Low-coordinate rare-earth complexes of the asymmetric 2,4-di-tert- butylphenolate ligand prepared by redox transmetallation/protolysis reactions, and their reactivity towards ring-opening polymerisation

Article abstract of DOI:10.1039/c0dt00023j

New trivalent lanthanoid aryloxide complexes have been prepared by redox transmetallation/protolysis (rtp) reactions using 2,4-di-tert-butylphenol (dbpH). Mononuclear octahedral complexes from tetrahydrofuran (thf) were of the type [Ln(dbp)₃(thf)₃] (Ln = La (1), Pr (2), Nd (3), Gd (4), Er (5)). The lanthanoid contraction results in the rather subtle change in stereochemistry from meridional (La, Pr, Nd, Gd) to facial (Er). An analogous reaction with neodymium in dimethoxyethane (dme), resulted in the isolation of the seven coordinate [Nd(dbp)₃(dme)₂] (6), and this is comparable with the thf complexes in terms of steric crowding. Dinuclear complexes of the type [Ln₂(dbp)₆(thf)₂], (Ln = Nd (7), Er (8)) were obtained when 1 and 5 were recrystallised from toluene. These dimeric complexes contain two bridging and four terminal phenolates, as well as a single coordinated molecule of thf at each metal. A similar structural motif was observed for the products when the reaction was performed in diethyl ether, and in the absence of a solvent, yielding [Nd₂(dbp) ₆(Et₂O)₂] (9) and [Nd₂(dbp) ₆(dbpH)₂] (10) respectively. Complexes 3 and 4 alone were efficient but poorly-controlled initiators for the ROP of rac-lactide, but with an excess of BnOH as a co-initiator they showed features consistent with immortal polymerisation. Use of BnNH₂ led to well-controlled, amine-initiated immortal ROP of rac-lactide, only the second report of this type of process for a group 3 or lanthanoid system. The Royal Society of Chemistry 2010.

Full text of DOI:10.1039/c0dt00023j

View Article Online / Journal Homepage / Table of Contents for this issue
This article is published as part of the Dalton Transactions themed issue entitled:
New Horizons in Organo-f-element Chemistry
Guest Editor: Geoff Cloke
University of Sussex, UK
Published in issue 29, 2010 of Dalton Transactions
Image reproduced with the permission of Tobin Marks
Articles in the issue include:
PERSPECTIVES:
Organo-f-element catalysts for efficient and highly selective hydroalkoxylation and hydrothiolation
Charles J. Weiss and Tobin J. Marks, Dalton Trans., 2010, DOI: 10.1039/c003089a
Non-classical divalent lanthanide complexes
François Nief, Dalton Trans., 2010, DOI: 10.1039/c001280g
COMMUNICATIONS:
A bimetallic uranium μ-dicarbide complex: synthesis, X-ray crystal structure, and bonding
Alexander R. Fox, Sidney E. Creutz and Christopher C. Cummins
Dalton Trans., 2010, DOI: 10.1039/c0dt00419g
PAPERS:
Coordination polymerization of renewable butyrolactone-based vinyl monomers by lanthanide and
early metal catalysts
Garret M. Miyake, Stacie E. Newton, Wesley R. Mariott and Eugene Y.-X. Chen,
Dalton Trans., 2010, DOI: 10.1039/c001909g
Visit the Dalton Transactions website for more cutting-edge inorganic and
organometallic chemistry research
www.rsc.org/dalton

PAPER
www.rsc.org/dalton | Dalton Transactions
Low-coordinate rare-earth complexes of the asymmetric
2,4-di-tert-butylphenolate ligand prepared by redox
transmetallation/protolysis reactions, and their reactivity towards
ring-opening polymerisation†
Lawrence Clark,b Glen B. Deacon,*a Craig M. Forsyth,a Peter C. Junk,*a Philip Mountford*b and
Josh P. Townleya
Received 25th February 2010, Accepted 19th May 2010
First published as an Advance Article on the web 10th June 2010
DOI: 10.1039/c0dt00023j
New trivalent lanthanoid aryloxide complexes have been prepared by redox transmetallation/protolysis
(rtp) reactions using 2,4-di-tert-butylphenol (dbpH). Mononuclear octahedral complexes from
tetrahydrofuran (thf) were of the type [Ln(dbp)3(thf)3] (Ln = La (1), Pr (2), Nd (3), Gd (4), Er (5)). The
lanthanoid contraction results in the rather subtle change in stereochemistry from meridional (La, Pr,
Nd, Gd) to facial (Er). An analogous reaction with neodymium in dimethoxyethane (dme), resulted in
the isolation of the seven coordinate [Nd(dbp)3(dme)2] (6), and this is comparable with the thf
complexes in terms of steric crowding. Dinuclear complexes of the type [Ln2(dbp)6(thf)2], (Ln = Nd (7),
Er (8)) were obtained when 1 and 5 were recrystallised from toluene. These dimeric complexes contain
two bridging and four terminal phenolates, as well as a single coordinated molecule of thf at each metal.
A similar structural motif was observed for the products when the reaction was performed in diethyl
ether, and in the absence of a solvent, yielding [Nd2(dbp)6(Et2O)2] (9) and [Nd2(dbp)6(dbpH)2] (10)
respectively. Complexes 3 and 4 alone were efficient but poorly-controlled initiators for the ROP of
rac-lactide, but with an excess of BnOH as a co-initiator they showed features consistent with immortal
polymerisation. Use of BnNH2 led to well-controlled, amine-initiated immortal ROP of rac-lactide,
only the second report of this type of process for a group 3 or lanthanoid system.
the ligand. For this reason, 2,4-di-tert-butylphenol was chosen for
Introduction
this investigation. The steric bulk in one ortho-position provides
the potential for some degree of saturation of the coordination
sphere, while the para-tert-butyl group should assist hydrocarbon
solubility.
Investigations into the structures of lanthanoid aryloxides have
given insight into the effect of substituents on coordination num-
ber, stereochemistry and nuclearity.1–3 For example, the 2,6-di-tert-
butyl-4-X-phenolates have had a pivotal role in establishing low
coordination numbers (e.g. 3) for large lanthanoid ions.4,5 By con-
trast, 3,5-di-tert-butylphenolates with distal substituents favours
di-nuclear and polynuclear complexes, e.g. [Nd3(OAr)9(thf)4],6
which have excellent hydrocarbon solubility, which is attractive
for sol–gel processes.1,7,8
A recent review by Boyle and Ottley2 is dominated by the
extensive investigations into symmetrically substituted phenox-
ides, particularly 2,6-disubstituted derivatives. A recent study of
complexes of the asymmetric 2-tert-butylphenolate ion9 (tbp), and
comprehensive reviews/books highlight a distinct lack of data
regarding lanthanoid complexes incorporating asymmetrically
substituted phenolates.1–3 However, an asymmetric steric bulk may
lead to interesting changes in the bonding or bridging ability of
Lanthanoid aryloxides are traditionally prepared by protolysis
of Ln(NR2)3 (R = SiMe3) with the phenol or by metathesis reac-
tions. However, the former method requires Ln(NR2)3 to be highly
pure, often requiring multiple crystallisations or sublimations.1,3
Metathesis reactions between a lanthanoid halide and an al-
kali metal phenolate can experience complications arising from
halide/alkali halide retention.2 Redox transmetallation/protolysis
reactions10–12 are emerging as a versatile alternative to traditional
metathesis/protolysis reactions especially for the synthesis of low
coordinate rare earth complexes.10,13–15
Such complexes have a number of potential applications,
including use of Ln(II) compounds as soluble one-electron re-
ducing agents,16 routes to lanthanoid doped semiconductors,17
MOCVD (metal organic chemical vapour deposition) and sol–
gel precursors1,7,8 and as catalysts.1
aSchool of Chemistry, Monash University, Clayton, Vic., 3800, Australia
Polyesters derived from lactic acid are currently a topic of much
interest due to their biocompatibility and biodegradability.18–31
Derived from 100% renewable resources including corn and
sugar beet, these polymers can act as replacements for oil-based
materials. These materials are commercially synthesised20 by the
metal catalysed ring-opening polymerization (ROP) of the cyclic
bChemistry Research Laboratory, Department of Chemistry, University of
Oxford, Mansfield Road, Oxford, UK OX1 3TA
† Electronic supplementary information (ESI) available: Selected bond
lengths and angles for [Nd2(dbp)6(dbpH)2)] and [Er(dbp)3(thf)3]. CCDC
reference numbers [CCDC NUMBER(S)]. For ESI and crystallographic
data in CIF or other electronic format see DOI: 10.1039/c0dt00023j
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 6693–6704 | 6693
©

diester lactide (LA) which exists as three stereoisomers, designated
meso, L and D (the 50 : 50 mixture of L- and D-lactide is known
as rac-lactide).
3 when in thf, x = 2 when in dme) in moderate to good yields as
shown in eqn (1).
2Ln(Hg) +6[dbpH]+3[Hg(C6F )2 ]
5
solv
⎯⎯→ 2[Ln(dbp)3(solv)x ]+6C6F H +3Hg
5
(1)
solv = thf, x = 3; Ln = La(1), Pr(2), Nd(3), Gd(4), Er(5),
solv = dme, x = 2; Ln = Nd(6)
A
synthesis was also attempted in the absence of
bis(pentafluorophenyl)mercury, but no reaction was detected, and
only unreacted phenol was recovered, suggesting that this phenol is
not sufficiently acidic to undergo a direct reaction with the metal
in polar solvents, even after activation with elemental mercury.
Accordingly, the reactions plausibly proceed through formation,
then protolysis, of pentafluorophenyllanthanoid species, a path
for which there is significant evidence.10,12 A polar solvent also
appears necessary to facilitate redox transmetallation/protolysis,
since an attempted synthesis in toluene failed. There is only
one example of redox transmetallation/protolysis in toluene,35
namely the generation of the homoleptic [La(Odpp)3] (Odpp = 2,6-
diphenylphenolate) complex, where p-Ph–La interactions detected
in the structure may have a role in the success of the synthesis.
In most cases the reaction in thf proceeds quickly, often with
a visible colour change occurring within seconds of initiation.
Reaction rates appeared to be dependent on the form of the
metal, with those that used fresh filings (ytterbium and lanthanum)
reacting much more rapidly than those with commercial powders
(neodymium, gadolinium, erbium). Praseodymium metal, in the
form of chips, resulted in the slowest reaction, presumably
due to the smaller surface area of the metal. The progress of
the reactions could be monitored by 19F NMR spectroscopy,
since bis(pentafluorophenyl)mercury is converted into pentafluo-
robenzene, which shows characteristic resonances at -139 ppm,
-162 ppm, and -154 ppm.36 In most cases the completion of
the reaction can also be seen in the infrared spectrum by the
disappearance of the OH stretching absorptions at 3610 cm-1
and 3534 cm-1. However, when Ln = Gd, the X-ray crystal
structure showed the presence of 0.5 dbpH of crystallisation
per molecule. The IR spectrum showed two n(OH) absorptions
(Experimental section), somewhat displaced from the free ligand
values, which were regenerated on exposure of the Nujol mull to
air. The phenolates could be isolated, in many cases, by simple
filtration and subsequent crystallisation. However, in reactions
with Sm, Dy, Ho, Y and Yb, although there was evidence of the
formation of the appropriate phenolates with some dissolution
of metal, observation of appropriately coloured solutions, and
formation of pentafluorobenzene, the products could not be
crystallised. The IR spectra of the oils showed no OH absorption,
consistent with phenolate formation. Overall, the results show that
redox transmetallation/protolysis is a suitable synthetic method
for phenolates without two ortho substituents to provide steric
stabilisation. While the paramagnetic property of most of the
lanthanoid complexes makes characterisation by 1H and 13C NMR
spectroscopy difficult due to dramatic shifting and broadening of
the resonances, assignments have been made for most compounds.
Elemental and/or metal analysis results are in agreement with the
crystal structure compositions, except for 5 where loosely held
solvent of crystallisation observed at 123 K is not detected on
analysis of the bulk sample at room temperature.
ROP catalysts are generally derived from Lewis acidic metals,
feature one or more alkoxide initiating groups, and operate
through a coordination–insertion chain growth mechanism.25–30
Group 3 and lanthanoid complexes have been at the forefront
of ROP catalysis for nearly two decades.28–30 While most of the
recent developments in this area have focused on ligand-supported
systems,27,28,30 homoleptic aryloxides of the type [Ln(OAr)3] were
among the first classes of rare earth lactone or lactide ROP catalyst
studied. Feijen et al. reported the ROP of L-LA using the tris(2,6-
di-tert-butylphenoxide) compounds [Ln(OAr*)3] (Ar* = O-2,6-
C6H3But2).32,33 When used alone, these are sluggish catalysts for
ROP owing to the presence of the two bulky ortho-positioned
tert-butyl groups. No specific end groups could be identified in
the resultant poly(L-LA) which had broad polydispersity indices
(PDIs) and large Mn (number average molecular weight) values
compared to those expected for one poly(L-LA) chain forming
per initiating Ln–OAr* group. However, in the presence of various
alcohols ROH (e.g., R = Bn, Pri or But), well-controlled ROP
was observed giving OR-terminated poly(L-LA) with narrow
molecular weight distributions. The Mn values were consistent
with one polymer chain forming per ROH added, and it was
proposed that [Ln(OAr*)3] was converted to the less sterically
encumbered, but symmetrically substituted [Ln(OR)3] in situ,
this in turn leading to poly(L-LA) by a coordination–insertion
mechanism.
Following on from Feijen’s seminal work with [Ln(OAr*)3],
Shen et al. recently reported the ROP of L-LA and rac-LA using
less sterically encumbered tris(phenoxide) complexes of the type
[Ln(O-4-C6H4But)3] and [Ln(O-2,6-C6H3Me2)3].34 In contrast to
Feijen’s results for [Ln(OAr*)3], these less bulky compounds gave
somewhat better controlled ROP, especially for the larger metals.
In these cases the reported number average molecular weights were
in good agreement with those expected for one poly(LA) chain per
Ln–OAr bond, although the PDIs at high conversions were still
relatively broad (ca. 1.5–1.6).
The new 2,4-di-tert-butylphenoxide complexes prepared in this
present study provide an opportunity to explore the effect of an
unsymmetrical substitution pattern on ring-opening polymerisa-
tion by lanthanoid tris(phenolates).
Results and discussion
Reactions were carried out as a one-pot synthesis, in which
bis(pentafluorophenyl)mercury12 and 2,4-di-tert-butylphenol
(dbpH) were added to an excess of lanthanoid metal (with 1–2
drops of mercury for activation) in a donor solvent, such as thf or
dme, to give compounds of the type [Ln(dbp)3(solv)x] (where x =
6694 | Dalton Trans., 2010, 39, 6693–6704
This journal is
The Royal Society of Chemistry 2010
©

In the cases of lanthanum to gadolinium, the product is a
six coordinate heteroleptic compound with distorted octahedral
geometry and a meridional ligand array, as shown in Fig. 1 for
Ln = La.
Fig. 2 Molecular structure of [Er(dbp)3(thf)3] (5). Ligands are arranged
facially. Hydrogen atoms omitted for clarity.
[Ce(tbp)2(thf)3], again reflecting the influence of the para-tert-butyl
group on the structural outcome.
As complexes of lanthanoid metals smaller than erbium could
not be crystallised, it is not known whether these complexes
would preserve the facial conformation, or possibly even adopt
a structure with a lower coordination number.
Fig. 1 Molecular structure of [La(dbp)3(thf)3] (1). Representative of
metals La–Gd. Ligands are arranged meridionally. Hydrogen atoms
omitted for clarity.
Some comparative bond lengths and angles between
[La(dbp)3(thf)3] (1), [Pr(dbp)3(thf)3] (2), [Nd(dbp)3(thf)3] (3), and
[Gd(dbp)3(thf)3] (4) are given in Table 1. Data for the less precise
[Er(dbp)3(thf)3] (5) are not included.
The structure resembles that of [Ce(tbp)3(thf)3]9 (tbp = 2-tert-
butylphenoxide) in terms of coordination number, albeit with
the important difference that this cerium complex exhibits facial
ligand coordination. Since the only difference in the ligands
is the tert-butyl group in the para-position, the capacity to
affect stereochemistry of the complexes in the solid state is
not limited to the ortho- and meta-substituents. A few similar
observations exist in the literature,15a,37 with some major structural
differences between lanthanoid or alkaline earth complexes of
2,6-dimethylphenolates and 2,4,6-trimethylphenolates. Significant
differences also exist between structures of [Yb(OC6H2But2-2,6-X-
4)2(thf)3] complexes from X = Me15b to X = But.15c
Five of the six Ln–O bond lengths contract by more (0.097–
3+
0.13 A) than the decrease in ionic radius from La to Gd3+
˚
38
˚
˚
(0.094 A), but Ln–O(2) contracts only by 0.073 A, reducing
˚
the average contraction to 0.11 A. The distance from the metal
to the quaternary carbon on the ortho-tert-butyl group remains
˚
essentially the same, approximately 4.9 A, and the lanthanoid
contraction is off-set by a decrease in the average Ln–O–C angle
from lanthanum to gadolinium. Presumably the shortening of the
metal–oxygen bond puts a greater steric strain on the system,
causing the average Ln–O–C angle to deviate from linearity with
each successively smaller metal (see Fig. 3). The change to fac-
stereochemistry places all the aryloxide ligands transoid to thf
ligands which have a weak trans influence in contrast to the
high trans influence of OAr ligands,39 and this arrangement may
enhance stabilisation. The clustering of common ligands is a
feature of high coordination number lanthanoid complexes.8 Six
is a relatively low coordination number for lanthanoid elements
and the sum of the steric coordination numbers of the ligands
is not excessive at 7.8.40 Comparisons with fac-[Ce(tbp)3(thf)3],9
show differences in ligand arrangement. The o-tert-butyl groups of
[Ce(tbp)3(thf)3] are oriented towards the centre of the molecule to
face one another, whereas in [Er(dbp)3(thf)3], they are positioned
to face away from one another, presumably to reduce repulsion
between But groups with a smaller lanthanoid ion in the solid
state.
Isolation of crystalline products from the reactions of smaller
metals (Dy, Ho, Er, Yb) with dbpH proved extremely difficult.
Only crystals of [Er(dbp)3(thf)3] (5) could be obtained. However,
complete solution of the structure was problematic, and so is
presented in terms of connectivity only. Numerous attempts to
produce more suitable crystals of 5 failed. When a concentrated
solution of 5 was layered with dry hexane in an attempt to
induce crystallisation, crystals of [Er2(dbp)6(thf)2] (8) were instead
obtained. The fact that these complexes lose coordinated thf so
readily may be a contributing factor to the difficulty of their
crystallisation. It was anticipated that there should be some change
as a result of the lanthanoid contraction, perhaps in degree of
solvent coordination or nuclearity. However, the observed change
from Gd to Er was much more subtle, with the erbium complex
showing a switch in ligand arrangement from meridional to facial
(Fig. 2). Thus, the smaller lanthanoid has a structure similar to
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 6693–6704 | 6695
©

Table 1 Selected bond lengths and angles of [Ln(dbp)3(thf)3] (1–4)
La (1)
Pr (2)
Nd (3)
Gd (4)
˚
Bond lengths/A
Ln–O(1)
Ln–O(2)
Ln–O(3)
Ln–O(4) (thf)
Ln–O(5) (thf)
2.262(3)
2.243(4)
2.283(3)
2.597(3)
2.623(3)
2.570(3)
1.032
2.209(5)
2.212(4)
2.213(4)
2.552(4)
2.605(5)
2.527(5)
0.990
2.207(2)
2.194(2)
2.206(2)
2.535(2)
2.580(2)
2.500(2)
0.983
2.165(4)
2.170(4)
2.162(4)
2.475(4)
2.515(4)
2.444(4)
0.938
Ln–O(6) (thf)
Ionic radii (metal)38
Ln–o-quaternary carbon (average)
4.939
4.916
4.904
4.922
Bond angles/◦
Ln–O(1)–C(1)
153.1(3)
156.6(3)
157.8(3)
157.8(4)
145.2(4)
153.6(4)
146.07(19)
158.34(19)
152.65(19)
135.40(8)
112.61(8)
106.35(8)
72.47(7)
120.24(7)
166.01(7)
100.72(7)
76.72(7)
161.7(4)
144.4(4)
142.8(4)
Ln–O(2)–C(15)
Ln–O(3)–C(29)
O(1)–Ln–O(2)
O(2)–Ln–O(3)
O(3)–Ln–O(1)
O(4)–Ln–O(5) (thf)
O(5)–Ln–O(6) (thf)
O(6)–Ln–O(4) (thf)
O(1)–Ln–O(4)
O(1)–Ln–O(5)
130.93(12)
110.71(12)
112.89(13)
68.72(11)
127.57(11)
163.42(11)
104.81(11)
77.96(12)
136.04(17)
106.27(17)
112.18(18)
72.50(15)
119.90(16)
166.11(16)
106.68(16)
78.95(17)
140.94(15)
100.80(16)
110.96(15)
76.69(14)
120.60(15)
159.83(14)
92.23(15)
84.19(14)
Fig. 3 Average Ln–ortho-quaternary carbon distance and Ln–O1–C1
angles of 1 and 4.
For comparison, the reaction between neodymium and 2,4-di-
tert-butylphenol was also carried out in 1,2-dimethoxyethane (eqn
(1)). The resulting complex was similar to the thf derivative, but
contained two molecules of chelating dme, making the metal seven
coordinate, rather than six (Fig. 4). The aryloxide ligands are
arranged pseudo-meridionally with respect to the dme.
Fig. 4 Molecular structure of [Nd(dbp)3(dme)2]. The dbp ligands are
arranged in a pseudo-meridional manner. Hydrogen atoms are omitted
for clarity.
Some comparative bond lengths and angles between the thf and
dme derivatives are given in Table 2. The dme derivative exhibits a
longer average Nd–O(solv.) bond, which is to be expected from
the change in coordination number. However, the increase of
<Nd–OAr> is less marked. Interestingly, the two longer Nd–OAr
bond lengths are transoid to each other reflecting the strong trans
influence of aryloxide ligands.39
Some structural similarity between the thf and dme derivatives is
in accordance with the steric coordination numbers40 of 1.21 (thf)
and 1.78 (dme), whereby three molecules of thf provide essentially
the same steric hindrance as two chelating dme molecules.
Due to the moderate bulkiness of the aryloxide ligand, a point
of interest was to determine whether a homoleptic complex was
achievable. A number of possible routes were explored. Since the
redox transmetallation/protolysis reaction would not proceed in
any non-coordinating solvents attempted (hexane, toluene), the
possibility that the thf could be removed by recrystallisation of 3
from toluene was studied. This proved to be partially successful,
giving the dinuclear species [Nd2(dbp)6(thf)2].2C7H8 (7.2C7H8)
(Fig. 5, eqn (2)), and only one molecule of thf per metal centre.
toluene
(2)
2[Ln(dbp)3(thf)3] ⎯⎯⎯→ [Ln2 (dbp)6 (thf)2 ]+ 4thf
The coordination number of each metal centre was reduced to
five. Consequently, <Nd–O> of 7 is shorter than that of 3, and
˚
Nd–O(thf) of 7 is shorter (0.1 A) than <Nd–O(thf)> of 3 despite
being transoid to an aryloxide. However, since it is bridging and
˚
is 0.2 A longer than Nd–OAr(terminal) (Table 3), it evidently has
little trans influence.
6696 | Dalton Trans., 2010, 39, 6693–6704
This journal is
The Royal Society of Chemistry 2010
©

Table 2 Selected bond lengths and angles of [Nd(dbp)3(thf)3] (3) and
[Nd(dbp)3(dme)2] (6)
[Nd(dbp)3(thf)3] [Nd(dbp)3(dme)2]
(3)
(6)
˚
Bond lengths/A
Nd–O(1)
Nd–O(2)
Nd–O(3)
2.207(2)
2.194(2)
2.206(2)
2.538
2.243(7)
2.210(6)
2.241(6)
2.614
Nd–O(solvent) (average)
Nd–o-quaternary carbon (average) 4.904
4.881
Bond angles/◦
Nd–O(1)–C(1)
Nd–O(2)–C(15)
Nd–O(3)–C(29)
O(1)–Nd–O(2)
O(2)–Nd–O(3)
O(3)–Nd–O(1)
O(4)–Nd–O(5) (thf)
O(5)–Nd–O(6) (thf)
O(6)–Nd–O(4) (thf)
O(1)–Nd–O(4)
146.07(19)
147.4(6)
162.5(6)
151.5(6)
94.4(2)
103.3(3)
162.1(2)
63.7(2)
135.9(2)
74.4(2)
81.0(2)
158.34(19)
152.65(19)
135.40(8)
112.61(8)
106.35(8)
72.47(7)
120.24(7)
166.01(7)
100.72(7)
76.72(7)
Fig. 5 Molecular structure of [Nd2(dbp)6(thf)2].2C7H8 (7.2C7H8). Hydro-
gen atoms omitted for clarity.
O(1)–Nd–O(5)
104.3(2)
lanthanoid aryloxides owing to its greater steric bulk (compared
to thf).40 The reaction was successful, and crystals were obtained
directly from the reaction flask. They were, however, not the
target homoleptic complex, but instead a dinuclear species,
[Nd2(dbp)6(Et2O)2] (9), very similar to 7 (eqn (3), Fig. 6). When
compared with 7, Nd–O(ether) increases for 9 due to the greater
bulk of diethyl ether compared with thf (steric coordination
numbers 1.44 vs. 1.21, respectively),40 and there is a shortening
of the average Nd–o-quaternary carbon distance of the terminal
aryloxides. This is offset, however, by a lengthening of this same
distance of the bridging aryloxides (Table 3).
Again the Ln–o-quaternary carbon distance is conserved in
the terminal aryloxides, with an average distance of 4.88 A,
˚
similar to those of 1–5 (Table 1). The bridging aryloxides exhibit
˚
significantly shorter distances (4.53 A, Table 3), but this position
may be dictated by the bulky o-But group of the nearby terminal
aryloxides. Selected bond lengths and angles are given in Table 3.
The 1H NMR spectrum of 7 in C6D6 showed two broad aromatic
resonances and two broad But resonances, both in a 2 : 1 intensity
ratio, attributable to terminal and bridging ligands respectively.
Recrystallisation of [Er(dbp)3(thf)3] from toluene (or
thf/hexane – above) yielded a structurally similar complex,
(8), albeit with no lattice toluene. Table 3 provides bond distance
and angle data, and shows similar features to those of 7 with the
terminal Er–O distances shortened by the difference between the
ionic radii of Nd3+ and Er3+.38 Surprisingly, Er–OAr(bridging) is
shorter than Nd–OAr(bridging) by twice the expected amount.
In a further attempt to obtain a homoleptic complex, the
redox transmetallation/protolysis reaction was performed in
diethyl ether (eqn (3)), which sometimes does not coordinate10 to
2Ln(Hg) +6[dbpH]+3[Hg(C6F )2 ]
5
(3)
Et
O
2
⎯⎯⎯→ [Ln2(dbp)6 (Et2O)2 ]+6C6F H +3Hg
5
An effort was made to synthesise a homoleptic neodymium
complex by direct reaction between neodymium and 2,4-di-tert-
butylphenol, in the absence of any solvent, with a drop of mercury
◦
for activation (eqn (4)). When heated to 120 C for 24 h, a blue
deposit was observed on the metal surface, as well as a few single
crystals. While the quality of the crystals for X-ray crystallography
Table 3 Selected bond lengths and angles of dimeric complexes 7–9
[Nd2(dbp)6(thf)2] (7)
[Er2(dbp)6(thf)2] (8)
[Nd2(dbp)6(Et2O)2] (9)
˚
Bond lengths/A
Ln(1)–Ln(1)¢
Ln(1)–O(1)
Ln(1)–O(2)
Ln(1)–O(3)
Ln(1)–O(4)
3.933(10)
2.432(4)
2.189(4)
2.167(4)
2.437(4)
4.881
3.6199(3)
2.243(2)
2.084(2)
2.064(2)
2.342(2)
4.824
3.9000(4)
2.374(2)
2.168(2)
2.163(3)
2.484(2)
4.768
Ln(1)–o-quaternary carbon (terminal) (average)
Ln(1)–o-quaternary carbon (bridging)
4.528
4.603
4.583
Bond angles/◦
Ln(1)–O(1)–C(1)
Ln(1)–O(2)–C(15)
Ln(1)–O(3)–C(29)
O(4)–Ln(1)–O(1)¢
O(4)–Ln(1)–O(2)
O(4)–Ln(1)–O(3)
O(2)–Ln (1)–O(3)
99.3(3)
157.8(3)
160.0(4)
175.91(14)
86.74(14)
84.26(15)
111.45(15)
111.04(17)
141.8(2)
153.6(2)
169.04(8)
88.52(8)
86.28(9)
120.86(9)
91.60(8)
156.4(2)
158.8(3)
174.44(8)
87.86(8)
87.63(9)
114.56(10)
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 6693–6704 | 6697
©

Table 4 Polymerisation of rac-lactide initiated by [Nd(dbp)3(thf)3] (3) and [Gd(dbp)3(thf)3] (4)a
c
Catalyst
Equivs. BnOH
Conversion (%) b
Mn (calcd)/g mol-1
Mn(GPC)c/g mol-1
Mw/Mn
[Nd(dbp)3(thf)3] (3)
[Gd(dbp)3(thf)3] (4)
[Nd(dbp)3(thf)3] (3)
[Gd(dbp)3(thf)3] (4)
0
0
5
5
92
91
96
95
4420d
4370d
2880e
2850e
7570
10 100
3530
3190
2.7
2.2
1.5
1.6
1
a Conditions: [rac-LA]0 : [Ln]0 = 100, 2 mL thf, 293 K, reaction time 30 or 60 mins. b Measured by H NMR in CDCl3. c Measured by GPC using the
appropriate Mark–Houwink corrections for poly(rac-LA) in thf. d Mn (calcd) = (conversion ¥ 144.13 ¥ 100)/3. e Mn (calcd) = [(conversion ¥ 144.13 ¥
100)/5] + 108.1.
Fig. 7 Molecular structure of [Nd2(dbp)6(dbpH)2] (10). Hydrogen atoms
are omitted for clarity.
polylactides.41–43 For both 3 and 4 the experimental molecular
Fig. 6 Molecular structure of [Nd2(dbp)6(Et2O)2] (9). Hydrogen atoms
weights of 7570 and 10 100 g mol-1 were significantly higher than
are omitted for clarity.
those predicted for one polymer chain growing per Ln–dbp bond.
A likely interpretation is that once one or more LA monomers
was relatively poor, the connectivity of the resulting solution
and refinement was unambiguous. In the absence of a solvent,
unreacted phenol is incorporated to saturate the coordination
sphere of the metal, yielding a complex with a structural motif
similar to 7, 8 and 9 (Fig. 7).
have inserted, the propagation rate for the so-formed Ln-[LA]x-
dbp moiety is significantly larger than that for initiation due to
reduced steric hindrance. The very broad PDIs of 2.7 and 2.2 are
also consistent with a poorly controlled polymerisation process,
with poly(rac-LA) chains of considerably different molecular
weights forming, perhaps from metal centres with one, two or
three propagating chains.
D
Mindful of Feijen’s report32,33 that addition of certain alcohols
to the very bulky [Ln(OAr*)3] systems could lead to better-behaved
catalyst systems, we carried out further ROP experiments with 3
and 4 in the presence of 5 equivs. BnOH (Table 4). The reduced
PDIs of 1.5 and 1.6 are consistent with better controlled sys-
tems. Furthermore, in both cases considerably reduced molecular
weights of 3530 and 3190 g mol-1 are observed which are close to
those expected for one poly(rac-LA) chain forming per added
equiv. of BnOH (2880 and 2850 g mol-1). This suggests that,
in addition to leading to the exchange of Ln–OAr groups for
Ln–OBn counterparts, the excess BnOH (as compared to Ln–
OAr bonds) is able to act as a chain transfer agent such that
5 poly(rac-LA) chains form per metal centre. This phenomenon
is well-established in alcohol co-initiated ROP and is the basis
of so-called immortal polymerisation processes.44 Consistent with
(4)
2Nd +8[dbpH] ⎯H⎯g→[Nd2(dbp)6 (dbpH)2 ]+3H2
Despite the imprecision of the determination, Nd–O(4) is clearly
much longer than <Nd–O(2,3)>, which are typical Nd–OAr bond
lengths. Hence O(4) is the phenol oxygen. There is also observation
of the proton of O4 interacting with the phenyl ring of O1¢, with a
˚
distance of only 2.99 A between O(3) and the centroid of the O(1)¢
phenolate ring.
Polymerisation studies
The compounds [Nd(dbp)3(thf)3] (3) and [Gd(dbp)3(thf)3] (4) were
selected as representative examples and their ability to effect
the ROP of rac-LA in thf at room temperature was assessed.
Both compounds gave >90% conversion of monomer to poly(rac-
1
the proposed mechanism, the H NMR and MALDI-ToF mass
1
LA) within minutes for [rac-LA]0 : [Ln]0 = 100 as judged by H
spectra of the resultant poly(rac-LA) featured only HOC(H)Me–
and BnOC(O)C(H)Me– end groups. Unfortunately, the poly(rac-
LA) formed in these and the other ROP experiments described
herein exhibited no isotactic or heterotactic bias. The D(m/z) of
NMR spectroscopy. The molecular weights and PDIs (Mw/Mn;
Table 4) were determined by gel permeation chromatography
(GPC) using the appropriate Mark–Houwink corrections for
6698 | Dalton Trans., 2010, 39, 6693–6704
This journal is
The Royal Society of Chemistry 2010
©

72 amu between the adjacent peak envelopes in the MALDI-ToF
mass spectra is indicative of extensive transesterification which is
typically found for lanthanoid-based ROP catalysts.
compared to the equivalent experiments with 5 equivs. BnOH
(PDI = 1.5) or 3 alone without co-initiator (PDI = 2.7). Analogous
experiments increasing [rac-LA]0 : [3] from 100 to 500 gave the
linear relationship with Mn illustrated in Fig. 8. The gradient of
the least-squares fit (R2 = 0.986) is 35(2) g mol-1 which compares
well with that expected (28.8 g mol-1) for 5 polymer chains growing
per metal centre in each case.
While this work was in progress, we reported the well-controlled
amine-coinitiated ROP of rac-LA in the presence of certain
ligand-supported dicationic, zwitterionic and neutral yttrium
complexes.45 This approach yields amide-terminated poly(rac-
lactide)s of the type RR¢N-[poly(rac-LA)]-H and so could provide
improved routes for the synthesis of dendrimetic, brush and star-
shaped poly(lactides). We were therefore interested to determine
whether complexes of the type Ln(OAr)3 would also catalyse the
amine-coinitiated ROP of rac-LA.
Using 3 as a representative example we carried out two series
of experiments with BnNH2 as the co-initiator. The results are
summarised in Fig. 8 and 9. For example, ROP of 100 equivs.
rac-LA in the presence of 5 equivs. BnNH2 in thf at room
temperature gave 94% conversion to benzyl amine-terminated
poly(rac-LA) within 10 min. The Mn of 3110 g mol-1 measured
by GPC was close to that expected (2820 g mol-1) for 5 BnNH-
[poly(rac-LA)]-H chains forming per metal centre. The PDI of
1.3 for this polymer indicates a much better defined ROP process
Fig. 9 shows a plot of experimental Mn (as measured by GPC)
vs. [rac-LA]t : [BnNH2]0 for the ROP of rac-LA (held constant at
200 equivs. at t = 0) using 3 and 3, 5, 7, or 10 equivs. BnNH2
co-initiator. The gradient of 176(5) g mol-1 for the fitted line (R2 =
0.985) is in reasonable agreement with that expected (144.1 g mol-1)
1
for one chain growing per added amine. Finally, the H NMR
and MALDI-ToF mass spectra of the poly(rac-LA) formed were
consistent with the exclusive formation of BnNH-[poly(rac-LA)]-
H, although once again the polymers were atactic and had D(m/z)
separations of 72 amu between adjacent peak envelopes in the
MALDI-ToF mass spectra.
Since metal alkoxides/aryloxides are not deprotonated by
BnNH2, the data in Fig. 8 and 9 suggest that in the pres-
ence of 3, BnNH2 rapidly (with respect to propagation) ring-
opens a molecule of rac-LA to form BnNH-C(O)CH(Me)OC-
(O)CH(Me)O-H (11) via an activated monomer mechanism (eqn
(5)).46 Thereafter, 11 (containing an acidic O–H group) acts in
an analogous manner to ROH (cf. Table 4 and Feijen’s initial
reports32,33), forming active species of the type [Nd{-poly(rac-
LA)-NHBn}3] under the polymerisation conditions. These would
then propagate the ROP through a coordination-insertion mecha-
nism, with the individual BnNH-[rac-LA]n-H chains equilibrating
rapidly as proposed in “conventional” immortal polymerisation
systems.44 An analogous mechanism was proposed previously for
the BnNH2 co-initiated ROP of rac-LA in the presence of certain
ligand-supported yttrium complexes.45
Fig. 8 Mn (as measured by GPC) vs. [rac-LA]t : [Nd]0 for the ROP of
rac-LA (between 100 and 500 equivs. at t = 0) using 3 and 5 equivs. BnNH2
co-initiator. The numbers in parentheses are the PDIs. The gradient of the
fitted line (R2 = 0.986) is 35(2) g mol-1 with an intercept of 107.1 g mol-1
(corresponding to the benzylamine-derived end groups).
(5)
Attempts to extend the amine co-initiated ROP of rac-LA
to ButNH2 or piperidine as representative bulky primary or
secondary amines were unsuccessful. Poly(rac-LA)s with broad
PDIs between 1.6 and 1.9 and poor agreement between found
and predicted molecular weights were obtained. It appears that
with these less nucleophilic co-initiators there may be competition
between the monomer-activated ROP initiation (cf . eqn (5)) and
the coordination-insertion initiation inherent in tris(phenolate)
complexes in general (cf. the first two entries in Table 4).
Conclusion
A number of new mononuclear, heteroleptic lanthanoid arylox-
ides, [Ln(dbp)3(thf)3] (1–5) and [Nd(dbp)3(dme)2] (6), with unsym-
metrical substitution has been synthesised by redox transmetalla-
tion/protolysis and fully characterised, further establishing the
generality of this synthetic method. Thf could be removed to an
extent by recrystallisation from toluene, resulting in the formation
of a five-coordinate dinuclear species, [Ln2(dbp)6(thf)2] (7, 8).
Fig. 9 Mn (as measured by GPC) vs. [rac-LA]t : [BnNH2]0 for the ROP
of rac-LA (200 equivs. at t = 0) using 3 and 3, 5, 7, or 10 equivs. BnNH2
co-initiator. The numbers in parentheses are the PDIs. The gradient of the
fitted line (R2 = 0.985) is 176(5) g mol-1 with an intercept of 107.1 g mol-1
(corresponding to the benzylamine-derived end groups).
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 6693–6704 | 6699
©

Similar species, 9 and 10, could be obtained when the reaction
was performed in diethyl ether, or in the absence of a solvent. The
coordination numbers, generally 5 and 6, in the structures of these
complexes are low, but higher than for aryloxides with two o-But
substituents.4,5,10,15b,15c,47–49
Complexes 3 and 4 were efficient but poorly-controlled initiators
for the ROP of rac-lactide, widening the scope of aryloxide initia-
tors to those with bulky unsymmetrical substitution. Improved
catalyst systems were obtained using an excess of BnOH as
a co-initiator which showed features consistent with immortal
polymerisation. Use of BnNH2 led to well-controlled, amine-
initiated immortal ROP of rac-lactide, only the second example
of this type of process for a group 3 or lanthanoid system.
lactide was kept constant (0.7 M). The mass of 3 and the volume
of BnNH2 were varied accordingly to observe the relationships
between Mn and [rac-lactide]0 : [BnNH2]0 : [3]0. Polymer molecular
weights (Mn, Mw) were determined by GPC using a Polymer
Laboratories Plgel Mixed-D column (300 mm length, 7.5 mm
diameter) and a Polymer Laboratories PLGPC50 Plus instrument
equipped with a refractive index detector. Thf (HPLC grade)
◦
was used as an eluent at 30 C with a rate of 1 mL min-1.
Linear polystyrenes were used as primary calibration standards,
and Mark–Houwink corrections41–43 for poly(rac-LA) in thf were
applied for the experimental samples.
MALDI-ToF mass spectra were measured using a Waters
MALDI micro equipped with a 337 nm nitrogen laser. An
accelerating voltage of 25 kV was applied. The polymer sam-
ples were dissolved in thf at a concentration of 1 mg mL-1.
The cationization agent used was potassium trifluoroacetate
(Fluka, > 99%) dissolved in thf at a concentration of 5 mg mL-1.
The matrix used was trans-2-[3-(4-tert-butylphenyl)-2-methyl-2-
propenylidene]malononitrile (DCTB) (Fluka) and was dissolved
in thf at a concentration of 40 mg mL-1. Solutions of matrix, salt
and polymer were mixed in a volume ratio of 4 : 1 : 4, respectively.
The mixed solution was hand-spotted on a stainless steel MALDI
target and left to dry. The spectra were recorded in the reflectron
mode.
Experimental
All products were air-sensitive and required the use of Schlenk and
vacuum line techniques, and were manipulated under a nitrogen
atmosphere. Solvents were dried by distillation over sodium or
sodium/benzophenone. Thf used for polymerisation was dried
over potassium at reflux and distilled. All solvents were stored
in J. Young Teflon valve ampoules. Elemental analyses (C, H)
were performed by the Campbell Microanalytical Laboratory,
University of Otago, Dunedin, New Zealand. Metal analyses were
by EDTA titrations with Xylenol Orange indicator, following acid
digestion (HCl or H2SO4/HNO3) and buffering with hexamine.
Infrared spectra were obtained as Nujol mulls with a Perkin
Elmer 1600 FTIR or Perkin Elmer Spectrum RX1 instrument.
1H and 19F NMR spectra were recorded on a Bruker AM300MHz
spectrometer.
2,4-di-tert-Butylphenol (dbpH) was purchased from Sigma-
Aldrich, and was not further purified. Bis(pentafluorophenyl)-
mercury was synthesised according to the published procedure.50
Lanthanoid metals were obtained fromRhone Poulenc or Santoku
and stored under nitrogen and freshly filed before use (where
indicated).
[La(dbp)3(thf)3] (1)
Lanthanum metal powder (freshly filed) (0.24 g, 1.7 mmol),
bis(pentafluorophenyl)mercury (0.50 g, 0.92 mmol), 2,4-di-tert-
butylphenol (0.48 g, 2.3 mmol), and Hg metal (~2 drops), were
added to a Schlenk flask along with dry thf (30 ml), and placed in
an ultrasound bath for ~144 h. A faint brown coloured solution
was observed. The solution was filtered and concentrated under
vacuum until a faint pink oil remained. Large clear block crystals
grew at -4 ◦C over a period of 30 days. (0.35 g, 47%); m.p: 108–
111 ◦C; Found: C, 65.92%; H, 8.77%; La, 14.00%. C54H87O6La
(971.2 g mol-1) requires: C, 66.80%; H, 9.03%; La, 14.30%; IR
(Nujol mull, cm-1): 1862 w, 1769 w, 1641 w, 1597 s, 1546 w, 1525
m, 1483 s, 1402 s, 1381 s, 1360 s, 1348 m, 1258 s, 1201 s, 1148 m,
1120 m, 1085 s, 1030 s, 907 s, 872 s, 834 s, 736 s, 670 s, 637 m; 1H
NMR (C6D6): 7.48 (s, 3H, ArH); 7.25 (d, 3H, ArH); 7.10 (s, 3H,
ArH); 3.48 (s, 12H, thf); 1.62 (s, 27H, p-But); 1.34 (s, 27H, o-But);
1.26 (s, 12H, thf).
Polymerisation reactions
All manipulations were performed in a glove-box under a dinitro-
gen atmosphere at ambient temperature. A catalyst solution was
prepared by dissolving the metal complex (1.39 ¥ 10-2 mmol) in
thf (0.5 mL). A monomer solution was prepared by dissolving
rac-lactide (100 equivs., 0.20 g, 1.39 mmol) in thf (1.5 mL).
BnOH (5 equivs., 6.95 ¥ 10-2 mmol) was added to the metal
complex solution where indicated. The monomer solution was
added to the catalyst solution with rapid stirring. An aliquot
of the mixture was taken after 30 min, quenched with a small
amount of wet thf, and evaporated to dryness under vacuum.
[Pr(dbp)3(thf)3] (2)
A similar preparation from praseodymium metal pieces (0.43 g, 3.1
mmol), bis(pentafluorophenyl)mercury (0.53 g, 1.0 mmol), 2,4-di-
tert-butylphenol (0.49 g, 2.4 mmol), and Hg metal (~2 drops), in
dry thf (30 ml) with ultrasonication for ~144 h gave a grey coloured
solution. After filtration and concentration under vacuum, an oil
remained, which gav◦e pale green block crystals in 1 week. (0.40 g,
52%) m.p: 178–182 C; Found: Pr, 14.21%. C54H87O6Pr (973.2 g
mol-1) requires: Pr, 14.48%; IR (Nujol mull, cm-1): 1862 w, 1769 w,
1641 w, 1597 m, 1525 m, 1402 m, 1360 m, 1348 m, 1257 m, 1202
m, 1147 m, 1120 m, 1085 m, 1030 m, 907 m, 886 m, 834 m, 735
m, 670 m, 637 m; 1H NMR (C6D6): 21.25 (br s, 3H, ArH); 13.71
(br s, 3H, ArH); 12.08 (br s, 3H, ArH); 9.16 (br s, 12H, thf); 3.41
(br s, 27H, p-But); -5.60 (br s, 27H, o-But); -9.16 (br s, 12H, thf).
1
The completeness was monitored by H NMR spectroscopy. A
representative procedure for the extended studies involving 3,
BnNH2 and rac-lactide is as follows. Rac-lactide (200 equivs.,
0.40 g, 2.78 mmol) was dissolved in thf (3.5 mL). The monomer
solution was added quickly to a mixture of 3 (1 equiv., 1.40 ¥ 10-2 g,
1.39 ¥ 10-2 mmol) and BnNH2 (3 equivs., 4.5 ¥ 10-3 mL, 4.16 ¥
10-2 mmol) dissolved in thf (0.5 mL) with rapid stirring. After
10 min the reaction was quenched with a small quantity of wet thf.
Aliquots were then taken and evaporated to dryness for analysis. In
subsequent polymerisation runs, the initial concentration of rac-
6700 | Dalton Trans., 2010, 39, 6693–6704
This journal is
The Royal Society of Chemistry 2010
©

[Nd(dbp)3(thf)3] (3)
vacuum giving pale blue block crystals over 7 days. (0.21 g, 26%);
(Found: Nd, 15.51%. C50H83O7Nd (940.43 g mol-1) requires Nd,
15.34%); IR (Nujol mull, cm-1): 1598 m, 1526 w, 1401 m, 1360 m,
1260 m, 1201 m, 1144 m, 1086 m, 1048 m, 1011 m, 971 m, 906
w, 887 w, 859 m, 820 m, 754 w, 733 m, 669 m, 638 w; H NMR
(C6D6): 18.98 (br s, 9H, ArH); 15.11 (br s, 20H, dme); 4.95 (br s,
54H, But)
Neodymium metal filings (0.40 g, 2.8 mmol), bis(penta-
fluorophenyl)mercury (0.53 g, 0.99 mmol), 2,4-di-tert-butylphenol
(0.49 g, 2.4 mmol), and Hg metal (~2 drops) in dry thf
(30 ml), ultrasonicated for 144 h gave a blue/grey coloured
solution. Filtration and concentration under vacuum gave an oil
from which large translucent bl◦ue block-like crystals (0.42 g, 54%)
grew in 2 days; m.p: 110–113 C; Found: C, 65.89%; H, 9.00%;
Nd, 14.20%. C54H87O6Nd (976.51 g mol-1) requires C, 66.40; H,
8.92%; Nd, 14.77%; IR (Nujol mull, cm-1): 1862 w, 1769 w, 1641
w, 1597 s, 1546 w, 1525 m, 1483 s, 1402 s, 1382 s, 1360 s, 1348 s,
1257 s, 1202 s, 1147 s, 1120 s, 1085 s, 1030 s, 907 s, 871 s, 834 s,
736 s, 670 s, 637 m; 1H NMR (C6D6): 13.59 (br s, 3H, ArH); 13.10
(br s, 3H, ArH); 12.20 (br s, 3H, ArH); 6.34 (br s, 12H, thf); 2.23
(br s, 27H, p-But); -1.87 (br s, 27H, o-But); -2.39 (br s, 12H, thf).
1
[Nd2(dbp)6(thf)2].2C7H8 (7.2C7H8)
Nd(dbp)3(thf)3 was dissolved in toluene (20 ml) and was placed
in an ultrasound bath for ca. 1 h. The solvent was then slowly
evaporated until crystals could be seen to form. Small pale blue
rectangular crystals grew over several days at -4 ◦C in quantitative
yield; m.p: 198–202 ◦C; Found: C, 65.96%; H, 8.67%; Nd, 16.90%.
C92H142O8Nd2 (1664.58 g mol-1) (loss of lattice toluene) requires
C, 66.30%; H, 8.65%; Nd, 17.33%; IR (Nujol mull, cm-1): 1602 m,
1529 w, 1486 s, 1402 s, 1361 s, 1260 s, 1226 s, 1202 m, 1177 w, 1151
m, 1124 m, 1085 s, 1018 m, 909 m, 887 m, 837 s, 822 s, 739 m, 727
m, 694 w, 674 m, 642 m; 1H NMR (C6D6): 32.61 (br s, 12H, ArH-
terminal phenolates); 13.74 (br s, 6H, ArH-bridging phenolates);
-2.66 (br s, 72H, But-terminal phenolates); -4.82 (br s, 36H, But-
bridging phenolates); -8.41 (br s, 16H, thf).
[Gd(dbp)3(thf)3].0.5hdbp (4.0.5hdbp)
Gadolinium metal powder (0.65 g, 4.1 mmol), bis(penta-
fluorophenyl)mercury (0.67 g, 1.3 mmol), 2,4-di-tert-butylphenol
(0.61 g, 3.0 mmol), and Hg metal (~2 drops), in dry thf (30 ml),
ultrasonicated for 5 days gave a red/brown coloured solution.
Filtration and concentration under vacuum yielded an oil from
which large irregular orange/brown crystals (0.53 g, 54%) grew
in 1 week; m.p: 263–268 ◦C; Found: C, 66.68%; H, 8.93%; Gd,
14.36%. C61H98O6.5Gd (1092.7 g mol-1) requires: C, 67.05%; H,
9.04%; Gd, 14.39%; IR (Nujol mull, cm-1): 3501 m(br), 3334 m(br),
2555 w, 1863 m, 1817 w, 1771 m, 1643 m, 1600 s, 1455 s, 1402 s,
1359 s, 1259 s, 1201 s, 1150 s, 1122 s, 1085 s, 1026 s, 955 m, 908 s,
887 s, 819 s, 737 s, 671 s, 638 m, 570 m. 1H NMR spectrum could
not be confidently assigned due to paramagnetic broadening.
[Er2(dbp)6(thf)2] (8)
Er(dbp)3(thf)3 was dissolved in toluene (20 ml) giving
a
pink/brown solution, and placed in the ultrasound bath for ca.
1 h. The solvent was then slowly evaporated until crystals could
be seen to be forming. Small pale pink rectangular crystals grew
over several days at -4 ◦C in quantitative yield; m.p: 195–201 ◦C;
(Found: Er, 19.79%. C92H142O8Er2 (1710.62 g mol-1) requires Er,
19.56%); IR (Nujol mull, cm-1): 2031 w, 1938 w, 1858 w, 1776 w,
1603 m, 1526 m, 1402 m, 1261 m, 1216 m, 1202 m, 1177 w, 1152 m,
1125 m, 1085 m, 1017 m, 911 m, 888 w, 868 w, 840 m, 819 m, 727 s,
694 m, 674 w, 641 m; 1H NMR spectrum could not be confidently
assigned due to paramagnetic broadening.
[Er(dbp)3(thf)3] (5)
Erbium metal powder (0.70 g, 3.9 mmol), 2,4-di-tert-butylphenol
(0.51 g, 2.47 mmol), bis-pentafluorophenyl mercury (0.64 g,
1.20 mmol), and mercury metal (~2 drops) in dry thf (30 ml),
ultrasonicated overnight gave a yellow/brown solution which
upon filtration and concentration under vacuum yielded large pink
block crystals. (0.34 g, 41%); m.p: 94–100 ◦C; Found: C, 63.34%;
H, 8.63%; Er, 16.59%. C54H87O6Er (999.52 g mol-1) requires: C,
64.89%; H, 8.77%; Er, 16.73%; IR (Nujol mull, cm-1): 1877 w,
1769 w, 1643 w, 1600 s, 1531 s, 1276 s, 1202 m, 1180 m, 1148 m,
1122 w, 1087 m, 1072 m, 1021 m, 956 m, 941 m, 910 m, 870 m, 839
m, 756 w, 739 m, 718 m, 670 m, 639 m, 584 m; 1H NMR (C6D6):
52.09 (br s, 9H, ArH); 25.89 (br s, 24H, thf); -9.31 (s, 27H, p-But);
-25.76 (br s, 27H, o-But). Although the single crystal retained
0.5thf of solvation at 123 K for its X-ray structure determination,
this was readily lost, e.g. at room temperature before weighing for
microanalysis or the metal analysis.
[Nd2(dbp)6(Et2O)2] (9)
Similarly, neodymium metal filings (0.50 g, 3.5 mmol),
bis(pentafluorophenyl)mercury (0.67 g, 1.3 mmol), 2,4-di-tert-
butylphenol (0.51 g, 2.5 mmol), and Hg metal (~2 drops), in
dry diethyl ether (30 ml), were stirred at room temperature for
12 h to give a dark blue coloured solution. Small pale blue
crystals suitable for X-ray crystallography grew directly from the
reaction flask. The solution was filtered and concentrated under
vacuum for further characterisation. (0.31 g, 32%); m.p: 192–
199 ◦C; (Found: Nd, 17.23%. C92H146O8Nd2 (1668.62 g mol-1)
requires Nd, 17.29%); IR (Nujol mull, cm-1): 1769 w, 1682 w,
1644 w, 1601 s, 1532 s, 1460 s, 1377 s, 1261 s, 1149 m, 1086 m,
[Nd(dbp)3(dme)2] (6)
1
Neodymium metal filings (0.4 g, 2.4 mmol), bis(penta-
fluorophenyl)mercury (0.51 g, 0.95 mmol), 2,4-di-tert-butylphenol
(0.54 g, 2.6 mmol), and Hg metal (~2 drops), were added to
a Schlenk flask along with dry dme (30 ml), and placed in an
ultrasound bath for 72 h. A dark blue/green coloured solution
was observed. The solution was filtered and concentrated under
955 m, 909 m, 887 m, 821 s, 727 s, 674 m, 641 m; H NMR
spectrum could not be confidently assigned due to broadening. It
has been reported that the paramagnetism of Nd3+ complexes is
often enhanced in dinuclear complexes.15a In other cases, similar
complexes exhibit differing degrees of broadening and shifting of
resonances.48
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 6693–6704 | 6701
©

[Nd2(dbp)6(dbpH)2] (10)
were initially processed and corrected for absorption using the
Bruker Apex II program suite,51 or the programs DENZO52 and
SORTAV.53 The structures were solved using direct methods,
and observed reflections were used in least squares refinement
on F2, with anisotropic thermal parameters refined for non-
hydrogen atoms. Hydrogen atoms were constrained in calculated
positions and refined with a riding model. Structure solutions
and refinements were performed using the programs SHELXS-
9754 and SHELXL-9755 through the graphical interface X-Seed,56
which was also used to generate the figures.
General Variata: All compounds exhibit a relatively large ratio
of ADP max/min for carbon atoms, due to the contrast between
rigid aromatic carbons and the carbons of the But groups which
have a greater potential for free rotation. To convey the true
nature of the ellipsoids of these atoms they have been refined
anisotropically with no restraints. Restraining these atoms would
not reflect the variable extremes of the thermal envelopes of all
carbon atoms in these compounds. In extreme cases, disorder has
been modelled as best as possible. Specific variata for compounds
are mentioned below.
Neodymium metal powder (0.52 g, 3.64 mmol), 2,4-di-tert-
butylphenol (0.50 g, 2.42 mmol) and mercury (~2 drops) were
heated to 130 ◦C under vacuum in a sealed glass tube for
24 h. Pale blue crystals were obtained directly from the tube.
After obtaining the crystal structure, the remaining crystals were
separated from excess metal by extracting into hot toluene,
filtration, and evaporation to dryness. (0.27 g, 47%); Found: Nd,
13.98%. C112H170O8Nd2 (1932.96 g mol-1) requires Nd, 14.92%; IR
(Nujol mull, cm-1): 3610 s, 3558 s, 3378 m, 1939 m, 1871 m, 1799
m, 1735 w, 1604 s, 1564 s, 1505 s, 1392 s, 1361 s, 1331 m, 1260 s,
1176 s, 1110 s, 1082 s, 1029 s, 920 m, 856 s, 828 s, 728 s, 694 s, 678 s,
1
644 m; H NMR (C6D6): 7.47 (s, 8H, ArH); 7.04 (m, 8H, ArH);
6.06 (s, 8H, ArH); 3.77 (brs, 2H, OH); 1.56 (s, 72H, p-But); 1.30
(s, 72H, o-But).
Additional experimental data
Attempted synthesis of [Nd(dbp)3(thf)3] in the absence of Hg(C6F5)2
Neodymium metal filings (0.27 g, 1.87 mmol), 2,4-di-tert-
butylphenol (0.51 g, 2.47 mmol), and Hg metal (~2 drops), were
added to a Schlenk flask along with dry thf (30 ml), and placed in
an ultrasound bath for 48 h. No colour change was observed in
the solution. Filtering and concentration under vacuum yielded
only 2,4-di-tert-butylphenol.
Crystal data for [La(dbp)3(thf)3]. C54H87LaO6, M = 971.15,
0.28 ¥ 0.20 ¥ 0.08 mm3, monoclinic, space group P21/c (No. 14),
◦
˚
a = 19.616(4), b = 28.198(6), c = 19.521(4) A, b = 93.23(3) , V =
3
-3
˚
10780(4) A , Z = 8, Dc = 1.197 g cm , F000 = 4128, Nonius KAPPA
CCD, 2qmax = 50.0◦, 27 592 reflections collected, 14 832 unique
(Rint = 0.0679). Final GooF = 1.026, R1 = 0.0519, wR2 = 0.1311,
R indices based on 11 587 reflections with I > 2s(I) (refinement on
F2), 1176 parameters, 0 restraints. Lp and absorption corrections
applied, m = 0.836 mm-1.
Attempted synthesis of [Nd(dbp)3] in toluene
Neodymium metal filings (0.50 g, 3.5 mmol), bis(penta-
fluorophenyl)mercury (0.67 g, 1.3 mmol), 2,4-di-tert-butylphenol
(0.51 g, 2.5 mmol), and Hg metal (~2 drops), were added to a
Schlenk flask along with dry toluene (30 ml), and placed in an
ultrasound bath for 48 h. No colour change was observed. 19F
NMR showed no conversion of bis(pentafluorophenyl)mercury
to pentafluorobenzene.
Variata: The crystals were of poor quality, possibly twinned,
resulting in a large number of rejected reflections and low
completeness (78%). However, this compound is isostructural with
compounds 2 and 3.
Crystal data for [Pr(dbp)3(thf)3]. C54H87O6Pr, M = 973.15,
green block, 0.08 ¥ 0.05 ¥ 0.05 mm3, monoclinic, space group
˚
P21/c (No. 14), a = 19.575(4), b = 28.185(6), c = 19.489(4) A, b =
Attempted synthesis of [Ln(dbp)3(thf)3]; Ln = Sm, Dy, Ho, Yb, Y
◦
3
-3
˚
93.19(3) , V = 10736(4) A , Z = 8, Dc = 1.204 g cm , F000 = 4144,
Nonius KappaCCD, 2qmax = 55.0◦, 112 224 reflections collected,
24 565 unique (Rint = 0.1339). Final GooF = 1.103, R1 = 0.0597,
wR2 = 0.1365, R indices based on 15 392 reflections with I >
2s(I) (refinement on F2), 1175 parameters, 0 restraints. Lp and
absorption corrections applied, m = 0.952 mm-1.
Lanthanoid metal powder (0.70 g), bis(pentafluoro-
phenyl)mercury (0.67 g, 1.25 mmol), 2,4-di-tert-butylphenol
(0.50 g, 2.42 mmol), and Hg metal (~2 drops), were added to
a Schlenk flask along with dry thf (30 ml), and placed in an
ultrasound bath for 48 h. The resulting solution (appropriately
coloured for Ln = Sm, Ho, Yb) was filtered and concentrated
under vacuum, yielding an oil that failed to crystallise; IR (Nujol
mull, cm-1): 1729 w, 1602 w, 1488 s, 1362 m, 1287 m, 1262 m, 1202
w, 1177 w, 1151 w, 1120 w, 1087 m, 1021 w, 913 w, 844 w, 821
m, 668 w. 19F NMR spectra of representative reaction mixtures
showed consumption of the mercurial and formation of C6F6H.
Crystal data for [Nd(dbp)3(thf)3]. C54H87NdO6, M = 976.48,
blue block, 0.25 ¥ 0.20 ¥ 0.08 mm3, monoclinic, space group
˚
P21/c (No. 14), a = 19.5150(4), b = 28.1539(7), c = 19.4626(5) A,
◦
3
-3
˚
b = 93.238(1) , V = 10676.1(4) A , Z = 8, Dc = 1.215 g cm ,
F000 = 4152, Nonius KappaCCD, 2qmax = 55.0◦, 117 868 reflections
collected, 24 463 unique (Rint = 0.0562). Final GooF = 1.125, R1 =
0.0447, wR2 = 0.0763, R indices based on 20 451 reflections with
I > 2s(I) (refinement on F2), 1176 parameters, 1 restraint. Lp and
absorption corrections applied, m = 1.017 mm-1.
X-Ray structure determinations†
Intensity data were collected using a Bruker X8 Apex II CCD or
an Enraf-Nonius KAPPA CCD at 123 K with Mo-Ka radiation
Variata: Elongated thermal ellipsoid of C98 indicates envelope
disorder. However, model was unsatisfactory when attempting to
model disorder. There is an abnormal thermal ellipsoid of C41,
but we deemed it non-sensible to model disorder of a single C
atom when But group can freely rotate. Occupancies of C110 and
C111 were fixed after trial refinement.
˚
(l = 0.71073 A). Suitable crystals were immersed in viscous
hydrocarbon oil and mounted on a glass fiber which was mounted
on the diffractometer. Using psi and omega scans, Nt (total)
reflections were measured, which were reduced to No. of unique
reflections, with Fo > 2s(Fo) being considered observed. Data
6702 | Dalton Trans., 2010, 39, 6693–6704
This journal is
The Royal Society of Chemistry 2010
©

Crystal data for [Gd(dbp)3(thf)3].0.5hdbp. C61H98GdO6.5, M =
1092.64, colourless block, 0.25 ¥ 0.15 ¥ 0.04 mm3, monoclinic,
space group P21/n (No. 14), a = 24.1258(14), b = 9.3193(6), c =
reflections with I > 2s(I) (refinement on F2), 506 parameters, 62
restraints. Lp and absorption corrections applied, m = 1.897 mm-1.
Crystal data for [Nd2(dbp)6(Et2O)2]. C92H146Nd2O8, M =
1668.57, blue block, 0.07 ¥ 0.05 ¥ 0.05 mm3, monoclinic, space
group P21/c (No. 14), a = 14.3151(4), b = 28.3247(7), c =
◦
3
˚
˚
27.4879(15) A, b = 105.778(2) , V = 5947.4(6) A , Z = 4, Dc =
1.220 g cm-3, F000 = 2320, Bruker X8 Apex II CCD, 2qmax = 50.0◦,
72 464 reflections collected, 10 451 unique (Rint = 0.0836). Final
GooF = 1.220, R1 = 0.0626, wR2 = 0.1288, R indices based on 8825
reflections with I > 2s(I) (refinement on F2), 729 parameters, 36
restraints. Lp and absorption corrections applied, m = 1.162 mm-1.
Variata: The lattice phenol is disordered over a symmetry site
such that the components overlap. Anisotropic refinement of C-
atoms restrained gave the best solution for modelling disorder.
◦
3
˚
˚
11.5266(3) A, b = 105.602(1) , V = 4501.5(2) A , Z = 2, Dc =
1.231 g cm-3, F000 = 1764, Bruker X8 Apex II CCD, 2qmax = 55.0◦,
30 589 reflections collected, 10 202 unique (Rint = 0.0492). Final
GooF = 1.207, R1 = 0.0501, wR2 = 0.0791, R indices based on 8730
reflections with I > 2s(I) (refinement on F2), 480 parameters, 0
restraints. Lp and absorption corrections applied, m = 1.191 mm-1.
Crystal data for [Nd2(dbp)6(dbpH)2]. C112H170Nd2O8, M =
Crystal data for [Er(dbp)3(thf)3].0.5thf. C56H91ErO6.5, M =
1932.96, blue block, 0.07 ¥ 0.05 ¥ 0.05 mm3, triclinic, space group
1035.55, pink block, 0.10 ¥ 0.07 ¥ 0.07 mm3, trigonal, space
¯
˚
P1 (No. 2), a = 11.472(2), b = 14.196(3), c = 18.464(4) A, a =
˚
group P3c1 (No. 158), a = b = 24.9188(2), c = 16.7294(3) A,
◦
3
˚
73.39(3), b = 82.08(3), g = 68.85(3) , V = 2685.4(9) A , Z = 1,
3
-3
˚
V = 8996.32(19) A , Z = 6, Dc = 1.147 g cm , F000 = 3282,
Bruker X8 Apex II CCD, 2qmax = 50.0◦, 81 382 reflections
collected, 9725 unique (Rint = 0.0453). Final GooF = 1.187, R1 =
0.0703, wR2 = 0.1387, R indices based on 8872 reflections with
I > 2s(I) (refinement on F2), 283 parameters, 29 restraints. Lp
and absorption corrections applied, m = 1.441 mm-1. Absolute
structure parameter = 0.00(4) (Flack, H. D. Acta Cryst. 1983,
A39, 876–881).
Dc = 1.195 g cm-3, F000 = 1026, Nonius KappaCCD, 2qmax = 50.0◦,
16 914 reflections collected, 7603 unique (Rint = 0.2217). Final
GooF = 1.044, R1 = 0.1350, wR2 = 0.2212, R indices based on 3405
reflections with I > 2s(I) (refinement on F2), 269 parameters, 0
restraints. Lp and absorption corrections applied, m = 1.007 mm-1.
Variata: The crystals were of poor quality, possibly twinned,
resulting in a large number of rejected reflections and low
completeness (71%). As such, this compound is presented in terms
of connectivity only. The lattice thf was refined with isotropic C
and O atoms due to high thermal motion. Selected bond lengths
and angles are given in supplementary data.†
Variata: Attempts to include more atoms in the anisotropic
model gave a less satisfactory refinement. Structure is presented
for connectivity only, so a full anisotropic refinement was deemed
unnecessary. Selected bond lengths and angles are given in
supplementary data. Repeated attempts to obtain good quality
crystals were unsuccessful. For example use of thf/hexane gave
compound 8 with loss of thf.
Acknowledgements
We are grateful to the Australian Research Council for support and
Dr Joaquim Marc¸alo (Instituto Tecnolo´gico e Nuclear, Sacave´m,
Portugal) for calculation of the steric coordination number of dbp.
Monash Research Graduate School provided travel funds for JPT
to visit Oxford University.
Crystal data for [Nd(dbp)3(dme)2]. C50H83NdO7, M = 940.40,
pale blue block, 0.20 ¥ 0.10 ¥ 0.10 mm3, monoclinic, space group
˚
P21 (No. 4), a = 15.866(3), b = 22.379(5), c = 16.364(3) A,
◦
3
-3
˚
b = 115.92(3) , V = 5225.7(18) A , Z = 4, Dc = 1.195 g cm ,
F000 = 1996, Nonius KappaCCD, 2qmax = 55.0◦, 55 029 reflections
collected, 23 311 unique (Rint = 0.1392). Final GooF = 1.032, R1 =
0.0726, wR2 = 0.1407, R indices based on 14 258 reflections with
I > 2s(I) (refinement on F2), 1084 parameters, 1 restraint. Lp
and absorption corrections applied, m = 1.038 mm-1. Absolute
structure parameter = -0.001(14) (Flack, H. D. Acta Cryst. 1983,
A39, 876–881).
References
1 D. C. Bradley, R. C. Mehrotra, I. P. Rothwell and A. Singh, Alkoxo
and Aryloxo Derivatives of Metals, Academic Press, London, 2001.
2 T. J. Boyle and L. A. M. Ottley, Chem. Rev., 2008, 108, 1896.
3 N. Y. Turova, E. P. Turevskaya, V. G. Kessler and M. I. Yanovskaya, The
Chemistry of Metal Alkoxides, Kluwer Academic Publishers, Boston,
2002.
4 P. B. Hitchcock, M. F. Lappert and A. Singh, J. Chem. Soc., Chem.
Crystal data for [Nd2(dbp)6(thf)2].2C7H8. C106H158Nd2O8, M =
Commun., 1983, 1499.
1848.80, blue block, 0.20 ¥ 0.07 ¥ 0.06 mm3, triclinic, space group
5 M. F. Lappert, A. Singh and R. G. Smith, Inorg. Synth., 1990, 27, 164.
6 G. B. Deacon, C. M. Forsyth, R. Harika, P. C. Junk, J. W. Ziller and
W. J. E va n s , J. Mater. Chem., 2004, 14, 3144.
7 M. N. Bochkarev, L. N. Zakharov and G. S. Kalinina, Organoderiva-
tives of the Rare Earth Elements, Kluwer, Dordrecht, 1995.
8 S. A. Cotton, in Comprehensive Coordination Chemistry II, ed. J. A.
McCleverty and T. J. Meyer, Elsevier, Oxford, 2004, vol. 3, ch. 3.2, p.
93.
¯
˚
P1 (No. 2), a = 11.226(2), b = 14.321(3), c = 16.832(3) A, a =
◦
3
˚
83.24(3), b = 73.97(3), g = 73.61(3) , V = 2492.9(9) A , Z = 1,
Dc = 1.231 g cm-3, F000 = 978, Nonius KappaCCD, 2qmax = 55.0◦,
34 572 reflections collected, 11 295 unique (Rint = 0.0844). Final
GooF = 1.172, R1 = 0.0630, wR2 = 0.1412, R indices based on 9523
reflections with I > 2s(I) (refinement on F2), 542 parameters, 0
restraints. Lp and absorption corrections applied, m = 1.082 mm-1.
9 T. J. Boyle, L. J. Tribby and S. D. Bunge, Eur. J. Inorg. Chem., 2006,
4553.
10 G. B. Deacon, G. D. Fallon, C. M. Forsyth, S. C. Harris, P. C. Junk,
B. W. Skelton and A. H. White, Dalton Trans., 2006, 802.
11 G. B. Deacon and C. M. Forsyth, in Inorg. Chem. Highlights, ed.
G. Meyer, D. Naumann and L. Wesemann, WILEY-VCH Verlag
CmbH & Co., 2002, ch. 7, pp. 139.
Crystal data for [Er2(dbp)6(thf)2]. C92H142Er2O8, M = 1710.58,
3
¯
pink block, 0.10 ¥ 0.07 ¥ 0.06 mm , triclinic, space group P1 (No.
˚
˚
2), a = 11.6509(2), b = 14.0152(2), c = 14.7135(2) A, a = 72.387(1),
◦
3
b = 82.619(1), g = 81.695(1) , V = 2256.75(6) A , Z = 1, Dc =
12 G. B. Deacon, C. M. Forsyth and S. Nickel, J. Organomet. Chem., 2002,
1.259 g cm-3, F000 = 894, Bruker X8 Apex II CCD, 2qmax = 55.0◦,
29 373 reflections collected, 10 298 unique (Rint = 0.0339). Final
GooF = 1.066, R1 = 0.0319, wR2 = 0.0746, R indices based on 9397
647, 50.
13 G. B. Deacon, E. E. Delbridge, B. W. Skelton and A. H. White,
Eur. J. Inorg. Chem., 1998, 543.
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 6693–6704 | 6703
©

14 G. B. Deacon, P. E. Fanwick, A. Gitlits, I. P. Rothwell, B. W. Skelton
and A. H. White, Eur. J. Inorg. Chem., 2001, 1505.
15 (a) G. B. Deacon, P. C. Junk and G. J. Moxey, Chem.–Asian J., 2009, 4,
1717; (b) G. B. Deacon, P. B. Hitchcock, S. A. Holmes, M. F. Lappert,
P. MacKinnon and R. H. Newnham, J. Chem. Soc., Chem. Commun.,
1989, 935; (c) G. B. Deacon, T. Feng, P. MacKinnon, R. H. Newnham,
S. Nickel, B. W. Skelton and A. H. White, Aust. J. Chem., 1993, 46, 387.
16 Z. M. Hou, A. Fujita, Y. G. Zhang, T. Miyano, H. Yamazaki and Y.
Wakatsuki, J. Am. Chem. Soc., 1998, 120, 754.
17 R. P. Davies, C. R. Abernathy, S. J. Pearton, D. P. Norton, M. P. Ivill
and F. Ren, Chem. Eng. Commun., 2009, 196, 1030.
18 C. Emo and S. Roberto, Adv. Mater., 1996, 8, 305.
19 K. E. Uhrich, S. M. Cannizzaro, R. S. Langer and K. M. Shakesheff,
Chem. Rev., 1999, 99, 3181.
20 R. E. Drumright, P. R. Gruber and D. E. Henton, Adv. Mater., 2000,
12, 1841.
21 I. Yoshito and T. Hideto, Macromol. Rapid Commun., 2000, 21, 117.
22 M. Stefan, Angew. Chem., Int. Ed., 2004, 43, 1078.
23 C. K. Williams and M. A. Hillmyer, Polym. Rev., 2008, 48, 1.
24 E. S. Place, J. H. George, C. K. Williams and M. M. Stevens, Chem.
Soc. Rev., 2009, 38, 1139.
35 G. B. Deacon, T. C. Feng, C. M. Forsyth, A. Gitlits, D. C. Hockless,
Q. Shen, B. W. Skelton and A. H. White, J. Chem. Soc., Dalton Trans.,
2000, 961.
36 S. Ando and T. Matsuura, Magn. Reson. Chem., 1995, 33, 639.
37 G. B. Deacon, P. C. Junk, G. J. Moxey, M. Guino-o and K. Ruhlandt-
Senge, Dalton Trans., 2009, 4878.
38 R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor.
Gen. Crystallogr., 1976, 32, 751.
39 G. B. Deacon, T. C. Feng, B. W. Skelton and A. H. White, Aust. J. Chem.,
1995, 48, 741.
40 J. Marc¸alo and A. P. de Matos, Polyhedron, 1989, 8, 2431; J. Marc¸alo,
unpublished results for dbp, personal communication.
41 A. Rudin and H. L. W. Hoegy, J. Polym. Sci., Part A1, 1972, 10, 217.
42 I. Barakat, P. Dubois, R. Jerome and P. Teyssie, J. Polym. Sci., Part A:
Polym. Chem., 1993, 31, 505.
43 J. R. Dorgan, J. Janzen, D. M. Knauss, S. B. Hait, B. R. Limoges and
M. H. Hutchinson, J. Polym. Sci., Part B: Polym. Phys., 2005, 43, 3100.
44 A. Amboune, C. M. Thomas and J.-F. Carpentier, Macromol. Rapid
Commun., 2007, 28, 693.
45 L. Clark, M. G. Cushion, H. D. Dyer, A. D. Schwarz, R. Duchateau
and P. Mountford, Chem. Commun., 2010, 46, 273.
25 B. J. O’Keefe, M. A. Hillmyer and W. B. Tolman, J. Chem. Soc., Dalton
Trans., 2001, 2215.
26 O. Dechy-Cabaret, B. Martin-Vaca and D. Bourissou, Chem. Rev., 2004,
104, 6147.
27 R. H. Platel, L. M. Hodgson and C. K. Williams, Polym. Rev., 2008,
48, 11.
28 S. Agarwal, C. Mast, K. Dehnicke and A. Greiner, Macromol. Rapid
Commun., 2000, 21, 195.
29 I. Palard, M. Schappacher, A. Soum and S. M. Guillaume, Polym. Int.,
2006, 55, 1132.
30 A. Amboune, C. M. Thomas and J.-F. Carpentier, Pure Appl. Chem.,
2007, 79, 2013.
31 N. E. Kamber, W. Jeong, R. M. Waymouth, R. C. Pratt, B. G. G.
Lohmeijer and J. L. Hedrick, Chem. Rev., 2007, 107, 5813.
32 W. M. Stevels, M. J. K. Ankone´, P. J. Dijkastra and J. Feijen,
Macromolecules, 1996, 29, 3332.
33 W. M. Stevels, M. J. K. Ankone´, P. J. Dijkastra and J. Feijen,
Macromolecules, 1996, 29, 6132.
46 BnNH2 alone does not polymerise rac-LA under these conditions.
47 G. B. Deacon, T. Feng, S. Nickel, M. I. Ogden and A. H. White,
Aust. J. Chem., 1992, 45, 671.
48 L. Zhang, Y. M. Yao, Y. J. Luo, Q. Shen and J. Sun, Polyhedron, 2000,
19, 2243; G. R. Giesbrecht, J. C. Gordon, D. L. Clark, B. L. Scott, J. G.
Watkin and K. J. Young, Inorg. Chem., 2002, 41, 6372.
49 Q. Z. Qi, Y. H. Lin, J. Y. Hu and Q. Shen, Polyhedron, 1995, 14, 413.
50 G. B. Deacon, J. E. Cosgriff, E. T. Lawrenz, C. M. Forsyth and
D. L. Wilkinson, in Synthetic Methods of Organometallic and Inorganic
Chemistry, ed. W. A. Herrmann, Theime, Stuttgart, 1997, vol. 6,
pp. 48–51.
51 Apex II Version 2.1, Bruker AXS Ltd., Madison, Wisconsin.
52 Z. Otwinowski and W. Minor, Methods Enzymol., 1997, 276, 307.
53 R. H. Blessing, Acta Crystallogr., Sect. A: Found. Crystallogr., 1995,
51, 33.
54 G. M. Sheldrick, SHELXS-97 Program for crystal structure solution,
University of Gottingen, Germany, 1997.
55 G. M. Sheldrick, SHELXL-97 Program for crystal structure refinement,
University of Gottingen, Germany, 1997.
56 L. J. Barbour, J. Supramol. Chem., 2001, 1, 189.
34 C. Yu, L. Zhang, X. Ni, Z. Shen and K. Tu, J. Polym. Sci., Part A:
Polym. Chem., 2004, 42, 6209.
6704 | Dalton Trans., 2010, 39, 6693–6704
This journal is
The Royal Society of Chemistry 2010
©