Ring-opening polymerization of cyclic esters by pincer complexes derived from alkaline earth metals

Article abstract of DOI:10.1002/aoc.3098

We report the synthesis and characterization of new Ca(II) and Mg(II) complexes of the type [Ph - PPP]MR (M = Ca, R = N(SiMe₃)₂ (1); M = Mg, R = ⁿBu (2)) where Ph - PPP^- is a tridentate monoanionic ligand (Ph - PPP - H = bis(2-diphenylphosphinophenyl)phosphine). Reaction of the opportune metal precursor (Ca[N(SiMe₃) ₂]₂.THF₂ or MgⁿBu₂) with 1 equiv. of the pro-ligand Ph - PPP_-H produces the corresponding calcium amido (1) or magnesium butyl (2) complex in high yield. Solution NMR studies show monomeric and kinetically stable structures for both species. The obtained complexes efficiently mediate the ring-opening polymerization of E-caprolactone showing a turnover frequency of 40 000 h^-1. In the presence of a hexogen alcohol, up to 2000 equiv. of monomer are converted by using a low loading of catalyst (5 μmol). Kinetic studies show a first-order reaction in monomer concentration. Copyright

Full text of DOI:10.1002/aoc.3098

Full Paper
Received: 18 September 2013
Revised: 14 October 2013
Accepted: 25 October 2013
Published online in Wiley Online Library: 29 January 2014
(wileyonlinelibrary.com) DOI 10.1002/aoc.3098
Ring-opening polymerization of cyclic esters
by pincer complexes derived from alkaline
earth metals
Marina Lamberti, Antonella Botta and Mina Mazzeo*
We report the synthesis and characterization of new Ca(II) and Mg(II) complexes of the type [Ph―PPP]MR (M = Ca, R = N(SiMe3)2
(1); M = Mg, R = nBu (2)) where Ph―PPPÀ is a tridentate monoanionic ligand (Ph―PPP―H = bis(2-diphenylphosphinophenyl)
phosphine). Reaction of the opportune metal precursor (Ca[N(SiMe3)2]2.THF2 or MgnBu2) with 1 equiv. of the pro-ligand
Ph―PPP―H produces the corresponding calcium amido (1) or magnesium butyl (2) complex in high yield. Solution NMR
studies show monomeric and kinetically stable structures for both species. The obtained complexes efficiently mediate
the ring-opening polymerization of ɛ-caprolactone showing a turnover frequency of 40 000 hÀ1. In the presence of a
hexogen alcohol, up to 2000 equiv. of monomer are converted by using a low loading of catalyst (5 μmol). Kinetic
studies show a first-order reaction in monomer concentration. Copyright © 2014 John Wiley & Sons, Ltd.
Keywords: calcium; magnesium; ring-opening polymerization; cyclic esters
corresponding homoleptic species.[27–30] Since the ligand redistri-
bution reactions become more probable with the increase of the
Introduction
Aliphatic polyesters such as polycaprolactone (PCL) and
polylactide (PLA) represent an important class of biomaterials.
Since the products following their hydrolysis can be metabolized
or excreted from the human body via normal metabolic
pathways, they have found successful applications in medical
and pharmaceutical fields for the preparation of sutures, dental
devices, orthopedic devices, drug delivery systems and scaffolds
for tissue engineering.[1,2]
For biomedical applications a fine tailoring of the mechanical
properties and degradation profiles of these materials is crucial.
The election route for producing aliphatic polyesters with
designed macromolecular architectures, in terms of chain end
groups, molecular weights, molecular weight distribution and
microstructure, is the ring-opening polymerization (ROP) of the
related cyclic esters promoted by single-site metal catalysts.[3–5]
Various discrete metal complexes have been employed in the
ROP of cyclic esters and these include both main group and tran-
sition metals.[6–8] Since residual trace amounts of the catalyst may
contaminate the polymeric product, the use of catalysts based on
biological benign metals such as group IV, Ca, Mg, Zn and Fe is
highly desirable for biomedical applications.[9–15]
ionic radius of the metal, while stable heteroleptic magnesium
complexes are easily accessible, the analogous calcium com-
plexes are extremely scarce and known essentially as amido
derivatives. To prevent detrimental rearrangements of ligands
and to guarantee a certain stability of heteroleptic calcium
complexes, an opportune design of the ancillary ligand is decisive.
In the last years, very few examples of heteroleptic calcium
complexes as initiators for the ROP of cyclic esters have been
reported.[20,31–36] The first were developed by Chisholm et al.,
which described calcium complexes supported by sterically
encumbered tris(pyrazolyl)borates.[37–39] However, ancillary li-
gands able to stabilize heteroleptic species are still very limited,
and the most popular examples are essentially restricted to
aminotrop(on)iminates,[40] β-diketiminates[41–43] or Schiff-base
ligands.[44,45] In the last few years, some examples of stable cationic
calcium complexes have been reported by Carpentier et al.[46–48]
Recently we explored the coordination chemistry of phosphido
pincer ligands toward different metals, such as zinc, aluminum
and group 3 metals.[49–51] In the coordination to hard metal
centers, such as aluminum and group 3 metals, complexes in
which the metal center is chelated by a κ3-PPP pincer ligand have
been obtained. Conversely, in the case of the zinc, a soft metal
center, the pincer ligand acts as a flexible coordination environ-
ment. Density functional theory (DFT) studies about the ring-
opening polymerization promoted by zinc complexes highlighted
the importance of the coordinative flexibility of the ancillary
Recently, special attention has been given to the exploration of
catalysts based on zinc[16–21] and magnesium[17,21–26] complexes.
Despite their dimensional similarities, in terms of ionic radii (0.74
and 0.72 Å, respectively) they exhibit rather different chemical
properties, since magnesium (2+) is considered a hard metal,
while zinc (2+) is soft.
Conversely, the organometallic and coordination chemistry of
calcium is much less developed than that of magnesium and
zinc. Calcium is hard like the Mg2+ ion but it is significantly larger
(ionic radius = 1.14 Å); therefore its heteroleptic complexes usu-
ally show very low stability as they readily decompose by ligand
redistribution reactions in Schlenk-type equilibria to generate the
*
Correspondence to: Mina Mazzeo, Department of Chemistry and Biology,
University of Salerno, Via Giovanni Paolo II, 132, Fisciano I-84084, Salerno, Italy.
Email: mmazzeo@unisa.it
Department of Chemistry and Biology, University of Salerno, Fisciano I84084,
Salerno, Italy
Appl. Organometal. Chem. 2014, 28, 140–145
Copyright © 2014 John Wiley & Sons, Ltd.

ROP of cyclic esters by group II metal complexes
ligand to promote the coordination of the monomer at the metal
reactive center.[52]
In this contribution we extend the use of a tridentate phospine-
based pincer ligand to the synthesis of new heteroleptic complexes
of calcium and magnesium.
The synthesized complexes have been tested as catalysts for
the ROP of ɛ-caprolactone (ɛ-CL) and lactides, and compared with
the related zinc complex previously reported.
In solution, complex 2 shows the same symmetry discussed for
complex 1. In the 31P{1H}NMR spectrum a doublet for the two
equivalent neutral phosphorus donors (δ = À10.7 ppm) and a
triplet for the anionic phosphido donor (at À41.4 ppm) are
observed. The 31P NMR chemical shift of the phosphido donor
moves toward higher field in comparison to that observed for
the larger calcium ion. This difference can be due to the higher
ionicity of the metal–phosphorus bond in complex 1, which leads
to enhanced charge on the phosphorus atom and to better
charge back-donation from phosphorus into the phenyl π
system. This effect is also evident in the 13C NMR spectrum of
complex 1, in which a downfield shift of the ipso-carbon atom
of approximately 20 ppm compared to the signal of the same
carbon of the neutral phosphine proligand is observed. This
trend is coherent with that observed for arylphosphanide
complexes of the alkaline earth metals.[54]
Results and Discussion
Synthesis and Characterization
The pro-ligand Ph―PPP―H has been synthesized following a
previously published synthetic procedure.[53]
Although, according to Pearson’s HSAB theory,[55] the forma-
tion of bonds between hard alkaline earth metal cations and soft
phosphorus donor ligands should be disfavored, in this case
complexes stable at room temperature are obtained. Probably,
the incorporation of phosphorus donor atoms into a chelating
structure of a pincer-type ligand mitigates this difficulty, stabiliz-
ing the coordination at the metal by the formation of two five-
membered metallacycles.
The reaction of Ca[N(SiMe3)2]2.THF2 with 1 equiv. of the
pro-ligand Ph―PPP―H in benzene at room temperature gives
the complex [Ph―PPP]CaN(SiMe3)2 (1) in quantitative yield
(see Scheme 1). After removal of the solvent, by evaporation
under reduced pressure, complex 1 has been isolated as a pale-
yellow extremely air-sensitive solid. It was fully soluble in ethers
and in aromatic solvents such as toluene, and mostly insoluble
in hexane.
The parent magnesium complex [Ph―PPP]MgnBu (2) has been
obtained by alkane elimination reaction of the proligand
Ph―PPP―H and 1 equiv. of MgnBu2 in benzene/THF solution
at room temperature.
Ring-Opening Polymerization Studies
The ability of the heteroleptic complexes 1 and 2 to promote the
ROP of ɛ-CL was examined (Scheme 2). The obtained polymers
have been characterized by 1H NMR and gel permeation chroma-
tography (GPC). The most representative results are summarized
in Table 1. These data have been compared with the results
previously reported for the analogous zinc complex 3 (see
Scheme 1).
Both complexes 1 and 2 have been characterized by
multinuclear NMR spectroscopy (1H, 13C, 31P NMR) and elemental
analysis.
NMR and elemental analysis of complexes 1 and 2 are coher-
ent with six-coordinate structures consisting of one phosphido
pincer ligand, one labile ligand (amido or alkyl group) and two
molecules of THF coordinated at the metal center.
Both complexes 1 and 2 are efficient initiators for the ROP of ε-CL
to give high-molecular-weight polymers.
In the 1H NMR spectrum of complex 1 (C6D6, 25°C) the methyl
protons of the bis(trimethylsilyl) amido group appear as a single
signal at 0.08 ppm. In the aromatic region, a single set of signals
is detected, suggesting that the two phenyl groups on each
phosphorus are equivalent. Coherently, in the 31P{1H} NMR spec-
trum two resonances are observed: a doublet for the two equiv-
alent neutral phosphine donors (δ = À4.96 ppm) and a triplet for
the anionic phosphorous donor (δ = À10.2 ppm). Both reso-
nances are shifted downfield from the corresponding ligand pre-
cursor (δ = À10.9 ppm for Ph2P― and δ = À53.8 ppm for Ar2PH).
All NMR data support the κ3 coordination mode of the ancillary
ligand to the metal center and suggest a highly symmetric
structure of complex 1 in solution coherent with an averaged-
C2v symmetry on the NMR timescale. Complex 1 is stable in
solution at room temperature for days, as no evidence of
detrimental Schlenk-type equilibrium phenomena involving
[Ph―PPP]2Ca and Ca[N(SiMe3)2]2 have been detected during
The calcium amide complex 1 shows high efficiency in the
ring-opening polymerization of ɛ-CL under mild reaction condi-
tions and with a very low loading of initiator (5 μmol). A turnover
frequency of 40 000 hÀ1 (run 1 of Table 1) is achieved at room
temperature.
The molecular weight distribution of the produced polymers is
monomodal, although the measured Mn values are twice the cal-
culated Mn values assuming the growth of one polymer chain per
calcium initiator. This suggests that, possibly, only a fraction of
the metal complex is involved in catalysis, most likely as a result
of an inefficient initiation by the poor nucleophilic amide group.
Upon addition of one equivalent of isopropanol, a more
efficient control on the molecular weights is achieved as demon-
strated by the good agreement between the theoretical and
experimental values (cf. runs 1 and 2 of Table 1). Reasonably,
the isopropoxide initiating group, formed by alcoholysis of the
amido group, mimics more efficiently the putative propagating
group of the presumed active species; thus polymeric chains of
predictable molecular weights and narrow molecular weight
distribution are produced.
1
monitoring by H NMR of a solution of 1 in C6D6 over a period
of several days at room temperature.
Scheme 1. Structures of the metal complexes 1–3.
Scheme 2. Ring-opening polymerization of ε-CL initiated by complexes 1–3.
Appl. Organometal. Chem. 2014, 28, 140–145
Copyright © 2014 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/aoc

M. Lamberti et al.
Table 1. Ring-opening polymerization of ε-CL initiated by complexes 1–3a
iPrOH (equiv.)
[ɛ-CL]/ [I]
Time (s)
Yieldb (%)
Mn GPC(×103)
c
Mtnh d (×103)
Mw/Mn
c
Run
Cat.
Solvent
1
1
1
1
1
1
1
1
1
1
2
2
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
THF
—
1
400
400
30
30
82
100
96
83.7
45.0
12.4
16.2
27.3
43.1
22.9
40.6
20.4
120.0
47.8
37.4
45.7
7.3
2.47
1.29
1.27
1.31
1.52
1.44
1.59
1.27
1.15
1.77
1.13
2
3
6
400
30
4
10
10
10
—
1
500
180
180
240
30
100
100
98
5.7
5
1000
2000
400
11.4
22.4
21.5
34.6
10.7
25.6
45.6
6
7
47
8
THF
400
40
76
11
12
13
THF
10
—
1
2000
400
240
30
47
Toluene
Toluene
100
100
400
30
aReaction conditions : [Ι] =2.5 mM , 2 ml solvent, at 25°C.
bConversion of ε-CL as determined by 1H NMR spectral data.
cExperimental Mn (gmolÀ1) and Mw/Mn (PDI) values were determined by GPC in THF using polystyrene standards and corrected using a factor of 0.56.
dMtnh ( gmolÀ1) = 114.14 × ([ε-CL]/[iPrOH ]) × conversion ε-CL.
Upon addition of an excess of iPrOH, complex 1 provides an ef-
ficient binary catalytic system for the ‘immortal’ ROP of ε-CL (runs
2–6, Table 1). The conversion of 2000 equiv. of monomer is
achieved in few minutes at room temperature by using only 5
μmol of metal catalyst.
The molar masses of the obtained PCLs depend on the mono-
mer/alcohol ratio, indicating that the number of polymer mole-
cules increases proportionally to the equivalents of added
alcohol. This suggests that the isopropyl alcohol acts as chain
transfer agent and that the rate constant ktr for transfer between
growing chains and resting alcohol molecules or the dormant
hydroxy-end-capped polymer chains was far greater than that
for chain propagation (kpr; Scheme 3). In all cases the distribu-
tion of molecular weights is monomodal but the molecular
weights are significantly higher than those theoretically
calculated. Plausibly, in the presence of more than 1 equiv. of
alcohol, the calcium complex 1 might lead to the formation
of polynuclear species.[56]
The polymerization of ɛ-CL shows a high solvent dependence.
The use of THF as polymerization solvent decelerates the
polymerization because it competes with monomer for metal
coordination (cf. runs 1 and 7, Table 1). As already observed
in toluene solution, in the presence of 1 equiv. of alcohol, a
lightly increased polymerization activity is observed (cf. runs
1 and 7, Table 1) while, in the presence of more than 1 equiv.,
the agreement between the experimental Mn values and the
expected values is lost as a consequence of a plausible forma-
tion of polynuclear species.
well-documented reactivity of calcium complexes toward chlori-
nated solvents. In general, calcium amide derivatives react with
chlorinated hydrocarbon solvents, leading to chloro-metal com-
plexes, presumably via initial deprotonation of the acidic protons
of CH2Cl2 by the strong amide base.[38,44]
1H NMR analysis of a low-molecular-weight PCL sample
obtained by carrying out a polymerization experiment in the
presence of 200 equiv. of monomer and 20 equiv. of isopropanol
discloses the presence of isopropyl ester end groups (―COOCH
(CH3)2; 1.26 and 5.00 ppm), generated via insertion of the mono-
mer unit into the Ca―OCH(CH3)2 bond, and hydroxyl end groups
(CH2CH2OH; 3.64 ppm), generated by hydrolysis of the growing
chain (see Fig. 1).
These data suggest that, in the presence of isopropanol, the
OiPr group is the only initiating moiety involved in the polymer-
ization process; therefore, a coordination–insertion mechanism
should be operative.
The parent magnesium complex 2 shows a similar activity to
the heavier congener calcium (cf. runs 1 and 2 with runs 12 and
13, Table 1). The zinc complex 3 is intrinsically less active than
its group II-based equivalents and the conversion of about 400
equiv. of monomer was obtained only after 6 h.[52] This is in line
with the trend previously observed for these metals in the ROP of
cyclic esters[37,38,46,57,58] and could be correlated with the
expected degree of ionic or polar character of the M-X bond (X =
initiating group). Moreover, according to DFT calculations, coordi-
nation of the monomer to the zinc reactive center is possible only
after dechelation of one arm of the pincer ligand, while for the
six-coordinated complexes 1 and 2 this might occur by an easy
substitution of the coordinated solvent.
Compound 1 shows no activity in CH2Cl2 solution under the
same polymerization conditions, which could be due to the
Polymerization of Lactide
Complexes 1 and 2 have also been assessed as
initiators for the ROP of L-, D- and rac-lactide
(Scheme 4). The obtained polymers have been
characterized by NMR and GPC. The most repre-
sentative results are summarized in Table 2.
In the ROP of lactide, the calcium complex 1
shows good activity. The conversion of 200 equiv.
Scheme 3. ROP of ε-CL promoted by the binary catalytic systems 1–2/iPrOH.
wileyonlinelibrary.com/journal/aoc
Copyright © 2014 John Wiley & Sons, Ltd.
Appl. Organometal. Chem. 2014, 28, 140–145

ROP of cyclic esters by group II metal complexes
narrow polydispersity indexes (PDIs ranging from 1.17 to 1.27)
and they are also in good agreement with the Mn values theoret-
ically calculated. Reasonably, this might be a consequence of a
lower propagation rate in comparison to the initiation rate.
The molecular weights Mn of the resulting polymers increase
linearly with the conversion (runs 14–16, Table 2). This further
proves the controlled character of the polymerization reaction.
The tacticity of the resulting polymers has been investigated
by examination of the methine region of the homonuclear
decoupled 1H NMR spectra. The polymers obtained from the
enantiopure monomers L- and D-LA (runs 14–18, Table 2) are
isotactic; thus no epimerization of the stereogenic centers occurs
in the course of polymerization. Atactic PLAs are obtained from
racemic lactide (runs 19 and 23, Table 2) when either toluene or
THF is used as polymerization solvent.
Figure 1. 1H NMR spectrum (400 MHz, CDCl3, 25°C) of low-molecular-weight
PCL. Reaction conditions: [ɛ-CL]/[1]/[iPrOH] = 200:1:20, T=25°C, t=30 s.
Kinetics studies show that the polymerization of L-LA by 1 follows
a first-order dependence on the concentration of lactide, with an
apparent propagation rate constant kapp =8.333 × 10À2 minÀ1 at
25 °C (Fig. 2).
of the monomer L-LA is achieved within 1 h, under mild polymer-
ization conditions (room temperature in toluene solution; see
run 14, Table 2). The same activity is observed in the polymer-
ization of the enantiomer D-lactide (see run 18, Table 2),
while a reduced reactivity is achieved in the presence of
the racemic mixture (see run 19, Table 2).
Both calcium and magnesium complexes show lower activity
in the ROP of lactide than that observed toward ɛ-CL, in agree-
ment with the behavior commonly observed in the ROP of these
monomers.[59,60] In the ROP of lactide, no significant differences,
either in terms of activity or molecular weight control, are
detected when an equivalent of alcohol is added to the initiator
(run 17, Table 2).
Surprisingly, in the ROP of lactide, the reactivity of the zinc
complex 3 is comparable to that of the group II metal initiators
and is higher than that observed toward ɛ-CL.[61] The activity is
further improved in the presence of 1 equiv. of alcohol. DFT cal-
culations performed for complex 3 suggest that the difference
in polymerization rates of the two monomers is attributable to
the different coordination steps. In fact, coordination of lactide
is energetically more favorable than that of caprolactone.
The parent complexes of the metals of group II are defi-
nitely higher Lewis acid than zinc and this should favor the
coordination of the monomer but, at the same time, the high
stability of the formed adducts should slow down the subse-
quent insertion step.
Differently from that observed for the polymerization of ε-CL,
the experimental Mn values of the PLAs produced by 1 show
Scheme 4. Ring-opening polymerization of lactide initiated by
complexes 1–3.
Table 2. Ring-opening polymerization of lactides by 1–3a
b
c
b
Run Cat. Monomer Time Yield Mn,GPC
Mn,th
Mw/Mn
(min)
(%)
(×103)
(×103)
14
1
1
1
1
1
1
2
3
3
1
L-LA
L-LA
15
30
60
60
30
30
30
60
20
60
30
61
95
97
50
24
94
70
91
12
7.5-
15.5
23.2
24.1
12.6
7.2
8.6
17.6
27.4
28.8
14.4
6.9
1.17
1.17
1.20
1.17
1.16
1.12
1.27
1.23
1.05
1.03
15
16d
L-LA
17
L-LA
18
D-LA
rac-LA
L-LA
19
20
29.8
389.2
8.8
26.5
20.2
8.7
21
L-LA
22d
23e
L-LA
rac-LA
3.0
3.1
aAll reactions were carried out at 25°C with [I] = 2.5 mM ([LA]/
[I] = 200 in 2 ml toluene.
bExperimental Mn (corrected using a factor of 0.58) and PDI
values were determined by GPC analysis in THF using
polystyrene standards.
cCalculated Mn of PLA (in gmolÀ1) =144.14 × ([LA]/[iPrOH] ×
conversion L-LA.
Figure 2. Pseudo-first-order kinetic plot for ROP of L-LA promoted by 1.
Pseudo-first-order rate constant kapp = 8.333 × 10À2 6.5 × 10À4 minÀ1
(R = 0.9976). Reaction conditions: [1] = 8.3 mM, [L-LA]/[1] = 100; C7D8;
T = 25 °C.
dOne equivalent of iPrOH was added.
eTHF was used as solvent.
Appl. Organometal. Chem. 2014, 28, 140–145
Copyright © 2014 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/aoc

M. Lamberti et al.
independent measurements. It was corrected using the factor
of 0.58 for polylactide and 0.56 for polycaprolactone according
to the literature.[63–65] Elemental analyses were performed in
the microanalytical laboratory of the institute.
Conclusions
In this paper we report the synthesis and characterization of new
calcium and magnesium complexes with a tridentate phosphido
pincer ligand. Both complexes are stable in solution and no
evidence of detrimental Schlenk-type equilibrium is detected.
This demonstrates the ability of an uncommon coordination, such
as a phosphorous-based pincer ligand, to efficiently coordinate
the hard Lewis acid metals of group II.
Complexes 1 and 2 show excellent catalytic activity in the ROP
of ɛ-CL in comparison to the parent zinc complex. Conversely, in
the ROP of lactide the difference of reactivity among all com-
plexes is less significant. Reasonably, because of the high Lewis
acidity of the group II metals the insertion step is not favored.
In all cases the polymerization reactions are well controlled and
produce high molecular weight polymers.
Synthesis of [(o-C6H4PPh2)2PCa(NSiMe3)2(THF)2] (1)
A benzene solution (3 ml) of Ph―PPP―H ligand (0.170 g, 0.31 mmol)
was added slowly to a benzene solution (2 ml) of [Ca(NSiMe3)2(THF)2]
(0.152 g, 0.31 mmol) to give a red solution. After being stirred for
60 min, all volatiles were removed under vacuum, and the resi-
due red solid was washed with cold hexane and then dried up
at reduced pressure for 3 h (yield 0.266 g, 98%. The crude solid
was recrystallized from hexane at À20°C to obtain a microcrys-
talline powder (yield after crystallization 63.4%).
1H NMR (400 MHz, C6D6, 298 K): δ 7.92 (m, 2H, CH phenyl (a)),
3
3
7.51 (t, JH–H = 8 Hz; 2H, CH phenyl,(b)), 7.24 (t, JH–H = 8 Hz; 8H,
CH, ortho Ph)), 7.10–7.01 (m, 2H, CH, phenyl (d)), 6.96–6.88
(m, 12 H, CH, para and meta Ph)), 6.66 (t, JH–H = 8 Hz; 2H, phenyl
(c)), 3.56 (m, 8H, α CH2 of THF), 1.30 (m, 8H, β CH2 of THF), 0.08
(s, 18H, Si(CH)3).
Experimental
General Procedures: Materials and Methods
All manipulations of air- and/or water-sensitive compounds were
carried out under dry nitrogen atmosphere using a Braun
Labmaster glovebox or standard Schlenk line techniques. Glass-
ware and vials used in the polymerization were dried in an oven
at 120°C overnight and exposed to vacuum–nitrogen cycle three
times. Hexane and THF (Carlo Erba) were purified by distillation
from sodium benzophenone ketyl. Toluene (Carlo Erba) was puri-
fied by distillation from sodium. Anhydrous dibutyl magnesium
(Aldrich) was used as received. All deuterated solvents were dried
using molecular sieves. The proligand Ph―PPP―H was prepared
according to published procedures.[53] Lactide (Aldrich) was puri-
fied by P2O5 and then by two subsequent crystallizations from
hot toluene. After purification, lactide was stored at – 20 °C under
the inert atmosphere of the glovebox. ε-Caprolactone (Aldrich)
was dried with CaH2 for 24 h at room temperature, then distilled
under reduced pressure and kept over activated 3 Å molecular
sieves. All other chemicals were commercially available and used
as received unless otherwise stated.
13C{1H} NMR (100.62 MHz, C6D6): δ 166.5 (td, JC–P = 43 Hz,
JC–P =22 Hz, Cipso―P (phosphido)), 138.1 (dt, JC–P =23 Hz, JC–P =4
Hz, CH, Ph), 136.2 (td, JC–P = 48 Hz, JC–P =20 Hz, Cipso―PPh2) 134.2
(t, JC–P =7 Hz, CH; Ph), 131.8 (d, JC–P =6 Hz, CH; Ph), 132.5 (t, JC–P =6
Hz, CH; Ph), 132.0 (m, Cipso―P (phosphine)), 130.2 (s, CH; Ph), 128.4
(d, JC–P =7 Hz, CH, Ph), 127.8 (s, CH; Ph), 70.1 (s, 4C, α-THF), 25.7
(s, 4C, β-THF), 2.3 (s, 6C, NSi(CH3)3).
31P NMR (161.97 MHz, C6D6, 298 K): δ À4.96 (d, 2 P, 3JP–P = 109
3
Hz), À10.2 (t, 1 P, JP–P = 109 Hz).
Elemental analysis (%) calcd for C50H62CaNO2P3Si2: C, 66.86; H,
6.96; N, 1.56; found: C 66.93, H 6.81; N, 1.52.
Synthesis of [(o-C6H4PPh2)2PMgnBu (THF)2 (2)
A benzene/THF solution (3 ml/1 ml) of Ph―PPP―H ligand
(0.518 g, 0.9 mmol) was added slowly to a solution (0.9 ml
of a 1 M solution in heptane) of MgnBu2 0.1 M, to give a
deep-orange solution. After stirring for 60 min, all volatiles
were removed under vacuum, and the residue red solid was
washed with cold hexane, and then dried up at reduced pres-
sure for 3 h (yield 0.820 g, 93%). The crude solid was
recrystallized from hexane at À20°C to obtain an orange crys-
talline powder.
The metal precursor Ca[N(SiMe3)2]2.THF2 was prepared as
described elsewhere.[57,62]
Instruments and Measurements
NMR spectra were recorded on Bruker Avance 400 spectrometer
(1H, 400.00 MHz; 13C, 100.62 MHz; 31P 161.97 MHz) at 25°C, unless
otherwise stated. Deuterated solvents were purchased from
Cambridge Isotope Laboratories, Inc., and degassed and dried
over activated 3 Å molecular sieves prior to use. Chemical shifts
(δ) are listed as parts per million and coupling constants (J) in
1H NMR (400 MHz, C6D6, 298 K): δ 7.80 (m, 2H, CH phenyl (a)),
3
7.47 (m; 8H, CH, ortho Ph), 7.21 (t, JH–H = 4 Hz; 4H, CH para Ph),
7.10–6.94 (m, 12H, CH, CH para Ph, CH phenyl (b and d)), 6.71
(t, JH–H = 7 Hz; 2H, phenyl (c)), 3.56 (m, 8H, α CH2 THF), 1.30
(m, 8H, β CH2 THF), 1.57 (m, 2H, Mg―CH2 CH2 CH2CH3), 1.12
(t, 3H, Mg―CH2 CH2 CH2CH3), 0.19 (m, 3H, Mg―CH2―CH2―CH2
CH3), À0.12 (t, 2H, Mg―CH2).
1
hertz. H NMR spectra are referenced using the residual solvent
peak at δ 7.16 for C6D6 and δ 7.27 for CDCl3. 13C NMR spectra
are referenced using the residual solvent peak at δ 128.39 for
C6D6 and δ 77.23 for CDCl3. 31P NMR spectra are referenced exter-
nally using 85% H3PO4 at δ = 0.00.
13C{1H} NMR (100.62 MHz, C6D6, 298 K): δ 152.3 (td, C―P,
JC–P = 33 Hz, JC–P = 21 Hz, Cipso―P (phosphido)), 136.2 136.1
136.2 (td, JC–P = 31 Hz, JC–P = 20 Hz, Cipso―PPh2), 134.6, 132.1,
130.4, 129.9, 128.7, 126.0, 67.9 (s, 4C, α-THF), 31.6
(Mg―CH2―CH2―), 31.4 (Mg―CH2―CH2―CH2), 25.4 (s, 4C, β-
THF), 14.2 (Mg―CH2―CH2―CH2―CH3), 7.8 (Mg―CH2).
Molecular weights (Mn and Mw) and the molecular mass distri-
bution (Mw/Mn) of polymer samples were measured by GPC at
30°C, using THF as solvent, flow rate of eluent 1 ml minÀ1 and
narrow polystyrene standards as reference. The measurements
were performed on a Waters 1525 binary system equipped with
a Waters 2414 RI detector using four Styragel columns (range
1000–1 000 000 Å). Every value was the average of two
31P NMR (161.97 MHz, C6D6, 298 K): δ À10.7 (d, 2 P, 3JP–P = 147
3
Hz), À41.4 (t, 1 P, JP–P = 147 Hz).
Elemental analysis (%) calcd for C50H62MgNO2P3Si2: C, 68.05; H,
7.08; N, 1.59; found: C 67.92, H 6.91; N, 1.53.
wileyonlinelibrary.com/journal/aoc
Copyright © 2014 John Wiley & Sons, Ltd.
Appl. Organometal. Chem. 2014, 28, 140–145

ROP of cyclic esters by group II metal complexes
[19] C. K. Williams, L. E. Breyfogle, S. K. Choi, W. Nam, V. G. Young Jr., M. A.
Hillmyer, W. B. Tolman, J. Am. Chem. Soc. 2003, 125, 11350.
[20] M. H. Chisholm, J. C. Gallucci, K. Phomphrai, Inorg. Chem. 2005, 44, 8004.
[21] M. H. Chisholm, N. W. Eilerts, J. C. Huffman, S. S. Iyer, M. Pacold,
K. Phomphrai, J. Am. Chem. Soc. 2000, 122, 11845.
[22] H.-Y. Tang, H.-Y. Chen, J.-H. Huang, C.-C. Lin, Macromolecules 2007,
40, 8855.
[23] E. L. Marshall, V. C. Gibson, H. S. Rzepa, J. Am. Chem. Soc. 2005,
127, 6048.
[24] T.-L. Yu, C.-C. Wu, C.-C. Chen, B.-H. Huang, J. Wu, C.-C. Lin, Polymer
2005, 46, 5909.
Polymerization Experiments
General procedures for the polymerization of caprolactone and lactides
In a Braun Labmaster glovebox, a 4 ml vial was charged with a
solution of metal initiator (5.0 μmol) in the appropriate solvent
(0.5 ml), to which the alcohol was eventually added, and the
resulting solution was rapidly added to a solution of the mono-
mer in the same solvent (1.5 ml). The mixture was immediately
stirred with a magnetic stir bar at 25°C for the predicted time.
After specified time intervals an aliquot of the crude material
was sampled by pipette and quenched in wet CDCl3. The sample
was subjected to monomer conversion determination, which was
monitored by integration of monomer versus polymer in 1H NMR
spectra. The reaction mixture was quenched with wet n-hexane.
The precipitates collected from the bulk mixture were dried
in air, dissolved with dichloromethane and sequentially precip-
itated into methanol. The obtained polymer was collected by
filtration and further dried in a vacuum oven at 60°C for 16 h.
[25] M.-L. Shueh, Y.-S. Wang, B.-H. Huang, C.-Y. Kuo, C.-C. Lin, Macromole-
cules 2004, 37, 5155.
[26] M. H. Chisholm, J. Gallucci, K. Phomphrai, Inorg. Chem. 2002, 41, 2785.
[27] T. P. Hanusa, Chem. Rev. 1993, 93, 1023.
[28] T. P. Hanusa, Coord. Chem. Rev. 2000, 210, 329.
[29] J. S. Alexander, K. Ruhlandt-Senge, Eur. J. Inorg. Chem. 2002, 2761.
[30] W. D. Buchanan, D. G. Allis, K. Ruhlandt-Senge, Chem. Commun.
2010, 46, 4449.
[31] Z. Zhong, S. Schneiderbauer, P. J. Dijkstra, M. Westerhausen, J. Feijen,
Polym. Bull. 2003, 51, 175.
[32] M. H. Chisholm, D. Navarro-Llobet, W. J. J. Simonsick, Macromolecules
2001, 34, 8851.
[33] H. Chen, H. Tang, C. Lin, Polymer 2007, 48, 2257.
[34] M.-W. Hsiao, C.-C. Lin, Dalton Trans. 2013, 42, 2041.
[35] C.-T. Chen, T.-L. Huang, Dalton Trans. 2013, 42, 12612.
[36] M. Bouyhayi, Y. Sarazin, O. L. Casagrande, J.-F. Carpentier, Appl.
Organomet. Chem. 2012, 26, 681.
[37] M. H. Chisholm, J. Gallucci, K. Phomphrai, Chem. Commun. 2003, 48.
[38] M. H. Chisholm, J. Gallucci, K. Phomphrai, Inorg. Chem. 2004, 43, 6717.
[39] M. H. Chisholm, J. C. Gallucci, G. Yaman, Dalton Trans. 2009, 368.
[40] S. Datta, M. T. Garner, P. W. Roesky, Dalton Trans. 2008, 2839.
[41] M. R. Crimmin, I. J. Casely, M. S. Hill, J. Am. Chem. Soc. 2005, 127, 2042.
[42] C. Ruspic, S. Nembenna, A. Hofmeister, J. Magull, S. Harder, H. W.
Roesky, J. Am. Chem. Soc. 2006, 15000.
Kinetic experiments
In a typical experiment carried out in a Braun Labmaster
glovebox, the initiator solution (4 μmol in 0.5 ml of C7D8) was
injected into a Teflon-valved J. Young NMR tube loaded with
the solid monomer (100 equiv.). The sample was thermostated
at the required temperature. The polymerization reaction was
1
monitored via H NMR analysis from the integration of polymer
and monomer signals. The characteristic chemical shift for lactide
monomer in deuterated toluene is 4.12 ppm (q, CH; lactide); for
the corresponding polymer it is 5.12 ppm (q, CH; polylactide).
[43] S. Harder, J. Brettar, Angew. Chem. Int. Ed. 2006, 45, 3474.
[44] D. J. Darensbourg, W. Choi, C. P. Richers, Macromolecules 2007,
40, 3521.
[45] D. J. Darensbourg, W. Choi, P. Ganguly, C. P. Richers, Macromolecules
2006, 39, 4374.
Acknowledgments
[46] V. Poirier, T. Roisnel, J.-F. Carpentier, Y. Sarazin, Dalton Trans.
2009, 9820.
[47] Y. Sarazin, D. Rosca, V. Poirier, T. Roisnel, A. Silvestru, L. Maron, J.-F.
Carpentier, Organometallics 2010, 29, 6569.
[48] Y. Sarazin, B. Liu, T. Roisnel, L. Maron, J.-F. Carpentier, J. Am. Chem.
Soc. 2011, 133, 9069.
The authors thank Dr Patrizia Oliva for NMR assistance, Dr Patrizia
Iannece for elemental analyses and Dr Ilaria D’Auria for GPC
measurements.
[49] M. Mazzeo, M. Lamberti, I. D’Auria, S. Milione, J. C. Peters, C.
Pellecchia, J. Polym. Sci. Part A 2010, 48, 1374.
[50] I. D’Auria, M. Mazzeo, D. Pappalardo, M. Lamberti, C. Pellecchia,
References
[1] K. E. Uhrich, S. M. Cannizzaro, R. S. Langer, K. M. Shakesheff, Chem.
Rev. 1999, 99, 3181.
J. Polym. Sci. Part A 2011, 49, 403.
[2] M. Martina, D. W. Hutmacher, Polym. Int. 2007, 56, 145.
[3] K. M. Stridsberg, M. Ryner, A.-C. Albertsson, Advances in Polymer
Science, Vol. 157, Springer, Heidelberg, 2002.
[51] I. D’Auria, M. Lamberti, M. Mazzeo, S. Milione, C. Pellecchia, J. Polym.
Sci. Part A 2013. doi:10.1002/pola.26969.
[52] I. D’Auria, M. Lamberti, M. Mazzeo, S. Milione, G. Roviello, C.
[4] R. H. Platel, L. M. Hodgson, C. K. Williams, Polymer Rev. 2008, 48, 11.
[5] A.-C. Albertsson, I. K. Varma, Biomacromolecules 2003, 4, 1466.
[6] M. H. Chisholm, Pure Appl. Chem. 2010, 82, 1647.
[7] J. Wu, T.-L. Yu, C.-T. Chen, C.-C. Lin, Coord. Chem. Rev. 2006, 250, 602.
[8] A. Sauer, A. Kapelski, C. Fliedel, S. Dagorne, M. Kol, J. Okuda, Dalton
Trans. 2013, 42, 9007.
[9] A. L. Zelikoff, J. Kopilov, I. Goldberg, G. W. Coates, M. Kol, Chem.
Commun. 2009, 6804.
[10] E. Sergeeva, J. Kopilov, I. Goldberg, M. Kol, Inorg. Chem. 2010,
49, 3977.
Pellecchia, Chem. Eur. J. 2012, 18, 2349.
[53] M. T. Whited, E. Rivard, J. C. Peters, Chem. Commun. 2006, 1613.
[54] M. Gärtner, H. Görls, M. Westerhausen, Inorg. Chem. 2008, 47, 1397.
[55] R. G. Pearson, Chemical Hardness, Wiley-VCH, Weinheim, 1997.
[56] S.-M. Ho, C.-S. Hsiao, A. Datta, C.-H. Hung, L.-C. Chang, T.-Y. Lee, J.-H.
Huang, Inorg. Chem. 2009, 48, 8004.
[57] Y. Sarazin, R. H. Howard, D. L. Hughes, S. M. Humphrey, M.
Bochmann, Dalton Trans. 2006, 340.
[58] B. Liu, T. Roisnel, J.-P. Guegan, J.-F. Carpentier, Y. Sarazin, Chem. Eur. J.
2012, 18, 6289.
[11] J.-C. Buffet, A. N. Martin, M. Kol, J. Okuda, Polym. Chem. 2011, 2, 2378.
[12] A. Stopper, J. Okuda, M. Kol, Macromolecules 2012, 45, 698.
[13] S. Gendler, S. Segal, I. Goldberg, Z. Goldschmidt, M. Kol, Inorg. Chem.
2006, 45, 4783.
[14] B. J. O’Keefe, S. M. Monnier, M. A. Hillmyer, W. B. Tolman, J. Chem.
Soc. Dalton Trans. 2001, 2215.
[59] M. H. Chisholm, K. Choojun, J. C. Gallucci, P. M. Wambua, Chem. Sci.
2012, 3, 3445.
[60] Z. Zhong, P. J. Dijkstra, C. Birg, M. Westerhausen, J. Feijen, Macromol-
ecules 2001, 34, 3863.
[61] K. Phomphrai, P. Chumsaeng, P. Sangtrirutnugul, P. Kongsaeree, M.
Pohmakotr, Dalton Trans. 2010, 39, 1865.
[15] C. A. Wheaton, P. G. Hayes, B. J. Ireland, Dalton Trans. 2009, 4832.
[16] H.-Y. Chen, H.-Y. Tang, C.-C. Lin, Macromolecules 2006, 39, 3745.
[17] B. M. Chamberlain, M. Cheng, D. R. Moore, T. M. Ovitt, E. B. Lobkovsky,
G. W. Coates, J. Am. Chem. Soc. 2001, 123, 3229.
[18] B.-H. Huang, C.-N. Lin, M.-L. Hsueh, T.-L. Yu, T. Athar, C.-C. Lin, Polymer
2006, 47, 6662.
[62] N. Ajellal, J.-F. Carpentier, C. Guillaume, S. M. Guillaume, M. Helou, V.
Poirier, Y. Sarazin, A. Trifonov, Dalton Trans. 2010, 39, 8363.
[63] I. Palard, A. Soum, S. M. Guillaume, Macromolecules 2005, 38, 6888.
[64] M. Save, M. Schappacher, A. Soum, Macromol. Chem. Phys. 2002, 203, 889.
[65] I. Barakat, P. Dubois, R. Jerome, P. Teyssié, J. Polym. Sci. Part A 1993,
31, 505.
Appl. Organometal. Chem. 2014, 28, 140–145
Copyright © 2014 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/aoc