Regarding Initial Ring Opening of Propylene Oxide in its Copolymerization with CO2 Catalyzed by a Cobalt(III) Porphyrin Complex

Article abstract of DOI:10.1002/chem.201404147

Cobalt(III) tetraphenylporphyrin chloride (TPPCoCl) was experimentally proved to be an active catalyst for poly(propylene carbonate) production. It was chosen as a model catalyst in the present work to investigate the initiation step of propylene oxide (PO)/CO₂ copolymerization, which is supposed to be the ring opening of the epoxide. Ring-opening intermediates (1-7) were detected by using ¹HNMR spectroscopy. A first-order reaction in TPPCoCl was determined. A combination of monometallic and bimetallic ring-opening pathways is proposed according to kinetics experiments. Addition of onium salts (e.g., bis(triphenylphosphine)iminium chloride, PPNCl) efficiently promoted the PO ring-opening rate. The existence of axial ligand exchange in the cobalt porphyrin complex in the presence of onium salts was suggested by analyzing collected ¹HNMR spectra. Monometallic or bimetallic? Ring-opening intermediates of propylene oxide (PO) catalyzed by a cobalt(III) porphyrin complex were identified by using ¹HNMR spectroscopy. The generated metal alkoxides underwent hydrolysis in the presence of trace amounts of water. A combined bimetallic and monometallic ring-opening pathway was suggested from the kinetics results. Addition of onium salts accelerated the PO ring opening. Ligand exchange on the cobalt center in the presence of onium salts was also detected.

Full text of DOI:10.1002/chem.201404147

DOI: 10.1002/chem.201404147
Full Paper
&
Ring-Opening Polymerization
Regarding Initial Ring Opening of Propylene Oxide in its
Copolymerization with CO2 Catalyzed by a Cobalt(III) Porphyrin
Complex
Wei Xia, Sergei I. Vagin, and Bernhard Rieger*[a]
Abstract: Cobalt(III) tetraphenylporphyrin chloride (TPPCoCl)
was experimentally proved to be an active catalyst for poly-
(propylene carbonate) production. It was chosen as a model
catalyst in the present work to investigate the initiation step
of propylene oxide (PO)/CO2 copolymerization, which is sup-
posed to be the ring opening of the epoxide. Ring-opening
intermediates (1–7) were detected by using 1H NMR spec-
troscopy. A first-order reaction in TPPCoCl was determined.
A combination of monometallic and bimetallic ring-opening
pathways is proposed according to kinetics experiments. Ad-
dition of onium salts (e.g., bis(triphenylphosphine)iminium
chloride, PPNCl) efficiently promoted the PO ring-opening
rate. The existence of axial ligand exchange in the cobalt
porphyrin complex in the presence of onium salts was sug-
1
gested by analyzing collected H NMR spectra.
Introduction
Epoxides are valuable intermediate compounds for organic
syntheses as they can be easily transformed by a wide range
of nucleophiles and electrophiles.[1] For instance, copolymeriza-
tion of epoxides with CO2 has not only been considered as an
excellent way to utilize this greenhouse gas, but also affords
biodegradable polymers.[2]
Scheme 1. Copolymerization of propylene oxide and CO2.
in previous publications, infers either that coordination to the
metal alters its Lewis acidity[9] or that it participates as an addi-
tional initiator, which starts an extra polymer chain.[10]
This copolymerization reaction was first discovered by Inoue
with a Zn catalyst prepared in situ from ZnEt2 and water.[3]
Since then, numerous heterogeneous and homogeneous cata-
lyst systems have been developed to improve the catalytic ac-
tivity and selectivity for the copolymer over byproducts (poly-
(propylene oxide) and cyclic propylene carbonate) and to fur-
ther investigate the polymerization mechanism.[2] As demon-
strated by substantial insightful previous mechanistic studies,[4]
the initiation step is generally considered to be the ring open-
ing of the epoxide by an initiator.
To further explore the mechanism of PO/CO2 copolymeriza-
tion, in the present work TPPCoCl is chosen as a model catalyst
to investigate the initial PO ring-opening step. A series of
onium salts is introduced into the ring-opening reaction
system to examine their influences on the ring-opening rate.
The obtained results are discussed in detail.
Although cobalt porphyrin was experimentally proved to Results and Discussion
have low affinity towards PO binding in the gas phase,[5] poly-
Synthesis and characterization of TPPCoCl
(propylene carbonate) (PPC) was eventually afforded through
a cobalt(III) tetraphenylporphyrin chloride (TPPCoIIICl)-catalyzed
PO/CO2 copolymerization reaction with either a Schiff base,
DMAP (4-dimethylaminopyridine),[6] or onium salts such as
PPNCl[7] as a cocatalyst (Scheme 1). The addition of a cocatalyst
into the polymerization system efficiently accelerates the copo-
lymer formation.[4d,8] A possible role of the cocatalyst, proposed
The method for the metalation of the tetraphenylporphyrin
ligand with cobalt acetate tetrahydrate was adopted from
Adler’s early work,[11] wherein DMF was chosen as a suitable
solvent. The air oxidation of TPPCoII to TPPCoIIICl was per-
formed in methanol with HCl (aq.) as a counter ion supplier.
Therefore, saturated NaHCO3 aqueous solution was subse-
quently utilized to neutralize the thus-obtained cobalt(III) por-
phyrin complex.
[a] W. Xia, Dr. S. I. Vagin, Prof. B. Rieger
WACKER-Lehrstuhl fꢀr Makromolekulare Chemie
Technische Universitꢁt Mꢀnchen
The UV/Vis spectrum of TPPCoCl in CH2Cl2 showed similar
characteristics compared to those reported in the literature
(Soret band: 406 nm, Q-band: 542.9 nm (br)).[12] Upon addition
of a coordinating solvent such as THF, the porphyrin Soret
band exhibited a 20 nm redshift and its intensity increased,
Lichtenbergstraße 4, 85747, Garching (Germany)
E-mail: rieger@tum.de
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201404147.
Chem. Eur. J. 2014, 20, 15499 – 15504
15499
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
and the porphyrin Q-band became narrower, which was ac-
companied by the appearance of a shoulder peak at 590 nm
(Figure S1 in the Supporting Information). A transformation
from five-coordinate TPPCoCl (in CH2Cl2) to a six-coordinate
species (in CH2Cl2 +THF) was invoked to explain the observed
change in the spectrum. A change in the UV/Vis spectrum was
also found for TPPZnII in non-coordinating and coordinating
solvents and was discussed in detail by Nappa and Valen-
tine.[13]
1
A change in the H NMR spectrum of TPPCoCl in non-coordi-
nating and coordinating solvents was also observed. Broad
proton signals of the cobalt porphyrin complex were found in
the NMR spectrum if CD2Cl2 was used as the NMR solvent (Fig-
ure S2 in the Supporting Information). Through addition of
[D8]THF, these proton resonances became sharp (Figure S3 in
the Supporting Information). A spin crossover between dia-
magnetic low-spin cobalt(III) and paramagnetic high-spin
cobalt(III) was assumed to interpret the phenomena in CD2Cl2.
In the presence of a coordinating solvent, a hexacoordinate
species was formed, stabilizing the cobalt center in a low spin
state. Huet attributed the broad proton resonances of TPPCoCl
in non-coordinating solvent to the paramagnetism of TPPCoII
and the TPPCoIII p-cation radical afforded from a disproportio-
nation reaction of TPPCoIIICl.[14] Unexpected hyperstoichiomet-
ric chlorine atom (ca. 1.3) bound to the metal was found for
the crystal of TPPCoCl grown from CH2Cl2/hexane (Figure S5 in
the Supporting Information), an observation that may be evi-
dential for such disproportionation.
Figure 1. 1H NMR spectrum collected after reaction time of 30 min (CD2Cl2,
300 MHz, 258C, *=polydimethylsiloxane residual, only the signals of methyl
groups are denoted).
Ring opening of PO in the absence of onium salts
TPPCoCl was first mixed with PO and 1H NMR spectroscopy
was utilized to probe if the ring opening of PO by TPPCoCl
1
could take place. A series of time-resolved H NMR measure-
ments was performed with TPPCoCl (5 mg) and PO (0.3 mL) in
CD2Cl2 (0.5 mL) at room temperature and NMR signals with
higher field chemical shifts compared with the signal of tetra-
methylsilane (TMS) were found in the collected spectra
(Figure 1), signals that can only be ascribed to the axially coor-
dinated species at the cobalt center owing to the shielding
effect of the p-electron ring current of the porphyrin macrocy-
cle.[15] After careful assignment of these peaks with further
help of COSY measurements (Figure S18 in the Supporting In-
formation), four cobalt porphyrin alkoxides were identified (1–
4), which proved the feasibility of PO ring opening by TPPCoCl
in the absence of onium salts.
Scheme 2. Reaction pathway of TPPCoCl-catalyzed propylene oxide ring
opening.
tive ring opening by ClÀ (Scheme 2, step (a)). Notably, the NMR
signal of the methine group in 1 (À7.31 ppm) was approxi-
mately 5 ppm upfield shifted compared with the correspond-
ing signal in TPPAl–chloropropoxide (À2.28 ppm),[4g] inferring
either a different conformational distortion of the porphyrin
ligand or a higher degree of out-of-plane orientation of the
aluminum center, both of which would strongly influence the
shielding effect of the porphyrin ring current on the axially co-
ordinated ligand.[16]
The first ring-opening intermediate produced was identified
as TPPCoOCH(CH3)CH2Cl (1) from logical analysis of the high-
field NMR peaks that grew initially. This species was considered
to arise from the nucleophilic attack of ClÀ (coordinated or dis-
sociated, we do not specify at this point) on the less bulky
carbon atom of PO. Furthermore, a doublet of very low intensi-
ty at À1.46 ppm displayed a similar coupling constant to the
methyl group peak in 1, and thereby, was assigned to the
methyl group in TPPCoOCH2CH(CH3)Cl (2). The molar ratio be-
tween 1 and 2 remained constant over the entire reaction
period and was calculated to be 95:5, indicating a highly selec-
In addition to the formation of complexes 1 and 2, trace
amounts of water (100 ppm) in the reaction system produced
two additional cobalt porphyrin alkoxides, 3 and 4, through
the hydrolysis of complex 1 (Scheme 2, step (b)). 1-Chloro-2-
propanol was found in the NMR spectrum as one of the hy-
drolysis products (Figures S7 and S8 in the Supporting Infor-
mation). A metal hydroxide, TPPCoOH, should form in the hy-
drolysis with the axial OHÀ, a hydroxide that would then serve
as a new nucleophile for PO ring opening (Scheme 2, step (c)).
Chem. Eur. J. 2014, 20, 15499 – 15504
www.chemeurj.org
15500
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Indeed, following the growth of the NMR signals of 1 and 2,
another series of peaks arises in the high-field region and was
attributed to TPPCoOCH2CH(CH3)OH (3) and TPPCoOCH-
(CH3)CH2OH (4). After a certain reaction time, the formation of
species 1 slowed down, followed by a decrease in its NMR sig-
nals owing to the continuing consumption of 1 through hy-
drolysis. The molar ratio between 3 and 4 remained constant
throughout the reaction and was calculated to be 85:15.
Higher amounts of water in the system induce hydrolysis of
complexes 3 and 4 as well (Scheme 2, step (d)). The corre-
sponding hydrolysis product, namely propylene glycol, was
clearly identified when [D6]DMSO was used as the NMR solvent
(Figures S9 and S10 in the Supporting Information). Propylene
glycol formation was further confirmed by the shift and accu-
mulation of both the OH signals and the residual water peak
upon D2O addition.
a b-H peak at 9.08 ppm (Figure S3 in the Supporting Informa-
tion) and during the ring-opening reaction of PO, this signal
gradually decreased and another two b-H signals grew at 9.03
and 9.01 ppm, corresponding to complexes 1 and 3, respec-
tively (Figure S7 in the Supporting Information). Molecular
sieves (3 ꢁ) were used in the PO ring-opening system to scav-
enge the residual water and an NMR spectrum was collected
at 30 min reaction time. This showed complexes 1 and 2 as
the exclusive ring-opening intermediates, which supports the
viewpoint that complexes 3 and 4 were produced from ring
opening of PO by one of the hydrolysis products (TPPCoOH) of
the initially formed TPPCo-chloropropoxides (Figure S12 in the
Supporting Information).
As shown in our previous work,[19] TPPCoCl had no activity
in the homopolymerization of PO to produce poly(propylene
oxide) even in the presence of a cocatalyst (PPNCl). In the pres-
ent work also, no repeating insertion of PO was found by NMR
measurements, explaining the uniformly high carbonate con-
tent in TPPCoCl-produced copolymer (up to >99%) as a conse-
quence of extremely slow ether linkage formation.
The hydrolysis of complex 1, followed by the formation of
species 3 and 4, may be an explanation for the typically ob-
served bimodal molecular weight distribution of PPC generat-
ed by TPPCoCl:[7] two distinct initiators, ClÀ and OHÀ, would
produce copolymer chains with terminating groups that allow
their propagation either at only one end (chloro initiated) or si-
multaneously at both ends (hydroxy initiated).[17] Such a mecha-
nism of polymer formation with bimodal distribution implies
a faster chain-transfer reaction, caused by the presence of hy-
droxyl groups, compared with the propagation rate. Indeed,
substitution of coordinated chloro-/hydroxy propoxide ligands
in 1–4 with formation of TPPCo–methoxide species readily
takes place upon addition of trace amounts of methanol
(0.3 mL, 1 equiv compared with TPPCoCl), as shown by NMR
measurements (Figure S11 in the Supporting Information).
Taking into account such easy hydrolysis and alcoholysis of
TPPCo–alkoxide species, the ratio found between the com-
plexes 3 and 4 should reflect their relative thermodynamic sta-
bility rather than the regioselectivity of the PO ring opening by
a OH nucleophile. Most probable is that the interconversion of
3 and 4 is predominantly intramolecular. It should be noted,
however, that under the applied conditions, this interconver-
sion is still slow with respect to the NMR time scale, as all the
TPPCo–alkoxide species, 1–4, appear in the NMR spectrum as
sets of well-resolved signals with their expected coupling pat-
terns, except that of the OH proton in 3 (compare the NMR
spectra in Figure 1 and Figures S9 and S10 in the Supporting
Information).
As a side reaction in our experiments, slow reduction of
TPPCoIII to TPPCoII in the presence of propylene oxide took
place simultaneously with PO ring opening.[20a] The detailed
study of this redox reaction is in progress in our laboratory.
An attempt to gain deeper insight into the mechanism of
initiation was undertaken by studying the reaction kinetics.
The determination of TPPCoCl reaction order in PO ring open-
ing was achieved by monitoring the decrease of the TPPCoCl
b-H peak (TPPCoCl=5 mg, PO=0.3 mL, CD2Cl2 =0.5 mL). A
pseudo-first-order reaction was established from a very good
linear fit of the reaction time (x axis) and Àln(ct/c0) (y axis;
Figure 2, Table S1 in the Supporting Information). Although
this result might be understood as participation of only one
metal center in the ring-opening event, other reaction models,
for example, the bimetallic mechanism in Scheme 3 would also
give a pseudo-first-order reaction for TPPCo(Cl)PO consump-
tion. Indeed, assuming nearly equal reactivity of coordinated
PO for all TPPCo(X)PO species, the concentration of activated
PO will remain equal throughout the whole reaction, rendering
the reaction rate law in Scheme 3. To verify this model, further
kinetic studies were conducted by varying the initial TPPCoCl
concentration (TPPCoCl=2.5, 1.25, 0.65, and 0.25 mg, PO=
0.3 mL, CD2Cl2 =0.5 mL). According to the reaction model, the
A singlet at À16.26 ppm was noted during the ring-opening
reaction period (Figure S9 in the Supporting Information), both
in CD2Cl2 and [D6]DMSO, although more pronounced in the
latter; this signal was assigned to the hydroxyl complex,
TPPCoOH. Such a high-field shifted signal can be either as-
cribed to a proton that is very close to the porphyrin core and
is strongly shielded by the ring current, or to a metal hydride.
As no aldehyde or ketone peaks were detected, b-hydride
elimination of the metal alkoxide to afford a metal hydride was
excluded,[18] and the formation of TPPCoOH as a hydrolysis
product of 1 was suggested (step (b) in Scheme 2).
The chemical shift of the b-H porphyrin peak was found to
be sensitive to the nature of the axial ligand. TPPCoCl showed
Figure 2. Pseudo-first-order reaction of TPPCoCl-catalyzed PO ring opening.
Chem. Eur. J. 2014, 20, 15499 – 15504
www.chemeurj.org
15501
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 3. Bimetallic ring opening in a pseudo-first-order reaction.
initial rate of TPPCo(Cl)PO conversion should increase as
a square function of the initial concentration of the complex.
Similar behavior was observed, for example, by Jacobsen and
co-workers in experiments on epoxide ring opening using
a salen–CrIII complex as the catalyst, which enabled the estab-
lishment of a bimetallic mechanism for this reaction.[4b] As
shown in Figure 3, the double logarithmic plot of the reaction
rate (ln(r)) versus initial concentration of TPPCoCl (ln(c0)) gives
a linear fit with a slope of 1.45, corresponding to the reaction
order in TPPCoCl.
Scheme 4. Monometallic and bimetallic PO ring-opening pathway (X=Cl,
OH, or alkoxides). Dashed line shows a primarily electrostatic interaction
within the contact ion pair.
ic anions, for example, bis(triphenylphosphine)iminium chlo-
ride (PPNCl) are commonly used as cocatalysts for PO/CO2 co-
polymerization as they efficiently increase the reaction
rate.[19,21] In the present work, upon addition of 1 equivalent of
À
extra nucleophilic anions (ClÀ, BrÀ, or N3 ), the acceleration
effect on the ring opening of PO was clear, in that the starting
TPPCoCl was completely converted into TPPCo–alkoxides
within ten minutes compared with more than one hour in
their absence. An acceleration of axial ligand exchange in
cobalt porphyrin complexes in the presence of onium salts is
1
suggested by analysis of the H NMR spectra (Scheme 5).
Figure 3. Kinetic data plot showing a 1.45 reaction order when the initial
TPPCoCl concentration is varied.
This value is indicative for only a partial involvement of the
bimetallic mechanism in PO ring opening by TPPCoCl, which
thus competes with a monometallic route (Scheme 4). The
monometallic route may also proceed via a formation of con-
tact ion pairs. Moreover, a potential change in the reaction
rate law from the presence of trace amounts of water was ex-
cluded through addition of molecular sieves (3 ꢁ) into the
ring-opening reaction system, as under such conditions, the
same reaction order of 1.5 in TPPCoCl was subsequently ob-
tained. Interestingly, a reaction order of 1.5 in catalyst was also
determined for propylene oxide ring-opening polymerization
by TPPCrCl.[20b]
Scheme 5. Coordination equilibrium between PO, axial chloride, and addi-
tional nucleophiles.
Upon addition of 1 equivalent of PPNCl into the reaction
system, line broadening of the porphyrin b-H peaks was ob-
served. Further addition to 10 equivalents intensified the line
broadening and shifted the above-mentioned peaks towards
trans chloride-coordinated porphyrin species ([TPPCoXCl]À)
(Figures S14 and S15 in the Supporting Information). As the
À
ClÀ-coordinated TPPCoCl2 [PPN+] complex formed by addition
of 3 equivalents of PPNCl did not exhibit broad NMR signals in
the absence of PO (Figure S15 in the Supporting Information),
paramagnetism of the cobalt center induced by trans ClÀ coor-
dination was thus excluded. It is probable that there exists
ligand exchange between PO and additional nucleophiles at
the cobalt center, a process that causes the broadening of the
porphyrin b-H signals. To further prove this point, a weakly co-
Ring opening of PO in the presence of onium salts
PO ring opening catalyzed by TPPCoCl was also explored in
the presence of a series of onium salts to examine their influ-
ence on the ring-opening rate. Onium salts bearing nucleophil-
Chem. Eur. J. 2014, 20, 15499 – 15504
www.chemeurj.org
15502
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ordinating anion, BF4À, was used. As expected, addition of 1–
10 equivalents of PPNBF4 did not induce a noticeable change
in the porphyrin b-H peaks, even in the presence of PO (Fig-
ure S16 in the Supporting Information).
should be involved, as has often been reported for related cat-
alytic systems for CO2/PO copolymerization reactions.[10,21d,22]
Conclusion
Ring-opening intermediates produced from the additional
anions were also detected by using 1H NMR spectroscopy
(Figure 4). The proton resonance of the methyl group in 1 at
TPPCoCl was synthesized and identified by several characteri-
zation methods. The obtained UV/Vis and 1H NMR spectra
were compared with literature results and were discussed in
detail. PO ring-opening intermediates (1–7) formed by attack
1
of various nucleophiles were identified by using H NMR spec-
troscopy. Alcoholysis and hydrolysis of these cobalt(III) porphy-
rin alkoxides were demonstrated. Addition of nucleophilic
anions accelerated the PO ring opening significantly, and influ-
enced its regioselectivity. Under such conditions, acceleration
of ligand exchange on the cobalt center via a hexacoordinated
anionic species is suggested from analysis of the 1H NMR
spectra.
Experimental Section
The 1D solution NMR spectra were collected on a Bruker AVIII-300
spectrometer. 2D COSY measurements were carried out on
a Bruker AVIII-500 spectrometer. Mass spectra were collected on
a Varian LC-MS 500 (50–2000 Da) spectrometer. UV/Vis spectra
were collected by a Varian Cary 50 Scan UV–visible spectrophoto-
meter. Elemental analysis was performed at the microanalytic labo-
ratory of the Department of Inorganic Chemistry at the Technical
University of Munich. Single-crystal X-ray crystallography was car-
ried out at the Department of Inorganic Chemistry at the Technical
University of Munich. All chemicals and solvents were utilized as re-
ceived from commercial suppliers.
Figure 4. 1H NMR spectrum collected during PO ring-opening reaction in the
presence of 1 equivalent of PPNCl (spectrum 1), 0.3 equivalent of PPNN3
(spectrum 2), or 10 equivalents of NBu4Br (spectrum 3). Conditions: CD2Cl2,
300 MHz, 258C. Formation of species 3 and 4 owing to residual water was
observed in all three reactions.
Acknowledgements
Dr. C. E. Anderson is thanked for general discussions. Sugges-
tions for crystal growth from Mr. A. Plikhta are appreciated.
Bundesministerium fꢂr Bildung und Forschung is acknowl-
edged for financial support of this work (033RC1015C).
À2.50 ppm did not appear when 1 equivalent of PPNN3 was
added into the reaction system. Instead,
a doublet at
Keywords: carbon dioxide · NMR spectroscopy · onium salts ·
porphyrinoids · ring-opening polymerization
À2.67 ppm was detected and was assigned to the protons of
the methyl group in TPPCoOCH(CH3)CH2N3 (5) (Figure S17 in
the Supporting Information). If the PPNN3 loading was reduced
to 0.3 equivalents, both 1 and 5 were found in the NMR spec-
trum recorded during the ring-opening reaction period
(Figure 4). Upon addition of 10 equivalents of NBu4Br, two new
doublets appeared at À1.29 ppm (J=6.27 Hz) and À2.45 ppm
(J=6.26 Hz) and were assigned to the protons of the methyl
groups in TPPCoOCH2CH(CH3)Br (6) and TPPCoOCH(CH3)CH2Br
(7), respectively (Figure 4), although the ClÀ-produced metal
alkoxides were still the dominant species in this reaction. The
molar ratio of 1:1 between complexes 6 and 7 indicated
a lower regioselectivity of PO ring opening by bromide in the
presence of the PPN cation. Further, the ratio of 2 to 1 in-
creased under such conditions, both for the reaction in the
presence of PPNBr or PPNCl and of PPNN3 compared with the
reaction without external nucleophile. A change in the mecha-
nism of PO ring opening under these distinct conditions
[1] J. G. Smith, Synthesis 1984, 629–656.
[2] S. Klaus, M. W. Lehenmeier, C. E. Anderson, B. Rieger, Coord. Chem. Rev.
2011, 255, 1460–1479.
[3] S. Inoue, H. Koinuma, T. Tsuruta, J. Polym. Sci. Part B 1969, 7, 287–292.
[4] a) G. A. Luinstra, G. R. Haas, F. Molnar, V. Bernhart, R. Eberhardt, B.
Rieger, Chem. Eur. J. 2005, 11, 6298–6314; b) E. N. Jacobsen, Acc. Chem.
Res. 2000, 33, 421–431; c) L. P. C. Nielsen, C. P. Stevenson, D. G. Black-
mond, E. N. Jacobsen, J. Am. Chem. Soc. 2004, 126, 1360–1362; d) C. T.
Cohen, T. Chu, G. W. Coates, J. Am. Chem. Soc. 2005, 127, 10869–10878;
e) D. J. Darensbourg, J. C. Yarbrough, J. Am. Chem. Soc. 2002, 124,
6335–6342; f) D. J. Darensbourg, J. C. Yarbrough, C. Ortiz, C. C. Fang, J.
Am. Chem. Soc. 2003, 125, 7586–7591; g) M. H. Chisholm, Z. Zhou, J.
Am. Chem. Soc. 2004, 126, 11030–11039.
[5] P. Chen, M. H. Chisholm, J. C. Gallucci, X. Zhang, Z. Zhou, Inorg. Chem.
2005, 44, 2588–2595.
[6] H. Sugimoto, K. Kuroda, Macromolecules 2008, 41, 312–317.
[7] Y. Qin, X. Wang, S. Zhang, X. Zhao, F. Wang, J. Polym. Sci. Part A 2008,
46, 5959–5967.
[8] K. Nakano, S. Hashimoto, K. Nozaki, Chem. Sci. 2010, 1, 369–373.
Chem. Eur. J. 2014, 20, 15499 – 15504
www.chemeurj.org
15503
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[9] T. Aida, S. Inoue, Acc. Chem. Res. 1996, 29, 39–48.
[10] D. J. Darensbourg, R. M. Mackiewicz, J. Am. Chem. Soc. 2005, 127,
14026–14038.
[20] a) C. Chatterjee, M. H. Chisholm, A. El-Khaldy, R. D. McIntosh, J. T. Miller,
T. Wu, Inorg. Chem. 2013, 52, 4547–4553; b) C. Chatterjee, M. H. Chis-
holm, Inorg. Chem. 2012, 51, 12041–12052.
[21] a) D. J. Darensbourg, R. M. Mackiewicz, J. L. Rodgers, A. L. Phelps, Inorg.
Chem. 2004, 43, 1831–1833; b) D. J. Darensbourg, R. M. Mackiewicz,
J. L. Rodgers, C. C. Fang, D. R. Billodeaux, J. H. Reibenspies, Inorg. Chem.
2004, 43, 6024–6034; c) R. L. Paddock, S. T. Nguyen, Macromolecules
2005, 38, 6251–6253; d) X.-B. Lu, L. Shi, Y.-M. Wang, R. Zhang, Y.-J.
Zhang, X.-J. Peng, Z.-C. Zhang, B. Li, J. Am. Chem. Soc. 2006, 128, 1664–
1674.
[11] A. D. Adler, F. R. Longo, F. Kampas, J. Kim, J. Inorg. Nucl. Chem. 1970, 32,
2443–2445.
[12] K. Yamamoto, M. Hoshino, M. Kohno, H. Ohya-Nishiguchi, Bull. Chem.
Soc. Jpn. 1986, 59, 351–354.
[13] M. Nappa, J. S. Valentine, J. Am. Chem. Soc. 1978, 100, 5075–5080.
[14] J. Huet, A. Gaudemer, C. Boucly-Goester, P. Boucly, Inorg. Chem. 1982,
21, 3413–3419.
[22] a) T. Aida, M. Ishikawa, S. Inoue, Macromolecules 1986, 19, 8–13; b) D.-Y.
Rao, B. Li, R. Zhang, H. Wang, X.-B. Lu, Inorg. Chem. 2009, 48, 2830–
2836; c) D. J. Darensbourg, A. I. Moncada, W. Choi, J. H. Reibenspies, J.
Am. Chem. Soc. 2008, 130, 6523–6533.
[15] R. J. Abraham, C. J. Medforth, Magn. Reson. Chem. 1987, 25, 432–438.
[16] C. J. Medforth, C. M. Muzzi, K. M. Shea, K. M. Smith, R. J. Abraham, S. Jia,
J. A. Shelnutt, J. Chem. Soc. Perkin Trans. 2 1997, 839–844.
[17] a) H. Sugimoto, H. Ohtsuka, S. Inoue, J. Polym. Sci. Part A 2005, 43,
4172–4186; b) S. Klaus, S. I. Vagin, M. W. Lehenmeier, P. Deglmann, A. K.
Brym, B. Rieger, Macromolecules 2011, 44, 9508–9516.
[18] H. E. Bryndza, J. C. Calabrese, M. Marsi, D. C. Roe, W. Tam, J. E. Bercaw, J.
Am. Chem. Soc. 1986, 108, 4805–4813.
Received: June 27, 2014
Published online on October 3, 2014
[19] C. E. Anderson, S. I. Vagin, W. Xia, H. Jin, B. Rieger, Macromolecules 2012,
45, 6840–6849.
Please note: Minor changes have been made to this manuscript since its
publication in Chemistry—A European Journal Early View. The Editor.
Chem. Eur. J. 2014, 20, 15499 – 15504
www.chemeurj.org
15504
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim