Bridged bis(amidinate) ytterbium alkoxide and phenoxide: Syntheses, structures, and their high activity for controlled polymerization of L-lactide and ε-caprolactone

Article abstract of DOI:10.1021/ic801189j

Bridged bis(amidinate) ytterbium aikoxide and phenoxide with diverse molecular structures were synthesized in high yields and confirmed by X-ray crystal structural analysis. The reaction of LYbCI(THF)₂ (L = Me ₃SiNC(Ph)N(CH₂

Full text of DOI:10.1021/ic801189j

Inorg. Chem. 2009, 48, 744-751
Bridged Bis(amidinate) Ytterbium Alkoxide and Phenoxide: Syntheses,
Structures, and Their High Activity for Controlled Polymerization of
L-Lactide and ε-Caprolactone
Junfeng Wang, Yingming Yao, Yong Zhang, and Qi Shen*
Key Laboratory of Organic Synthesis Jiangsu ProVince, Chemistry and Chemical Engineering of
Suzhou UniVersity, Suzhou, 215123, China
Received June 30, 2008
Bridged bis(amidinate) ytterbium alkoxide and phenoxide with diverse molecular structures were synthesized in
high yields and confirmed by X-ray crystal structural analysis. The reaction of LYbCl(THF)2 (L )
Me3SiNC(Ph)N(CH2)3NC(Ph)NSiMe3) with 1 equiv of NaOAr (ArO ) 2,6-diisopropylphenoxo) afforded the mononuclear
complex LYb(OAr)DME 1 with a seven-coordinated ytterbium atom surrrounded by one chelating bis(amidinate)
ligand, one phenoxo group, and one DME (dimethoxyethane) molecule. The same reaction with 1 equiv of NaOiPr
yielded the binuclear complex Yb(µ2-L)2(µ2-OiPr)2Yb, 2, with two equivalent six-coordinate metal centers connected
by two linked bis(amidinate)s and two OiPr bridges formed via a ligand redistribution reaction that occurred during
the metathesis reaction. Both 1 and 2 initiated the ring-opening polymerization of L-lactide, as well as ε-caprolactone
(ε-CL), in a controlled manner with high reactivity, as indicated by a linear relationship between Mn and conversion
and by narrow molecular weight distributions (PDI ) 1.15-1.25) up to 100% conversion. The differences in catalytic
performance between complexes 1 and 2 are discussed.
Introduction
containing tetradentated bis(phenolate) ligands are of par-
ticular interest for this aim. Diverse lanthanide alkoxides and
other derivatives bearing these “ONNO”, “ONOO”, or
“OSSO” ligands allow provision of the polymerization in a
living fashion5-7 and even production of the polylactides
(PLAs) with high stereoselectivity.5,6
There is continuing interest in the synthesis of aliphatic
polyesters, particularly, of polylactide, due to their biode-
gradable and biocompatible characteristics and the renewable
nature of their feedstock.1 Controlled ring-opening polym-
erization of lactones is a convenient route for the synthesis
of aliphatic polyesters with predictable molecular weight,
low polydispersity indices, and control over end groups.
Lanthanide alkoxides are known to be highly active initiators.
However, not all of the alkoxides are capable of initiating
the polymerization in a living fashion.2 Consequently,
significant effort over the past few decades has been placed
toward the design and synthesis of well-characterized single-
site lanthanide alkoxides by use of an appropriate ancillary
ligand to improve their performance. Various ligand sets,
such as bulky cyclopentadienyl3 and noncyclopentadienyl
ligands,4-7 have been developed. Among them, heteroatom-
Bridged bis(amidinate) dianions can be viewed as another
kind of noncyclopentadienyl tetradentated “NNNN” ligand,
(2) (a) Giesbrecht, G. R.; Whitener, G. D.; Arnold, J. J. Chem. Soc., Dalton
Trans. 2001, 923. (b) Stevels, W. M.; Ankone´, M. J. K.; Dijkstra,
P. J.; Feijen, J. Macromolecules 1996, 29, 6132. (c) Stevels, M. W.;
Ankone, M. J. K.; Dijkstra, P. J.; Feijen, J. Macromolecules 1996,
29, 8296. (d) Akatsuka, M.; Aida, T.; Inoue, S. Macromolecules 1995,
28, 1320. (e) Nishiura, M.; Hou, Z. M.; Koizumi, T.; Imamoto, T.;
Wakatsuki, Y. Macromolecules 1999, 32, 8245. (f) Shen, Z. Q.; Chen,
X. H.; Shen, Y. Q.; Zhang, Y. F. J. Polym. Sci., Part A: Polym. Chem.
1994, 32, 597. (g) Shen, Y. Q.; Shen, Z. Q.; Zhang, Y. F.; Yao, K. M.
Macromolecules 1996, 29, 8289. (h) Shen, Y. Q.; Shen, Z. Q.; Zhang,
Y. F.; Yao, K. M. Macromolecules 1996, 29, 3441. (i) Shen, Y. Q.;
Shen, Z. Q.; Zhang, F. Y.; Zhang, Y. F. J. Appl. Polym. Sci. 1997,
64, 2131. (j) Ling, J.; Shen, Z. Q. Macromolecules 2001, 34, 7613.
(k) Ling, J.; Zhu, W. P.; Shen, Z. Q. Macromolecules 2004, 37, 758.
(l) Yao, Y.; Ma, M.; Xu, X.; Zhang, Y.; Shen, Q.; Wong, W.-T.
Organometallics 2005, 24, 4014. (m) Sheng, H.; Zhou, L.; Zhang,
Y.; Yao, Y.; Shen, Q. J. Polym. Sci., Part A: Polym. Chem. 2007, 45,
1210. (n) Kerton, F. M.; Whitwood, A. C.; Willans, C. E. Dalton
Trans. 2004, 2237. (o) Yu, C.; Zhang, L.; Ni, X.; Shen, Z.; Tu, K. J.
Polym. Sci., Part A: Polym. Chem. 2004, 42, 6209.
* Author to whom correspondence should be addressed. E-mail:
qshen@suda.edu.cn.
(1) (a) Okada, M Prog. Polym. Sci. 2002, 27, 87. (b) Dechy-Cabaret, O.;
Martin-Vaca, B.; Bourissou, D. Chem. ReV. 2004, 104, 6147. (c) Wu,
J.; Yu, T.-L.; Chen, C.-T.; Lin, C.-C. Coord. Chem. ReV. 2006, 250,
602. (d) O’Keefe, B. J.; Hillmyer, M. A.; Tolman, W. B. J. Chem.
Soc., Dalton Trans. 2001, 2215.
744 Inorganic Chemistry, Vol. 48, No. 2, 2009
10.1021/ic801189j CCC: $40.75  2009 American Chemical Society
Published on Web 12/18/2008

Bridged Bis(amidinate) Ytterbium Alkoxide and Phenoxide
Scheme 1
Scheme 2
which has the advantage of tunable electronic and steric
effects. These tetradentated “NNNN” groups were first
introduced in lanthanide chemistry by Hessen et al. in 2001.8
However, their usage in the design of lanthanide catalysts
as a supporting ligand has not attracted more attention yet,
although alkoxide and amide derivatives of yttrium supported
by unbridged arylamidinated were reported to be active
catalysts for the polymerization of D,L-lactone.9 Very re-
cently, we reported successful catalyst systems of ytterbium
amides supported by a bis(amidinate) L (L ) Me3SiNC-
(Ph)N(CH2)3NC(Ph)NSiMe3) for the controlled polymeriza-
tion of L-lactide.10 Further study revealed that the ytterbium
amide complex LYb(NHAr)(DME) (Ar ) 2,6-iPr2C6H3) can
also serve as an efficient catalyst for the addition of amines
to nitriles affording, selectively, monosubstituted N-aryla-
midinates.11 These results promoted us to design well-defined
lanthanide alkoxides using these ligands for the controlled
polymerization of lactones.
well-characterized binuclear lanthanide alkoxide-catalyzed,
well-controlled polymerization of lactones. Moreover, the
difference in performance between the two complexes is also
discussed.
In this contribution, we describe the synthesis and char-
acterization of a new ytterbium alkoxide and phenoxide with
the bridged bis(amidinate) ligand L, and their high activity
for the controlled ring-opening polymerization of L-lactide,
as well as ε-caprolactone, presenting the first example for a
Results and Discussion
Syntheses of Complexes 1 and 2. The reaction of
LYbCl(THF)2, which was prepared by a published proce-
dure,12 with 1 equiv of NaOAr (OAr ) 2,6-diisopropylphe-
noxo) was conducted in THF. The reaction went smoothly,
and the chlorine can be simply replaced by the OAr group
to give the corresponding aryloxide LYb(OAr)(DME), 1, as
colorless crystals upon the addition of DME (dimethoxyle-
thane) in good yield. The monomeric nature of 1 was
indicated by an X-ray diffraction study (Scheme 1).
In contrast, the reaction of LYbCl(THF)2 with 1 equiv of
NaOiPr afforded the binuclear complex Yb(µ2-L)2(µ2-
OiPr)2Yb, 2, which was confirmed by a crystal structure
determination, instead of the mononuclear complex LY-
b(OiPr), as the case of aryloxide (Scheme 2). The reaction
is reproducible with the yields ranging from 48 to 53%, and
no mononuclear complex LYb(OiPr) has ever been isolated,
indicating that 2 is more thermodynamically stable than
expected for LYb(OiPr). The formation of 2 demonstrated
that a ligand redistribution reaction occurred concomitantly
during the metathesis reaction. Such a ligand redistribution
reaction with a binuclear complex was seen previously for
the related reaction of LYbCl(THF)2 with NaN(TMS)210 and
Li2L.13 Here, the occurrence of a ligand redistribution may
be attributed to the encapsulation of two OiPr bridges, which
makes the complex more stable. The reaction pathway for
the formation of 2 can be supposed as follows: the attack of
NaOiPr results in the cleavage of one of the linked
bis(amidinate)-Yb bonds, followed by redistribution to
another Yb atom.
(3) (a) Yamashita, M.; Takemoto, Y.; Ihara, E.; Yasuda, H. Macromol-
ecules 1996, 29, 1798. (b) Yasuda, H.; Tamai, H. Prog. Polym. Sci.
1993, 18, 1097. (c) Yasuda, H.; Yamamoto, H.; Yokota, K.; Miyake,
S.; Nakamura, A. J. Am. Chem. Soc. 1992, 114, 4908.
(4) (a) Xue, M.; Sun, H.; Zhou, H.; Yao, Y.; Shen, Q.; Zhang, Y. J. Polym.
Sci., Part A: Polym. Chem. 2006, 44, 1147. (b) Yao, Y. M.; Xu, X. P.;
Zhang, Y.; Shen, S.; Wong, W. T. Inorg. Chem. 2005, 44, 5133. (c)
Simic, V.; Spassky, N.; Hubert-Pfalzgraf, L. G. Macromolecules 1997,
30, 7338. (d) Simic, V.; Girardon, V.; Spassky, N.; Hubert-Pfalzgraf,
L. G.; Duda, A. Polym. Degrad. Stab. 1998, 59, 227. (e) Save, M.;
Schappacher, M. le; Soum, A. Macromol. Chem. Phys. 2002, 203,
889. (f) Spassky, N.; Simic, V.; Montaudo, M. S.; Hubert-Pfalzgraf,
L. G. Macromol. Chem. Phys. 2000, 201, 2432. (g) Agarwal, S.; Mast,
C.; Dehnike, K.; Greiner, A. Macromol. Rapid Commun. 2000, 21,
195. (h) Chamberlain, B. M.; Jazdzewski, B. A.; Pink, M.; Hillmyer,
M. A.; Tolman, W. B. Macromolecules 2000, 33, 3970. (i) Stevels,
W. M.; Ankone´, M. J. K.; Dijkstra, P. J.; Feijen, J. Macromolecules
1996, 29, 3332. (j) Alaaeddine, A.; Amgoune, A.; Thomas, C. M.;
Dagorne, S.; Bellemin-Laonnaz, S.; Carpentier, J.-F. Eur. J. Inorg.
Chem. 2006, 3652.
(5) (a) Ovitt, T. M.; Coates, G. W. J. Am. Chem. Soc. 1999, 121, 4072.
(b) Amgoune, A.; Thomas, C. M.; Roisnel, T.; Carpentier, J.-F.
Chem.sEur. J. 2006, 12, 169. (c) Ma, H.; Spaniol, T. P.; Okuda, J.
Inorg. Chem. 2008, 47, 3328. (d) Yao, Y. M.; Xu, X. P.; Zhang, Y.;
Shen, Q.; Wong, W. T. Inorg. Chem. 2005, 44, 5133.
(6) (a) Cai, C.-X.; Amgoune, A.; Lehmann, C. W.; Carpentier, J.-F.
Chem.Commun. 2004, 330. (b) Ma, H.; Spaniol, T. P.; Okuda, J.
Angew. Chem., Int. Ed. 2006, 118, 7982. (c) Bonnet, F.; Cowley, A. R.;
Mountford, P. Inorg. Chem. 2005, 44, 9046. (d) Amgoune, A.; Thomas,
C. M.; Hinca, S.; Roisnel, T.; Carpentier, J.-F. Angew. Chem., Int.
Ed. 200645, 2782. (e) Amgoune, A.; Thomas, C. M.; Carpentier, J.-
F. Macromol. Rapid Commun. 2007, 28, 693. (f) Ma, H.; Spaniol,
T. P.; Okuda, J. Angew. Chem., Int. Ed. 2006, 45, 7818.
(7) (a) Liu, X.; Shang, X.; Tang, T.; Hu, N.; Pei, F.; Cui, D.; Chen, X.;
Jing, X. Organometallics 2007, 26, 2747. (b) Kerton, F. M.; Whitwood,
A. C.; Willams, C. E. Dalton Trans. 2004, 2237. (c) Ma, H.; Okuda,
J. Macromolecules 2005, 38, 2665. (d) Ma, H.; Spaniol, T. P.; Okuda,
J. Dalton Trans. 2003, 4770. (e) Ma, H.; Melillo, G.; Oliva, L.; Okuda,
J. Dalton Trans. 2005, 721. (f) Schaverien, C. J.; Meijboom, N.; Orpen,
A. G. J. Chem. Soc., Chem. Commun. 1992, 124. (g) Gribkov, D. V.;
Hultzch, K. C.; Hampel, F. Chem.sEur. J. 2003, 9, 4796.
(8) Bambirra, S.; Meetsma, A.; Hessen, B.; Teuben, J. H. Organometallics
2001, 20, 782.
Both complexes 1 and 2 are soluble in THF, DME, and
even in toluene. A strong absorption of the CdN stretch at
approximately 1640 cm-1 was observed for both complexes
in their IR spectra, indicating the delocalized double bond
(9) Aubrecht, K. B.; Chang, K.; Hillmyer, M. A.; Tolman, W. B. J. Polym.
Sci., Part A: Polym. Chem. 2001, 39, 284.
(10) Wang, J.; Cai, T.; Yao, Y.; Zhang, Y.; Shen, Q. Dalton Trans. 2007,
5275.
(12) Wang, J. F.; Yao, Y. M.; Cheng, J. L.; Pang, X. A.; Zhang, Y.; Shen,
Q. J. Mol. Struct. 2005, 743, 229.
(13) Wang, J.; Sun, H.; Yao, Y.; Zhang, Y.; Shen, Q. Polyhedron 2008,
27, 1977.
(11) Wang, J.; Xu, F.; Cai, T.; Shen, Q. Org. Lett. 2008, 10, 445.
Inorganic Chemistry, Vol. 48, No. 2, 2009 745

Wang et al.
Figure 1. Molecular structure of complex 1 drawn with 20% thermal
ellipsoids. Hydrogen atoms are omitted for clarity.
Table 1. Selected Bond Lengths and Bond Angles in 1
Yb(1)-N(1)
2.485(4)
2.283(5)
2.477(4)
2.296(5)
2.805(5)
2.075(3)
55.66(16)
150.75(14)
Yb(1)-C(2)
N(1)-C(1)
N(2)-C(1)
N(3)-C(2)
N(4)-C(2)
2.782(6)
1.326(7)
1.309(7)
1.326(7)
1.331(8)
Yb(1)-N(2)
Figure 2. Molecular structure of complex 2 drawn with 30% thermal
ellipsoids. Hydrogen atoms are omitted for clarity.
Yb(1)-N(3)
Yb(1)-N(4)
Table 2. Selected Bond Lengths and Bond Angles in 2
Yb(1)-C(1)
Yb(1)-O(1)
Yb(1)-N(1)
2.337(3) Yb(1A)-N(3)
2.346(3) Yb(1A)-N(4)
2.352(3) Yb(1A)-N(1A)
2.304(3) Yb(1A)-N(2A)
2.212(2) Yb(1A)-O(1)
2.186(2) Yb(1A)-O(1A)
2.352(3)
2.304(3)
2.337(3)
2.346(3)
2.186(2)
2.212(2)
N(2)-Yb(1)-N(1)
O(1)-Yb(1)-O(2)
N(4)-Yb(1)-N(3)
56.01(17)
Yb(1)-N(2)
Yb(1)-N(3A)
Yb(1)-N(4A)
Yb(1)-O(1)
Yb(1)-O(1A)
N(1)-Yb(1)-N(2)
of N-C-N linkage.14 The acceptable NMR spectra for them
are difficult to obtain because of the paramagnetic behavior
of ytterbium.
57.88(10) N(1A)-Yb(1A)-N(2A) 57.88(10)
N(4A)-Yb(1)-N(3A) 57.83(10) N(4)-Yb(1A)-N(3)
57.83(10)
X-Ray Crystal Structures of Complexes 1 and 2. Single
crystals of 1 and 2 suitable for X-ray diffraction were
obtained by cooling a mixture of DME and toluene. The
X-ray diffraction analysis confirmed complex 1 being a
monomer (Figure 1). The selected bond lengths and angles
are listed in Table 1. The central metal coordinated to one
chelating bis(amidinate) ligand and three oxygen atoms from
one -OAr group and one DME molecule, forming a distorted
trigonal bipyramid, if each amidinate ligand is considered
to occupy one coordination vertex. Each center of the two
amidinate groups and one oxygen atom (O3) occupy equato-
rial positions, while two oxygen atoms (O1 and O2) locate
on axial sites. The coordination geometry around the central
metal in 1 is similar to those of the related alkyl,8 amide,10
chloride, and cyclopentadienyl derivatives12 reported previ-
ously. The bond parameters within the two amidinate ligands
are not dramatically different. But there is a substantial
difference in Yb-N(amidinate) distance for the amidinate
nitrogens attached to the bridge compared to those attached
to the SiMe3 group (the former being 0.202 Å and 0.181 Å
shorter), which is similar to those found in the related
derivatives.8,10,12 The bond parameters within Yb-N-C-N
fall in the normal range of those found in the published
complexes. The C-N bond distances within the chelating
N-C-N unit are nearly equal, and the average value, 1.33
Å, reflected the delocalization of the π bond in the N-C-N
unit. Because no aryloxide analogous with the related ligand
was reported, the bond length of Yb-O(OAr) can only be
compared with those of other lanthanide aryloxide com-
plexes. The Ln-O(OAr) bond length of 2.075(3) Å in 1 is
somewhat shorter than the corresponding values observed
in L1Yb (OAr)Cl(THF) [L1 ) N,N′-bis(2,6-dimethylphenyl)-
2,4-pentanediiminato, OAr ) 2,6-ditert-butylphenoxo-],
2.039(3) Å,4a and unbridged bis(amidinate) aryloxide
t
L2TMSY(OC6H2 Bu2Me) (LTMS ) bis[N,N′-bis(trimethylsilyl-
)benzamidinate]), 2.075(2) Å,9 when taking into account the
difference in ionic radii of the metals and the different
coordination numbers of the Yb complexes.
The molecular structure of 2 is shown in Figure 2. The
selected bond lengths and angles are listed in Table 2.
Complex 2 has a centro-symmetric bimetallic structure in
which the coordination sphere around each Yb center is
composed of two amidinate groups from two bridged
bis(amidinate) ligands and two bridged oxygen atoms from
OiPr groups forming a distorted tetrahedron, if each amidinate
ligand is considered to occupy a single coordination vertex.
In the Yb2O2 core, the two bridging oxygen atoms and two
lanthanide atoms are exactly coplanar, as required by the
crystallographic symmetry. The dihedral angle of the Yb2O2
plane and the plane defined by Yb1-N2-N4-Yb1A equals
84.10°. Two OiPr groups are unsymmetrically coordinated
to the central metals, but the bond lengths of Yb-OiPr are
equivalent within the error limits. The mean bond length of
Yb-OiPr is 2.199(2) Å, which is comparable with the
values reported in [(C5H5)2Lu(µ-OCH2CH2CH2CH2Ph)]2,15a
[(Me3CC5H4)2Ce(µ-OiPr)]2,15b (C5H5)2Yb(µ-OiPr)]2,15c and
[(MBMP)Yb(µ-OiPr2)(THF)2]2 [MBMP ) 2,2-methylenebis(6-
(15) (a) Stults, S. D.; Anderson, R. A.; Zalkin, A. Organometallics 1990,
9, 1623. (b) Schumann, H.; Palamidis, E.; Loebel, J. J. Organomet.
Chem. 1990, 384, C49. (c) Wu, Z. Z.; Xu, Z.; You, X. Z.; Zhou, X. G.;
Jin, Z. S. Polyhedron 1992, 11, 2673.
(14) Giesbrecht, G. R.; Sharfir, A.; Arnold, J. Dalton Trans. 1999, 3601.
746 Inorganic Chemistry, Vol. 48, No. 2, 2009

Bridged Bis(amidinate) Ytterbium Alkoxide and Phenoxide
Table 3. Polymerization of L-Lactide Initiated by 1 and 2a
run
init.
temp. (°C)
[M]0/[I]0
t (min)
conv. (%)b
Mn (104)c
Mn (104)d
Mn (104)e
PDI
1.19
1
1
1
1
1
2
2
2
2
2
2
2
70
70
70
90
70
70
70
70
70
200
300
200
200
200
500
800
1500
1500
1500
2000
10
60
10
5
1
1
100
trace
trace
100
100
100
100
100
trace
100
29
7.42
4.30
2.88
2
3f
4
7.35
1.40
3.40
5.80
10.8
4.26
0.81
1.97
3.36
6.26
2.88
1.44
3.61
5.75
10.8
1.22
1.26
1.21
1.22
1.26
5
6
7
5
8
20
20
10
120
9f
10
11
100
100
10.3
4.71
5.97
2.73
10.8
4.18
1.29
1.25
a General polymerization conditions: in toluene, [LLA] ) 1 mol L-1 b Yield: weight of polymer obtained/weight of monomer used. c Measured by GPC
.
relative to polystyrene standards. d Measured by GPC relative to polystyrene standards with Mark-Houwink corrections17 for Mn [Mn(obsd) ) 0.58Mn(GPC)].
e Calculated value ) 144.13 × [M]0/[I]0 × conv. (for initiator 1), and ) 144.13 × [M]0/[I]0 × 0.5 × conv. (for initiator 2). f In THF.
tert-butyl-4-methylphenoxo)]4b when the difference in ionic
radii of the metals and the different coordination number of
these metals are considered. It is worth noticing that the
distance of Yb-Yb is 3.45 Å in complex 2, which is much
shorter than the 6.54 Å in [YbN(TMS)2]2(µ-L)2 that has been
published.10 Such a short distance may be attributed to the
existence of two oxygen bridges. However, the value is
longer than that estimated from twice the single bond radius
of the Yb atom. In the present spanning ligation, the bonding
parameters within the amidinate ligands are comparable to
those found in [YbN(TMS)2]2(µ-L)2.10
Polymerization of L-Lactide Initiated by 1 and 2. The
polymerization of L-lactide with complexes 1 and 2 was
examined. The results obtained under various conditions are
listed in Table 3. Both complexes were found to be efficient
initiators to give a complete conversion in a short time in
toluene at 70 °C. The complex with alkoxide group 2 is more
active than the complex with a bulky phenoxide group 1, as
isopropoxide is usually a better initiating group than the
phenoxide group.4i For example, complete conversion of 200
equiv of monomer was achieved in only 1 min for 2 (Table
3, entry 5), but in 10 min for 1 (Table 3, entry 1).
Furthermore, the yield reached still as high as 100% when
the molar ratio of monomer to initiator (M/I) increased to
1500 for 2 (Table 3, entry 8). To the best of our knowledge,
such a high reactivity for lactide polymerization shown by
a well-defined lanthanide alkoxide has never been reported
yet. However, polymerization with 1 went sluggishly, when
M/I increased to 300 (Table 3, entry 2). The experiment was
reproducible. Such a dramatic decrease in activity with the
increasing of M/I from 200 to 300 for 1 (the same situation
was also found for the system with 2 (entries 10 and 11))
may be attributed to the existence of a trace of lactide acid
contaminated in the monomer and moisture in the solvent,
since 1 and 2 are highly sensitive to any proton source. Thus,
the catalyst will be decomposed if the loading of the catalyst
is too low to tolerate this impurity.
because of competitive coordination of the coordinating
solvent.16
All of the PLAs obtained with complexes 1 and 2 showed
monomodal GPC traces with narrow polydispersity values
(Mw/Mn ) 1.15-1.29). In the case of 2, the experimental
(uncorrected) number-average molecular mass (Mn) values
are very close to the theoretical ones (calculated on the
assumption that each OiPr group initiates the polymerization),
and the corrected values with Mark-Houwink corrections17
are proportionately lower (Figure 3a). In addition, the larger
experimental values for both uncorrected and corrected Mn
related to calculated ones were observed for the case of 1
(Figure 3b). Most likely it is because some transfer occurred
in the system with 2 and both the difficulty of initiation by
the hindered aryloxide and some transfer reaction for 1.
Although there is deviation in Mn between corrected experi-
mental and calculated values for both systems, the polym-
erization occurs in a living fashion over the entire conversion
range, as confirmed by narrow polydispersities of the
polymers. The same situation was also found for the lactide
polymerization with the catalysts published.6c,7c The living
character was further confirmed by a proportional increase
in number averaged molecular weight (Mw) with the ratio
of monomer to initiator for complex 2 (Table 3, entries 5-8)
and a linear increase in Mn with monomer conversion
recorded with both 1 and 2, while the molecular weight
distributions (Mw/Mn) of the resulting polymers were kept
narrow (Mw/Mn e 1.29) and intact, as shown in Figure 3.
Polymerization kinetics at 70 °C in toluene with a molar
ratio of monomer/initiator of 200 for 1 and 1500 for 2 were
further measured. The pseudo first-order kinetic plots of
ln([M]0/[M]) versus polymerization time were observed for
both systems, as shown in Figure 4. The linear semi
logarithmic plots indicate that the concentration of the
catalytic active species kept constant throughout the polym-
erization.
It is worth noticing that the polydispersity index of the
resulting polymers keeps narrow and almost unchanging,
even when the polymerization temperature is raised from
70 to 100 °C (Table 3, entries 1 and 4; 8 and 10).
Prereacting phenoxide complex 1 with 2-propanol, to
generate in situ the corresponding isopropoxide complex, was
The polymerizations with 1 and 2 proceeded much more
slowly in THF than in toluene; only trace polymers were
obtained in THF (Table 3, entries 3 and 9), presumably
(16) Heck, R.; Schulz, E.; Collin, J.; Carpentier, J.-F. J. Mol. Catal. A 2007,
268, 163.
(17) Kowalski, A.; Duda, A.; Penczek, S. Macromolecules 1998, 31, 2114.
Inorganic Chemistry, Vol. 48, No. 2, 2009 747

Wang et al.
Figure 3. Mn and Mw/Mn versus conversion plots for PLAs obtained
using complexes 1 (a) and 2 (b). Conditions: [M]0/[I]0 ) 200 (for
complex 1) and [M]0/[I]0 ) 1500 (for complex 2), [LA]0 ) 1 mol/L, 70
°C, toluene.
Figure 4. Polymerization time vs ln[M]0/[M] plots for the polymerization
of L-lactide initiated with complexes 1 (a) and 2 (b). Conditions: [M]0/[I]0
) 200 (for complex 1), [M]0/[I]0 ) 1500 (for complex 2), [LA]0 ) 1 mol/
L, 70 °C toluene.
tried since the treatment of amide complexes with 2-propanol
is well-known to be a useful strategy for improving the
reactivity and control of polymerization. Surprisingly, the
pretreatment led to the deactivation of 1, and almost no
polymerization was observed under the same conditions. The
reason for it may be that 1 was decomposed by free
2-propanol via the protonolysis of the amidinate-metal bond.
Attempts to isolate the product from the reaction of 1 with
2-propanol failed.
Polymerization of ε-CL Initiated by 1 and 2. The ring-
opening polymerizations of ε-CL were examined by usage
of 1 and 2 as initiators. The representative results are listed
in Table 4. For both catalysts, the polymerization depends
upon the solvent greatly, and the activity in toluene is much
higher than that in THF (Table 4, entries 3 and 4, and 7 and
9).15 As shown in Table 4, the polymerization with 1
proceeded rapidly at room temperature in toluene: complete
conversions of 800 equivalents of ε-CL to polymers were
achieved within 20 min, and the resulting polymer with
narrow molecular weight distributions was obtained (Mw/
Mn ) 1.37, Table 4 entry 3). However, the polymerization
with 2 is not so efficient at room temperature, as confirmed
by the low activity and the rather broad polydispersity of
the resulting polymer: only 50% of the conversion of 400
equivalents of the monomer to polymers was reached in 90
min, and the Mw/Mn of the polymer was as high as 1.79
(Table 4, entry 5). Surprisingly, raising the polymerization
temperature to 40 °C led to a dramatic increase in the activity
of 2: complete conversion of as high as 2000 equivalents of
ε-CL was reached within 60 min, and the resulting polyca-
prolactone had a high molecular weight with a narrow
molecular weight distribution of 1.29 (Table 4 entry 8). In
contrast, raising the temperature is not favorable for the
NMR spectroscopy has been used to identify the stereo-
1
sequence in PLAs. H NMR spectra of the PLAs obtained
with 2 showed a nearly ideal quartet for the methine proton
at δ 5.17 ppm and a doublet for the methyl proton at δ 1.58
ppm, indicating no inversion of the configuration of asym-
metric carbon atoms of the monomer under the reaction
conditions.
The end groups of the oligomer have been analyzed by
1H NMR, leading to the conclusion that the PLA chains are
capped by an isopropoxyester (δ 5.07 ppm) at one end and
an alcohol at the other end (δ 4.30 and 2.70 ppm) assigned
to the methine proton,2n,18 which results from the hydrolytic
deactivation of the ytterbium alkoxide growing species.
(18) Stevels, W. M.; Ankone, M. J. K.; Dijkstra, P. J.; Feijen, J. Macromol.
Chem. Phys. 1995, 196, 1153.
748 Inorganic Chemistry, Vol. 48, No. 2, 2009

Bridged Bis(amidinate) Ytterbium Alkoxide and Phenoxide
Table 4. Polymerization of ε-CL initiated by 1 and 2a
run
init.
temp. (°C)
[M]0/[I]0
t (min)
conv. (%)b
Mn (104)c
Mn (104)d
Mn (104)e
PDI
1
2
3
4f
5
6
7
8
9f
1
1
1
1
2
2
2
2
2
40
40
rt
rt
rt
40
40
40
40
1000
1000
800
800
400
1000
1500
2000
1500
90
15
20
20
90
20
50
60
50
100
85
100
trace
50
100
100
100
trace
8.57
7.10
14.5
4.80
3.98
8.12
11.41
9.70
9.13
1.77
1.67
1.37
4.21
10.2
15
1.32
5.71
8.40
2.28
11.41
17.12
22.83
1.79
1.25
1.26
1.29
23.0
12.88
a General polymerization conditions: in toluene. b Yield: weight of polymer obtained/weight of monomer used. c Measured by GPC relative to polystyrene
standards. d Measured by GPC relative to polystyrene standards with Mark-Houwink corrections for Mn [Mn(obsd) ) 0.56Mn(GPC)]. e Calculated value )
114.14 × [M]0/[I]0 × conv. (for initiator 1), and ) 112.17 × [M]0/[I]0 × 0.5 × conv. (for initiator 2). f In THF.
polymerization with 1: the polymerization is out of control
at 40 °C, as confirmed by the fact that the polymers have
much lower molecular weights than the theoretical one and
broad molecular weight distributions of 1.77 and 1.67 (entries
1 and 2). These differences in activity and control of the
polymerization between 1 and 2 may be attributed to a crucial
influence on the catalytic behavior resulting from the
coordinated environment around the central metal. The
difficulty of initiation by the hindered aryloxide group,
resulting in slower initiation in comparison with the propa-
gating reaction and a faster transesterification reaction at 40
°C compared to that at room temperature, may be the reason
for the uncontrolled polymerization of 1 at 40 °C. The
bridged OiPr group was supposed to remain in 2 at room
temperature, and the µ-OiPr group is less active in the
polymerization of ε-CL. Thus, the polymerization with 2
proceeded at room temperature sluggishly. The bridged OiPr
group of 2 was supposed to change to a terminal group under
the attack of the monomer at 40 °C, and a the terminal OiPr
group is highly active. So far, we were not able to isolate
any pure complex, but the suggestion seems reasonable.
Both 1 and 2 were capable of initiating controlled
polymerization of ε-CL under their own optimizing condi-
tions, as shown by linear increases of Mn with conversion
for both 1 and 2 (Figure 5), and also by a proportional
increase of Mn with the monomer-to-initiator ratio for 2
(Table 4, entries 6-8) and narrow molecular weight distribu-
tions. The polymerization kinetics at room temperature in
toluene with an M/I of 800 for 1 and at 40 °C with an M/I
Figure 5. Mn and Mw/Mn vs conversion plots for poly(ε-CL) obtained using
complexes 1 (a) and 2 (b). Conditions: [M]0/[I]0 ) 800, toluene, rt (for
complex 1); [M]0/[I]0 ) 1500, toluene, 40 °C (for complex 2).
of 1500 for 2 gave the pseudo first-order kinetic plots of
ln([M]0/[M]) versus polymerization time, as shown in Figure
6.
the polymerization of L-lactide and ε-CL in a controlled
manner. The most notable feature of the binuclear ytterbium
isopropoxide initiator described in this report is high po-
lymerization activity in conjunction with good controllability
for the polymerization of both L-lactide and ε-CL. It can be
anticipated that the best combination of a bridged bis(amidi-
nate) with an active group in group 3 metal complexes should
provide new and interesting reactivity. Work to develop new
catalysts with this kind of ligand set is proceeding in our
laboratory.
End-group analysis of the oligomer obtained by polym-
erization at an M/I of 10 with 2 revealed the existence of an
isopropoxyester (δ 5.07 ppm) and an alcohol (δ 3.57 ppm)
1
in H NMR. The results indicated the participation of only
the isopropoxy group in initiation.
Conclusion
We have synthesized ytterbium phenoxide complex 1 and
isopropoxide complex 2 supported by a bridged bis(amidi-
nate) ligand. X-ray structural analysis showed that 1 is a
mononuclear complex with a chelating bis(amidinate) group,
while 2 is a binuclear complex in which two Yb atoms are
connected by two bridged bis(amidinate) groups and two
bridged OiPr groups. Both 1 and 2 were found to initiate
Experimental Section
General Procedures. All manipulations and reactions were
performed under a purified argon atmosphere using standard
Schlenk techniques. Solvents were degassed and distilled from
Inorganic Chemistry, Vol. 48, No. 2, 2009 749

Wang et al.
Table 5. Details of the Crystallographic Data and Refinements for 1
and 2a
1
2
formula
Mr
C39H61N4O3Si2Yb
863.14
C52H82N8O2Si4Yb2
1309.70
temp/K
cryst size/mm
cryst syst
space group
a/Å
213(2)
193(2)
0.50 × 0.30 × 0.30 0.30 × 0.20 × 0.10
monoclinic
P21/c
triclinic
j
P1
24.764(2)
8.7551(6)
20.0979(16)
90
106.648(2
90
4174.7(6)
4
1.373
10.5426(3)
10.7834(5)
15.2118(6)
109.501(9)
99.684(9)
104.004(12)
1521.96
1
b/Å
c/Å
R/deg
ꢀ/deg
γ/deg
V/Å3
Z
Dcalcd/g cm-3
µ/mm-1
F(000)
1.429
3.174
662
2.336
1780
θ range for collection/deg 3.08-25.35
3.02-25.35
no. of reflns collected
no. of independent reflns
R [(I > 2σ(I)]
wR
39562
7645
0.0431
0.0798
1.165
15165
5525
0.0269
0.0618
1.091
GOF
a The structures were solved by direct methods and refined by full-matrix
least-squares procedures based on F2. All the non-hydrogen atoms were
refined anisotropically. The hydrogen atoms were all generated geo-
metrically, assigned appropriate isotropic thermal parameters, and allowed
to ride on their parent carbon atoms. All of the H atoms were held stationary
and included in the structure factor calculation in the final stage of full-
matrix least-squares refinement. The structures were solved and refined using
SHELEXL-97 program.
20.23%. IR (KBr, cm-1): 2957 (m), 1637 (s), 1624 (s), 1436 (m),
1252 (m), 1127 (s), 1084 (s), 1003 (w), 833 (w), 750 (m), 700
(m).
Yb(µ2-L)2(µ2-OiPr)2Yb (2). A solution of NaOiPr (2 mmol, 1
M) in THF was added to a stirring solution of LYbCl(THF)2 (1.55
g, 2 mmol) in THF (30 mL). The reaction mixture was stirred for
48 h at room temperature. The volatiles were removed under
reduced pressure, and the residue was extracted with toluene (2 ×
80 mL). The volume of the extract was reduced to 10 mL, followed
by an addition of 1 mL of DME. Cooling the solution to 0 °C
afforded the product as colorless crystals (0.69 g, 53% based on
Yb). Mp: 92-95 °C. Anal. calcd for C52H82N8O2Si4Yb2: C, 47.69;
H, 6.31; N, 8.56; Yb, 26.42%. Found: C, 47.84; H, 6.01; N, 8.85;
Yb, 26.33%. IR (KBr, cm-1): 2955 (m), 1635 (s), 1566 (s), 1489
(m), 1381 (m), 1250 (m), 1157 (w), 1003 (w), 918 (w), 833 (m),
779 (m), 748 (m), 702 (m).
Figure 6. Polymerization time versus ln[M]0/[M] for the polymerization
of ε-CL initiated by complexes 1 (a) and 2 (b). Conditions: [M]0/[I]0
800, toluene, rt (for complex 1); [M]0/[I]0 ) 1500, toluene 40 °C (for
complex 2).
)
sodium benzophenone ketyl before use. ε-CL and L-lactide were
purchased from Acros and purified by distillation in CaH2 for ε-CL
and recrystallization from a mixture of THF and toluene for
L-lactide. LYbCl(THF)211 was prepared according to the literature
procedure. All other reagents were purchased from Acros and were
used as received without further purification. The IR spectra were
recorded on a Magna-IR 550 spectrometer as KBr pellets. Melting
points were determined in sealed Ar-filled capillary tubes and are
uncorrected. Lanthanide analyses were carried out by a complexo-
metric titration. Carbon, hydrogen, and nitrogen analyses were
preformed by direct combustion using a Carlo-Erba EA 1110
instrument. Molecular weight and molecular weight distributions
were determined against a polystyrene standard by gel permeation
chromatography on a Waters 1515 apparatus with three HR columns
(HR-1, HR-2, and HR-4) at 30 °C; THF was used as an eluent.
LYb(OAr)(DME) (Ar ) 2,6-iPr2C6H3) (1). A solution of
NaOAr (2 mmol, 1 M) in THF was added to a stirring solution of
LYbCl(THF)2 (1.55 g, 2 mmol) in THF (30 mL). The reaction
mixture was stirred for 48 h at room temperature. The volatiles
were removed under reduced pressure, and the residue was extracted
with DME (2 × 25 mL) and reduced to 20 mL. Cooling the solution
to 0 °C afforded the colorless crystals (0.86 g, 50% based on Yb).
Mp: 149-151 °C. Anal. calcd for C39H61N4O3Si2Yb: C, 54.27; H,
7.12; N, 6.49; Yb, 20.05%. Found: C, 54.18; H, 7.02; N, 6.58; Yb,
General Procedure for the Polymerization of ε-CL and
L-Lactide. The procedures for the polymerization of ε-CL and
L-lactide initiated by complexes 1 and 2 are similar, and a typical
polymerization procedure is given below. To a stirred solution of
ε-CL or L-lactide in toluene was added a toluene solution of the
initiator using a syringe. The polymerization mixture was stirred
for a definite time at the desired temperature and then quenched
with an ethanol solution containing a small amount of hydrochloric
acid (5%). The polymer was precipitated from ethanol and washed
with ethanol three times and dried under a vacuum.
Oligomer for End Group Analysis. The oligomerization of
ε-CL (L-lactide) was carried out with 2 in toluene at 25 °C under
a [ε-CL]/[initiator] ([L-lactide]/[initiator]) molar ratio of 10. The
reaction was terminated by adding 1 mL of 5% HCl/hexane after
2 h. The oligomer was precipitated from ethanol. The product was
dissolved in THF, followed by precipitation in ethanol. After
filtration, the white product was dried in a vacuum.
750 Inorganic Chemistry, Vol. 48, No. 2, 2009

Bridged Bis(amidinate) Ytterbium Alkoxide and Phenoxide
X-Ray Crystallography. Crystals suitable for X-ray diffraction
of complexes 1 and 2 were sealed in a thin-walled glass capillary
filled with argon for structural analysis. Diffraction data were
collected on a Rigaku Mercury CCD area detector in the ω scan
mode using Mo KR radiation (λ ) 0.71070 Å). The diffracted
intensities were corrected for Lorentz polarization effects and
empirical absorption corrections. Details of the intensity data
collection and crystal data are given in Table 5.
National Natural Science Foundation of Jiangsu province
(BK2007505) and the Major Basic Research Project of the
National Natural Science Foundation of the Jiangsu Higher
Education Institutions (07KJA15014) is gratefully acknowl-
edged.
Supporting Information Available: Crystallographic file in CIF
format. This material is available free of charge via the Internet at
http://pubs.acs.org.
Acknowledgment. Financial support from the National
Natural Science Foundation of China (20632040) and the
IC801189J
Inorganic Chemistry, Vol. 48, No. 2, 2009 751