Stable divalent germanium, tin and lead amino(ether)-phenolate monomeric complexes: Structural features, inclusion heterobimetallic complexes, and ROP catalysis

Article abstract of DOI:10.1039/c3dt51681d

Stable germanium(ii) and lead(ii) amido complexes {LOⁱ} M(N(SiMe₃)₂) (M = Ge^II, Pb^II) bearing amino(ether)phenolate ligands are readily available using the proteo-ligands {LOⁱ}H of general formula 2-CH₂NR ₂-4,6-tBu₂-C₆H₂OH (i = 1, NR ₂ = N((CH₂)₂OCH₃)₂; i = 2, NR₂ = NEt₂; i = 3, NR₂ = aza-15-crown-5) and M(N(SiMe₃)₂)₂ precursors. The molecular structures of these germylenes and plumbylenes, as well as those of {LO ³}GeCl, {LO³}SnCl and of the congeneric {LO ⁴}Sn^II(N(SiMe₃)₂) where NR ₂ = aza-12-crown-4, have been determined crystallographically. All complexes are monomeric, with 3-coordinate metal centres. The phenolate systematically acts as a N^O_phenolate bidentate ligand, with no interactions between the metal and the O_side-arm atoms in these cases (for {LO¹}^-, {LO³}^- and {LO ⁴}^-) where they could potentially arise. For each family, the lone pair of electrons essentially features ns² character, and there is little, if any, hybridization of the valence orbitals. Heterobimetallic complexes {LO³}M(N(SiMe₃)₂)·LiOTf, where the Li⁺ cation sits inside the tethered crown-ether, were prepared by reaction of {LO³}M(N(SiMe₃)₂) and LiOTf (M = Ge^II, Sn^II). The inclusion of Li⁺ (featuring a close contact with the triflate anion) in the macrocycle bears no influence on the coordination sphere of the divalent tetrel element. In association with iPrOH, the amido germylenes, stannylenes and plumbylenes catalyse the controlled polymerisation of l- and racemic lactide. The activity increases linearly according to Ge^II ? Sn^II ? Pb^II. The simple germylenes generate very sluggish catalysts, but the activity is significantly boosted if the heterobimetallic complex {LO ³}Ge(N(SiMe₃)₂)·LiOTf is used instead. On the other hand, with 10-25 equiv. of iPrOH, the plumbylenes afford highly active binary catalysts, converting 1000 or 5000 equiv. of monomer at 60 °C within 3 or 45 min, respectively, in a controlled fashion.

Full text of DOI:10.1039/c3dt51681d

View Article Online
View Journal
Dalton
Transactions
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: Y. Sarazin, L. Wang, S. Rosca, S.
Sinbandhit, V. Poirier, V. Dorcet, J. Carpentier and T. Roisnel, Dalton Trans., 2013, DOI: 10.1039/C3DT51681D.
This is an Accepted Manuscript, which has been through the RSC Publishing peer
Dalton
review process and has been accepted for publication.
Transactions
Volume 39 | Number 3 | 21 January 2010 | Pages 657–964
Accepted Manuscripts are published online shortly after acceptance, which is prior
to technical editing, formatting and proof reading. This free service from RSC
Publishing allows authors to make their results available to the community, in
citable form, before publication of the edited article. This Accepted Manuscript will
be replaced by the edited and formatted Advance Article as soon as this is available.
An international journal of inorganic chemistry
www.rsc.org/dalton
To cite this manuscript please use its permanent Digital Object Identifier (DOI®),
which is identical for all formats of publication.
More information about Accepted Manuscripts can be found in the
Information for Authors.
Please note that technical editing may introduce minor changes to the text and/or
graphics contained in the manuscript submitted by the author(s) which may alter
content, and that the standard Terms & Conditions and the ethical guidelines
that apply to the journal are still applicable. In no event shall the RSC be held
responsible for any errors or omissions in these Accepted Manuscript manuscripts or
any consequences arising from the use of any information contained in them.
ISSN 1477-9226
COMMUNICATION
PAPER
Bu et al.
Manzano et al.
Zinc(ii)-boron(iii)-imidazolate
Experimental and computational study framework (ZBIF) with unusual
of the interplay between C–H/p and
anion–p interactions
pentagonal channels prepared from
deep eutectic solvent
1477-9226(2010)39:1;1-K
www.rsc.org/dalton
Registered Charity Number 207890

Page 1 of 55
Dalton Transactions
View Article Online
DOI: 10.1039/C3DT51681D
Stable Divalent Germanium, Tin and Lead Amino(ether)-phenolate Monomeric
Complexes: Structural Features, Inclusion Heterobimetallic Complexes, and ROP
Catalysis
a
a
a
b
Lingfang Wang, Sorin-Claudiu Roşca, Valentin Poirier, Sourisak Sinbandhit, Vincent
c
c
a
a
Dorcet, Thierry Roisnel, Jean-François Carpentier and Yann Sarazin*
a
Organometallics: Materials and Catalysis, Institut des Sciences Chimiques de Rennes, UMR
226 CNRS – University of Rennes 1, 35042 Cedex, Rennes, France
6
b
Centre Régional des Mesures Physiques de l’Ouest, University of Rennes 1, 35042 Cedex,
Rennes, France
c
Centre de Diffractométrie des Rayons X, Institut des Sciences Chimiques de Rennes, UMR
6
226 CNRS – University of Rennes 1, 35042 Cedex, Rennes, France
1

Dalton Transactions
Page 2 of 55
View Article Online
DOI: 10.1039/C3DT51681D
Abstract
i
II II
Stable germanium(II) and lead(II) amido complexes {LO }M(N(SiMe
bearing amino(ether)phenolate ligands are readily available using the proteo-ligands {LO }H
of general formula 2-CH NR -4,6-tBu -C OH (i = 1, NR = N((CH OCH ; i = 2, NR
NEt ; i = 3, NR = aza-15-crown-5) and M(N(SiMe precursors. The molecular structures
of these germylenes and plumbylenes, as well as those of {LO }GeCl, {LO }SnCl and of the
3
)
2
) (M = Ge , Pb )
i
2
2
2
6
H
2
2
2
)
2
3
)
2
2
=
2
2
3
)
)
2 2
3
3
4
II
congeneric {LO }Sn (N(SiMe
3
)
2
) where NR = aza-12-crown-4, have been determined
2
crystallographically. All complexes are monomeric, with 3-coordinate metal centres. The
phenolate systematically acts as a N^Ophenolate bidentate ligand, with no interactions between
1
−
3
−
4 −
the metal and the Oside-arm atoms in these cases (for {LO } , {LO } and {LO } ) where they
2
could potentially arise. For each family, the lone pair of electrons essentially features ns
character, and there is little, if any, hybridization of the valence orbitals. Heterobimetallic
3
+
complexes {LO }M(N(SiMe
3
) )·LiOTf, where the Li cation sits inside the tethered crown-
2
3
II
II
ether, were prepared by reaction of {LO }M(N(SiMe
3
) ) and LiOTf (M = Ge , Sn ). The
2
+
inclusion of Li (featuring close contact with the triflate anion) into the macrocycle bears no
influence on the coordination sphere of the divalent tetrel element. In association with iPrOH,
the amido germylenes, stannylenes and plumbylenes catalyse the controlled polymerisation of
II
II
II
L- and racemic lactide. The activity increases linearly according to Ge << Sn << Pb . The
simple germylenes generate very sluggish catalysts, but the activity is significantly boosted if
3
the heterobimetallic complex {LO }Ge(N(SiMe
3
2
) )·LiOTf is used instead. On the other hand,
with 10−25 equiv of iPrOH, the plumbylenes afford highly active binary catalysts, converting
,000 or 5,000 equiv of monomer at 60 °C within 3 or 45 min, respectively, in a controlled
fashion.
1
2

Page 3 of 55
Dalton Transactions
View Article Online
DOI: 10.1039/C3DT51681D
Introduction
Poly(L-lactide) is a biopolymer used for a variety of specialty applications and as a bulk
1
polymer. It is conveniently prepared through metal-mediated ring-opening polymerisation
(
ROP) of L-lactide. Following a decade of intense research, a great diversity now exists in the
2
range of metal catalysts available for ROP reactions and related ring-opening processes, with
3
4
5
6
zinc, aluminium and rare-earth elements attracting the most interest. Industrial plants still
7
rely on the use of simple tin(II) systems such as the versatile tin(II) bis(octanoate), an
inexpensive and robust compound considered safe by the US Food and Drug Administration.
8
Considering the popularity of this and other group 14 polymerisation catalysts, it is
9
6e-g,10
11
surprising that only a handful of germanium(II), tin(II)
or even lead(II) ROP catalysts
1
2
have been reported. The canon of ligands employed to tailor ROP catalysts is virtually
boundless, with prominent examples including bulky -diketiminates or a range of phenolate
2
-6
(
salen, salan, amino-phenolates…) ligands. We embarked a few years ago upon the design
of ROP catalysts built on somewhat unconventional metals supported by multidentate
6
j,6l,13
amino(ether)-phenolate ligands,
and explored ROP mechanisms using tin(II)
6
g,14
complexes.
In the course of these investigations, we became involved in the coordination
chemistry of tin(II) and related germanium(II) and lead(II) amino(ether)-phenolate
complexes.
Phenolates are amenable to the tuning of their electronic and steric properties through
modification of the substituents at the ortho and para positions of the aromatic ring, and this
1
5
has led to a rich coordination chemistry. Yet, phenolates have seldom been applied to the
stabilisation of singlet germylene, stannylene or plumbylene species, the heavier homologues
1
6
of divalent carbenes. Unlike alkoxide M(OR)
2
species that are often polymetallic (R =
alkyl), sterically stabilised (and significantly less basic) homoleptic M(OAr) phenolate
or 2,6-
1
7
2
1
2
8
complexes are monomeric for M = Ge and Sn (Ar = 2,6-tBu
2
-4-MeC
6
H
3

Dalton Transactions
Page 4 of 55
1
9
View Article Online
Mes
2
C
6
H
3
). On the other hand, structural information for analogous plumbylenes are scarce,
DOI: 10.1039/C3DT51681D
and it is just recently that the monomeric Pb(OC
6
H
3
-2,6-(2,6-iPr
2
C
6
H
3
)
2
2
) has been
2
0
authenticated in the solid state. The presence of side-arms containing heteroatoms (N, O) is
a contributing factor towards the kinetic stability of three- or four-coordinate divalent
homoleptic complexes formed through intramolecular coordination of the heteroatom(s) onto
2
1
the metal centre. Tetracoordinated complexes (N^O)
2
M supported by chelating
dimethylaminoethoxide or 2,4,6-tris[(dimethylamino)methyl]phenolate ligands were found to
2
2
be monomeric for M = Ge, Sn, but the solid-state structure of the lead(II) derivative was not
2
3
available. The bidentate 2-[(dimethylamino)methyl]phenolate yielded the monomeric
germylene (N^O) Ge, but the tin(II) and Pb(II) derivatives could not be structurally
characterised; the parent heteroleptic complex (N^O)SnCl was dimeric, with Ophenolate atoms
2
2
4
bridging two four-coordinate tin(II) centres. Dimers are commonly observed for tin(II) when
the germanium analogues are monomeric (lead congeners are seldom mentioned), e.g. in
2
5
{
M(OCH
2
CH
2
NMe
2
)(N(SiMe
3
)
2
}
n
(M = Ge, n = 1; M = Sn, n = 2).
We report here the synthesis and characterisation of monomeric, heteroleptic, 3-
coordinate complexes of germanium(II), tin(II) and lead(II) incorporating multidentate,
monoanionic amino(ether)-phenolate ligands, and their behaviour towards the ROP of L-
lactide is introduced. Heterobimetallic complexes prepared by inclusion of alkali salts in the
side-arm of these ligands containing an aza-crown-ether side-arm are also presented.
4

Page 5 of 55
Dalton Transactions
View Article Online
DOI: 10.1039/C3DT51681D
Results and Discussion
Stable, monomeric divalent metal aminophenolates
1
The heteroleptic amido complexes {LO }M(N(SiMe
3
2
) ) (M = Ge, 1; Sn, 5; Pb, 9),
2
3
{
LO }M(N(SiMe
3
)
2
) (M = Ge, 2; Sn, 6; Pb, 10), {LO }M(N(SiMe
3
2
) ) (M = Ge, 3; Sn, 7; Pb,
4
1
1) and {LO }M(N(SiMe
)
3 2
) (M = Sn, 8) where the metal is supported by a monoanionic
1
−
chelating ligand chosen from an amino(ether)-phenolate (as in {LO } ), an amino-phenolate
2
−
3
−
4 −
(
as in {LO } ) or an amino(crown-ether)-phenolate (as in {LO } and {LO } ) are available
in good isolated yields upon protonolysis of the homoleptic divalent metal-amido precursors
i
26
M(N(SiMe
3
)
2
) with the corresponding proteo-ligand {LO }H in diethyl ether (Scheme 1).
2
Scheme 1. Synthesis of the germylenes, stannylenes and plumbylenes 1−11; complexes 5−7
are taken from references 6g and 14a.
5

Dalton Transactions
Page 6 of 55
View Article Online
DOI: 10.1039/C3DT51681D
The convenient protonolysis procedure offered better yields than one-pot salt
2
7
metathesis reactions. For instance, the germylene 3 was isolated after a tedious work-up in
3
only 46% yield from the one-pot reaction of {LO }H, GeCl
2
·dioxane and two equiv of
3
KN(SiMe
3
)
2
. The stoichiometric reaction of {LO }GeCl (a colourless solid accessible upon
3
treatment of GeCl
2
·dioxane with fresh {LO }K) with KN(SiMe
3
2
) ) brought no improvement.
3
3
Similarly, {LO }SnCl was obtained from SnCl
2
and {LO }K but its reaction with
KN(SiMe
3
2
) only led to partial formation of 7 (ca. 60%) together with unidentified species.
3
3
Complexes 1−3, 5−10, {LO }GeCl and {LO }SnCl were isolated as analytically pure
colourless solids. The plumbylene 11 could not be obtained entirely free of the homoleptic
3
{
LO }
2
Pb (a complex otherwise cleanly synthesized by reaction of Pb(N(SiMe
3
)
2
2
) with 2
3
equiv of {LO }H). No reliable synthesis to the germylene 4 could be designed, as intractable
2
8
mixtures were repeatedly recovered. All complexes are soluble in aromatic hydrocarbons
and ethers, and are sparingly so in petroleum ether; they are fully soluble in dichloromethane
2
9
and do not react with this solvent through acid-base reaction. All are stable in aromatic
1
solvents as indicated by H NMR monitoring, bar the kinetically labile 11 which rapidly
3
evolves to generate {LO }
2
Pb and Pb(N(SiMe
3
)
2
2
) . The complexes were characterised by 1D
2
9
1
119
1
and 2D solution NMR methods, including Si{ H} and, where relevant, Sn{ H} and
2
07
1
Pb{ H} NMR spectroscopy. Except for 2, 6 and 10 incorporating the simplest amino-
2
−
1
phenolate {LO } , the H NMR spectra of all complexes in toluene-d
8
or benzene-d
6
showed
high levels of fluxionality at 298 K, which hindered detailed assignment of the resonances for
the side-arm hydrogens atoms; low temperature NMR in toluene-d
8
provided little help.
at 298 K for the amido
groups onto the side-arm
nitrogen atom are not equivalent. In 10 (Figure 1, top), the four N(CH CH methylene
hydrogens are magnetically distinct, each being characterised by its own broad resonance (1H
1
In the H NMR spectra recorded in benzene-d
6
or toluene-d
CH
8
14a
2
14a
complexes 2, 6, 10 and for {LO }SnCl, the two CH
2
3
2
)
3 2
6

Page 7 of 55
Dalton Transactions
View Article Online
=
3.28, 2.84, 2.50 and 2.39 ppm), whereas two broad signals exchanging slowly at 298 K are
DOI: 10.1039/C3DT51681D
found for the two N(CH
2
CH
3
)
2
methyl groups (centred on 1H = 0.74 and 0.57 ppm); this is
indicative of overall C symmetry, which was corroborated by crystallographic studies (vide
1
infra). The fluxionality in 10 could be frozen at low temperature: sharp resonances were
1
detected for all hydrogen atoms in the H NMR spectrum recorded at 263 K (Figure 1,
bottom), and they were readily assigned. At 368 K, the ArCH
2
, NCH
2
CH
3
and NCH
2
CH
3
1
moieties in 10 each gives rise to a single resonance and the H NMR spectrum agrees with
pseudo C symmetry.
s
1
2
Fig. 1 H NMR spectra (400.13 MHz, toluene-d
8
) for {LO }Pb(N(SiMe
3
2
) ) (10) recorded at
2
63 K (bottom) and 298 K (top). Solvent resonances are identified by *.
7

Dalton Transactions
Page 8 of 55
View Article Online
Manual NMR line-shape analysis was performed for 10 in the temperature range 263–
DOI: 10.1039/C3DT51681D
3
63 K, using a 0.1 M solution in toluene-d
reversible on return to 298 K. An overlay of the 0.301.80 ppm region of the H NMR spectra
400.13 MHz) is displayed in Figure 2. Coalescence of the two resonances attributed to the
two non-equivalent methyl groups in the N(CH CH moiety (two triplets centred on  =
.64 and 0.46 ppm in the slow regime at 263 K) was observed at T = 304 K.
8
. All changes observed in this range were
1
(
2
)
3 2
0
c
Fast regime
Intermediate regime
Slow regime
1
Fig. 2 Stack of the 0.301.80 ppm region of the H NMR spectra (400.13 MHz, toluene-d
8
) of
2
{
LO }Pb(N(SiMe
3
2
) ) (10) recorded in the temperature range 263363 K.
‡
–1
Using  = 71 Hz (determined at 263 K) leads to an estimate of G = 14.6 kcal·mol
for the free energy of activation associated to the exchange between the ethyl groups. The
8

Page 9 of 55
Dalton Transactions
‡
–1
Vie w‡ Article Online
corresponding enthalpy and entropy of activation H = 14.8(0.5) kcal·mol and S =
DOI: 10.1039/C3DT51681D
–
1
–1
+
0.7(1.6) cal·K ·mol were estimated by Eyring treatment of exchange rates determined by
line-shape analysis (Figure 3). The variation of entropy associated to this process, which is
assumed to proceed via dissociation-recoordination of the amino moiety, is surprisingly small.
–
1
Arrhenius analysis led to E
a
= 15.5(0.5) kcal·mol . The equation
‡
E
a
= H + mRT
(1)
where m is the order of the reaction corresponding to the fluxional changes, gives a kinetic
order of 1 for this process for T = 304 K. A single resonance at ca. +2100 ppm is seen in the
Pb{ H} NMR spectra of 10 in this temperature range, which confirms the existence of a
c
2
07
1
single environment.
Fig. 3 Eyring treatment of exchange rates determined by line-shape analysis (T = 263−363 K)
2
for the dynamic behaviour of NCH
2
CH
3
hydrogens in {LO }Pb(N(SiMe
3
2
) ) (10).
2
Eyring and Arrhenius analysis were performed for {LO }SnCl (in the range 298368
K; T
c
= 322 K), but treatment of the data for 2 (T = 233−363 K; T = 318 K) could not be
c
9

Dalton Transactions
Page 10 of 55
e
0
View Article Online
DOI: 10.1039/C3DT51681D
performed as relevant parameters (ν, ν
½
and ν , see the Supporting information) could
½
2
not be determined accurately. Comparative data for 2, 6, 10 and {LO }SnCl are collected in
Table 1; they all are commensurate. Although performed over a limited temperature range, the
2
data for {LO }SnCl compare with those reported for 6, confirming the reliability of the
procedure. Identical phenomena, equivalent to a dynamic racemization process at a pseudo-
chiral 3-coordinate metal centre, must be at work in both cases.
Table 1. Summary of Eyring and Arrhenius analyses for fluxional processes in the
2
a
N(CH
2
CH
3
) fragments of 2, 6, 10 and {LO }SnCl.
2
b
c
‡
‡
‡
Complex
T
c

E
a
G
H
S
−
1
−1
−1
−1
−1
(
°C) (Hz) (kcal·mol ) (kcal·mol ) (kcal·mol ) (cal·K ·mol )
2
b
b
b
b
{
{
{
{
LO }Ge(N(SiMe
3
)
2
) (2)
318 n/a
318 28
n/a
+16.3
+16.0
+14.6
+16.4
n/a
n/a
2
LO }Sn(N(SiMe
3
)
2
2
) (6)
+13.6(0.3)
+15.5(0.5)
+13.2(0.1)
+13.0(0.3)
+14.8(0.5)
+12.5(0.1)
–9.6(1.0)
+0.7(1.6)
−9.8(10.3)
2
LO }Pb(N(SiMe
3
)
) (10) 304 71
2
LO }SnCl
322 62
a
b
c
NMR data recorded in toluene-d
8
; data for 6 taken from reference 14a. Coalescence temperature. Difference of
b
frequencies for the separated methyl groups at the lowest available temperature. Could not be determined accurately.
Heteronuclear NMR data recorded at 298 K in aromatic solvents are collated in Table
2
9
1
2
. The Si{ H} chemical shifts for 1−11 all fall in the same narrow range; the slight shift
towards high fields on moving from germylenes to stannylenes and plumbylenes is consistent
1
19
1
with increasing ionicity of the corresponding M−Namido bond. The Sn{ H} chemical shift
1
19
1
−3
(
Sn: I = ½, natural abundance = 8.6%, receptivity relative to H = 4.4·10 ) for the new
stannylene 8 (−49.9 ppm) is nearly identical to those measured for 5−7, and is diagnostic of a
1
4a
207
1
3
-coordinate, monomeric tin(II) centre. Since Pb{ H} chemical shifts for lead(II)
2
07
compounds spread in the range −6000 to +3000 ppm ( Pb: I = ½, natural abundance =
1
−3
2
2.1%, receptivity relative to H = 2.1·10 ), the similar resonances detected for the
10

Page 11 of 55
Dalton Transactions
View Article Online
plumbylenes 9−11 (singlets between +2000 and +2150 ppm) testify to near-identical chemical
DOI: 10.1039/C3DT51681D
and magnetic environments for the three metal centres. Since the aminophenolate ligand in 10
is devoid of side-arm oxygen atom, we concluded that the tethered Oside-arm atoms in 9 and 11
do not interact with the metal in solution. This postulate is in agreement with structural
2
07
1
information obtained from XRD crystallography, and Pb{ H} chemical shifts in the range
+2000 to +2200 ppm are thus indicative of 3-coordinate, monomeric amino-phenolate lead(II)
3
amides. By comparison, the homoleptic {LO }
2
Pb, featuring a 4-coordinate metal in the solid
state (Supporting information), exhibits a shielded resonance at 207Pb = −367 ppm.
3
2
3
a
Table 2. NMR data for 1−11, {LO }
2
Pb, {LO }SnCl and {LO }SnCl.
2
9
1
119
1
207
1
Complex Solvent
Si{ H}
Sn{ H}
Pb{ H} Reference
(
ppm)
(ppm)
(ppm)
1
{
LO }Ge(N(SiMe
3
3
3
)
)
)
2
2
2
) (1)
) (2)
) (3)
toluene-d
toluene-d
toluene-d
toluene-d
8
8
8
8
0.41
0.06
2.37
–0.66
–0.63
−0.49
−0.34
-
-
-
-
-
-
-
-
-
-
-
-
-
this work
this work
this work
14a
2
{
{
{
{
{
{
{
{
{
{
{
{
LO }Ge(N(SiMe
3
LO }Ge(N(SiMe
1
LO }Sn(N(SiMe
3
)
2
2
2
2
) (5)
–63.8
2
LO }Sn(N(SiMe
3
3
3
)
)
)
) (6) benzene-d
) (7) benzene-d
6
6
–62.8
14a
3
LO }Sn(N(SiMe
−55.0
6g
4
LO }Sn(N(SiMe
) (8)
toluene-d
toluene-d
toluene-d
toluene-d
8
−49.9
this work
14a
2
LO }SnCl
8
8
8
8
–218.1
3
LO }SnCl
-
−385.0
this work
this work
this work
this work
this work
1
LO }Pb(N(SiMe
3
3
3
)
)
)
2
2
2
) (9)
–3.35
–2.35
−3.29
-
-
-
-
-
2007
2
LO }Pb(N(SiMe
) (10) toluene-d
2135
2027
−367
3
LO }Pb(N(SiMe
) (11) benzene-d
benzene-d
6
6
3
LO }
2
Pb
a
NMR data recorded at 25 °C.
11

Dalton Transactions
Page 12 of 55
3
View Article Online
DOI: 10.1039/C3DT51681D
The molecular structures of the germylenes 1−3 and {LO }GeCl, stannylenes 8 and
3
{
LO }SnCl, and plumbylenes 9−11 were determined by X-ray diffraction measurements
(
Figures 4−12). Independently of the identity of the metal, all these complexes are monomeric
in the solid state and feature 3-coordinate metal atoms. All amino(ether)-phenolate ligands
lead to the formation of a 6-membered metallacycle through sole coordination of the Ophenolate
1
−
and Nside-arm atoms as also observed for the amino-phenolate {LO } ; where they could
potentially occur, interactions between Oside-arm atoms and the metal were never detected. The
−
−
environment about the metal is otherwise completed by Cl or the bulky amide N(SiMe
3
) .
2
Pertinent metric parameters for these complexes as well as 5−7 are displayed in Table 3.
1
Fig. 4 ORTEP diagram of the molecular structure of {LO }Ge(N(SiMe
3
2
) ) (1). Ellipsoids are
drawn at the 50% probability level. Hydrogen atoms are omitted for clarity. Selected bond
lengths (Å) and angles (deg): Ge(1)−O(1) = 1.876(2), Ge(1)−N(1) = 1.901(2), Ge(1)−N(20) =
2
1
.319(3); O(1)−Ge(1)−N(1) = 96.35(9), O(1)−Ge(1)−N(20) = 88.33(9), N(1)−Ge(1)−N(20) =
00.4(1).
12

Page 13 of 55
Dalton Transactions
View Article Online
DOI: 10.1039/C3DT51681D
2
Fig. 5 ORTEP diagram of the molecular structure of {LO }Ge(N(SiMe
3
2
) ) (2). Ellipsoids are
drawn at the 50% probability level. Hydrogen atoms are omitted for clarity. Selected bond
lengths (Å) and angles (deg): Ge(1)−O(1) = 1.872(1), Ge(1)−N(2) = 1.907(1), Ge(1)−N(27) =
2
.294(1); O(1)−Ge(1)−N(2) = 94.52(6), O(1)−Ge(1)−N(27) = 90.22(6), N(2)−Ge(1)−N(27) =
00.60(6).
1
3
Fig. 6 ORTEP diagram of the molecular structure of {LO }Ge(N(SiMe
3
2
) ) (3). Ellipsoids are
drawn at the 50% probability level. Hydrogen atoms are omitted for clarity. Selected bond
lengths (Å) and angles (deg): Ge(1)−O(28) = 1.891(1), Ge(1)−N(21) = 1.913(1), Ge(1)−N(1)
=
2.318(1); O(28)−Ge(1)−N(21) = 95.72(5), O(28}−Ge(1)−N(1) = 90.07(4),
N(21)−Ge(1)−N(1) = 100.14(5).
13

Dalton Transactions
Page 14 of 55
View Article Online
DOI: 10.1039/C3DT51681D
3
Fig. 7 ORTEP diagram of the molecular structure of {LO }GeCl. Ellipsoids are drawn at the
5
0% probability level. Hydrogen atoms are omitted for clarity. Only the main component of
the disordered segment in the heterocycle side-arm (viz O9A) is represented. Selected bond
lengths (Å) and angles (deg): Ge(1)−(O3)1 = 1.860(2), Ge(1)−N(15) = 2.189(2), Ge(1)−Cl(1)
=
2.301(7); O(31)−Ge(1)−N(15) = 92.42(7), O(31)−Ge(1)−Cl(1) = 95.72(6),
N(15)−Ge(1)−Cl(1) = 97.95(6).
4
Fig. 8 ORTEP diagram of the molecular structure of {LO }Sn(N(SiMe
3
2
) ) (8). Ellipsoids are
drawn at the 50% probability level. Hydrogen atoms are omitted for clarity. Selected bond
lengths (Å) and angles (deg): Sn(1)−O(1) = 2.064(1), Sn(1)−N(29) = 2.115(1), Sn(1)−N(17) =
2
.419(1); O(1)−Sn(1)−N(29) = 93.78(5), O(1)−Sn(1)−N(17) = 86.55(4), N(29)−Sn(1)−N(17)
95.62(5).
=
14

Page 15 of 55
Dalton Transactions
View Article Online
DOI: 10.1039/C3DT51681D
3
Fig. 9 ORTEP diagram of the molecular structure of {LO }SnCl. Ellipsoids are drawn at the
5
0% probability level. Hydrogen atoms are omitted for clarity. Selected bond lengths (Å) and
angles (deg): Sn(1)−O(36) = 2.072(2), Sn(1)−N(1) = 2.357(2), Sn(1)−Cl(2) = 2.462(1);
O(36)−Sn(1)−N(1) = 86.17(6), O(36)−Sn(1)−Cl(2) = 92.96(5), N(1)−Sn(1)−Cl(2) = 97.50(5).
1
Fig. 10 ORTEP diagram of the molecular structure of {LO }Pb(N(SiMe
3
2
) ) (9). Ellipsoids are
drawn at the 50% probability level. Hydrogen atoms are omitted for clarity. Selected bond
lengths (Å) and angles (deg): Pb(1)−O(1) = 2.220(4), Pb(1)−N(20) = 2.537(5), Pb(1)−N(31) =
2
.243(5); O(1)−Pb(1)−N(31) = 93.7(2), O(1)−Pb(1)−N(20) = 83.9(2), N(31)−Pb(1)−N(20) =
5.8(2).
9
15

Dalton Transactions
Page 16 of 55
View Article Online
DOI: 10.1039/C3DT51681D
2
Fig. 11 ORTEP diagram of the molecular structure of {LO }Pb(N(SiMe
3
2
) ) (10). Ellipsoids
are drawn at the 50% probability level. Hydrogen atoms are omitted for clarity. Selected bond
lengths (Å) and angles (deg): Pb(1)−O(1) = 2.186(2), Pb(1)−N(1) = 2.218(3), Pb(1)−N(20) =
2
9
.536(3); O(1)−Pb(1)−N(1) = 92.61(9), O(1)−Pb(1)−N(20) = 83.97(9), N(1)−Pb(1)−N(20) =
7.9(1).
3
Fig. 12 ORTEP diagram of the molecular structure of {LO }Pb(N(SiMe
3
2
) ) (11). Ellipsoids
are drawn at the 50% probability level. Hydrogen atoms are omitted for clarity. Selected bond
lengths (Å) and angles (deg): Pb(1)−O(11) = 2.227(2), Pb(1)−N(1) = 2.237(2), Pb(1)−N(31) =
16

Page 17 of 55
Dalton Transactions
View Article Online
2
.543(2); O(11)−Pb(1)−N(1) = 91.51(8), O(11)−Pb(1)−N(31) = 84.68(8), N(1)−Pb(1)−N(31)
DOI: 10.1039/C3DT51681D
=
97.97(8).
Except for the M−heteroatom interatomic distances which increase regularly with the
II
II
size of the metal (effective ionic radii for a coordination number of 6: Ge , 0.73 Å; Sn ,
II
II
II
II
30
unspecified; Pb , 1.19 Å; empirical atomic radius: Ge , 1.25 Å; Sn , 1.45 Å; Pb , 1.80 Å),
the geometric features of all complexes are very similar (Table 3). All
heteroatom−metal−heteroatom angles are fairly close to 90 °. This suggests very limited or
absence of hybridization between s and p valence orbitals, and, for a given metal, the
2
31
character of the orbital for the lone pair of electrons is essentially ns . All bond lengths fall
in the expected range for such compounds. For a given ligand framework, there is no notable
modification of the structural features for Ge/Sn/Pb complexes beyond the normal extension
of the three M−heteroatom distances. For each family built on a same metal, bond distances
and angles vary little between complexes with the exception of the Namine−Sn−Cl angle for
3
2
{
LO }SnCl (entry 10, 97.5 °) and{LO }SnCl (entry 9, 89.5 °). This latter complex is rather
peculiar, as the Namine−Sn−Cl angle is also much smaller than the corresponding Namine−Sn−
Namine angle found in the congeneric amido complex 6 (entry 6, 97.7 °), whereas no such
3
discrepancy was found between the analogous pair of complexes {LO }SnCl and 7.
−
−
Comparison of entries 3 and 4 emphasizes that the nature of the X co-ligand, where X is
−
−
either Cl or N(SiMe
3
) , should bear little influence of the geometric patterns around the
2
metal.
17

Dalton Transactions
Page 18 of 55
2
3
a
Table 3. Relevant metric parameters for 1−3, 5−11, {LO }SnCl and {LO }SnCl.
Entry
Complex
M−Ophenolate
M−X
(Å)
M−Namine
O
phenolate−M−Namine
O
phenolate−M−X
N
amine−M−X Reference
(
Å)
(Å)
(deg)
(deg)
(deg)
1
1
2
3
4
{LO }Ge(N(SiMe
3
3
3
)
)
)
2
2
2
)
)
)
(1)
(2)
(3)
1.876(2)
1.872(1)
1.891(1)
1.860(2)
1.901(2)
1.907(1)
1.913(1)
2.301(7)
2.319(3)
2.294(1)
2.318(1)
2.189(2)
88.33(9)
90.22(6)
90.07(4)
92.42(7)
96.35(9)
94.52(6)
95.72(5)
95.72(6)
100.4(1)
100.60(6)
100.14(5)
97.95(6)
this work
2
{LO }Ge(N(SiMe
this work
this work
this work
3
{LO }Ge(N(SiMe
3
{LO }GeCl
1
5
6
7
8
9
{LO }Sn(N(SiMe
3
)
2
)
(5)
(6)
(7)
(8)
2.077(1)
2.066(3)
2.074(1)
2.064(1)
2.036(2)
2.072(2)
2.128(2)
2.102(4)
2.112(1)
2.115(1)
2.506(7)
2.468(1)
2.469(2)
2.435(3)
2.437(1)
2.419(1)
2.393(2)
2.357(2)
85.43(5)
86.6(1)
91.24(5)
94.0(1)
95.95(6)
97.7(1)
14a
14a
2
{LO }Sn(N(SiMe
3
3
3
)
)
)
2
2
2
)
)
)
3
{LO }Sn(N(SiMe
86.08(4)
86.55(4)
86.18(7)
86.17(6)
94.06(4)
93.78(5)
93.79(6)
92.96(5)
96.78(4)
95.62(5)
89.46(5)
97.50(5)
6g
4
{LO }Sn(N(SiMe
this work
14a
2
{LO }SnCl
3
1
0
{LO }SnCl
this work
1
1
1
1
1
2
3
{LO }Pb(N(SiMe
3
3
3
)
)
)
2
2
2
)
)
)
(9)
2.220(4)
2.186(2)
2.227(2)
2.243(5)
2.218(3)
2.237(2)
2.537(5)
2.536(3)
2.543(2)
83.9(2)
83.97(9)
84.68(8)
93.7(2)
92.61(9)
91.51(8)
95.8(2)
97.9(1)
97.97(8)
this work
this work
this work
2
{LO }Pb(N(SiMe
(10)
(11)
3
{LO }Pb(N(SiMe
a
II
II
II
3 2
M = Ge , Sn or Pb ; X = Cl or N(SiMe )

Page 19 of 55
Dalton Transactions
View Article Online
DOI: 10.1039/C3DT51681D
Heterobimetallic inclusion complexes
3
The structural features of family {LO }M(N(SiMe
3
2
) (M = Ge, 3; Sn, 7; Pb, 11) where the
amino(ether)-phenolate incorporates the aza-15-crown-5 tether are of particular interest.
Because none of the Oside-arm atoms is coordinated onto the metal, we postulated they could be
employed for further coordination chemistry involving an additional metallic centre. The high
affinity of (aza-)crown ethers for cationic metals has been demonstrated, and can be exploited
3
2
to design ion sensors acting through selective ligation of metal ions. Macrocycles containing
5
heteroatoms such as 15-crown-5 and 1-aza-15-crown-5 are ideally suited to the binding of
+
+
32b
the small Li and Na alkali ions, and we reasoned that salts of these metals could be
combined with 3, 7 and/or 11 to prepare heterobimetallic complexes through inclusion of the
3
−
hard cation in the anchored macrocycle of the {LO } ligand. A related approach was
implemented by Jurkschat and co-workers, who provided elegant spectroscopic evidence for
the formation of tin(IV)-halide bimetallic species upon addition of various alkali halides to
solutions of their bis(crown ether)-substituted organostannanes X
2
Sn(CH
Br, I). Also, Batten and co-workers have just reported manganese- or cuprous-potassium
heterobimetallic coordination polymers using a functionalised diaza-18-crown-6 ligand
2
-[16]-crown-
5
) (X
2
3
3
=
3
4
possessing pendant p-pyridylpyrazole side-arms.
3
35
In a preliminary reaction, the proteo-ligand {LO }H (a colourless oil) was reacted
1
13
1
with LiOTf in diethyl ether (Scheme 2). The H and C{ H} NMR data for the white solid
3
(
{LO }H·LiOTf) obtained quantitatively after evaporation was different from those for
3
{
LO }H, suggesting that the lithium cation was ligated by the macrocyclic heteroatoms. This
was confirmed by X-ray diffraction crystallography, which shows the alkali metal to sit in the
pocket formed by the four Oside-arm atoms and to be further coordinated by one oxygen atom
from the triflate counter-ion, whereas the Nside-arm atom is not involved in the coordination
sphere of the metal (Figure 13). The geometry about the metal constitutes a distorted square
19

Dalton Transactions
Page 20 of 55
3
6
View Article Online
DOI: 10.1039/C3DT51681D
pyramidal ( = 0.36), with the tightly bound Otriflate atom occupying the apical position and
the Oside-arm atoms being more remote from the metal.
Scheme 2. Synthesis of heterobimetallic compounds
3
As the molecular structures of 3 and {LO }H·LiOTf offered the structural features
required to the formation of an heterobimetallic complex, the reaction of the latter with
II
Ge(N(SiMe
3
)
2
)
2
was attempted, but it failed to yield the mixed Ge −Li species (Scheme 2).
Instead, the desired complex 3·LiOTf was obtained by equimolar reaction between the
20

Page 21 of 55
Dalton Transactions
View Article Online
DOI: 10.1039/C3DT51681D
germylene 3 and lithium triflate. It is a colourless solid soluble in ethers and aromatic
hydrocarbons, but insoluble in light petroleum. It was characterised by NMR spectroscopy
3
and X-ray diffraction crystallography, but the presence of residual {LO }H·LiOTf could not
be avoided, which precluded good combustion analysis.
3
Fig. 13 ORTEP diagram of the molecular structure of {LO }H·LiOTf. Ellipsoids drawn at the
5
0% probability level. Hydrogen atoms omitted for clarity. Selected bond lengths (Å) and
angles (deg): Li(1)−O(103) = 1.940(5), Li(1)−O(10) = 2.105(5), Li(1)−O(7) = 2.105(5),
Li(1)−O(13) = 2.138(5), Li(1)−O(4) = 2.159(5); O(103)−Li(1)−O(10) = 119.5(3),
O(103)−Li(1)−O(7) = 100.3(2), O(10)−Li(1)−O(7) = 77.86(18), O(103)−Li(1)−O(13) =
1
05.5(2), O(10)−Li(1)−O(13) = 77.46(19), O(7)−Li(1)−O(13) = 150.7(3), O(103)−Li(1)−O(4)
=
=
109.5(3), O(10)−Li(1)−O(4) = 128.9(2), O(7)−Li(1)−O(4) = 80.27(18), O(13)−Li(1)−O(4)
103.5(2).
The molecular solid state structure of 3·LiOTf is remarkable (Figure 14). It can be
divided into two fragments which virtually do not interact with each other: one pertaining to
the amino-phenolate Ge(II) amide, and the other relating to the a polyether-LiOTf moiety.
−
The triflate anion tightly bound to the Li atom and the bulky amido group N(SiMe
3
)
2
are
21

Dalton Transactions
Page 22 of 55
View Article Online
DOI: 10.1039/C3DT51681D
located in trans position with respect to the plane defined by the five heteroatoms of the
macrocycle, so that they impart minimal steric congestion to the coordination spheres of
either of the two metals. All bonding patterns and metric parameters for the 3-coordinate Ge
centre in 3·LiOTf match closely those described for 3 alone, whereas those measured around
II
36
3
the 5-coordinate Li centre ( = 0.33) are very similar to those found for {LO }H·LiOTf. The
II
large Ge ···Li distance (5.85 Å) rules out the existence of metallophilic interactions.
Fig. 14 ORTEP diagram of the molecular structure of 3·LiOTf. Ellipsoids are drawn at the
5
0% probability level. Hydrogen atoms are omitted for clarity. Selected bond lengths (Å) and
angles (deg): Ge(1)−O(11) = 1.871(1), Ge(1)−N(1) = 1.902(1), Ge(1)−N(31) = 2.335(1),
O(34)−Li(51) = 2.081(3), O(37)−Li(51) = 2.096(3), O(40)−Li(51) = 2.034(3), O(43)−Li(51) =
2
.081(3), Li(51)−O(52) = 1.885(3); O(11)−Ge(1)−N(1) = 96.23(6), O(11)−Ge(1)−N(31) =
9.49(5), N(1)−Ge(1)−N(31) = 99.04(6), O(52)−Li(51)−O(40) = 117.82(16),
8
O(52)−Li(51)−O(43) = 105.45(15), O(40)−Li(51)−O(43) = 78.32(12), O(52)−Li(51)−O(34) =
11.26(16), O(40)−Li(51)−O(34) = 130.15(16), O(43)−Li(51)−O(34) = 97.17(14),
O(52)−Li(51)−O(37) = 102.34(16), O(40)−Li(51)−O(37) = 78.46(12), O(43)−Li(51)−O(37) =
49.81(17), O(34)−Li(51)−O(37) = 83.41(12).
1
1
22

Page 23 of 55
Dalton Transactions
II
View Article Online
DOI: 10.1039/C3DT51681D
The heterobimetallic Sn −Li complex 7·LiOTf was also prepared by reaction of 7 and
3
lithium triflate, since the reaction of {LO }H·LiOTf and Sn(N(SiMe
3
)
2
2
) proved unsuccessful
II
(
Scheme 2). The solid state structure of the Sn −Li bimetallic complex 7·LiOTf is depicted in
Figure 15. All attempts to obtain the lead(II) analogue of 3·LiOTf and 7·LiOTf failed: the
3
kinetic lability of 11 and its contamination by {LO }
2
Pb (vide supra) preclude its use as an
3
efficient precursor, while no reaction took place between {LO }H·LiOTf and Pb(N(SiMe
3
)
2
2
) .
Fig. 15 ORTEP diagram of the molecular structure of 7·LiOTf. Ellipsoids are drawn at the
0% probability level. Hydrogen atoms are omitted for clarity. Selected bond lengths (Å) and
angles (deg): Sn(1)−O(31) = 2.0880(15), Sn(1)−N(32) = 2.1137(17), Sn(1)−N(15) =
5
2
.4860(17), Li−O(45) = 1.893(4), Li−O(6) = 2.025(4), Li−O(12) = 2.040(4), Li−O(9) =
.111(4), Li−O(3) = 2.142(4); O(31)−Sn(1)−N(32) = 96.86(6), O(31)−Sn(1)−N(15) =
4.41(6), N(32)−Sn(1)−N(15) = 99.47(6), O(45) −Li−O(6) = 105.55(19), O(45)−Li−O(12) =
17.6(2), O(6)−Li−O(12) = 136.29(19), O(45)−Li−O(9) = 107.82(19), O(6)−Li−O(9) =
9.19(15), O(12)−Li−O(9) = 81.14(15), O(45)−Li−O(3) = 100.42(18), O(6)−Li−O(3) =
1.44(15), O(12)−Li−O(3) = 96.91(17), O(9)−Li−O(3) = 149.11(19).
2
8
1
7
8
23

Dalton Transactions
Page 24 of 55
View Article Online
DOI: 10.1039/C3DT51681D
The molecular structure of 7·LiOTf resembles closely that of 3·LiOTf, with a 3-
coordinate tin(II) centre and a 5-coordinate lithium atom in a square pyramidal environment
3
6
II
(
 = 0.21). The geometries and interatomic distances around the Sn and Li atoms in
3
7
·LiOTf match those found in the parent compounds 7 and {LO }H·LiOTf, even if Li−Oside-
arm and Li−Otriflate bond lengths in this latter compound are a little longer than in the
II
heterobimetallic complex. There is no Sn ···Li interaction on account of the long
intermetallic distance (6.06 Å).
On the whole, in the solid state, the inclusion of the small alkali metal bears no impact
on the coordination sphere about the p-block metal in these amino(crown-ether)-phenolate
complexes, be it with germanium(II) or the larger tin(II). This is also likely so in solution, as
3
7
119
1
indicated by heteronuclear NMR spectroscopy. The Sn{ H} data recorded for 7 and
·LiOTf (Sn = −55.0 and −45.8 ppm, respectively) in dichloromethane-d (owing to limited
solubility of the latter in aromatic hydrocarbons) are nearly identical; the corresponding
7
2
2
9
1
7
1
Si{ H} chemical shifts are also very similar (Si = −0.75 and −0.31 ppm), and the Li{ H}
3
chemical shift for 7·LiOTf (Li = −0.56 ppm) matched that for the tin-free {LO }H·LiOTf
7
1
(
Li = −0.84 ppm) in this solvent. Moreover, in benzene-d
6
or toluene-d , the Li{ H} (Li =
8
2
9
1
−
0.74 ppm) and Si{ H} (Si = +0.70 ppm) resonances for 3·LiOTf are comparable to those
3
for {LO }H·LiOTf (Li = −0.96 ppm) and 3 (Si = 2.37 ppm), respectively.
Although the size of the macrocyclic side-arm is in principle suited to the binding of
3
2
sodium ions, attempts to such insertions using sodium triflate or the
+
−
38
[
Na(OEt
2
)
4
] ·[H
2
N{B(C
6
F
5
)
3
} ] loose ion pair, failed to deliver heterobimetallic complexes
2
with either 3 or 7. Efforts to prepare bimetallic species starting from complex 8 where the
macrocycle contains only four heteroatoms also met no success, although NMR data showed
4
3 −
that {LO }H·LiOTf could be synthesised. The ability of the highly chelating {LO } to yield
polymetallic alkali species had previously been highlighted through ready formation of
24

Page 25 of 55
Dalton Transactions
3
13i
3
13d
View Article Online
DOI: 10.1039/C3DT51681D
{
LO }Li·LiN(SiMe
2
H)
2
and [{LO }K·KN(SiMe
2
H) ], although the challenges overcome
2
for the preparation of 3·LiOTf and 7·LiOTf were greater than those associated to these
homobimetallic alkali complexes. We have in the past failed to prepare the Zn−Li equivalent
to 3·LiOTf and 7·LiOTf, perhaps because the ionic nature of the Zn−O and Li−O bonds led
3
−
to deleterious redistribution reactions. The ability of the tethered side-arm in {LO } to
+
3
perfectly host Li salts also precludes the use of {LO }Li species for salt metathesis reactions;
3
as intractable mixtures are always obtained from such reactions; this is why {LO }K was used
3
3
instead to obtain {LO }GeCl and {LO }SnCl (vide supra).
Ring-opening polymerisation studies
The performance of complexes 1−3 and 5−10 in the catalysis of the immortal ring-opening
polymerisation (iROP) of L-lactide (L-LA) or racemic-lactide (rac-LA) upon addition of
2
c,39
iPrOH was probed (Scheme 3).
The heterobimetallic 3·LiOTf and 7·LiOTf were also
assessed to gauge the influence of the additional alkali metal, but the heteroleptic chloro
derivatives were not interrogated because (i) their behaviour in the presence of (excess)
alcohol is often erratic, and (ii) chloride is a very poor initiating group. Complex 11, which
could not be obtained free of impurity, was also excluded from this screening. Reactions were
typically performed in toluene at 60−100 °C, using 500−1,000 equiv of lactide and 10−25
equiv of alcohol vs. metal, and [lactide] = 2.0 M.
0
Scheme 3
25

Dalton Transactions
Page 26 of 55
a
Table 4. iROP of lactide promoted by 1−3, 5−10, 3·LiOTf or 7·LiOTf in association with iPrOH.
re
b
c
d
d
e
f
Entry
Precat.
LA [LA]
0
/[Precat]
0
/[iPrOH]
0
T
(
time
(min)
Yield
(%)
M
n,theo
M
n,SEC
1
M
w
/M
n
M
n,NMR
P
r
1
1
°C)
(g·mol ) (g·mol )
(g·mol )
1
2
3
4
5
6
7
8
1
2
3
3
L-
L-
L-
L-
L-
L-
L-
L-
L-
L-
L-
L-
L-
L-
L-
L-
L-
500:1:10
500:1:10
500:1:10
1,000:1:10
500:1:10
1,000:1:10
1,000:1:10
1,000:1:10
1,000:1:10
1,000:1:10
1,000:1:10
1,000:1:10
1,000:1:10
1,000:1:10
5,000:1:25
500:1:10
100
100
100
100
100
100
60
60
60
60
60
60
60
60
60
100
100
100
60
360
360
360
360
360
360
180
180
180
180
180
3
12
60
45
180
180
360
180
12
74
82
35
15
57
34
88
88
87
95
22
92
93
92
96
87
87
66
92
92
5,400
6,000
2,600
2,200
4,100
5,000
12,700
12,700
12,500
13,800
3,200
13,300
13,400
13,300
27,700
6,300
6,300
4,800
13,300
13,300
c
7,100
8,100
3,600
2,900
5,600
1.21
1.13
1.09
1.07
1.06
1.06
1.07
1.06
1.11
1.06
1.08
1.09
1.21
1.30
1.10
1.44
1.43
1.15
1.13
1.23
4,800
5,200
2,900
2,700
4,800
4,000
11,700
9,100
13,600
9,700
2,000
12,400
12,500
13,800
26,000
6,500
5,600
4,500
8,200
13,800
3·LiOTf
3·LiOTf
5,600
5
6
7
8
12,900
13,300
11,900
14,800
4,500
14,000
13,200
15,000
26,200
11,500
9,900
g
9
1
1
1
1
1
1
1
1
1
1
2
0
1
2
3
4
5
6
7
8
9
0
7·LiOTf
9
9
9
9
9
10
2
7
h
500:1:10
500:1:10
1,000:1:10
1,000:1:10
rac-
rac-
rac-
7,000
10,300
9,200
0.68
0.61
0.48
d
g
9
60
a
b
Polymerisations in toluene, [lactide]
0
= 2.0 M unless otherwise stated. Isolated yield of PLA after precipitation. Mn,theo = [lactide]
0
/[iPrOH]
0
× yield × 144.13 + MiPrOH
.
4
0 e
f
1
Determined by SEC vs. polystyrene standards, and corrected by a factor of 0.58. Determined by end-group analysis. Determined by homodecoupled H NMR
g
h
0
spectroscopy. From reference 6g. [L-lactide] = 4.0 M.

Page 27 of 55
Dalton Transactions
View Article Online
DOI: 10.1039/C3DT51681D
Several trends emerge rapidly from examination of the data collected in Table 4. In
combination with iPrOH (10 equiv), which acts both as a co-catalyst and a chain transfer
agent, all tested complexes afford binary catalytic systems for ROP reactions presenting a
high level of control over the macromolecular features (qualitatively measured by the good
agreement between theoretical and experimentally determined (by SEC or NMR) molecular
weights, and by the narrow polydispersity index M
w
/M ). Yet, the activity changes drastically
n
II
II
with the size of the metal, as reaction rates increase with metal size according to Ge << Sn
II
<
< Pb . Where germylenes require 6 h to convert only partly 500 equiv of monomer at 100 °C
entries 1−3), nearly full conversion is achieved within 3 h at 60 °C with the stannylenes
entries 7−10) while the plumbylene 9 fully converts 1,000 equiv of monomer in as little as 3
(
(
min at 60 °C (entry 12). Increase of ROP catalytic activity with metal size has already been
6
j,6l,41
reported for alkaline-earth metals,
but this cannot be extended to all groups has for
2
a,2g-h
instance such relationship cannot be drawn for metals of groups 4 and 13.
In this series
of compounds, for a given metal, the identity of the ligand bears limited influence, if any, on
the final outcome of the polymerisation; compare for instance entries 1−3, 5−8 and 16−17, an
observation which has already been discussed elsewhere in details in the case of tin(II)
6
g
precatalysts. The polymerisation of rac-lactide proceeds with rates and control comparable
to those of L-lactide, but the resulting polymers are essentially atactic (entries 18−20). End-
group analysis (NMR and MALDI-ToF MS) confirmed the identity of the expected termini
(
−CH(CH
3
)OH and (CH
3
2
) CHOC(=O)−).
Although they afford excellent control over the reactions parameters, the germylenes
1−3 are crippled by excessively low reaction rates, which in practise rules out their use as
good ROP precatalysts, at least for the polymerisation of lactide. The presence of LiOTf had a
beneficial effect on the catalytic activity of the germylene 3, as 3·LiOTf proved substantially
more active under otherwise identical experimental conditions (entries 3 vs. 5 and 4 vs. 6).
27

Dalton Transactions
Page 28 of 55
View Article Online
DOI: 10.1039/C3DT51681D
Although the nature of the ROP mechanism mediated by 3·LiOTf/iPrOH has not been
elucidated, a possible intuitive explanation can be proposed: one may envisage that the strong
+
Lewis acid Li (although tamed by coordination of the crown ether) acts as an activator for
2
c
the incoming monomer in a way reminiscent of the so-called “activated monomer” or “dual-
4
2
catalyst” ROP mechanisms. The fact that, by contrast, 7·LiOTf afforded lower conversion
than 7 (entries 9 and 11) probably arises from the much greater sensitivity of the former
compared to 3·LiOTf, which may result in rapid catalyst decomposition under catalytic
conditions. The catalytic performances of 5−7 (that of 8 is strictly analogous) have already
6
g
been discussed elsewhere and will not be further detailed here.
The lead-based binary system 9/iPrOH proved most effective. Under controlled
conditions, it afforded very rapidly narrowly dispersed polymers of predictable lengths
II
(
entries 12, 15 and 20). Large quantities of monomer (5,000 equiv vs. Pb ) were fully
converted into medium molecular weight material within 45 min under mild conditions, and
the resulting material exhibits excellent control over the molecular masses. In terms of
combined productivity and activity, this stands on equal foot with performances achieved with
1
3b,13e
highly effective zinc-based systems for the polymerisation of lactide.
The rapid increase
in polydispersity observed after full conversion, which results from deleterious
transesterification reactions, further testifies to the high reactivity of the binary catalyst
9
/iPrOH (entries 12−14). The utilisation of the other plumbylene, 10, was not investigated in
details, but based on the limited role of the ligand under the chosen experimental conditions
vide supra), similar result may be anticipated.
(
Conclusion
Complete families of stable monomeric germylenes, stannylenes and plumbylenes supported
by multidentate amino-, amino(ether)- and amino(crown-ether)-phenolate ligands are now
28

Page 29 of 55
Dalton Transactions
View Article Online
DOI: 10.1039/C3DT51681D
available. The combination of crystallographic and heteronuclear NMR studies shows that
independently of the nature of the metal centre and that of the co-ligand, the metal centre
systematically exists in a 3-coordinate environment. Therefore, from a strict coordination
2
−
point-of-view, the simple amino-phenolate {LO } is as good a ligand as the more
1
−
encumbered and electron-donating amino(ether)- and amino(crown-ether)-phenolates {LO }
3
−
and {LO } , respectively. The NMR signature of these complexes containing NMR-active
1
19
1
207
1
metal centres is readily provided by Sn{ H} and Pb{ H} NMR spectroscopies.
The fact that the metal in these divalent group 14 metallenes is satisfied in a 3-
coordinate coordination environment enables the preparation of heterobimetallic complexes
3
−
by inclusion of lithium salts in the crown-ether side-arm of the ligand {LO } , at least with
germanium(II) and tin(II) for which the metal−Ophenolate bond is fairly covalent. So far only
LiOTf has been used for this purpose with success, but several other complexes could in
principle be obtained upon expanding the size of the crown-ether, and future efforts could aim
at chelating a variety of monocations of alkali, coinage or triel metals.
If the catalytic activity of the simple germylenes, and in particular that supported by
3
−
the amino-crown ether-phenolate {LO } , for the polymerisation of lactide was disappointing,
preliminary results suggest it may be possible to boost their performance by inclusion of
judicious cations in the macrocyclic tether. On the other hand, the plumbylenes have revealed
excellent ability for the ROP of L-lactide, both in terms on control and reaction rates. Of
course, the toxicity of lead is a major liability that under normal circumstances would
immediately rule it out as a potential candidate for catalyst development in this field.
However, maximising reaction rates and monomer loadings to the point where only ppm
levels of metal catalyst are required should alleviate partly this issue, and in this aim we are
now trying to develop other lead(II) precatalysts for immortal ROP catalysis.
29