Lactide polymerization catalyzed by Mg and Zn diketiminate complexes with flexible ligand frameworks

Article abstract of DOI:10.1039/c3dt53278j

Diketimine ligands bearing N-benzyl, N-9-anthrylmethyl and N-mesitylmethyl substituents (nacnac^BnH, nacnac^AnH, and nacnac ^MesH) were prepared from condensation of the amine with either acetyl acetone or its ethylene glycol monoketal. Chlorination with N-chlorosuccinimide in the 3-position yielded Clnacnac^BnH and Clnacnac^AnH. The ligands were reacted with Zn(TMSA)₂ (TMSA = N(SiMe ₃)₂) to yield nacnac^AnZn(TMSA) and Clnacnac^BnZn(TMSA). Protonation with isopropanol afforded nacnac ^AnZnOiPr and Clnacnac^BnZnOiPr. Reaction of the diketimines with Mg(TMSA)₂ afforded nacnac^AnMg(TMSA), nacnac ^MesMg(TMSA), Clnacnac^BnMg(TMSA) and Clnacnac ^AnMg(TMSA). Subsequent protonation with tert-butanol produced nacnac^MesMgOtBu and Clnacnac^BnMgOtBu, but only decomposition was observed with N-anthrylmethyl substituents. Most complexes were characterized by X-ray diffraction studies. TMSA complexes were monomeric and alkoxide complexes dimeric in the solid state. All alkoxide complexes, as well as nacnac^AnMg(TMSA)/BnOH and Clnacnac^AnMg(TMSA)/BnOH were moderately to highly active in rac-lactide polymerization (90% conversion in 30 s to 3 h). nacnac^AnZnOiPr produced highly heterotactic polymers (P_r = 0.90), Clnacnac^BnMgOtBu/BnOH produced slightly isotactic polymers at -30°C (P_r = 0.43), and all other catalysts produced atactic polymers with a slight heterotactic bias. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c3dt53278j

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Lactide polymerization catalyzed by Mg and Zn
diketiminate complexes with flexible ligand
frameworks†
Cite this: Dalton Trans., 2014, 43,
6339
Todd J. J. Whitehorne, Boris Vabre and Frank Schaper*
Diketimine ligands bearing N-benzyl, N-9-anthrylmethyl and N-mesitylmethyl substituents (nacnacBnH,
nacnacAnH, and nacnacMesH) were prepared from condensation of the amine with either acetyl acetone
or its ethylene glycol monoketal. Chlorination with N-chlorosuccinimide in the 3-position yielded Clnac-
nacBnH and ClnacnacAnH. The ligands were reacted with Zn(TMSA)2 (TMSA = N(SiMe3)2) to yield nac-
nacAnZn(TMSA) and ClnacnacBnZn(TMSA). Protonation with isopropanol afforded nacnacAnZnOiPr and
ClnacnacBnZnOiPr. Reaction of the diketimines with Mg(TMSA)2 afforded nacnacAnMg(TMSA), nacnac-
MesMg(TMSA), ClnacnacBnMg(TMSA) and ClnacnacAnMg(TMSA). Subsequent protonation with tert-butanol
produced nacnacMesMgOtBu and ClnacnacBnMgOtBu, but only decomposition was observed with
N-anthrylmethyl substituents. Most complexes were characterized by X-ray diffraction studies. TMSA
complexes were monomeric and alkoxide complexes dimeric in the solid state. All alkoxide complexes, as
well as nacnacAnMg(TMSA)/BnOH and ClnacnacAnMg(TMSA)/BnOH were moderately to highly active in
rac-lactide polymerization (90% conversion in 30 s to 3 h). nacnacAnZnOiPr produced highly heterotactic
Received 29th January 2013,
Accepted 2nd February 2014
polymers (Pr = 0.90), ClnacnacBnMgOtBu/BnOH produced slightly isotactic polymers at −30 °C (Pr
=
DOI: 10.1039/c3dt53278j
www.rsc.org/dalton
0.43), and all other catalysts produced atactic polymers with a slight heterotactic bias.
to α-olefin polymerization, for example, general structure–reac-
tivity/selectivity relationships for the polymerization of lactide
and other cyclic esters are mostly lacking. We investigated in
recent years the influence of general catalyst geometry on cata-
lyst performance using structurally similar N-alkyl diketimi-
nates (nacnacR) as spectator ligands. Octahedral zirconium
Introduction
Biodegradable polymers,1–3 i.e. polymers which degrade com-
pletely into CO2 and water in the environment, are gaining
importance or at least interest due to concerns about the
impact of macroscopic plastic debris and microplastics on
wildlife,4 in particular in the marine environment.5–7 Poly-
esters are attractive targets for biodegradable polymers, since
the ester bond can be cleaved by hydrolysis. Of these, polylactic
acid (PLA), which is currently used in medical applications
and produced on an industrial scale for packing applications,
offers the additional advantage that the monomer is obtained
from a renewable feedstock.3,8–10 PLA is obtained industrially
from Sn-catalyzed polymerization of lactide, the dimeric anhy-
dride of lactic acid. Catalyst development for the coordination–
insertion polymerization of lactide has become a highly fre-
quented research area due to the challenge of polymerizing
lactide with high isotacticity and high activity.8,11–17 Compared
bisdiketiminate
complexes,
( )-C6H10-(nacnacR)2Zr(OEt)2
(Scheme 1), showed by far the highest activity obtained with
group 4 metal catalysts, but did not provide selectivity and
were unsatisfactory catalysts due to transesterification side
reactions and catalyst decomposition.18 Square-planar copper
diketiminate complexes, {nacnacR2Cu(µ-OiPr)}2 (Scheme 1),
also showed extremely high activities, which were orders of
magnitude above other Cu(II) catalysts, but showed no stereo-
selectivity.19,20 In contrast to the octahedral Zr catalyst, the
Cu catalyst was highly stable, able to produce block-copolymers,
did not undergo either transesterification reactions or chain-
transfer even in the absence of monomer, and was suitable for
immortal polymerization. Zn and Mg catalysts, 1–3
(Scheme 1), having the same N-alkyl diketimine ligands in a
tetrahedral coordination geometry, displayed only moderate
activity in comparison to other complexes with the same
central metal.21,22 While the Zn catalysts provided heterotactic
polymers, Mg catalyst 3 showed a slight preference for isotactic
monomer enchainment at low temperatures. The source of the
Centre in Green Chemistry and Catalysis (CGCC), Département de chimie, Université
de Montréal, 2900 Boul. E.-Montpetit, Montréal, QCH3T 1J4, Canada.
E-mail: Frank.Schaper@umontreal.ca
†Electronic supplementary information (ESI) available: Tables S1–S4. Lactide
conversion/time plots (Fig. S1–S10). Single-crystal diffraction data (CIF). CCDC
972589–972593, 972595–972599. For ESI and crystallographic data in CIF or
other electronic format see DOI: 10.1039/c3dt53278j
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6339–6352 | 6339

View Article Online
Paper
Dalton Transactions
Scheme 1
preference for isotactic enchainment is not clear at the
moment, but a ligand-mediated chain-end control mechanism
was proposed based on the available data.21
Zinc22–39 and magnesium21,24–31,40–43 diketiminate complexes
have been widely used in lactide polymerization, starting with
the seminal work of Coates using nacnacdippZnOiPr,44 4,
(dipp = diisopropylphenyl).23,24 In all cases, monomer insertion
occurred either unselectively or with heterotactic preference,
following a chain-end control mechanism. The latter can be
very efficient, provided that the ligand is sufficiently bulky,
and Pr-values of >90% were obtained with several catalysts
(Pr = probability of alternating enantiomer insertion). Non-
symmetric, mono-substituted N-aryl diketiminate ligands have
been used in several cases to yield C2- or C1-symmetric Mg or
Zn diketiminate complexes.27–29,37–39 However, catalyst geo-
metry did not impact on stereoselectivity and the catalysts
continued to show only heterotactic preference, if any,
for monomer insertion in the polymerization of rac-
lactide.27–29,37,38 Complex 3 thus represents so far the only
example of a diketiminate catalyst with an isotactic preference
and only few magnesium complexes with other ligand
systems have been reported to produce isotactically enriched
Scheme 2
acetone ethylene glycol monoketal (Scheme 2) with two equiv.
of amine in the absence of solvent50,51 finally yielded 5a in
62% yield (see Fig. S1† for the crystal structure of 5a).
Reaction of 5a with Zn[N(SiMe3)2]2 cleanly yielded the
respective Zn amide complex, 5b (Scheme 2). Further protona-
tion with isopropanol afforded nacnacAnZnOiPr, 5c. As typically
observed for nacnacZnN(SiMe3)2 complexes, 5b crystallized as a
monomeric complex with a trigonal-planar geometry around
the zinc atom and a perpendicular orientation of the amide
ligand versus the ligand mean plane (Fig. 1). Bond distances
and angles in 5b are in the range typically observed for
nacnacZnN(SiMe3)2 (Table S1†).22,28,29,52,53 Alkoxide complex 5c
crystallized as an alkoxide bridged dimer (Fig. 2), as do
all other similar complexes with the exception of monomeric
nacnacdippZnOtBu.26 The zinc atom is found in a distorted
tetrahedral coordination geometry. Bond distances and angles
are comparable to other alkoxide bridged zinc diketiminates
(Table S1†),22–24,39,52,54–57 but on the lower end of the
range observed for Zn–N distances and the higher end for the
N–Zn–N angle, indicating rather low steric pressure from
the ligand despite the annulated aromatic ring. The
orientation of the anthryl moieties in 5c is in agreement
with π-stacking interactions: the anthryl moieties are nearly
coplanar (angle between mean planes: 7°), with a slight offset
PLA (Pr
=
0.36,45 0.35–0.45,46 and 0.30–0.4147). Herein
we report further investigations into Mg and Zn diketiminate
catalysts with primary alkyl substituents on the nitrogen
atoms to explore the potential benefits of a flexible ligand
framework.
Results and discussion
Zinc complexes. Initial variations of the ligand substitution
pattern concentrated on increasing the steric bulk of the
N-benzyl substituent by replacement with N-anthrylmethyl
(5a, Scheme 2). The condensation of anthrylmethylamine with
acetyl acetone, a route successfully employed for other N-alkyl
diketimines,48 yielded only the monocondensation product
(Scheme 2), regardless of reaction conditions (1–2 equiv. HCl,
TsOH or a mixture of both). Condensation in ethanolic
HCl at reflux49 likewise failed to provide 5a. Reaction of acetyl
6340 | Dalton Trans., 2014, 43, 6339–6352
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Fig. 1 X-ray structure of 5b. Most hydrogen atoms, the second inde-
pendent molecule of comparable geometry and co-crystallized solvent
were omitted for clarity. Thermal ellipsoids are drawn at 50% probability.
Bond distances/Å: Zn1–N1: 1.935(1) & Zn2–N5: 1.935(1), Zn1–N2: 1.986(1)
&
Zn2–N4: 1.992(1), Zn1–N3: 1.897(1) & Zn2–N6: 1.902(1). Bond
Fig. 2 X-ray structure of 5c. Most hydrogen atoms and co-crystallized
solvent omitted for clarity. Thermal ellipsoids are drawn at 50% prob-
ability. The molecule has crystallographic inversion symmetry. Bond dis-
tances/Å: Zn1–N1: 1.972(2), Zn1–N2: 1.959(2), Zn1–O1: 1.975(2), Zn1–
O1A: 1.976(2), Zn1–Zn1A: 2.982(1). Bond angles/°: N1–Zn1–N2: 100.65
(7), O1–Zn1–O1A: 82.01(6), N1–Zn1–O1: 117.76(7), N2–Zn1–O1: 121.15
(7), N1–Zn1–O1A: 126.23(7), N2–Zn1–O1A: 110.25(7), Zn1–O1–Zn1A:
97.99(6).
angles/°: N1–Zn1–N2: 100.11(6) & N4–Zn2–N5: 99.60(6), N1–Zn1–N3:
142.41(6) & N5–Zn2–N6: 144.00(6), N2–Zn1–N3: 117.30(6) & N4–Zn2–
N6: 116.33(6).
placing a carbon atom in the middle of the aromatic rings and
distances of 3.3–3.7 Å.
Complex 5c proved to be a moderately active and controlled
catalyst for the solution polymerization of rac-lactide (Table 1,
Fig. S2†). After a short induction period, the reaction was first-
order in lactide concentration with an apparent rate constant
of kapp = 0.012–0.016 min−1 at 2 mM catalyst concentration.
The activity is very similar to the rate constants observed for
the respective N-benzyl complex 1 (kapp = 0.02–0.04 min−1) and
is oriented nearly symmetrically with regard to the ligand
mean plane (Fig. 1), the other is turned with the anthryl
moiety towards the backbone of the ligand. The latter orien-
tation is also observed for one of the N-substituents in dimeric
5c (Fig. 2). In both orientations, complex geometry becomes
essentially CS-symmetric and no catalytic site control by a puta-
tive C2-symmetric rotamer is observed. Polymer molecular
weight control with 5c was rather poor and polymer molecular
weights significantly higher than expected were obtained.
Neither a notable induction period nor catalyst decomposition
was observed, in agreement with overall narrow polydispersi-
ties. Prolonged reaction times after completion of polymeriz-
ation increased the Pr value as well as the discrepancy between
N-methylbenzyl complex 2 (kapp = 0.01–0.02 min−1 22
and only
)
slightly lower than nacnacdippZnOiPr, 4 (kapp = 0.05 min−1).24
The obtained polymer is highly heterotactic, with Pr values
around 90% (Table 1, Pr). Complex 5c thus shows an
even higher heterotactic preference than the corresponding
benzyl complex 2. A closer look at the crystal structures of
5b and 5c offers a tentative explanation for the lack of
catalytic-site control. In 5b, one anthrylmethyl substituent
Table 1 rac-Lactide polymerization with Zn complexes
Mn (g mol−1 b
)
Chains per Znb
Mw/Mn
a
#
Catalyst
[Zn] : lactide (: BnOH)
Conversion/time
Final conversion
Pr
1
2
3
4
5c
5c
5c
6c
1 : 300
1 : 300
1 : 300 : 1
1 : 300
40%/30 min
33%/30 min
19%/30 min
80%/3 h
93%/3 h
95%/6 h
87%/3 h
80%
0.88
0.93
0.87
0.59
66 500
118 000
14 900
n.d.
0.6
0.3
2.3
1.14
1.16
1.04
Conditions: CH2Cl2, ambient temperature (23 °C), 2 mM [Zn]. Conversion determined from 1H NMR. a Pr determined from decoupled 1H NMR
by Pr = 2·I1/(I1 + I2), with I1 = 5.20–5.25 ppm (rmr, mmr/rmm), I2 = 5.13–5.20 ppm (mmr/rmm, mmm, mrm). b Mw and Mn determined by size
exclusion chromatography vs. polystyrene standards, with an MH correction factor of 0.58. Number of chains per catalyst determined from
(conversion·mlactide/nZn + MiPrOH)/Mn.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6339–6352 | 6341

View Article Online
Paper
Dalton Transactions
Scheme 3
expected and obtained polymer molecular weight (cf. #1 vs.
#2). This might indicate transesterification side-reactions as
possible sources for the lack of polymer molecular weight
control (see ESI†).22 In the presence of one equiv. of benzyl
alcohol, the expected molecular weight for two chains per
metal center is obtained with narrow polydispersities (Table 1,
#3), indicating some stability of 5c under immortal polymeriz-
ation conditions. Zinc (and magnesium) diketiminate com-
plexes with N-alkyl substituents are, however, generally labile
towards ligand protonation by excess alcohol.
To investigate electronic effects on catalyst performance,
we prepared diketimine 6a with a chlorine substituent in the
3-position of the ligand backbone by reaction of the parent
diketimine with N-chlorosuccinimide (Scheme 3). Reaction
with Zn[N(SiMe3)2]2 yielded the amide complex 6b as an
oil, which was reacted, without further purification, with iso-
propanol to yield alkoxide complex 6c. Polymerization of rac-
lactide with 6c (Table 1, #4) showed an activity comparable
to that of its non-chlorinated analogue 2 and a Pr value of 0.59,
only slightly reduced compared to 2 (0.65–0.71).22
Fig. 3 X-ray structure of 5d. Hydrogen atoms and co-crystallized
solvent omitted for clarity. Thermal ellipsoids are drawn at 50% prob-
ability. Mg1–N2: 2.026(2) Å, Mg1–N3: 2.029(2) Å, Mg1–N1: 1.972(2) Å,
N2–Mg1–N3: 96.80(8)°, N1–Mg1–N2: 126.55(8)°, N1–Mg1–N3:
125.34(9)°.
Magnesium complexes
Since magnesium complex
3 was the first diketiminate
complex with an isotactic preference for monomer insertion, it
was of interest to see how slight variations of the substitution
pattern influenced stereoselectivity in lactide polymerization.
Reaction of Mg[N(SiMe3)2]2 with 5a yielded the respective
amide complex 5d (Scheme 4). Similarly to other diketiminate
metal amides, 5d is monomeric in the solid state (Fig. 3). In
comparison to the corresponding Zn amide 5b (Fig. 1), the
more ionic bonds in the Mg complex allow a stronger deviation
from a trigonal-planar geometry and the Mg atom is bent by
29° out of the plane of the diketiminate ligand. This deviation
is higher than in the corresponding N-aryl complex nac-
nacdippMgN(SiMe3)2,58 where this deviation is only 16°. The
difference can be ascribed to the higher flexibility of N-alkyl
diketiminates, which allow an easier out-of-plane bending of
the metal atom.59 Other geometrical parameters in 5d are com-
parable to those of nacnacdippMgN(SiMe3)2 (Table S2†).
Attempts to introduce an alkoxide substituent by reaction
of 5d with isopropanol led to decomposition. In the case of 3,
similar problems had been circumvented by employing tert-
butanol,21 but the reaction proved less controlled with 5d:
When an C6D6 solution of 5d was titrated with tert-butanol,
resonances assigned to the target complex 5e were obtained
upon addition of 0.25 equiv. of tert-butanol (Scheme 4,
Fig. S3†). Further addition of alcohol, however, led to the pres-
ence of increasing amounts of the ligand 5a and unknown by-
products. (The expected by-product, homoleptic (nacnacAn)2-
Mg, seems not to be present in the mixture. Likewise, we
were unable to prepare the homoleptic complex by reaction of
magnesium amide with two equiv. of 5a.) Although tert-buto-
xide complex 5e was the major species present, we were unable
to obtain a pure product despite extensive recrystallisation
experiments. Reaction of 5d with several other alcohols, such
as ethanol, benzyl alcohol, methyl lactate, or phenol, likewise
failed to yield an isolatable pure product. In the case of
Scheme 4
6342 | Dalton Trans., 2014, 43, 6339–6352
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
phenol, single crystals of the phenolate complex 5f suitable for complexes (Table S3†).21,24,29,60 The structure of 5f illustrates
X-ray diffraction were obtained (Fig. 4), but only in quantities the high flexibility of N-alkyl diketiminate ligands: all the
insufficient for further characterization. Bond distances anthryl orientations described for 5b and 5c are realized in 5f.
and angles in 5f are comparable to other [nacnacMg(µ-OR)]2 Two anthryl groups are arranged in a π-stacking orientation
and display the forward orientation of the anthryl moiety tar-
geted for effective stereocontrol, the other two do not show
suitable orientations for enforcing C2-symmetry: one anthryl
group is nearly perpendicular to the mean ligand plane, and
the other one is rotated towards the ligand backbone.
In the absence of an isolated alkoxide complex, the amide
complex 5d was employed for rac-lactide polymerizations. Sur-
prisingly, 5d proved to be highly active (Table 2, #1; Fig. S4†),
slightly faster than complex 3 with N-benzyl substituents.
None of the polymerizations performed reached completion,
indicating instability of the catalyst under polymerization con-
ditions. As with 3, polymerization at room temperature yielded
an atactic polymer. A notable induction period (Fig. S4†), a low
number of polymer chains per catalyst center and high poly-
dispersities above 1.8 indicate slow initiation of polymerization.
Addition of one equiv. benzyl alcohol to polymerizations with
5d removed the induction period and polymerization was com-
plete in less than one minute (Table 2, #2–#8; Fig. S5 and S6;†
cf.: nacnacdippMgOR, 97% in 2 min24), rendering 5d a highly
active catalyst with kobs = 2.8(3) min−1 at 2 mM catalyst concen-
tration. If two equiv. BnOH were used (Table 2, #5; Fig. S5†), or
at an increased lactide : Mg ratio of 900 : 1 (Table 2, #8;
Fig. S6†), catalyst decomposition prevented complete polymer-
ization. In all cases, the obtained polymer was atactic and, con-
trary to 3, polymerization at reduced temperatures did not
Fig. 4 X-ray structure of 5f. Most hydrogen atoms and co-crystallized
solvent omitted for clarity. Thermal ellipsoids are drawn at 50% prob-
ability. Bond distances/Å: Mg1–N1: 2.0559(15), Mg1–N2: 2.0406(15),
Mg2–N3: 2.0281(16), Mg2–N4: 2.0172(16), Mg1–O1: 1.9862(13), Mg1–
O2: 1.9940(13), Mg2–O1: 1.9700(13), Mg2–O2: 1.9845(13), Mg1–Mg2:
3.0122(8). Bond angles/°: N1–Mg1–N2: 94.88(6), N3–Mg2–N4: 95.67(6),
O1–Mg1–O2: 80.65(5), O1–Mg2–O2: 81.29(5).
affect stereoselectivity (Table 2, #9). Polydispersities varied
between 1.2 and 1.7, narrower than for polymerizations
without additional alcohol, but nevertheless indicative of poor
polymer weight control. In addition, obtained polymer weights
were now much lower than expected with 1.5 to 3.9 polymer
chains produced per metal center. Participation of (Me3Si)2NH
Table 2 rac-Lactide polymerization with Mg catalysts
Mn (g mol−1 b
)
Chains per Mgb
Mw/Mn
a
#
Catalyst
[Mg] : lactide (: BnOH)
Conversion/time
Final conversion
Pr
1
2
3
4
5
6
7
8
5d
5d
5d
5d
5d
5d
5d
5d
1 : 100
20–30%/1 minc
80%/30 s
60%–65%c
95%
>95%
95%
70%
95%
>95%
65%
85%/40 min
80%
92%/17 h
75–90%c
30–95%c
90–99%/50 minc
40%
0.51–0.52
0.52
0.49
0.49
0.48
0.51
0.49
0.50
0.52
15 100–15 700
24 800
15 500–16 300
13 300
7900
9500
27 200
36 800
n.d.
45 400
0.6
1.6
2.5–2.7
3.1
3.9
1.4
3.1
2.2
1.79–1.87
1.65
1.39–1.60
1.49
1.15
1.29
1 : 300 : 0.7
1 : 300 : 1
1 : 300 : 1.3
1 : 300 : 2
1 : 100 : 1
1 : 600 : 1
1 : 900 : 1
1 : 100 : 1
1 : 300
1 : 300
1 : 300
1 : 300 : 1
1 : 300 : 1
1 : 300 : 1
75–95%/30 sc
95%/30 s
50%/30 s
90%/30 s
95%/30 s
60%/30 s
1.60
1.45
9
5d, −30 °C
7e
7e, −30 °C
6e
10
11
12
13
14
15
70%/5 min
0.53
0.55
0.8
0.7
0.8–1.6
0.7–2.1
1.1–1.2
1.62
1.98
1.34–1.71
1.04–1.20
1.08–1.11
55 500
70–80%/10 minc
15–95%/10 minc
0.53–0.57d
0.49–0.53d
0.43d
0.48
23 300–39 000
16 600–27 800
31 900–37 900
n.d.
6e
6e, −30 °C
8d
30%/1 min
Conditions: CH2Cl2, ambient temperature, 2 mM catalyst concentration. Final conversion without a given time is the maximum conversion,
when the conversion/time plot plateaued. See ESI for approximate rate constants. a Pr determined from decoupled 1H NMR by Pr = 2·I1/(I1 + I2),
with I1 = 5.20–5.25 ppm (rmr, mmr/rmm), I2 = 5.13–5.20 ppm (mmr/rmm, mmm, mrm). b Mn and Mw determined by size exclusion chromatography
vs. polystyrene standards, with an MH correction factor of 0.58. Number of chains per Mg center determined from (conversion·mlactide/nMg
+
MiPrOH)/Mn. c Minimum and maximum values of multiple experiments. d From multiple aliquots/polymerizations (see ESI).
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6339–6352 | 6343

View Article Online
Paper
Dalton Transactions
Scheme 5
as a chain-transfer agent seems unlikely, given its low acidity
and slow insertion rates. Consequently, no notable presence of
1
polymer bound N(SiMe3)2 was detected in H NMR spectra of
polymers obtained with 5d/BnOH. Variations in the catalyst :
monomer ratio (1 : 100–1 : 900; #3, #6–#8) did not correlate
with the number of polymer chains obtained per catalyst, thus
excluding impurities in the monomer acting as chain-transfer
agents. Increasing amounts of benzyl alcohol (0.7–2 equiv.;
#2–#5) led to narrower polydispersities, in agreement with its
activating effect on initiation, but they still remained above
1.15. Higher amounts of alcohol increased, as expected, the
number of chains per catalyst center, but the obtained
polymer molecular weight remained lower than expected in all
cases. GPC traces did not indicate the presence of a low-mole-
cular-weight fraction that would be indicative of intramolecu-
lar transesterifications to yield cyclic oligomers. However, the
Pr value correlates – somewhat noisily – with the discrepancy
of expected and obtained molecular weight, i.e. the number of
chains per catalyst center produced (Fig. S7†). Since a likely
cause for variations in Pr are transesterification reactions
(the appearance of rr-triads causes an artificial reduction of
Pr, if calculated as described in Table 2), the observed dis-
crepancy in expected and obtained molecular weight might be
attributed to transesterification (see ESI†).
Given the high amount of π-stacking observed in the X-ray
structures of 5c and 5f, which might stabilize the undesired
meso-rotamer, we investigated ligand 7a, which is sterically
similar to 5a, but should show a reduced tendency for π–π-in-
teractions. Ligand 7a was obtained similarly to 5a, but the
presence of molecular sieves was essential to obtain yields
above 20% (Scheme 5). Reaction with Mg[N(SiMe3)2]2 yielded
the amide complex 7d. The crystal structure of amide complex
7d is nearly identical to that of 5d (Fig. 5, Table S2†) with the
only difference being that the near Cs-symmetry in 5d was
observed as crystallographic Cs-symmetry in 7d. Complex 7d
reacted with tert-butanol to yield 7e. Alternatively, direct
addition of tert-butanol to the crude reaction mixture contain-
ing 7d in a one-pot, two-step reaction yielded 7e in comparable
yields. The crystal structure of 7e shows a higher than usual
bending of the Mg atom out of the mean plane of the diketimi-
nate ligand and a likewise higher distortion from tetrahedral
symmetry, which are both indicative of the steric strain intro-
duced by the mesitylmethyl substituent, but otherwise geo-
metric parameters are comparable to 5f, 6e and other dimeric
{nacnacMg(µ-OR)}2 complexes (Fig. 6, Table S3†).21,24,29,60
Satisfyingly, the π-stacking interactions observed in 5f are
absent in 7e.
Fig. 5 X-ray structure of 7d. Hydrogen atoms omitted for clarity.
Thermal ellipsoids are drawn at 50% probability. Mg1–N1: 2.036(1) Å,
Mg1–N2: 1.975(2) Å, N1–Mg1–N1A: 95.28(6)°, N1–Mg1–N2: 129.39(3)°.
The performance of 7e in lactide polymerization was disap-
pointing (Table 2, #10, #11). Activities are lower than those of
5d/BnOH, but are still significantly higher than those observed
for 3.21 ortho-Substitution of the N-benzyl substituent thus
increased activity in 7e as well as 5d. However, catalyst
decomposition led to incomplete monomer consumption at
room temperature and polydispersities were above 1.6 at room
temperature and at −30 °C, despite a very minor induction
period (Fig. S8†). In addition, the catalyst showed no prefer-
ence for isotactic monomer insertion even at low
temperatures.
Magnesium complexes with β-chloro-diketiminate ligands
The effect of ligand chlorination was studied using β-chloro-
diketimines 6a and 8a (Scheme 6). The reaction of Mg[N-
(SiMe3)2]2 with 6a yielded mixtures of Mg[N(SiMe3)2]2, the
heteroleptic amide complex 6d and the homoleptic bisdiketimi-
nate complex 6f. To corroborate the assignment, 6f was
prepared independently and characterized by X-ray diffraction
(Fig. 7). Comparison of 6f with the corresponding homoleptic
complex lacking the chlorine substituent, nacnacBn2Mg,21
allows the delineation of the structural impact of introducing a
substituent on the Cβ atom of the diketiminate ligand
(Table 3). Both structures show a distorted tetrahedral geome-
try, with similar orientations of the benzyl substituents. Elec-
tronic influences are rather small, leading to
a minor
contraction of the Mg–N bond lengths in 6f (0.1 Å, Table 3).
6344 | Dalton Trans., 2014, 43, 6339–6352
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Fig. 7 X-ray structure of 6f. Hydrogen atoms omitted for clarity.
Thermal ellipsoids are drawn at 50% probability. Mg1–N1: 2.023(1) Å,
Mg1–N2: 2.024(1) Å, Mg1–N3: 2.038(1) Å, Mg1–N4: 2.045(1) Å,
N1–Mg1–N2: 90.39(5)°, N3–Mg1–N4: 89.93(5)°.
Fig. 6 X-ray structure of 7e. Hydrogen atoms and co-crystallized
solvent omitted for clarity. Thermal ellipsoids are drawn at 50% prob-
ability. Bond distances/Å: Mg1–N1: 2.0720(18), Mg1–N2: 2.0476(17),
Mg2–N11: 2.0664(17), Mg2–N12: 2.0460(17), Mg1–O1: 1.9747(14), Mg1–
O2: 1.9611(14), Mg2–O1: 1.9585(14), Mg2–O2: 1.9842(14), Mg1–Mg2:
2.9710(9). Bond angles/°: N1–Mg1–N2: 90.57(7), N11–Mg2–N12: 90.96(7),
O1–Mg1–O2: 82.18(6), O1–Mg2–O2: 82.00(6).
favoured its formation sufficiently for NMR characterization.
The alkoxide complex 6e was obtained by addition of tert-
butanol to a 1 : 1 mixture of 6a and Mg[N(SiMe3)2]2. The crystal
structure of 6e shows the expected dimeric structure with the
Mg atom in a tetrahedral environment (Fig. 8). In contrast to
its non-chlorinated analog 3,21 the meso-rotamer is observed in
the solid state, with both N-benzyl substituents on the same
side of the ligand plane. The syn-orientation of the benzyl sub-
stituents results in a smaller distortion of the tetrahedral
coordination geometry around Mg than in 3, where the benzyl
groups are found in the anti-orientation of the C2-symmetric
rotamer (angle of the MgN2 and MgO2 mean planes: 6e: 87°,
3: 80°). In other aspects, 6e is very similar to 3, with the same
minor steric impact of the chlorine substituent as discussed
for 6f (Table 3).
Reactions of 8a with magnesium amide mirror the reactivity
observed for the non-chlorinated analog 5a in that preparation
of the amide complex 8d was straightforward, but we were
unable to obtain an isolable alkoxide complex (Scheme 6).
Single crystals were obtained for 8d, but the obtained structure
suffered from weak diffraction intensities and the presence of
two independent molecules, one of which was severely dis-
Scheme 6
More notable is the steric impact of the chlorine substituent. ordered. We thus refrain from a discussion of this structure,
the overall geometry of which is very similar to that of 5d.61
By way of the methyl groups at the ligand backbone (widening
A single crystal of the homoleptic bis(diketiminate) complex,
of the Cβ–Cα–CMe angles by 3–4°, Table 3), substitution at Cβ
pushes the N-substituents closer to the front of the complex (Cl-nacnacAn)2Mg, 8f, was obtained as a minor byproduct with
clearly different morphology in one recrystallisation of 8d. The
structure of 8f (Fig. 9) shows π-stacking interactions between
(reduction of CN–Cβ–CN by 4° and of N–Cβ–N by 2°). Other
geometric data are consistent with a minor increase of steric
pressure in 6f (Table 3), in particular the slight reduction of the anthryl moiety and the chloro-diketiminate backbone
for two anthrylmethyl substituents, while the remaining two
the distances between Mg and the N-substituents.
are in a coplanar arrangement with each other. Mg–N bond
Although heteroleptic 6d could not be isolated from the
obtained product mixtures, the use of excess Mg[N(SiMe3)2]2 distances in 8f are slightly longer by 0.04° and N–Mg–N bite
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6339–6352 | 6345

View Article Online
Paper
Dalton Transactions
Table 3 Geometric impact of a chlorine substituent at the diketiminate Cβ atoma
6f
8f
(nacnacBn)2Mg (ref. 21)
6e
3 (ref. 21)
Mg–N
Mg–CN
Cβ–Cα–CMe
N–Mg–N
CN–Cβ-CN
N–Cβ–N
2.03(1) Å
2.98(6) Å
118°–119°
90°
95°–96°
74°
2.07(1) Å
3.05(5) Å
117°–119°
88°
97°–98°
73°–74°
2.04(1) Å
3.02(2) Å
114°–116°
94°
100°
76°
2.037(2) Å
2.95 Å
118°
92°
97°
2.041(3) Å
3.01–3.06 Å
115°
94°
99°
75°
76°
a Errors are standard deviations of the average of equivalent bond lengths or angles, not experimental uncertainties. CN: benzylic carbon atom;
Cβ: central carbon atom of the ligand backbone; CMe: diketiminate methyl group.
Fig. 8 X-ray structure of 6e. Hydrogen atoms omitted for clarity.
Thermal ellipsoids are drawn at 50% probability. Bond distances/Å:
Mg1–N1: 2.035(2), Mg1–N2: 2.038(2), Mg1–O1: 1.952(2), Mg1–O1A:
Fig. 9 X-ray structure of 8f. Hydrogen atoms omitted for clarity.
1.964(2), Mg1–Mg1A: 2.915(1). Bond angles/°: N1–Mg1–N2: 91.92(8),
Thermal ellipsoids are drawn at 50% probability. Bond distances/Å: Mg–
O1–Mg1–O1A: 83.79(7), N1–Mg1–O1: 125.38(8), N1–Mg1–O1A:
N: 2.065(2)–2.076(2). Bond angles/°: N1–Mg1–N2: 88.04(7), N3–Mg1–
116.45(8), N2–Mg1–O1: 122.26(8), N2–Mg1–O1A: 120.25(8).
N4: 87.20(7), NLig1–Mg1–NLig2: 108.30(8)–128.65(7).
angles are slightly smaller by 2° than in the respective period was observed in conversion vs. time plots which,
N-benzyl complex 6f (Table 3), indicating some small increase accompanied by broadened polydispersities, indicated
in steric bulk. It should be noted, however, that replacing slow initiation through insertion into the Mg–OtBu bond
phenyl by anthryl seems to have only a minor steric impact, (Fig. S10†). Addition of benzyl alcohol removed the induction
comparable to that of chlorination of the β-position.
period and narrowed polydispersities to 1.0–1.2 (Fig. S10†).
In the absence of an isolable alkoxide complex, lactide However, observed activities, number of polymer chains
polymerization was again performed with the amide complex obtained per Mg and final conversions varied widely, the latter
8d in the presence of benzyl alcohol. An activity lower than being as low as 30%. This indicates that tert-butanol, liberated
that of 5d/BnOH, significant catalyst decomposition and an by reaction with BnOH, remained active as a chain transfer
uninteresting Pr-value of 0.48 were observed (Table 2, #15; reagent and that catalyst decomposition occurs in the presence
Fig. S9†). Polymerizations with 8d/BnOH were thus not investi- of free alcohol. Side reactions can be suppressed by lowering
gated further. Complex 6e, on the other hand, was very active the polymerization temperature to −30 °C, yielding reproduci-
in lactide polymerization, reaching conversions above 70–80% ble results, polydispersities of 1.1 and only one polymer chain
after 10 min (Table 2, #12–#14). As for 3, a notable induction per metal center.
6346 | Dalton Trans., 2014, 43, 6339–6352
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
While 6e, as 3, has a slight heterotactic bias at room temp-
erature (Pr = 0.54), addition of benzyl alcohol inexplicably
reduced the Pr value to 0.46–0.49. Chain-transfer reactions due
Experimental part
General
to the free alcohol present should not affect stereoselectivity in All reactions were carried out using Schlenk and glove
the case of a highly flexible ligand. However, catalyst decompo- box techniques under a nitrogen atmosphere. NacnacBnH was
sition (indicated by the lack of complete conversion) might prepared according to literature,48,63 as was (9-anthryl)methyl-
yield species which are polymerization-active, but lack amine,64 acetyl acetone ethylene glycol monoketal,51 and
the slight heterotacticity of 6e. Alternatively, decomposition Zn(N(SiMe3)2)2.65 Solvents were dried by passage through acti-
products might be capable of catalyzing transesterification to vated aluminum oxide (MBraun SPS), de-oxygenated by
a higher degree than 6e, which would lower the apparent repeated extraction with nitrogen, and stored over molecular
Pr, since the statistical method used for its calculation sieves. C6D6 was dried over sodium and degassed by three
does not allow the presence of rr-triads.22 Lowering freeze–pump–thaw cycles. CDCl3 and CD2Cl2 were dried over
the polymerization temperature to −30 °C reduced Pr to 3 Å molecular sieves. rac-Lactide (98%) was purchased from
0.43, which is a slightly higher isotactic preference than Sigma-Aldrich, purified by 3× recrystallisation from dry ethyl
observed for 3 under the same conditions (Pr = 0.46). Since acetate and kept at −30 °C. All other chemicals were purchased
chain transfer and catalyst decomposition are absent at this from common commercial suppliers and used without further
temperature, this reflects a small, but reproducible isotactic purification. 1H and 13C NMR spectra were acquired on a
preference of 6e, the highest observed so far for a diketiminate Bruker AVX 400 spectrometer. The chemical shifts were refer-
complex.
enced to the residual signals of the deuterated solvents (C6D6:
1H: δ 7.16 ppm, 13C: δ 128.38 ppm, CDCl3: 1H: δ 7.26 ppm,
CD2Cl2: 1H: δ 5.32 ppm). Elemental analyses were performed
by the Laboratoire d’analyse élémentaire (Université de Mon-
tréal). Molecular weight analyses were performed on a Waters
Conclusions
For zinc-based diketiminate catalysts, the flexible ligand frame- 1525 gel permeation chromatograph equipped with three
work provided by N-CH2R substituents does not seem to offer Phenomenex columns and a refractive index detector at 35 °C.
a significant advantage in polymerization performance. As THF was used as the eluent at a flow rate of 1.0 mL min−1 and
observed for the more rigid N-aryl substituents, activities in polystyrene standards (Sigma-Aldrich, 1.5 mg mL−1, prepared
rac-lactide polymerization decrease with increasing substi- and filtered (0.2 mm) directly prior to injection) were used
tution of the diketimine,23,24,26,28–30,37 and the activity of for calibration. Obtained molecular weights were corrected
sterically demanding 6c (kp ≈ 0.1 M−1
of the range obtained with N-alkyl substituted diketimines
(kp ≈ 0.1–0.5 M−1 −1).22 Stereoselectivities follow the general
s
−1) is at the lower end by a Mark–Houwink factor of 0.58.66
N,N′-Di(9-anthrylmethyl)-2-amino-4-imino-2-pentene,
nacnacAnH, 5a
s
trend for zinc diketiminate complexes: a heterotactic preference
reinforced by sterically demanding ligands.
In a sealed vessel acetylacetone ethylene glycol monoketal
For magnesium complexes, on the other hand, N-alkyl (0.70 g, 4.39 mmol) and (9-anthryl)methylamine (1.82 g,
diketiminate ligands showed significant improvements 8.79 mmol) were added. The vessel was closed and its contents
when compared to their more rigid N-aryl analogues. With stirred at 110 °C for 4 hours or until the mixture had comple-
bulky 5d, complete conversion of the monomer was achieved tely solidified. The solid was washed with cold diethyl
in less than one minute (kp ≈ 50 M−1 −1), faster than its ether, dried under reduced pressure and used without
s
analogue with the sterically bulky 2,6-diisopropylphenyl sub- further purification (1.30 g, 62%, 90% purity according to
stituent (kp = 11 M−1 s−1) and thus one of the fastest mag- NMR).
nesium based catalysts reported.42 Even more importantly,
1H NMR (400 MHz, C6D6): δ 11.70 (s, 1H, NH), 7.88 (s, 2H,
the isotactic preference previously observed for 3 was again Ar), 7.68 (m, 8H, Ar), 7.09 (m, 4H, Ar), 6.81 (m, 4H, Ar), 4.73
obtained (slightly increased) in its chlorinated analogue 6e. (s, 5H, CH, NCH2), 1.86 (s, 6H, CH3). 13C{1H} NMR (100 MHz,
This is in contrast to the respective N-aryl complexes which C6D6): 194.3 (CvN), 131.5 (Ar), 131.2 (Ar), 130.1 (Ar), 128.0
display slight to strong heterotactic preferences. The reversal (Ar), 127.3 (Ar), 125.2 (Ar), 124.5 (Ar), 124.4 (Ar), 94.9
of stereopreference might be associated with the ability of (HC(CN)2), 43.4 (NCH2), 19.8 (Me). EI-HR-MS (m/z): calcd for
the ligand framework to adapt itself to the last inserted C35H30N2 [M + H]+: 479.2482; found: 479.2485.
monomer unit.62 The few magnesium complexes which show
The main contamination was the monocondensation
isotactic preference for lactide insertion all contain tridentate product 4-(9-anthryl)methylamino-3-penten-2-one, acnacAnH,
ligands, which require partial dissociation prior to lactide which was also the main product in unsuccessful conden-
coordination and might thus offer a similar degree of sation attempts described in the text. 1H-NMR (CDCl3,
ligand flexibility.45–47 However, significant increases in stereo- 400 MHz): δ 10.97 (s, 1H, NH), 8.48 (s, 1H, Ar), 8.20 (m, 2H,
selectivity will be required to enable us to investigate a poten- Ar), 8.01 (m, 2H, Ar), 7.57 (m, 2H, Ar), 7.48 (m, 2H, Ar), 5.37 (d,
tial mechanistic relationship between ligand flexibility and 2H, NCH2), 5.05 (s, 1H, CH), 2.29 (s, 1H, CH3), 1.90 (s, 1H,
stereocontrol.
CH3). 13C{1H} NMR (100 MHz, CDCl3): 195.1 (CvO), 162.2
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6339–6352 | 6347

View Article Online
Paper
Dalton Transactions
(CvN), 131.7 (Ar), 130.4 (Ar), 129.5 (Ar), 128.7 (Ar), 127.4 (Ar), Calcd for C41H47MgN3Si2: C, 74.35; H, 7.15; N, 6.34. Found C,
126.9 (Ar), 125.2 (Ar), 123.5 (Ar), 95.7 (HCCN), 39.8 (NCH2), 74.45; H, 7.25; N, 6.20.
28.9 (Me), 19.8 (Me). Anal. Calcd for C20H19NO: C, 83.01; H,
nacnacAnMgOtBu, 5e
6.62; N, 4.84. Found: C, 82.84; H, 6.54; N, 4.80.
A solution of 5d in C6D6 was slowly titrated with a solution of
nacnacAnZnN(SiMe3)2, 5b
tert-butanol in C6D6. After addition of 0.25 equiv. the solution
1
contained 5d and 5e. H-NMR (C6D6, 400 MHz): δ 8.82 (d, 4H
5a (0.34 g, 0.71 mmol) was added to
a solution of
Ar), 8.24 (d, 4H Ar), 8.20 (s, 2H Ar), 8.10 (s, 2H Ar), 7.82 (d, 4H
Ar), 7.76 (d, 4H Ar), 5.80 (s, 4H, NCH2), 4.43 (s, 1H, CH), 1.53
(s, 6H, (NC)Me), 1.10 (s, 9H, OCMe3). Further addition of tert-
butanol led to the formation of the ligand 5a and unidentified
side-products.
Zn(N(SiMe3)2)2 (0.27 g, 0.71 mmol) in toluene or hexane in the
glove box, forming a suspension. Stirring was continued for
4 days or until the mixture became homogeneous. Solvent
and formed amine were removed under reduced pressure.
The crude solid was used in further reactions without purifi-
cation. Recrystallisation from toluene yielded analytically pure,
colourless crystals (0.38 g, 76%).
N,N′-Dibenzyl-2-amino-3-chloro-4-imino-2-pentene,
ClnacnacBnH, 6a
1H NMR (400 MHz, C6D6): 8.33 (s, 2H, Ar), 8.20 (m, 8H, Ar),
7.91 (m, 4H, Ar), 7.37 (m, 4H, Ar), 5.54 (s, 4H, NCH2), 4.41 (s,
1H, CH), 1.71 (s, 6H, Me), −0.59 (s, 18H, SiMe3). 13C{1H} NMR
(100 MHz, C6D6): 169.5 (CvN), 131.9 (Ar), 131.7 (Ar), 130.4
(Ar), 129.4 (Ar), 127.9 (Ar), 126.0 (Ar), 124.9 (Ar), 124.8 (Ar),
96.2 (HC(CN)2), 48.7 (NCH2), 24.7 (Me), 4.6 (SiMe3). Anal.
Calcd for C41H47N3Si2Zn: C, 70.01; H, 6.74; N, 5.97. Found:
C, 69.72; H, 6.57; N, 5.73.
To a solution of nacnacBnH (5.48 g, 19.7 mmol) in dry THF
(150 mL) was added N-chlorosuccinimide (3.00 g, 22.4 mmol).
After stirring at room temperature for 45 minutes, a white pre-
cipitate formed, which was removed by filtration. H2O
(500 mL) was then added. The product was extracted using
hexanes (2 × 600 mL). After drying over Na2SO4, the solvent
was evaporated. The obtained yellow oil was crystallized from
dry ethanol at −80 °C, washed with cold dry ethanol and
recrystallized from refluxing ethanol. The eluate yielded a
second fraction at −80 °C (colourless crystals, 3.41 g, 55%).
1H-NMR (CDCl3, 400 MHz, 298 K): δ 12.22 (bs, 1H, NH),
7.25–7.19 (m, 10H, Ph), 4.49 (s, 4H, NCH2), 2.18 (s, 6H, Me).
13C{1H} NMR (CDCl3, 75 MHz, 298 K): δ 160.3 (CvN), 140.5
(ipso Ph), 128.6 (ortho or meta Ph), 127.3 (ortho or meta Ph),
126.8 (para Ph), 127.7 (ClC), 51.5 (NCH2), 17.3 (Me). Anal.
Calcd for C19H21ClN2: C, 72.95; H, 6.77; N, 8.95. Found
C, 72.89; H, 6.66; N, 9.17.
nacnacAnZnOiPr·CH2Cl2, 5c
To a solution of 5b (0.38 g, 0.54 mmol) in toluene (5 mL), iso-
propanol (41 µL, 0.54 mmol) was added and allowed to react
for 3 h during which time a white precipitate appeared. The
solvent was removed under reduced pressure, yielding the
crude product as a white solid. Purification was carried out by
recrystallisation in dichloromethane (0.11 g, 34%).
1H NMR (400 MHz, CDCl3): 8.35 (m, 4H, Ar), 8.22 (s, 2H,
Ar), 7.82 (m, 4H, Ar), 7.25 (m, 8H, Ar), 5.37 (s, 4H, NCH2), 4.12
(s, 1H, CH), 3.89 (sept., 1H, CHMe2), 1.39 (s, 6H, Me), 1.07 (d, ClnacnacBnZnOiPr, 6c
6H, CHMe2). 13C{1H} NMR (100 MHz, CDCl3): 169.4 (CvN),
Zn(N(SiMe3)2)2 (0.86 g, 2.2 mmol) was added to a solution of
133.4 (Ar), 131.6 (Ar), 130.3 (Ar), 129.1 (Ar), 127.1 (Ar),
125.6 (Ar), 125.4 (Ar), 124.7 (Ar), 94.3 (HC(CN)2), 53.6 (CHMe2),
49.5 (NCH2), 28.1 (Me), 24.7 (Me). Anal. Calcd for
C38H36N2OZn·CH2Cl2: C, 68.18; H, 5.57; N, 4.08. Found:
C, 68.08; H, 5.29; N, 4.10. (One equivalent of dichloromethane
per Zn was also observed in the crystal structure of 5c.)
6a (0.70 g, 2.2 mmol) in toluene (30 mL) over 5 min. The
solvent was removed under reduced pressure, yielding Clnac-
nacBnZnN(SiMe3)2, 6b, as an orange oil (1.25 g, 2.21 mmol,
1H NMR (400 MHz, C6D6): 7.22–7.14 (m, 10H, Ph), 4.85 (s, 4H,
NCH2), 2.21 (s, 6H, Me(CvN)2), 0.20 (s, 18H, SiMe3).
Toluene (30 mL) followed by isopropanol (0.13 g,
2.20 mmol) were added. After 5 min of stirring, a white pre-
cipitate formed, which was isolated by filtration and washed
nacnacAnMgN(SiMe3)2, 5d
5a (1.07 g, 2.2 mmol) was added to
Mg(N(SiMe3)2)2 (0.86 g, 2.5 mmol) in toluene or hexane,
a
solution of with toluene (15 mL). 0.30 g, 26%.
1H NMR (400 MHz, C6D6): δ 7.16–7.05 (m, 10H, Ph), 4.57 (s,
forming a suspension. The mixture was allowed to stir over- 4H, NCH2), 3.85 (sep., 1H, OC(H)Me2), 2.11 (s, 6H, Me(CvN)2),
night, after which the solvent and formed amine were removed 0.96 (d, 6H, OC(H)Me2). 13C{1H} NMR (75 MHz, C6D6): δ 169.5
under reduced pressure. The crude solid could be used (CvN), 140.6 (Ph), 128.3 (Ph), 126.5 (Ph), 126.4 (Ph), 101.2
either directly for further reactions or recrystallized from (ClC), 66.2 (OCHMe2 or NCH2), 55.0 (OCHMe2 or NCH2), 27.9
toluene (colourless crystals, 1.21 g, 82%).
(Me), 20.8 (Me). Anal. Calcd for C22H27ClN2OZn: C, 60.56;
1H NMR (400 MHz, C6D6): δ 8.27 (d, 4H, Ar), 8.11 (s, 2H, H, 6.24; N, 6.42. Found C, 60.10; H, 6.21; N, 6.35.
Ar), 7.76 (d, 4H, Ar), 7.27 (8H, Ar), 5.25 (s, 4H, NCH2), 4.69 (s,
ClnacnacBnMgOtBu, 6e
1H, CH), 1.83 (s, 6H, (NC)Me), −0.50 (s, 18H, SiMe3). 13C{1H}
NMR (100 MHz, C6D6): δ 170.2 (CvN), 132.3 (Ar), 131.5 (Ar), To a solution of 6a (0.100 g, 0.32 mmol) in anhydrous THF
131.2 (Ar), 129.8 (Ar), 128.6 (Ar), 126.8 (Ar), 125.3 (Ar), 124.7 (6 mL), Mg(N(SiMe3)2)2 (0.110 g, 0.32 mmol) was added.
(Ar), 95.8 (CH), 47.9 (NCH2), 24.1 ((NC)Me), 4.4 (SiMe3). Anal. The yellow-golden solution was stirred overnight followed
6348 | Dalton Trans., 2014, 43, 6339–6352
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
by solvent evaporation under reduced pressure yielding a ArMe), 1.80 (s, 6H, ArMe) 0.00 (s, 18H, SiMe3). 13C{1H} NMR
yellow solid identified as a mixture of ClnacnacBnMgOtBu, 6d, (75 MHz, C6D6): δ 169.7 (CvN), 137.1 (Ar), 137.0 (Ar), 134.3
and (ClnacnacBn)2Mg, 6f. After re-dissolution in THF (6 mL), (Ar), 130.4 (Ar), 95.1 (CH), 48.8 (NCH2), 23.4 (Me), 20.9 (Me),
tert-butanol (0.024 g, 0.32 mmol) was added. The reaction 20.7 (Me), 4.6 (SiMe3). Anal. Calcd for C31H51MgN3Si2: C,
mixture was stirred for 3 hours. Evaporating the solvent 68.16; H, 9.41; N, 7.69. Found C, 0.23; H, 9.50; N, 7.70.
yielded a yellow solid, which was washed with hexane and
nacnacMesMgOtBu·1/2CH2Cl2, 7e
dried under reduced pressure. Recrystallisation from CH2Cl2–
hexane at room temperature afforded colourless crystals after As in 5c, substituting 5b for 7d (0.13 g, 0.24 mmol), isopro-
24 h (0.04 g, 33%).
1H-NMR (C6D6, 400 MHz): δ 7.35–6.95 (m, 10H, Ph), 4.68 (s, after recrystallisation from dichloromethane (72 mg, 66%).
4H, NCH2), 2.19 (s, 6H, (NC)Me), 0.91 (s, 9H, OCMe3). 13C{1H} 1H-NMR (400 MHz, C6D6): δ 6.82 (s, 4H, Ar), 4.85 (s, 4H,
panol for tert-butanol (23 µL, 0.24 mmol). Colourless crystals
NMR (100 MHz, C6D6): δ 170.6 (CvN), 141.0 (ipso Ph), 128.7 NCH2), 4.67 (s, 1H, CH), 2.48 (s, 12H, ArMe), 2.16 (s, 6H, (CN)
(ortho or meta Ph), 128.1 (ortho or meta Ph), 127.8 (para Ph), Me), 1.82 (s, 6H, ArMe), 1.06 (s, 9H, OCMe3). 13C{1H} NMR
126.7 (CCl), 67.5 (OCMe3), 53.5 (NCH2), 33.8 (OCMe3), 20.9 (100 MHz, C6D6): δ 169.9 (CvN), 136.5 (ipso Mes), 135.7
((NC)Me). Anal. Calcd for C23H29ClMgN2O: C, 67.50; H, 7.14; (ortho/meta Mes), 135.3 (ortho/meta Mes), 130.5 (para Mes),
N, 6.85. Found C, 66.28; H, 7.28; N, 6.74.
80.9 (CH), 67.2 (OC(CH3)3), 44.9 (NCH2), 33.1 (Me, ortho), 23.5
(Me, para), 21.6 (OCMe3), 20.9 (NCMe). Anal. Calcd for
C29H42MgN2O·1/2CH2Cl2: C, 70.66; H, 8.64; N, 5.59. Found C,
(ClnacnacBn)2Mg, 6f
To a solution of 6a (0.100 g, 0.32 mmol) in anhydrous toluene 69.65; H, 8.68; N, 5.68. (Half an equivalent of CH2Cl2 per Mg
(5 mL), Mg(N(SiMe3)2)2 (0.055 g, 0.16 mmol) was added. The was observed in the crystal structure.)
yellow-golden solution was stirred overnight. After evaporation
N,N′-Di(9-anthrylmethyl)-2-amino-3-chloro-4-imino-2-pentene,
of the solvent, the product, a bright solid, was identified by
NMR spectroscopy. The obtained solid was re-dissolved in a
ClnacnacAnH, 8a
minimum of toluene and crystallized at −35 °C to yield yellow To a solution of 5a (1.00 g, 2.1 mmol) in anhydrous THF
crystals (0.09 g, 91%).
(40 mL) was added N-chlorosuccinimide (0.30 g, 2.3 mmol).
1H-NMR (C6D6, 400 MHz): δ 7.28–6.89 (m, 20H, Ph), 4.15 (s, After stirring at room temperature for 45 minutes, a white pre-
8H, NCH2), 2.05 (s, 12H, Me). 13C{1H} NMR (100 MHz, C6D6): cipitate formed, which was removed by filtration. Water was
δ 169.3 (CvN), 140.9 (ipso Ph), 128.5 (ortho/meta Ph), 128.1 then added. The product was extracted using dichloromethane
(ortho/meta Ph), 126.8 (para Ph), 126.6 (CCl), 53.3 (NCH2), 20.1 (2 × 10 mL) and dried over Na2SO4. Evaporation of the solvent
(Me). Anal. Calcd for C38H40Cl2MgN4: C, 70.44; H, 6.22; N, yielded a yellow oil, which was crystallized from anhydrous
8.65. Found C, 70.33; H, 6.37; N, 8.48.
dichloromethane (0.45 g, 42%).
1H NMR (400 MHz, C6D6): 12.32 (s, 1H, NH), 7.76 (s, 2H,
Ar), 7.56 (m, 8H, Ar), 7.06 (m, 4H, Ar), 6.87 (m, 4H, Ar), 4.57 (s,
5H, CH, NCH2), 2.12 (s, 6H, Me). 13C{1H} NMR (100 MHz,
N,N′-Di(mesitylmethyl)-2-amino-4-imino-2-pentene,
nacnacMesH, 7a
In a sealed vessel with a stir bar, acetylacetone ethylene glycol C6D6): 159.6 (CvN), 131.1 (Ar), 129.8 (Ar), 129.1(Ar), 127.5 (Ar),
monoketal (0.19 g, 1.68 mmol) and mesitylmethylamine (0.5 g, 125.3 (Ar), 125.1 (Ar), 124.4 (Ar), 124.0 (Ar), 100.6 (CCl), 44.1
3.35 mmol) were added along with an activated 3 Å molecular (NCH2), 17.2 (Me). Anal. Calcd for C35H29ClN2: C, 81.93; H,
sieve. The vessel was closed and heated to 110 °C for 4 hours 5.70; N, 5.46. Found C, 81.73; H, 5.70; N, 5.47.
or until the reaction had completely solidified. The solid
ClnacnacAnMgN(SiMe3)2, 8d
product was dissolved in ethanol and crystallized (0.32 g,
52%).
Following the procedure described for 5d, 5a (0.12 g,
1H NMR (400 MHz, C6D6): 11.17 (s, 1H, NH), 6.72 (s, 4H, 0.23 mmol), Mg(N(SiMe3)2)2 (0.08 g, 0.23 mmol), (crude 0.14 g,
Ar), 4.63 (s, 1H, CH), 4.10 (s, 4H, NCH2), 2.28 (s, 6H, ArCH3), 87%) colourless crystals (0.05 g, 28%).
2.00 (s, 12H, ArCH3), 1.80 (s, 6H, CH3). 13C{1H} NMR
1H NMR (400 MHz, C6D6): δ 8.13 (s, 2H, Ar), 8.10 (d, 4H,
(100 MHz, C6D6): 159.9 (CvN), 136.8 (ipso Mes), 135.5 (ortho/ Ar), 7.75 (d, 4H, Ar), 7.31–7.21 (m, 8H, Ar), 5.16 (s, 4H, NCH2),
meta Mes), 134.3 (ortho/meta Mes), 129.2 (para Mes), 94.7 (HC 2.31 (s, 6H, (NC)Me), −0.57 (s, 18H, SiMe3). 13C{1H} NMR
(CN)2), 44.9 (NCH2), 21.2 (Me), 19.5 (Me), 19.4 (Me). Anal. (100 MHz, C6D6): δ 169.3 (CvN), 132.2 (Ar), 131.2 (Ar), 130.5
Calcd for C25H42N2: C, 82.82; H, 9.45; N, 7.73. Found C, 82.84; (Ar), 129.9 (Ar), 129.1 (Ar), 127.0 (Ar), 125.4 (Ar), 124.4 (Ar),
H, 9.46; N, 7.89.
102.4 (ClC(CN)2), 48.9 (NCH2), 21.4 (Me), 4.4 (SiMe3). Anal.
Calcd for C41H46ClMgN3Si2: C, 70.68; H, 6.65; N, 6.03. Found
C, 71.27; H, 6.47; N, 5.71.
nacnacMesMgN(SiMe3)2, 7d
Following the procedure described for 5d, 7a (0.30 g,
0.83 mmol), Mg(N(SiMe3)2)2 (0.29 g, 0.83 mmol) to yield
colourless crystals (0.17 g, 37%).
X-ray diffraction studies
Single crystals were obtained as described in Table S4.† Diffr-
1H NMR (400 MHz, C6D6): δ 6.74 (s, 4H, Ar), 4.64 (s, 1H, action data were collected with Cu Kα radiation on a Bruker
CH), 4.28 (s, 4H, NCH2), 2.28 (s, 6H, (CN)Me), 2.09 (s, 12H, Smart 6000 or a Bruker Microstar/Proteum, both equipped
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6339–6352 | 6349

View Article Online
Paper
Dalton Transactions
Table 4 Details of X-ray diffraction studies
5b
5c
5d
5f
6e
Formula
C41H47N3Si2Zn·C7H8 C76H72N4O2Zn2·2CH2Cl2 C41H47MgN3Si2·2CH2Cl2 C82H68Mg2N4O2·2.5CH2Cl2 C46H58Cl2Mg2N4O2
Mw (g mol−1);
795.50; 1.246
1373.97; 1.387
747.23; 1.224
1402.33; 1.314
818.48; 1.236
dcalcd (g cm−3
)
Crystal size (mm) 0.15·0.15·0.18
0.27·0.27·0.36
150; 716
Triclinic
ˉ
P1
11.8196(4)
12.4402(5)
12.4806(5)
95.050(1)
115.269(2)
92.814(1)
0.16·0.09·0.09
200; 792
Triclinic
ˉ
P1
11.4629(5)
12.8758(6)
14.0056(6)
92.665(3)
98.376(3)
96.507(3)
0.18·0.14·0.11
100; 1466
Triclinic
ˉ
P1
14.4200(5)
16.6394(6)
17.0800(6)
72.642(2)
73.431(2)
67.467(2)
3544.4(2); 2
3–73; 0.96
0.14·0.14·0.14
150; 872
Monoclinic
P21/c
9.1499(2)
22.1705(3)
10.8809(2)
T (K); F(000)
Crystal system
Space group
Unit cell: a (Å)
b (Å)
c (Å)
α (°)
150; 1688
Triclinic
P1
ˉ
13.0063(3)
15.7472(4)
20.9693(5)
84.7981(11)
86.8527(12)
83.0374(10)
4241.4(2); 4
2–73; 0.96
β (°)
95.103(1)
γ (°)
V (Å3); Z
1645.4(1); 1
4–72; 0.95
2027.74(16); 2
3–73; 0.96
2198.53(7); 2
4–73; 0.97
θ range (°);
completeness
Collected
reflections; Rσ
Unique
56 188; 0.026
16 111; 0.034
1.246; multiscan
0.040; 0.114
0.045; 0.118
0.91
21 587; 0.021
6223; 0.035
26 826; 0.070
7708; 0.089
47 715; 0.039
13 541; 0.035
2.447; multiscan
0.049; 0.138
28 207; 0.038
4271; 0.072
reflections; Rint
μ (mm−1); Abs.
Corr.
2.790; multiscan
0.046; 0.114
0.047; 0.115
2.402; multiscan
0.054; 0.144
0.070; 0.153
1.926; multiscan
0.058; 0.155
0.066; 0.183
R1(F); wR(F2)
(I > 2σ(I))
R1(F); wR(F2)
(all data)
GoF(F2)
0.062; 0.147
0.86
1.01
1.08
1.15
Residual electron 1.08; −0.73
density
0.57; −0.80
0.71; −0.55
0.41; −0.42
0.47; −0.72
6f
7d
7e
8f
Formula
C38H40Cl2MgN4
647.9; 1.290
0.18·0.16·0.16
150; 1368
C31H51MgN3Si2
546.24; 1.111
0.14·0.11·0.07
100; 1192
C29H43MgN2O·0.5CH2Cl2
501.42; 1.167
0.15·0.15·0.15
100; 2168
C70H56Cl2MgN4
1048.39; 1.224
0.13·0.08·0.08
100; 1100
Mw (g mol−1); dcalcd (g cm−3
Crystal size (mm)
T (K); F(000)
)
Crystal system
Space group
Unit cell: a (Å)
b (Å)
c (Å)
α (°)
Monoclinic
P21/c
14.5416(5)
11.1177(4)
20.7327(7)
Orthorhombic
Pnma
17.4882(2)
18.0767(2)
Monoclinic
P21/n
15.0133(8)
17.0252(9)
22.8695(13)
Triclinic
P1
ˉ
12.5041(8)
14.3017(8)
18.1609(10)
104.305(2)
109.464(2)
100.227(2)
2844.6(3); 2
3–58; 0.98
52 174; 0.046
7765; 0.029
1.484; multiscan
0.048; 0.136
0.051; 0.139
1.04
10.3297(1)
β (°)
95.640(2)
102.436(2)
γ (°)
V (Å3); Z
3335.6(2); 4
3–73; 0.97
3265.52(6); 4
5–72; 0.98
5708.4(5); 8
3–72; 0.99
θ Range (°); completeness
Collected reflections; Rσ
Unique reflections; Rint
μ (mm−1); Abs. Corr.
R1(F); wR(F2) (I > 2σ(I))
R1(F); wR(F2) (all data)
GoF(F2)
40 209; 0.021
6473; 0.037
2.185; multiscan
0.039; 0.107
0.043; 0.111
1.05
45 504; 0.011
3233; 0.025
1.334; multiscan
0.034; 0.095
0.035; 0.096
1.06
80 320; 0.026
11 142; 0.042
1.567; multiscan
0.054; 0.150
0.059; 0.155
1.05
Residual electron density
0.28; −0.43
0.40; −0.28
1.01; −0.71
0.30; −0.52
with Helios MX mirror optics and rotating anode sources, or solution after application of SQUEEZE. No notable structural
on a Bruker Microsource/APEX2 using the APEX2 software difference was found between solutions and thus the latter was
package.67 Data reduction was performed with SAINT,68 and used. Further experimental details can be found in Table 4
absorption corrections with SADABS.69 Structures were solved and in the ESI† (CIF).
with direct methods (SHELXS97).70 All non-hydrogen atoms
were refined anisotropically using full-matrix least-squares on
F2 and hydrogen atoms refined with fixed isotropic U using a
Acknowledgements
riding model (SHELXL97).70 In 8f, disordered solvent was
found around the inversion center (93 electrons per unit cell),
but the best modeled solution remained 12% higher in wR2
(with appr. twice as high errors in bond lengths) than the
Johannes T. Wendler’s, Simon Cassegrain’s and Marine Cros’
contributions to these studies during their internships are
gratefully acknowledged. We thank F. Bélanger for support
6350 | Dalton Trans., 2014, 43, 6339–6352
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
with X-ray diffraction studies and Prof. R. E. Prud’homme and 22 F. Drouin, P. O. Oguadinma, T. J. J. Whitehorne,
P. Ménard-Tremblay for access to GPC. Funding was provided
by NSERC, Centre en chimie verte et catalyse (CCVC) and
FQRNT (Ph.D. stipend for T. W.).
R. E. Prud’homme and F. Schaper, Organometallics, 2010,
29, 2139–2147.
23 M. Cheng, A. B. Attygalle, E. B. Lobkovsky and
G. W. Coates, J. Am. Chem. Soc., 1999, 121, 11583–
11584.
24 B. M. Chamberlain, M. Cheng, D. R. Moore, T. M. Ovitt,
E. B. Lobkovsky and G. W. Coates, J. Am. Chem. Soc., 2001,
123, 3229–3238.
References
1 R. G. Patel and D. J. Sen, Int. Pharm. Sci., 2011, 1, 29.
2 K. Kishan and S. Carmen, in Degradable Polymers and
25 M. H. Chisholm, J. C. Huffman and K. Phomphrai,
J. Chem. Soc., Dalton Trans., 2001, 222–224.
Materials: Principles and Practice, 2nd edn, American 26 M. H. Chisholm, J. Gallucci and K. Phomphrai, Inorg.
Chemical Society, 2012, vol. 1114, pp. 3–10. Chem., 2002, 41, 2785–2794.
3 G. E. Luckachan and C. K. S. Pillai, J. Polym. Environ., 2011, 27 M. H. Chisholm and K. Phomphrai, Inorg. Chim. Acta,
19, 637–676. 2003, 350, 121.
4 R. C. Thompson, C. J. Moore, F. S. v. Saal and S. H. Swan, 28 A. P. Dove, V. C. Gibson, E. L. Marshall, A. J. P. White and
Philos. Trans. R. Soc. London, Ser. B, 2009, 364, 2153–2166. D. J. Williams, Dalton Trans., 2004, 570–578.
5 J. Hammer, M. S. Kraak and J. Parsons, in Reviews of 29 M. H. Chisholm, J. C. Gallucci and K. Phomphrai, Inorg.
Environmental Contamination and Toxicology, ed. Chem., 2005, 44, 8004–8010.
D. M. Whitacre, Springer, New York, 2012, vol. 220, pp. 30 C. N. Ayala, M. H. Chisholm, J. C. Gallucci and
1–44. C. Krempner, Dalton Trans., 2009, 9237–9245.
6 D. K. A. Barnes, F. Galgani, R. C. Thompson and M. Barlaz, 31 R. Tong and J. Cheng, Angew. Chem., Int. Ed., 2008, 47,
Philos. Trans. R. Soc. London, Ser. B, 2009, 364, 1985–1998. 4830–4834.
7 M. Cole, P. Lindeque, C. Halsband and T. S. Galloway, Mar. 32 K. Yu and C. W. Jones, J. Catal., 2004, 222, 558.
Pollut. Bull., 2011, 62, 2588–2597.
8 C. K. Williams and M. A. Hillmyer, Polym. Rev., 2008, 48,
1–10.
9 J. Ahmed and S. K. Varshney, Int. J. Food Properties, 2011,
14, 37–58.
10 S. Inkinen, M. Hakkarainen, A.-C. Albertsson and
A. Södergård, Biomacromolecules, 2011, 12, 523–532.
33 K. Yu and C. W. Jones, Polym. Prepr. (Am. Chem. Soc., Div.
Polym. Chem.), 2004, 45, 1005.
34 M. Kroeger, C. Folli, O. Walter and M. Doering, Adv. Synth.
Catal., 2006, 348, 1908–1918.
35 M. Helou, O. Miserque, J.-M. Brusson, J.-F. Carpentier and
S. M. Guillaume, Macromol. Rapid Commun., 2009, 30,
2128–2135.
11 S. Dutta, W.-C. Hung, B.-H. Huang and C.-C. Lin, in Syn- 36 V. Poirier, M. Duc, J.-F. Carpentier and Y. Sarazin, Chem-
thetic Biodegradable Polymers, ed. B. Rieger, A. Künkel, SusChem, 2010, 3, 579–590.
G. W. Coates, R. Reichardt, E. Dinjus and T. A. Zevaco, 37 H.-Y. Chen, B.-H. Huang and C.-C. Lin, Macromolecules,
Springer-Verlag, Berlin, 2011, pp. 219–284. 2005, 38, 5400–5405.
12 P. J. Dijkstra, H. Du and J. Feijen, Polym. Chem., 2011, 2, 38 T. Athar, A. Hakeem and N. Topnani, J. Chil. Chem. Soc.,
520–527. 2011, 56, 887–890.
13 J.-F. Carpentier, Macromol. Rapid Commun., 2010, 31, 1696– 39 H.-Y. Chen, Y.-L. Peng, T.-H. Huang, A. K. Sutar,
1705.
S. A. Miller and C.-C. Lin, J. Mol. Catal. A: Chem., 2011, 339,
61–71.
14 N. Ajellal, J.-F. Carpentier, C. Guillaume, S. M. Guillaume,
M. Helou, V. Poirier, Y. Sarazin and A. Trifonov, Dalton 40 L. F. Sanchez-Barba, D. L. Hughes, S. M. Humphrey and
Trans., 2010, 39, 8363. M. Bochmann, Organometallics, 2006, 25, 1012–1020.
15 C. A. Wheaton, P. G. Hayes and B. J. Ireland, Dalton Trans., 41 R. Tong and J. Cheng, Bioconjugate Chem., 2010, 21, 111–
2009, 4832–4846. 121.
16 R. H. Platel, L. M. Hodgson and C. K. Williams, Polym. 42 M. H. Chisholm, K. Choojun, J. C. Gallucci and
Rev., 2008, 48, 11–63. P. M. Wambua, Chem. Sci., 2012, 3, 3445–3457.
17 M. H. Chisholm and Z. Zhou, J. Mater. Chem., 2004, 14, 43 R. A. Collins, J. Unruangsri and P. Mountford, Dalton
3081. Trans., 2013, 42, 759–769.
18 I. El-Zoghbi, T. J. J. Whitehorne and F. Schaper, Dalton 44 Independent of their solid state structure, nacnacZn(OR)
Trans., 2013, 42, 9376–9387.
19 T. J. J. Whitehorne and F. Schaper, Chem. Commun., 2012,
48, 10334–10336.
20 T. J. J. Whitehorne and F. Schaper, Inorg. Chem., 2013, 52,
13612–13622.
complexes might be dimeric or monomeric in solution.
Since solution structures of the (pre-)catalysts influence
only the very first monomer insertion, they were not investi-
gated in detail and all complexes are presented as mono-
mers in the text to facilitate readability.
21 F. Drouin, T. J. J. Whitehorne and F. Schaper, Dalton Trans., 45 J.-C. Buffet, J. P. Davin, T. P. Spaniol and J. Okuda, New
2011, 40, 1396–1400.
J. Chem., 2011, 35, 2253–2257.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6339–6352 | 6351

View Article Online
Paper
Dalton Transactions
46 L. Wang and H. Ma, Macromolecules, 2010, 43, 6535– 58 A. P. Dove, V. C. Gibson, P. Hormnirun, E. L. Marshall,
6537.
J. A. Segal, A. J. P. White and D. J. Williams, Dalton Trans.,
2003, 3088–3097.
59 S. Latreche and F. Schaper, Inorg. Chim. Acta, 2011, 43,
49–53.
60 P. J. Bailey, R. A. Coxall, C. M. Dick, S. Fabre,
L. C. Henderson, C. Herber, S. T. Liddle, D. Loroño-Gonzá-
lez, A. Parkin and S. Parsons, Chem.–Eur. J., 2003, 9, 4820–
4828.
47 S. Song, H. Ma and Y. Yang, Dalton Trans., 2013, 42, 14200–
14211.
48 I. El-Zoghbi, A. Ased, P. O. Oguadinma, E. Tchirioua and
F. Schaper, Can. J. Chem., 2010, 88, 1040–1045.
49 J. Feldman, S. J. McLain, A. Parthasarathy, W. J. Marshall,
J. C. Calabrese and S. D. Arthur, Organometallics, 1997, 16,
1514–1516.
50 K. J. Fisher, Tetrahedron Lett., 1970, 2613–2616.
51 J. Vela, L. Zhu, C. J. Flaschenriem, W. W. Brennessel,
61 T. J. J. Whitehorne and F. Schaper, Private Communication
to the Cambridge Structural Database, 2013, CCDC 972594.
R. J. Lachicotte and P. L. Holland, Organometallics, 2007, 62 H. Ma, T. P. Spaniol and J. Okuda, Angew. Chem., Int. Ed.,
26, 3416–3423. 2006, 45, 7818–7821.
52 M. Cheng, D. R. Moore, J. J. Reczek, B. M. Chamberlain, 63 P. O. Oguadinma and F. Schaper, Organometallics, 2009, 28,
E. B. Lobkovsky and G. W. Coates, J. Am. Chem. Soc., 2001,
123, 8738–8749.
53 D. F. J. Piesik, S. Range and S. Harder, Organometallics,
2008, 27, 6178–6187.
54 D. R. Moore, M. Cheng, E. B. Lobkovsky and G. W. Coates,
Angew. Chem., Int. Ed., 2002, 41, 2599–2602.
6721–6731.
64 D. E. Stack, A. L. Hill, C. B. Diffendaffer and N. M. Burns,
Org. Lett., 2002, 4, 4487–4490.
65 D. Rivillo, H. Gulyás, J. Benet-Buchholz, E. C. Escudero-
Adán, Z. Freixa and P. W. N. M. van Leeuwen, Angew.
Chem., Int. Ed., 2007, 46, 7247–7250.
55 R. Boese, D. Blaser, T. Eisenmann and S. Schulz, Private 66 M. Save, M. Schappacher and A. Soum, Macromol. Chem.
communication to the Cambridge Structural Database, 2009,
Phys., 2002, 203, 889–899.
CCDC 730654.
67 APEX2, Bruker AXS Inc., Madison, USA, 2006.
56 S. Schulz, T. Eisenmann, D. Bläser and R. Boese, Z. Anorg. 68 SAINT, Bruker AXS Inc., Madison, USA, 2006.
Allg. Chem., 2009, 635, 995–1000.
57 G. Bendt, S. Schulz, J. Spielmann, S. Schmidt, D. Bläser
and C. Wölper, Eur. J. Inorg. Chem., 2012, 3725–3731.
69 G. M. Sheldrick, SADABS, Bruker AXS Inc., Madison, USA,
1996 & 2004.
70 G. M. Sheldrick, Acta Crystallogr., 2008, A64, 112–122.
6352 | Dalton Trans., 2014, 43, 6339–6352
This journal is © The Royal Society of Chemistry 2014