A simple and efficient phosphorescent probe for iodide-specific detection based on crystallization-induced phosphorescence of organic ionic crystals

Article abstract of DOI:10.1039/c8tc04781b

It is very challenging to develop luminescent detection methods for iodide ions due to their significant fluorescence quenching of general fluorophores induced by the heavy-atom effect. Herein, we reveal a unique crystallization-induced bright phosphorescence of tetraphenylphosphonium iodide (TPP I) and develop a novel time-gated detection method for iodide ions to minimize the autofluorescence from complex biological samples. Both the involvement of heavy iodide ion and accessible ionic interaction-induced organic crystals contribute to the accomplishment of a very high phosphorescence quantum yield (0.42) for TPP I. This specific iodide-triggered bright organic phosphorescence enables the development of a novel time-gated detection method for iodide ions in the luminescence turn-on manner and further offers an opportunity to establish a facile test strip for the visual detection of iodide ions based on solid-state phosphorescence on a solid substrate. Excellent performance of this probe in imaging iodide in live cells and intriguing use in double encryption illustrate the versatile and broad application prospects of highly efficient organic ionic crystals in various areas.

Full text of DOI:10.1039/c8tc04781b

Received: 14 September 2018
DOI: 10.1002/aoc.4770
Revised: 19 November 2018
Accepted: 23 November 2018
F U L L P A P E R
Homoleptic, bis‐ligated magnesium complexes for ring‐
opening polymerization of lactide and lactones: Synthesis,
structure, polymerization behavior and mechanism studies
Xiao‐Lu Chen1 | Bin Wang1
| Li Pan1 | Yue‐Sheng Li1,2
1 Tianjin Key Lab of Composite and
Functional Materials, School of Materials
Science and Engineering, Tianjin
University, Tianjin 300072, China
2 Collaborative Innovation Center of
Chemical Science and Engineering
(Tianjin), Tianjin 300072, China
Bis‐ligated, homoleptic magnesium complexes 1–3 were synthesized through
the reaction of 1 equiv. dibutyl magnesium with 2 equiv. β‐ketiminato ligands
bearing different substituents on the nitrogen atom and 8 position on
benzocyclohexanone. All of the complexes were identified by nuclear magnetic
resonance (NMR) and X‐ray crystallography. Complexes 2 and 3 adopted
distorted tetrahedral geometry around Mg, by chelating of two ancillary
Correspondence
ligands, while complex
1 adopted a dimeric structure with penta‐
Bin Wang, Tianjin Key Lab of Composite
and Functional Materials, School of
Materials Science and Engineering,
Tianjin University, Tianjin 300072, China.
Email: binwang@tju.edu.cn
coordination around Mg. These complexes can be used as efficient catalysts
for the ring‐opening polymerization of L‐lactide, ε‐caprolactone, δ‐
valerolactone (δ‐VL) and trimethylene carbonate in the presence of alcohol
as a co‐initiator. With the increasing steric bulk of the ancillary ligands, the
catalytic activity of Mg complexes was improved significantly. Particularly,
complex 3 having the largest steric hindrance showed excellent catalytic perfor-
mance for the polymerization of δ‐VL. It could polymerize 800 equiv. δ‐VL in
10 min, and produce polyvalerolactone with narrow molecular weight distribu-
tions (Mw/Mn < 1.2) at 35°C or higher temperature. No transesterification side
reaction was observed. Moreover, complex 3 exhibited good tolerance to exces-
sive alcohol and an immortal polymerization characteristic. The mechanism
studies by in situ NMR demonstrated a coordination‐insertion process. Besides,
it revealed that the steric bulky substituents in the active species derived from
the complex and alcohol prevented the metal center from deactivation.
Funding information
National Natural Science Foundation of
China, Grant/Award Number: 21474100
KEYWORDS
bis‐ligated, magnesium complex, polyester, ring‐opening polymerization, δ‐valerolactone
1 | INTRODUCTION
because of their particular and remarkable biodegradabil-
ity and biocompatibility.[1–3] Therefore, fundamental
Over recent decades, biodegradable polymers have been
widely applied in the biomedical field and industries
research on the synthesis of aliphatic polyester, which is
an important biodegradable polymer, has attracted atten-
tion and developed rapidly.[4–6] Ring‐opening polymeriza-
tion (ROP) initiated by organometallic catalysts is one of
the most effective methods to synthesize polyester‐like
polylactide (PLA), polycaprolactone (PCL) and other
polyesters with high molecular weight (MW) and narrow
Electronic supplementary information (ESI) available: characterization
of ligand and binuclear complex, crystal data and structure refinements
of complexes 1–3, and crystallographic data in CIF format. CCDC:
1855155 for 1, 1855633 for 2 and 1855636 for 3.
Appl Organometal Chem. 2019;e4770.
wileyonlinelibrary.com/journal/aoc
© 2019 John Wiley & Sons, Ltd.
1 of 11
https://doi.org/10.1002/aoc.4770

2 of 11
CHEN ET AL.
molecular weight distribution (MWD).[7–12] A variety of
complexes with different metal centers have been studied
and proved to be effective catalysts. As it is hard to
remove metallic residues in polymers completely, the
resultant polyesters initiated by the complexes of toxic
metals such as tin and aluminum may have negative
effects on the human body, causing their limited applica-
tion in the field of green packaging and medical
devices.[13–17] At this point, the catalysts with non‐toxic
and biologically benign metals like magnesium,[18–26]
zinc,[27–33] calcium[26] and iron[34] are more suitable and
superior catalysts.[35,36]
In most organometallic complexes of divalent metals
such as magnesium and zinc for the ROP of cyclic esters,
the metal centers are supported by a multi‐dentate ligand
and the complexes are formed as type of LM‐X [where X
= Et, OBn, OSiPh3, OtBu2Ph, N (SiMe3)2 or NPr2].[37,38]
Generally, the efficient catalysts for the ROP of cyclic
esters were reported to be these heteroleptic complexes
(LM‐X). They are always more reactive than the corre-
sponding homoleptic complexes, because of the differ-
ence in electron unsaturation and Lewis acidity of the
magnesium center.[37,39–41] Most of the bis‐chelated com-
plexes ML2 were inactive or exhibited negligible activity,
sometimes because of their poor solubility.[41–46]
ligated zinc complexes, which could catalyze the ROP of
L‐lactide (L‐LA) at high temperature with a more acidic
phenol as an initiator (Scheme 1b).[37] Nonetheless, the
transesterification was still detected and hard to be elim-
inated, leading to the uncontrolled polymerization pro-
cess. The homoleptic zinc and magnesium complexes
synthesized by the Ejfler group in 2014 showed high effi-
ciency for ROP of L‐LA at 58°C, which completed the
polymerization almost in 15 min (Scheme 1c).[13] The
Dagorne group reported that the homoleptic bis
(phosphinophenolate) Zn (II) complex (Scheme 1d) in
combination with BnOH initiated the ROP of 100 equiv.
rac‐LA at room temperature via an ‘activated monomer’
mechanism.[39] To initiate the polymerization effectively,
it was necessary to feed rather a high ratio of catalyst
(over 1/50 equiv. to monomer), and the resultant PLA
showed low MW. The homoleptic zinc complex
(Scheme 1e) based on salicylaldiminato ligands synthe-
sized by the Wu group needed to be activated by raising
the temperature to 130°C.[47] Some homoleptic magne-
sium complexes (Scheme 1f–h) produced PCLs in 15–60
min, with MWs diverging from theoretical MWs or broad
MWD.[10,48,49] Though the homoleptic metal complexes
were not traditionally regarded as effective initiators for
ROP of cyclic esters, these elegant studies showed that
the homoleptic, bis‐ligated organometallic complexes
can be a potentially effective catalytic system in ROP of
lactides and lactones, especially in the presence of exter-
nal alcohol. As aforementioned, previous reported mono-
mers initiated by homoleptic complexes were mainly LA
and ε‐caprolactone (ε‐CL). ROP of other cyclic esters
was seldom reported.
However, previous reports showed that some
homoleptic complexes were indeed active in ROP of
cyclic esters in the presence of external alcohol
(Scheme 1).[10,13,37,40,47–49] The Ejfler group developed a
kind of hetero‐ and homoleptic molecular zinc complexes
with one or two equivalents of aminophenolate ligand in
2008 (Scheme 1a).[40] In 2014, Gerling et al. reported bis‐
SCHEME 1 The typical homoleptic, bis‐ligated organometallic complexes

CHEN ET AL.
3 of 11
Herein, we synthesized a series of magnesium bis
(chelate) complexes supported by two bidentate cyclic β‐
ketiminato ligands with different substituents attaching
proligands. Compared with proligand L3H, the proton
resonances of methyl groups in 2,6‐iPr2C6H3 belonging
to complex 3 split into four doublets, suggesting that
methyl groups become chemically inequivalent after
coordination with magnesium atom. The broad singlet
peaks at δ ∼ 11.3 ppm corresponding to the exchangeable
proton (‐OH) in the ketiminato ligands disappear. Addi-
tionally, there are no signals of the n‐butyl protons bound
to the magnesium metal center in the upfield region. The
successful coordination of the ligands with magnesium
and formation of bis (chelate) complexes are confirmed
by the NMR spectroscopic data and elemental analyses.
Single crystals of complexes 1–3 suitable for X‐ray crystal-
lographic analysis were prepared via the diffusion of hex-
anes into a toluene solution at ambient temperature.
Crystallographic and refinement data are summarized in
Table S1 (see ESI), and the selected bond lengths and
angles are listed in Table 1.
to the nitrogen atom and the
8
position on
benzocyclohexanone (Scheme 2). They were found to be
a kind of effective catalytic system in the ROP of lactides
and lactones, combining with benzyl alcohol (BnOH) and
isopropyl alcohol (iPrOH). Among them, the catalytic sys-
tem composed of complex 3 and alcohol displayed prom-
ising performance for ROP of various monomers,
including L‐LA, ε‐CL, δ‐valerolactone (δ‐VL) and
trimethylene carbonate (TMC) under mild conditions,
showing living and controlled nature. These homoleptic
L2Mg complexes were not stable in the presence of exter-
nal alcohol, and the real catalytic species were presumed
to be heteroleptic LMgOR complexes obtained by in situ
alcoholysis.
The magnesium atoms in complexes 2 and 3 are coor-
dinated with two bidentate N,O‐ligands, adopting a
monomeric structure with distorted tetrahedral geometry
around Mg. In contrast, complex 1 adopts a dimeric struc-
ture (Figure 1; Figures S4 and S5 of ESI), in which two
magnesium centers are penta‐coordinated with four
bidentate N,O‐ligands symmetrically and bridged via the
oxygen atoms of the ketiminato ligands. The formation
of dimer is ascribed to the low steric hindrance of the
proligand L1. Complexes 2 and 3 have very similar Mg‐
O bond lengths. The lengths of Mg1‐O1 and Mg1‐O2 are
1.9185(15) and 1.9174(16) Å for complex 2, and
1.9260(9) and 1.9260(9) Å for complex 3, respectively.
The bond length values are in the normal range of Mg–
O bond lengths (generally ranging from 1.889 to 1.999
Å).[24,51] The bond lengths between magnesium and
nitrogen (Mg1‐N1, Mg1‐N2) in complex 3 [2.0511(11)
2 | RESULTS AND DISCUSSION
2.1 | Synthesis and characterization of
bis‐ligated magnesium complexes
Cyclic β‐ketiminato proligands with different substituents
(L1H–L3H) were prepared according to the literature
methods.[50] The structures of the resultant proligands
were characterized spectroscopically and crystallographi-
cally. Complexes 1–3 were synthesized by the reaction
of MgnBu2 (2 M in heptane) and 2 equiv. ligands L1H–
1
L3H, and were characterized clearly by H‐nuclear mag-
netic resonance (NMR), 13C‐NMR spectra and elemental
1
analysis (see Figures S1–S3 of ESI). In the H‐NMR spec-
tra, the proton resonances of the Mg complex exhibit
apparent chemical shifts from those of the corresponding
TABLE 1 Selected bond lengths and angles in complexes 1–3
1
2
3
Bond lengths in Å
Mg1‐N1
2.1165(12)
2.1302(12)
2.0268(10)
2.0344(10)
2.0852(18)
2.0705(18)
1.9185(15)
1.9174(16)
2.0511(11)
2.0511(11)
1.9260(9)
1.9260(9)
Mg1–N2
Mg1–O1
Mg1–O2
Bond angles in °
O1–Mg1–N1
O1–Mg1–N2
O2–Mg1–O1
O2–Mg1–N1
O2–Mg1–N2
N2–Mg1–N1
97.73(4)
113.64(4)
173.16(4)
87.50(4)
89.54(4)
127.26(5)
104.63(7)
89.53(7)
134.90(8)
89.22(7)
104.90(7)
142.91(8)
114.41(4)
90.32(4)
121.54(6)
90.32(4)
114.41(4)
129.20(7)
SCHEME 2 Synthesis of complexes 1–3

4 of 11
CHEN ET AL.
the large steric bulk of the substituents in complex 3
makes the space around the magnesium center crowded
and 2,6‐iPr2C6H3 groups difficult to rotate. The steric bulk
around the metal center is able to prevent complex 3 from
deactivation by protons and transesterification side reac-
tion in the ROP process, which is going to be investigated
further in the mechanism study.
2.2 | ROP of lactides and lactones by
complexes 1–3
i
Combined with PrOH and BnOH, the homoleptic com-
plexes were proved to be effective catalysts for ROP of
L‐LA, ε‐CL, δ‐VL and TMC. All of the polymerizations
i
were explored at 80°C or 35°C in the presence of PrOH
or BnOH (1 equiv. to Mg complex), and the typical results
were summarized in Table 2. Complexes 1–3 were able to
polymerize L‐LA, ε‐CL and δ‐VL effectively, achieving
high monomer conversion easily in 1 hr under optimized
conditions.
FIGURE 1 ORTEP drawing of complex 3 with thermal ellipsoids
at the 30% probability level. Hydrogen atoms have been omitted for
clarity
and 2.0511(11) Å] are a little shorter than those of com-
plex 2 [2.0852(18) and 2.0705(18) Å]. Additionally, the
Mg‐N and Mg‐O bonds in complex 1 are the longest
[Mg1‐N1, Mg1‐N2, Mg1‐O1, Mg1‐O2 are 2.1165(12),
2.1302(12), 2.0268(10), 2.0344(10) Å, respectively]. This
implies the possible stronger coordination of magnesium
atom with nitrogen atom from L3 and the loose coordina-
tion around the magnesium atom in complex 1, on
account of the dimeric nature of complex 1 where each
magnesium center is penta‐coordinated.
Without any substituents in the 8 position of the phe-
nyl ring and aromatic amine moiety of β‐ketiminato
ligands in complex 1, the least steric hindrance leads to
the dimeric behavior. In complex 1, the angle between
O2–Mg1–O1 is 173.16(4), while the angles in complexes
2 and 3 are 134.90(8) and 121.54(6), respectively. This
demonstrates that the magnesium center is nearly co‐
planar with these two O atoms in comlex 1. Without pro-
tection from the steric hindrance of the bis‐chelated
bidentate proligands, it is liable to be attacked and
inactivated. The six‐membered Mg‐O‐C‐C‐C‐N ring is
nearly planar in complexes 1 and 2, but is more distorted
in complex 3. The torsion angles around the magnesium
center for 1 (Mg‐N1‐C7‐C8), 2 and 3 (Mg‐N1‐C17‐C14)
are 9.8°, 8.9° and −12.5°, respectively. The dihedral
angles between the aromatic moieties bonding to the
nitrogen atom and the six‐membered ring are −58.9° in
complex 1 (C7‐N1‐C6‐C5), 50.9° in complex 2 (C40‐N1‐
C41‐C42) and −81.63° in complex 3 (C17‐N1‐C26‐C25).
The aromatic ring linked with the nitrogen atom in com-
plex 3 is much more tilted and nearly perpendicular to
the organometallic skeleton. These results indicate that
Complexes 1 and 2 showed no activity in ROP of L‐LA
and ε‐CL at 35°C in the presence of BnOH (Table 2, entries
1, 3, 8 and 10). However, when increasing the temperature
to 80°C, they were able to polymerize L‐LA and ε‐CL rap-
idly (Table 2, entries 2, 4, 9 and 11). The conversion initi-
ated by complex 1 was up to 92% in 1 hr, producing PLA
with moderate MW (Mn = 24.2 kDa) and narrow MWD
(Mw/Mn = 1.12; Table 2, entry 2). The experimental MW
was consistent with the theoretical MW well, indicating
the controlled performance in ROP of L‐LA. However,
the resultant PCL produced by complexes 1 and 2 had
much lower experimental MW (Mn = 4.67 and 10.4 kDa)
with relatively broad MWD (Mw/Mn = 1.34 and 1.36) com-
pared with theoretical values (Table 2, entries 4 and 11).
Here, we hypothesized that there may be depolymeriza-
tion, transesterification and accelerated chain transfer in
the ROP of ε‐CL at high temperature. Interestingly, com-
plex 1 can prompt the ROP of δ‐VL at 35°C in the presence
of iPrOH or BnOH. Compared with iPrOH, BnOH can acti-
vate complex 1 more effectively (Table 2, entry 5 vs. 6). The
structure of the alcohol had great influence on ROP of δ‐
VL initiated by complex 1. The conversion was 49% in 1
hr with the co‐initiation of BnOH at 35°C, while it was
i
29% with PrOH, producing PVL with Mn = 7.31 kDa. In
the ROP of δ‐VL, complex 2 showed much higher activity
than complex 1 and was less dependent on alcohol co‐
initiator. It could initiate the ROP of 400 equiv. δ‐VL at
mild conditions and the monomer conversion was 69%
i
and 62% with PrOH and BnOH, respectively (Table 2,
entries 12 and 13).
The deviation between the experimental MW and cal-
culated MW was observed in ROP of ε‐CL and δ‐VL

CHEN ET AL.
5 of 11
TABLE 2 ROP of lactides and lactones catalyzed by complex 1–3/ROH systema
Entry Complex Monomer ROH T (°C) [M]/[I]/ROH t (min) Conv.b (%) Mn,calcd (× 103) Mn (× 103) Mw/Mn
c
d
d
1
1
1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3
3
L‐LA
L‐LA
ε‐CL
ε‐CL
δ‐VL
δ‐VL
δ‐VL
L‐LA
L‐LA
ε‐CL
ε‐CL
δ‐VL
δ‐VL
L‐LA
L‐LA
ε‐CL
ε‐CL
δ‐VL
δ‐VL
TMC
BnOH 35
BnOH 80
BnOH 35
BnOH 80
iPrOH 35
BnOH 35
iPrOH 80
BnOH 35
BnOH 80
BnOH 35
BnOH 80
iPrOH 35
BnOH 35
BnOH 35
iPrOH 35
BnOH 35
iPrOH 35
iPrOH 35
BnOH 35
iPrOH 35
200/1/1
200/1/1
200/1/1
200/1/1
400/1/1
400/1/1
400/1/1
200/1/1
200/1/1
200/1/1
200/1/1
400/1/1
400/1/1
200/1/1
400/1/1
200/1/1
400/1/1
400/1/1
400/1/1
400/1/1
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
10
10
10
10
10
trace
92
–
–
–
2
26.5
–
24.2
–
1.12
–
3
trace
92
4
21.0
11.6
19.6
35.2
–
4.67
n.d.
7.31
6.18
–
1.34
n.d.
1.22
1.37
–
5
29
6
49
7
88
8
trace
> 99
trace
94
9
28.8
–
33.2
–
1.26
–
10
11
12
13
14
15
16
17
18
19
20
21.5
27.6
24.8
22.5
37.5
22.1
30.1
37.6
37.2
35.9
10.4
8.68
13.3
28.3
33.3
20.3
30.2
28.7
31.4
10.7
1.36
1.52
1.54
1.17
1.11
1.09
1.07
1.11
1.09
1.39
69
62
78
65
97
66
94
93
88
aAll reactions were conducted in toluene at 80°C or 35°C, [M]0 = 1.0 M.
bMonomer conversions were calculated by 1HNMR.
cCalculated by ([Mono.]0/[OH]0) × Mw (Mono.) × conv. (%) + Mw (iPrOH) or Mw (BnOH).
dDetermined by GPC analysis (THF).
catalyzed by complexes 1 and 2. According to the
coordination‐insertion mechanism, the metal center at
the end of the active chain can attack not only the carbonyl
groups of the monomers but also those of active chains,
causing transesterification and depolymerization.[9] It is
assumed that complexes 1 and 2, with less steric hindrance
and weak protection around the metal center, can poly-
merize the lactones and depolymerize the polymer at the
same time. At 80°C, complex 1 could polymerize 400
equiv. δ‐VL in 1 hr and the conversion was up to 88%
(Table 2, entry 7), giving PVL with MW = 6.18 kDa. The
significant divergence of experimental Mn from the theo-
retical value (Mn,calcd = 35.2 kDa) indicated chain transfer
and depolymerization reaction under higher temperature.
To prove the transesterification or the depolymerization
reaction in complex 1, a sequential copolymerization of
200 equiv. ε‐CL first and then 200 equiv. δ‐VL was further
conducted by 1 equiv. complex 1 at 80ºC. The resultant
copolymer was analyzed by 13C‐NMR (Figure S6a). Four
resonance peaks corresponding to CL‐CL, VL‐VL, VL‐CL
and CL‐VL, respectively, were clearly detected, which con-
firmed the depolymerization and transesterification reac-
tion in the ROP of CL catalyzed by complex 1. By sharp
contrast, the majority of continuous CL‐CL and VL‐VL of
homodyads and a very small fraction of VL‐CL and CL‐
VL heterodyads were observed in the copolymer obtained
by complex 3 (Figure S6b). Apparently, complex 3 with the
largest steric bulk showed superior catalytic performances
for ROP of L‐LA, ε‐CL and δ‐VL at mild conditions. Com-
plex 3 was able to polymerize 200 equiv. L‐LA at 35°C and
the monomer conversion was up to 78% in 1 hr, while
complexes 1 and 2 did not exhibit satisfying activity under
the same conditions (Table 2, entries 1, 8 and 14). Com-
pared with the reported bis‐ligated complexes, where
higher temperature (70°C, 100°C or 130°C) or catalyst
loading (monomer/catalyst = 50/1) were necessary, com-
plex 3 with the highest activity was able to polymerize
200–400 equiv. L‐LA in 60 min and ε‐CL in 10 min at
35°C, giving PLAs and PCLs with moderate MW (Mn = 20–
35 kDa) and narrow MWD (Mw/Mn < 1.2).

6 of 11
CHEN ET AL.
It was noted that complex 3 also displayed excellent
reactivity than ε‐CL with a structure of a seven‐
membered ring. As complex 3 showed outstanding activ-
ity in ROP of δ‐VL, polymerization under different condi-
tions was studied in detail. The typical results were
summarized in Table 3. In most of the experiments, the
conversion reached up to 90% rapidly with the co‐
activity for ROP of δ‐VL, which can convert 400 equiv.
monomers in 10 min at 35°C in the presence of iPrOH or
BnOH (Table 2, entries 16–19). The polymerization rate
was rather fast compared with the reported N‐heterocyclic
olefins (NHOs)/Lewis acid pairs, which polymerize 100
equiv. δ‐VL in 80 min,[52] and the Mw/Mn was still less
than 1.1. In addition, complex 3 could polymerize 400
equiv. TMC almost completely in 10 min, affording PTMC
with relatively higher Mw/Mn (Table 2, entry 20). Taking
all of the results into account, it clearly demonstrated that
complex 3 possessed the highest activity in the ROP of L‐
LA and other lactones. With the increasing steric bulk of
the substituent in the ligands, the catalytic performance
of the complexes enhanced significantly. Moreover, the
catalytic activity of complex 1 was highly dependent on
the structure of the co‐initiator, while no significant differ-
ence in catalytic activity was observed for complex 3 by
using different alcohols. We envisioned that the bulky ste-
ric hindrance in active species protected the central mag-
nesium from deactivation and improved the stability of
active species during the polymerization process. We will
give a detailed discussion later.
i
initiation of PrOH or BnOH, affording polyvalerolactone
(PVL) with consistently narrow polydispersity. Complex 3
was capable of completing the polymerization of 800
equiv. δ‐VL in 10 min in toluene solution (M0 = 2.0 M)
under mild conditions (Table 3, entries 8 and 9). Increas-
ing the temperature resulted in the decrease of MW nota-
bly, which might be ascribed to the more rapid chain
transfer (Table 3, entry 1 vs. 3, 2 vs. 4). The MW increased
with a higher monomer feed ratio (Table 3, entry 3 vs. 6, 4
vs. 7, 9 vs. 10). For example, complex 3 can polymerize
1000 equiv. δ‐VL and afford PVL with Mn = 76.2 kDa,
while it produced PVL with Mn = 42.0 kDa when feeding
800 equiv. δ‐VL (Table 3, entry 9 vs. 10). The experimen-
tal MWs were relatively lower than the calculated MWs.
This divergence may be ascribed to the presence of trace
impurities as other chain transfer agents (CTAs). The
trace impurities might have more influence and make
the divergences more serious when less catalyst was
fed.[54] At high catalyst loading (the monomer to catalyst
ratio was under 400/1), the experimental MWs matched
better with the calculated values than at lower catalyst
loading (Table 3, entries 4 and 7). Besides, when the load-
ing of ROH was higher, the MWs were closer to the the-
oretical MW (Table 4), in agreement with the previous
2.3 | ROP of δ‐VL catalyzed by complex 3
There have been lots of investigation about the ROP of ε‐
CL, and a lack of study on the ROP of δ‐VL,[8,53] which
was a six‐membered ring and had lower polymerization
TABLE 3 ROP of δ‐VL catalyzed by complex 3/ROH systema
Time
(min)
Conv.b
(%)
Mn,calcd
Mn
c
d
d
Entry
ROH
T(°C)
[M]/[I]/ROH
(× 103)
(× 103)
Mw/Mn
1
iPrOH
iPrOH
iPrOH
iPrOH
BnOH
iPrOH
iPrOH
iPrOH
BnOH
BnOH
80
50
35
35
35
35
35
35
35
35
200/1/1
400/1/1
200/1/1
400/1/1
400/1/1
800/1/1
800/1/1
800/1/1
800/1/1
1000/1/1
60
10
60
10
10
60
10
10
10
20
91
94
92
94
93
93
78
96
90
83
18.2
37.6
18.4
37.6
37.2
74.5
62.4
76.9
72.1
83.1
4.52f
22.0
10.2
28.7
31.4
37.5
37.6
40.6
42.0
76.2
1.03f
1.14
1.10
1.11
1.09
1.09
1.14
1.13
1.12
1.27
2
3
4
5
6
7
8e
9e
10e
aAll reactions were conducted in toluene at 35°C, where [M]0 = 1.0 M.
bMonomer conversions were measured by 1H‐NMR.
cCalculated by ([VL]0/[OH]0) × Mw (VL) × conv. (%) + Mw (iPrOH) or Mw (BnOH).
dDetermined by GPC analysis (THF).
ePolymerization condition: [M]0 = 2.0 M.
fThe obtained polyester was tested by MALDI‐TOF MS.

CHEN ET AL.
7 of 11
TABLE 4 ROP of δ‐VL catalyzed by complex 3 with excessive alcohola
Time
(min)
Conv.b
(%)
Mn,calcd
(× 103)
Mn
c
d
d
Entry
ROH
[M]/[I]/ROH
(× 103)
Mw/Mn
1
2
3
4
5
6
7
iPrOH
iPrOH
iPrOH
BnOH
BnOH
BnOH
BnOH
800/1/5
10
10
10
5
96
85
95
64
88
95
94
15.4
6.87
3.80
10.3
7.05
3.91
1.99
10.4
7.13
4.8
1.15
1.11
1.17
1.06
1.10
1.09
1.02e
800/1/10
800/1/20
800/1/5
14.3
7.44
4.33
2.70e
800/1/10
800/1/20
1000/1/50
10
10
20
aAll reactions were conducted in toluene at 35°C, where [M]0 = 2.0 M.
bMonomer conversions were measured by 1H‐NMR.
cCalculated by ([VL]0/[OH]0) × Mw (VL) × conv. (%) + Mw (iPrOH) or Mw (BnOH).
dDetermined by GPC analysis (THF).
eThe obtained polyester was detected by MALDI‐TOF MS.
literature.[54] It was ascribed to the higher reactivity of
ROH or the polymeric alcohol in the transfer and
exchange than other impurities functioned as CTAs.[53]
As the ROP of δ‐VL cannot proceed smoothly without
alcohol as co‐initiator, the alcohol plays a significant role
in the initiation of polymerization. The effect of alcohol
loading content was also investigated, and the results
were displayed in Table 4. Most of the heteroleptic organ-
ometallic complexes were sensitive to the excess of exter-
nal alcohol.[10,13] However, complex 3 still exhibited high
catalytic activity when the ROH/Mg ratio varied from 5 to
50, with monomer conversion up to 90% in 10 min. The
molecular weight of PVLs was basically proportional to
the amount of ROH. This result showed an immortal
nature of the complex 3/ROH catalytic system in ROP
of δ‐VL, where each magnesium metal center could initi-
FIGURE 2 MALDI‐TOF MS analysis for the resulting PVL by
complex 3/iPrOH in toluene at 80°C ([M]/[I]/ROH = 200/1/1,
[M]0 = 1.0 M, t = 60 min)
ate at most 50 PVL chains with overloading of ROH. The
narrow MWD confirmed the reversible and fast chain
transfer caused by the excessive ROH as CTA. In addi-
tion, a linear relationship between Mn and monomer con-
version was found with consistently narrow MWD (see
Figure S7 of ESI). This ensured the living and controlled
nature of this binary catalytic system composed of the
magnesium complex 3 and alcohol. The kinetics study
moiety (d) and the aromatic ring belonging to the BnO‐
group (f) were found to be 3.56 ppm, 5.11 ppm and
7.35 ppm, respectively, indicating PVL macromolecular
chains were capped with the hydroxyl group at one end
and RO‐ group at the other end. The PVLs with low
MW produced by complex 3/iPrOH or BnOH binary sys-
tem (Table 3, entry 1; Table 4, entry 7) were tested by
MALDI‐TOF MS (Figure 2, and Figure S10 of ESI). All
of the calculated masses were in agreement with the
value of sodium‐complexed linear PVL. The MALDI‐
TOF analysis indicated that there was no back‐biting of
the terminal active ‐O* to ester groups along the polymer
chain, as evidenced by the absence of cyclic PVL
(Figure 2), even when the polymerizations were carried
out at high temperature (80°C). The MALDI‐TOF also
displayed BnO‐capped PVL with narrow MWD, in
i
was conducted in toluene in the presence of PrOH at
35°C with [VL] = 1.0 M and [M] = [I] = 1.25 mM. The
polymerization reaction was found to be first‐order
dependent on the monomer concentration with an appar-
ent rate constant of 0.135 min−1 (see Figure S8 of ESI).
The structures of PVLs produced by complex 3/ROH
were characterized by 1H‐NMR and MALDI‐TOF MS
1
(Figure 2, and Figures S9 and S10 of ESI). In the H‐
NMR spectra (see Figure S9 of ESI), the corresponding
resonance peaks assigned to the methene group in the
terminal hydroxyl part (peak e in Figure S9), the methene

8 of 11
CHEN ET AL.
agreement with the 1H‐NMR spectrum (see Figure S10 of
ESI). No cyclic PVLs were observed under the
overloading of ROH (BnOH/[Mg] = 50). According to
the extremely narrow MWD (Mw/Mn = 1.02), no
transesterification side reactions were detected, in spite
i
of PrOH or BnOH as co‐initiator. This ensured the
reversible and fast chain transfer reaction and the immor-
tal nature in the excess of ROH functioned as CTA again.
2.4 | Mechanism of investigation for ROP
of δ‐VL initiated by bis‐ligated
complexes/ROH system
The coordination‐insertion and the activated monomer
were two main mechanisms for the ROP of cyclic
esters.[10,55–58] To explore the initiation and polymeriza-
tion mechanism for complexes 1–3, the interactions
between monomer, Mg complexes and ROH (iPrOH or
BnOH) were investigated by in situ NMR at ambient tem-
perature, respectively.
FIGURE 3 1H‐NMR spectrum of the mixture of complex 3 and
iPrOH in tol‐d8 (reaction for 12 min) compared with the spectra of
complex 3 and ligand L3‐H (iPrOH: [Mg] = 1:1)
Generally, when the activated monomer mechanism
is operative, monomer is directly coordinated to the metal
center and activated by the complex, facilitating the sub-
sequent attack by external alcohol. Coordination of
monomer with metal complex will lead to peak shifts or
complex 3 and 1 equiv. BnOH (see Figure S12 of ESI).
Therefore, the reaction between complex 3 and 1 equiv.
ROH produced 1 equiv. alkoxide complexes LMgOR
(LMgOiPr or LMgOBn) as shown in equation 1, Scheme 3.
And they were the real active species that can initiate the
ROP of VL by inserting monomer into the metal‐alkoxide
bond.[55]
The reaction of complex 3 with different amounts of
BnOH was also traced in C6D6 (Figure 4). Different from
the mixture of complex 3 and BnOH with 1:1 molar ratio,
the spectrum of the 1:2 mixture appeared a new broad
peak near 4.3 ppm. This new resonance signal can be
ascribed to the methene protons of inactive BnO‐ from
Mg (OBn)2 in equation 2. The same signal was observed
in the mixture of MgnBu2 and 2 equiv. BnOH. This result
indicated that the active alkoxides LMgOBn derived from
equation 1 were not stable enough in excess of BnOH,
leading to deactivation and transformation into simple
alkoxides Mg (OBn)2. According to the polymerization
initiated by MgnBu2 under overloading of alcohol (see
Table S2 of ESI), the Mg (OiPr)2 were evidently inactive
species and Mg (OBn)2 showed negligible activity, proba-
bly because of their different steric hindrance and lack of
1
occurrence of new resonances peaks in H‐NMR spec-
tra.[10,56] In this work, no new resonance peaks or obvi-
1
ous chemical shifts were observed in H‐NMR spectra of
complex 3/δ‐VL mixture, and the resonance peaks still
kept intact when prolonging reaction time (see Figure
S11 of ESI). As aforementioned, no polymers were
obtained in the absence of ROH. This result strongly indi-
cated that the bis‐ligated complex alone had no interac-
tion with monomer, excluding the possibility of
activated monomer mechanism.[10,56,57]
i
Thus, PrOH and BnOH were added to complex 3,
respectively, to testify the coordination‐insertion mecha-
i
nism. When 1 equiv. PrOH was mixed with complex 3
in tol‐d8, the typical peaks of L3H were detected, indicat-
ing the formation of proligands. Meanwhile, new reso-
nances peaks at the range of 0.5–4.0 ppm were found
(Figure 3). The new resonance peak that showed up at
3.03 ppm (Figure 3) was ascribed to methine of the iso-
propyl groups on the amine moiety of LMgOiPr. In addi-
tion, four characteristic peaks attributed to the methyls of
isopropyl in complex 3 shifted from 0.24, 0.91, 0.96 and
1.11 ppm to 0.61, 1.07, 1.13 and 1.16 ppm. It was notewor-
thy that a new peak was detected at 0.92 ppm (f in
i
Figure 3), corresponding to the methyl groups of PrO‐
in alkoxides LMgOiPr. The integral areas of the alkoxide
LMgOiPr to the generated proligand were almost 1:1.
The same results were observed in the mixture of
SCHEME 3 The reaction equation between complex 3 and
different amounts of BnOH

CHEN ET AL.
9 of 11
FIGURE 5 1H‐NMR spectra of mixture of VL, complex 3 and
BnOH in C6D6 at room temperature (VL:[Mg]:[OH] = 5:1:1), and
the mixture of complex 3 and BnOH ([Mg]:[OH] = 1:1)
FIGURE 4 The reaction of complex 3 and 2 equiv. BnOH in
C6D6, compared with the reaction of complex 3 with 1 equiv.
BnOH and MgnBu2 with 2 equiv. BnOH (reaction time = 100 min)
on the stability of the alkoxide (see Figure S14 of ESI).
In sharp contrast, only the signal of proligands L1H and
no characteristic peaks of L1MgOiPr were observed. This
indicated that the LMgOiPr alkoxide derived from com-
plex 1, with the least steric bulk, was unstable and easy
to transform into inactive alkoxide Mg (OiPr)2. Combin-
ing with the experimental results (Table 1, entries 5, 6,
18 and 19), it was convinced that the steric hindrance of
alkoxide, LMgOR, including the steric bulk from the
ancillary ligand segment (L) and the alcohol segment
(RO‐), can stabilize the active species effectively, resulting
in higher apparent catalytic activity.
effective protection on the metal center. Nonetheless,
complex 3 still maintained catalytic activity and worked
well with excessive alcohol (even up to 50 equiv. BnOH).
In combination with the results in Table 4, we could
deduce that in the presence of monomer, the active mag-
nesium alkoxide (LMgOR) still initiated ROP of δ‐VL and
produced active chains rapidly. The redundant alcohol
actually functioned as an effective CTA instead of
deactivating the active species, determining the MW of
corresponding PVLs.
The influence of the different external alcohols on the
stability of LMgOR was investigated. It was observed that
the active alkoxides LMgOiPr decreased continuously in
In order to gain more insight into the polymerization
mechanism by alkoxides, a polymerization system with 5
equiv. VL, 1 equiv. complex 3 and 1 equiv. BnOH was
tracked by 1H‐NMR spectra in C6D6 at room temperature
(Figure 5). Besides the resonance signals of proligands,
some new additional signals were found near the charac-
teristic peaks of alkoxides LMgOBn, which could be
assigned to the active chain LMg‐(VL)n‐OBn. The
methine in isopropyl groups of L segment in LMg‐ was
detected at 3.02 ppm. Other newly formed signals that
appeared at 5.00 ppm and 3.95 ppm, with 1:1 integral
ratio, could be attributed to the CH2‐ in BnO‐ group and
CH2‐ in LMg‐ group in the active chain end, respectively.
It was deduced that during the ROP of δ‐VL, the mono-
mers inserted into the Mg‐O bond in alkoxide and formed
the active chain terminated by BnO‐ and LMg‐ groups.
They only reduced 9.6% after 80 min, exhibiting good sta-
bility of the active chain (see Figure S15 of ESI).
i
the presence of 2 equiv. PrOH (see Figure S13 of ESI).
All of the alkoxide complexes finally turned into
i
proligands. The broad signal of methine proton in PrO‐
shifted gradually from 3.63 ppm to 3.68 ppm. As the
i
methine proton in PrOH displayed a strong multiple
peak at 3.70 ppm, it suggested the transformation from
LMgOiPr into Mg (OiPr)2 as shown in equation 2,
Scheme 3. This implied that the active alkoxides
(LMgOiPr) were unstable in excess of iPrOH. By contrast,
the active alkoxides derived from BnOH were apparently
more stable. They still existed in excess of BnOH after
100 minutes, while the alkoxides from iPrOH were
completely consumed at the same time. This was ascribed
to the steric hindrance or the electronic effect of the BnO‐
group in alkoxide.
i
The reaction between complex 1 and 2 equiv. PrOH
Taking all into account, the coordination‐insertion
mechanism and immortal polymerization were strongly
was also explored, to test the steric effect of the ligand

10 of 11
CHEN ET AL.
SCHEME 4 Proposed polymerization
process and mechanism in ROP of δ‐
VL catalyzed by complex 3 and R‐OH
presumed in the ROP of δ‐VL initiated by homoleptic
complex MgL2, as shown in Scheme 4.[51,54] At first, the
complexes reacted with ROH and turned into active alk-
oxide, LMgOR. Next, the monomer coordinated to Lewis
acidic metal center, and then was attacked by RO‐group,
forming active chain ended with RO‐ and LMg‐. The
chain transfer rate and the initiation rate were much
faster than the propagation rate, producing PVLs with
ROH‐dependent MW and consistently narrow MWD.
According to the mechanism study, the stability of the
alkoxide and active chains were basically dependent on
the steric hindrance of organometallic complexes and
alcohol.
complex and alcohol was an efficient way to produce
PVL with different MWs and quite narrow MWD.
ACKNOWLEDGEMENTS
The authors are grateful for financial support by the National
Natural Science Foundation of China (No. 21474100).
ORCID
Bin Wang https://orcid.org/0000-0002-3130-3229
Yue‐Sheng Li https://orcid.org/0000-0003-4637-4254
REFERENCES
[1] G. L. Gregory, E. M. Lopez‐Vidal, A. Buchard, Chem. Commun.
2017, 53, 2198.
3 | CONCLUSIONS
[2] Y. Zhu, C. Romain, C. K. Williams, Nature 2016, 540, 354.
In conclusion, we developed a new series of bis‐ligated,
homoleptic magnesium complexes, displaying favorable
catalytic activity toward ROP of L‐LA, ε‐CL and δ‐VL.
Remarkably, complex 3, with the largest steric hindrance,
exhibited excellent catalytic performance for the ROP of
δ‐VL. The MALDI‐TOF analysis demonstrated that there
was no transesterification side reaction, leading to the
formation of linear PVL capped with RO‐ and hydroxy
group. Moreover, complex 3 retained its high catalytic
activity in excess of alcohol, affording PVLs with con-
trolled MWs and consistently narrow MWD. It showed
an immortal polymerization nature with rapid and
reversible chain transfer during the polymerization. The
monitoring of the polymerization process reflected a
coordination‐insertion mechanism by the magnesium
alkoxides (LMgOR) derived from the reaction between
the complex MgL2 and ROH. It also manifested the effect
of steric bulk on the stability of different alkoxides,
explaining the maintained high catalytic activity under
the overloading of alcohol. Above all, the new binary cat-
alytic system combined with the homoleptic magnesium
[3] D. K. Schneiderman, M. A. Hillmyer, Macromolecules 2017,
50, 3733.
[4] V. H. Sangeetha, H. Deka, T. O. Varghese, S. K. Nayak, Polym.
Compos. 2018, 39, 81.
[5] X. Zhang, M. Fevre, G. O. Jones, R. M. Waymouth, Chem. Rev.
2018, 118, 839.
[6] H.‐Y. Ji, B. Wang, L. Pan, Y.‐S. Li, Green Chem. 2018, 20, 641.
[7] S. M. Guillaume, E. Kirillov, Y. Sarazin, J. F. Carpentier, Chem.
Eur. J. 2015, 21, 7988.
[8] E. Stirling, Y. Champouret, M. Visseaux, Polym. Chem. 2018,
9, 2517.
[9] M. Labet, W. Thielemans, Chem. Soc. Rev. 2009, 38, 3484.
[10] I. D'Auria, C. Tedesco, M. Mazzeo, C. Pellecchia, Dalton Trans.
2017, 46, 12217.
[11] S. Dagorne, M. Normand, E. Kirillov, J.‐F. Carpentier, Coordin.
Chem. Rev. 2013, 257, 1869.
[12] S. Dagorne, C. Fliedel, Top. Organomet. Chem. 2013, 41, 125.
[13] J. Wojtaszak, K. Mierzwicki, S. Szafert, N. Gulia, J. Ejfler, Dal-
ton Trans. 2014, 43, 2424.

CHEN ET AL.
11 of 11
[14] C. S. Hege, S. M. Schiller, Green Chem. 2014, 16, 1410.
[41] M. Huang, C. Pan, H. Ma, Dalton Trans. 2015, 44, 12 420.
[15] S. P. Parwe, S. D. Warkad, M. V. Mane, P. S. Shedage, B.
[42] R. Petrus, P. Sobota, Organometallics 2012, 31, 4755.
Garnaik, Polymer 2017, 111, 244.
[43] S. Song, X. Zhang, H. Ma, Y. Yang, Dalton Trans. 2012, 41,
[16] C. Duval, C. Nouvel, J.‐L. Six, J. Polym. Sci., Part a: Polym.
Chem. 2014, 52, 1130.
3266.
[44] C. C. Roberts, B. R. Barnett, D. B. Green, J. M. Fritsch, Organ-
ometallics 2012, 31, 4133.
[17] C. Exley, Morphologie 2016, 100, 51.
[18] I. E. Nifant'ev, A. V. Shlyakhtin, V. V. Bagrov, M. E. Minyaev,
A. V. Churakov, S. G. Karchevsky, K. P. Birin, P. V. Ivchenko,
Dalton Trans. 2017, 46, 12 132.
[45] H.‐J. Chuang, H.‐L. Chen, B.‐H. Huang, T.‐E. Tsai, P.‐L.
Huang, T.‐T. Liao, C.‐C. Lin, J. Polym. Sci., Part a: Polym.
Chem. 2012, 51, 1185.
[19] A. Garcés, L. F. Sánchez‐Barba, J. Fernández‐Baeza, A. Otero, A.
Lara‐Sánchez, A. M. Rodríguez, Organometallics 2017, 36, 884.
[46] V. Poirier, T. Roisnel, J. F. Carpentier, Y. Sarazin, Dalton
Trans. 2011, 40, 523.
[20] J. Hu, C. Kan, H. Wang, H. Ma, Macromolecules 2018, 51, 5304.
[47] Z.‐R. Dai, C.‐F. Yin, C. Wang, J.‐C. Wu, Chinese Chem. Lett.
2016, 27, 1649.
[21] I. E. Nifant'ev, A. V. Shlyakhtin, A. N. Tavtorkin, P. V.
Ivchenko, R. S. Borisov, A. V. Churakov, Catal. Commun.
2016, 87, 106.
[48] B. Raghavendra, K. Bakthavachalam, B. Ramakrishna, N.
Dastagiri Reddy, Organometallics 2017, 36, 4005.
[22] I. E. Nifant'ev, A. V. Shlyakhtin, V. V. Bagrov, P. D. Komarov,
M. A. Kosarev, A. N. Tavtorkin, M. E. Minyaev, V. A.
Roznyatovsky, P. V. Ivchenko, Polym. Chem. 2017, 8, 6806.
[49] B. Raghavendra, P. V. S. Shashank, M. K. Pandey, N. D. Reddy,
Organometallics 2018, 37, 1656.
[50] D.‐P. Song, Y.‐X. Wang, H.‐L. Mu, B.‐X. Li, Y.‐S. Li, Organome-
tallics 2011, 30, 925.
[23] T. Rosen, I. Goldberg, W. Navarra, V. Venditto, M. Kol, Angew.
Chem. Int. Ed. 2018, 57, 7191.
[51] K. A. Frazier, R. D. Froese, Y. He, J. Klosin, C. N. Theriault, P.
C. Vosejpka, Z. Zhou, K. A. Abboud, Organometallics 2011, 30,
3318.
[24] H. Wang, J. Guo, Y. Yang, H. Ma, Dalton Trans. 2016, 45, 10
942.
[25] Y. Yang, H. Wang, H. Ma, Polyhedron 2016, 117, 569.
[52] S. Naumann, D. Wang, Macromolecules 2016, 49, 8869.
[26] N. Romero, Q. Dufrois, L. Vendier, C. Dinoi, M. Etienne,
[53] N. Ajellal, J. F. Carpentier, C. Guillaume, S. M. Guillaume, M.
Helou, V. Poirier, Y. Sarazin, A. Trifonov, Dalton Trans. 2010,
39, 8363.
Organometallics 2017, 36, 564.
[27] D. Jedrzkiewicz, D. Kantorska, J. Wojtaszak, J. Ejfler, S.
Szafert, Dalton Trans. 2017, 46, 4929.
[54] C. K. Williams, L. E. Breyfogle, S. K. Choi, W. Nam, V. G.
Young, M. A. Hillmyer, W. B. Tolman, J. Am. Chem. Soc.
2003, 125, 11 350.
[28] C. Kan, J. Hu, Y. Huang, H. Wang, H. Ma, Macromolecules
2017, 50, 7911.
[55] V. Poirier, T. Roisnel, J. F. Carpentier, Y. Sarazin, Dalton
Trans. 2009, 44, 9820.
[29] M. Keram, H. Ma, Appl. Organomet. Chem. 2017, 31, e3893.
[30] K. S. Kwon, S. Nayab, H.‐I. Lee, J. H. Jeong, Polyhedron 2017,
126, 127.
[56] Y. Sun, Y. Cui, J. Xiong, Z. Dai, N. Tang, J. Wu, Dalton Trans.
2015, 44, 16 383.
[31] S. Mantri, A. Routaray, N. Nath, K. Sutar Alekha, T. Maharana,
[57] C. Bakewell, A. J. P. White, N. J. Long, C. K. Williams, Inorg.
Chem. 2013, 52, 12 561.
Polym. Int. 2016, 66, 313.
[32] A. Otero, J. Fernandez‐Baeza, L. F. Sanchez‐Barba, S. Sobrino,
A. Garces, A. Lara‐Sanchez, A. M. Rodriguez, Dalton Trans.
2017, 46, 15 107.
[58] T. Rosen, I. Goldberg, V. Venditto, M. Kol, J. Am. Chem. Soc.
2016, 138, 12 041.
[33] D. Jędrzkiewicz, G. Adamus, M. Kwiecień, Ł. John, J. Ejfler,
Inorg. Chem. 2017, 56, 1349.
SUPPORTING INFORMATION
[34] R. Duan, C. Hu, X. Li, X. Pang, Z. Sun, X. Chen, X. Wang, Mac-
romolecules 2017, 50, 9188.
Additional supporting information may be found online
in the Supporting Information section at the end of the
article.
́
[35] C. Gallegos, V. Tabernero, F. M. Garcıa‐Valle, M. E. G.
Mosquera, T. Cuenca, J. Cano, Organometallics 2013, 32, 6624.
[36] Y. Huang, X. Kou, Y. L. Duan, F. F. Ding, Y. F. Yin, W. Wang,
Y. Yang, Dalton Trans. 2018, 47, 8121.
How to cite this article: Chen X‐L, Wang B, Pan
L, Li Y‐S. Homoleptic, bis‐ligated magnesium
complexes for ring‐opening polymerization of
lactide and lactones: Synthesis, structure,
polymerization behavior and mechanism studies.
Appl Organometal Chem. 2019;e4770. https://doi.
org/10.1002/aoc.4770
[37] K. A. Gerling, N. M. Rezayee, A. L. Rheingold, D. B. Green, J.
M. Fritsch, Dalton Trans. 2014, 43, 16 498.
[38] C. A. Wheaton, P. G. Hayes, B. J. Ireland, Dalton Trans. 2009,
25, 4832.
[39] C. Fliedel, V. Rosa, F. M. Alves, A. M. Martins, T. Avilés, S.
Dagorne, Dalton Trans. 2015, 44, 12 376.
[40] J. Ejfler, S. Szafert, K. Mierzwicki, L. B. Jerzykiewicz, P. Sobota,
Dalton Trans. 2008, 46, 6556.