Synthesis, structure, and reactivity of tris(amidate) mono(amido) and tetrakis(amidate) complexes of group 4 transition metals

Article abstract of DOI:10.1039/c3dt51868j

The syntheses of a series of tris(amidate) mono(amido) titanium and zirconium complexes are reported. The binding motif of the amidate ligand has been determined to depend on the size of the metal centre for these sterically demanding N,O-chelating ligands; the larger zirconium metal centre supports three κ²-(N,O) bound amidate ligands while the titanium analogue has one ligand bound in a κ¹-(O) fashion to alleviate steric strain. Reactivity studies indicate that, despite high steric crowding about the tris(amidate) mono(amido) zirconium metal centre, transamination of the reactive dimethylamido ligand can be achieved using aniline. This complex is also an active precatalyst for intramolecular alkene hydroamination, in which protonolysis of one amidate ligand in the presence of excess amine is observed as an initiation step prior to catalytic turnover. Eight-coordinate homoleptic κ²-amidate complexes of zirconium and hafnium have also been prepared. This journal is The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c3dt51868j

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Synthesis, structure, and reactivity of tris(amidate)
mono(amido) and tetrakis(amidate) complexes of
group 4 transition metals†
Cite this: Dalton Trans., 2013, 42, 15670
Philippa R. Payne, Robert K. Thomson, Diane M. Medeiros, Geoff Wan and
Laurel L. Schafer*
The syntheses of a series of tris(amidate) mono(amido) titanium and zirconium complexes are reported.
The binding motif of the amidate ligand has been determined to depend on the size of the metal centre
for these sterically demanding N,O-chelating ligands; the larger zirconium metal centre supports three
κ2-(N,O) bound amidate ligands while the titanium analogue has one ligand bound in a κ1-(O) fashion to
alleviate steric strain. Reactivity studies indicate that, despite high steric crowding about the tris(amidate)
mono(amido) zirconium metal centre, transamination of the reactive dimethylamido ligand can be
achieved using aniline. This complex is also an active precatalyst for intramolecular alkene hydroamina-
tion, in which protonolysis of one amidate ligand in the presence of excess amine is observed as an
initiation step prior to catalytic turnover. Eight-coordinate homoleptic κ2-amidate complexes of zirconium
and hafnium have also been prepared.
Received 10th July 2013,
Accepted 9th September 2013
DOI: 10.1039/c3dt51868j
www.rsc.org/dalton
Introduction
In recent years, we and others have reported the synthesis of a
variety of monometallic mono and bis(N,O)-ligated group 3, 4,
and 5 metal complexes with broad applications in catalysis.1–10
These complexes include mono and bis(amidate) tantalum(V)
complexes for the hydroaminoalkylation of amines,10,11 mono
and bis(amidate) yttrium(III) complexes for the hydroamination
of aminoalkenes9,12 and the amidation of aldehydes,13 bis-
(2-pyridonate) titanium(IV) complexes for the random copolymeri-
zation of ε-caprolactone and rac-lactide,14 bis(ureate) zirco-
nium(IV) complexes15,16 as well as bis(amidate) zirconium(IV)
and titanium(IV) complexes for hydroamination7,12,15–22 and
olefin polymerization.2 These complexes can be generated
from the homoleptic metal amide or alkyl starting materials
(M(NR2)x or MRx) via a protonolysis reaction (Scheme 1) with
the organic proligand. In most cases, merely adjusting the
ligand to metal stoichiometry allows for the selective gen-
eration of targeted metal complexes.
Scheme 1 Protonolysis methodology using homoleptic metal amido or alkyl
starting materials.
Tris(amidate) yttrium complexes have been used to catalyze
the polymerization of ε-caprolactone23 and rac-lactide14 as well
as the amidation of aldehydes.13 However, to date, tris(N,O)-
ligated group 4 systems have not been investigated. In an effort
to further explore the coordination chemistry of group 4
amidate complexes, here we investigate the synthesis, charac-
terization, and reactivity of sterically crowded tris(amidate)
mono(amido) complexes. In some cases, even tetrakis(amidate)
complexes can be readily prepared.
Amidate ligands can adopt a variety of coordination modes
when binding to a metal; the ligands can bind in a monoden-
tate fashion, either κ1-(O) or κ1-(N), or in a bidentate κ2-(N,O)
motif, generating a four-membered metallacycle. When the
chelating binding mode is achieved, despite the large steric
requirements of the ligand and the small size of the group 4
d0 metal centres, complexes with expanded coordination
numbers often result. The structure and coordination geome-
try of these complexes in both the solution phase and the
solid state are of particular interest. Here we report trends in
amidate group 4 metal coordination and we show how both
metal and ligand size can impact observed bonding modes.
The resultant reactivity of these sterically congested complexes
Department of Chemistry, University of British Columbia, 2036 Main Mall,
Vancouver, BC, Canada, V6 T 1Z1. E-mail: schafer@chem.ubc.ca;
Fax: +1 (604)-822-2847; Tel: +1 (604)-822-9264
†Electronic supplementary information (ESI) available: Details of the crystal
structure determination of 1, 4, 5, 6, and 7, variable temperature NMR spectra
for 1, and representative spectra for the intramolecular hydroamination. CCDC
958398, 943792, 943794–943796. For ESI and crystallographic data in CIF or
other electronic format see DOI: 10.1039/c3dt51868j
15670 | Dalton Trans., 2013, 42, 15670–15677
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
is reported and most importantly, the reactivity of these
systems in catalytic hydroamination is explored.
The amidate ligands in this species are non-equivalent,
with two bound in a chelating κ2 fashion while the third
amidate ligand is monodentate. The κ1-amidate ligand is
bound via the more basic oxygen atom to this highly oxophilic
hard metal centre. The formation of this six-coordinate tita-
nium complex is not surprising, as expanded coordination
numbers for titanium are less common, and the steric require-
ments of these amidate ligands are significant for this small
metal centre.
Results and discussion
Coordination chemistry
Previous work has described the synthesis of bis(N-(2,6-diiso-
propylphenyl)benzamidate) bis(diethylamido) titanium,
a
robust catalyst for the hydroamination of alkynes,17,18,24,25 via
a protonolysis reaction of titanium tetrakis(diethylamido) with
the amide proligand. The amidate ligands in this monometal-
lic complex are bound in a κ2 fashion in the solid state and the
steric bulk of the nitrogen substituent helps to avoid bridging
interactions. Application of this strategy for the synthesis of
mono(amidate) tris(amido) complexes of group 4 metals has
not been successful, and instead complex product mixtures
result. However, such protonolysis reactions may be used for
the synthesis of tris(amidate) complexes. N-(2,6-Dimethylphe-
nyl)benzamide is an ideal proligand for this application as the
slightly reduced steric bulk of the nitrogen-bound aryl substi-
tuent is attractive for potentially accessing three coordinated
amidate ligands, while being large enough to promote the for-
mation of discrete monometallic species by discouraging brid-
ging interactions.
Complex 1 exhibits distorted octahedral geometry with the
amido ligand (N4) and the oxygen (O1) of one of the κ2-bound
amidates trans to one another. The steric properties of the
amidate ligands affect the geometric isomer observed in the
solid state, as to orient the bulky N-aryl substituents of the
κ2 ligands trans to one another. The Ti–O bond length of the
κ1-(O) amidate (1.857(1) Å) is considerably shorter than those of
the κ2 ligands (2.137(1) and 2.059(1) Å). While this short bond
length could result from steric effects, the near linear angle
(C31–O3–Ti1 151.3(1)°) of this κ1-(O) bound ligand is consist-
ent with a strongly π-donating alkoxide ligand. The Ti–N4
amido bond length (1.889(2) Å) is within the range expected
for M–N multiple bonds.26 The solid-state molecular structure
of the related 16e− bis(N-(2,6-dimethylphenyl)benzamidate)
bis(diethylamido) titanium complex has been reported pre-
viously.26 Comparison of the bond lengths in the bis(amidate)
structure to those of 1 indicate that addition of the third
amidate ligand results in a shortening of the average κ2
amidate Ti–N (2.116 cf. 2.293 Å) and lengthening of the Ti–O
bonds (2.098 cf. 2.040 Å), presumably due to exchange of the
4e− donating amido ligand with the more strongly π-donating
alkoxide ligand of the κ1-(O) amidate as well as the increased
steric congestion. The Ti–N4 amido bond length (1.889(2) Å)
is very similar to those found in the bis(amidate) analogue
(1.894(1) and 1.901(1) Å).26
The reaction of tetrakis(dimethylamido) titanium with
three equivalents of N-(2,6-dimethylphenyl)benzamide yields
an orange solid after removal of the volatiles (Scheme 2).
Single crystals suitable for X-ray crystallography can be
obtained from a layered toluene–pentane solution thereby con-
firming the monometallic tris(amidate) structure (Fig. 1).
1
The solution-phase H NMR spectroscopy of 1 is also con-
sistent with more than one ligand binding mode and displays
very broad signals at room temperature highlighting the fluxio-
nal behaviour of the ligands on the NMR timescale. The broad
signals can be attributed to rapid interconversion between the
κ1- and κ2-binding modes in the solution phase. Variable
temperature 1H NMR spectroscopy (25–80 °C, see ESI†)
resulted in coalescence and sharpening of the amidate signals,
consistent with increased fluxionality and the appearance of
averaged signals.
Scheme 2 Synthesis of tris(N-(2,6-dimethylphenyl)benzamidate) mono-
(dimethylamido) titanium 1.
Zirconium, as a larger metal centre, is better suited for
accommodating such steric congestion and is known to have
expanded coordination numbers.15,27,28 Indeed, the reaction of
three equivalents of the amide proligand with one equivalent
of tetrakis(dimethylamido) zirconium generated tris(amidate)
zirconium complexes 2–4 (Scheme 3).
These complexes have a variety of steric and electronic pro-
perties in the distal backbone position (R1, Scheme 3) as well
as on the nitrogen-bound aryl substituent (R2, Scheme 3).
Complexes 2 and 3 have also been prepared from the
corresponding bis(amidate) bis(dimethylamido) zirconium
Fig. 1 ORTEP depiction of solid-state molecular structure of tris(amidate)
complex 1. Ellipsoids are plotted at 50% probability and hydrogen atoms are
omitted for clarity, simplified core structure shown on the right. Selected bond
lengths [Å] and angles [°]: Ti–N1 2.082(1), Ti–N2 2.149(2), Ti–N4 1.889(2), Ti–O1
2.137(1), Ti–O2 2.059(1), Ti–O3 1.857(1), C31–O3–Ti1 151.3(1), sum about N4 360°. complexes via addition of another equivalent of the amide
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 15670–15677 | 15671

View Article Online
Paper
Dalton Transactions
aryl substituent at δ 2.31 ppm. In contrast, 3, with the more
sterically demanding isopropylphenyl groups on the nitrogen-
bound aryl substituent, exhibits hindered rotation leading to
inequivalent isopropyl signals in the 1H NMR spectrum.
The observed ligand fluxionality can result in disproportio-
nation reactions in the solution phase. Indeed, while com-
plexes 2–4 can be isolated readily, allowing the reaction
(Scheme 3) to proceed for extended periods of time (>24 hours)
can result in the formation of the bis(amidate) bis(amido) zirco-
nium and tetrakis(amidate) complexes. This is particularly
evident for complexes 3 and 4 which have the bulky diisopropyl-
phenyl substituents on the nitrogen atom. This ligand redistri-
bution is proposed to proceed via κ1-bound amidate ligands
that may engage in bridging interactions between metal centres.
These examples of ligand redistribution point toward potential
decomposition pathways for such systems. Notably, homo-
leptic amidate complexes can result in very sterically crowded
8-coordinate zirconium and hafnium complexes. To date,
related homoleptic Ti species have not been observed.
Scheme 3 Synthesis of tris(amidate) mono(dimethylamido) zirconium com-
plexes 2–4.
Fig. 2 ORTEP depiction of the solid-state molecular structure of tris(amidate)
complex 4. Ellipsoids plotted at 50% probability. The backbone naphthyl group
and hydrogen atoms are omitted for clarity. Selected bond lengths [Å]: Zr–N1
2.431(3), Zr–N2 2.403(3), Zr–N3 2.271(3), Zr–N4 2.018(3), Zr–O1 2.151(2),
Zr–O2 2.136(2), Zr–O3 2.208(2), sum about N4 360°.
The homoleptic complexes of both zirconium and hafnium
can be synthesized with excellent yields of 85% and 94%
respectively via protonolysis with the tetra(benzyl) metal pre-
cursors (Scheme 4). It is also possible via protonolysis with the
tetrakis(dimethylamido) zirconium starting material, as has
been shown for complex 5. To date, no notable reactivity has
proligand. A solid-state molecular structure was obtained for been observed with these complexes; however, similar homo-
the sterically demanding complex 4 confirming that zirco- leptic small-bite-angle ligated early transition metals have
nium, as a larger metal centre, is capable of having all three been used as MOCVD precursors.29–31 Solid-state structural
amidates bound in a κ2 manner (Fig. 2).
data can be obtained for both Zr and Hf homoleptic com-
Complex 4 adopts a distorted pentagonal bipyramidal geo- pounds and show that 5 and 6 are isostructural, both adopting
metry in the solid state. The axial positions are occupied by a dodecahedral geometry in the solid state with pseudo-D2d
the dimethylamido ligand (N4) and the oxygen (O1) of an symmetry (Fig. 3). For example, in complex 5 there exists the
amidate ligand. The remaining amidate donors (N1, N2, O2, C2-axis bisecting both N1 and N3, and N2 and N4. This axis
N3, and O3) define the distorted pentagonal plane. The Zr–N4 also lies at the intersection of the two pseudo-mirror planes
amido bond (2.018(3) Å) is consistent with a Zr–N double within the molecule. The two perpendicular C2 axes of sym-
bond.26 Comparison to the related bis(N-(2,6-dimethylphenyl)- metry lie between the aforementioned mirror planes, bisecting
pivalamidate) bis(dimethylamido) zirconium6 indicates that O1 and O2, as well as O1 and O4. It must be noted that homo-
while the Zr–N dimethylamido bond length is shorter in the leptic hafnium complex 6 displays equivalent binding of all of
tris(amidate) 4 (2.018(3) cf. 2.038(3) and 2.044(2) Å), the the amidate ligands while those of complex 5 are slightly dis-
amidate bond lengths on average are similar. Each of the torted. Each of the amidate ligands is once more bound in an
amidate ligands in complex 4 are in different coordination asymmetric fashion to the metal centre with a shorter Zr–O
environments. However, in every case the ligand is bound in bond compared with the Zr–N bond lengths (Δ 0.16–0.21 Å).
an asymmetric fashion to the metal centre with a shorter Zr–O The average metal to ligand bond lengths in the homoleptic
bond than Zr–N. The most symmetrically bound amidate (N3, structures (Zr–N 2.309, Zr–O 2.120 and Hf–N 2.325, Hf–O
O3) has a Zr–N and Zr–O bond length difference of 0.19 Å and 2.163 Å) are slightly shorter than those in tris(amidate) 4 (Zr–N
the most asymmetric (N1, O1) has a difference of 0.28 Å. The 2.368, Zr–O 2.164 Å).
bonding can therefore be most accurately described as an
alkoxy–imine bonding mode, and 4 is considered to be a 16e−
complex.
While the solid-state molecular structure of 4 demonstrates
that this tris(amidate) complex is C1-symmetric in the solid
state, the room temperature 1H NMR spectrum of 4 indicates a
C3-symmetric structure in solution, consistent with three equi-
valent κ2-bound amidate ligands on the NMR timescale. The
solution-phase spectroscopy of 2 is analogous, with one signal
Scheme 4 Synthesis of homoleptic amidate complexes of zirconium and
observed for the methyl substituents on the nitrogen-bound hafnium.
15672 | Dalton Trans., 2013, 42, 15670–15677
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Reactivity and catalysis
These high-coordinate metal complexes have a large amount
of steric bulk shielding the metal centre which could be detri-
mental for productive reactivity. Gratifyingly, reaction of the
coordinatively saturated tris(amidate) complex 3 with aniline
results in the formation of new tris(amidate) mono(anilido)
complex 7 as a yellow microcrystalline solid (Scheme 5).
1
The H NMR spectrum for complex 7 is very complicated,
with hindered rotation about all three amidate N–Cipso bonds
which results in twelve distinct doublets and six septet reso-
nances for the isopropyl methyl and methine protons respect-
ively. Variable temperature 1H NMR spectroscopy indicates a
coalescence of the isopropyl methyl signals at high tempera-
tures, and no significant change at low temperatures. Single
crystals of 7 suitable for solid-state molecular structure deter-
mination can be isolated from a 1 : 1 toluene–hexanes mixture
(Fig. 4).
The geometry of this compound is very similar to that of 4,
existing as a 7-coordinate complex best described as distorted
pentagonal bipyramidal. In complex 7 however, the amidate
axial position is occupied by the N2 donor, instead of an
oxygen donor. The remaining amidate donors (N1, O1, N3, O3,
and O2) define the pentagonal plane. Once again, the Zr–N4
amido bond (2.071(2) Å) is consistent with a Zr–N double bond
and the κ2-amidate ligands are bound to zirconium in an
alkoxy–imine bonding mode.
Fig. 3 ORTEP depiction of solid-state molecular structures of homoleptic
amidate complexes 5 and 6. Ellipsoids are plotted at 50% probability, and
hydrogen atoms are omitted for clarity. Selected bond lengths for complex 5
[Å]: Zr–N1 2.323(1), Zr–N2 2.323(1), Zr–N3 2.295(2), Zr–N4 2.288(1), Zr–O1
2.120(1), Zr–O2 2.114(1), Zr–O3 2.124(1), Zr–O4 2.123(1). Selected bond
lengths for complex 6 [Å]: Hf–N1 2.325(3), Hf–O1 2.163(3).
This transamination reaction indicates that the amido
group of these bulky systems is a reactive ligand and perhaps
productive catalytic reactivity could be observed with these
tris(amidate) zirconium complexes. One potential catalytic
The most notable features of these complexes are the
π-stacking interactions stabilizing the N-cis conformation,
which would otherwise be sterically unfavourable (Table 1).
The inter-ring distance from the ring centroids in complex 5
(3.610(3) Å) falls in the range typically expected for π-stacking
(∼3.6 Å)32,33 while that of 6 (4.023(5) Å) is slightly longer. This
is likely due to crystal packing resulting in the tilted nature of
the aryl rings in complex 6, demonstrated by the plane to
plane angle of 15.4(3)°. It is also possible that the additional
steric bulk of tert-butyl substituent on the backbone of 5 may
force the two π-stacked rings closer together. High symmetry is
observed in the solution-phase; the 1H NMR spectra of 5 and 6
are consistent with all four ligands being equivalent. Variable
Scheme 5 Generation of 7 via reaction of 3 with aniline.
1
temperature H NMR spectroscopy of complex 6 did not show
any discernible change over a wide range in temperatures.
With a thorough understanding of the solution-phase and
the solid-state behaviour of these complexes, preliminary
investigations into reactivity have been performed.
Table 1 π-Stacking interactions for tetrakis(amidate) complexes 5 and 6a
5 (Zr)
6 (Hf)
Centroid–centroid (Å)
3.610(3)
3.615(3)
3.96(15)
4.05(15)
1.207(7)
1.217(7)
4.023(5)
15.4(3)
1.583(9)
Plane–plane angle (°)
Fig. 4 ORTEP depiction of solid-state molecular structure of tris(amidate)
complex 7. Ellipsoids are plotted at 50% probability and hydrogen atoms are
omitted for clarity. Selected bond lengths [Å]: Zr–N1 2.269(2), Zr–N2 2.340(2),
Zr–N3 2.362(2), Zr–N4 2.071(2), Zr–O1 2.200(2), Zr–O2 2.144(2), Zr–O3 2.142(2),
sum about N4 360°.
Plane–plane shift distance (Å)
a Calculated using OLEX2.34
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 15670–15677 | 15673

View Article Online
Paper
Dalton Transactions
require a vacant coordination site in addition to the reactive
Zr–N bond for coordinating either the incoming alkene (direct
insertion) or the acidic amine (proton-assisted C–N bond for-
mation). Precatalyst 4 does not afford a vacant coordination
site in addition to the reactive Zr–N bond. However, such reac-
tive species could be generated in situ by the ligand redistribu-
tion reactions detailed earlier or by substitution of an amidate
ligand in the presence of excess amine.
To probe the active species promoting the observed cyclo-
hydroamination reaction, the reaction can be monitored by 1H
NMR spectroscopy as a function of time. After only 30 minutes
at 110 °C there is the clear appearance of free ligand in the
reaction mixture, as evidenced by the presence of a signal at
δ 8.81 ppm, corresponding to the ortho-naphthyl signal of the
N-(2,6-diisopropylphenyl)naphthylamide. Integration against
the 1,3,5-trimethoxybenzene internal standard shows that one
equivalent of free amide proligand is generated in situ. This
protonolysis of an amidate ligand in the presence of excess
amine would therefore generate bis(amidate) bis(amido) zirco-
nium species in solution, which have been proposed to be
active catalysts for hydroamination via the cycloaddition
mechanistic pathway.39 These observations are consistent with
the observed induction period prior to productive reactivity
(see ESI†) and can be attributed to the time required to gen-
erate active bis(amidate) complexes in solution.
Scheme 6 Catalytic cyclohydroamination with precatalyst 4.
application is hydroamination, as bis(amidate) supported
group 4 systems for hydroamination have precedence in the
literature.4,6,8,21,24,35,36 To the best of our knowledge, tris-
(amidate) complexes of early transition metals have not been
reported for this catalytic transformation.
For these studies we have focused on the use of precatalyst
4 (Scheme 4, Fig. 2), as the ortho-naphthyl 1H NMR signal
at
δ 9.11 ppm of the N-(2,6-diisopropylphenyl)naphthyl-
amidate ligand provides a convenient spectroscopic handle for
monitoring ligand coordination. Interestingly, complex 4 is a
successful, albeit sluggish, precatalyst for the cyclohydroami-
nation of 2,2-diphenylpent-4-en-1-amine, a commonly used
aminoalkene substrate37 (Scheme 6). This C–N bond formation
could be achieved with this very sterically shielded and coordina-
tively saturated metal centre in 41% NMR yield after 24 hours.38
This result was surprising considering that the majority of
group 4 catalyzed hydroamination reactivity has been proposed
to occur via a [2 + 2] cycloaddition between an imido–
zirconium intermediate and the alkene.39–45 Indeed early
investigations are consistent with this cycloaddition-based
mechanism for related bis(amidate) zirconium catalysts
(Fig. 5).21,22,39,44,45 However, the formation of the imido–zirco-
nium intermediate requires the availability of two reactive
ligand sites, which is not obvious with the three κ2-bound
amidate ligands in 4.
Alternatively, there have been kinetic mechanistic as well as
computational reports detailing zirconium precatalysts that
promote this transformation via alternate mechanisms such as
proton-assisted C–N bond formation or direct alkene insertion
into the Zr–N bond of zirconium bis(amido) intermediates.
These include zirconium complexes with constrained-
geometry ligands,46 zwitterionic complexes,47 or a tethered
bis(ureate) system.15,48 However, such mechanistic pathways
Conclusions
The synthesis of tris(amidate) mono(amido) complexes of tita-
nium and zirconium can be accomplished through protonoly-
sis of tetrakis(amido) metal starting materials with amide
proligands. These sterically demanding complexes have
differing coordination modes depending on the size of the
metal centre. When titanium is used, one of the amidate
ligands is bound in a κ1-(O) manner and the amidate ligands
exhibit exchange between κ1 and κ2 modes on the 1H NMR
timescale. Zirconium and hafnium, as larger metal centres,
can better accommodate the sterically demanding ligands,
resulting in tris(amidate) and tetrakis(amidate) complexes
with all amidate ligands bound in a κ2 motif. The tris(amidate)
complexes are susceptible to ligand exchange and dispropor-
tionation reactions and highlight the ligand fluxionality of
amidate ligands in the solution phase. The propensity for
ligand exchange to alleviate steric crowding can be used to
access reactive complexes in situ for catalytic applications. The
observed disproportionation into unreactive homoleptic
species is noteworthy, and represents a potential catalyst
decomposition pathway for amidate precatalyst systems.
Experimental procedures
General methods
All synthetic manipulations were conducted using convention-
al Schlenk line techniques or a glovebox under an atmosphere
Fig. 5 Cycloaddition mechanism for zirconium catalyzed hydroamination sup-
ported by sterically bulky amidate ligands. [Zr] = bis(amidate) zirconium.
15674 | Dalton Trans., 2013, 42, 15670–15677
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
of dry dinitrogen unless stated otherwise. Anhydrous hexanes which were used in all calculations. The final wR(F2) was 0.116
and toluene were purchased from Aldrich, sparged with dry, (all data).
degassed dinitrogen, and passed over an activated aluminum
Synthesis of tris(N-(2,6-dimethylphenyl)pivalamidate) mono-
oxide column and degassed prior to use. Anhydrous benzene, (dimethylamido) zirconium, 2. N-(2,6-Dimethylphenyl)pivala-
diethyl ether, tetrahydrofuran, and pentanes were purchased mide (2.30 g, 11.2 mmol) was dissolved in ∼120 mL of toluene
from Aldrich, sparged with dry, degassed dinitrogen, and in a 250 mL round bottomed Schlenk flask and cooled to
purified by passage through an Innovative Technologies SPS- −78 °C while stirring. In a separate flask, Zr(NMe2)4 (1.00 g,
PureSolv-400-4 apparatus. d6-Benzene and d8-toluene were 3.73 mmol) was dissolved in ∼30 mL of Et2O prior to addition
purchased from Cambridge Isotopes Ltd, degassed by 3 to the solution of the amide via cannula. This solution was
freeze–pump–thaw cycles, and stored overnight over molecular allowed to warm to room temperature overnight while stirring,
sieves prior to use. Reagents for amide synthesis were used as resulting in a clear colourless solution. Excess solvent was
received from Aldrich without further purification. The amides removed in vacuo to generate an off-white waxy solid. The
were synthesized from the corresponding commercial acid crude product was dissolved in 25 mL of pentane and filtered
chlorides and amines and rigorously dried by heating to through Celite™ to remove traces of unreacted proligand. The
100 °C under vacuum. The following compounds were syn- solution was concentrated to dryness to give 2 (2.48 g, 89%).
thesized as reported in the literature: N-(2,6-dimethylphenyl)- δH (300 MHz, C6D6) 1.02 (27H, s, C(CH3)3), 2.31 (18H, br s, Ph-
benzamide,49 N-(2,6-dimethylphenyl)pivalamide,23 N-(2,6-di- (CH3)2), 2.96 (6H, s, N(CH3)2), 6.78 (9H, m, Ar-H). δC (75 MHz,
isopropylphenyl)benzamide,50 N-(2,6-diisopropylphenyl)naphthyl- C6D6): 19.6, 27.8, 41.3, 43.2, 124.3, 127.8, 132.4, 143.7, 189.6.
amide,23 2,2-diphenylpent-4-en-1-amine,51 Zr(CH2Ph)4,52 and
Synthesis of tris(N-(2,6-diisopropylphenyl)benzamidate)
Hf(CH2Ph)4.52 Ti(NMe2)4, Zr(NMe2)4, and HfCl4 were pur- mono(dimethylamido) zirconium, 3. N-(2,6-Dimethylphenyl)-
chased from Strem Chemicals and used as received. All NMR benzamide (3.15 g, 11.2 mmol) was dissolved in ∼100 mL of
spectra were recorded on Bruker 300 MHz or 400 MHz Avance Et2O in a 250 mL round bottomed Schlenk flask and cooled to
spectrometers in non-spinning mode and referenced to the −78 °C while stirring. In a separate flask, Zr(NMe2)4 (1.00 g,
residual solvent signal. Single crystal X-ray structure determi- 3.73 mmol) was dissolved in ∼30 mL of Et2O prior to addition
nations, mass spectral, and elemental analyses determinations to the solution of the amide via cannula. This solution was
were performed at the Department of Chemistry, University of allowed to warm to room temperature overnight while stirring,
British Columbia. Mass spectra were measured on a Kratos resulting in a clear colorless solution. Excess Et2O was
MS-50. Elemental analyses were performed on a Carlo Erba removed in vacuo resulting in a white solid residue. The crude
Elemental Analyzer EA 1108. All X-ray crystallographic product was dissolved in 20 mL of pentane and filtered
measurements were made on Rigaku ADSC, Rigaku AFC7, through Celite™ to remove traces of unreacted proligand. The
Bruker X8 APEX CCD, or Bruker APEX DUO area detectors with solution was concentrated to dryness to give 3 (3.13 g, 86%).
graphite-monochromated Mo Kα radiation. Crystals for X-ray (Found: C, 71.47; H, 8.91; N, 5.20%. C59H72N4O3Zr requires
diffraction studies were recrystallized from the indicated C, 72.57; H, 7.43; N, 5.74%.) δH (300 MHz, C6D6) 0.88 (18H, d,
3
solvent, mounted in inert oil, and transferred to the cold gas 3JHH = 6.4 Hz, CH(CH3)2), 1.07 (6H, d, JHH = 6.6 Hz, C(CH3)2),
3
stream of the diffractometer.
1.26 (12H, d, JHH = 6.7 Hz, C(CH3)2), 3.29 (6H, s, N(CH3)2),
3
Synthesis of tris(N-(2,6-dimethylphenyl)benzamidate) 3.75 (6H, sept, JHH = 6.7 Hz, CH(CH3)2), 6.72–7.60 (24H, m,
mono(dimethylamido) titanium, 1. Ti(NMe2)4 (0.0668 g, Ar-H). δC (75 MHz, C6D6): δ 24.9, 25.5, 28.7, 43.7, 124.8, 126.4,
0.2981 mmol) was weighed out into a large vial equipped with 128.3, 130.8, 131.9, 132.9, 142.6, 143.2, 178.1; MS(EI) m/z
a stir bar and ∼3 mL of hexanes was added. A suspension of 930 ([M+] − NMe2).
N-(2,6-dimethylphenyl)benzamide (0.2015 g, 0.8944 mmol) in
Synthesis of tris(N-(2,6-dimethylphenyl)naphthylamidate)
hexanes was added slowly to the Ti(NMe2)4 solution and mono(dimethylamido) zirconium, 4. N-(2,6-Dimethylphenyl)-
stirred for 2 h. The clear, orange/red solution was then concen- naphthylamide (0.40 g, 1.21 mmol) was dissolved in ∼5 mL of
trated to dryness in vacuo resulting in an orange powder. THF in a 20 mL vial. In a separate Schlenk flask, Zr(NMe2)4
Recrystallization from a layered toluene–pentane solution gave (0.109 g, 0.4 mmol) was dissolved in ∼20 mL of THF prior to
1 as red prisms (0.210 g, 92%). (Found: C, 74.00; H, 6.29; the addition of the ligand solution via pipette. The solution
N, 7.32%. C47H48N4O3Ti requires C, 73.81; H, 6.33; N, 7.33%); was warmed to 60 °C overnight and then filtered through a
δH (400 MHz, C6D6, 85 °C), 2.26 (18H, s, Ar(CH3)2), 3.43 (6H, pipette plug of Celite™. The solution was concentrated to
br s, N(CH3)2), 6.80–6.99 (18H, m, Ar-H), 7.65 (6H, br s, Ar-H); dryness, and washed with pentanes. The pentanes were then
δC (125 MHz, C6D6, 85 °C)53 19.2, 51.0, 124.8 (br), 128.9, 129.2, decanted off resulting in 4 as a white solid. (Found: C, 76.04;
131.3 (br), 134.6 (br), 145.6 (br); MS(EI) m/z 764 ([M+]), H, 7.36; N, 4.80%. C71H78N4O3Zr requires C, 75.69; H, 6.98; N,
3
721 ([M+] − NMe2), 540 ([M+] − amidate), 496 ([M+] − NMe2 − 4.97%); δH (300 MHz, C6D6) 0.60 (18H, d, JHH = 6.1 Hz, CH
3
amidate), 225 (amidate).
(CH3)2), 1.16 (18H, d, JHH = 6.6 Hz, CH(CH3)2), 3.55 (6H, br s,
3
Crystal data. C47H48N4O3Ti,
M = 764.79, monoclinic,
N(CH3)2), 3.72–3.82 (6H, m, CH(CH3)2), 6.77 (3H, t, JHH
7.5 Hz, Ar-H), 7.00 (9H, br s, Ar-H), 7.16–7.20 (3H, m, Ar-H),
7.33–7.47 (12H, m, Ar-H), 9.11 (3H, d, JHH = 8.6 Hz, Ar-H);
=
a = 24.942(3), b = 10.581(1), c = 31.085(3) Å, V = 8203(2) Å3,
T = 90 K, space group C2/c, Z = 8, μ(Mo-Kα) = 0.25 mm−1
,
3
40 442 reflections measured, 10 639 unique (Rint = 0.042)
δC (75 MHz, C6D6) 22.7, 24.0, 28.3, 44.2, 123.8, 124.2, 125.7,
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 15670–15677 | 15675

View Article Online
Paper
Dalton Transactions
126.0, 127.0, 127.3, 128.6, 130.2, 131.4, 132.0, 134.5, 141.2, was dissolved in 100 mL Et2O and cooled to −78 °C in a
142.7, 180.8; MS(EI) m/z 1125 ([M+]), 1081 ([M+] − NMe2), 750 250 mL round bottomed Schlenk flask. To this flask was
([M+] − amidate − NMe2).
added (0.1 mL, 1.35 mmol) aniline which had been dissolved
in 10 mL of Et2O. This solution was allowed to warm to room
temperature overnight while stirring, resulting in a bright
yellow solution. Excess Et2O was removed in vacuo to generate
a yellow solid residue. The crude product was dissolved in
15 mL of pentane and filtered through Celite™ to remove a
small amount of pale yellow insoluble material. The solution
was concentrated to dryness to give 5 as a bright yellow solid
(0.700 g, 51%). Single crystals were grown from a saturated
toluene solution at room temperature. δH (300 MHz, C6D6)
Crystal data. C71H78N4O3Zr, M = 1126.59, triclinic, a =
11.9659(8), b = 12.5520(8), c = 23.178(2) Å, V = 3199.2(4) Å3, T =
100 K, space group P1, Z = 2, μ(Mo-Kα) = 0.22 mm−1, 20 144
ˉ
reflections measured, 8118 unique (Rint = 0.059) which were
used in all calculations. The final wR(F2) was 0.094 (all data).
Synthesis of tetrakis((N-2,6-dimethylphenyl)pivalamidate)
zirconium, 5. N-(2,6-Dimethylphenyl)pivalamide (0.77 g,
3.76 mmol) and Zr(CH2Ph)4 (0.25 g, 0.94 mmol) were com-
bined in a foil wrapped 250 mL round-bottomed Schlenk flask
equipped with a stir bar. To this flask was added 50 mL of
THF which had been cooled to −78 °C. The cloudy white
mixture was allowed to warm to room temperature and stirred
overnight. The clear, colourless solution was concentrated to
dryness in vacuo to give a white solid residue. The crude
material was triturated with ∼30 mL of hexanes, and the
product was isolated by filtration and dried in vacuo to give 6
as a white powder (0.72 g, 85%). Single clear colourless crystals
suitable for X-ray crystallographic analysis were grown from
a saturated hexane/benzene solution at room temperature.
δH (300 MHz, C6D6) 1.02 (36H, s, C(CH3)3) 2.40 (24H, s,
Ar(CH3)2), 6.62–6.69 (12H, m, Ar-H); δC (100 MHz, C6D6) 20.5,
28.3, 41.6, 124.7, 128.0, 132.8, 143.3, 189.9.
3
3
−0.08 (3H, d, JHH = 6.7 Hz, CH(CH3)2), 0.18 (3H, d, JHH
6.5 Hz, CH(CH3)2), 0.24 (3H, d, JHH = 6.6 Hz, CH(CH3)2), 0.48
=
3
3
3
(3H, d, JHH = 6.8 Hz, CH(CH3)2), 0.59 (3H, d, JHH = 6.5 Hz,
3
CH(CH3)2), 0.82 (3H, d, JHH = 6.6 Hz, CH(CH3)2), 0.89 (3H,
d, JHH = 6.5 Hz, CH(CH3)2), 1.08 (3H, d, JHH = 6.5 Hz,
CH(CH3)2), 1.14 (3H, d, JHH = 6.7 Hz, CH(CH3)2), 1.30 (3H,
d, JHH = 6.5 Hz, CH(CH3)2), 1.48 (3H, d, JHH = 6.7 Hz,
CH(CH3)2), 1.81 (3H, d, JHH = 6.8 Hz, CH(CH3)2), 2.69 (1H,
sept, JHH = 6.8 Hz, CH(CH3)2), 3.21 (1H, sept, JHH = 6.8 Hz,
CH(CH3)2), 3.58 (1H, sept, JHH = 6.9 Hz, CH(CH3)2), 3.60 (1H,
sept, JHH = 6.7 Hz, CH(CH3)2), 3.69 (1H, br s, CH(CH3)2), 4.54
3
3
3
3
3
3
3
3
3
3
3
(1H, sept, JHH = 7.6 Hz, CH(CH3)2), 6.57–7.71 (29H, m, Ar-H),
8.15 (1H, s, Zr-NH); MS(EI) (m/z): 930 ([M+] − NHPh).
Crystal data. C52H72N4O4Zr, M = 908.36, triclinic, a =
Crystal data. C70H80N4O3Zr, M = 1116.60, triclinic, a =
11.901(2), b = 11.908(2), c = 18.614(3) Å, V = 2568.1(8) Å3, T =
13.661(5), b = 15.314(5), c = 18.604(5) Å, V = 3395.0(19) Å3, T =
173 K, space group P1, Z = 2, μ(Mo-Kα) = 0.26 mm−1, 33 550
173 K, space group P1, Z = 2, μ(Mo-Kα) = 0.21 mm−1, 41 052
ˉ
ˉ
reflections measured, 12 691 unique (Rint = 0.031) which were
used in all calculations. The final wR(F2) was 0.107 (all data).
Synthesis of tetrakis((N-2,6-dimethylphenyl)benzamidate)
hafnium, 6. N-(2,6-Dimethylphenyl)benzamide (3.32 g,
14.7 mmol) and Hf(CH2Ph)4 (2.00 g, 3.68 mmol) were com-
bined in a foil wrapped 500 mL round-bottomed Schlenk flask
equipped with a stir bar. To this flask was added 100 mL of
THF which had been cooled to −78 °C. The cloudy white
mixture was allowed to warm to room temperature and stirred
for 2 h. The clear, colourless solution was then concentrated to
dryness in vacuo to give a white solid residue. The crude
material was washed with ∼50 mL of hexanes and dried
in vacuo to yield 7 as a white powder (3.72 g, 94%). Single clear
colourless crystals suitable for X-ray crystallographic analysis
were grown from a saturated hexanes solution at room temp-
erature. (Found: C, 67.26; H, 5.37; N, 5.09%. C60H56N4O4Hf
requires C, 67.00; H, 5.25; N, 5.21%); δH (300 MHz, C6D6) 2.51
reflections measured, 11 859 unique (Rint = 0.053) which were
used in all calculations. The final wR(F2) was 0.115 (all data).
General procedure for the intramolecular hydroamination of
primary aminoalkenes with complex 4
Complex 4 (10.0 mg, 0.0089 mmol), 2,2-diphenylpent-4-en-1-
amine (21.1 mg, 0.089 mmol), 1,3,5-trimethoxybenzene (5 mg,
0.0296 mmol), and d6-benzene (∼0.5 g) were weighed into a
small vial and then transferred into a J. Young NMR tube
equipped with a Teflon cap. The reaction mixture was heated
in a pre-heated 110 °C oil bath for the indicated time and
monitored by 1H NMR spectroscopy.
Acknowledgements
(24H, s, Ar(CH3)2), 6.77–6.87 (24H, m, Ar-H), 7.70 (8H, d, 3JHH
=
The authors gratefully acknowledge Boehringer Ingelheim
(Canada) Ltd and NSERC for supporting this work. PRP thanks
NSERC for a graduate scholarship. The authors thank Scott
Ryken, Jacky Yim, and Dr Brian O. Patrick for help with the
X-ray crystallography studies.
6.5 Hz, Ar-H); δC (75 MHz, C6D6) 19.8, 125.3, 128.4, 128.8,
129.3, 132.0, 133.4, 134.5, 143.4, 179.2; MS(EI) m/z 1076 ([M+]),
852 ([M+] − [amidate]).
Crystal data. C60H56HfN4O4, M = 1075.56, tetragonal, a =
11.0080(7), b = 11.0080(7), c = 21.318(2) Å, V = 2583.2(5) Å3, T =
173 K, space group P421c, Z = 2, μ(Mo-Kα) = 2.07 mm−1, 20 445
ˉ
reflections measured, 3104 unique (Rint = 0.072) which were
used in all calculations. The final wR(F2) was 0.082 (all data).
Synthesis of tris(N-(2,6-diisopropylphenyl)benzamidate)
mono(anilido) zirconium, 7. Complex 3 (1.00 g, 1.35 mmol)
References
1 P. Arndt, C. Lefeber, R. Kempe, A. Tillack and U. Rosenthal,
Chem. Ber., 1996, 129, 1281–1285.
15676 | Dalton Trans., 2013, 42, 15670–15677
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
2 G. R. Giesbrecht, A. Shafir and J. Arnold, Inorg. Chem., 28 R. Hart, W. Levason, B. Patel and G. Reid, J. Chem. Soc.,
2001, 40, 6069–6072. Dalton Trans., 2002, 3153–3159.
3 D. V. Gribkov and K. C. Hultzsch, Angew. Chem., Int. Ed., 29 S. V. Pasko, A. Abrutis and L. G. Hubert-Pfalzgraf, Mater.
2004, 43, 5542–5546. Lett., 2005, 59, 261–265.
4 P. D. Knight, I. Munslow, P. N. O’Shaughnessy and P. Scott, 30 Y. F. Loo, R. O’Kane, A. C. Jones, H. C. Aspinall, R. J. Potter,
Chem. Commun., 2004, 894–895.
P. R. Chalker, J. F. Bickley, S. Taylor and L. M. Smith,
J. Mater. Chem., 2005, 15, 1896–1902.
31 G. I. Spijksma, H. J. M. Bouwmeester, D. H. A. Blank,
A. Fischer, M. Henry and V. G. Kessler, Inorg. Chem., 2006,
45, 4938–4950.
5 D. A. Kissounko, I. A. Guzei, S. H. Gellman and S. S. Stahl,
Organometallics, 2005, 24, 5208–5210.
6 A. V. Lee and L. L. Schafer, Eur. J. Inorg. Chem., 2007, 2243–
2255.
7 A. L. Reznichenko and K. C. Hultzsch, Organometallics, 32 H. W. Roesky and M. Andruh, Coord. Chem. Rev., 2003, 236,
2010, 29, 24–27. 91–119.
8 P. Eisenberger and L. L. Schafer, Pure Appl. Chem., 2010, 33 Q. F. Mokuolu, P. A. Duckmanton, P. B. Hitchcock,
82, 1503–1515.
C. Wilson, A. J. Blake, L. Shukla and J. B. Love, Dalton
Trans., 2004, 1960–1970.
34 O. V. Dolomanov, L. J. Bourhis, R. J. Gildea, J. A. K. Howard
and H. Puschmann, J. Appl. Crystallogr., 2009, 42, 339–341.
35 R. O. Ayinla and L. L. Schafer, Inorg. Chim. Acta, 2006, 359,
3097–3102.
9 F. Zhang, J. Zhang, H. Song and G. Zi, Inorg. Chem.
Commun., 2011, 14, 72–74.
10 G. Zi, F. Zhang and H. Song, Chem. Commun., 2010, 46,
6296–6298.
11 P. Eisenberger, R. O. Ayinla, J. M. P. Lauzon and
L. L. Schafer, Angew. Chem., Int. Ed., 2009, 48, 8361– 36 R. O. Ayinla, T. Gibson and L. L. Schafer, J. Organomet.
8365. Chem., 2011, 696, 50–60.
12 L. J. E. Stanlake and L. L. Schafer, Organometallics, 2009, 37 S. Hong and T. J. Marks, Acc. Chem. Res., 2004, 37, 673–686.
28, 3990–3998.
38 Determined using 1,3,5-trimethoxybenzene as an internal
standard.
39 R. K. Thomson, J. A. Bexrud and L. L. Schafer, Organo-
metallics, 2006, 25, 4069–4071.
13 J. A. Thomson and L. L. Schafer, Dalton Trans., 2012, 41,
7897–7904.
14 R. L. Webster, N. Noroozi, S. G. Hatzikiriakos,
J. A. Thomson and L. L. Schafer, Chem. Commun., 2013, 49, 40 P. J. Walsh, A. M. Baranger and R. G. Bergman, J. Am.
57–59. Chem. Soc., 1992, 114, 1708–1719.
15 D. C. Leitch, P. R. Payne, C. R. Dunbar and L. L. Schafer, 41 F. Pohlki and S. Doye, Angew. Chem., Int. Ed., 2001, 40,
J. Am. Chem. Soc., 2009, 131, 18246–18247. 2305–2308.
16 P. R. Payne, J. A. Bexrud, D. C. Leitch and L. L. Schafer, 42 J. S. Johnson and R. G. Bergman, J. Am. Chem. Soc., 2001,
Can. J. Chem., 2011, 89, 1222–1229.
123, 2923–2924.
17 Z. Zhang and L. L. Schafer, Org. Lett., 2003, 5, 4733–4736.
18 C. Li, R. K. Thomson, B. Gillon, B. O. Patrick and
L. L. Schafer, Chem. Commun., 2003, 2462–2463.
19 A. L. Gott, A. J. Clarke, G. J. Clarkson and P. Scott, Organo-
metallics, 2007, 26, 1729–1737.
43 Y. H. Li, Y. H. Shi and A. L. Odom, J. Am. Chem. Soc., 2004,
126, 1794–1803.
44 A. L. Gott, A. J. Clarke, G. J. Clarkson and P. Scott, Chem.
Commun., 2008, 1422–1424.
45 L. E. N. Allan, G. J. Clarkson, D. J. Fox, A. L. Gott and
P. Scott, J. Am. Chem. Soc., 2010, 132, 15308–15320.
46 B. D. Stubbert and T. J. Marks, J. Am. Chem. Soc., 2007, 129,
6149–6167.
47 K. Manna, S. Xu and A. D. Sadow, Angew. Chem., Int. Ed.,
2011, 50, 1865–1868.
48 D. C. Leitch, R. H. Platel and L. L. Schafer, J. Am. Chem.
Soc., 2011, 133, 15453–15463.
49 R. K. Thomson, B. O. Patrick and L. L. Schafer,
Can. J. Chem., 2005, 83, 1037–1042.
50 C. A. Nijhuis, E. Jellema, T. J. J. Sciarone, A. Meetsma,
P. H. M. Budzelaar and B. Hessen, Eur. J. Inorg. Chem.,
2005, 2089–2099.
20 J. A. Bexrud and L. L. Schafer, Dalton Trans., 2010, 39,
361–363.
21 M. C. Wood, D. C. Leitch, C. S. Yeung, J. A. Kozak and
L. L. Schafer, Angew. Chem., Int. Ed., 2007, 46, 354–358.
22 M. C. Wood, D. C. Leitch, C. S. Yeung, J. A. Kozak and
L. L. Schafer, Angew. Chem., Int. Ed., 2009, 48, 6938.
23 L. J. E. Stanlake, J. D. Beard and L. L. Schafer, Inorg. Chem.,
2008, 47, 8062–8068.
24 Z. Zhang, D. C. Leitch, M. Lu, B. O. Patrick and
L. L. Schafer, Chem.–Eur. J., 2007, 13, 2012–2022.
25 H. Zhai, A. Borzenko, Y. Y. Lau, S. H. Ahn and L. L. Schafer,
Angew. Chem., Int. Ed., 2012, 51, 12219–12223.
26 R. K. Thomson, F. E. Zahariev, Z. Zhang, B. O. Patrick, 51 S. W. Hong, S. Tian, M. V. Metz and T. J. Marks, J. Am.
Y. A. Wang and L. L. Schafer, Inorg. Chem., 2005, 44, 8680–
8689.
27 G. Willey, T. Woodman, M. Fisher and M. B. Drew, Tran-
sition Met. Chem., 1998, 23, 467–471.
Chem. Soc., 2003, 125, 14768–14783.
52 U. Zucchini, E. Albizzati and U. Giannini, J. Organomet.
Chem., 1971, 26, 357–372.
53 Carbonyl carbon signal was not observed.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 15670–15677 | 15677