyclohexane.
2 ðmE ꢀ mDÞ
ð1Þ
ð2Þ
Dn ¼
Df ¼
Df þ const:
hc
a3
e ꢀ 1
n2 ꢀ 1
ꢀ
2e þ 1 2n2 þ 1
isomers have essentially identical spectra irrespective of the
stereochemistry. In the fluorescence spectra, cis- and trans-2
exhibit a violet fluorescence with the maximum wavelength
(lem) at 387 nm, whereas the quantum yield is low (FF = 0.03).
Notably, both the absorption and fluorescence of 2 are
considerably longer than those of dibenzophosphole oxide 1c
(R = Ph) and bis(silicon)-bridged biphenyl 3a. These results
are attributable to the narrow HOMO–LUMO gap of 2 due
to their low-lying LUMO, supported by the quantum
chemical calculations using a time-dependent density func-
tional theory (TD-DFT) method (see the Supporting Infor-
mation).
in which mE and mG are the dipole moments in the excited state
and ground state, respectively, Dn is the Stokes shift, and Df is
the orientation polarizability. The solvent polarity parameter
is given by Equation (2), in which e is the dielectric constant
and n is the refractive index.[24] For the compounds trans-8b
and 9, linear relationships were observed for the plots of Dn as
a function of Df (Figure 3). The slopes obtained for trans-8b
and 9 are 16900 and 13400 cmꢀ1, respectively. These results
indicate that the extent of the electron-accepting character of
bis(phosphoryl)-bridged biphenyl is considerably larger than
that of the monophosphoryl-bridged analogue. In addition,
little deviation was observed in the Lippert–Mataga plots for
both 8a and 9 in iPrOH, indicating that the specific solvent
effect due to the hydrogen bonding is likely negligible in
iPrOH for both 8a and 9.
As for the extended analogues, phenyl- and diphenylami-
nophenyl-substituted derivatives trans-8a and trans-8b have
labs at 375 nm and 399 nm, and lem at 422 nm and 547 nm,
12096 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 12094 –12098

is one of the simplest skeletons and therefore should find
many applications as a building block in organic devices and
molecular electronics.
Received: June 15, 2011
Published online: October 25, 2011
Keywords: biaryls · fluorescence · phospholes · conjugation ·
.
radical arylation
[1] J. E. Anthony, A. Facchetti, M. Heeney, S. R. Marder, X. Zhan,
Adv. Mater. 2010, 22, 3876 – 3892, and references therein.
[2] A. Ajayaghosh, Chem. Soc. Rev. 2003, 32, 181 – 191, and
references therein.
~
*
Figure 3. The Lippert–Mataga plots for trans-8b ( ) and 9 ( ).
[3] A. Fukazawa, S. Yamaguchi, Chem. Asian J. 2009, 4, 1386 – 1400.
[4] K. L. Chan, M. J. McKiernan, C. R. Towns, A. B. Holmes, J. Am.
Chem. Soc. 2005, 127, 7662 – 7663.
Importantly, the Ph2NC6H4-substituted BPB trans-8b
retains its fluorescence quantum yield even in the solid state
(FF = 0.30). This is in contrast to the fact that the FF of
dibenzophosphole oxide analogue 9 in the solid state is much
smaller than that in CH2Cl2 solution (FF = 0.18 in solid, 0.89
in solution). This is likely due to the increased steric demand
of the two phosphoryl groups in trans-8b, which prohibit
intermolecular interactions in the solid state, although the
reason is still unclear at the moment.
[5] Recent examples: a) S. Zhang, R. Chen, J. Yin, F. Liu, H. Jiang,
N. Shi, Z. An, C. Ma, B. Liu, W. Huang, Org. Lett. 2010, 12,
3438 – 3441; b) N. Fukawa, T. Osaka, K. Noguchi, K. Tanaka,
Org. Lett. 2010, 12, 1324 – 1327; c) J. Yin, S.-L. Zhang, R.-F.
Chen, Q.-D. Ling, W. Huang, Phys. Chem. Chem. Phys. 2010, 12,
15448 – 15458; d) K. Geramita, J. McBee, T. D. Tilley, J. Org.
Chem. 2009, 74, 820 – 829; e) H.-C. Su, O. Fadhel, C.-J. Yang, T.-
Y. Cho, C. Fave, M. Hissler, C.-C. Wu, R. Rꢂau, J. Am. Chem.
Soc. 2006, 128, 983 – 995.
[6] Recent examples: a) I. I. Perepichka, I. F. Perepichika, M. R.
Bryce, L.-O. Pꢃlsson, Chem. Commun. 2005, 3397 – 3399; b) C.
Wang, H. Dong, H. Li, H. Zhao, Q. Meng, W. Hu, Cryst. Growth
Des. 2010, 10, 4155 – 4160.
We also evaluated the electrochemical properties of these
two p-electron systems. The cyclic voltammograms of trans-
8b and 9 are shown in Figure 4. Ph2NC6H4-substituted BPB
[7] Reviews on P-containing p-conjugated materials: a) T. Baum-
gartner, R. Rꢂau, Chem. Rev. 2006, 106, 4681 – 4727; Correction:
T. Baumgartner, R. Rꢂau, Chem. Rev. 2007, 107, 303; b) F.
Mathey, Angew. Chem. 2003, 115, 1616 – 1643; Angew. Chem.
Int. Ed. 2003, 42, 1578 – 1604.
[8] Recent examples: a) Y. Ren, T. Baumgartner, J. Am. Chem. Soc.
2011, 133, 1328 – 1340; b) Y. Liu, J. Feng, A. Ren, J. Comput.
Chem. 2007, 28, 2500 – 2509; c) T. Baumgartner, T. Neumann, B.
Wirges, Angew. Chem. 2004, 116, 6323 – 6328; Angew. Chem. Int.
Ed. 2004, 43, 6197 – 6201.
[9] a) A. Fukazawa, T.-a. Murai, L. Li, Y. Chen, S. Yamaguchi, C. R.
Chim. 2010, 13, 1082 – 1090; b) A. Fukazawa, Y. Ichihashi, Y.
Kosaka, S. Yamaguchi, Chem. Asian J. 2009, 4, 1729 – 1740; c) A.
Fukazawa, M. Hara, T. Okamoto, E.-C. Son, C. Xu, K. Tamao, S.
Yamaguchi, Org. Lett. 2008, 10, 913 – 916.
Figure 4. Cyclic voltammograms of trans-8b (c) and 9 (a) in
THF containing Bu4N+PF6 (0.1m) with a scan rate of 100 mVsꢀ1
.
ꢀ
[10] M. Shimizu, H. Tatsumi, K. Mochida, K. Oda, T. Hiyama, Chem.
Asian J. 2008, 3, 1238 – 1247.
[11] T. Kimura, Y. Izumi, N. Furukawa, Y. Minoshima, Heteroat.
Chem. 1996, 7, 143 – 147.
[12] T. Kimura, Y. Ishikawa, N. Furukawa, Chem. Lett. 1993, 635 –
636.
[13] T. Kimura, Y. Ishikawa, S. Ogawa, T. Nishio, I. Iida, N.
Furukawa, Tetrahedron Lett. 1992, 33, 6355 – 6358.
[14] B. M. Trost, P. L. Kinson, J. Am. Chem. Soc. 1970, 92, 2591 –
2593.
[15] Indeed, only dilithiation of 4 with excess butyllithium has been
reported, see: Q. Perron, J. Praz, A. Alexakis, Tetrahedron:
Asymmetry 2009, 20, 1004 – 1007.
[16] a) S. Vaillard, C. Mꢄck-Lichtenfeld, S. Grimme, A. Studer,
Angew. Chem. 2007, 119, 6653 – 6656; Angew. Chem. Int. Ed.
2007, 46, 6533 – 6536; b) A. Bruch, A. Ambrosius, R. Frohlich, A.
Studer, D. B. Guthrie, H. Zhang, D. P. Curran, J. Am. Chem. Soc.
2010, 132, 11452 – 11454; c) M.-C. Lamas, A. Studer, Org. Lett.
2011, 13, 2236 – 2239.
trans-8b and dibenzophosphole oxide 9 show the first
reversible reduction waves with the reduction potentials of
ꢀ2.15 and ꢀ2.36 V (vs. Fc/Fc+), respectively. These observa-
tions also demonstrate that the two phosphoryl bridges in
trans-8b are effective in enhancing the electron-accepting
character.
In summary, we demonstrated that the radical phospha-
nylation of tetrabromobiphenyls furnishes highly strained
bis(phosphoryl)-bridged biphenyls. The incorporation of two
phosphoryl groups as the bridging moieties is an effective way
to construct a rigid biphenyl skeleton with a highly electron-
accepting character. Recently, several fascinating phosphoryl-
bridged p-conjugated skeletons have been developed and
applied as useful electron-accepting scaffolds. Among these
examples, the bis(phosphoryl)biphenyl disclosed in this study
Angew. Chem. Int. Ed. 2011, 50, 12094 –12098
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12097

Communications
[17] Crystal data for trans-2 and cis-2: See the Supporting Informa-
Ishiyama, J. Takagi, J. F. Hartwig, N. Miyaura, Angew. Chem.
2002, 114, 3182; Angew. Chem. Int. Ed. 2002, 41, 3056.
[22] Z. H. Li, M. S. Wong, Y. Tao, M. DꢁIorio, J. Org. Chem. 2004, 69,
921 – 927.
[23] J. R. Lakowicz, Principles of Fluorescence Spectroscopy, 3rd ed.,
Springer, Berlin, 2006, pp. 205 – 213.
tion. CCDC-829426 (trans-2) and CCDC-829425 (cis-2) 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.
[18] G. B. Robertson, Nature 1961, 191, 593 – 594.
[24] Although the Lippert–Mataga equation originally assumes the
presence of a molecular dipole, it has been demonstrated to be
applicable to several quadrupole p-conjugated compounds; a) B.
Strehmel, A. M. Sarker, J. H. Malpert, V. Strehmel, H. Seifert,
D. C. Neckers, J. Am. Chem. Soc. 1999, 121, 1226 – 1236; b) F. C.
Krebs, H. Spanggaared, J. Org. Chem. 2002, 67, 7185 – 7192;
c) C.-H. Zhao, E. Sakuda, A. Wakamiya, S. Yamaguchi, Chem.
Eur. J. 2009, 15, 10603 – 10612.
[19] A. Decken, CCDC-254271 contains the supplementary crystallo-
graphic 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.
[20] a) W. Egan, R. Tang, G. Zon, K. Mislow, J. Am. Chem. Soc. 1971,
93, 6205 – 6216; b) M. Widhalm, L. Brecker, K. Mereiter,
Tetrahedron: Asymmetry 2006, 17, 1355 – 1369.
[21] a) T. Ishiyama, J. Takagi, K. Ishida, N. Miyaura, N. R. Anastasi,
J. F. Hartwig, J. Am. Chem. Soc. 2002, 124, 390 – 391; b) T.
12098 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 12094 –12098