Soluble conjugated microporous polymers

Article abstract of DOI:10.1002/anie.201205521

Soluble and porous: Soluble conjugated microporous polymers (SCMPs) can be prepared by synthesizing discrete hyperbranched molecules. The materials can be cast from solution as thin films (see picture), suggesting a range of processing options that are not available for insoluble CMP networks. Soluble, conjugated dendrimers are porous and give insight into the structural origin of microporosity. Copyright

Full text of DOI:10.1002/anie.201205521

Angewandte
Chemie
DOI: 10.1002/anie.201205521
Porous Polymers
Soluble Conjugated Microporous Polymers**
Ge Cheng, Tom Hasell, Abbie Trewin, Dave J. Adams, and Andrew I. Cooper*
In memoriam Jon Weaver
In the last five years, conjugated microporous polymers
(CMPs)[1] and other polymer networks formed by carbon–
carbon coupling chemistry[2] have emerged as an important
platform in amorphous porous materials. CMPs are the first
synthetic networks that combine permanent microporosity
(pores < 2 nm) with extended p-conjugation. Building on
reports of tunable pore sizes,[1a,c] structural modularity,[1e]
record surface areas (5000–6500 m2 gÀ1),[2b,c] and exceptional
physicochemical stability,[2b] new materials have been devel-
oped for applications such as catalysis,[1h,i,2e] light harvest-
ing,[1g] carbon dioxide capture,[1k,2d] superhydrophobic sepa-
rations,[1m] luminescence,[1l] sensors,[1n] and supercapacitors.[1j]
All of these materials are insoluble networks. This insolubility
limits the range of processing options for some of the more
interesting applications of CMPs that seek to exploit the
unique combination of porosity, conjugation, and synthetic
diversity. Indeed, processability and device integration were
highlighted as a challenge of “paramount importance” in
a recent review of this area.[10]
molecular weight of the material and to incorporate solubi-
lizing alkyl groups. To prepare the soluble CMPs, a two-step
(A4 + B2)-type Suzuki-catalyzed aryl–aryl coupling copoly-
merization was performed (Scheme 1). In the first step,
Unlike CMP networks, some other porous polymers are
solution processable. In particular, rigid and contorted
“polymers of intrinsic microporosity” (PIMs)[3] can be
dissolved in organic solvents and fabricated, for example,
into microporous membranes.[3,4] To date, however, soluble
linear PIMs have been prepared by condensation chemistry
that introduces heteroatoms into the polymer chain and that
does not introduce extended p-conjugation. Processing of
microporous polymers has been described as an important
target.[5] However, the closest reported step towards soluble
CMPs have been solution-dispersible CMP nanoparticles
formed by emulsion techniques.[6]
Here, we report the first example of a soluble conjugated
microporous polymer (SCMP). The synthesis is based on
hyperbranching, as used previously, for example, to prepare
soluble hyperbranched polyphenylenes.[7] Initially, we focused
on 1,3,6,8-tetrabromopyrene as an A4 monomer, building on
our studies of insoluble pyrene CMP networks.[8] Here, a tert-
butyl-functionalized B2 monomer is introduced to limit the
Scheme 1. Two-step, one-pot synthesis of a soluble conjugated micro-
porous polymer, SCMP1. The resulting material is a statistical hyper-
branched copolymer that is soluble in common organic solvents. A
solution of SCMP1 in THF shows green luminescence under UV
irradiation (l=254 nm; image below the scheme). Reagents and
conditions: a) bis(pinacolato)diboron, Pd(OAc)2, KOAc, anhydrous
DMF, 908C and b) Pd(PPh3)4, K2CO3, anhydrous DMF, 1108C.
palladium acetate (Pd(OAc)2)[9] catalyzes an aryl halide/
diboron coupling to generate arylboronates of both the A4
monomer, 1,3,6,8-tetrabromopyrene, and the B2 monomer,
1,3-dibromo-7-tert-butylpyrene,[10] in a one-pot “prepolyme-
rization” reaction. Without isolating the arylboronate species,
statistical copolymerization of the two monomers was then
carried out in a second step by addition of Pd(PPh3)4 and
K2CO3. After purification by antisolvent reprecipitation, the
polymer was isolated as a deep yellow film. These materials
were dissolved in common organic solvents such as tetrahy-
drofuran (THF), CH2Cl2, and toluene to give homogeneous
green-luminescent solutions (Scheme 1 and Table S1 in the
Supporting Information).
[*] Dr. G. Cheng, Dr. T. Hasell, Dr. A. Trewin, Dr. D. J. Adams,
Prof. A. I. Cooper
Department of Chemistry and Centre for Materials Discovery
University of Liverpool, Crown Street, Liverpool L69 3BX (UK)
E-mail: aicooper@liv.ac.uk
SCMP1 is porous and the porosity depends on the method
by which the material is isolated from solution. In particular,
the porosity was different for materials precipitated rapidly in
antisolvents in comparison with films prepared by slow
solvent evaporation. A wide range of conditions and solvents
were investigated (see the Supporting Information), but here
Homepage: http://www.liv.ac.uk/chemistry/res/coopergroup
[**] The authors gratefully acknowledge the EPSRC for funding (EP/
H000925/1). A.I.C. is a Royal Society Wolfson award holder. A.T. is
a Royal Society URF.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201205521.
Angew. Chem. Int. Ed. 2012, 51, 12727 –12731
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12727

.
Angewandte
Communications
we discuss two examples: antisolvent precipitation in a poor
solvent (petroleum ether) and solution casting from a good
solvent (dichloromethane, DCM). In both cases, SCMP1 was
dissolved initially in DCM. For antisolvent precipitation, this
DCM solution was added dropwise into excess petroleum
ether. The rapidly precipitated SCMP1 powder was then
removed by centrifugation (Figure 1a). For solvent casting,
By contrast, the solvent-cast SCMP1 film is effectively
nonporous to nitrogen at 77 K (BET surface = 12 m2 gÀ1).
Both materials, however, have a similar H2 uptakes (about
4 mmolgÀ1 at 1 bar, 77 K), although greater desorption
hysteresis is observed for the solvent-cast film. The difference
in gas selectivity for the two samples may arise from the
packing of the polymer molecules in the solid state, with the
rapidly precipitated SCMP1 sample vitrifying into a less
densely packed molecular structure. The solvent-evaporated
SCMP1 sample forms a film that is selectively porous to
hydrogen, suggesting potential in applications as coatings for
gas separations. The solvent-cast SCMP1 film also adsorbs
significant quantities of other gases such as CO2, methane,
and xenon at 273 K (Figure 2).
Figure 2. Nitrogen, methane, xenon, and carbon dioxide isotherms,
recorded at 273 K, for DCM-cast SCMP1 films. Adsorption/desorption
curves are shown as closed/open symbols, respectively.
Figure 1. a) Photograph of antisolvent-precipitated SCMP1 powder.
b) Photograph of a SCMP1 film prepared by slow evaporation. c) SEM
image showing fused nanospheres in the precipitated polymer. d) SEM
image showing the smooth surface of a cast SCMP1 film. e) Gas
sorption isotherms for the precipitated powder, measured at 77 K, for
nitrogen (blue) and hydrogen (red); desorption curves shown as open
symbols. f) Equivalent gas sorption isotherms for the solvent-evapo-
rated film. Note different vertical scales in (e) and (f).
The weight-averaged molecular weight of SCMP1, as
measured by gel permeation chromatography (GPC) was
5.316 gmolÀ1. However, accurate molecular weight determi-
nation for highly branched polymers is challenging using
GPC, which uses linear polymers as calibration standards. To
address this, two pyrene dendrimers[10] were synthesized as
control molecules with defined mass and structure. This also
allowed comparison of the sorption properties of SCMP1 with
those of analogous branched molecules with precisely con-
trolled composition and mass. Two hybrid polyhedral oligo-
meric silsesquioxane (POSS)–polypyrene dendrimers were
synthesized with peripheral dendrons that reflect the struc-
ture of the hyperbranched copolymer, SCMP1 (Figure 3).
These dendrimers were synthesized by a convergent cross-
metathesis pathway using Grubbꢀs catalyst[11] (see details in
the Supporting Information).
the DCM was simply allowed to evaporate slowly on a glass
slide, leaving the solid SCMP1 as a transparent, yellow film
(Figure 1b). Scanning electron microscope (SEM) images
reveal the rapidly precipitated powder is comprised of fused
spheres of around 100 nm in diameter (Figure 1c), while the
DCM-cast film has a smooth and uniform surface (Figure 1d).
The film is uniform and coherent but does not, in this first
example, have sufficient mechanical strength to be self-
supporting upon removal.
The nitrogen and hydrogen sorption isotherms for SCMP1
are shown in Figure 1e and 1 f. The rapidly precipitated
SCMP1 shows a type II nitrogen isotherm and a clear micro-
pore step at low relative pressures. The Brunauer–Emmett–
Teller surface area (SABET) is 505 m2 gÀ1; that is, at the low end
of the range for our first generation of insoluble CMP
networks.[1a] The upturn in the nitrogen isotherm at higher
relative pressures indicates meso/macroporosity, presumably
from the nanoscopic particles (Figure 1c) and associated
interparticle voids.
POSS-dend-1 (Figure 3a) has a calculated molecular
weight, confirmed by mass spectrometry, of 5952 gmolÀ1
.
POSS-dend-2 has additional bulky groups in the dendrons
(shown in pink Figure 3b) and a higher mass of 10.053 gmolÀ1
relative to POSS-dend-2. The dendrimers were dissolved in
DCM and precipitated into petroleum ether in identical
manner to the SCMP1 antisolvent process. POSS-dend-
1 shows low nitrogen porosity in comparison to SCMP1,
with an apparent BET surface area of 28 m2 gÀ1 (Figure 3c).
POSS-dend-1 does adsorb H2 at 77 K, but the isotherm shows
12728 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 12727 –12731

Angewandte
Chemie
surface area of 498 m2 gÀ1 and no hysteresis in the hydrogen
isotherm (Figure 3d). We suggest that POSS-dend-2 is less
interpenetrated in the solid state than POSS-dend-1 as
a result of the additional bulky pyrene groups (Figure 3b),
leading to a significant enhancement in microporosity. A
structural model for POSS-dend-2 was constructed (Fig-
ure 3e). A Connolly surface for the dendrimer (Figure 3 f)
highlights its irregular shape and the existence of cavities
extending deep within the dendrimer. It is likely that these
cavities contribute to the permanent porosity of the rigid
dendrimer in the solid state. An analogous type of micro-
porosity can be envisaged in SCMP1, but the latter material is
more difficult to simulate because it is not possible to define
a single molecular building block.
The GPC elution curves for SCMP1 and the two pyrene
dendrimers are shown in Figure 4 and the data are summar-
ized in Table 1. The molecular weight distribution for SCMP1
is, unsurprisingly, broader than the dendrimers which are
single-molecule species. GPC underestimates the true molec-
ular weights of the dendrimers. Overall, the GPC data suggest
that the molecular weight for SCMP1 falls in the same range
as the two dendrimers. The porous properties of antisolvent-
Figure 4. GPC chromatograms of octavinylsilsesquioxane (OVS),
POSS-dend-1, POSS-dend-2, and SCMP1.
Table 1: Molecular weight and sorption properties for SCMP1 and pyrene
dendrimers.[a]
[d]
Sample
Mw [gmolÀ1
]
Mn [gmolÀ1
]
PDI BET[b] [m2 gÀ1
]
N2,[c] H2
[mmolgÀ1
]
Figure 3. a) Structure of POSS-dend-1. b) Structure of POSS-dend-2,
additional bulky groups shown in pink. c) Nitrogen and hydrogen
isotherms, measured at 77 K, for POSS-dend-1. d) Comparable nitro-
gen and hydrogen isotherms for POSS-dend-2. Note different vertical
scales in (c) and (d). e) Molecular model for POSS-dend-2. f) Model
with Connolly surface shown in blue, probe radius=1.82 ꢀ.
SCMP1
OVS
POSS-
dend-1
POSS-
dend-2
5316
560
4709
4340
551
4651
1.22 505
5.6, 3.8
–
0.3, 1.3
1.01
–
1.01 28
6115
6040
1.01 498
5.5, 3.5
[a] Sorption data given for the antisolvent precipitated form. [b] Apparent
BETsurface area calculated over the range P/P0 =0.01–0.1. [c] N2 uptake
at P/P0 =0.1, 77 K. [d] H2 uptake at 1 bar, 77 K.
hysteresis. POSS-dend-2, however, is much more porous with
nitrogen sorption similar to SCMP1, with an apparent BET
Angew. Chem. Int. Ed. 2012, 51, 12727 –12731
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12729

.
Angewandte
Communications
polymer product was isolated by centrifugation and dried in vacuum
at 1208C to give 3.2 g of a light yellow powder.
precipitated POSS-dend-2 and SCMP1 are also very similar,
and their N2 sorption isotherms overlay almost exactly
(Figure S4 in the Supporting Information).
Step 2: This light yellow powder was dissolved in DCM (20 mL)
and absorbed on 10 g silica gel and air dried, followed by Soxhlet
extraction with hot hexane, a poor solvent for the SCMP, for
three days. The hexane solution was replaced with THF, a good
solvent for the SCMP, to extract the polymer from the silica gel over
two days. THF was removed by rotary evaporation to give 2.6 g of the
product, SCMP1, as a deep yellow film (yield = 81% by weight). GPC
analysis: Mw = 5316 gmolÀ1, Mn = 4340 gmolÀ1, PDI = 1.22 (Mw =
weight-averaged molecular weight, Mn = number-averaged molecular
weight, and PDI = polydispersity index). 1H NMR (400 MHz,
CDCl3): d = 9.1–7.3 (br, -pyrenyl) and 1.9–0.3 ppm (br, -CH3).
Assuming no end groups, a ratio of aromatic/tert-butyl groups of
1.11:1 would be expected. For SCMP1, an integration of 0.625:1 is
found. After hydrolysis of boronic ester end groups using BBr3, a ratio
of 1.08:1 was measured, close to the theoretical value. Hence, the feed
ratio is maintained, but the polymer also contains a significant
number of end groups, as expected from the relatively low molecular
weight.
Typical antisolvent reprecipitation conditions: SCMP1 was dis-
solved in CH2Cl2 (1 mL) at 80 mgmLÀ1 concentration and added
dropwise to petroleum ether (10 mL, b.p. 40–608C). The resulting
precipitated material was separated by centrifugation for 5 minutes at
5000 revolutions per minute (r.p.m.) before decanting the super-
natant.
Film casting: SCMP1 was dissolved in CH2Cl2 (1 mL) at
80 mgmLÀ1 concentration. The CH2Cl2 was subsequently allowed to
evaporate under nitrogen flow, leaving the polymer as a coherent film
on the glass surface of the containment vessel.
Both POSS-dend-1 and POSS-dend-2 display strong blue
luminescence when their solutions are irradiated by UV light
(Figure S5 in the Supporting Information).[12] Solutions of
SCMP1 are also photoluminescent because of the conjugated
structure of the polymer (Scheme 1). Absorption and emis-
sion spectra are included in Figures S6–S8 in the Supporting
Information. It was shown previously, for pyrene-based
dendrimers and polymers, that an increase in extended
conjugation causes a red-shift in fluorescence.[8,10]
Here, fluorescence in SCMP1 is more red-shifted because
the POSS core breaks the conjugation in the dendrimers. The
larger red-shift in fluorescence for POSS-dend-1 with respect
to POSS-dend-2 is not at present understood, but could stem
from reduction in conjugation arising from steric constraints
in the larger dendrimer.
To conclude, we have demonstrated for the first time that
soluble conjugated microporous polymers, SCMPs, can be
prepared by adapting the synthesis conditions to form discrete
hyperbranched chains rather than extended networks. These
materials can be processed from solution to form films, and
the resultant porosity is a function of the processing
conditions. Soluble conjugated dendrimers can also exhibit
microporosity, and we suggest that the structural origin of
microporosity—rigidity combined with non-interpenetrating
cavities—is probably similar in both cases. From a practical
viewpoint, however, SCMPs are preferable to dendrimers
because they can be prepared in a simple two-step, one-pot
procedure. Our first SCMPs involve pyrene monomers and
Suzuki cross-coupling chemistry, but it is likely that analogous
hyperbranching strategies can be developed for the wide
range of insoluble CMP networks,[1] as well-related, non-
conjugated porous networks,[2] that have been reported
recently.
Dendrimers: The syntheses and purification processes of POSS-
dend-1 and POSS-dend-2 are detailed in the Supporting Information.
Received: July 13, 2012
Published online: November 9, 2012
Keywords: absorption · dendrimers · membranes · polymers ·
.
selectivity
[1] a) J. X. Jiang, F. Su, A. Trewin, C. D. Wood, N. L. Campbell, H.
Niu, C. Dickinson, A. Y. Ganin, M. J. Rosseinsky, Y. Z. Khi-
myak, A. I. Cooper, Angew. Chem. 2007, 119, 8728 – 8732;
Angew. Chem. Int. Ed. 2007, 46, 8574 – 8578; b) J. Weber, A.
Thomas, J. Am. Chem. Soc. 2008, 130, 6334 – 6335; c) J. X. Jiang,
F. Su, A. Trewin, C. D. Wood, H. Niu, J. T. A. Jones, Y. Z.
Khimyak, A. I. Cooper, J. Am. Chem. Soc. 2008, 130, 7710 –
7720; d) A. I. Cooper, Adv. Mater. 2009, 21, 1291 – 1295; e) R.
Dawson, A. Laybourn, R. Clowes, Y. Z. Khimyak, D. J. Adams,
A. I. Cooper, Macromolecules 2009, 42, 8809 – 8816; f) J.
Schmidt, M. Werner, A. Thomas, Macromolecules 2009, 42,
4426 – 4429; g) L. Chen, Y. Honsho, S. Seki, D. L. Jiang, J. Am.
Chem. Soc. 2010, 132, 6742 – 6748; h) L. Chen, Y. Yang, D. L.
Jiang, J. Am. Chem. Soc. 2010, 132, 9138 – 9143; i) J. X. Jiang, C.
Wang, A. Laybourn, T. Hasell, R. Clowes, Y. Z. Khimyak, J. L.
Xiao, S. J. Higgins, D. J. Adams, A. I. Cooper, Angew. Chem.
2011, 123, 1104 – 1107; Angew. Chem. Int. Ed. 2011, 50, 1072 –
1075; j) Y. Kou, Y. H. Xu, Z. Q. Guo, D. L. Jiang, Angew. Chem.
2011, 123, 8912 – 8916; Angew. Chem. Int. Ed. 2011, 50, 8753 –
8757; k) R. Dawson, D. J. Adams, A. I. Cooper, Chem. Sci. 2011,
2, 1173 – 1177; l) Y. H. Xu, L. Chen, Z. Q. Guo, A. Nagai, D. L.
Jiang, J. Am. Chem. Soc. 2011, 133, 17622 – 17625; m) A. Li,
H. X. Sun, D. Z. Tan, W. J. Fan, S. H. Wen, X. J. Qing, G. X. Li,
S. Y. Li, W. Q. Deng, Energy Environ. Sci. 2011, 4, 2062 – 2065;
n) X. Liu, Y. Xu, D. Jiang, J. Am. Chem. Soc. 2012, 134, 8738 –
8741; o) F. Vilela, K. Zhang, M. Antonietti, Energy Environ. Sci.
2012, 5, 7819 – 7832.
Experimental Section
Synthesis of SCMP1: Step 1 (pre-polymerization): To an oven-dried
500 mL round-bottom flask equipped with a reflux condenser were
added 1,3,6,8-tetrabromopyrene (A4, 2.58 g, 5.0 mmol), 1,3-dibromo-
7-tert-butylprene[10] (B2, 4.16 g, 10.0 mmol), bis(pinacolato)diboron
(C, 8.00 g, 31.5 mmol), palladium acetate, Pd(OAc)2 (240 mg,
1.07 mmol), potassium acetate, KOAc (5.80 g, 59.10 mmol), and
anhydrous dimethylformamide, DMF (275 mL) under a nitrogen
atmosphere. After the mixture was degassed, it was heated and stirred
at 908C for 22 h.
Step 2 (polymerization): The pre-polymerized mixture was
cooled down to room temperature and Pd(PPh3)4 (680 mg,
0.59 mmol), K2CO3 (4.80 g, 34.73 mmol), and H2O (25 mL) were
added and the solution was degassed. The mixture was then heated to
1208C and stirred for five days under a nitrogen atmosphere.
Purification of SCMP1: Step 1: The resulting deep green mixture
was diluted with DCM (500 mL), washed with a 20% HCl solution
followed by brine until the green organic layer changed to brown; it
was then washed with water and dried over MgSO4. The clear solution
was concentrated at reduced pressure and any Pd-black particles were
removed by passing through a short silica gel column, followed by
elution with THF. The organic solution was then concentrated and
precipitated twice from DCM (40 mL) into MeOH (320 mL). The
12730 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 12727 –12731

Angewandte
Chemie
[2] a) M. Rose, W. Bohlmann, M. Sabo, S. Kaskel, Chem. Commun.
2008, 2462 – 2464; b) T. Ben, H. Ren, S. Q. Ma, D. P. Cao, J. H.
Lan, X. F. Jing, W. C. Wang, J. Xu, F. Deng, J. M. Simmons, S. L.
Qiu, G. S. Zhu, Angew. Chem. 2009, 121, 9621 – 9624; Angew.
Chem. Int. Ed. 2009, 48, 9457 – 9460; c) D. Yuan, W. Lu, D. Zhao,
H.-C. Zhou, Adv. Mater. 2011, 23, 3723 – 3725; d) W. Lu, J. P.
Sculley, D. Yuan, R. Krishna, Z. Wei, H.-C. Zhou, Angew. Chem.
2012, DOI: 10.1002/ange.201202176; Angew. Chem. Int. Ed.
2012, DOI: 10.1002/anie.201202176; e) Z. G. Xie, C. Wang, K. E.
deKrafft, W. B. Lin, J. Am. Chem. Soc. 2011, 133, 2056 – 2059.
[3] P. M. Budd, B. S. Ghanem, S. Makhseed, N. B. McKeown, K. J.
Msayib, C. E. Tattershall, Chem. Commun. 2004, 230 – 231.
[4] a) P. M. Budd, E. S. Elabas, B. S. Ghanem, S. Makhseed, N. B.
McKeown, K. J. Msayib, C. E. Tattershall, D. Wang, Adv. Mater.
2004, 16, 456 – 459; b) N. B. McKeown, P. M. Budd, Macro-
molecules 2010, 43, 5163 – 5176.
[6] a) A. Patra, J.-M. Koenen, U. Scherf, Chem. Commun. 2011, 47,
9612 – 9614; b) P. Zhang, Z. Weng, J. Guo, C. Wang, Chem.
Mater. 2011, 23, 5243 – 5249.
[7] a) Y. H. Kim, O. W. Webster, J. Am. Chem. Soc. 1990, 112, 4592 –
4593; b) Y. H. Kim, O. W. Webster, Macromolecules 1992, 25,
5561 – 5572.
[8] J. X. Jiang, A. Trewin, D. J. Adams, A. I. Cooper, Chem. Sci.
2011, 2, 1777 – 1781.
[9] L. Zhu, J. Duquette, M. B. Zhang, J. Org. Chem. 2003, 68, 3729 –
3732.
[10] T. M. Figueira-Duarte, S. C. Simon, M. Wagner, S. I. Drtezhinin,
K. A. Zachariasse, K. Muellen, Angew. Chem. 2008, 120, 10329 –
10332; Angew. Chem. Int. Ed. 2008, 47, 10175 – 10178.
[11] a) Y. Itami, B. Marciniec, M. Kubicki, Chem. Eur. J. 2004, 10,
1239 – 1248; b) G. Cheng, N. R. Vautravers, R. E. Morris, D. J.
Cole-Hamilton, Org. Biomol. Chem. 2008, 6, 4662 – 4667.
[12] P. Andrꢁ, G. Cheng, A. Ruseckas, T. van Mourik, H. Frꢂchtl,
J. A. Crayston, R. E. Morris, D. Cole-Hamilton, I. D. W. Samuel,
J. Phys. Chem. B 2008, 112, 16382 – 16392.
[5] A. Thomas, Angew. Chem. 2010, 122, 8506 – 8523; Angew. Chem.
Int. Ed. 2010, 49, 8328 – 8344.
Angew. Chem. Int. Ed. 2012, 51, 12727 –12731
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12731